Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies on enzymes of sialic acid and pseudaminic acid biosynthesis Chou, Wayne Kuo Wei 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2006-199565.pdf [ 6.96MB ]
Metadata
JSON: 831-1.0061142.json
JSON-LD: 831-1.0061142-ld.json
RDF/XML (Pretty): 831-1.0061142-rdf.xml
RDF/JSON: 831-1.0061142-rdf.json
Turtle: 831-1.0061142-turtle.txt
N-Triples: 831-1.0061142-rdf-ntriples.txt
Original Record: 831-1.0061142-source.json
Full Text
831-1.0061142-fulltext.txt
Citation
831-1.0061142.ris

Full Text

Studies on Enzymes of Sialic Acid and Pseudaminic Acid Biosynthesis by Wayne Kuo Wei Chou B.Sc, The University of British Columbia, 2000 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Chemistry) THE UNIVERSITY OF BRITISH C O L U M B I A August 2006 © Wayne Kuo Wei Chou, 2006 / 11 Abstract The first two steps in the mammalian biosynthesis of sialic acid are catalyzed by a bifunctional hydrolyzing U D P - G l c N A c 2-epimerase/ManNAc kinase. The hydrolyzing epimerase domain, which catalyzes the conversion of U D P - G l c N A c to M a n N A c and U D P , was studied mechanistically. Incubation of U D P - G l c N A c in deuterated buffer produced a - [2- 2 H]ManNAc indicating that the epimerase reaction proceeds with a net retention of configuration at C - l and that C-2 is deprotonated and reprotonated with a solvent-derived deuterium atom. The enzymatic incubation of a potential reaction intermediate, 2-acetamidoglucal, resulted in a slow hydration reaction, forming M a n N A c . A positional isotope exchange and kinetic isotope effect experiment demonstrated that U D P loss occurs through a C - 0 bond cleavage and that C-2 deprotonation is not rate-limiting. These results support a mechanism involving the anri-elimination of U D P , forming 2-acetamidoglucal, followed by its syn-hydration, forming M a n N A c . SiaA, a monofunctional hydrolyzing U D P - G l c N A c 2-epimerase from Neisseria meningitidis, was recently identified to catalyze a mechanistically similar reaction to the mammalian epimerase and its study was undertaken to identify catalytic residues. Mutagenesis produced three SiaA mutants, D100N, E122Q and D131N. In each case catalysis was severely impaired, however, products were detected upon extended incubations. E122Q and D131N catalyzed the release of 2-acetamidoglucal before the formation of M a n N A c , demonstrating that hydrolyzing epimerases are able to catalyze the direct formation of 2-acetamidoglucal. Pseudaminic acid synthase is proposed to catalyze the condensation of PEP and 6-deoxy A l t d i N A c to form pseudaminic acid and phosphate. The activity of a pseudaminic Ill acid synthase has never before been observed, but a gene candidate, (neuB3 from Campylobacter jejuni N C T C 11168) potentially encoding a protein fulfilling this role, was identified and cloned. The activity of NeuB3 was studied through the synthesis and enzymatic incubation of its potential substrate, 6-deoxy A l t d i N A c . N M R spectroscopic analysis verified the activity of NeuB3 as a pseudaminic acid synthase and that the absolute stereochemistry of the pseudaminic acid produced had the L-glycero-L-manno configuration. Kinetic analysis revealed that NeuB3 requires a divalent metal ion for 1 8 catalysis and that optimal catalysis occurs at p H 7.0. Incubation with [2- 0 ] P E P demonstrated that NeuB3 operates via a C - 0 bond cleavage mechanism utilized in P E P -condensing synthases studied to date. IV Table of Contents A b s t r a c t i i T a b l e o f C o n t e n t s iv L i s t o f T a b l e s v i i i L i s t o f F i g u r e s ix L i s t o f S y m b o l s a n d A b b r e v i a t i o n s x i i i A c k n o w l e d g e m e n t s xx D e d i c a t i o n xx i C h a p t e r 1: S i a l i c A c i d a n d S i a l i c A c i d - R e l a t e d M o l e c u l e s 1 1.1 Sialic Acids and Sialic Acid-Related Molecules 2 1.2 Biosynthesis of Sialic Ac id in Mammals and Bacteria 6 1.3 U D P - G l c N A c 2-Epimerases 9 1.3.1 Epimerase Mechanisms 11 1.3.2 Non-Hydrolyzing Epimerase Mechanism 15 1.3.3 Hydrolyzing Epimerase Mechanism 17 1.4 /V-acetylneuraminic A c i d Synthases 22 1.5 Biosynthesis of Pseudaminic A c i d 28 1.6 Project Goals 31 C h a p t e r 2 : M e c h a n i s t i c S tud ies on H y d r o l y z i n g U D P - G l c N A c 2 - E p i m e r a s e s 33 2.1 Introduction 34 2.2 Expression and Purification of Mammalian Hydrolyzing U D P - G l c N A c 2 -Epimerase 36 2.2.1 Baculovirus Expression System 36 V 2.2.2 Expression and Purification of the Mammalian Hydrolyzing U D P - G l c N A c 2-Ep unerase 37 2.3 Characterization of Mammalian Hydrolyzing Epimerase Activity 39 2.3.1 Stereochemistry of the Reaction and Solvent Isotope Incorporation 39 2.3.2 Catalytic Competence of 2-Acetamidoglucal 42 2.3.3 Catalytic Competence of the Oxazoline of M a n N A c and oc-F G l c N A c ... 45 2.4 Kinetic Characterization of Mammalian Hydrolyzing Epimerase Activity 47 2.4.1 Enzyme Allostery 47 2.4.2 Test for a Primary K I E 52 2.4.3 Inhibition Studies with Glycosidase Inhibitors 56 2.5 Test for C - 0 vs. P-O Bond Cleavage and Positional Isotope Exchange Studies.. 58 2.5.1 Test for C-O vs. P-O Bond Cleavage 58 2.5.2 PLX Introduction 59 2.5.3 PIX Experiment on the Mammalian Epimerase 63 2.6 Conclusions 66 2.7 Future Directions and Related Studies 69 2.8 Bacterial Hydrolyzing U D P - G l c N A c 2-Epimerases 72 2.9 Project Goals 77 2.10 Site-Directed Mutagenesis of SiaA 78 2.11 Over-Expression of Mutant siaA Genes and Affinity Purification of SiaA Mutants 80 2.12 Characterization of SiaA Mutant Activity 82 2.13 Conclusions 85 2.14 Future Directions and Related Studies 87 VI 2.15 Experimental Procedures 89 2.15.1 Materials 89 2.15.2 General Methods 89 2.15.3 Over-Expression and Purification of the Mammalian Hydrolyzing U D P -G l c N A c 2-Epimerase 92 2.15.4 Characterization of Mammalian Hydrolyzing Epimerase Activity 93 2.15.5 Kinetic Characterization of Mammalian Hydrolyzing U D P - G l c N A c 2-Epimerase 94 2.15.6 Test for C O versus P-O Bond Cleavage and PLX Experiment 96 2.15.7 Synthesis of Substrate Analogs ,97 2.15.8 Materials and Methods 100 2.15.9 Site-Directed Mutagenesis of siaA 102 2.15.10 Over-expression and Purification of Hexa-Histidine Tagged SiaA Mutants 104 2.15.11 Characterization of SiaA Mutant Activi ty 105 Chapter 3: Identification and Characterization of a Pseudaminic Acid Synthase (NeuB3) from Campylobacter jejuni 106 3.1 Introduction :.. 107 3.2 Expression and Purification of NeuB3 109 3.3 Characterization of NeuB3 Activity 110 3.3.1 Synthesis of 6-deoxy A l t d i N A c 110 3.3.2 Test for NeuB3 Activity 113 3.4 Isolation and Characterization of Pseudaminic A c i d 116 3.5 Kinetic Characterization of NeuB3 124 3.6 Test for C - 0 versus P-O Bond Cleavage 128 Vll 3.7 Conclusions.. '. 132 3.8 Future Directions and Related Studies 134 3.9 Experimental Procedures 137 3.9.1 Materials and General Methods 137 3.9.2 Over-expression of neuB3 and Purification of NeuB3 137 3.9.3 N M R Incubation Studies of NeuB3 138 3.9.4 Isolation and Characterization of Pseudaminic A c i d 138 3.9.5 Kinetic Characterization of NeuB3 139 3.9.6 C-O versus P-O Bond Cleavage 140 3.9.7 Synthesis of 2,4-diacetamido-2,4,6-trideoxy-L-altrose (6-deoxy Al td iNAc) 141 3.9.8 Synthesis of [2 - 1 8 0]PEP : 142 B i b l i o g r a p h y 143 vm List of Tables Table 2.1 Summary of wild-type and mutant SiaA protein activities determined using ' H N M R spectroscopy 84 Table 2.2 Primers used in the S D M of p A M 0 4 103 Table 3.1 *H N M R spectral assignments of pseudaminic acid 121 Table 3.2 1 3 C N M R spectral assignments of pseudaminic acid 121 Table 3.3 Metal dependence experiment with NeuB3 127 ix List of Figures Figure 1.1 Sialic acids and sialic acid related molecules 2 Figure 1.2 Proposed biosyntheses of NeuN A c in mammals and bacteria 6 Figure 1.3 Alternate proposed synthesis of M a n N A c in bacteria 8 Figure 1.4 U D P - G l c N A c 2-epimerases 9 Figure 1.5 Epimerization mechanisms in "activated" subtrates 12 Figure 1.6 Epimerization mechanisms in "unactivated" substrates , 14 Figure 1.7 Reaction and mechanism catalyzed by the bacterial "non-hydrolyzing" U D P -G l c N A c 2-epimerase 15 Figure 1.8 Positional isotopic exchange catalyzed by Rffe 16 Figure 1.9 Proposed mechanisms of Rffe elimination 17 Figure 1.10 Three potential mechanisms of the mammalian hydrolyzing U D P - G l c N A c 2-epimerase 18 Figure 1.11 Reaction mechanism catalyzed by GDP-mannose mannosyl hydrolase 19 Figure 1.12 Previous experiments performed on the mammalian hydrolyzing epimerase. 20 Figure 1.13 Reactions catalyzed by PEP-condensing synthases 23 Figure 1.14 Proposed mechanisms of NeuN A c formation by NeuB 25 Figure 1.15 Inhibitors of K D 0 8 P synthase ..26 Figure 1.16 si- Face attack of PEP catalyzed by NeuN A c synthase 26 Figure 1.17 Potential reaction catalyzed by a pseudaminic acid synthase 27 Figure 1.18 Proposed biosynthesis of pseudaminic acid in C. jejuni 29 Figure 2.1 Baculovirus infection phases 37 Figure 2.2 S D S - P A G E gel of the hydrolyzing epimerase 38 X Figure 2.3 ' H N M R spectra monitoring the reaction of U D P - G l c N A c with the mammalian epimerase 41 Figure 2.4 *H N M R spectra monitoring the enzymatic conversion of 2-acetamidoglucal to [2- 2H] M a n N A c 44 Figure 2.5 Mechanism of the mammalian hydrolyzing epimerase 45 Figure 2.6 Potential substrate analogs of the mammalian hydrolyzing epimerase 46 Figure 2.7 Kinetic plots of velocity vs. substrate concentration in non-allosterically regulated enzymes and allosterically regulated enzymes 48 Figure 2.8 M W C model for a tetrameric enzyme 50 Figure 2.9 K N F model for a tetrameric enzyme 51 Figure 2.10 Reaction coordinate diaGram of C - H vs. C - D bond breaking 53 Figure 2.11 Enzymatic synthesis of [ 2 " - 2 H ] U D P - G l c N A c using Rffe and U D P - M a n N A c dehydrogenase 54 Figure 2.12 Continuously coupled U D P assay used to measure kinetic parameters for the hydrolyzing epimerase 55 Figure 2.13 Kinetic plots of initial velocity (vo) vs. U D P - G l c N A c concentration with unlabelled U D P - G l c N A c (O) and [ 2 " - 2 H ] U D P - G l c N A c (x) 55 Figure 2.14 Known inhibitors of glycosidases 57 Figure 2.15 C-O vs P - 0 bond cleavage experiment 59 Figure 2.16 PIX experiment of Rffe , 60 Figure 2.17 3 1 P N M R spectrum of 44 % 1 8 0-labeled inorganic phosphate 61 Figure 2.18 3 1 P N M R spectra of the P IX reaction catalyzed by RffE 62 Figure 2.19 Synthesis of [ l - 1 8 0 ] U D P - G l c N A c 64 Figure 2.20 Enzymatic conversion of [ l " - 1 8 0 ] U D P - G l c N A c to [ 1 8 0 ] U D P and ManNAc. 65 Figure 2.21 Mechanism of the mammalian hydrolyzing U D P - G l c N A c 2-epimerase 67 Figure 2.22 Assay used by Petersen et al. to detect the activity of a G l c N A c 6-phosphate 2-epimerase 73 Figure 2.23 Production of mutant plasmids using the QuikChange® Site-Directed Mutagenesis K i t 78 Figure 2.24 Portion of pET-30 vector in recombinant siaA plasmids 80 Figure 2.25 SDS P A G E of affinity purified hexa-histidine tagged SiaA and mutant proteins 81 Figure 2.26 ' H N M R spectra monitoring the reaction of D131N with U D P - G l c N A c (5 mM) 83 Figure 2.27 *H N M R spectra monitoring the reaction of E122Q with 2-acetamidoglucal and U D P (5 m M each) : 83 Figure 2.28 Potential inhibitors of the hydrolyzing epimerases 88 Figure 3.1 SDS P A G E gel of NeuB3 109 Figure 3.2 Reaction catalyzed by pseudaminic acid synthase and Fischer projections of 6-deoxy A l t d i N A c and pseudaminic acid 110 Figure 3.3 Synthesis of 6-deoxy A l t d i N A c 112 Figure 3.4 3 1 P N M R spectra monitoring the reaction of PEP and 6-deoxy A l t d i N A c with NeuB3 113 Figure 3.5 Selective region of the ' H N M R spectra monitoring the incubation of PEP and 6-deoxy A l t d i N A c with NeuB3 115 Figure 3.6 Ring conformations of 6-deoxy A l t d i N A c 116 Figure 3.7 ID T O C S Y N M R spectra of pseudaminic acid 118 Figure 3.8 400 M H z ' H N M R spectrum of pseudaminic acid in D 2 0 119 Figure 3.9 400 M H z 1 3 C N M R spectrum of pseudaminic acid in D 2 0 120 Figure 3.10 Structure of pseudaminic acid and compounds used in the assignment of its stereochemistry 123 Figure 3.11 Coupled phosphate assay used in the kinetic analysis of NeuB3 124 Figure 3.12 Kinetic plots of initial velocities vs. substrate concentration 125 X l l Figure 3.13 p H vs. rate profile of NeuB3 126 Figure 3.14 Proposed C-O vs. P-O bond cleavage mechanisms for NeuB3 129 Figure 3.15 Synthesis of [2 - 1 8 0]PEP : 130 Figure 3.16 3 1 P N M R spectra monitoring the incubation of NeuB3 with [2 - 1 8 OjPEP and 6-deoxyAltdiNAc 131 Figure 3.17 Legionaminic acid and two epimeric derivatives 134 Figure 3.18 Revised biosynthestic pathway of pseudaminic acid 136 List of Symbols and Abbreviations 8 chemical shift (ppm) A c acetyl A c 2 0 acetic anhydride A c O H acetic acid A g 2 0 silver(I) oxide A 5 P D-arabinose 5-phosphate B E V S Baculovirus Expression Vector System B G T (3-D-galactosyl transferase B n benzyl B r T M S trimethyl silyl bromide B S A bovine serum albumin C. jejuni Campylobacter jejuni C M P - N e u 5 A c cytidine 5'-monophospho-A^-acetylneuraminic acid CMP-Pse cytidine 5'-monophospho-pseudaminic acid C O S Y correlation spectroscopy C T P cytosine triphosphate C V column volume(s) d doublet ( N M R ) dq doublet of quartets ( N M R ) D deuterium ( 2 H) D a Dalton D A H 7 P 2-keto-3-deoxy-D-arabino-2-octulosonic acid 7-phosphate D C I - M S desorption chemical ionization mass spectrometry dd doublet of doublets ( N M R ) XIV 6-deoxy A l t d i N A c 2,4-diacetamido-2,4,6-trideoxy A l t d i N A c D M F A^A^-dimethylformamide D N A deoxyribonucleic acid D R u 5 P D-ribulose-5-phosphate D T T 1,4-dithio-DL-threitol D X u 5 P D-xylulose-5-phosphate E. coli Escherichia coli, strains AD202 , BL21 (DE-3), K l and X L - 1 Blue E D T A ethylenediaminetetraacetate, disodium salt E 4 P D-erythrose 4-phosphate E S I - M S electrospray ionization mass spectrometry E t O H ethanol Et2NP (OBn)2 di-benzyl-A^/V-di-ethyl phosphoramidite Gal galactose G D P - M a n guanosine 5'-diphospho-D-mannose G D P - G a l guanosine 5'-diphospho-D-galactose Glc glucose G l c N A c ./V-acetylglucosamine G l c N A c 6-P TV-acetylglucosamine 6-phosphate Gne gene encoding the mammalian hydrolyzing U D P - G l c N A c 2-epimerase G P glycogen phosphorylase H 1 B M human inclusion body myopathy H. pylori Helicobacter pylori H M Q C heteronuclear multiple-quantum coherence N M R spectroscopy IPTG isopropyl P-D-thiogalactopyranoside J coupling constant ( N M R ) ; subscripts indicate coupling partners X V £ c a t catalytic rate constant, turnover number kcJKm specificity constant; second-order rate constant k D a kiloDalton K D N 2-keto-3-deoxy-D-g/ycero-D-g<3/<2cto-2-nonulosonic acid K D N 9-P 2-keto-3-deoxy-D-g/ycero-D-ga/acto-2-nonulosonic acid 9-phosphate K D O 2-keto-3-deoxy-D-A77a«no-2-octulosonic acid K D 0 8 P 2-keto-3-deoxy-D-ma«no-2-octulosonic acid 8-phosphate Ky enzyme-inhibitor complex dissociation constant K I E kinetic isotope effect Km Michaelis constant, substrate concentration at half the maximal velocity ( V m a x ) K N F Koshland, Nemethy and Filmer model for enzyme allostery L B Luria-Bertani medium L P S lipopolysaccharide L S I - M S liquid secondary ionization mass spectrometry m multiplet ( N M R ) m 6 A jV-6-methyl deoxyadenosine M A L D I matrix-assisted laser desorption ionization Man mannose M a n N A c TV-acetylmannosamine M a n N A c 6-P iV-acetylmannosamine 6-phosphate M e O H methanol M E S 2-(Af-morpholino)ethanesulfonic acid M E S G 2-amino-6-mercapto-7-methylpurine ribonucleoside M W C Monod, Wyman and Changeux model for enzyme allostery XVI M W C O molecular weight cut-off n H i l l coefficient NA, Nc, NQ, Ny number of given nucleotide (A, C, G , or T) present in a sequence N A D + nicotinamide adenine dinucleotide, oxidized form N A D H nicotinamide adenine dinucleotide, reduced form NAG-thiazoline ./V-acetylglucosamine thiazoline nanE gene encoding a G l c N A c 6 P 2-epimerase in E. coli K l N a O M e sodium methoxide N C A M neural cell adhesion molecule neuB gene encoding the protein NeuB in E. coli K l NeuB N e u N A c synthase from E. coli K l neuB3 gene encoding NeuB3 in C. jejuni NeuB3 pseudaminic acid synthase from C. jejuni neuC gene encoding the protein NeuC in E. coli K l NeuC hydrolyzing U D P - G l c N A c 2-epimerase from E. coli K l NeuNAc neuraminic acid or sialic acid N. meningitidis Neisseria meningitidis N M R nuclear magnetic resonance OD600 optical density at 600 nm P C R polymerase chain reaction P E P phosphoenolpyruvate Pi inorganic phosphate PIX positional isotope exchange P L P pyridoxal phosphate P M S F phenylmethylsulfonyl fluoride xvi i P N P purine nucleoside phosphorylase P ( O M E ) 3 trimethoxy phosphine ppm parts per million P S A a-(2-8)-linked polysialic acid Pse pseudaminic acid psi pounds per square inch pyr pyridine RffE non-hydrolyzing U D P - G l c N A c 2-epimerase from E. coli x g revolutions per minute rt room temperature [S] substrate concentration S D M site-directed mutagenesis S D S - P A G E sodium dodecylsulfate polyacrylamide gel electrophoresis SF9 Spodoptera frugiperda ovarian cells strain 9 siaA gene encoding SiaA in N. meningitidis SiaA hydrolyzing U D P - G l c N A c 2-epimerase from N. meningitidis T H F tetrahydrofuran T O C S Y total correlated spectroscopy T O F time of flight Tris 2-amino-2-(hydroxymethyl)-l ,3-propanediol U enzyme unit U D P uridine diphosphate U D P D H U D P - M a n N A c dehydrogenase U D P - G l c N A c UDP-iV-acetylglucosamine U D P - M a n N A c UDP-Af-acetylmannosamine xvi i i U D P - M a n N A c U A UDP-A^-acetylmannosaminuronic acid U M P uridine 5'-monophophosphate U T P uridine 5'-triphosphate U V ultraviolet v initial reaction velocity (rate) Kmax maximal reaction velocity (rate) Common Amino A c i d Abbreviations A A l a alanine C Cys cysteine D Asp aspartate E G lu glutamate F Phe phenylalanine G G l y glycine H His histidine I lie isoleucine K Lys lysine L Leu leucine M Met methionine N Asn asparagine P Pro proline Q Gin glutamine R A r g arginine S Ser serine T Thr threonine V V a l valine W Trp tryptophan Y Tyr tyrosine Nucleotide Base Abbreviations A adenine C cytosine G guanine T thymine U uracil Acknowledgements XX M y success is completely attributable to Dr. Martin Tanner, as well as past and present members of the Tanner lab. I would like to thank Dr. Martin Tanner for his friendship, guidance and patience over the past few years, my predecessors for making the Tanner lab what it is today and current members for teaching me something new everyday (whether I ask for it or not). I would like to acknowledge the shops and services in the Chemistry department at U B C . In particular, I would like to thank Dr. Elena Polishchuk, Helen Wright and Candice Martins for maintaining the Biological Services Laboratory and for providing a friendly work atmosphere. I would also like to thank Dr. Werner Reutter and Dr. Stephan Hinderlich for donating the recombinant baculovirus encoding the mammalian epimerase, and Dr. Warren W . Wakarchuk and Dr. Scott Dick for donating the recombinant plasmid encoding neuB3. Dedicated to all the lab heroes. Chapter 1: Sialic Acid and Sialic Acid-Related Molecules 2 1.1 Sialic Acids and Sialic Acid-Related Molecules Sialic acids and sialic acid-related molecules are 2-keto-3-deoxy polyhydroxylated acids (or a-keto acids) found on the surfaces of eukaryotic and prokaryotic cells (Figure l . l ) . 1 The sialic acid family is composed of over 50 different isolated derivatives of jV-acetylneuraminic acid (NeuNAc) and 2-keto-3-deoxy-D-glycero-D-galacto-2-nonuloson\c acid ( K D N ) . Derivatives include modifications at the C-4, C-7, C-8 and C-9 carbon with acetate, lactate, phosphate, sulphate and methyl ethers. Sialic acid-related molecules are structurally similar to sialic acids yet differ in stereochemistry, substitution patterns or carbon chain length from NeuN A c and K D N . Examples of these molecules include 2-keto-3-deoxy-L-g/y6'ero-L-m<af««o-2-nonulosonic acid (pseudaminic acid) and 2-keto-3-deoxy-D-/77a««o-2-octulosonic acid ( K D O ) which are both found in bacteria, and 2-keto-3-deoxy-D-<3raZ>/no-2-octulosonic acid 7-phosphate (DAH7P) , found as an intermediate in the shikimic acid pathway for the formation of aromatic amino acids in plants, fungi and bacteria. COOH KDN DAH7P Figure 1.1 Sialic acids and sialic acid related molecules. Sialic acids and related molecules contain a carboxylate functionality which is negatively charged at physiological p H . In mammals, these negatively charged sialic 3 acids are displayed at the distal ends of cell surface carbohydrate chains and serve important functions in regulating cellular processes, such as cellular recognition and adhesion processes. 2 ' 3 , 4 Neural cell adhesion molecules ( N C A M s ) are cell surface glycoproteins found in glial, muscle and neuronal cells. N C A M s on the surface of a cell can bind homophilically (intracellularly) and heterophilically (intercellularly) adhering cells together. Polysialic acid (PSA), a homo-polymer of up to 200 residues of a-(2-8)-linked sialic acid, is found to glycosylate N C A M s during early development, growth and tissue regeneration. Sialylation of N C A M s inhibits adhesion through the high charge, ionic strength and hydration associated with P S A . During later stages of growth de-sialylation of N C A M s allows cells to adhere. Both forms of N C A M s are implicated in regulation of cellular migration, growth and differentiation.5'6 Regulation via sialylation is also seen in erythrocytes, leukocytes, thrombocytes and certain glycoconjugates.3 In the case of erythrocytes, their cell surfaces are polysialylated when first produced. The sialic acids are removed via sialidases or spontaneously in solution over their 120 day life span. Eventually host macrophages recognize, bind and phagocytose the exposed erythrocytes. This also demonstrates that the sialylation of mammalian cells is necessary for the host immune system to not recognize its own tissues. Interruptions in cell sialylation have been linked to the causes of several diseases. Sialuria is a human genetic disease causing severe mental retardation.7 Patients with sialuria exhibit the overproduction and accumulation of sialic acid in the cytoplasm. Human inclusion body myopathy ( H I B M ) is another set of diseases characterized by the neuromuscular degeneration of leg muscles mainly in adults of Middle-eastern and 4 Jewish descent and is thought to be caused by deficiencies in the biosynthesis of sialic acid. 8 Ce l l surface sialylation has also been correlated to the metastatic potential of tumor cells. 9 In some bacteria sialic acids are also exposed on the cell surface. The display of sialic acid in pathogenic bacteria is thought to lead to evasion and protection from the host immune system through molecular mimicry of host sialylated cel ls . 1 0 The main 11 12 bacteria responsible for the disease meningitis in humans are Neisseria meningitidis ' and Escherichia coli K l . 1 3 These Gram negative bacteria form a protective extracellular capsule composed of a-(2-8)-linked P S A similar to the P S A found in mammalian cells. The ability of P S A to protect these bacteria from the host immune system is demonstrated by the poor antibody response elicited by their capsular polysaccharides in mammals. 1 4 Gram negative bacteria also possess lipopolysaccharide (LPS), a glycolipid covering the outer membrane of the cell. LPS is thought to provide a highly charged hydrophilic layer on the outer surface of the cell, conveying protection from the host immune system and implicated in the virulence of certain bacteria.1 Sialylation of L P S with NeuNAc has also been observed in pathogenic bacteria such as E. coli and N. meningitidis. In addition, the sialic acid related molecule K D O is an important component of the inner core region of L P S , which anchors the bulk of the polysaccharide chain to lipid A , the membrane bound anchor of L P S . 1 5 Interruptions in the formation of K D O have been demonstrated to lead to an arrest in bacterial cell growth 1 6" 1 8 and cells with compromised L P S chains have been demonstrated to be less virulent and more susceptible to the action of antibiotics. 1 9 ' 2 0 5 Another example of sialic acid display in bacteria is seen with the flagellated Gram negative bacteria Campylobacter jejuni and Helicobacter pylori. C. jejuni is a main 21 23 * 24 cause of bacterial diarrhea " and H. pylori causes duodenal ulcers in humans. In addition to the display of sialic acid on their L P S , their flagella are also heavily glycosylated with pseudaminic a c i d . 2 5 , 2 6 Two flagellin proteins, F laA and FlaB, comprise the flagella in these organisms, of which the former protein is the main component. In H. pylori flagellin, up to 10 serine and threonine residues are glycosylated, contributing up to 3 kDa of modification. In C. jejuni flagellin, glycosylation has been found on up to 19 serine and threonine residues accounting for 6 kDa of modification. Most of the modified residues are located on the surface-exposed regions of the flagellin. The importance of the production of this a-keto acid is demonstrated by aflagellate C. jejuni and H. pylori26 mutants unable to produce pseudaminic acid. These results indicate that pseudaminic acid glycosylation of flagellin proteins is necessary for flagellar assembly and function. These bacteria also require the motility conferred by their flagella to 21 28 30 colonize the viscous gastrointestinal tract of host organisms. ' A step towards understanding the biological relevance of sialic acids and sialic acid related compounds is to study their biosynthetic formation. In mammals, this may lead to reasons for the occurrence of certain genetic diseases and in bacteria allow for the development of novel drugs to combat pathogenic micro-organisms. The well-studied biosynthesis of N e u N A c wi l l be the focus of the next section, as it shares aspects common in mammals and bacteria, as well as in the biosyntheses of sialic acid-related molecules. 