UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanistic studies of the enzymes involved in the biosynthesis of CMP-N, N'-diacetyllegionaminic acid… Glaze, Pavel Alexander 2009

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

Item Metadata

Download

Media
24-ubc_2009_fall_glaze_pavel.pdf [ 4.17MB ]
Metadata
JSON: 24-1.0061207.json
JSON-LD: 24-1.0061207-ld.json
RDF/XML (Pretty): 24-1.0061207-rdf.xml
RDF/JSON: 24-1.0061207-rdf.json
Turtle: 24-1.0061207-turtle.txt
N-Triples: 24-1.0061207-rdf-ntriples.txt
Original Record: 24-1.0061207-source.json
Full Text
24-1.0061207-fulltext.txt
Citation
24-1.0061207.ris

Full Text

MECHANISTIC STUDIES ON THE ENZYMES INVOLVED IN THE BIOSYNTHESIS OF CMP-N,N’-DIACETYLLEGIONAMINIC ACID AND UDP-D-APIOSE by  Pavel Alexander Glaze B.Sc., The University of British Columbia, 2003 A THESIS SUBMITTED iN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES  (Chemistry)  THE UMVERSITY OF BRITISH COLUMBIA  (Vancouver)  May, 2009 © Pavel Alexander Glaze, 2009  Abstract This thesis focuses on the biosynthesis of two sugar nucleotides. The enzymes responsible for the biosynthesis of N,N’-diacetyllegionaminic acid in Legionellapneumophila are identified for the first time. All three genes (neuA,B, C) demonstrated sequence homology to the genes involved in sialic acid biosynthesis. The first gene neuC encodes a hydrolyzing 2epimerase, which is found to catalyze the conversion of UDP-N,N’-diacetylbacillosamine (UDP Bac2,4diNAc) into 2,4-diacetamido-2,4,6-trideoxymannose (6-deoxyMandiNAc) and uridine diphosphate (UDP). The incubation of UDP-Bac2,4diNAc with NeuC in deuterated buffer H]-6-deoxyMandiNAc. This indicates that the reaction catalyzed by the 2 generated c-[2hydrolyzing 2-epimerase proceeds with a net retention of configuration at C-i, and that C-2 is deprotonated and reprotonated with a solvent-derived deuterium atom. An enzymatic reaction in 18 demonstrated that the loss of UDP occurs through a C-O bond cleavage process. These 2 H 0 results support a mechanism involving the anti-elimination of UDP, forming a 6-deoxy-2,4diacetamidoglucal intermediate, followed by a syn-hydration, to generate 6-deoxyMandiNAc. N,N’-diacetyllegionaminic acid synthase (NeuB) is a potential phosphoenolpyruvate-condensing synthase involved in the biosynthesis of N,N’-diacetyllegionaminic acid (Leg5Ac7Ac). This enzyme is proposed to catalyze the condensation of phosphoenolpyruvate (PEP) and 6deoxyMandiNAc to form Leg5Ac7Ac and phosphate. NMR spectroscopic analysis confirmed that NeuB is an acid synthase and that the N,N’-diacetyllegionaminic acid product has the D glycero-D-galacto configuration. Incubation with [2-’ 0]-PEP demonstrated that NeuB operates 8 via a C-O bond cleavage mechanism. Finally, the NeuA homolog was demonstrated to possess  11  CMP-N,N’-diacetyllegionaminic acid synthetase activity generating CMP-Leg5Ac7Ac, which is activated for use in lipopolysaccharide biosynthesis. UDP-D-apiose is biosynthesized from UDP-D-glucuronic acid by a bifunctional enzyme UDP-D-apiose/UDP-D-xylose synthase (AXS 1). NMR spectroscopic analysis confirmed that AXS 1 produces a roughly 1:1 mixture of UDP-D-apiose (UDP-Api) and UDP-D-xylose (UDP Xyl). Incubation of a potential reaction intermediate, UDP-4-ketoxylose, resulted in the slow formation of either UDP-Xyl and possibly UDP-Api. AXS 1 catalyzed the formation of exclusively UDP-2-deoxy-2-fluoroxylose when incubated with UDP-2-deoxy-2-fluoroglucuronic acid, while fluoride, UDP and CO 2 was formed when AXS1 was incubated with UDP-3-deoxy3-fluoroglucuronic acid. The enzymatic incubation with these substrate analogs provided further evidence for the retro-aldol mechanism.  