6 1.2 Biosynthesis of Sialic Acid in Mammals and Bacteria HO v HO ° \ H O — " ^ A AcHN I O-UDP H 2 0 UDP-GlcNAc HO UDP-GlcNAc 2-Epimerase HO " T V H0 NHAc Q ManNAc Kinase ATP UDP OH 0 3 P O v NHAc H 1 ^ \ OH ManNAc ManNAc 6-P Bacteria PEP OCMP NeuNAc Synthase Mammals PEP NeuNAc 9-P Synthase COOH CMP-NeuNAc Synthetase CTP CMP-NeuNA Transferases NeuNAc NeuNAc 9-P Phosphatase NeuNAc 9-P yOOOOSA Figure 1.2 Proposed biosyntheses of NeuNAc in mammals and bacteria. The outer pathway is the biosynthesis of NeuNAc in mammals, while the inner pathway is the biosynthesis of NeuNAc in bacteria. In mammals, the biosynthesis of NeuNAc begins with a bifunctional enzyme, UDP-Af-acetylglucosamine 2-epimerase/A /-acetylmannosamine kinase ( U D P - G l c N A c 2-epimerase/ManNAc kinase) (Figure 1.2, outer pathway).1 This enzyme catalyzes the C-2 inversion and release of U D P from UDP-A^-acetylglucosamine ( U D P - G l c N A c ) to form A^-acetylmannosamine (ManNAc) . It then phosphorylates M a n N A c at C-6, producing N-acetylmannosamine 6-phosphate (ManNAc 6-P). The next enzyme in the pathway, N-acetylneuraminic acid 9-phosphate synthase (NeuNAc 9-P synthase), condenses phosphoenolpyruvate (PEP) with M a n N A c 6-P forming iV-acetylneuraminic acid 9-phosphate (NeuNAc 9-P), the 9-carbon precursor of N e u N A c . N e u N A c is then formed by dephosphorylation at C-9 via iV-acetylneuraminic acid 9-phosphate phosphatase 7 (NeuNAc 9-P phosphatase) and activated for transfer as CMP-TV-acetylneuraminic acid ( C M P - N e u N A c ) by a C M P - N e u N A c synthetase with C T P . C M P - N e u N A c is the final product in this pathway and is the substrate for all of the 18 known NeuNAc transferases. The bacterial biosynthetic pathway for sialic acid appears to be different than the mammalian version in that phosphorylated intermediates are not formed and M a n N A c is directly converted into N e u N A c . At the time the work in this thesis was initiated, the source of M a n N A c in bacteria was not clear, however, N e u N A c synthase had been identified (Figure 1.2, inner pathway).1 N e u N A c synthase is a bacterial homolog of the mammalian N e u N A c 9-phosphate synthase, but, condenses P E P with M a n N A c instead of M a n N A c 6-phosphate forming N e u N A c . A bacterial C M P - N e u N A c synthetase homolog, catalyzing the formation of C M P - N e u N A c with C T P and a bacterial C M P - N e u N A c transferase homolog, catalyzing the transfer of N e u N A c , have also been identified. In question is the formation of M a n N A c in this pathway. A reasonable proposition, related to the mammalian pathway, is the formation of M a n N A c from U D P - G l c N A c through the action of a mono-functional bacterial U D P - G l c N A c 2-epimerase (shown in Figure 1.2). A n alternate proposal by Petersen et al. (2000) is that the formation of M a n N A c occurs through the action of a G l c N A c 6-phosphate 2-epimerase, which interconverts the two epimers G l c N A c 6-phosphate and M a n N A c 6-phosphate, and a M a n N A c 6-phosphate phosphatase, which then hydrolyzes the C-6 phosphate group from M a n N A c 6-phosphate to form M a n N A c (Figure 1.3).31 This alternative pathway is thought to be involved in the catabolism of sialic acid rather than its anabolism. 3 2 These two possibilities are further discussed in section 2.7. The first portion of this chapter wi l l discuss the first enzyme in the N e u N A c biosynthetic pathway, U D P - G l c N A c 2-epimerase. Discussion wi l l focus on previous mechanistic studies of this enzyme and a related bacterial U D P - G l c N A c 2-epimerase, RffE, catalyzing a similar reaction. "03PO v O3PO \ NHAc HO-_ n GlcNAc 6-P I n ManNAc 6-P H O H O ^ ^ \ 2 - e p i m e r 3 S e "9Z^C^L°\ Phosphatase HO^ AcHN ^ *> \ -OH OH HPO4 OH GlcNAc 6-P ManNAc 6-P ManNAc Figure 1.3 Alternate proposed synthesis of ManNAc in bacteria. The second part of this chapter w i l l focus on PEP-condensing synthases including the two NeuNAc synthases, mechanisms of the reactions they catalyze, and the potential for the existence of an unidentified PEP-condensing enzyme in the biosynthesis of pseudaminic acid. 1.3 UDP-GlcNAc 2-Epimerases The reaction catalyzed by the U D P - G l c N A c 2-epimerases involved in sialic acid biosynthesis is not a true epimerization, since it is effectively irreversible and the substrate and products are not epimeric in relation. These enzymes w i l l therefore be referred to as "hydrolyzing" U D P - G l c N A c 2-epimerases to indicate that U D P release accompanies the inversion of stereochemistry (Figure 1.4A). A bacterial true "non-hydrolyzing" U D P - G l c N A c 2-epimerase (Rfffi), which catalyzes the reversible epimerization of U D P - G l c N A c and U D P - M a n N A c without the loss of U D P , does exist (Figure 1.4B). 3 3 ' 3 4 Since both types of epimerases catalyze similar reactions on the same substrate, and often share significant sequence homology, it is reasonable to expect that they utilize similar catalytic mechanisms. hydrolyzing HO UDP-GlcNAc 2-epimerase AcHN O-UDP H , 0 UDP UDP-GlcNAc ManNAc non-hydrolyzing HO UDP-GlcNAc 2-epimerase ^ C V Q AcHN O-UDP UDP-GlcNAc Rffe O - U D P UDP-ManNAc Figure 1.4 UDP-GlcNAc 2-epimerases. A) hydrolyzing UDP-GlcNAc 2-epimerase and B) non-hydrolyzing UDP-GlcNAc 2-epimerase (RffE). The bifunctional enzyme responsible for the first two reactions in the mammalian biosynthetic pathway of N e u N A c is the hydrolyzing U D P - G l c N A c 2-epimerase/ManNAc 10 kinase. 3 5 , 3 6 This is a hexameric enzyme with a monomelic mass of 75 kDa. The monomer is 722 amino acids long and composed of two domains. 3 7 The amino acids in the N -terminal domain share sequence similarities to bacterial non-hydrolyzing U D P - G l c N A c 2-epimerases (the first 378 residues of the rat enzyme share 22 % sequence identity with E. coli RffE), while residues in the C-terminal domain share sequence similarities to proteins of the sugar kinase superfamily. This bi-functionality was also confirmed through the generation of mutations in either domain, resulting in the expected selective loss of each of the predicted activities. Although the monomer contains regions for both activities, the hexameric state seems to be required for U D P - G l c N A c 2-epimerase activity, while only the dimeric state is required for M a n N A c kinase activity. The hydrolyzing epimerase is considered the key regulatory enzyme of cell 38 surface sialylation and catalyzes the rate limiting step in the biosynthesis of sialic acid. This is demonstrated by strong positive correlations of the activity of the hydrolyzing epimerase to the amount of cellular C M P - N e u N A c and sialic acid in human hematopoeitic cells, and by the fact that the hydrolyzing epimerase is feed-back inhibited by the final product in the pathway, C M P - N e u N A c . 3 9 Mutations in the hydrolyzing epimerase leading to a loss of feed back inhibition by C M P - N e u N A c are thought to be the cause of sialuria. 7 The importance of this enzyme is further demonstrated by murine (mouse) embryos engineered to lack a functional hydrolyzing U D P - G l c N A c 2-epimerase which suffered early embryonic death. 4 0 11 1.3.1 Epimerase Mechanisms Insight into the operative mechanism of the hydrolyzing 2-epimerase can be derived from studies on the bacterial non-hydrolyzing 2-epimerase. RffE provides a source of M a n N A c residues which are used in the biosynthesis of a variety of cell surface glycoconjugates in Gram negative and positive bacteria. 4 1" 4 4 As mentioned earlier, this enzyme catalyzes the reversible epimerization of U D P - G l c N A c and U D P - M a n N A c . Since this enzyme catalyzes a true epimerization and was the subject of extensive mechanistic study prior to this work, we wi l l now present a general introduction to epimerase mechanisms in the context of the non-hydrolyzing enzyme. A common strategy used by carbohydrate epimerases with "activated" substrates is a deprotonation/reprotonation mechanism (Figure 1 .5A) . 4 1 ' 4 5 ' 4 6 These substrates are "activated" in the sense that the stereocenter being inverted is adjacent to a carbonyl and bears a relatively acidic proton. Deprotonation of the acidic center forms an enol/enolate intermediate, and reprotonation from the opposite face inverts the stereocenter. Examples of enzymes which catalyze inversions in "activated" substrates include D-ribulose-5-phosphate-3-epimerase, which catalyzes the C-3 epimerization of D-ribulose-5-phosphate (DRu5P) and D-xylulose-5-phosphate (DXu4P) (Figure 1.5B), and cellobiose 2-epimerase, which catalyzes the C-2 epimerization of cellobiose and (3-D-glucopyranosyl-(l—*4)-D-mannose ((3-D-Glc/?-(l-^4)-D-Man) (Figure 1.5C). The operative mechanism of D-ribulose-5-phosphate-3-epimerase involves deprotonation at the C-3 center, forming an enolate intermediate which is then protonated on the opposite face inverting the center. 4 7 ' 4 8 A divalent metal ion is further proposed to activate the carbonyl for nucleophilic attack and to stabilize the enolate intermediate. The C-2 epimerization 12 catalyzed by cellobiose 2-epimerase is proposed to proceed with the same deprotonation/reprotonation mechanism through an enol intermediate with its substrate in the open-chain fo rm. 4 9 ' 5 0 B Asp178-< Hi§ ,JHis HO X* His \ HO . JHis ;zh 2 + His \ HO y\\s Zn 2 + OH x o — o>-Asp38 - < q ^ _ — O' y- -< HO O" b x 0 HO^f 2 H DRu5P "OH enolate intermediate " / OH R DXu5P CH 2OP0 3 : B-Enz HB-Enz OH OH cellobiose OH HB-Enz HB-Enz HB-Enz OH P-D-Glcp-(1-4)-D-Man HB-Enz *B-Enz XB-Enz O-H HB-Enz R= HO OH Figure 1.5 Epimerization mechanisms in "activated" subtrates. A) General deprotonation/ reprotonation mechanism, B) mechanism of D-ribulose-5-phosphate-3-epimerase and C) mechanism of cellobiose 2-epimerase. 13 With the non-hydrolyzing epimerase, Rffe, inversion occurs at a stereocenter which is "unactivated" in the sense that the inverted center does not bear an acidic proton. In this case, it is unlikely that Rffe utilizes a deprotonation/reprotonation mechanism for catalysis. A common strategy utilized by carbohydrate epimerases whose substrates are "unactivated" is through a transient oxidation with the co-factor N A D + ( F i g u r e 1.6A and 41,45,46 ^ n i n i t i a i oxidation of an adjacent alcohol to a ketone with N A D + activates the stereocenter being inverted. A deprotonation/reprotonation mechanism, followed by reduction of the ketone with N A D H effectively inverts the stereocenter (Figure 1.6A). This mechanism is employed by the enzyme GDP-mannose 3,5-epimerase, which interconverts GDP-D-mannose (GDP-D-Man) to GDP-L-galactose (GDP-L-Gal) (Figure 1.7 C ) . 5 1 A n initial oxidation at C-4, followed by two deprotonation/reprotonations at C-3 and C-5, and reduction at C-4 forms the product. Another carbohydrate epimerization strategy involves the use of N A D + to directly oxidize the stereocenter, and then reduce the same center from the opposite face (Figure 1.6B). UDP-galactose 4-epimerase uses this direct oxidation/reduction mechanism to catalyze the C-4 epimerizations of U D P -galactose (UDP-Gal) and UDP-glucose (UDP-Glc) (Figure 1.6D). 5 2 Oxidation at C-4 generates a ketone intermediate and rotation of the sugar moiety exposes the opposite face of the ketone to reduction by N A D H . Since the non-hydrolyzing epimerase does not 53 require a co-factor for catalysis, these mechanisms also cannot be applied. 14 D '-Glc Figure 1.6 Epimerization mechanisms in "unactivated" substrates. General transient oxidation/ epimerization mechanisms A) deprotonation/reprotonation mechanism and B) oxidation/reduction mechanism, C) mechanism of GDP-mannose 3,5-epimerase and D) mechanism of UDP-galactose 4-epimerase. 15 1.3.2 Non-Hydrolyzing Epimerase Mechanism Earlier studies suggested that the operative mechanism of the non-hydrolyzing epimerase is the ^ / - e l im ina t i on of U D P to form 2-acetamidoglucal, followed by the .^ -addi t ion of U D P back to 2-acetamidoglucal, forming U D P - M a n N A c (Figure 1.7). 5 3 , 5 4 Enz—Bl--. H 0 £ HO—v ' HO—v NHAc ur\~~~^-\~-\-0 anti-elimination HO'^V^-^' 0 \ syn-addition HO— ^ — H * \ A T = — Q H AcHN^ O H o u M p AcHN J X p / O U M P O ^ O U M P UDP-GlcNAc II E n Z _ B H A UDP-ManNAc » O ° Figure 1.7 Reaction and mechanism catalyzed by the bacterial "non-hydrolyzing" UDP-GlcNAc 2-epimerase. Evidence for this glycal mechanism was obtained from incubations of RffE with isotopically labeled substrates. 5 3 ' 5 4 A primary kinetic isotope effect of 1.8 was observed when rates with [ 2 " - 2 H ] U D P - G l c N A c and unlabeled U D P - G l c N A c were compared. This indicated that deprotonation at C-2 is the rate-determining step of the reaction. Further, the incubation of RffE and U D P - G l c N A c in D 2 0 led to the formation of [2" - 2 H]UDP-G l c N A c and [2" - 2 H]UDP-ManNAc. This indicated that a proton is removed from C-2 and is replaced with a solvent-derived deuterium atom during catalysis. Evidence for the formation of U D P was obtained when [ l " - 1 8 0 ] U D P - G l c N A c (with the 1 8 0- label in the bridging position between C - l and the p-phosphorus) was incubated with RffE. A positional isotope exchange (PIX) was observed, where the 1 8 0- label was found to scramble between the bridging and non-bridging positions in the p-phosphate of both epimers (Figure 1.8). The observation of a PIX, indicates that the C - 0 bond is transiently cleaved allowing the p-phosphate to rotate and the 1 8 0- label to scramble into a non-bridging position. Key evidence implicating the formation of both intermediates, 2-16 acetamidoglucal and U D P , came upon extended incubation of RffE and U D P - G l c N A c . This incubation eventually led to the release and direct detection of 2-acetamidoglucal and U D P in solution, indicating that these intermediates are thermodynamically more stable than the substrates. Figure 1.8 Positional isotopic exchange catalyzed by RffE. The l s O label is represented by the darkened oxygen atom. The detailed mechanisms of the syn- and arc^'-elimination steps of RffE are not known although the elimination steps have been suggested to proceed with an E l mechanism (Figure 1.9A) or an E l - l i k e E2 mechanism involving an oxocarbenium ion-like transition state (Figure 1.9B). 5 3 A structure of RffE co-crystallized with U D P -G l c N A c was solved using X-ray crystallography. 5 5 The structure of RffE was found to share structural homology to the glycosyltransferases, glycogen phosphorylase (GP) and T-4 phage P-glucosyltransferase (BGT) . GP catalyzes the reversible C - l phosphorolytic cleavage of the terminal a- l ,4-l inked glucose residue of glycogen, producing a-D-17 glucose 1-phosphate and B G T catalyzes the C - l electrophilic transfer of glucose, from UDP-glucose, to 5-hydroxymethylcytosine in D N A . The structural homology between these enzymes supports the notion of an oxocarbenium ion-like transition state in the elimination steps of RffE, since these glycosyltranferases are also proposed to proceed 56 58 through a similar transition state (Figure 1.9C). Figure 1.9 Proposed mechanisms of RffE elimination. A) E l mechanism, B) El-like E 2 mechanism and C) mechanism of BGT. Note that the ^/-elimination step of UDP is shown. 1.3.3 Hydrolyzing Epimerase Mechanism Previous work on the mammalian hydrolyzing 2-epimerase involved in vitro studies on partially purified rat liver enzyme. 5 9" 6 3 Kinetic analysis indicated the ordered 18 release of U D P followed by the irreversible formation of M a n N A c . 6 3 Three general strategies for catalysis were investigated (Figure 1.10). The first mechanism involves an initial hydrolysis of U D P - G l c N A c to form U D P and G l c N A c , followed by epimerization at C-2 to generate M a n N A c (path A) . The second begins with epimerization at C-2, forming U D P - M a n N A c , followed by hydrolysis of U D P - M a n N A c , forming U D P and M a n N A c (path B) . The third is the ^/ / -e l iminat ion o f U D P , forming the intermediate 2-acetamidoglucal, followed by the addition of water to form M a n N A c (path C). 2-Acetamidoglucal Figure 1.10 Three potential mechanisms of the mammalian hydrolyzing UDP-GlcNAc 2-epimerase. A) Hydrolysis, followed by C-2 epimerization, B) C-2 epimerization, followed by hydrolysis and C) anti-elimination of UDP, followed by addition of water. The hydrolysis steps in paths A and B could occur via the direct nucleophilic attack of water at the anomeric carbon. This is similar to the C - l nucleophilic attack seen in the tranferases mentioned above and with GDP-mannose mannosyl hydrolase, which catalyzes the C - l hydrolysis of GDP-mannose (GDP-Man) , forming G D P and mannose 19 (Man) (Figure 1.1 1 ) . 6 4 - 6 6 The C-2 epimerization in the second step of path A would presumably proceed through the open chain aldehyde form of G l c N A c and involve a deprotonation/reprotonation mechanism. The C-2 epimerization in the first step of path B could occur identically to the mechanism of RffE. Path C is also similar to the glycal mechanism of RffE except that water is added across the double bond of 2-acetamidoglucal instead of U D P . Figure 1.11 Reaction mechanism catalyzed by GDP-mannose mannosyl hydrolase. When the mammalian hydrolyzing epimerase was incubated with U D P - G l c N A c in tritiated water, the M a n N A c produced contained tritium at the C-2 position (Figure 1.12A). 6 0 This indicated that catalysis was proceeding with a deprotonation and reprotonation event at the C-2 position, as expected with all three mechanisms. When the putative intermediate of path A , G l c N A c , was incubated with the epimerase, no M a n N A c was generated, arguing against this mechanism (Figure 1.12B). 5 9 When U D P - M a n N A c was incubated with the epimerase, M a n N A c was produced, initially suggesting that path B could be the operative mechanism (Figure 1.12C). 6 1 When the same incubation was carried out in tritiated water, tritium was incorporated at the C-2 position of M a n N A c , however, the extent of incorporation was lower than when U D P - G l c N A c was the substrate (87 % incorporation compared to an identical incubation with U D P - G l c N A c ) . 20 Further, an incomplete incubation of U D P - M a n N A c and the epimerase in tritiated water did not produce any trace amounts of U D P - G l c N A c or any tritium incorporation at the C-2 position in recovered U D P - M a n N A c . These results argue against path B since the reversible epimerization of U D P - M a n N A c to U D P - G l c N A c could explain the incorporation of tritium at the C-2 position of M a n N A c . U D P - M a n N A c was deemed to be an alternative substrate of the hydrolyzing epimerase as opposed to a reaction intermediate. Initial evidence supporting the glycal mechanism (path C) came from the incubation of the hydrolyzing epimerase with 2-acetamidoglucal. The enzymatic products were analyzed using paper chromatography and a spot co-migrating with M a n N A c indicated that path C was likely the operative mechanism (Figure 1.12D). 6 3 A HO—v HO—\ NHAc B HO—v HO—v NHAc * C H N i u D P T OH A c H N \ H \ h UDP-GlcNAc ManNAc GlcNAc ManNAc C HO—v NHAc HO—v NHAc D HO—v HO—v NHAc H«5Si2i _ H T O _ h 8 O ^ \  H ^ X T ^ . " K X ^ O-UDP T OH ~{ OH UDP-ManNAc ManNAc _ A A c . M N . . . . ManNAc 2-Acetamidoglucal Figure 1.12 Previous experiments performed on the mammalian hydrolyzing epimerase. No further mechanistic experiments on the mammalian hydrolyzing U D P -G l c N A c 2-epimerase have been performed since the early 1970's. This is likely due to the documented instability of the mammalian enzyme and the problems associated with obtaining it from natural sources. 6 0 ' 6 1 However, with the recent cloning of the hydrolyzing epimerase gene from rat into a recombinant D N A construct by the Reutter lab, sufficient quantities of pure epimerase can be obtained. 3 5 This recombinant gene would be ideal for the further study of the mechanism catalyzed by the hydrolyzing 21 epimerase. Additional experiments on the complete characterization of the products from the 2-acetamidoglucal incubation would further address the potential of 2-acetamidoglucal as an intermediate. Similar experiments to those conducted on the non-hydrolyzing epimerase could determine whether the hydrolyzing epimerase does utilize a similar mechanism. 22 1.4 /V-acetymeuraminic Acid Synthases Two sialic acid synthases are responsible for the formation of the nine carbon frame of NeuNAc . N e u N A c 9-P synthase catalyzes the condensation of P E P with M a n N A c 6-phosphate, forming N e u N A c 9-phosphate in mammals (Figure 1.13A) and N e u N A c synthase (NeuB*) catalyzes the condensation of P E P and M a n N A c , forming NeuNAc in bacteria (Figure 1.13B). Several mammalian N e u N A c 9-P synthases have been characterized from rat 6 7, hog 6 8 , mouse 6 9 and human 7 0 and the latter two have been cloned. A l l mammalian synthases have been found to require a divalent metal ion, such as M n 2 + and M g 2 + , for catalysis. Interestingly, the human N e u N A c 9-P synthase also exhibits K D N 9-P synthase activity, and can catalyze the condensation of P E P with mannose 6-phosphate producing K D N 9-P. Rat and mouse synthases have been shown to lack this ability and it has been suggested that a separate pathway involving a K D N 9-P synthase is needed in other organisms producing K D N . 7 1 N o mechanistic studies have been conducted on the mammalian N e u N A c 9-P synthases, although its mechanism should resemble that of the bacterial enzymes since the human enzyme shares 33 % sequence identity with the E. coli enzyme. 6 9 ' 7 0 Other bacterial neuB1 genes have been identified and cloned from C. jejuni?1 N. meningitidis72 and S. agalactiae.73 These bacterial synthases also require a divalent metal ion for catalysis, and do not catalyze the condensation of PEP with mannose. 2-Keto-3-deoxy-D-manno-2-octulosonic acid 8-phosphate synthase ( K D 0 8 P synthase) (Figure 1.13C) 7 4 ' 7 5 and 2-keto-3-deoxy-D-ara6wo-2-heptulosonic acid 7-1 neuB (lower case and italicized) refers to the gene encoding the enzyme NeuB (upper case and non-italicized). This convention will be used throughout the thesis to label genes and their corresponding gene products. 23 phosphate synthase ( D A H 7 P synthase) (Figure 1.13D) 7 6 ' 7 7 are two other PEP-condensing enzymes which also catalyze the formation of a-keto acids. K D 0 8 P synthase catalyzes the condensation of PEP with D-arabinose 5-phosphate (A5P), forming 2-keto-3-deoxy-D-mtfn/TO-2-octulosonic acid 8-phosphate ( K D 0 8 P ) , and is involved in the biosynthesis of K D O . D A H 7 P synthase catalyzes the condensation of PEP with D-erythrose 4-phosphate (E4P), forming 2-keto-3-deoxy-D-ara6mo-2-heptulosonic acid 7-phosphate (DAH7P) , and is involved in the shikimic acid pathway. HO HO ManNAc 6-P ManNAc 0,PO OH OH arabinose 5-P NeuNAc 9P 0 3 PO synthase PEP HP0 4 NeuNAc synthase PEP HPO4 KD08P synthase PEP HPO4 COOH COOH NeuNAc HO / / / / ( / C H 2 O P 0 3 HO" H O - - ^ \ - ^ » ^ Y ^ C O O H KD08P OH OH erythrose 4-P DAH7P synthase PEP HPO4 O3PO COOH DAH7P Figure 1.13 Reactions catalyzed by PEP-condensing synthases. Reaction catalyzed by A) NeuNAc 9P synthase, B) NeuNAc synthase, C) KDO8P synthase and D) DAH7P synthase. 24 A key mechanistic experiment was recently conducted on NeuB from N. meningitidis involving the use of [2 - 1 8 0]PEP as a substrate in enzymatic incubations. 7 2 This experiment was used to differentiate between two mechanisms, a C - 0 bond cleavage mechanism and a P-O bond cleavage mechanism. The C - 0 bond cleavage mechanism is proposed to proceed with the initial attack of the C-3 of PEP onto the open chain aldehyde of M a n N A c , followed by an attack of water at the C-2 position of PEP to form a tetrahedral intermediate (Figure 1.14A). The tetrahedral intermediate then collapses releasing phosphate and forming open chain N e u N A c which cyclizes to the pyranose form in solution. The formation of the tetrahedral intermediate could occur in a concerted fashion or in a stepwise fashion through an oxocarbenium ion intermediate (the latter is shown in F ig 1.14A). Enzymatic precedence for the C - 0 bond cleavage mechanism comes from K D 0 8 P synthase and D A H 7 P synthase which both utilize a C - 0 bond cleavage mechanism for catalysis. ' In this case, the incubation of [2- 0 ] P E P would lead to the formation of 1 8 0-labeled phosphate. The second mechanism is a P-O bond cleavage mechanism proposed to proceed with the attack of water to the phosphate group of PEP, releasing phosphate and forming the enolate anion of pyruvate (Figure 1.14B). The enolate then adds to the C - l aldehyde of the open chain form of M a n N A c , producing open chain NeuNAc , which cyclizes to the pyranose form in solution. 70 Enzymatic precedence for this reaction comes from pyruvate kinase and PEP O A carboxykinase where catalysis is thought to proceed with a nucleophilic attack at the phosphate group. In this case, incubation with [2 - 1 8 0]PEP would lead to the formation of 1 8 0-labeled NeuNAc . This [2 - 1 8 0]PEP experiment was also conducted with K D 0 8 P synthase and D A H 7 P synthase, and in all experiments 1 8 0-labeled phosphate was formed, 25 indicating that the operative mechanism of these PEP-condensing enzymes is the C - 0 bond cleavage mechanism. = OV H 2 ° ^ " ' O H ^ - " O H Y - - O H M 2 + R R R ManNAc (open chain) oxocarbenium ion tetrahedral NeuNAc + P E P intermediate intermediate (open chain) D - 18-^ - 18 - 1R O 2 C ^ « 2 P O 3 = o 2 c Y « j o2c -H R R OH R ManNAc (open chain) aldose (open chain) NeuNAc + P E P + enolate (open chain) Figure 1.14 Proposed mechanisms of NeuNAc formation by NeuB. A) C-O bond cleavage mechanism and B) P-O bond cleavage mechanism. The 1 8 0 isotope is represented by the darkened oxygen atom. Evidence for the formation of the oxocarbenium ion intermediate in the C - 0 bond cleavage mechanism comes from studies on K D 0 8 P synthase with cationic inhibitors (Figure 1.15). 8 1 - 8 4 These mechanism-based inhibitors are bi-substrate inhibitors designed to mimic the oxocarbenium ion intermediate of K D 0 8 P and had inhibition constants (KCs) in the micromolar range. Evidence for the formation of the tetrahedral intermediate comes from time resolved electrospray ionization-time of flight-mass spectrometry (ES1-T O F - M S ) studies on the K D 0 8 P synthase reaction. 8 5 B y monitoring the enzymatic reaction of K D 0 8 P synthase and its natural substrates using ES1-MS on the millisecond time scale, the mass of the tetrahedral intermediate/enzyme complex was observed. 26 Figure 1.15 Inhibitors of KD08P synthase. The stereochemical course of the addition of P E P was also recently addressed in C. jejuni N e u B . 8 6 This was accomplished using C-3 deuterated and fluorinated P E P analogs. The separate incubations of NeuB with Z-[3- 2 H]PEP and £ - [3 - 2 H]PEP led to the formation of (3>S)-3-deutero-NeuNAc and (3/?)-3-deutero-NeuNAc, respectively (Figure 1.16). These results in conjunction with the observation that the C-4 of N e u N A c possesses the (^-configuration, indicates that the si-face of P E P is attacking the si-face of the aldehyde of M a n N A c during NeuB catalysis. Similar results were obtained with Z-[3-F]PEP although 2s-[3-F]PEP was not accepted as a substrate. These experiments have been conducted with K D 0 8 P synthase and D A H 7 P synthase also showing that catalysis proceeds with the si-face attack of P E P . 8 7 " 9 0 A difference with these synthases, however, is that the re-faces of the aldehyde substrates are attacked during catalysis. Figure 1.16 si-face attack of PEP catalyzed by NeuNAc synthase. In PEP, Hz and He refer to the protons which are in a cis and trans orientation to the phosphate group, respectively. In pseudaminic acid, He and Hz end up as the axial and equatorial protons, respectively. The term si-face refers to the pro-chiral sinister face of PEP and the aldehyde of ManNAc. 27 The PEP-condensing synthases discussed all catalyze a mechanistically similar condensation reaction of P E P to the open chain aide hyde of monosaccharide substrates, forming oc-keto acids. To date these are the only PEP-condensing synthases that have been identified. It is reasonable to assume that other PEP-condensing synthases exist which catalyze the formation of other a-keto acids, such as pseudaminic acid. The biosynthetic formation of pseudaminic acid is unknown, although it is possible that a pseudaminic acid synthase exists which would catalyze the condensation of PEP with its putative substrate, 2,4-diacetamido-2,4,6-trideoxy L-altrose (6-deoxy A l t D i N A c ) (Figure 1.17). AcHN OH OH AcHN Pseudaminic Acid 6-deoxy AltdiNAc Figure 1.17 Potential reaction catalyzed by a pseudaminic acid synthase. 28 1.5 Biosynthesis of Pseudaminic Acid None of the enzymes in the biosynthetic pathway of pseudaminic acid (PSE) have been identified, however, with the recent sequencing of the complete genome of C. jejuni N C T C 11168, 9 3 genes in this organism have been tentatively assigned roles in the biosynthesis of pseudaminic acid based on homology to the sialic acid biosynthetic pathways and mutational knockout studies (Figure 1.18). 2 5 ' 2 7 ' 9 1 ' 9 2 Our collaborators from the Wakarchuk lab have cloned five genes potentially involved in the biosynthesis of pseudaminic acid. The key enzyme proposed in the pathway is a pseudaminic acid synthase. This enzyme is expected to condense P E P with 2,4-diacetamido-2,4,6-trideoxy L-altrose (6-deoxy A l t d i N A c ) , forming pseudaminic acid directly. A gene encoding a putative CMP-pseudaminic acid synthetase ( C M P - P S E synthetase) was also identified via its homology to C M P - N e u N A c synthetases. Differences in the pseudaminic acid and NeuNAc pathways are evident in the substrates of the PEP-condensing synthases. 6-deoxy A l t d i N A c differs from the M a n N A c substrates in stereochemistry at the C-2, C-4 and C-5 positions, has an additional C-5 acetamido group, and lacks a C-6 hydroxyl group. The pseudaminic acid biosynthetic pathway is also proposed to start from U D P - G l c N A c and mass spectral studies support the notion that UDP-linked intermediates are involved. 9 2 ' 9 3 The first enzyme of the pseudaminic acid biosynthetic pathway is proposed to be a U D P - G l c N A c C-6 dehydratase, catalyzing an oxidation at C-4 and deoxygenation at C-6, forming a UDP-2-acetamido-2,6-dideoxy-4-keto hexose (UDP-6-deoxy-4-keto HexNAc) . The next enzyme is proposed to be a C-4 PLP-dependent aminotransferase, which catalyzes the transfer of an amino group to C-4, generating UDP-2-acetamido-4-amino-2,4,6-trideoxy L-altrose 29 (UDP-4-amino-6-deoxy A l t N A c ) . The activities of the first two enzymes have inverted the stereochemistry, and introduced an amino functionality at the C-4 position. They must also invert the C-5 center, although it is mechanistically reasonable to assume that this inversion could occur with either enzyme. The next enzyme is proposed to be an acetyltransferase which acetylates the C-4 amino group forming UDP-2,4-diacetamido-2,4,6-trideoxy L-altrose (UDP-6-deoxy Al td iNAc) . N o gene candidate encoding a possible acetyl transferase has yet been identified. A hydrolase is then proposed to hydrolyze the UDP-glycosidic linkage, forming 6-deoxy A l t d i N A c , the substrate for pseudaminic acid synthase. H O — v H ° H O - \ - » « - ^ « \ HNAcI O-UDP UDP-G lcNAc ° ^ ^ j ^ - _ - - 0 dehydratase \ — ~ \ transferase i Q H 2 0 HNAcI L - G l u a - k e t o g l u H O ^ ^ - ^ * ^ PLP-amino N H 2 UDP-GlcNAc O-UDP UDP-6-deoxy-4-keto HexNAc HNAc Pseudaminic acid (PSE) pseudaminic acid OH synthase HNAc (NeuB3) C 0 2 Pi P E P H O — ^ U D P H 3 C HNAc O H UDP-diNAc-altrose H N A c H,C HNAcI 3 O-UDP UDP-4-aminc-6-deoxy Al tNAc acyl transferase o h < y d r o l a s e (unidentified) C M P - P S E sSynthetase CTP^ 6-deoxy AltdiNAc H,C HNAcI 3 O-UDP UDP-6-deoxy AltdiNAc AcNH HNAc C M P - P S E Figure 1.18 Proposed biosynthesis of pseudaminic acid in C. jejuni. When the neuB gene in C. jejuni was identified, two other genes showing homology to N. meningitidis neuB were also identified. 2 7 The activities of the products of these two genes, NeuB2 and NeuB3, were unidentified but thought to modify the 30 flagellin proteins in C. jejuni. NeuB3 was found to have 35 % sequence identity to NeuB from TV. meningitidis and an insertional mutation in this gene resulted in C. jejuni which were aflagellate. neuB3 (cjl317) is the gene proposed to encode a pseudaminic acid synthase, which catalyzes the condensation of P E P with 6-deoxy A l t d i N A c to form pseudaminic acid. 31 1.6 Project Goals The aim of this thesis is to study two different enzymes catalyzing two unique reactions, yet, which are related through the biosyntheses of a-keto acids. The hydrolyzing U D P - G l c N A c 2-epimerase catalyzes the first step in the biosynthesis of N e u N A c and its mechanism is studied in Chapter 2. Pseudaminic acid synthase is potentially involved in the biosynthesis of pseudaminic acid and its discovery and study are the focus of Chapter 3. With the recent cloning of the hydrolyzing epimerase gene into a recombinant form and our previous experience in studying the non-hydrolyzing epimerase, we undertook the mechanistic study of the hydrolyzing epimerase in collaboration with the Reutter lab. In Chapter 2, the glycal mechanism is further investigated through enzymatic incubations with 2-acetamidoglucal and isotopically labeled substrates. In addition, site-directed mutagenesis studies on a newly discovered bacterial hydrolyzing U D P - G l c N A c 2-epimerase are presented. Pseudaminic acid synthase is a potential PEP-condensing synthase homologous to the N e u N A c synthases. The activity of this enzyme has never been observed, but is postulated to catalyze the condensation of 6-deoxy A l t d i N A c and P E P to make pseudaminic acid and phosphate. With the recent discovery and cloning of the potential gene candidate neuB3 by our collaborators in the Wakarchuk lab, we decided to identify and study the activity of the gene product of neuB3. In Chapter 3, the putative substrate of NeuB3, 6-deoxy A l t d i N A c , is synthesized and tested enzymatically. The reaction and products are identified and characterized using N M R spectroscopy, mass spectral 32 analysis and kinetics. The mechanism of this enzyme is further probed with isotopically labeled PEP. 33 Chapter 2: Mechanistic Studies on Hydrolyzing UDP-GlcNAc 2-Epimerases 34 2.1 Introduction2 A s mentioned in Chapter 1, three potential mechanisms were considered for the reaction catalyzed by the mammalian hydrolyzing epimerase. Evidence to date suggests that the glycal mechanism, involving the arc/z-elimination of U D P to form the intermediate 2-acetamidoglucal, followed by the addition of water, forming M a n N A c , is the operative mechanism. The most convincing evidence in support of this mechanism was the formation of a product which co-migrated with M a n N A c when 2-acetamidoglucal was incubated with the enzyme and the products were analyzed using high voltage paper electrophoresis.6 3 While this is an important observation that is relevant to the mechanistic analysis, the absence of any further characterization of the reaction products or kinetics requires that the experiment be revisited in more detail before it is taken for fact. Additional mechanistic evidence is the C-2 incorporation of 3 H in the M a n N A c formed from enzymatic incubations in tritiated water. 6 0 This indicates that C-2 is deprotonated and reprotonated with a solvent-derived proton during catalysis and is consistent with the glycal mechanism. The first part of Chapter 2 describes mechanistic studies which further probe the glycal mechanism using the recombinant gene encoding the mammalian hydrolyzing epimerase from rat. These studies identify the product from the incubation of 2-acetamidoglucal as M a n N A c using ' H N M R spectroscopy and demonstrate that the elimination of U D P occurs through a C - 0 bond cleavage process. They also demonstrate that the epimerase reaction proceeds with a net retention of configuration at C - l , 2 Versions of the work in this chapter have been published. 1) Chou, W. K., Hinderlich, S., Reutter, W., Tanner, M . E. (2003) Sialic acid biosynthesis: Stereochemistry and mechanism of the reaction catalyzed by the mammalian UDP-jV-acetylglucosamine 2-epimerase. J. Am. Chem. Soc. 125: 2455-2461. 2) Murkin, A. S., Chou, W. K., Wakarchuk, W. W., Tanner, M . E. (2004) Identification of a bacterial hydrolyzing UDP-N-acetylglucosamine 2-epimerase. Biochemistry, 43: 14290-14298. 35 indicating that the anti-elimination of U D P , forming 2-acetamidoglucal, and the syn-addition of water, forming M a n N A c , is the operative mechanism. A positional isotope exchange experiment and a kinetic isotope effect experiment further address the rate-determining step of the mammalian hydrolyzing epimerase. In the second part of Chapter 2, the discovery of a bacterial hydrolyzing U D P -G l c N A c 2-epimerase, SiaA, which catalyzes a reaction identical to the mammalian epimerase, leads to the study of active site residues in hydrolyzing epimerases. Guided by the X-ray crystal structure of the non-hydrolyzing epimerase, RffE, three active site residues that are conserved in both hydrolyzing and non-hydrolyzing epimerases are targeted for mutagenesis. The resulting conservative mutants, D100N, E122Q and D131N, were found to be severely catalytically impaired and ' H N M R spectroscopic studies provide further evidence for the glycal mechanism utilized by hydrolyzing epimerases. 36 Section A: Mechanistic Studies on the Mammalian Hydrolyzing UDP-GlcNAc 2-Epimerase from Rattus norvegicus 2.2 Expression and Purification of Mammalian Hydrolyzing UDP-GlcNAc 2-Epimerase 2.2.1 Baculovirus Expression System The production of a recombinant source of eukaryotic proteins often poses a much more challenging task than the corresponding prokaryotic proteins. Eukaryotic proteins are commonly post-translationally modified with such modifications as methylation, phosphorylation, sulphation, l ipid addition and glycosylation. 9 4 These modifications are often necessary for the protein to be fully active and properly folded. A method developed for the production of recombinant sources of post-translationally modified proteins is through the use of insect cells and baculoviruses (insect viruses). 9 5" 9 7 Insect cells post-translationally modify proteins similarly to mammalian cells. This method involves the generation of a recombinant baculovirus, which is used to infect and insert the desired gene into insect cells, which then express and produce the gene product. The cloning of the mammalian epimerase was conducted using the Bac-to-Bac® Baculovirus Expression System ( B E V S , Invitrogen) to generate a recombinant baculovirus (Autographica californica nuclear polyhedrosis virus) for the infection and production of the hydrolyzing epimerase in Spodoptera frugiperda ovarian cells (Sf9, fall army worm). 3 5 37 Baculovirus infection of insect cells involves three phases of growth (Figure 2.1). 9 5 In the early phase, the baculoviruses attach to insect cells and insert their D N A . The insect cells begin to express the viral D N A . In the late phase, late viral genes are expressed which code for replication of viral D N A and viral assembly. In the very late phase, virus particles are assembled and cell lysis occurs, releasing baculovirus particles. In the case of a recombinant baculovirus, the gene of interest w i l l be expressed upon infection of the virus. The gene product can be harvested after several days, before the very late phase, or more recombinant baculoviruses can be harvested during the very late phase. Figure 2.1 Baculovirus infection phases. 2.2.2 Expression and Purification of the Mammalian Hydrolyzing UDP-GlcNAc 2 -Epimerase Production of the hydrolyzing epimerase involved culturing Sf9 cells in a shaker flask and infection with the recombinant baculovirus encoding the mammalian hydrolyzing epimerase (donated from the lab of Dr. Werner Reutter). After an optimal infection time, the cells were harvested via centrifugation. The resulting cell pellet was lysed and purified using anion exchange chromatography. The resulting protein was early phase late phase very late phase 38 - 8 0 % pure as estimated using S D S - P A G E (Figure 2.2). Similar to previous reports, the mammalian epimerase was found to be unstable, with significant loss of activity within 48 hours with storage at 4 ° c . 6 0 ' 6 1 The enzyme was also unstable to freezing and concentration using centrifugal protein filters. A s a result, the epimerase was used directly in all assays. In order to prepare sufficient quantities of enzyme in deuterated buffer, the cell pellet was lysed and purified in deuterated buffer. A B C D E 66 kDa tfmm Figure 2.2 SDS-PAGE gel of the hydrolyzing epimerase. Lane A) molecular weight standards of 66 kDa (bovine serum albumin) and 29 kDa (carbonic anhydrase), B) baculovirus in media, C) crude cell lysate of uninfected Sf9 cells D) crude cell lysate of infected cells and E) purified hydrolyzing epimerase. 39 2.3 Characterization of Mammalian Hydrolyzing Epimerase Activity 2.3.1 Stereochemistry of the Reaction and Solvent Isotope Incorporation The first study conducted on the mammalian epimerase was determining the true reaction product of the enzymatic reaction. Previous studies on the hydrolyzing epimerase did not take into account the mutarotation of M a n N A c in aqueous solution, which interconverts the a - and (3-anomers through the open-chain form. The enzymatic formation of one anomer of M a n N A c would quickly lead to the formation of the other anomer via mutarotation. In order to determine which anomer is actually produced by the enzyme, a sample of the substrate was incubated with a large amount of enzyme and examined immediately using ' H N M R spectroscopy. It was necessary to show that the rate of mutarotation was slow enough to obtain a spectrum before equilibration occurred. Therefore the rate of non-enzymatic mutarotation was assessed by dissolving a commercial sample of M a n N A c , existing mainly as the p-anomer, in deuterated buffer (10 m M phosphate buffer, pD 7.9) and monitoring this process using ' H N M R spectroscopy. Since buffer can catalyze mutarotation, a low concentration was employed. B y observing the anomeric proton signals of M a n N A c , the equilibration of M a n N A c anomers was found to have a half-life of 10 minutes and rested as a ~1:1 ratio at equilibrium. The identity of the ' H N M R anomeric proton signals of M a n N A c also had to be assigned. In the case of sugars with an axial proton at C-2 (such as glucose), the H - l to H-2 coupling constant (JHI,H2) values can be used to assign the anomeric proton signals (•/HI,H2 for the a-anomer is larger than the p-anomer). In the case of sugars with an 4 0 equatorial proton at C - 2 , the JHI,H2 values cannot be used to assign the anomeric proton signals because of the similar values for both anomers. Bock and Pedersen reported that the C - l to H - l coupling constant (Jc\,m) values in a ] H coupled 1 3 C N M R spectrum are 98 1 0 Hz larger in the a-anomer than the p-anomer for a variety of D-monosaccharides. A ' H coupled 1 3 C N M R spectrum of an equilibrated sample of M a n N A c was taken, identifying the a-anomeric 1 3 C signal at 9 4 . 5 0 ppm with a J C I , H I of 171 H z and the p-anomeric 1 3 C signal at 9 4 . 3 8 ppm with a Jc\,m o f 161 Hz. A 2 - D heteronuclear N M R spectrum ( H M Q C ) taken of the same sample of M a n N A c , correlated the proton signal at 5 . 0 2 ppm to the a-anomeric 1 3 C signal and the proton signal at 4 . 9 2 ppm to the P-anomeric 1 3 C signal. The enzymatic conversion of U D P - G l c N A c to U D P and M a n N A c was monitored using ] H N M R spectroscopy in deuterated buffer. The initial *H N M R spectrum shows the anomeric proton signal of U D P - G l c N A c at 5 . 4 1 ppm in the absence of enzyme (Figure 2 . 3 , t = 0 min). This signal appears as a doublet of doublets due to coupling to the p-phosphorus atom and H - 2 " . In the presence of enzyme, the N M R spectral time course reveals the appearance of signals corresponding to M a n N A c and U D P , with full conversion to products after 4 5 min. This first spectrum, taken 5 min after the addition of enzyme, shows a non-equilibrium ratio of a- to P -ManNAc of 3:1 (Figure 2 . 3 , t = 5 min). This indicates that the true reaction product of the hydrolyzing epimerase is a - M a n N A c . Later spectra show the full conversion of substrate to products and that non-enzymatic mutarotation has equilibrated the anomers of M a n N A c to a 1:1 ratio (Figure 2 . 3 , t = 9 hrs). The observation that a - M a n N A c is the true reaction product of the mammalian 41 epimerase also indicates that the stereochemical course of the reaction proceeds with a net retention of configuration at C - l ( a - U D P - G l c N A c to a -ManNAc) . OUDP OH H UDP-GlcNAc a-ManNAc (J-ManNAc t = 9 hrs 5740 ^530 ' ' 5.20" ' £ l c P 1 . 0 0 ' ' 4.90 ppm Figure 2.3 'H NMR spectra monitoring the reaction of UDP-GlcNAc with the mammalian hydrolyzing epimerase. A) Enzymatic conversion of UDP-GlcNAc to a-[2- 2H]ManNAc, followed by mutarotation to (J-[2-2H]ManNAc and B) ' H N M R spectra monitoring the enzymatic conversion of UDP-GlcNAc. The actual product formed from this enzymatic incubation is a - [2- 2 H]ManNAc. Where previous studies demonstrated the C-2 incorporation of a solvent derived tritium 42 atom, 6 0 this study also demonstrates that C-2 is deprotonated and reprotonated with a solvent derived deuterium atom during catalysis. Since the tritium incorporation experiment uses the label in trace quantities and is subject to isotope effects, it is difficult to quantitate the extent of incorporation. In this experiment deuterium incorporation was found to be greater than 95 % demonstrating that incorporation is essentially quantitative. Deuterium incorporation is observed from the disappearance of the H-2 signals in the *H N M R spectrum (not shown), and by the appearance of the anomeric proton signals as singlets (due to small coupling constants to the C-2 deuterium atom and loss of coupling to the (3-phosphorus atom of UDP) . B y observing the remaining U D P - G l c N A c , after 80 % of the reaction had gone to completion (not shown), it was seen that no deuterium was incorporated at C-2". This would be observed by a change in the coupling pattern of the anomeric proton, since coupling to H - 2 " would be lost. These results further confirm previous observations that no label is incorporated into starting material and demonstrate that the reaction is essentially irreversible. 6 0 2.3.2 Catalytic Competence of 2-Acetamidoglucal In order to positively confirm previous results that 2-acetamidoglucal is converted to M a n N A c by the mammalian hydrolyzing epimerase, 6 3 ] H N M R spectroscopy was used. 2-Acetamidoglucal was synthesized according to literature described procedures" and incubated with the hydrolyzing epimerase in the presence and absence of U D P in deuterated buffer. The formation of M a n N A c was confirmed by monitoring the appearance of the two anomeric proton signals of M a n N A c . The first ' H N M R spectrum, in the absence of enzyme, shows the H - l signal of 2-acetamidoglucal at 6.59 ppm (Figure 43 2.4A). In the presence of enzyme, the ' H N M R spectroscopic time course reveals that the glycal is hydrated slowly to M a n N A c with a conversion of 6 % after 29 hours (Figure 2.4 B) . The conversion of the glycal to M a n N A c in the presence of 5 m M U D P was 12 % after 29 hours (Figure 2.4C). The increased formation of M a n N A c in the presence of U D P was reproducible (performed in duplicate). As a control reaction, 2-acetamidoglucal in deuterated buffer was also monitored, demonstrating that the non-enzymatic hydration of the glycal in solution does not occur. It is possible that the hydration of the glycal is due to glycosidase impurities in the enzyme preparation, since some glycosidases are known to hydrate their corresponding g lyca l s . 1 0 0 , 1 0 1 The strict formation of M a n N A c , however, argues against these impurities since G l c N A c would likely be formed as a result of the hydration from glycosidases. 44 B 8.0 7.6 7.2 6.8 6.4 6.0 5.6 5.2 4.8 |c oc-[2-2H]ManNAc p-[2-2H]ManNAc 5 03 4:98 4;93, 8.0 7.6 7.2 6.8 6.4 6.0 5.6 5.2 4.8 U D P i m p i 1: 5.03,4,98 4.93 : -.8 8.0 7.6 7.2 6.8 6.0 5.6 5.2 4 Figure 2.4 'H NMR spectra monitoring the enzymatic conversion of 2-acetamidoglucal to [2-2H]ManNAc. A) In the absence of enzyme, B) in the presence of enzyme after 29 h, C) in the presence of enzyme and 5 mM UDP after 29 h. This result positively confirms that the hydrolyzing epimerase is able to hydrate 2-acetamidoglucal to M a n N A c and strongly supports 2-acetamidoglucal as an intermediate in the hydrolyzing epimerase reaction. Although 2-acetamidoglucal is catalytically competent to serve as a substrate, it is not kinetically competent due to the slow rates of hydration. The slightly higher conversion of 2-acetamidoglucal to M a n N A c in the presence of U D P may indicate that U D P is enzyme bound during the hydration step of the reaction. 45 Since it has also been demonstrated that the hydrolyzing epimerase reaction proceeds with a net retention of configuration at C - l , then the operative mechanism of the epimerase is the ^ / - e l im ina t i on of U D P , followed by the ^y/7-hydration of 2-acetamidoglucal, forming a - M a n N A c (Figure 2.5). RffE also catalyzes a net-retention of configuration at C - l , although the second step involves the syrc-addition of U D P instead of water. Figure 2.5 Mechanism of the mammalian hydrolyzing epimerase. 2.3.3 Catalytic Competence of the Oxazoline of ManNAc and a-F GlcNAc Following the results obtained for the hydration of 2-acetamidoglucal, substrate analogs were tested for turnover (Figure 2.6). 2-acetamido-l,2-dideoxy-a-D-glucopyranosyl fluoride (a-F G l c N A c ) could act as a substrate analog for U D P - G l c N A c , with the elimination of fluoride occurring in a similar manner to the elimination of U D P . Fluorinated substrates have been used as inexpensive alternative donor substrates to study the activity of glycosyltransferases. 1 0 2 The oxazoline of M a n N A c , on the other hand, could be a potential reaction intermediate in the hydration step of the reaction. Recall the E l mechanism of RffE involving the oxocarbenium ion intermediate (Chapterl, Figure 1.10, p. 17). Potentially, the elimination and addition steps catalyzed by the mammalian epimerase could proceed through an oxocarbenium ion intermediate or transition state. 46 The oxocarbenium ion intermediate could be stabilized as an oxazolinium ion intermediate via neighboring group participation from the 2-acetamido group. a-F G l c N A c was synthesized according to the procedure of Micheel and Wul f f . 1 0 3 Per-acetylated oxazoline of M a n N A c was synthesized by Dr. Jomy Samuel, which was then de-acetylated to form the oxazoline of M a n N A c , according to the procedure of Khorl in et al..m In order to test the substrate analogs for turnover, incubations identical to those conducted with 2-acetamidoglucal were initiated. Each substrate analog (5 m M ) was incubated individually, in the presence and absence of U D P , for 2 days and monitored using ' H N M R spectroscopy. Unfortunately, no reaction was observed in either case. C H 3 a-F GlcNAc oxazoline of ManNAc Figure 2.6 Potential substrate analogs of the mammalian hydrolyzing epimerase. 47 2.4 Kinetic Characterization of Mammalian Hydrolyzing Epimerase Activity A s previously mentioned, the mammalian hydrolyzing epimerase is a hexameric enzyme. Previous kinetic analysis by Hinderlich et al. revealed that the mammalian epimerase exhibits strong negative allostery in substrate binding (Hi l l coefficient of 0.45). 3 6 This indicates that the binding of U D P - G l c N A c to one active site in one subunit hinders the subsequent binding of U D P - G l c N A c to the other subunits. The same study revealed that the mammalian epimerase is also feed-back inhibited by the final product in the pathway, C M P - N e u N A c . This inhibition was also found to be strongly cooperative (Hi l l coefficient of 4.1), indicating that the binding of C M P - N e u N A c to one subunit increases the binding of C M P - N e u N A c to other subunits. 3 6 In order to understand allosteric regulation in the mammalian epimerase a brief introduction to enzyme allostery w i l l be given. 2.4.1 Enzyme Allostery The Michaelis-Menten equation (Equation 2.1) is a quantitative description of the reaction rate (v) versus substrate concentration ([S]) catalyzed by a non-allosterically regulated monomeric enzyme, where F m a x is the maximum reaction rate (or turnover number) and Km is the Michaelis constant. 1 0 5 A plot of v versus [S] in such an enzyme wi l l yield a hyperbolic curve (Figure 2.7). v = [S] Equation 2.1 Michaelis-Menten equation 4 8 Similarly, the rates of reaction in oligomeric enzymes with identical subunits which act independently of one another are also described by the same equation. In this case, the enzyme subunits are acting independently as though they were monomeric. Vmax 1 Figure 2.7 Kinetic plots of velocity vs. substrate concentration in non-allosterically regulated enzymes and allosterically regulated enzymes. • non-allosterically regulated enzyme, • positive allosterically regulated enzyme and A negative allosterically regulated enzyme. Note that a negatively allosterically regulated enzyme will reach V ^ x at higher substrate concentrations. Figure was taken from Segel et al.}05 If the subunits in an oligomeric enzyme do not act independently of one another, such that the binding of one regulator molecule to one subunit affects the binding of substrates or catalysis in other subunits, then the enzyme is allosterically regulated. Allosteric regulation could alter the binding affinity (KM) of other subunits (referred to as 49 K systems) or the catalytic rate (kcat) of other subunits (referred to as V systems). Only K systems wi l l be discussed as they pertain to the mammalian hydrolyzing epimerase. In K systems, allosteric regulation could be positive, such that the binding of a molecule to one subunit increases the binding affinity in other subunits, or negative, such that the binding of a molecule to one subunit decreases the binding affinity in other subunits. A positive allosterically regulated enzyme exhibits a sigmoidal curve in a v vs. [S] kinetic plot, while a negative allosterically regulated enzyme exhibits a plateaued curve in the same plot (Figure 2.7). The sigmoidal curve can be explained by a smaller change in substrate concentration which results in a faster rate due to increased binding affinities of the subunits for their subsequent substrates (positive allostery) and the plateued curve can be explained a larger change in substrate concentration to reach the same increase in rate due to decreased binding affinities of the subunits for their subsequent substrates (negative allostery). In order to understand allosteric regulation, the Monod, Wyman, Changeux (rvlVVC) model, also referred to as the symmetry driven model, wi l l be discussed using a generic tetrameric protein and positive cooperativity as an example (Figure 2.8) . 1 0 6 In the M W C model, the enzyme subunits exist in equilibrium between two conformational states. There is a tense state (T state), which has a low affinity for substrates (the dissociation constant, Kj, is large) and a relaxed state (R state), which has a high affinity for substrates (the dissociation constant, KR, is small). The subunits in the oligomeric enzyme are all identical in their binding affinities in each state and also exist in the same state as the other subunits (hence A T and KR are identical for each subunit in their T and R state, respectively). In the absence of substrate, the initial ratio of enzyme in the To 50 state to Ro state (L) favors the To state (L = To/Ro » 1). A t low substrate concentrations, the substrates bind favorably to the enzymes in the Ro state (since KR < Kj), resulting in the R i state, while keeping the same ratio of enzyme in the To state to the Ro state. Since there is little substrate, the chances of another substrate binding to the R i state is low and so catalysis is slow. A s the substrate concentration increases, the substrates bind to enzymes in the higher R states favorably, resulting in the shift of the T states to the R states (since the equilibrium ratios of higher T to R states is reflective of L multiplied by c, the ratio of KR/KJ). This results in a large change in rate over a small change in substrate concentration. T states T 2 T, cLi c2L ^(sXs) R c5L R 3 c4L _ ( s X s ) R 4 R states Figure 2.8 MWC model for a tetrameric enzyme. T and R are the tense and relaxed states of the subunits. KT and Kg. are the dissociation constants of the tense and relaxed states. L is the initial ratio of the tense and relaxed states. S denotes substrate. The example given explains homotropic allostery where the substrate is the allosteric regulator. Heterotropic allostery also exists where another molecule is the allosteric regulator. With heterotropic allostery, binding of the allosteric regulator could occur in the substrate binding site or another site. In the M W C model, binding of a heterotropic allosteric regulator can shift the conformation of the enzyme to the R state 51 (activating the enzyme) or to the T state (inhibiting the enzyme). Binding of the heterotropic regulator to a different site can also occur in a similar manner to substrate binding and show positive cooperativity. Although the M W C model can explain positive cooperativity well , a shortcoming of this model is that it does not explain negative cooperativity. Another model to consider is the Koshland, Nemethy, Filmer (KNF) model (Figure 2.9). 1 0 7 The enzyme subunits in this model exist in the same state in the absence of substrate. When a substrate molecule or allosteric regulator is bound in one subunit, a conformational change in the other subunits occurs, such that varying degrees of substrate affinity exist ( K a * Kb * K c * Kd). The binding affinities in these subunits could be increased (positive cooperativity) or decreased (negative cooperativity) since they do not have identical binding affinities as in the M W C model. Although the K N F model is more complicated, it is thought to be a 108 better representation of enzyme behavior. ~ T ~ | *~a, ( S ) Q _ K b _ ( D ( D *~c_ g x f ) I T ~~ C Q ^ O O ^ © O © © Figure 2.9 KNF model for a tetrameric enzyme. Ka.d represent the dissociation constants for each state. S denotes substrate. The reaction rate versus substrate concentration catalyzed by allosteric enzymes can be described using the H i l l equation (Equation 2.2). 1 0 5 y = [ST K'+[S]" Equation 2.2 Hill equation In the H i l l equation, " « " is the Hi l l coefficient and reflects the degree of cooperativity in an allosteric enzyme and A"' is an apparent overall binding dissociation constant. A H i l l 52 coefficient greater than 1 indicates positive cooperativity and less than one indicates negative cooperativity. Using the H i l l equation, kinetic constants for allosteric enzymes can be determined as well as the degree of cooperativity between subunits. 2.4.2 Test for a Primary KIE Isotopic substitution can be used to probe the rate-limiting step in a reaction. If the substituted isotope is involved in bond breaking during the rate-limiting step of the reaction, a decrease in the reaction rate, compared to unlabeled material, w i l l be observed. This is termed a primary kinetic isotope effect (KIE). The reduction in rate occurs because the substituted isotope has a greater mass, resulting in an increased amount of dissociation energy required to break the bond. As an example, i f a C - H bond is broken during the rate-limiting step of a reaction and the protium atom is substituted for a deuterium atom, the C - D bond wi l l have a lower vibrational zero-point energy compared to the C - H bond (Figure 2.10). Breaking the C - D bond wi l l require more energy to reach the transition state than the C - H bond, resulting in a decreased reaction rate. B y comparing the reaction rate of isotopically labeled to unlabeled material (ko/ku), the magnitude of the K I E can be determined. 53 Reaction coordinate Figure 2.10 Reaction coordinate diagram of C-H vs. C-D bond breaking. A E D and A E H are the activation energies required to reach the transition state for bond dissociation ([C—L—B] *). The C-2" incorporation of a solvent-derived deuterium atom prompted a K I E study with the non-hydrolyzing epimerase, R f f E . 5 3 When the catalytic efficiencies (kcJKu) of [ 2 " - 2 H ] U D P - G l c N A c and unlabeled U D P - G l c N A c were compared, a primary K I E of 1.8 was observed, indicating that C-2 deprotonation was the rate-limiting step. Similarly, the mammalian hydrolyzing epimerase was tested to determine i f C-2 deprotonation is rate-limiting during catalysis. [ 2 " - 2 H ] - U D P - G l c N A c was prepared identically to the procedure used with the non hydrolyzing epimerase (Figure 2.11). 5 3 This procedure exploits the 10:1 equilibrium mixture of U D P - G l c N A c to U D P - M a n N A c and C-2" deuterium incorporation with RffE. Incubation of U D P - G l c N A c in deuterated buffer, produced a 10:1 mixture of [ 2 " - 2 H ] U D P - G l c N A c to [2" - 2 H]UDP-ManNAc . After the heat inactivation of RffE, U D P - M a n N A c dehydrogenase ( U D P D H ) and N A D + was added, converting [ 2 " - 2 H ] U D P - M a n N A c to [2"- 2H]UDP-mannosaminuronic acid (UDP-M a n N A c U A ) . Due to the difference in charges, [ 2 " - 2 H ] U D P - G l c N A c was easily separated using anion-exchange chromatography. The position of the label was 54 determined using *H N M R spectroscopy and the incorporation of the deuterium label in the [ 2 " - 2 H ] U D P - G l c N A c produced was determined to be greater than 95 % by mass spectral analysis. H 0 V H RffE H ° \ NHAc RffE H O " " f i o ^ X k e p i m e r a s % H r t o X ± k D 2 ° - " T H ^ . AcHNI D 2 ° D1 37°C, 2h; AcHN . OUDP OUDP 5 0 o C i 1 h OUDP [2"-1H]UDP-GlcNAc [2"-2H]UDP-ManNAc ' [2"-2H]UDP-GlcNAc 9% 90% v v ' I UDPDH, NAD + H ° 2 < \ NHAc ' H ° - A D D I AcHN I OUDP OUDP [2"-2H]UDP-ManNAcUA k v ' Separate by ion-exchange chromatography on DE52 Figure 2.11 Enzymatic synthesis of [2"-2H]UDP-GlcNAc using RffE and UDP-ManNAc dehydrogenase. A continuously coupled U D P assay, using pyruvate kinase and lactate dehydrogenase was employed to measure kinetic parameters (Figure 2.12). 1 0 9 The first coupling enzyme, pyruvate kinase, takes phosphoenolpyruvate and U D P to form U T P and pyruvate. The second coupling enzyme reduces pyruvate to lactate while consuming 1 equivalent of N A D H , which is monitored spectrophotometrically at 340 nm (e = 6220 M " 1 cm"1). Using the H i l l equation, the kinetic parameters obtained with U D P - G l c N A c as a substrate were a £ c a t of 0.33 ± 0.01 s"1, a A^M of 26 ± 4 | i M and a H i l l coefficient of 0.83 ± 0.17 and that with [ 2 " - 2 H ] U D P - G l c N A c were a kcat of 0.32 ± 0.01 s"1, a KM of 27 ± 4 pJVl and a H i l l coefficient of 0.79 ± 0.16 (Figure 2.13). These values give a &cat isotope 55 effect of 1.0 ± 0.1 and a kcat/KM isotope effect of 1.0 ± 0.4 and demonstrate that cleavage of the C-2/H-2 bond is not rate-limiting during catalysis. hydrolyzing epimerase UDP-GlcNAc • ManNAc + UDP " 0 2 C ^ ^ O P 0 3 2 pyruvate | kinase PEP UTP NADH NAD+ OiC^^P V " Q 2 C . ^ O H lactate | CH 3 dehydrogenase CH 3 pyruvate lactate Figure 2.12 Continuously coupled UDP assay used to measure kinetic parameters for the hydrolyzing epimerase. E 2 1 1 1 1 -- 7%x -r i i i i i i i -0 100 200 300 400 [UDP-GlcNAc] (uM) Figure 2.13 Kinetic plots of initial velocity (v<>) vs. UDP-GlcNAc concentration with unlabelled L D P -GlcNAc (O) and |2"- 2HlUDP-GlcNAc (x). The line was fitted to the plot obtained with unlabeled UDP-GlcNAc. Each kinetic point was measured in triplicate. 56 2.4.3 Inhibition Studies with Glycosidase Inhibitors The structural homology of RffE to the glycosyltransferases, G P and B G T (mentioned in Chapter 1, pp. 16-17), 5 5 prompted a previous member of the Tanner lab, Dr. Jomy Samuel, to investigate the inhibition of RffE with glycosidase inhibitors. 1 1 0 Glycosyltransferases and glycosidases catalyze the transfer of a donor sugar to an acceptor nucleophile. 1 1 1 In the former case, transfer is typically to a hydroxyl group of another sugar and in the latter case, transfer is to water. Known inhibitors of glycosyltransferases are not as numerous as those of glycosidases, presumably because it is difficult to mimic the diphosphonucleotide leaving group. However, several glycosidase inhibitors are commercially available. The inhibitors, 1-deoxy-nojirimycin, 2-acetamido-l,2-dideoxy-nojirimycin, iV-acetylglucosamine thiazoline (NAG-thiazoline) and 2-acetamido-2-deoxy-D-glucono-l,5-lactone, were all tested for inhibiton with RffE (Figure 2.14). 1-deoxy nojirimycin and 2-acetamido-l,2-dideoxy nojirimycin contain a ring nitrogen group which is protonated at physiological pH. These inhibitors are designed to mimic the oxocarbenium ion intermediates/transition states of glycosidases. 1 1 2 ' 1 1 3 2-acetamido-2-deoxy-D-glucono-1,5-lactone is also proposed to mimic the oxocarbenium ion intermediate/transition state through the flattened chair geometry of the lactone functionali ty. 1 1 4 , 1 1 5 NAG-thiazoline is a potential non-hydrolyzable, mimic of the oxazolinium intermediate. 1 1 6 In all cases, no inhibition was observed with RffE. Considering that glycosidases catalyze the transfer of water to a sugar substrate and that some glycosidases can catalyze the hydration of their corresponding g lyca l s , 1 0 1 , 1 1 1 it is possible that inhibition with glycosidase inhibitors w i l l be observed 57 with the mammalian hydrolyzing epimerase, where it was not with RffE. N A G thiazoline assay, 1 m M of the same inhibitors were kinetically analyzed at saturating concentrations of U D P - G l c N A c (1 mM) . No inhibition was observed with any of the compounds. Although inhibition with the glycosidase inhibitors would have given evidence for an oxcarbenium ion intermediate/transition state, this result does not entirely rule out the possibility. One explanation is that the U D P portion contributes most of the binding interactions, which these inhibitors lack. This would be supported by the observation that the presence of U D P increases the hydration of 2-acetamidoglucal. Given the nature of the kinetic assay used, U D P was not tested with the glycosidase inhibitors. It should also be noted that, while there are many known glycosidase inhibitors, success in designing potent glycosyltransferase inhibitors has been much rarer. was synthesized according to the method of Knapp et al.n6 Using the same U D P coupled OH AcHN 1-deoxy-nojirimycin 2-acetamido-1,2-dideoxy-nojirimycin NAG-thiazoline C H 3 Figure 2.14 Known inhibitors of glycosidases. 58 2.5 Test for C-O vs. P-O Bond Cleavage and Positional Isotope Exchange Studies 2.5.1 Test for C-O vs. P-O Bond Cleavage To determine whether the loss of U D P proceeds with a C-O or P-O bond cleavage, a labeling experiment with H 2 1 8 0 was conducted. A C - O bond cleavage could occur through the a^ri-elimination of U D P (Figure 2.15A) and a P-O bond cleavage could occur through the nucleophilic attack of water at the phosphorus atom of the P-phosphate (Figure 2.15B). In the former case, i f the mammalian epimerase and U D P - G l c N A c were incubated in H 2 1 8 0 , then the 1 8 0 label would reside in the M a n N A c product. In the latter case, the same incubation would result in 1 8 0-labeled U D P . The mammalian epimerase and U D P - G l c N A c were incubated in phosphate buffer (pH 7.5) containing 50% H 2 1 8 0 . The reaction was run until completion and the U D P produced was isolated using anion-exchange chromatography. Mass spectral analysis of the isolated U D P demonstrated the absence of the 1 8 0 label. The resulting M a n N A c produced was not isolated, nor analyzed because of the position of the 1 8 0 label. The 1 8 0 label would reside at the anomeric position of M a n N A c and it is possible that this label would non-enzymatically exchange with solvent, giving false positive or negative results. The absence of the label in the U D P produced is consistent with a C-O bond cleavage mechanism, yet, is a negative result confirming the absence of label. In order to obtain positive conformation that U D P is lost via a C-O bond cleavage, a positional isotope exchange (PIX) experiment was conducted. 59 1 8 « H 2 Figure 2.15 C-O vs P-O bond cleavage experiment. The l 8 0 label is represented by the darkened oxygen atom. 2.5.2 PIX Introduction The isotopic labeling of enzyme substrates or products is a common method of elucidating the chemical events which occur during enzyme catalysis. The primary K I E observed with RffE demonstrates that C-2" deprotonation is rate-limiting and the incorporation of deuterium into the products of RffE and the mammalian epimerase demonstrate that the C-2/H-2 bond is broken during catalysis. Another isotope, used to 18 18 elucidate chemical events in enzymes with phosphorylated substrates is O. With an O-labeled substrate, the fate of the label can be traced in the enzymatic products. In the case of the PEP-condensing synthases described in Chapter 1, incubation with [1 - 1 8 0]PEP and their corresponding co-substrates led to the formation of their corresponding a-keto acids and 1 8 0-labeled inorganic phosphate. The conclusion from these results is that these PEP-condensing synthases catalyze the loss of phosphate through a C - 0 bond cleavage. The fate of the , 0 0 label can also be used to determine enzyme-bound reaction intermediates 60 i f this label scrambles to another position during catalysis . 1 1 7 ' 1 1 8 In this case, the scrambling of a label is termed a positional isotope exchange (PIX). The mechanics of a P I X can be explained using the experiment previously conducted with RffE (Figure 2.16). 5 4 When [ l - , 8 0 ] U D P - G l c N A c (with 1 8 0 in the bridging position between the anomeric carbon and the pVphosphorus) was incubated with RffE, this produced the intermediates, 2-acetamidoglucal and l s O-labeled U D P . The three free P-O bonds in the P-phosphate of U D P are torsionally equivalent by rotation about the fourth P-O bond (connected to the a-phosphate), and re-addition to 2-acetamidoglucal can occur equally with any of the three oxygen atoms. In this experiment, the intermediate must have a lifetime greater than or equal to that of rotation about the fourth P-O bond or no P I X wi l l be observed. A P I X was observed, with scrambling of the label into non-bridging positions in the substrate and product. The P IX experiment with Rffe provided evidence for the intermediate, U D P , and showed that it must have been generated via a C - O bond cleavage mechanism. Figure 2.16 PIX experiment of RffE. The l s O label is represented by the darkened oxygen atom. 61 To monitor the P IX, 3 1 P N M R spectroscopy was used. 3 1 P N M R spectroscopy was first used by Cohn and Hu to observe the enzymatic exchange of 1 6 0 for 1 8 0 in inorganic phosphate and phosphate esters. 1 1 9 , 1 2 0 In an enzymatically equilibrated sample of 44 % 1 8 0-labeled inorganic phosphate, it was demonstrated by 3 1 P N M R spectroscopy that each 1 8 0 isotope incorporated produced a ~0.2 ppm upfield shift in the phosphorus signal, corresponding to the 5 species of non-labeled to fully labeled phosphate (Figure 2.17). 1 1 9 The ratios of the phosphate species followed a normal distribution representing the percentage of 1 8 0 to 1 6 0 . The experiments of Cohn and H u also demonstrated that the isotopically induced chemical shifts could be used to determine the P-O bond order in labeled phosphate ester species. 1 2 0 If an 1 8 0 label is in a bridging position, singly bonded to phosphorus (bond order of 1), then the induced upfield shift is slightly smaller than that produced i f the label is at a non-bridging position (bond order greater than 1). On the average, the former induced upfield shifts were -0.015 ppm and the latter ~0.020-0.030 ppm (bond order greater than 1). 1 6 o 2 1 8 o 2 i 6 o ; 8 o I -0.04 0.00 0.04 0.08 0.12 ppm Figure 2.17 3 1 P NMR spectrum of 44 % 1 O-labeled inorganic phosphate. Figure was taken from Cohn and H u . 1 1 9 62 A) JLJ u (unlabelled) = sugar—G—|—O—UMP O " O b (bridging) = sugar—*—P—O—UMP O" II n (nonbridging) = sugar—O—fj>—O—UMP iwyi%< # * V** <+**\ ^nfH 4 y» •• hjm b' ik. -12.6 -12.8 -13,0 -13,2 -13.4 -13.6 ppm Figure 2.18 3 , P NMR spectra of the PIX reaction catalyzed by RffE. A) Before the addition of RffE and B) after the addition of RffE. "u" and "u"'denotes unlabeled, "b" and "b"' denotes bridging labeled and "n" and "n" ' denotes non-bridging labeled UDP-GlcNAc and UDP-ManNAc, respectively. Figure taken from Morgan et al.53 The initial 3 1 P N M R spectrum of the P I X experiment with RffE shows the (3-phosphorus signals for 81 % labeled [ l " - 1 8 0 ] U D P - G l c N A c (Figure 2 .18A) . 5 4 The smaller doublet (due to coupling to the a-phosphorus), corresponds to unlabeled U D P - G l c N A c and the larger doublet, shifted slightly upfield by 0.013 ppm, corresponds to bridging 1 8 0-labeled U D P - G l c N A c . After the addition of enzyme and full equilibration, a new signal, corresponding to non-bridging 1 8 0-labeled U D P - G l c N A c appears, shifted 0.029 ppm upfield compared to unlabeled U D P - G l c N A c (Figure 2.18B). The scrambling of the " T 1 -13.8 -14.0 63 1 8 0 label is also observed in the U D P - M a n N A c produced (P-phosphorus signals centered at -13.8 ppm). The intensities of the signals also correspond to the statistical 2:1 distribution of non-bridging to bridging sites. 2.5.3 PIX Experiment on the Mammalian Epimerase In order to obtain a positive confirmation that the loss of U D P occurs via a C-O bond cleavage, a P I X experiment using [ l " - 1 8 0 ] U D P - G l c N A c was conducted with the mammalian epimerase (Figure 2.20A). [ l " - 1 8 0 ] U D P - G l c N A c was synthesized according to literature procedures (Figure 2.19). 5 4 The 1 8 0 label was incorporated by heating compound 1 in 95 % enriched H 2 1 8 0 . Compound 2 was then phosphorylated using the phosphoramidite method. 1 2 1 Complete deprotection of compound 3 gave compound 4, and morpholidate coupling resulted in [ l " - 1 8 0 ] U D P - G l c N A c . Incorporation of the 1 8 0 label was determined to be 68 % by mass spectral analysis and the position of the label was determined using 3 1 P N M R spectroscopy (Figure 2.20B and inset). In the initial 3 1 P N M R spectrum, before the addition of the epimerase, the 1 8 0-labeled P-phosphorus signal of U D P - G l c N A c is shifted 0.013 ppm upfield from the unlabeled P-phosphorus signal in a 2:1 ratio. After the addition of the mammalian epimerase, signals corresponding to free U D P appear (Figure 2.20C and inset). The signal corresponding to the P-phosphorus atom of unlabeled U D P is centered at -8.001 ppm, while the signal corresponding to the p-phosphorus atom of 1 8 0-labeled U D P is shifted 0.029 ppm upfield. The ratio of 1 8 0-labeled to unlabeled U D P was found to be 2:1, and was confirmed through its isolation and mass spectral analysis. Since the label is fully retained in the 64 U D P produced, it can be concluded that the loss of U D P occurs through a C-O bond cleavage process. AcO-AcO' AcO HO-HO HO H 2» * ^ > ^ l j CH3CN A * N ° H 80-85 °C AcO AcO AcO-1) Et3NP(OBn)2, AcHN n «H 2) H 20 2 , THF 2 AcO—. triazole, CH2CI2 A$£$^k AcHN 1 3 . »PO(OBn)2 AcHN O 1) H 2, Pd/C 2) NaOMe, MeOH ONa ONa , 0 ^ :X>. [1"-180]UDP-GlcNAc 0 H 0 H "NH UMP-morpholidate, HO-^ L 1H-tetrazole, HO-^V~^—"°\ ^ H O ^ - T - A trioctylamine, AcHN | Pyridine •PO(ONa)2 Figure 2.19 Synthesis of [l-1801UDP-GlcNAc. The l s O label is represented by the darkened oxygen atom. B y observing the remaining U D P - G l c N A c in the reaction (after 80 % conversion), it can be seen that there is no positional isotopic scrambling of the 1 8 0- label (Figure 2.20C). Scrambling would be evidenced by the appearance of an additional P-phosphorus signal shifted 0.029 ppm from the P-phosphorus signal of unlabeled U D P - G l c N A c . A control reaction with RffE was conducted to ensure that positional isotopic scrambling is observable under these conditions. A P I X was observed in this control reaction (not shown). Although the absence of a P IX suggests that C-O bond cleavage is irreversible, it is possible that the P-phosphate is tightly bound by the enzyme, restricting bond rotation. If this were the case, no scrambling would be observed. A P I X was observed, however, with RffE under single turnover conditions, when U D P - M a n N A c dehydrogenase was added to convert any U D P - M a n N A c formed into U D P - M a n N A c U A . 65 H O ^ HNAc -Q B • o o HNAc I II II • - P - O - P - O U r d B i eel OH OH B O - U D P - G I C N A C a-ManNAc O O ll II + H#^P -0 -P -OUrd B i a i OH OH 1 8 0 - U D P UDP-GlcNAc-a-P UDP-GlcNAc-P-P '-8.0 -8.4 -8.8 -9.2 -9^6-10.0-10.4-10.8-11.2-11.6-12.'o UDP-p-P -7'.9 -8'.0 : 8 r T UDP-a-P -1117 -1118 -11'9 ULl J J L -8.0 -8.4 -8.8 -9.2 -9.6-10^-10.4-10.8-11.2-11.6-12.0 Figure 2.20 3 1 P NMR spectra monitoring the enzymatic conversion of |l"- 1 8OlUDP-GlcNAc to | l ! fO]UDP and ManNAc. A) Reaction of [ l " - ' 8 0]UDP-GlcNAc and the mammalian epimerase, B) 3 1 P N M R spectra before enzyme addition and C) after enzyme addition and 80 % conversion, (a) denotes labeled material and (b) unlabeled material. * Signal corresponds to a 3 % impurity of UDP dimer, formed from the morpholidate coupling reaction. 66 2.6 Conclusions The mammalian hydrolyzing U D P - G l c N A c 2-epimerase is the key regulatory enzyme involved in the biosynthesis of sialic acids. 3 8 Its catalytic mechanism had not been fully characterized, but a glycal mechanism similar to that of the non-hydrolyzing epimerase, RffE, was implicated. With the recent cloning of the gene into a recombinant form 3 5 and previous success in studying RffE in the Tanner lab, a collaboration with the Reutter lab was initiated to elucidate its catalytic mechanism. The incubation of the mammalian epimerase in deuterated buffer, identified the product of the mammalian epimerase as oc-[2- 2H]ManNAc. This experiment demonstrated that the true product of the mammalian epimerase is a - M a n N A c and that the reaction proceeds with a net retention of configuration at C - l . The C-2 incorporation of deuterium re-confirmed previous observations that C-2 is deprotonated and reprotonated with a solvent derived isotope and showed that exchange occurred with each turnover. 6 0 The reaction products from the incubation of the mammalian epimerase with 2-acetamidoglucal were re-examined using ' H N M R spectroscopy. Previous analysis of the products utilized paper chromatography, which could have given a false positive result for the identification of a product which co-migrated with M a n N A c . Analysis of the N M R spectroscopic time course incubations, revealed the slow hydration of the glycal to M a n N A c . Apparently, 2-acetamidoglucal is catalytically competent to serve as a substrate but not kinetically competent due to the slow rate of hydration observed. In the presence of U D P , hydration of the glycal was found to be slightly enhanced. This may indicate that U D P is present in the active site during the hydration of the glycal. The 67 slightly higher conversion of the glycal in the presence of U D P indicates that U D P does not inhibit the hydration step of the reaction. The slow rates of hydration may be due to minimal amounts of enzyme in the correct protonation state or conformation required for hydration 1 2 2 and the presence of U D P may stabilize the active conformation of the active site to some extent for hydration. The bond cleavage experiment and P IX experiment demonstrate that the loss of U D P occurs through a C-O bond cleavage. This result and the fact that C-2 is deprotonated during catalysis, indicate that an a«//-elimination of U D P is occurring. The intermediacy of 2-acetamidoglucal is implicated by its hydration to M a n N A c . Since catalysis occurs with a net retention of configuration at C - l , the second step of the reaction involves a syrc-hydration of 2-acetamidoglucal. These results support the operative mechanism of the mammalian hydrolyzing epimerase as the <?«//-elimination of U D P , followed by the ^ - h y d r a t i o n of 2-acetamidoglucal, forming a - M a n N A c (Figure 2.21). Figure 2.21 Mechanism of the mammalian hydrolyzing UDP-GlcNAc 2-epimerase. In the case of the mammalian epimerase, the lack of a K I E demonstrated that C-2 deprotonation is not the rate-limiting step during catalysis. This is opposite to the result found with RffE, where C-2 deprotonation was demonstrated to be rate-limiting. Further, a P I X was not observed with the mammalian epimerase, although it was observed with 68 RffE. These results imply that the mammalian epimerase is proceeding through an E l mechanism, where C-O bond cleavage, forming an oxocarbenium ion intermediate, is slower than C-2 deprotonation, forming 2-acetamidoglucal. This may also indicate that C-O bond cleavage is the rate-determining step in the reaction. However, it is possible that the hydration of 2-acetamidoglucal or substrate binding/product release is rate-determining. More studies wi l l have to be conducted to explore these possibilities. 69 2.7 Future Directions and Related Studies Mutagenesis experiments on the mammalian hydrolyzing epimerase were conducted in the study by Reutter et al. that confirmed its bifunctionality. 3 7 Five histidine residues, conserved in hydrolyzing and non-hydrolyzing epimerases, were mutated in the ./V-terminal region of the mammalian hydrolyzing epimerase. The resulting mutants, H45A, H110A, H132A, H155A and H157A, were all decreased dramatically in hydrolyzing epimerase activity. The loss of activity in the latter 4 mutants was thought to be caused by a change in the oligomerization states. The wild-type mammalian epimerase exists mainly as the hexameric state, whereas the 4 histidine mutants were shown to exist mainly in a trimeric state (H155A, H157A) or in a trimeric/hexameric state (H110A and H132A) using size-exclusion chromatography. The H45 residue was speculated to be important in catalysis, due to the decreased activity of the H 4 5 A mutant and its existence as a hexamer. Although these residues were conserved in hydrolyzing and non-hydrolyzing epimerases, they may not play key roles in catalysis or reside in the active site of the mammalian epimerase, since no X-ray crystal structures of the mammalian epimerase exist. However, to further probe active site residues of the hydrolyzing epimerase, the X-ray crystal structure of RffE could be used as a guide to identify conserved key catalytic residues. Although crystallized with U D P - G l c N A c , the active site in the crystal structure of RffE shows only bound U D P . 5 5 It is possible that the sugar moiety is disordered or the sugar was converted to 2-acetamidoglucal which migrated away. In any event, mutagenesis studies were conducted by Dr. Jomy Samuel based on this structure. 1 2 3 Three acidic residues, conseved in hydrolyzing and non-hydrolyzing U D P - G l c N A c 2-70 epimerases, located where the G l c N A c portion is predicted to bind, were targeted for mutagenesis. The corresponding mutants, D 9 5 N , E112Q, D131N, showed a dramatic decrease in activity (>10 000 fold decrease in fccat) while binding was not significantly affected (KM values of ~3 m M for all three mutants versus 0.6 m M for wild-type enzyme) indicating that the three residues are important for catalysis. Since wild type RffE was found to release U D P and 2-acetamidoglucal into solution, the rates of U D P release were also measured. In two mutants, U D P release was too slow to quantify. In the E122Q mutant, U D P release was comparable to that in wild type RffE (1 X 10"3 M min"1 vs. 1 X 10"2 M min" 1, respectively). Thus this residue was thought to be important in the second, addition step of the reaction. Three likely candidates for mutagenesis studies in the mammalian hydrolyzing epimerase, homologous to the three acidic residues in RffE, are D112, E134 and D143. These mutants could be probed kinetically to determine their importance in catalysis and binding. Unlike RffE, the mammalian epimerase is able to catalytically convert 2-acetamidoglucal to the more stable product, M a n N A c . It may be possible to use this ability to identify the role of the mutated residue in the individual steps of the reaction. If a mutant protein is able to catalyze the hydration of 2-acetamidoglucal, but not the conversion of U D P - G l c N A c , then the mutated residue is important in the a«ri-elmination of U D P . It may also be possible for a mutation to affect the hydration of 2-acetamidoglucal, yet still catalyze the ^/^/-elimination of U D P forming 2-acetamidoglucal. The production of these site directed mutants is ongoing in the Reutter lab, although as of yet, unsuccessfully. 1 2 4 71 A n X-ray crystal structure of the mammalian hydrolyzing epimerase would also be desirable. This would aid in identifying key residues in the active site of the epimerase, and give insight into the interdomain interactions and their roles in regulating hydrolyzing epimerase activity. Structural comparisons may also further our understanding of the mammalian epimerase. Attempts to crystallize the mammalian hydrolyzing epimerase, however, have also met with no success. 1 2 4 72 Section B: Production and Analysis of Site Directed Mutants of the Bacterial Hydrolyzing UDP-GlcNAc 2-Epimerase, SiaA, from Neisseria meningitidis MC58 Group B. 2.8 Bacterial Hydrolyzing UDP-GlcNAc 2-Epimerases The genes involved in the biosynthesis of sialic acids have been identified in several bacterial species. Much work has focused on the sialic acid genes in the bacteria, E. coli K l and N. meningitidis B , involved in the production of the oc-(2-8)-linked P S A capsule. 1 2 5" 1 2 8 In E. coli K l and N. meningitidis, the sialic acid genes are all localized in a single gene cluster, named neu and sia, respectively. The gene products NeuB, NeuD and NeuS in E. coli K l and SiaB, SiaC, and SiaD in N. meningitidis have been characterized as N e u N A c synthases, C M P - N e u N A c synthetases and C M P - N e u N A c transferases, respectively. The gene products, NeuC and SiaA, share 32 % sequence identity and are thought to catalyze the first committed step in the biosynthesis of sialic acids in bacteria. The reaction catalyzed by these enzymes has been in debate. Pedersen et al. reported that SiaA was a G l c N A c 6-phosphate 2-epimerase, which converts G l c N A c 6-phosphate to M a n N A c 6-phosphate.31 The action of an unidentified M a n N A c 6-phosphate phosphatase was also proposed, which dephosphorylates M a n N A c 6-phosphate forming M a n N A c . Evidence for the activity of a G l c N A c 6-phosphate 2-epimerase was obtained when G l c N A c 6-phosphate was incubated with E. coli crude cell lysates, supplemented with a recombinant over-expressed siaA gene (cloned from N. meningitidis). After the incubation, the products were analyzed by incubating the reaction mixture with acid phosphatase (to dephosphorylate C-6), followed by incubation with 73 N e u N A c aldolase (converting M a n N A c and pyruvate to NeuNAc) and detection of N e u N A c using a colorimetric assay (Figure 2.22). The formation of M a n N A c 6-phosphate was detected, although at very low levels (0.1 U/g of cells). When the same assay was done with purified SiaA, G l c N A c 6-phosphate 2-epimerase activity was diminished (0.0002 U/g of cells). The authors speculated that the purified enzyme was unstable. O , P O — v H O — A c H N O H G l c N A c 6 P acid phosphatase S iaA H O v A c H N O H G l c N A c 0 , P O H O H O N H A c 1-0 H P 0 4 H O H O H O O H N H A c pyruvate H 0 ' ° v M a n N A c 6 P neuraminic acid A c H N "OH aldolase t M a n N A c H O H O N e u N A c C O , H Figure 2.22 Assay used by Petersen et al. to detect the activity of a GlcNAc 6-phosphate 2-epimerase. The differing opinion is that NeuC and SiaA are hydrolyzing U D P - G l c N A c 2-epimerases catalyzing the same reaction as the mammalian epimerase. 3 2 , 1 2 9 NeuC and SiaA share 27 % and 25 % sequence identity with the N-terminal region of the mammalian hydrolyzing epimerase, respectively. Evidence that NeuC is a hydrolyzing epimerase came from the incubation of radiolabeled U D P - [ 1 4 C ] G l c N A c with purified N e u C . 1 3 0 When the products of the incubation were analyzed using high voltage paper electrophoresis, two spots appeared. A major spot was found to co-migrate with G l c N A c , thought to be hydrolyzed non-enzymatically from U D P - G l c N A c , and a minor spot was found to co-migrate with M a n N A c , thought to be formed enzymatically from NeuC. 74 Similar to the case of the mammalian hydrolyzing epimerase, more evidence was needed to identify the products of the reaction catalyzed by NeuC. At this point a collaboration between Dr. Vann and co-workers with the Tanner lab ensued. Dr. Andrew Murkin , a previous graduate student of the Tanner lab, proceeded to identify the substrate of NeuC using ! H N M R spectroscopy and 3 1 P N M R spectroscopy. 1 2 9 G l c N A c , G l c N A c 6-phosphate and U D P - G l c N A c were tested as substrates. N o reaction was observed with G l c N A c or G l c N A c 6-phosphate. When U D P -G l c N A c was incubated with NeuC, rather than observing the formation of M a n N A c and U D P , the slow formation of 2-acetamidoglucal and U D P were observed (19 % conversion after 43.5 hours). The radiolabeled paper electrophoresis experiment was conducted again and the major spot, thought to be G l c N A c , was found to co-migrate with 2-acetamidoglucal. The fact that M a n N A c was not observed by ' H N M R spectroscopy was explained by the insensitivity of the technique which would not detect minute amounts of M a n N A c compared to the radiolabeled experiment. Complementation studies were also conducted in E. coli lacking a functional neuC gene, and unable to produce a capsular polysaccharide. When a recombinant plasmid containing neuC was supplied to this deficient strain, capsular polysaccharide synthesis was restored. Based on these results, it was concluded that NeuC is a hydrolyzing epimerase necessary for the biosynthesis of P S A and under the in vitro assay conditions the activity of NeuC is "crippled" and may require a regulatory molecule or the action of another enzyme, found in vivo, for full act ivi ty. 1 2 9 The results found with NeuC argued against SiaA being a G l c N A c 6-phosphate 2-epimerase. Due to the "crippled" activity of NeuC, the study of SiaA was taken on by Dr. 75 Andrew M u r k i n . 1 3 1 The siaA gene was first cloned from TV. meningitidis genomic D N A donated by a collaborator, Dr. Warren W. Wakarchuk. The activity of purified SiaA was then analyzed in a similar fashion to NeuC using ! H and 3 1 P N M R spectroscopy. G l c N A c and G l c N A c 6-phosphate were tested and demonstrated not to be substrates of SiaA. This time when SiaA was incubated with U D P - G l c N A c (5 m M ) , the N M R spectra showed the formation of M a n N A c and U D P (but not 2-acetamidoglucal) with full conversion after 30 minutes. This result confirmed that SiaA was a hydrolyzing U D P - G l c N A c 2-epimerase. The observations made by Pedersen et al. may have been due to residual G l c N A c 6-phosphate 2-epimerase activity in the E. coli crude cell extracts, not encoded by siaA. The diminished G l c N A c 6-phosphate 2-epimerase activity, observed upon purification of SiaA, would also support this conclusion. A similar proposal was given by Ringenberg et al. who reported the cloning of nanE, from E. coli, and demonstrated that the in vitro activity of its gene product is a G l c N A c 6-phosphate 2-epimerase using radiolabeled H P L C assays. 3 2 The success of these experiments led Dr. Murkin to pursue mechanistic studies on SiaA similar to those conducted with the mammalian hydrolyzing epimerase. ' H N M R spectral time course experiments identified the first formed enzymatic product as oc-[2-2 H ] M a n N A c indicating that the reaction catalyzed by S iaA proceeds with a net retention of configuration at C - l , and with a deprotonation and solvent deuterium incorporation at C-2. The intermediacy of 2-acetamidoglucal was demonstrated when 2-acetamidoglucal was shown to be hydrated to M a n N A c using ' H N M R spectroscopy. Unlike the mammalian epimerase, enzymatic hydration of 2-acetamidoglucal by S iaA required the strict presence of U D P , with no reaction occurring in the absence of U D P . The rate of 76 hydration by S iaA was also much faster with 37 % conversion after 4.5 hours and full conversion after two days. The incubation of [ l " - 1 8 0 ] U D P - G l c N A c with SiaA, led to the formation of 1 8 0-labeled U D P , demonstrating that the loss of U D P occurs through a C - 0 bond cleavage. These results clearly showed that SiaA, like the mammalian hydrolyzing epimerase, catalyzes the a«//-elimination of U D P , followed by the ^ - h y d r a t i o n of 2-acetamidoglucal. The similarity between the reactions catalyzed by SiaA and the mammalian hydrolyzing epimerase was further demonstrated by the absence of a K1E, with [ 2 " - 2 H ] U D P - G l c N A c , and the lack of an observed PLX, with [ l " - 1 8 0 ] U D P - G l c N A c . 77 2.9 Project Goals The production of the key active site mutants of the mammalian hydrolyzing epimerase has not been accomplished. Also attempts to crystallize and solve the X-ray crystal structure of the mammalian epimerase have met with failure. With the fortunate discovery of the bacterial hydrolyzing U D P - G l c N A c 2-epimerase, SiaA, an alternate source is accessible for the generation of site directed mutants and X-ray crystallographic studies. The next section describes the generation and characterization of site directed mutants of SiaA. Three residues, D100, E122 and D131, homologous to the acidic residues mutated in the study of RffE, are chosen for mutagenesis. The study of the resulting mutant proteins, D100N, E122Q and D131N, provides further insight into the mechanism utilized by hydrolyzing U D P - G l c N A c 2-epimerases. 78 2.10 Site-Directed Mutagenesis of SiaA The recombinant plasmid encoding siaA, p A M 0 4 , was obtained from Dr. Murkin and the site directed mutants were cloned using the QuikChange® Site-Directed Mutagenesis K i t from Stratagene (Figure 2.23). The mutagenesis kit takes advantage of the fact that plasmids isolated from E. coli cells are methylated at the N-6 position of deoxyadenosine residues (denoted m 6 A ) . After the methylated parental plasmid is subjected to amplification using the polymerase chain reaction (PCR), the parental D N A is digested using the restriction enzyme DpnI. Dpn l cleaves parental D N A at the target sequence 5 ' - G m 6 A T C - 3 ' leaving the newly amplified plasmid intact. The resulting plasmid, which is nicked, is then transformed into E. coli cells where it is ligated, replicated and then isolated. primer with ^ mutation Psienta.1 plasmid! Mutant Plasmid Mutant Ftasmicf (nidted) Figure 2.23 Production of mutant plasmids using the QuikChange Site-Directed Mutagenesis Kit. The three residues in the mammalian hydrolyzing epimerase, D100, E l 22 and D131, were chosen for mutagenesis based on the homologous acidic residues found in the active site of the non-hydrolyzing epimerase, R f f E . 5 5 The recombinant siaA plasmid was amplified with complementary primers containing the desired mutations. The resulting mixtures of wild type siaA plasmid and mutant siaA plasmid were subjected to treatment with D p n l and then transformed into E. coli X L - 1 blue supercompetent cells. Mutant recombinant siaA plasmids were then purified from the cells. The complete 79 corresponding mutant siaA genes were sequenced in order to determine that the correct mutations were made with no errors. 80 2.11 Over-Expression of Mutant siaA Genes and Affinity Purification of SiaA Mutants The siaA gene was cloned by Dr. Andrew Murkin into the pET-30 vector from Novagen (Figure 2.24). Consequently, the mutant siaA genes are also contained in this vector. When the siaA genes are over-expressed, the resulting proteins contain the additional amino acids in the TV-terminal region shown in Figure 2.24. The hexa-histidine tag (His-tag) is used to purify proteins using affinity chromatography, while the S-tag is used in the fluorescence detection and quantification of proteins. Although the His-tag was used in the purification of SiaA proteins, it can be cleaved at the thrombin cut site or the additional residues can be cleaved at the X a cut site, using the appropriate restriction enzymes (thrombin or factor X a , respectively). ATG CAC CAT CAT CAT CAT CAT TCT TCT GGT CTG GTG CCA CGC GGT TCT GGT ATG Met | His His His His His His per Ser Gly |Leu Val Pro Arg |Gly Ser | Gly Met His-tag thrombin cut site AAA GAA ACC GCT GCT GCT AAA TTC GAA CGC CAG CAC ATG GAC AGC CCA GAT CTG Lys Glu Thr Ala Ala Ala Lys Phe Glu Arg Gin His Met Asp Ser Pro Asp Leu I I S-tag GGT ACC GGT GGT GGC TCC GGT ATT GAG GGT CGC Gly Thr Gly Gly Gly Ser Gly He Glu Gly Arg i_fC site L_ : ^ Xa cut site Figure 2.24 Portion of pET-30 vector in recombinant siaA plasmids. The individual recombinant mutant siaA genes were transformed into E. coli B L -21 (DE3) cells for over-expression. After induction with 0.5 m M IPTG and further growth for three hours, the cells were harvested, lysed, and purified using affinity chromatography (Nickel-Sepharose® resin). The six histidine residues in the His-tag chelate to divalent nickel ions bound on the column. Un-tagged proteins are eluted off in 81 buffer without imidazole, while the mutant proteins were eluted off with 500 m M imidazole. After dialysis of the mutant proteins in buffer without imidazole (three times), the resulting pure desalted proteins were then lyophilized and stored at -80 °C. The mutant proteins were estimated to be greater than 90 % pure using SDS P A G E (Figure 2.25) and used in assays by re-dissolution in buffer or deuterated buffer. Figure 2.25 SDS PAGE of affinity purified hexa-histidine tagged SiaA and mutant proteins. Lane A) molecular mass standards of 66 kDa and 29 kDa, B) SiaA, C) D100N, D) E122Q and E) D131N. 82 2.12 Characterization of SiaA Mutant Activity Kinetic studies are often used to probe the impact of the mutation on catalytic rates and binding. Unfortunately, kinetic analysis of the three SiaA mutant proteins using the coupled U D P assay revealed that they were severely impaired catalytically (greater than 1000 times compared to wild-type SiaA), so that accurate kinetic parameters could not be obtained. It was mentioned previously, however, that mutation of a key catalytic residue may lead to the identification of its role in the individual chemical steps of the hydrolyzing epimerase reaction (^//-elimination or syn-hydraXion). Thus mutant S iaA activities, incubated with U D P - G l c N A c , and 2-acetamidoglucal and U D P , were monitored using ' H N M R spectroscopy. Each mutant was re-hydrated in deuterated buffer and their activities tested with either U D P - G l c N A c or 2-acetamidoglucal and U D P . The formation of M a n N A c was again observed by monitoring the appearance of the anomeric proton signals at 4.92 and 5.02 ppm. The ] H N M R spectroscopic time courses of D100N, E122Q and D131N, all showed the slow, partial conversion of U D P - G l c N A c to M a n N A c after seven days of incubation (3.5 %, 7.2 % and 11.8 %, respectively). A s a comparison, full conversion of U D P - G l c N A c to products for wild type SiaA (prepared similarly to the mutant proteins) was observed after 30 minutes with one third o f the amount o f protein used in the mutant incubations (665 p:g vs. 2 mg). Interestingly, the enzymatic incubations of E122Q and D131N revealed the appearance of a new signal at 5.59 ppm, before the formation of M a n N A c was detectable (Figure 2.26, t = 24 hours). This peak corresponds to the H - l signal of 2-acetamidoglucal, indicating that the mutants are able to catalyze its direct formation where the wild-type enzyme does not. Further conversion to 2-acetamidoglucal 83 and its hydration are observed at later time points (Figure 2.26, t = 7 days). A s expected, incubations with 2-acetamidoglucal and U D P demonstrated that E122Q and D131N were able to catalyze the hydration of 2-acetamidoglucal to M a n N A c (7.4 % and 9.6 % conversion after seven days, respectively) (Figure 2.27). Mutant hydration rates of 2-acetamidoglucal were also slower than the wi ld type enzyme (which showed 100 % conversion after two days). N o glycal hydration was observed with the D100N mutant or a control lacking enzyme under similar conditions. Table 2.1 summarizes the results for the incubation experiments. t = o Uridine UDP-GlcNAc \ t = 1 day 2-acetamidoglucal / t = 7 days 6.50 6.00 p p m 5.50 5.00 Figure 2.26 'H NMR spectra monitoring the reaction of D131N with UDP-GlcNAc (5 mM). t = o Uridine 2-acetamidoglucal t = 7 days ManNAc 6.50 6.00 ppm 5.50 5.00 Figure 2.27 'H NMR spectra monitoring the reaction of E122Q with 2-acetamidoglucal and UDP (5 mM each). 84 Table 2.1 Summary of wild-type and mutant SiaA protein activities determined using 'H NMR spectroscopy. "Sum of integrations from a- and B-anomers.b Not applicable. protein substrate(s) time % conversion to intermediate or product Glycal ManNAc" wild type UDP-GlcNAc 30 min 0 100 glycal and UDP 4.5 h N / A b 37 2 days N / A b 100 D100N UDP-GlcNAc 24 h 0 0 7 days 0 4 glycal and UDP 7 days N / A b 0 E122Q UDP-GlcNAc 24 h 9 0 7 days 14 7 glycal and UDP 7 days N / A b 7 D131N UDP-GlcNAc 24 h 3 0 7 days 9 11 glycal and UDP 7 days N / A b 10 85 2.13 Conclusions Following the discovery that S iaA is a bacterial hydrolyzing U D P - G l c N A c 2-epimerase, mutagenesis experiments were initiated to probe the role of active site residues in hydrolyzing epimerases. The resulting mutants were found to be severely catalytically impaired to the point where no reliable kinetic data could be obtained using the coupled U D P assay. Incubations of the mutant SiaA proteins were then monitored using ' H N M R spectroscopy. A l l mutants catalyzed the extremely slow conversion of U D P - G l c N A c to M a n N A c and U D P . Prior to the formation of M a n N A c , the E122Q and D131N mutants catalyzed the direct formation and release of 2-acetamidoglucal. This result provides direct evidence for the formation of 2-acetamidoglucal as an intermediate in hydrolyzing epimerase reactions. The impaired activity of these mutants implicates these residues as important for catalysis. Although no kinetic data are available, the mutation of homologous residues in RffE also resulted in a significant loss of activity without affecting substrate binding. Possible roles for these residues are the bases which deprotonate U D P - G l c N A c or activate water for hydration. These residues could also stabilize an oxocarbenium ion intermediate/transition state in each step of the reaction. The E l 2 2 and D131 residues could be important in the hydration step of the reaction due to the observed premature release of 2-acetamidoglucal in their corresponding mutants. The residue homologous to E l 2 2 in RffE, E l 12, was also proposed to be important in the second ^ - a d d i t i o n step of the reaction. The E112Q mutant was severely impaired catalytically, yet, was able to catalyze U D P release at a rate comparable to that of the wild type enzyme. Interestingly, 86 in NeuC, the residue which aligns with D131 is a histidine (HI31). NeuC was shown only to catalyze the slow formation of 2-acetamidoglucal and U D P . These observations further implicate the role of E l 2 2 and D131 in the hydration step of the reaction. Another possibility is that these acidic residues play an important role in allosteric regulation. Kinetic analysis of wi ld type SiaA, demonstrated that catalysis was strongly cooperative (Hi l l coefficient of 1.9) suggesting that S iaA may exist as a dimer. It is possible that these mutations also affect the cooperativity observed. Further studies, however, are needed to address these issues. 87 2.14 Future Directions and Related Studies Current work is ongoing in the lab of Dr. Natalie Strynadka in solving the X-ray crystallographic structure of SiaA by a graduate student, Ho Jun Lee. This structure would provide a picture of the active site in hydrolyzing epimerases and provide insight into the roles of the active site residues. This structure could also be used for the rational design of hydrolyzing epimerase inhibitors. Untagged and His-tagged versions of S iaA have been crystallized. Unfortunately, these crystals gave poor diffraction patterns insufficient for further analysis. 1 3 2 Variants of the three mutant SiaA proteins are also being crystallized and it is only a matter of time before the crystal structure of SiaA is obtained. Another goal is to co-crystallize S iaA bound to substrate analogs. The solved crystal structure of RffE, co-crystallized with U D P - G l c N A c , only revealed the presence of U D P in the active site. Although the sugar moiety may have been disordered, it is also possible that during the crystallization process the substrate was turned over to 2-acetamidoglucal which migrated away from the active site. Co-crystallization of SiaA with substrate analogs containing non-hydrolyzable glycosidic linkages could be the solution to determining which residues are near the sugar moiety. In a recent study, initial success with inhibitors of the mammalian epimerase has been achieved with C - l glycosidic-linked analogues (Figure 2.27, compounds 1-6). These compounds are designed to incorporate the sugar portion and charged U D P portion of potential intermediate and transition state structures of the hydrolyzing epimerases. The C - l glycosidic linkages are non-hydrolyzable, while the glycal and exo-glycal components could mimic potential oxocarbenium ion intermediates/transition states or 88 the 2-acetamidoglucal intermediate. Preliminary studies with the mammalian epimerase have demonstrated that these compounds inhibit enzyme activity 40 - 60 % when equimolar amounts of the analogues and U D P - G l c N A c are used (1.25 m M each). Several other methylene-linked analogues have been synthesized and their inhibition studies are pending (Figure 2.27, compounds 7-11).1 3 4 , 1 3 5 OMe 6 Figure 2.28 Potential inhibitors of the hydrolyzing epimerasesi 89 2.15 Experimental Procedures Section A : Mammalian Hydrolyzing Epimerase Experiments 2.15.1 Materials A l l materials and chemicals used in experiments were purchased from Sigma-Aldrich or Fisher Scientific unless otherwise stated. 1 8 0 enriched water (95 %) and D 2 0 were purchased from Icon Isotopes. 1-deoxy nojirimycin, 2-acetamido-l,2-dideoxy nojirimycin and 2-acetamido-2-deoxy-D-glucono-l,5-lactone were purchased from Toronto Research Chemicals Inc.. Sf9 cells, SF9 S F M II media, 2 m M L-glutamine, 100 X antibiotic/antimycotic solution and Trypan Blue stain were purchased from Invitrogen. Acrodisc syringe filters (0.2 um pore, 25 mm, H T Tuffryn Membrane) were purchased from Pall Corp. High-Trap™ Q-Sepharose HP columns (1 mL) were purchased from Amersham Pharmacia Biotech. Chelex® 100 resin (200-400 mesh, N a + form), AG® 1-X8 resin (100-200 mesh, formate form), and Bio-Gel® P-2 resin were purchased from B io -Rad Laboratories. Amicon Ultra-4 centrifugal protein filters (4 m L , 10 000 M W C O ) were purchased from Millipore. 2.15.2 General Methods 2.15.2.1 General Insect Cell Culturing and Virus Production Methods 50 m L of Sf9 S F M II media, supplemented with 20 u\M of L-glutamine and 1 % antibiotic/antimycotic solution, were inoculated with 2 m L of Sf9 cells cryogenically 90 stored in liquid N2. Sf9 cell cultures were incubated at 120 x g and 27 °C. Viable cell densities were determined by diluting the cell culture in additional media and counted using a hemocytometer, a compound microscope (Olympus 1X70) and Trypan Blue stain (2X dilution to insect cell media/virus culture). Insect cell cultures were infected with recombinant virus at a density of 1 X 10 6 cell mL" 1 (see mammalian epimerase expression below), or maintained indefinitely by re-inoculating 50 m L of fresh media at a cell concentration of 1 X 10 4 cell mL" 1 . 2 m L of 1 X 10 8 cell mL" 1 cultures were also flash frozen and stored in liquid N2 as stock cultures. Recombinant virus was produced by infecting cells with virus at a cell concentration of 1 X 10 6 cell mL" 1 , followed by incubation at 120 x g, 27 °C and 8 days. The cells were pelleted via centrifugation at 5000 x g for 30 min, and the resulting supernatant containing baculovirus was sterilized by syringe filtration and stored at -20 °C for 1 year or at -80 °C indefinitely. 2.15.2.2 General Enzyme Methods A l l proteins were handled at 4 °C unless otherwise stated. Protein concentrations were determined according to the method of Bradford 1 3 6 using a Cary3E U V / V i s spectrophotometer with bovine serum albumin as the standard. Protein purity was estimated using SDS P A G E gel electrophoresis and visualized using Coomassie Blue stain according to the method of Laemml i . 1 3 7 Standard molecular masses used in SDS P A G E were B S A (66 kDa) and carbonic anhydrase (29 kDa). A l l buffers were prepared and adjusted to the correct p H using HC1 or N a O H as the acid or base. Deuterated buffers were prepared by evaporating buffers prepared in 91 H2O to complete dryness and reconstitution to the initial volume using D2O, twice. The pD of the solution was determined by adding 0.4 p H units to the reading from the pH reading according to Glasoe and Long (pD = p H + 0.4). 1 3 8 One unit (U) of enzyme activity is defined as the amount of enzyme which converts 1 u M of substrate into product in one minute at 37 °C. 2.15.2.3 General Synthetic Methods Reaction progress was monitored using thin-layer chromatography on aluminum backed sheets of silica gel 60, 0.2 mm thickness, with fluorescent indicator UV254 (Machery-Nagel, Germany). U V absorbing compounds were first visualized using 254 nm of U V light. Non U V absorbing compounds were visualized by spraying the plate with a solution of ammonium molybdate tetrahydrate (((NH4)6Moy0 24-4H20, 24.0 g), ammonium cerium (IV) nitrate ((NH 4)2Ce(N0 3)6, 0.5 g), H 2 S 0 4 (28 mL) , and water (500 mL) , followed by charring with a heat gun. Flash chromatography was performed using silica gel (230-400 mesh, Silicycle). 2.15.2.4 N M R and Mass Spectrometry } H N M R spectra were obtained on Bruker A V 3 0 0 or A V 4 0 0 spectrometers at field strengths of 300 or 400 M H z , respectively. 1 3 C N M R spectra were obtained at field strengths of 75 or 100 M H z , respectively. Proton-decoupled 3 1 P N M R spectra were recorded on either spectrometer at field strengths of 121.5 or 162 M H z , respectively. Mass spectrometry was performed by the Mass Spectrometry Centre at the University of British Columbia ( U B C ) by liquid secondary ionization mass spectrometry (LSI -MS) , 92 using a Kratos Concept LI H Q mass spectrometer or electrospray ionization mass spectrometry (ESL-MS), using a Waters Micromass L C T mass spectrometer. Protein mass spectrometry was performed by Shouming He using a Perkin-Elmer Sciex APL300 electrospray mass spectrometer. 2.15.3 Over-Expression and Purification of the Mammalian Hydrolyzing UDP-GlcNAc 2-Epimerase The recombinant baculovirus encoding the mammalian hydrolyzing epimerase gene, gne, was obtained from the Reutter lab. 3 5 Mammalian U D P - G l c N A c 2-epimerase was over-expressed by inoculating 50 m L cultures of Sf9 Serum Free Medium II media, supplemented with 20 ( i M of L-glutamine and 1 % antibiotic/antimycotic solution, with 2 m L of Sf9 cells from -80 °C storage. Cells were grown to a cell density of 2 x 10 6 cells mL"' at 120 x g and 27 °C. The cell cultures were then infected with the recombinant baculovirus at a multiplicity of infection of 1 (1 viral particle per cell). After an optimal infection period of 60 h, the cell culture was harvested by centrifugation at 5000 x g for 20 min at 4 °C. The resulting cell pellet was resuspended in 10 m L of phosphate buffer (10 m M N a H 2 P 0 4 , pH 7.5, containing 1 m M dithiothreitol, 1 m M E D T A , and 1 m M phenylmethylsulfonyl fluoride) and the cells lysed by drawing the cell suspension in and out of a syringe-needle (PrecisionGlide Needle, 26 gauge, 5/8" length, Becton Dickinson & Co.). The crude cell lysate was clarified by centrifugation (5000 x g, 60 min) and syringe filtration. The filtrate was loaded onto an ion-exchange column (1 mL, High-Trap™ Q-Sepharose HP column) pre-equilibrated with column buffer (10 m M NaH2P04, p H 7.5, containing 1 m M D T T , and 1 m M E D T A ) . The column was eluted with a linear 93 gradient of 0-1 M N a C l in column buffer (20 m L of total buffer) and monitored at 280 nm. Fractions containing the enzyme were used directly in experiments. Enzyme in deuterated buffer was prepared as above, except that both the lysis and the column buffers were prepared using D 2 O (pD 7.9). 2.15.4 Characterization of Mammalian Hydrolyzing Epimerase Activity 2.15.4.1 Stereochemistry and Solvent Deuterium Isotope Incorporation Studies A sample of U D P - G l c N A c 2-epimerase in deuterated column buffer (950 U.L, 950 p:g of protein) was added to an N M R tube, and a ' H N M R spectrum was taken. A sample of U D P - G l c N A c prepared in the same buffer (50 f iL, 100 m M ) was added, and the resulting solution (5 m M U D P - G l c N A c , 1 mL) was incubated for 10 h at 25 °C. ' H N M R spectra were taken at timed intervals. A control reaction of U D P - G l c N A c in the same buffer (5 m M , 1 mL) was also monitored under identical conditions. To assess mutarotation, a commercial sample of M a n N A c (5 m M ) was dissolved in 1 ml of deuterated phosphate buffer and placed in an N M R tube. ' H N M R spectra were taken over the course of 30 minutes at which point no change in the ratio of the anomeric 1 ^ signals was evident. The same equilibrated sample was characterized further using C N M R , 'H-coupled 1 3 C N M R and H M Q C N M R spectroscopy. 2.15.4.2 Catalytic Competence of 2-Acetamidoglucal 2-Acetamidoglucal was synthesized by the procedure outlined by Pravdic et a l . " A solution of U D P - G l c N A c 2-epimerase in deuterated buffer (950 p:L, 950 fig of protein) 94 was prepared, and a ' H N M R spectrum was collected. A sample of 2-acetamidoglucal, prepared in the same buffer, was added and the resulting solution (5 m M 2-acetamidoglucal, 1 mL) was incubated for 29 hours. *H N M R spectra were taken at timed intervals. Identical incubations were conducted in the presence of 1, 5 and 10 m M U D P . 2.15.4.3 Catalytic Competence of the Oxazoline of ManNAc and 1-F GlcNAc The oxazoline of M a n N A c was synthesized by the procedure of Khorl in et al.]04 and 1-F G l c N A c was synthesized according to the method of Micheel and Wul f f . 1 0 3 'ft N M R spectroscopic time course incubations were conducted similarly to those with 2-acetamidoglucal, except that 5 m M of each substrate analog (1 m L total volume), in the presence and absence of 5 m M U D P , was incubated for 2 days. 2.15.5 Kinetic Characterization of Mammalian Hydrolyzing UDP-GlcNAc 2-Epimerase 2.15.5.1 KIE Experiment using [2" 2H]UDP-GlcNAc [ 2 " - 2 H ] - U D P - G l c N A c was prepared according to the procedure by Morgan et a l . 1 6 ' H N M R spectroscopy and mass spectrometry confirmed the extent of deuterium incorporation to be >95%: -ve E S I M S mlz 607 ( M - H + , 2 " - 2 H , 100). Concentrations of stock substrate solutions were calculated using the uridine chromophore at A2a with an £262 value of 9890 M " 1 cm" 1. Kinetic parameters were measured using a continuously coupled assay for U D P formation. 1 0 9 Each assay contained 50 m M N a H 2 P 0 4 buffer (pH 7.5), 10 m M M g C l 2 , 2 m M PEP, 0.2 m M N A D H , 188 units of pyruvate kinase (Type II 95 from rabbit muscle, Sigma-Aldrich), 250 units of lactate dehydrogenase (Type II from rabbit muscle, Sigma-Aldrich), and 5-300 u M U D P - G l c N A c or [ 2 " - 2 H ] - U D P - G l c N A c (800 p X total assay volume). U D P - G l c N A c 2-epimerase (10 (ig) was added to initiate the enzymatic reaction. Reaction rates were measured at 37 °C by monitoring the decrease in absorbance at 340 nm. The resulting calculated enzyme rate was plotted as a function of substrate concentration and the kinetic parameters were determine by a direct fit of the data to the H i l l equation using the computer program GraFi t . 1 3 9 This program performs a non-linear regression analysis on the data following the method of Marquart, 1 4 0 and reports an error for the data based on the deviation of the data from the calculated curve-of-best-fit. With different enzyme preparations, the error between different data sets did not vary by more than 10 %. 2.15.5.2 Inhibition Studies with Glycosidase Inhibitors The same kinetic assay conditions used in the K I E experiment were used in the inhibition studies with the glycosidase inhibitors, 1-deoxy nojirimycin, 2-acetamido-l,2-dideoxy nojirimycin, N A G thiazoline and 2-acetamido-l,2-dideoxy-D-glucono-l,5-lactone. 1 m M of each inhibitor was assayed with saturating conditions of 1 m M U D P -G l c N A c . 96 2.15.6 Test for C-O versus P-O Bond Cleavage and PIX Experiment 2.15.6.1 Negative Test for C-O vs. P-O Bond Cleavage A sample of U D P - G l c N A c 2-epimerase in column buffer (300 uL, 300 | lg of protein) was mixed with 300 u L of 95% 1 80-enriched H 2 0 and incubated at 30 °C for 10 h. The reaction mixture was then passed through a protein centricon (Amicon Ultra-4, 4 mL, 10 000 M W C O , Millipore) at 5000 x g for 30 min and the eluate was applied to a size exclusion column (Biogel P-2, 200-400 mesh, 2.5 cm x 70 cm column dimensions) and eluted with distilled water. Fractions exhibiting absorbance at 254 nm (UDP 1 31 chromophore) were collected and lyophilized to dryness. H and proton-decoupled P N M R spectra of relevant fractions revealed that they contained U D P . Mass spectral analysis indicated the absence of the 1 8 0- label in the U D P produced: -ve L S I - M S (thioglycerol) mlz 425 ( M - FT (monosodium salt), 1 6 0 , 100). 2.15.6.2 PIX Experiment and Positive Test for C-O vs. P-O Bond Cleavage [ l " - 1 8 0 ] U D P - G l c N A c was synthesized according to literature procedures. 5 3 A solution of [ l " - 1 8 0 ] U D P - G l c N A c in deuterated buffer (450 (iL, 17 m M ) was placed in an N M R tube, and Chelex-100 resin was added (20 mg of 200-400 mesh, N a form, previously rinsed with D2O) . ' H and proton-decoupled 3 1 P N M R spectra were taken. The scrambling experiment was initiated by the addition of U D P - G l c N A c 2-epimerase in deuterated buffer (200 uX, 200 \ig of protein), and the resulting solution was incubated for 24 h at 25 °C. ' H and proton-decoupled 3 1 P N M R and were taken at timed intervals. The acquisition parameters for the proton-decoupled 3 1 P N M R spectra were a sweep 97 width = 2437 Hz , acquisition time = 13.4 s, delay between pulses = 2.0 s, and pulse width = 10 [is. The final solution was then passed through a protein centricon (Amicon Ultra-4, 4 mL, 10 000 M W C O , Millipore) at 5000 x g for 30 min and the eluant was applied to a size exclusion column (Biogel P-2, 200-400 mesh, 2.5 cm x 70 cm) and eluted with distilled water. Fractions exhibiting absorbance at 254 nm (uridine chromophore) were collected and lyophilized to dryness. *H N M R spectra of relevant fractions revealed that they contained U D P . Mass spectral analysis confirmed the presence of the O-label in the U D P produced: -ve E S I - M S mlz 405 ( M - H + , 1 8 0 , 100), 403 ( M - H + , i e O , 46.5). 3 1 P N M R spectroscopy confirmed that the 1 8 0- label was incorporated into the (3-phosphate: 3 1 P N M R ( D 2 0 ) 5-8.001 (d, Jp.p = 21.0 Hz , 0.32P, p -P- 1 6 0) , -8.030 (d, Jp.p = 21.0 Hz, 0.68P, p -P- 1 8 0) , -9.645 (d, J p _ p = 21.0 Hz, IP, oc-P). 2.15.7 Synthesis of Substrate Analogs 2.15.7.1 Synthesis of 2-acetamido-l,2-dideoxy-D-arabino-l-enopyranose (2-acetamidoglucal) 2-Acetamidoglucal was synthesized according to the method of Pravdic et al" ' H N M R ( D 2 0 ) : 5 6.59 (s, 1H, H - l ) ; 4.14 (d, 1H, JJA = 6.6 Hz , H-3); 3.87 (m, 1H, J4,s= 9.0 Hz, y 5 , 6 = Js,6- = 4.0 Hz , H-5); 3.75 (d, 2H , J5fi = y 5 > 6 . = 4.0 Hz , H-6, H-6 ' ) ; 3.65 (dd, 1H, J3.4 = 6.6 Hz, J4,s = 9.0 Hz , H-4); 1.93 (s, 3H, H 3 C C O N H ) . +ve E S I - M S mlz 226 ( M + Na + ) . 98 2.15.7.2 Synthesis of 2-acetamido-l,2-dideoxy-a-D-glucopyranosyl fluoride (<x-F-GlcNAc) oc-F-GlcNAc was synthesized according to the method of Micheel and Wul f f . 1 0 3 ' H N M R ( D 2 0 ) : 5 5.65 (dd, 1H, Jx,i = 2.7 Hz, Jxf = 52.8 Hz , H - l ) , 3.91 (ddd, 1H, Jh2 = 2.6 Hz , y2,3 = 10.6 Hz, J2,F = 28, H-2), 3.82-3.64 (m, 4H), 3.50 (m, 4H,J= 9.3 Hz, H-4), 1.96 (s, 3H, CH 3-acetyl). 1 9 F N M R ( D 2 0 ) -146.2 (s, IF, F - l ) . +ve ESI -MS mlz 246 ( M + Na + ) . 2.15.7.3 Synthesis of 2-methyl-4,5-dihydro-(l,2-dideoxy-fi-D-mannopyranoso)[2,l-<f|-l,3-oxazole (ManNAc oxazoline) Per-acetylated M a n N A c oxazoline was synthesized by Dr. Jomy Samuel according to the method of Khorl in et al. and de-acetylated in this study using the same procedure. 1 0 4 ' H N M R ( D 2 0 ) : 5 5.27 (d, 1H, J 1 > 2 = 1 Hz, H - l ) , 5.24 (dd, 1H, J2,3 = 7.5 Hz, J 3 )4 = 3.8 Hz , H-3), 4.53 (dd, 1H, H-2), 4.23 (dd, 1H, J 4 j 5 = 9.0 Hz , H-4), 3.86-3.75 (m, 2H , H-5, H-6), 3.64 (dd, J6,6- = 12.1 Hz, H-6'). +ve E S I - M S mlz 226 ( M + Na + ) . 2.15.7.4 Synthesis of N A G Thiazoline N A G thiazoline was synthesized according to the method of Knapp et al.u6 ! H N M R ( D 2 0 ) 5 6.34 (d, 1H, J , , 2 = 7.0 Hz, H - l ) , 4.33-4.26 (m, 1H, H-2), 4.10 (dd, 1H, J 2 , 3 = 4.8 Hz, J 3 , 4 = 3.7 Hz , H-3), 3.72 (dd, 1H, J5fi = 2.5 Hz, = 12 Hz , H-6), 3.57 (dd, 1H, J5,6 = 6.3 Hz, y6,6' = 12 Hz, H-6 ' ) , 3.54 (ddd, 1H, / = 1 Hz , J 3 j 4 = 3.7 Hz , J 4 , 5 = 9.1 Hz , H-4), 3.31 (ddd, 1H, J 4 , 5 = 9.1 Hz, J5>6 = 2.5 Hz, J5 ;6> = 6.3 Hz , H-5), 2.24 (d, 3H, / = 2 Hz, CH 3 -oxazoline). +ve ESI -MS mlz 242 ( M + Na + ) . 99 2.15.7.5 Preparation of disodium uridine 5'-(2"-acetamido-2"-deoxy-[2"-2H]-a-D-glucopyranosyl diphosphate) ([2"- 2H]UDP-GlcNAc) [ 2 " - 2 H ] U D P - G l c N A c was prepared according to the procedure by Morgan et al: The plasmid pK186 (encoding RffE) and pUSOl (encoding U D P - M a n N A c dehydrogenase) were obtained from Dr. Jomy Samuel. ] H N M R spectra were identical to those previously reported. ' H N M R spectroscopy and mass spectrometry confirmed the extent of deuterium incorporation to be >95%: -ve E S I - M S m/z 607 ( M - H + , 2 H - 2 " , 100). 2.15.7.6 Synthesis of disodium uridine 5'-(2"-acetamido-2"-deoxy-[l"- 1 80]-a-D-glucopyranosyl diphosphate) ([l"- l sO]UDP-GlcNAc) [ l " - 1 8 0 ] U D P - G l c N A c , was synthesized according to the procedure by Morgan et a l . 5 3 ' H N M R spectra were identical to those reported and to commercially available unlabeled U D P - G l c N A c . The extent of 1 8 0 incorporation was determined to be 68% by mass spectral analysis: -ve L S I - M S (thioglycerol) m/z 614 ( M - H + (mono-l i thium salt), 1 8 0 , 100), 612 ( M - H + (mono-lithium salt), 1 6 0 , 46.5), 608 ( M - H + , 1 8 0 , 46.5), 606 ( M -H + , 1 6 0 , 21.6). The location of the l s O-labe l was confirmed by 3 1 P N M R spectroscopy: 3 , P N M R ( D 2 0 ) 8 -10.095 (d, IP, y p . p = 21.0 Hz , a-P), -11.785 (d, 0.32P, J p . p = 21.0 Hz, 8-P- 1 6 0),-11.805 (d, 0.68P, J p _ p = 21.0 Hz, B-P- 1 8 0) . 100 Section B: Mutant SiaA Experimental 2.15.8 Materials and Methods The same materials and methods apply as above with the following additions. 2.15.8.1 Materials The plasmid encoding siaA, p A M 0 4 , was obtained from Dr. Andrew Murkin. The Quik Change® Site-Directed Mutagenesis Kit was purchased from Stratagene. Luria Bertani media components (Tryptone and Yeast Cel l Extract) were purchased from Difco Laboratories. Kanamycin sulfate was purchased from Gibco. Aprotonin and pepstatin A were purchased from Boehringer Ingelheim. Spectra/Por® Biotech R C dialysis tubing (12 000 - 14 000 M W C O ) was purchased from Spectrum Laboratories, Inc. Chelating Sepharose® Fast Flow resin was purchased from Amersham Biotech. 2.15.8.2 General DNA Methods Primer concentrations were determined at A260 using Equation 2.3, where ' W A " , 'Wc", "No" and " A V represent the number of the respective bases in the primers. 1 4 1 %(\M) - A26Q 5 N ^ + Q 7 N c + l 2 N G +0.84iV T Equation 2.3 Equation used to determine primer concentration. D N A plasmid concentrations were determined at ^260 using Equation 2.4. 1 4 1 ^(ug/mL) = ^ 2 6 0 x 50 Equation 2.4 Equation used to determine plasmid concentration. 101 D N A purity was determined by agarose gel electrophoresis (1 % cross-linked agarose gel) using 100 bp and 1 kbp D N A standards (Invitrogen) and visualized by staining with ethidium bromide and U V light. Plasmid preparation and purification from 10 m L cell cultures were done using the Wizard® Plus Miniprep D N A Purification System (Promega) according to the manufacturer's protocol. D N A sequencing and production of primers were done by the Nucleic Acids and Protein Sequencing (NAPS) Unit at U B C . A l l plasmids and primers were stored in water at -20 °C. 2.15.8.3 Preparation of Chemically Competent E. coli cells and Transformation of Plasmids E. coli (strains BL-21 (DE3) or X L - 1 Blue) competent cells were prepared using a modified procedure by Inoue et al.142 10 m L of S O B culture was inoculated with the E. coli of interest and incubated overnight at 225 x g and 37 °C. A 250 m L S O B culture was then inoculated with the 10 ml cell culture, incubated for 1 hour at 225 x g and 37 °C and then at 200 x g and 22 °C until an OD6oo of 0.6. The cell culture was chilled on ice for 10 min, centrifuged at 2500 x g at 4 °C for 5 min. The supernatant was poured off and the resulting cell pellet was resuspended in 80 ml of transformation buffer (10 m M pipes buffer, 15 m M C a C l 2 , 250 m M KC1, adjusted to p H 6.7 with 1 M K O H , followed by the addition of 55 m M M n C l 2 ) chilled on ice for 10 min and centrifuged at 2500 x g at 4 °C for 5 min. The supernatant was again poured off and the cell pellet was resuspended in 20 ml of transformation buffer with 7 % D M S O , aliquoted into 500 ml portions, flash frozen using liquid N 2 and stored at -80 °C. 102 Transformation of recombinant plasmids was performed by thawing the frozen competent cells on ice for 30 min, adding 100 j i l of the competent cells to a pre-chilled polypropylene tube and adding 5 | i l o f plasmid solution. The resulting mixture was swirled gently 10 times, chilled on ice for 30 min, heat shocked at 42 °C in a water bath for 45 sec and further chilled on ice for 2 minutes. 1 ml of L B media, pre-warmed to 37 °C, was added to the tube and the mixture incubated at 225 x g and 37 °C for 1 hour. A n agarose plate containing the media of choice was then inoculated with 25, 50 and 100 ( iL of the cell mixture and the plates incubated overnight at 37 °C. Cells transformed with the recombinant plasmid are selected for by adding the appropriate antibiotic for which the plasmid has resistance. 2.15.9 Site-Directed Mutagenesis of siaA Mutant recombinant plasmids of the recombinant plasmid encoding wild-type siaA, p A M 0 4 , were prepared according to the protocol of the Quik Change® Site-Directed Mutagenesis Ki t from Stratagene. p A M 0 4 was subjected to amplification using the polymerase chain reaction (PCR) with oligonucleotide pairs which contain a mutation site and which are complementary to nucleotides surrounding the mutation site. Oligonucleotides were designed to have a melting temperature (T m ) >78 °C.with their target sequence and were calculated using Equation 2.5, where " % G C " is the percentage of guanosine and cytosine nucelotides in the primer, " N " is the length of the primer and " % mismatch" is the percentage of nucleotides in the primer which do not match the original sequence. 1 4 3 Oligonucleotide pairs that were used to introduce mutations in the 103 forward and reverse directions are listed in Table 2.2, with the mutated nucleotides underlined. T m = 81.5 + 0.41 (%GC) - 675/N - % mismatch Equation 2.5 Equation used to estimate melting temperature of primers. Table 2.2 Primers used in the SDM of pAM04. Primers for D100N WC01: 5 ' - G G T C A T G A T T C A C G G C A A C C G T T T A G A A G C A C T A G C - 3 ' (forward) WC02: 5 ' - G C G C C T G C T A G T G C T T C T A A A C G G T T G C C G T G - 3 ' (reverse) Primers For E122Q WC03: 5 ' - G C A G C C G T T T A G T T T G C C A T A T C C A A G G T G G T G - 3 ' (forward) WC04: 5 '-CTGT A C C A G AT A G T T C A C C T T G G ATATGGC-3 ' (reverse) Primers for D131N WC05: 5 ' - G G T G G T G A A C T A T C T G G T A C A G T A A A T G A C T C C A T T C G - 3 ' (forward) WC06: 5 ' - G T T T A C T A A T A G A A T G A C G A A T G A A G T C A T T T A C T G T A C C A G - 3 ' (reverse) P C R mixtures containing 5 u L of 10X reaction buffer, 1 u L of 25 m M dNTP mix, 120 ng of each primer, 20 ng of p A M 0 4 and 2.5 U of PfuTurbo® D N A polymerase, in a total volume of 50 j lL , were subjected to amplification using the iCycler iQ™ P C R machine (Bio-Rad). The cycles used were 1 cycle of 1 min at 95 °C, 20 cycles of 30 s at 95 °C, 30 s at 55 °C and 5 min at 72 °C, and 1 cycle of 7 min at 72 °C. The resulting P C R mixtures were incubated for 1 hour at 37 °C with 10 U of Dpnl restriction enzyme to digest the parental p A M 0 4 plasmid. 5 U.L of this mixture was transformed into X L l - B l u e supercompetent cells, which were then plated on L B agar plates containing 30 mg mL" 1 of kanamycin sulphate (which the plasmid has resistance for) and incubated overnight at 37 °C. Isolated colonies were used to inoculate 10 m L of L B media, which were grown overnight at 37 °C. The cells were then harvested via centriguation at 5000 x g for 10 min and the plasmids isolated using the Wizards® D N A Miniprep Purification System. The 104 whole mutant gene was sequenced to ensure the proper mutation had been introduced with no errors. 2.1.5.10 Over-expression and Purification of Hexa-Histidine Tagged SiaA Mutants The recombinant mutant plasmids were individually transformed into BL21 (DE3) chemically competent E. coli, which were incubated overnight at 37 °C with shaking at 225 x g in 10 m L of L B media containing 30 jig mL" 1 of kanamycin sulfate. The overnight cultures were poured into 500 m L cultures of L B medium containing 30 ug/mL of kanamycin sulphate and grown at 37 °C with shaking at 225 x g until an OD600 of 0.6-0.7 had been reached. Cells were induced for over-expression by the addition of 0.5 m M isopropyl P-D-thiogalactopyranoside (IPTG), and the cultures were allowed to continue to grow until an OD600 of 1.6-1.8 had been reached (approximately 4 h). Cells were harvested via centrifugation at 5000 x g for 30 min, resuspended in lysis buffer (20 m M sodium phosphate (pH 7.5), 10 m M imidazole, 0.4 M N a C l , 1 | lg /mL pepstatin A , and 1 (ig/mL aprotinin), and lysed at 20 000 psi in an ice-cooled French Press cell. The cell lysates were clarified by centrifugation at 6000 x g for 40 min and syringe filtration. A 10 m L column containing Chelating Sepharose Fast Flow resin (Amersham Biotech) was charged with 2 column volumes (CV) of 100 m M MSO4, followed by washing with 2 C V of distilled H 2 0 and 3 C V of start buffer (20 m M sodium phosphate, 10 m M imidazole and 0.4 M N a C l , p H 7.5). The clarified cell lysate was loaded at a rate of 2 m L min" 1, and start buffer (8 C V ) was passed through the column at a rate of 3 m L min"1 until no more flow-through protein eluted, as determined by monitoring at A2X0. To remove nonspecifically bound proteins, a stepwise gradient was then applied, consisting 105 of 2 C V each of 5 and 25 % elution buffer (20 m M sodium phosphate, 500 m M imidazole, and 0.4 M N a C l , p H 7.5). The histidine-tagged mutant SiaA protein was finally eluted with 3-4 C V of 100% elution buffer. Fractions containing the enzyme were pooled and dialyzed overnight against a 1:100 volume of dialysis buffer (20 m M sodium phosphate (pH 7.5)). The enzyme solution was then divided into 1.0 m L aliquots, flash-frozen in liquid N 2 , lyophilized and stored at -80 °C. Samples were reconstituted as needed by adding H 2 O or D 2 O and gently inverting several times. 2.15.11 Characterization of SiaA Mutant Activity Individual solutions of purified SiaA mutants in deuterated buffer (700 ul , 2 mg of protein) were added to an N M R tube. The enzymatic reaction was then initiated by the addition of a solution of U D P - G l c N A c (35 [iL, 100 m M ) and the resulting mixture (5 m M U D P - G l c N A c , 700 uL) incubated for 7 days at 37 °C. lH N M R spectra were taken at timed intervals. Similar incubation mixtures with 2-acetamidoglucal and U D P (both at 5 m M , 700 (iL) were prepared and monitored using ' H N M R spectroscopy. Enzymatic rates of the mutant proteins were determined using the same conditions as the coupled U D P assay used in the mammalian hydrolyzing epimerase. Maximal velocity (Fm a x) rates were measured under saturating conditions for the wild-type SiaA protein (10 m M U D P - G l c N A c ) and 2 mg of mutant protein. 106 Chapter 3: Identification and Characterization of a Pseudaminic Acid Synthase (NeuB3) from Campylobacter jejuni 107 3.1 Introduction3 Pseudaminic acid is an a-keto acid found as a post-translational modification on the flagellin proteins of the bacteria C. jejuni and H. pylori25'26 In these bacteria, this modification is necessary for the assembly of functional flagellae, and is therefore 21 26 30 required for their invasiveness. ' ' A t the time the work described in this thesis was begun, the biosynthetic pathway of pseudaminic acid was unknown, although a pseudaminic acid synthase was proposed to catalyze the key step via the condensation of P E P with 6-deoxy A l t d i N A c . A candidate gene, neuB3 from C. jejuni N C T C 11168, encoding a potential pseudaminic acid synthase, was identified by an insertional mutation that resulted in aflagellate Cjejuni21 Its gene product, NeuB3, was also found to share 35 % sequence identity with the N. meningitidis sialic acid synthase. With the cloning of neuB3 into a recombinant form in the lab of Dr. Wakarchuk, a collaborative effort ensued to study its gene product, NeuB3. In Chapter 3, the potential substrate, 6-deoxy A l t d i N A c , is synthesized and incubated with NeuB3 and PEP. The enzymatic incubation, monitored using 3 1 P and ' H N M R spectroscopy, reveals that NeuB3 catalyzes the formation of phosphate and pseudaminic acid. The isolation and characterization of the enzymatically produced pseudaminic acid confirms its stereochemistry to be identical to previously characterized derivatives. Further kinetic analysis of NeuB3 indicates that optimal catalysis proceeds at pH 7.0 and that NeuB3 requires a divalent metal ion for activity. A n additional 3 A version of this chapter has been published. Chou, W. K. , Dick, S., Wakarachuk, W. W., Tanner, M . E. (2005) Identification and characterization of NeuB3 from Campylobacter jejuni as a pseudaminic acid synthase. J. Biol. Chem. 280: 35922-35928. 108 experiment with NeuB3 and [2 - 1 8 0]PEP indicates that the operative mechanism of this enzyme is similar to that used by PEP-condensing synthases. 109 3.2 Expression and Purification of NeuB3 The neuB3 gene, Q1317, from Campylobacter jejuni N C T C 11168 was cloned by Dr. Scott Dick in the lab of Dr. Warren W. Wakarchuk and donated to the Tanner lab for study. The recombinant plasmid, pre-transformed into E. coli strain AD202 , was over-expressed by induction with 0.5 m M IPTG. After further growth for five hours, the cells were harvested by centrifugation and lysed using a French press. The resulting protein was purified using a single anion exchange chromatographic step, buffer exchanged using centrifugal protein filters, and reconstituted in fresh buffer or deuterated buffer. The purified protein was used directly in studies or stored in buffer containing 10 % glycerol at -80 °C. The protein was estimated to be greater than 90 % pure by SDS P A G E analysis (Figure 3.1) and the molecular mass of NeuB3 was also determined to be 38 670 ± 30 D a using E S I M S (38 647 D a predicted). 1 2 3 Figure 3.1 SDS P A G E gel of NeuB3. Lane A) molecular mass standards of 66 kDa and 29 kDa, B) crude cell lysate and C) anion-exchange purified NeuB3. 110 3.3 Characterization of NeuB3 Activity 3.3.1 Synthesis of 6-deoxy AltdiNAc The putative substrate of NeuB3 was identified by examining the stereochemistry of pseudaminic acid derivatives previously identified and characterized in the literature. 2 5 ' 1 4 4" 1 4 7 These derivatives were found to have the L-glycero-L-manno configuration shown in Figure 3.2. L-glycero refers to the stereochemical configuration of C-8, which is the same as that of L-glyceraldehyde, while L-manno refers to the stereochemical configurations of C-4, C-5, C-6 and C-7, which are the same as that of L -mannose. If NeuB3 is a PEP-condensing pseudaminic acid synthase, then the first three carbons would be derived from PEP , while the last 6 carbons would be derived from the second substrate, 2,4-diacetamido-2,4,6-trideoxy-L-altrose (6-deoxy Al td iNAc) . NHAc CKPO C O O pseudaminic acid \ / H O -N H A c o H 6-deoxy AltdiNAc C H O H-H O -A c H N -H O -NHAc H H H C H 3 6-deoxy AltdiNAc (open chain) C II C H 2 P E P synthase HPO4 C H O H O -C H 2 O H A c H N A c H N pseudaminic acid v c ° ° • |2 H H HO A c H N HO 3 ' 4 • 5 6 7 8 C H 3 L-glyceraldehyde p s e u d a m i n i c a c i d (open chain) C H O - O H H -NHAc H -H -H H H O -H O --OH - O H -H -H C H 2 O H L-mannose (open chain) Figure 3.2 Reaction catalyzed by pseudaminic acid synthase and Fischer projections of 6-deoxy AltdiNAc and pseudaminic acid. I l l A difficulty in studying the activity of NeuB3 is that 6-deoxy A l t d i N A c is not commercially available and had to be synthesized using a twelve step literature-reported synthesis (Figure 3 .3) . 1 4 8 , 1 4 9 Starting from L-fucose, this synthetic scheme inverts the C-2, C-3 and C-4 stereocentres, and introduces the two acetamido groups at C-2 and C-4 by several protection, deprotection and activation steps. L-Fucose was first per-acetylated with sodium acetate and acetic anhydride to yield compound 2. Compound 2 was then converted to its oc-benzyl glycoside 4 , v ia bromination with 45 % hydrobromic acid in acetic acid, yielding 3, and then reaction with silver(I) oxide in benzyl alcohol. Compound 4 was de-acetylated with sodium methoxide in methanol to give compound 5, which was protected as a C-3/C-4 isopropylidene with sulfuric acid and acetone, giving compound 6. C-2 acetylation with acetic anhydride and pyridine, gave compound 7. The C-3 and C-4 hydroxyl groups were then deprotected with acidic resin in water to give compound 8, followed by activation with methanesulfonyl chloride and pyridine to give compound 9. De-acetylation of C-2, formed the C-2/C-3 epoxide 10, which was then substituted with sodium azide to give the C-2, C-4 diazide 11. The azide groups were then reduced using lithium aluminum hydride and per-acetylated using acetic anhydride and pyridine, yielding compound 12. Following de-acetylation and C - l de-benzylation, the final product, 6-deoxy A l t d i N A c , was produced. The reaction steps used to synthesize pseudaminic acid were either quantitative or modestly yielding. Two reactions, however, gave particularly low yields. The azide substitution reaction (forming compound 11) was low yielding, due to the fact that azide substitution occurs at either the C-2 or C-3 position of the epoxide and also produces two C-4 epimeric products. 1 4 9 Compound 11 and its C-4 epimer were not separable by flash 112 chromatography and so were carried forward to the next step. The subsequent azide reduction and acetylation reactions (forming compound 12) also gave low yields. This was partly due to the purification of compound 12 from its C-4 epimer, via flash chromatography, which was particularly challenging and required 3 consecutive columns for purification. In an attempt to increase the reaction yield, the Staudinger reduction was employed in the reduction of the azides. 1 5 0 Reaction yields, however, were identical to that with lithium aluminum hydride. 6-deoxy AltdiNAc Figure 3.3 Synthesis of 6-deoxy AltdiNAc. 113 3.3.2 Test for NeuB3 Activity Following the synthesis of the putative substrate, the activity of NeuB 3 was tested by incubation with 6-deoxy A l t d i N A c and PEP , and monitored using 3 1 P N M R spectroscopy (Figures 3.4). The initial 3 1 P N M R spectrum, before the addition of NeuB3, showed a single phosphorus signal at -0.140 ppm corresponding to the phosphate group of P E P (Figure 3.4A). After incubation with NeuB3, a new signal appeared at 2.870 ppm corresponding to inorganic phosphate (Figure 3.4B). This indicates thatNeuB3 is able to catalyze the formation of phosphate from PEP. PEP and phosphate standards were spiked into the reaction mixture to confirm the identity of the signals. Control reactions lacking enzyme or 6-deoxy A l t d i N A c did not produce any phosphate under similar conditions. Incubations of NeuB3 with M a n N A c and PEP, confirmed previous reports that NeuB3 does not possess sialic acid synthase activity. 2 7 3.0 2.0 1.0 0.0 ppm Figure 3.4 3 1 P NMR spectra monitoring the reaction of PEP and 6-deoxy AltdiNAc with NeuB3. A) Before the addition of NeuB3 and B) after the addition of NeuB3. 114 The enzymatic reaction was also monitored by ' H N M R spectroscopy and showed the conversion of 6-deoxy A l t d i N A c and PEP into pseudaminic acid. The ' H N M R spectral signals, in the region of the spectrum between 1.5 and 2.0 ppm, were particularly informative for observing this conversion (Figure 3.6). The initial ' H N M R spectrum, taken before the addition of NeuB3, showed peaks which corresponded to the acetamido proton peaks for the P-anomer and oc-anomer of 6-deoxy A l t d i N A c at 1.88 and 1.93 ppm, and 1.89 and 1.90 ppm, respectively (Figure 3.5 A ) . This sugar existed as a 3:1 ratio of the p-anomer to a-anomer in solution, which was confirmed by integrating the anomeric proton signals at 5.15 ppm (p-anomeric proton, J\>2 - 1.8 Hz) and 4.92 ppm (a-anomeric proton, J i ; 2 = 2.8 Hz) (not shown). After the addition of NeuB3, the signals corresponding to 6-deoxy A l t d i N A c and P E P were replaced by signals corresponding to pseudaminic acid (Figure 3.5A). Pseudaminic acid existed predominantly as a single anomer under these conditions, which was observable by the two acetamido proton peaks at 1.81 ppm and 1.85 ppm. In this region two other signals, corresponding to the C-3 methylene protons of pseudaminic acid, appeared at 1.77 ppm and 1.62 ppm! The doublet of doublets at 1.77 ppm corresponded to the H-3 equatorial proton (H-3eq) and is characterized by a large Js^^x value (13.3 Hz) and small Ji^,4 value (4.5 Hz). The doublet of doublets at 1.62 ppm corresponded to the H-3 axial proton (H-3ax), yet appears as a triplet due to strong geminal coupling to the H-3eq proton (J3ax,3eq =13.3 Hz) and strong coupling to the H-4 proton (J3ax,4= 12.2 Hz). The large J^A value indicated that H-3ax and H-4 have a trans-diax\a\ relationship and that C-4 has the (S)-configuration. This demonstrated that NeuB3 is catalyzing the C-3 attack of PEP to the si-face of the open chain aldehyde of 6-deoxy A l t d i N A c . 115 It had been previously demonstrated that the absolute stereochemistry of pseudaminic acid isolated from Pseudomonas aeruginosa is of the L-glycero-L-manno configuration. 1 4 4 ' 1 4 5 However, this has not been proven in the case of C. jejuni and reasonable biosynthetic pathways for either enantiomer of the 6-deoxy A l t d i N A c substrate (L- or D-) can be proposed. Since this work shows that the L-enantiomer of 6-deoxy A l t d i N A c is the substrate for NeuB3, the absolute configuration of the pseudaminic acid produced is correct as the L-glycero-L-manno configuration. Figure 3.5 Selective region of the *H NMR spectra monitoring the incubation of PEP and 6-deoxy AltdiNAc with NeuB3. A) Before the addition of NeuB3 and B) after the addition of NeuB3. 116 3.4 Isolation and Characterization of Pseudaminic Acid The synthesis described in section 3.2.1 produced a substrate which was accepted by NeuB3. While the literature synthesis assigned the stereochemistry as that of L-6-deoxy A l t d i N A c , it was conceivable that this assignment was incorrect. 6-deoxy A l t d i N A c not only exists as two anomers in solution, but each anomer reportedly exists as two ring conformations ( ' C 4 and 4 C i ) (Figure 3.6). 1 4 9 The rapid ring flipping complicated the stereochemical assignments of the chiral centers in 6-deoxy A l t d i N A c . In order to further confirm that the pseudaminic acid produced was of the L-glycero-L-manno configuration, the product from the enzymatic incubations was isolated and characterized. NHAc HO~k^-^X^-OH= AcHN a-6-deoxy AltdiNAc NHAc OH OH J A c H N - ^ ^ J U 2 V NHAc a-6-deoxy AltdiNAc 1 C 4 AcHN J,H P-6-deoxy AltdiNAc AcHN OH NHAc p-6-deoxy AltdiNAc '4 1c,Figure 3.6 Ring conformations of 6-deoxy AltdiNAc. Enzymatically produced pseudaminic acid was isolated by first removing NeuB3 using centrifugal protein filters and then by passage through a single anion-exchange chromatographic column. The isolated pseudaminic acid was then characterized using ESI mass spectrometry and N M R spectroscopy (Figures 3.8 and 3.9, and Tables 3.1 and 117 3.2). ESI mass spectral analysis of pseudaminic acid showed two major parental ion masses of m/z 333 D a ( M - F f ) and m/z 315 D a ( M - H 3 0 + ) . The ' H N M R spectral signals in the ' H N M R spectrum of pseudaminic acid were well separated, except for three signals in the region between 3.95 and 4.08 ppm (Figure 3.8). Analysis by 2D C O S Y N M R spectroscopy (correlated spectroscopy) indicated that these proton signals corresponded to the H-4, H-7 and H-8 protons (not shown). To determine the coupling constants of these protons I D T O C S Y N M R spectroscopy (total correlated spectroscopy) was employed. T O C S Y N M R spectroscopy is used to determine the proton signals in a given system through spin-spin coupling. 1 5 1 B y selectively irradiating a proton signal, the magnetization is transferred through the bonds to the adjacent protons in the molecule. Since this is a through bond process, the protons nearest to the irradiated proton have larger signal intensities than protons which are farther away. B y selectively irradiating the H-9 signal of pseudaminic acid at 0.99 ppm, only the H-8 and H-7 signals appeared in the T O C S Y N M R spectrum (Figure 3.7A). The H-8 signal is a doublet of quartets centered at 4.01 ppm, while the H-7 signal is a doublet of doublets centered at 4.03 ppm. While these two signals, do overlap, the H-8 proton signal was sufficiently resolved to obtain coupling constant values. The H-3ax proton was also irradiated at 1.67 ppm to obtain coupling constants for the H-4 proton (Figure 3.7B). This T O C S Y N M R spectrum showed the H-4 signal, which appeared as a doublet of doublets of doublets centered at 4.05 ppm. The H-4 signal was well resolved with the H-7 and H-8 proton signals absent. 118 H-8 H-7 t 1 , . , , 1 , . , . 1 , , , . 1 . , , , 1 r 4.150 4.100 4.050 4.000 3.950 ppm Figure 3.7 ID TOCSY NMR spectra of pseudaminic acid. A) TOCSY N M R spectrum irradiated at 0.99 ppm (H-9 proton signal) and B) TOCSY N M R spectrum irradiated at 1.67 ppm (H-3ax proton signal). 4.50 4.00 3.50 3 00 2.50 2.00 1.50 1.00 Figure 3.8 400 MHz NMR spectrum of pseudaminic acid in D 20. 120 Figure 3.9 400 MHz C NMR spectrum of pseudaminic acid in D 20. 121 Table 3.1 'rl NMR spectral assignments of pseudaminic acid. Proton Signal 5 (ppm) muliplicity coupling (integration) constants (Hz) H-3eq (1 H ) . 1.81 dd JltqA ~ 4.6 H-3ax (1 H) 1.67 dd •/3ax,3eq 13.3 H-4( l H) 4.05 ddd Jzsx.A 12.2 H-5 (1 H) 4.13 dd =5.5 H-6 (1 H) 3.91 dd A e = 1 . 7 H-7 (1 H) 4.03 dd J 6 , 7 = 10.3 H-8 (1 H) 4.01 dq A s = 3.7 H-9 (1 H) 0.99 d Jgfi- 6.5 N - A c (3 H) 1.86 s -( 3 H ) 1.89 s -Table 3.2 C NMR spectral assignments of pseudaminic acid. Carbon Signal 5 (ppm) C-2 96.4 C-3 34.8 C-4 65.2 C-5 48.8 C-6 70.0 C-7 52.9 C-8 66.8 C-9 15.3 C H 3 - N A c 21.8 22.0 c = o 173.7 174.6 176.3 The N M R spectral signals of enzymatically produced pseudaminic acid were similar to those obtained from previously characterized derivatives of pseudaminic acid, isolated from the lipopolysaccharride of P. aeruginosa, Shigella boydii, Vibrio cholerae, Proteus vulgaris and the flagellin proteins of C. jejuni25'W'U1 B y examining the ' H N M R coupling constants, the relative configuration of the ring protons was assigned (Table 3.1). A s previously stated, the large J^XA value indicates that H-4 is axial. The smaller ^4,5 (4.5 122 Hz) and Js,6 (1.2 Hz) values indicate that H-5 is equatorial and H-6 is axial. The large Je,i of 10.3 Hz indicates that C-6 and C-7 have the erythro configuration, in contrast to N e u N A c where C-6 and C-7 have the threo configuration and a small Jej of 1.2 H z (Figure 3.10). 1 4 5 In addition, the difference in the ' H N M R chemical shifts of the H-3ax and H-3eq proton signals is small (0.14 ppm) and indicates that pseudaminic acid exists predominantly as the a-anomer in solution. In C-2 glycosylated pseudaminic acid, which is P-linked, the difference in chemical shifts is larger (0.77 ppm) . 2 5 ' 1 4 4 " 1 4 7 It was previously reported that the pseudaminic acid isolated from P. aeruginosa had the erythro configuration at C-7 and C - 8 . 1 4 5 This assignment was made by comparing the C-9 1 3 C N M R chemical shift of the pseudaminic acid isolated, to the C-4 chemical shifts of iV-acetyl L-tf//o-threonine, which has the erythro configuration, and /V-acetyl L-threonine, which has the threo configuration (Figure 3.10). Since the C-9 chemical shift of pseudaminic acid was closer to the C-4 chemical shift of /V-acetyl L-«//o-threonine (17.5 ppm) than that of N-acetyl L-threonine (19.5 ppm), it was concluded that the C-7 and C-8 carbons had the erythro configuration. The C-9 1 3 C N M R chemical shift for the pseudaminic acid produced in this experiment was 15.3 ppm. Since this chemical shift closely resembles that of the one previously reported, we also conclude that the C-7 and C-8 carbons have the erythro configuration. The N M R assignments obtained are in close agreement with those previously reported and indicate that the pseudaminic acid, produced from NeuB3, also possesses the L-glycero-L-manno configuration. These results also confirm that 6-deoxy A l t d i N A c was synthesized with the proper stereochemistry. 123 H 8 OH H3(9)C O. COO V 12 AcHN H4 COO H3eq AcHN H3ax pseudaminic acid H H HO AcHN HO C H ? 3 4 5 6 7 8 -OH -NHAc -H -H -H CHo pseudaminic acid (open chain) CHO HO-HO--H -H CH2OH L-erythrose CHO H-HO--OH -H CH2OH L-threose COO AcHN-HO--H -H CH, N-acetyl L-a//o-threonine COO AcHN-H--H -OH CHq A/-acetyl L-threonine Figure 3.10 Structure of pseudaminic acid and compounds used in the assignment of its stereochemistry. 124 3.5 Kinetic Characterization of NeuB3 Since this is the first discovery of a pseudaminic acid synthase, the activity of NeuB3 was kinetically characterized using a continuously, coupled phosphate assay (Figure 3 . I I ) . 1 5 2 The release of phosphate from the NeuB3 reaction is coupled to the enzymatic reaction of purine nucleoside phosphorylase (PNP), which uses phosphate to displace 2-amino-6-mercapto-7-methylpurine (thio-guanine) from the substrate analog 2-amino-6-mercapto-7-methylpurine ribonucleoside ( M E S G ) and forms ribose 1-phosphate: Although both the substrate and product of P N P have overlapping absorptions, the difference in their molar absorbtivities at 360 nm is 11 000 M " 1 cm" 1, which is sufficiently large to allow for the detection of the thio-guanine product. pseudaminic acid synthase 6-deoxy Al td iNAc — ^ — ^ * pseudaminic A c i d P E P HPO4 OH M E S G ribose 1-P thio-guanine Figure 3.11 Coupled phosphate assay used in the kinetic analysis of NeuB3. Initial reaction velocities (v) were measured for each substrate by varying the concentration of one substrate in the presence of saturating amounts of the other substrate (1 m M for each substrate). The kinetic data were plotted using Michaelis-Menten kinetics, giving kinetic constants of &Cat- 0.65 ± 0.01 s"1, A ^ P E P of 6.5 ± 0.4 | i M and /Cm6-deoxy A l t d i N A c of 9.5 ± 0.7 pJVi (Figure 3.12). The catalytic rate of NeuB3 is comparable to that observed for N e u N A c synthase from N. meningitidis (A;cat - 0.9 s"1), while the Km 125 values were found to be significantly lower (Km?E? = 0.25 m M and ZJVIanNAc =- 9.4 m M ) . 7 2 The Km values observed for NeuB3 may reflect a lower concentration of 6-deoxy A l t d i N A c available in the cell. [6-deoxy AltdiNAc] (|iM) Figure 3.12 Kinetic plots of initial velocity vs. substrate concentration and reciprocal plots. A) 1 m M 6-deoxy AltdiNAc and varying concentrations of PEP and B) 1 m M PEP and varying concentrations of 6-deoxy AltdiNAc. The kinetic parameters were measured in 100 m M Tris-HCl buffer and 10 mM M n C l 2 . Each kinetic point was performed in duplicate. In addition to measuring the kinetic parameters of NeuB3, a p H versus rate profile was constructed (Figure 3.13). The maximal rates were measured with saturating concentrations of substrates (1 m M each) at various p H values in a buffer composed of 50 m M Tr is -HCl and 50 m M sodium M E S . The p H versus rate profile indicates that 126 optimal catalysis for NeuB3 is observed at pH 7.0. The bell shaped profile also implicates an acid and a base in catalysis. pH vs. rate profile pH Figure 3.13 pH vs. rate profile of NeuB3. Since the N e u N A c synthases are all metalloenzymes requiring a divalent metal co-factor for catalysis, 2 7 ' 7 2 ' 7 3 the metal ion requirements of NeuB3 were also tested. The maximal rates of NeuB3 were measured with the addition of various divalent metal ions, without the addition of metals and with the addition of E D T A (Table 3.3). Low rates were obtained with the enzyme as isolated (no metals added) and the addition of E D T A abolished enzymatic activity. The addition of divalent metal ions restored the activity of the enzyme with maximal rates obtained in the presence of C o 2 + and M n 2 + . This demonstrates that NeuB3 requires a divalent metal ion for catalysis. 127 Additive (10 m M ) % Activity Co2+ 100 M n 2 + 91 M g 2 + 48 C a 2 + 30 N i 2 + 26 enzyme as isolated 4 E D T A 0 Table 3.3 Metal dependence experiment with NeuB3. Rates were normalized to the fastest rate (Co ). Each metal was assayed in duplicate. 128 3.6 Test for C-O versus P-O Bond Cleavage A s mentioned in Chapter 1 (pp. 23-25), the incubation of [2 - l s O]PEP with PEP-condensing synthases was used to differentiate between a C - 0 bond cleavage mechanism and a P-O bond cleavage mechanism. 7 2 , 7 7 ' 7 8 Potentially, NeuB3 catalysis could proceed through a C-O or P-O bond cleavage mechanism and incubation with [2- OJPEP could also be used to probe the operative mechanism (Figure 3.14). The C-O bond cleavage mechanism starts with the C-3 attack of PEP to the open chain aldehyde of 6-deoxy A l t d i N A c , forming an oxocarbenium ion intermediate (Figure 3.14A). Water then attacks the C-2 position, forming a tetrahedral intermediate. The tetrahedral intermediate collapses, releasing phosphate and forming the open chain form of pseudaminic acid, which cyclizes to the pyranose form in solution. The P-O bond cleavage mechanism begins with the attack of water to the phosphate group of PEP, releasing phosphate and forming the enolate anion of pyruvate (Figure 3.14B). The enolate then attacks the open chain aldehyde of 6-deoxy A l t d i N A c , forming the open chain form of pseudaminic acid, which cyclizes to the pyranose form in solution. If NeuB3 operates via a C - 0 bond cleavage mechanism, then its incubation with [2 - 1 8 0]PEP would lead to the formation of 1 8 0-labeled phosphate. If NeuB3 operates via a P-O bond cleavage mechanism then its incubation with [2 - 1 8 0]PEP would lead to C - l 1 8 0-labeled pseudaminic acid. 129 18 • - P D H A 0 2 C . J*-P03= - 1 8 + - - n 1 8 3 ^ H 2 0 \ . . „ O H V - O H Y - O H R' M 2 + R R R R: 6-deoxy AltdiNAc oxocarbenium ion tetrahedral pseudaminic acid H k j « V (open chain) + P E P intermediate intermediate (open chain) ~ \ _ O H AcHN"-<T^ ~o2c '•%o3= "o2c -o 2c. 1 - H O H ^ " O H 2 Hf P ° 4 H -R X R ^ 'OH R 6-deoxy AltdiNAc aldose (open chain) pseudaminic acid (open chain) + P E P + enolate (open chain) Figure 3.14 Proposed C-O vs. P-O bond cleavage mechanisms for NeuB3. A) C-O bond cleavage mechanism and B) P-O bond cleavage mechanism. 18 [2- 0 ] P E P was synthesized according to literature described procedures (Figure 3.15) . ' 5 3 ' 1 5 4 The 1 8 0 label was incorporated by stirring compound 15 in a mixture of l s O -enriched water, methanol and acetic acid. Treatment of l 8 0-labeled 15 with trimethoxy phosphine gave the protected form of phosphenolpyruvate 16. A t this point the 1 8 0 label was rendered non-exchangeable with aqueous solvent. Deprotection and de-esterification of compound 16 with trimethyl si lyl bromide and potassium hydroxide, followed by 18 cation exchange, produced [2- OJPEP. Incorporation of the label was determined to be 54 % using ESI mass spectral analysis. The position of the label was determined by 3 1 P N M R spectroscopy, since the substitution of 1 6 0 for 1 8 0 in a singly bonded position to phosphorus results in a small upfield shift for the labeled phosphorus atom. 1 5 5 130 o EtO ~Br H 18 0 EtO O ethyl-3-bromopyruvate 14 C H 3 C O O H MeOH 80% O Br p—(-OMe) 3 EtO" Dioxane 45% Bo. MeO 16 X .OMe O 1) BrTMS, CH 2 CI 2 N a ° " 2) KOH 3) Dowex (Na+ form) 28% O Na Na O u 18, [2-laO]PEP Figure 3.15 Synthesis of [2-l!tO]PEP. The mechanism of NeuB3 was tested by incubating 6-deoxy A l t d i N A c and [2-1 8 0 ] P E P with NeuB 3 and monitoring the reaction by 3 1 P N M R spectroscopy. The initial 3 1 P N M R spectrum of 54% labeled [2 - 1 8 0]PEP showed two phosphorus signals (Figure 3.16A). The signal at -0.140 ppm corresponds to unlabeled P E P and the signal at -0.159 ppm corresponds to 1 8 0-labeled P E P : After the addition of NeuB3, two new phosphorus signals appeared, corresponding to unlabeled phosphate at 2.870 ppm and 1 8 0-labeled phosphate at 2.848 ppm (Figure 3.16B). The ratio of 1 6 0 to 1 8 0 was the same as in [2-1 8 0 ] P E P and indicates that the 1 8 0 label was fully retained in the phosphate produced. This result demonstrates that NeuB3 is utilizing a C-O bond cleavage mechanism during catalysis. 131 180-PEP-1 60-PEP-H ^ W f r ^ * l f r i ' » i | | » ^ f i M ^ i | i U M « « K ^ ^ ****** 16O-PO,2-J L i - i 1 1 1 1 1 r — | — i — i — i — i — | 1 — i — i — i — | — i 1 1 — i — | 1 1—i—i—|— i—i— i—i—f 2.50 2.00 1.50 1.00 0.50 0.00 ppm Figure 3.16 3 1 P NMR spectra monitoring the incubation of NeuB3 with [2-1 80|PEP and 6-deoxyAltdiNAc. A) Before the addition of NeuB3 and B) after the addition of NeuB3. 132 3.7 Conclusions These are the first experiments which identify and characterize the activity of NeuB 3 from C. jejuni N C T C 11168 as a pseudaminic acid synthase, catalyzing the condensation of P E P with 6-deoxy A l t d i N A c , forming phosphate and pseudaminic acid. The acceptance of the synthetically produced 6-deoxy A l t d i N A c as a substrate by NeuB3, indicates that the stereochemistry of the pseudaminic acid produced is the L-glycero-L-manno configuration. This assignment was also verified by the isolation and characterization of the enzymatically produced pseudaminic acid. The kinetic characterization of NeuB3 provided kinetic constants and indicated that this enzyme operates at an optimal p H of 7.0. A metal dependence study also revealed that NeuB3 is a metalloenzyme, requiring a divalent metal ion for catalysis. The role of the divalent metal ion may be in the electrophilic activation of the aldehyde of open chain 6-deoxy A l t d i N A c during catalysis. This is also proposed for N e u N A c synthase from A7, meningitidis, where an X-ray crystal structure of N e u N A c synthase, co-crystallized with PEP , M n 2 + and N-acetyl mannosaminitol (open chain M a n N A c reduced at the C - l position) was recently solved. 7 2 The crystal structure showed that M n 2 + was 2.5 A away from the C - l hydroxyl group, indicating that the metal ion is at a favorable distance for the electrophilic activation of the aldehyde of M a n N A c during catalysis. The incubation of [2 - 1 8 0]PEP with NeuB3 resulted in the formation of 1 8 Q -labeled phosphate, indicating that NeuB3 utilizes a C - O bond cleavage mechanism. The P E P condensing synthases, K D 0 8 P synthase, D A H 7 P synthase and N e u N A c synthase all utilize a C - O bond, cleavage mechanism, which are proposed to proceed through oxocarbenium ion and tetrahedral intermediates. In light of these results, the proposed 133 mechanism of NeuB3 is the C-3 attack of PEP forming an oxocarbenium ion intermediate, followed by the C-2 attack of water forming a tetrahedral intermediate (Figure 3.14A). The tetrahedral intermediate then collapses, releasing phosphate and forming the open chain form of pseudaminic acid, which cyclizes to the pyranose form in solution. 134 3.8 Future Directions and Related Studies Several derivatives of pseudaminic acid with differing JV-acyl substituents have also been isolated and identified in various bacter ia . 2 5 ' 2 6 ' 1 4 4 " 1 4 7 ' 1 5 6 These include N-formamido /V-acetamidino and yV-hydroxypropionyl derivatives. In a recent study, N-propionyl, /V-butanoyl and /V-pentanoyl substituted M a n N A c derivatives were tolerated as substrates for N e u N A c synthase from C. jejuni86 The TV-substituted derivatives of pseudaminic acid are in all likelihood produced from NeuB3 from their corresponding N-acylated derivatives of 6-deoxy A l t d i N A c . Other a-keto acids have also been isolated from P. aeruginosa and Legionella pneumophilia.i57A60 A few examples are legionaminic acid (5,7-diacetamido-8-0-acetyl-3,5,7,9-tetradeoxy-D-g/ycero-D-go/acto-non-2-ulosonic acid), a C-4 epimeric derivative (5,7-diacetamido-8-<9-acetyl-3,5,7,9-tetradeoxy-D-g/ycero-D-to/o-non-2-ulosonic acid) and a C-8 epimeric derivative of legionaminic acid (5,7-diacetamido-3,5,7,9-tetradeoxy-L-g/ycero-D-ga/acto-non-2-ulosonic acid) (Figure 3.17). Biosynthetic pathways with PEP-condensing synthases are potentially involved for the formation of these a-keto acids. OAc OH OAc OH legionaminic acid C-4 epimer OH OH AcHN OH C-8 epimer Figure 3.17 Legionaminic acid and two epimeric derivatives. 135 . Recent work on the pseudaminic acid biosynthetic pathway in C. jejuni has led to the identification and characterization of three more enzymes in the pathway (Figure 3.18). The first enzyme in the pathway, PseB, has been identified and characterized as a U D P - G l c N A c C-6 dehydratase, which catalyzes an oxidation at C-4, dehydration at C-6 and an inversion of stereochemistry at C-5 , producing UDP-2-acetamido-2,6-dideoxy-P-L-arabino-4-hexulose (UDP-6-deoxy-4-keto-L-HexNAc). 1 6 1 The inversion of stereochemistry at C-5 produces an L-sugar required for the following enzymes in the pathway. The next enzyme in the pathway, PseC, has also been identified and characterized to be a PLP-dependent aminotransferase, catalyzing the transfer of an amino group to the C-4 position, forming UDP-4-amino-2,4,6-trideoxy-pV A l t N A c ( U D P -4-amino-6-deoxy A l t N A c ) . 1 6 1 The fourth enzyme in the pathway, PseG, is currently being studied in the Tanner lab by a graduate student, Feng L i u and was demonstrated to be an inverting UDP-6-deoxy A l t d i N A c hydrolase, catalyzing the hydrolysis of the glycosidic bond and forming a-6-deoxy A l t d i N A c and U D P . 1 6 2 The substrate for this enzyme was obtained chemo-enzymatically, by incubating U D P - G l c N A c with the first two enzymes in the pathway, followed by a chemical A r-acetylation step. Although studies demonstrating the activities proposed for the acetyl transferase and C M P - P S E synthetase enzymes in the biosynthetic pathway of pseudaminic acid in C. jejuni are lacking, a recent study of the genes in C. jejuni strain 81-176 implicated 3 additional genes, involved in the biosynthetic pathway of pseudaminic acid, named pseE, pseF and pseHm Mutation of these three genes all resulted in aflagellate bacteria, implicating their importance in the formation of flagella. The middle gene, pseF is proposed to encode a C M P - P S E synthetase, based on sequence homology to the E. coli 136 C M P - N e u N A c synthetase. Since a chemo-enzymatic route for the synthesis of pseudaminic is now available, it wi l l be possible to study the activity of PseF to determine i f this enzyme is in fact a C M P - P S E synthetase. The identities of the genes pseE and pseH are unknown, although these genes could potentially encode the unidentified acetyltransferase or a C M P - P S E transferase. Since enzymatic routes to their potential substrates are now available, it wi l l also be possible to study the activities of their respective gene products. PseB dehydratase H ° — \ O H O ~ \ - ^ - - , , T , » « \ H N A c O - U D P H'2° H W)-UDP T PseC aminotransferase NH, n m o i r a i i s T e r a s e i O L-GuTot-ketoglu H O - \ ^ - - * , ^ » \ UDP-GlcNAc UDP-6-deoxy-4-keto-L-HexNAc H * C H N A c O - U D P UDP-4-amino-6-deoxy AltNAc NeuB3 (Psel) O H pseudaminic acid "NAc _ synthase co2 ' f ^ Pi PEP HO PseG HNAc hydrolase acyl transferase (unidentified) O H UDP H 2 0 H O HNAc Pseudaminic acid (PSE) CMP-PSE synthetase H 3 C HNAc a-6-deoxy AltdiNAc H 3 C HNAcI O-UDP UDP-6-deoxy AltdiNAc HO HNAc CMP-PSE Figure 3.18 Revised biosynthetic pathway of pseudaminic acid. 137 3.9 Experimental Procedures 3.9.1 Materials and General Methods The previous materials and general methods from Chapter 2 also apply in this study with the following additions. Purine nucleoside phosphorylase and phosphoenolpyruvate were purchased from Sigma-Aldrich. 2-amino-6-mercapto-7-methylpurine ribonucleoside ( M E S G ) was purchased from Berry and Associates. 3.9.2 Over-expression of neuB3 and Purification of NeuB3 E. coli AD202 cells, pre-transformed with the recombinant plasmid encoding neuB3 (both donated from the lab of Dr. Warren W. Wakarchuk), were grown at 37 °C in 500 ml of L B medium supplemented with 100 ug m l - 1 of ampicillin. Overexpression of NeuB3 was induced through the addition of 0.5 m M isopropyl (3-D-thiogalactopyranoside at an OD^o of 0.60, with growth at 37 °C for 6 hours. Cells were harvested via centrifugation at 5000 x g for 30 minutes, resuspended in Tr i s -HCl buffer (20 m M , pH 7.0) containing pepstatin A (1 mg/L) and aprotonin (1 mg/L), and lysed with two passes through an ice-chilled French pressure cell. The resulting cell lysate was centrifuged at 7,000 x g for 1.5 hours, passed through a syringe filter, and loaded directly onto a 5 ml High-Trap™ Q-Sepharose H P column (Amersham Biosciences) pre-equilibrated with 20 m M Tr i s -HCl buffer (pH 7.0). NeuB3 was eluted using a linear gradient of 0 M to 1 M N a C l in 20 m M Tr i s -HCl buffer (pH 7.0). Fractions containing active enzyme were desalted 2 times by concentration at 5000 x g for 30 min and reconstitution with 20 m M Tr i s -HCl buffer (pH 7.0) using protein centricons, and then flash frozen with 10% 138 glycerol followed by storage at -80 °C. Protein samples were determined to be >90% pure by S D S - P A G E and were used directly in all assays. The molecular mass of NeuB3 was determined to be 38 670 D a using -ve E S I M S (38 647 D a predicted). Deuterated NeuB3 was prepared by taking column purified protein fractions and desalting 2 times and reconstituting in deuterated buffer (20 m M Tr i s -DCl buffer, p D 7.4). 3.9.3 N M R Incubation Studies of NeuB3 2,4-Diacetamido-2,4,6-trideoxy-L-<3^ras ,e (6-deoxy Al td iNAc) was synthesized according to literature procedures. 1 4 8 , 1 4 9 A solution containing 6-deoxy A l t d i N A c (10 m M ) and PEP (20 m M ) in deuterated buffer (700 U.L, 10 m M , pD 7.4) was placed in an N M R tube. Initial ' H and proton-decoupled 3 1 P N M R spectra were taken. The solution was removed from the tube and mixed with 50 mg of NeuB3 and 1 m M MgCL. in the same deuterated buffer (1 m L total volume). After incubation of the reaction mixture for 5 min at 25 °C, Chelex-100 resin (-20 mg) was added, incubated for 1 hour at room temperature and ' H and 3 1 P N M R spectra were retaken. 3.9.4 Isolation and Characterization of Pseudaminic Acid NeuB3 was removed from the enzymatic reactions by centrifugal concentration (5000 x g for 30 min) and the resulting filtrate was loaded onto a 10 ml column of D o w e x - A G l X 8 resin (formate form, 100-200 mesh, Bio-Rad) pre-equilibrated with water. A stepwise gradient of 0 - 1.0 M formic acid in water with 0.2 M increments (50 m L per increment) was used to elute the pseudaminic acid from the column. Pseudaminic acid eluted from the column in the 0.2 M and 0.4 M fractions which were 139 concentrated in vacuo and then lyophilized. Pseudaminic acid was characterized using ] H , 1 3 C , 2-D C O S Y , H M Q C and 1-D T O C S Y N M R spectroscopy (in 10 m M deuterated phosphate buffer pD 7.4), and -ve E S I - M S mass spectrometry. ' H ( D 2 0 ) 8 0.99 (d, 3 H , J 8 , 9 6 .5 Hz, H-9), 1.95 (dd, 1 H , J3axA 12.2 Hz, y 3 a X ; 3 e q 13.3 Hz , H-3ax), 1.99 (dd, 1 H , Jseqaax 13.3, ^ 4 . 6 Hz , H-3eq), 1.86 (s, 3 H , CH3) , 1.89 (s, 3 H , CH3) , 3.91 (dd, 1 H , J5,6 1.7 Hz , J 6 ) 7 10.3 Hz , H-6), 4.01 (dq, 1 H , / 7 , 8 3.7 Hz , J 8 , 9 6.5 Hz, H-8), 4.03 (dd, 1 H , J 6 , 7 10.3 , y 7,s3.7, H-7), 4.05 (ddd, 1 H , J3eqA 4.6, J 3ax,4 12.2, JA>5 5.5, H-4), 4.13 (dd, 1 H , J 4 ) 5 5.5, J5fi 1-7, H-5) . 1 3 C N M R ( D 2 0 ) 5 15.3 (C-9), 21.8 ( C H 3 - N A c ) , 22.0 ( C H 3 - N A c ) , 34.8 (C-3), 48.8 (C-5), 52.9 (C-7), 65.2 (C-4), 66.8 (C-8), 69.9 (C-6), 96.4 (C-2), 173.7, 174.6, 176.3 (C=0). -ve ESI -MS ( H 2 0 ) m/z 333 ( M - H + ) , m/z 315 ( M - H ^ 4 ) . 3.9.5 Kinetic Characterization of NeuB3 Enzyme kinetics were measured by a continuous coupled phosphate assay. ~ A cuvette containing 100 m M Tr i s -HCl buffer (pH 7.0), 6-deoxy A l t d i N A c (up to 1 m M ) , P E P (up to 1 m M ) , M E S G (200 u M ) , purine nucleoside phosphorylase (5 units buffer exchanged twice into 100 m M Tr i s -HCl buffer, pH 7.0), and M n C l 2 (10 m M ) was thermally equilibrated for 5 min at 37 °C. The enzymatic reaction was initiated by the addition of NeuB3 (3.2 ug) for a total assay volume of 500 p X and the enzymatic rate was calculated from the observed increase in absorption at Aieo (using Ae = 11 000 M " 1 cm"1). Enzymatic rates for 6-deoxy A l t d i N A c were measured in the presence of 1 m M PEP (saturating), and those for P E P were measured in the presence of 1 m M 6-deoxy A l t d i N A c (saturating). The resulting calculated enzyme rate was plotted as a function of substrate concentration and the kinetic parameters were determine by a direct fit of the 140 139 data to the Michealis-Menten equation using the computer program GraFit. This program performs a non-linear regression analysis on the data following the method of Marquart, 1 4 0 and reports an error for the data based on the deviation of the data from the calculated curve-of-best-fit. With different enzyme preparations, the error between different data sets did not vary by more than 10 %. Concentrations of stock substrate solutions were determined enzymatically using the assay conditions described with high concentrations of NeuB3 (64 p.g) and excess co-substrate. The p H vs. rate profile was constructed using a mixture of 50 m M Tr i s -HCl and 50 m M sodium M E S buffer at p H 6-9. Purine nucleoside phosphorylase was buffer exchanged twice in the same buffer at the different p H values. Saturating 6-deoxy A l t d i N A c (1 m M ) and saturating P E P (1 m M ) were used with the kinetic assay above. Initial rates were plotted against p H . The metal dependence experiment was carried out with saturating 6-deoxy A l t d i N A c (1 m M ) and saturating PEP (1 m M ) at p H 7.0 using the kinetic assay above. Different divalent metal cations were independently added at a concentration of 10 m M and the initial velocities were determined. The rates were normalized to the fastest rate (Co 2 + ) . Controls included adding 10 m M E D T A or no additives (enzyme as isolated) to the kinetic assay. 3.9.6 C-O versus P-O Bond Cleavage [2 - 1 8 0]PEP disodium salt was prepared according to literature reported procedures. 1 5 3 ' 1 5 4 A solution containing 10 m M 6-deoxy A l t d i N A c and 20 m M [2-1 8 0 ] P E P prepared in deuterated buffer (700 uL, 10 m M , pD 7.4) was placed in an N M R 141 tube, and Chelex-100 resin (-20 mg previously washed with D2O) was added. A n initial proton-decoupled 3 1 P N M R spectrum was obtained with the following parameters: spectral frequency of 121.5 M H z , sweep width of 2437 Hz , acquisition time of 13.4 s, pulse delay of 2 sec, and pulse width of 10 [is. The solution was decanted from the Chelex resin and mixed with 50 mg of NeuB3 and I m M M g C h in deuterated buffer (total volume of 1 mL) . The reaction was incubated for 5 min at 25 °C and Chelex-100 resin (-20 mg previously washed with D2O) was added. After a 1 h time period to allow complete complexation of the metals, another proton-decoupled 3 1 P N M R spectrum was acquired with the same parameters: 3 1 P N M R 8 2.870 (s, P i - 1 6 0 ) , 2.848 (s, P i - l s O ) , -0.140 (s, P - 1 6 0 , PEP), -0.159 (s, P - 1 8 0 , PEP). 3.9.7 Synthesis of 2,4-diacetamido-2,4,6-trideoxy-L-altrose (6-deoxy AltdiNAc) 6-Deoxy A l t d i N A c was synthesized according to literature described procedures. 1 4 8 ' 1 4 9 A variation in this procedure was in the reduction reaction of the diazide 11, forming compound 12 (Figure 3.3). In addition to using literature procedures, a Staudinger reduction was employed. To a solution of 700 mg (2.3 mmol) of compound 11 dissolved in 31.5 m L of T H E , was added 20.7 m L (10.4 mmol) of a 1 M solution of P M e 3 in T H F and 0.5 m L of 0.1 M N a O H . The reaction mixture was stirred at room temperature for 2 hours and concentrated in vacuo. Following concentration, 50 m L of pyridine and 3.0 m L (31.7 mmol) of acetic anhydride were added to the resulting syrup. This reaction mixture was stirred overnight, concentrated in vacuo and chromatographed on silica gel in a 19:1 solvent mixture of C H C L v M e O H , affording compound 12 in 25 % 1 8 yield. H N M R spectra matched those in the literature. 142 The ' r l N M R spectral assignments for the final product, 6-deoxy A l t d i N A c are below. ' H N M R ( D 2 0 ) : p-anomer 5 5.15 (d, 1H, y 1 > 2 = 1.8 Hz , p-H-1), 3.96 (dd, 1H, y 1 > 2 = 1.8 Hz, J 2 , 3 '= 2.2 Hz, p-H-2), 3.805 (dq, 1H, J4,5 = 10.4 Hz , J5y6 = 6.2 Hz, p-H-5), 3.79 (dd, 1H, J2,3 = 2.2 Hz, J3,4 = 2.8 Hz,p-H-3), 3.68 (dd, 1H, J3A = 2.8 Hz , J4,s = 10.4 Hz , p-H-4), 1.93 (s, 3H , p-NAc) , 1.88 (s, 3H , p-NAc) , 1.09 (d, 3H , J5,6 = 6.2 Hz, p-H-6): a-anomer 5 4.92 (d, 1H, J 1 ) 2 = 2.8 Hz, a-H-1), 4.12 (dq, 1H, J4,5 = 8.6 Hz, J5>6 = 6.8 Hz, a-H-5), 3.93 (dd, 1H, Jl>2 - 2.8 Hz , J2,3 = 2.9 Hz, a-H-2), 3.82 (m, a-H-4), 3.79 (m, a-H-3), 1.90 (s, 3H, a -NAc) , 1.89 (s, 3H, a - N A c ) , 1.12 (d, 3H, J5,6 = 6.8 Hz , a-H-6). +ve E S I M S m/z 269 ( M + Na + ) . 3.9.8 Synthesis of [2- 1 80]PEP [2 - 1 8 0]PEP disodium salt was prepared according to literature reported procedures. 1 5 3 ' 1 5 4 The extent of 1 8 0 incorporation at the C-2 position was determined to be 54% by mass spectrometry: -ve E S I - M S (MeOH) m/z 169 ( M - H + , 1 8 0 , 100), 167 ( M - H + , 1 6 0 , 85). 3 1 P N M R spectroscopy was used to determine that the position of the label was at the C-2 position: 3 1 P N M R 8 -0.140 (s, P - 1 6 0 , PEP), -0.159 (s, P - 1 8 0 , PEP). *H N M R ( D 2 0 ) : 5 5.30 (s, 1H, C H 2 proton), 5.15 (s, 1H, C H 2 proton). 143 Bibliography (1) Angata, T.; Vark i , A . Chem. Rev. 2002,102, 439-69. (2) Varki , A . FASEB J. \991,11, 248-55. (3) Traving, C ; Schauer, R. Cell. Mol. Life Sci. 1998, 54, 1330-49. (4) Kleene, R.; Schachner, M . Nat. Rev. Neurosci. 2004, 5, 195-208. (5) Muller, D . ; Stoppini, L . ; Wang, C ; Kiss, J. Z. Neuroscience 1994, 61, 441-5. (6) Yamamoto, N . ; Inui, K . ; Matsuyama, Y . ; Harada, A . ; Hanamura, K . ; Murakami, F.; Ruthazer, E . S.; Rutishauser, U . ; Seki, T. J. Neurosci. 2000, 20, 9145-51. (7) Weiss, P.; Tietze, F.; Gahl, W. A . ; Seppala, R.; Ashwell , G . J. Biol. Chem. 1989, 264, 17635-6. (8) Eisenberg, I.; Avidan, N . ; Potikha, T.; Hochner, H . ; Chen, M . ; Olender, T.; Barash, M . ; Shemesh, M . ; Sadeh, M . ; Grabov-Nardini, G . ; Shmilevich, I.; Friedmann, A . ; Karpati, G . ; Bradley, W. G . ; Baumbach, L . ; Lancet, D . ; Asher, E . B . ; Beckmann, J. S.; Argov, Z. ; Mitrani-Rosenbaum, S. Nat. Genet. 2001, 29, 83-7. (9) Yogeeswaran, G . ; Salk, P. L . Science 1981, 212, 1514-6. (10) Preston, A . ; Mandrell, R. E . ; Gibson, B . W. ; Apicella, M . A . Crit. Rev. Microbiol. 1996,22,139-180. (11) Edwards, U . ; Muller, A . ; Hammerschmidt, S.; Gerardy-Schahn, R.; Frosch, M . Mol. Microbiol. 1994, 14, 141-9. (12) Ganguli, S.; Zapata, G . ; Wallis, T.; Reid, C ; Boulnois, G . ; Vann, W. F.; Roberts, I. S. J. Bacteriol. 1994, 776", 4583-9. (13) Annunziato, P. W. ; Wright, L . F. ; Vann, W. F.; Silver, R. P. J. Bacteriol. 1995, 777, 312-19. 144 (14) Finne, J. Trend. Biochem. Sci. 1985, 10, 129-32. (15) Amor, K . ; Heinrichs, D . E . ; Frirdich, E . ; Ziebell, K . ; Johnson, R. P.; Whitfield, C . Infect. Immun. 2000, 68, 1116-24. (16) Rick, P. D . ; Osborn, M . J. Proc. Natl. Acad. Sci. USA 1972, 69, 3756-60. (17) Rick, P. D . ; Osborn, M . J. J. Biol. Chem. 1977, 252, 4895-903. (18) Rick, P. D . ; Young, D . A . J. Bacteriol. 1982, 150, 447-55. (19) Roantree, R. J.; Kuo, T. T.; Macphee, D . G . J. Gen. Microbiol. 1977,103, 223-34. (20) Kropinski, A . M . B . ; Chan, L . ; Milazzo, F. H . Antimicrob. Agents Chemother. 1978,13, 494-9. (21) Black, R. E . ; Levine, M . M . ; Clements, M . L . ; Hughes, T. P.; Blaser, M . J. J. Infect. Dis. 1988, 157, 472-9. (22) Butzler, J. P.; Skirrow, M . B . Clin. Gastroenterol. 1979, 8, 737-65. (23) Skirrow, M . B . Brit. Med. J. 1977, 2, 9-11. (24) . Cover, T. L . ; Blaser, M . J. Ann. Rev. Med. 1992, 43, 135-45. (25) Thibault, P.; Logan, S. M . ; Kel ly, J. F.; Brisson, J. R.; Ewing, C. P.; Trust, T. J.; Guerry, P. J. Biol. Chem. 2001, 276, 34862-70. (26) Schirm, M . ; Soo, E . C ; Aubry, A . J.; Austin, J.; Thibault, P.; Logan, S. M . Mol. Microbiol. 2003, 45,1579-92. (27) Linton, D . ; Karlyshev, A . V . ; Hitchen, P. G . ; Morris, H . R.; De l l , A . ; Gregson, N . A . ; Wren, B . W. Mol. Microbiol. 2000, 35, 1120-34. (28) Morooka, T.; Umeda, A . ; Amako, K. J. Gen. Microbiol. 1985, 131, 1973-80. 145 (29) Caldwell, M . B . ; Guerry, P.; Lee, E . C ; Burans, J. P.; Walker, R. I. Infect. Immun. 1985, 50, 941-3. (30) Josenhans, C ; Labigne, A . ; Suerbaum, S. J. Bacteriol. 1995,177, 3010-20. (31) Petersen, M . ; Fessner, W. ; Frosch, M ; Luneberg, E . FEMSMicrobiol. Lett. 2000, 184, 161-4. (32) Ringenberg, M . A . ; Steenbergen, S. M ; V imr , E . R. Mol. Microbiol. 2003, 50, 961-75. (33) Kawamura, T.; Kimura, M . ; Yamamori, S.; Ito, E . J. Biol. Chem. 1978, 253, 3595-601. (34) Kawamura, T.; Ishimoto, N . ; Ito, E . Meth. Enzymol. 1982, 83, 515-19. (35) Stasche, R.; Hinderlich, S.; Weise, C ; Effertz, K ; Lucka, L . ; Moormann, P.; Reutter, W. J. Biol. Chem. 1997, 272, 24319-24. (36) Hinderlich, S.; Stasche, R.; Zeitler, R.; Reutter, W . J. Biol. Chem. 1997, 272, 24313-8. (37) Effertz, K ; Hinderlich, S.; Reutter, W. J. Biol. Chem. 1999, 274, 28771-8. (38) Keppler, O. T.; Hinderlich, S.; Langner, J.; Schwartz-Albiez, R.; Reutter, W. ; Pawlita, M . Science 1999, 284, 1372-6. (39) Kikuchi , K ; Tsuiki, S. Biochim. Biophys. Acta 1973, 327, 193-206. (40) Schwarzkopf, M . ; Knobeloch, K . P.; Rohde, E . ; Hinderlich, S.; Wiechens, N . ; Lucka, L . ; Horak, I.; Reutter, W. ; Horstkorte, R. Proc. Natl. Acad. Sci. USA 2002, 99, 5267-70. (41) Tanner, M . E . Accounts of Chemical Research 2002, 35, 237-246. (42) Kawamura, T.; Ishimoto, N . ; Ito, E . J. Biol. Chem. 1979, 254, 8457-65. (43 940 (44 (45 (46 (47; (48 35. (49 (so; (51 (52; (53 (54 (55 (56; (57 146 Yokoyama, K . ; Mizuguchi , H . ; Araki , Y . ; Kaya, S.; Ito, E . J. Bacteriol. 1989, 171, 6. Harrington, C. R.; Baddiley, J. Eur. J. Biochem. 1985,153, 639-45. Tanner, M . E . Curr. Org. Chem. 2001, 5, 169-92. Samuel, J.; Tanner, M . E . Nat. Prod. Rep. 2002,19, 261-77. Kopp, J.; Kopriva, S.; Suss, K . H . ; Schulz, G. E . J. Mol. Biol. 1999, 287, 761-71. Jelakovic, S.; Kopriva, S.; Suss, K . H . ; Schulz, G . E . J. Mol. Biol. 2003, 326, 127-Amein, M . ; Leatherwood, J. M . Biochem. Biophys. Res. Comm. 1969, 36, 223-7. Luchansky, S. J.; Yarema, K. J.; Takahashi, S.; Bertozzi, C. R. J. Biol. Chem. 2003,275, 8035-42. Barber, G . A . J. Biol. Chem. 1979, 254, 7600-3. Frey, P. A . FASEB J. 1996,10, 461-70. Morgan, P.; Sala, R.; Tanner, M . J. Am. Chem. Soc. 1997,119, 10269-77. Sala, R.; Morgan, P.; Tanner, M . E . J. Am. Chem. Soc. 1996,118, 3033-4. Campbell, R. E . ; Mosimann, S. C ; Tanner, M . E . ; Strynadka, N . C. Biochemistry 2000, 39, 14993-5001. Street, I. P.; Rupitz, K . ; Withers, S. G . Biochemistry 1989, 28, 1581-7. Mitchell , E . P.; Withers, S. G . ; Ermert, P.; Vasella, A . T.; Garman, E . F.; Oikonomakos, N . G . ; Johnson, L. N . Biochemistry 1996, 35, 7341-55. 147 (58) Morera, S.; Imberty, A . ; Aschke-Sonnenborn, U . ; Ruger, W. ; Freemont, P. S. J. Mol. Biol. 1999, 292, 717-30. (59) Comb, D . G . ; Roseman, S. Biochim. Biophys. Acta. 1958, 29, 653-4. (60) Glaser, L . Biochim. Biophys. Acta 1960, 41, 534-6. (61) Salo, W. L . ; Fletcher, H . G . , Jr. Biochemistry 1970, 9, 882-5. (62) Sommar, K . M . ; El l is , D . B . Biochim. Biophys. Acta 1972, 268, 581-9. (63) Sommar, K . M ; El l is , D . B . Biochim. Biophys. Acta 1972, 268, 590-5. (64) Frick, D . N . ; Townsend, B . D . ; Bessman, M . J. J. Biol. Chem. 1995, 270, 24086-91. (65) Legler, P. M . ; Massiah, M . A . ; Bessman, M . J.; Mildvan, A . S. Biochemistry 2000, 39, 8603-8. (66) Legler, P. M . ; Massiah, M . A . ; Mildvan, A . S. Biochemistry 2002, 41, 10834-48. (67) Chen, H . ; Blume, A . ; Zimmermann-Kordmann, M . ; Reutter, W. ; Hinderlich, S. Glycobiology 2002,12, 65-71. (68) Watson, D . R.; Jourdian, G . W. ; Roseman, S. J. Biol. Chem. 1966, 241, 5627-36. (69) Nakata, D . ; Close, B . E . ; Colley, K . J.; Matsuda, ! . ; Kitajima, K . Biochem. Biophys. Res. Comm. 2000, 273, 642-8. (70) Lawrence, S. M . ; Huddleston, K . A . ; Pitts, L . R.; Nguyen, N . ; Lee, Y . C ; Vann, W. F.; Coleman, T. A . ; Betenbaugh, M . J. J. Biol. Chem. 2000, 275, 17869-77. (71) Angata, T.; Nakata, D . ; Matsuda, T.; Kitajima, K . Biochem. Biophys. Res. Comm. 1999, 267,326-31. 148 (72) Gunawan, J.; Simard, D . ; Gilbert, M . ; Lovering, A . L . ; Wakarchuk, W. W. ; Tanner, M . E . ; Strynadka, N . C. J. J. Biol. Chem. 2005, 280, 3555-63. (73) Suryanti, V . ; Nelson, A . ; Berry, A . Prot. Express. Purifn. 2003, 27, 346-56. (74) Levin, D . H . ; Racker, E . J. Biol. Chem. 1959, 234, 2532-9. (75) Shulami, S.; Furdui, C. ; Adir , N . ; Shoham, Y . ; Anderson, K . S.; Baasov, T. J. Biol. Chem. 2004, 279, 45110-20. (76) DeLeo, A . B . ; Sprinson, D . B . Biochem. Biophys. Res. Commun. 1968, 32, 873-7. (77) Furdui, C. ; Zhou, L . ; Woodard, R. W, ; Anderson, K . S. J. Biol. Chem. 2004, 279, 45618-25. (78) Hedstrom, L . ; Abeles, R. Biochem. Biophys. Res. Commun. 1988, 757, 816-20. (79). Seeholzer, S. H . ; Jaworowski, A . ; Rose, I. A . Biochemistry 1991, 30, 727-32. (80) Matte, A . ; Tari, L . W. ; Goldie, H . ; Delbaere, L . T. J. Biol. Chem. 1997, 272, 8105-8. (81) Asojo, O.; Friedman, J.; Adir , N . ; Belakhov, V . ; Shoham, Y . ; Baasov, T. Biochemistry 2001, 40, 6326-34. (82) Wang, J ; Duewel, H . S.; Woodard, R. W. ; Gatti, D . L . Biochemistry 2001, 40, 15676-83. (83) Kaustov, L . ; Kababya, S.; Belakhov, V . ; Baasov, T.; Shoham, Y . ; Schmidt, A . J. Am. Chem. Soc. 2003,125, 4662-9. (84) Belakhov, V . ; Dovgolevsky, E . ; Rabkin, E . ; Shulami, S.; Shoham, Y . ; Baasov, T. Carbohydr. Res. 2004, 339, 385-92. (85) L i , Z . ; Sau, A . K . ; Shen, S.; Whitehouse, C ; Baasov, T.; Anderson, K . S. J. Am. Chem. Soc. 2003,125, 9938-39. 149 (86) Sundaram, A . K . ; Pitts, L . ; Muhammad, K . ; W u , J.; Betenbaugh, M . ; Woodard, R. W. ; Vann, W. F. Biochem. J. 2004, 383, 83-9. (87) Dotson, G . D . ; Nanjappan, P.; Reily, M . D . ; Woodard, R. W. Biochemistry 1993, 32, 12392-7. (88) Sundaram, A . K . ; Woodard, R. W. J. Org. Chem. 2000, 65, 5891-7. (89) Onderka, D . K . ; Floss, H . G . J. Am. Chem. Soc. 1969, 91, 5894-6. (90) Floss, H . G . ; Carroll, M . ; Onderka, D . K . J. Biol. Chem. 1972, 247, 736-44. (91) Parkhill, J.; Wren, B . W. ; Mungall , K . ; Ketley, J. M . ; Churcher, C ; Basham, D . ; Chillingworth, T.; Davies, R. M . ; Feltwell, T.; Holroyd, S.; Jagels, K . ; Karlyshev, A . V . ; Moule, S.; Pallen, M . J.; Penn, C. W. ; Quail, M . A . ; Rajandream, M . A . ; Rutherford, K . M . ; van Vliet, A . H . M . ; Whitehead, S.; Barrell, B . G. Nature 2000, 403, 665-8. (92) Goon, S.; Kel ly, J. F.; Logan, S. M . ; Ewing, C. P.; Guerry, P. Mol. Microbiol. 2003,50,659-71. (93) Creuzenet, C ; Schur, M . J.; L i , J.; Wakarchuk, W. W.; Lam, J. S. J. Biol. Chem. 2000,275, 34873-80. (94) Krishna, R.; Wold, F. Identification of common post-translational modifications. In Protein Structure - A Practical Approach; 2 ed.; Oxford University Press, New York, 1997. (95) Invitrogen Instruction Manual: Guide to Baculovirus Expression Vector Systems (BEVS) and Insect Cell Culture Techniques. (96) Smith, G . E.; Fraser, M . J.; Summers, M . D . J. Virol. 1983, 46, 584-93. (97) Luckow, V . A . ; Lee, S. C ; Barry, G . F.; Olins, P. O. J. Virol. 1993, 67,4566-79. (98) Bock, K . ; Pedersen, C. J. J. Chem. Soc, Perkin Trans. 2 1974, 2, 293-7. (99) Pravdic, N . ; Franjic-Mihalic, 1.; Danilovic, B . Carbohydr. Res. 1975, 45, 302-6. 150 (100) Legler, G . Adv. Carbohydr. Chem. Biochem. 1990, 48, 319-84. (101) La i , E. C. ; Withers, S. G . Biochemistry 1994, 33, 14743-9. (102) Lougheed, B . ; Ly , H . D . ; Wakarchuk, W. W.; Withers, S. G . J. Biol. Chem. 1999, 274,37717-22.. (103) Micheel, F. ; Wulff, H . Chem. Ber. 1956, 89, 1521-30. (104) Khorlin, A . Y . ; Shulman, M . L . ; Zurabyan, S. E . ; Privalova, I. M . ; Kopaevich, Y . L. Izv. Akad. Nauk. SSSR, Ser. Khim. 1968, 9, 2094-8. (105) Segel, I. H . Enzyme Kinetics: Behavior and Analysis of Rapid Equilibrium and Steady-State Enzyme Systems; Wiley-Interscience: New York, 1975. (106) Monod, J.; Wyman, J.; Changeux, J. P. J. Mol. Biol. 1965, 12, 88-118. (107) Koshland, D . E . ; Nemethy, G . ; Filmer, D . Biochemistry 1966, 5, 365-85. (108) Fersht, A . Enzyme Structure and Mechanism; 2nd ed.; W. H . Freeman and Company: New York, 1985. (109) Gosselin, S.; Alhussaini, M . ; Streiff, M . B . ; Takabayashi, K . ; Palcic, M . M . Anal. Biochem. 1994, 220, 92-7. (110) Samuel, J., Ph. D . Thesis, University of British Columbia, 2004. (111) L y , H. D . ; Withers, S. G . Ann. Rev. Biochem. 1999, 68, 487-522. (112) Cambers, J. P.; Elbein, A . D . ; Will iams, J. C. Biochem. Biophys. Res. Commun. 1982, 107, 1490. (113) Dong, W. L . ; Jespersen, T.; Bols, M . ; Skrydstrup, T.; Sierks, M . R. Biochemistry 1996,55,2788-95. (114) Conchie, J.; Hay, A . J.; Strachan, I.; Levvy, G. A . Biochem. J. 1967,102, 929-41. 151 115) Gold, A . M . ; Legrand, E . ; Sanchez, G . R. J. Biol. Chem. 1971, 246, 5700-6. 116) Knapp, S.; Vocadlo, D . ; Gao, Z. N . ; Ki rk , B . ; Lou, J. P.; Withers, S. G . J. Am. Chem. Soc. 1996,118, 6804-5. 117) Rose, I. A . Ann. Rev. Biochem. 1966, 35, 23-56. 118) Raushel, F. M . ; Villafranca, J. 1. CRC Crit. Rev. Biochem. 1988, 23, 1-26. 119) Cohn, M . ; Hu , A . Proc. Natl. Acad. Sci. USA 1978, 75, 200-3. 120) Cohn, M . ; Hu, A . J. Am. Chem. Soc. 1980,102, 913-6. 121) Sim, M . M ; Kondo, H . ; Wong, C . - H . J. Am. Chem. Soc. 1993,115, 2260-7. 122) Cleland, W. W. Biochemistry 1990, 29, 3194-7. 123) Samuel, J.; Tanner, M . E . Biochim. Biophys. Acta 2004,1700, 85-91. 124) Reutter, W. , personal communication to Tanner, M . E . , 2002. 125) Vimr , E . R.; Steenbergen, S. M ; Cieslewicz, M . J. Ind. Microbiol. 1995, 15, 352-60. 126) Bliss, J. M . ; Silver, R. P. Mol. Microbiol. 1996, 27, 221-31. 127) Whitfield, C ; Roberts, I. S. Mol. Microbiol; Blackwell Publishing Limited, 1999; V o l . 31, p 1307. 128) Frosch, M . ; Edwards, U . ; Bousset, K . ; Krausse, B . ; Weisgerber, C. Mol. Microbiol. 1991, 5, 1251-63. 129) Vann, W. 'F . ; Daines, D . A . ; Murkin , A . S.; Tanner, M . E . ; Chaffin, D . O.; Rubens, C. E . ; Vionnet, J.; Silver, R. P. J. Bacteriol. 2004,186, 706-12. 152 130 131 2004 132 133 69, 665-79. Chem. 2004,2004,3304-12. 134 135 136 137 138 139 140 141 142 143 144 Murkin , A . S., Ph. D.Thesis, University of British Columbia, 2004. Murkin , A . S.; Chou, W. K . ; Wakarchuk, W. W.; Tanner, M . E . Biochemistry 43, 14290-14298. Lee, H . J.; Strynadka, N . C , personal communication to Tanner, M . E . , 2006. Stoltz, F.; Reiner, M . ; Blume, A . ; Reutter, W. ; Schmidt, R. R. J. Org. Chem. 2004, Schafer, A . ; Thiem, J. J. Org. Chem. 2000, 65, 25-9. Stoltz, F.; Blume, A . ; Hinderlich, S.; Reutter, W. ; Schmidt, R. R. Eur. J. Org. Bradford, M . M . Anal. Biochem. 1916, 72, 248-54. Laemmli, U . K . Nature, 227, 680-5. Glasoe, P. K . ; Long, F. A . J. Am. Chem. Soc. 1960, 60, 188-90. GraFit, v., Erithacus Software, Ltd. Marquart, D . W . J. Soc. Ind. Appl. Math. 1963,11, 431-41. Labarca, C ; Paigen, K . Anal. Biochem. 1980,102, 344-52. Inoue, H . ; Nojima, H . ; Okayama, H . Gene 1990, 96, 23-8. Stratagene Instruction Manual: QuikChange® Site-Directed Mutagenesis Kit. Knirel , Y . A . ; Vinogradov, E . V . ; Lvov, V . L . ; Kocharova, N . A . ; Shashkov, A . S.; Dmitriev, B . A . ; Kochetkov, N . K . Carbohydr. Res. 1984, 133, C5-C8. 153 (145) Knirel , Y . A . ; Kocharova, N . A . ; Shashkov, A . S.; Dmitriev, B . A . ; Kochetkov, N . K . ; Stanislavsky, E . S.; Mashilova, G . M . Eur. J. Biochem. 1987,163, 639-52. (146) Kenne, L . ; Lindberg, B . ; Schweda, E . ; Gustafsson, B . ; Holme, T. Carbohydr. Res. 1988, 285-94. (147) Kondakova, A . N . ; Perepelov, A . V . ; Bartodziejska, B . ; Shashkov, A . S.; Senchenkova, S. N . ; Wykrota, M . ; Knirel , Y . A . ; Rozalski, A . Carbohydr. Res. 2001, 333, 241-9. (148) Flowers, H . M . ; Levy, A . ; Sharon, N . Carbohydr. Res. 1967, 189-95. (149) Liav, A . L . ; Sharon, N . Carbohydr. Res. 1973, 30, 109-26. (150) Simonsen, K. B . ; Ayida , B . K . ; Vourloumis, D . ; Takahashi, M . ; Winters, G. C ; Barluenga, S.; Qamar, S.; Shandrick, S.; Zhao, Q.; Hermann, T. ChemBioChem 2002, 3, 1223-8. (151) Lambert, J. B . ; Mazzola, E . P. In Nuclear Magnetic Resonance Spectroscopy: An Introduction to Principles, Applications and Experimental Methods; Pearson Prentic Hal l : New Jersey, 2004, p 270-2. (152) Webb, M . R. Proc. Natl. Acad. Sci. USA 1992, 89, 4884-7. (153) Bondinell, W . E . ; Vnek, J.; Knowles, P. F.; Sprecher, M . ; Sprinson, D . B . J. Biol. Chem. 1971, 246, 6191-6. (154) Vialletelle, V . ; Rabiller, C. ; Heisler, A . ; Levayer, F. Tetrahedron-Asymmetry 1992, 3, 673-6. (155) Cohn, M . ; Hu , A . Proc. Natl. Acad. Sci. USA 1978, 75, 200-3. (156) Logan, S. M . ; Kel ly , J. F. ; Thibault, P.; Ewing, C. P.; Guerry, P. Mol Microbiol 2002, 46, 587-97. (157) Knirel , Y . A . ; Vinogradov, E . V . ; Shashkov, A . S.; Dmitriev, B . A . ; Kochetkov, 1M. K . ; Stanislavsky, E . S.; Mashilova, G . M . Eur. J. Biochem. 1987,163, 627-37. 154 (158) Knirel , Y . A . ; Rietschel, E . T.; Marre, R.; Zahringer, U . Eur. J. Biochem. 1994, 221,239-45. (159) Knirel , Y . A . ; M o l l , H . ; Helbig, J. H . ; Zahringer, U . Carbohydr. Res. 1997, 304, 77-9. (160) Tsvetkov, Y . E . ; Shashkov, A . S.; Knirel , Y . A . ; Zahringer, U . Carbohydr. Res. 2001,357,233-7. (161) Schoenhofen, I. C ; McNal ly , D . J.; Vinogradov, E . ; Whitfield, D . ; Young, N . M . ; Dick, S.; Wakarchuk, W. W.; Brisson, J. R.; Logan, S. M . J. Biol. Chem. 2006, 257, 723-32. (162) L i u , F.; Tanner, M . E . J. Biol. Chem. in press 2006. (163) Guerry, P.; Ewing, C. P.; Schirm, M . ; Lorenzo, M . ; Kel ly, J . ; Pattarini, D . ; Majam, G. ; Thibault, P.; Logan, S. Mol. Microbiol. 2006, 60, 299-311. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0061142/manifest

Comment

Related Items