111  Table of Contents Abstract . ii iv Table of Contents viii List of Figures xv List of Symbols and Abbreviations xxi Acknowledgements xxii Dedications 1 Chapter One: Sugar Nucleotide-Modifying Enzymes 2 Sugar Nucleotides 1.1 4 Epimerases and Racemases 1.2 5 Short-Chain Dehydrogenase/Reductase (SDR) Family 1.3 7 SDR Sugar Nucleotide Epimerases 1.4 7 1.4.1 UDP-Galactose 4-Epimerase (GalE) 12 1.4.2 GDP-mannose 3,5-Epimerase (GME) 14 1.4.3 GDP-fucose Synthase (GFS) 17 Cofactor-Independent Sugar Nucleotide Epimerases 1.5 17 1.5.1 Epimerases Operating at Substituents Bearing Activated Stereocenters 19 1.5.2 UDP-N-acetylglucosamine 2-Epimerases 23 Sugar Nucleotide Deoxygenation 1.6 23 1.6.1 SDR Sugar Nucleotide 4,6-Dehydratases 25 1.6.2 Sugar Nucleotide Aminotransferases 27 Sugar Nucleotide Dehydrogenases 1.7 29 Sugar Nucleotide Decarboxylases 1.8 30 Biosynthesis of Higher Order Sugars 1.9 30 1.9.1 Sialic Acid Biosynthesis 34 1.9.2 Pseudaminic Acid Biosynthesis 36 1.9.3 Legionaminic Acid Biosynthesis 41 Branched-Chain Sugar Nucleotides 1.10 41 1.10.1 Methylation Reaction in the Biosynthesis of Mycarose 42 1.10.2 Biosynthesis of UDP-D-Apiose 51 Project Goals 1.11 53 Chapter Two: Biosynthesis of CMP-N,N’-Diacetyllegionaminic Acid 54 Introduction 2.1 55 Preparation of UDP-Bac2,4diNAc 2.2 56 2.2.1 Expression and Purification of Pg1F, Pg1E and Pg1D enzymes 57 2.2.2 Chemoenzymatic synthesis of UDP-Bac2,4diNAc 22.3 Identification and Mechanistic Studies on the Hydrolyzing UDP-Bac2,4diNAc 60 Epimerase from Legionella pneumophila 2.3.1 Expression and Purification of the Hydrolyzing TJDP-Bac2,4diNAc 260 epimerase iv  61 Testing the Activity of the Hycirolyzing UDP-Bac2,4diNAc 2-epimerase Kinetic Characterization of the Hydrolyzing UDP-Bac2,4cIiNAc 2-epimerase.. 63 64 2.3.4 Stereochemical Analysis and Solvent Isotope Incorporation 69 2.3.5 Test for C-O vs. P-O Bond Cleavage 2.4 Identification and Mechanistic Studies on N,N’-Diacetyllegionaminic Acid 72 Synthase (NeuB) 2.4.1 Expression and Purification of N,N’-Diacetyllegionaminic Acid Synthase. 73 75 2.4.2 Test for N,N’-Diacetyllegionaminic Acid Synthase (NeuB) Activity 80 2.4.3 Isolation and Characterization of N,N’-Diacetyllegionaminic Acid 82 2.4.4 Potential Synthase Mechanisms and Test for C-O vs P-O Bond Cleavage 85 CMP-N,N’-Diacetyllegionaminic Acid Synthetase 2.5 2.5.1 Identification, Expression and Purification of CMP-N,N’-Diacetyllegionaminic 85 Acid Synthetase 86 2.5.2 Activity of CMP-N,N’-Diacetyllegionaminic Acid Synthetase 87 Conclusions 2.6 90 Future Directions 2.7 93 2.8 Experimental 93 2.8.1 Materials and General Methods 94 2.8.2 Cloning of L. pneumophilia neuA, neuB and neuC 2.8.3 Over-expression and Purification of L. pneumophila NeuA, NeuB and NeuC. 95 2.8.4 Sub-cloning of L pneumophila neuB for Fusion Protein, Overexpression and 96 Purification 97 2.8.5 Over-expression and Purification of His-Tagged Pg1F and PglE 98 2.8.6 Chemo-enzymatic Synthesis of UDP-Bac2,4diNAc 99 2.8.7 Characterization of Hydrolyzing 2-Epimerase Activity 99 2.8.7.1. NeuC Homolog Activity Assay 100 2.8.7.2. NeuC Kinetic Studies 2.8.7.3. Stereochemistry and Solvent Deuterium Isotope Incorporation Studies 100 101 2.8.7.4. Metal Dependency of NeuC 101 2.8.8 Test for C-O vs. P-O Bond Cleavage Mechanism 102 2.8.9 Characterization of N,N’-Diacetyllegionaminic Acid Synthase 102 2.8.10 Isolation and Characterization of the N,N’-Diacetyllegionamirilc Acid 103 2.8.11 Test for C-O vs. P-O Bond Cleavage Mechanism 2.8.12 Characterization of CMP-N,N’-Diacetyllegionaminic Acid Synthetase 104 Activity 105 Chapter Three: Mechanistic Studies on UDP-D-Apiose Synthase 106 Introduction 3.1 2.3.2 2.3.3  .  V  Cloning, Expression and Purification of UDP-D-Apiose/UDP-D-Xylose 107 Synthase 107 3.2.1 Preparation of cDNA from Arabidopsis thaliana 109 3.2.2 Expression and Purification ofAXS1 113 Testing the Activity ofAXS1 3.3 113 P NMR Spectroscopy 3.3.1 Monitoring the Activity ofAXS1 Using ‘H and 31 119 Cj-D-Glucuronic Acid 3 3.3.2 Preparation of UDP-[U-’ 121 C]-Glucuronic Acid 3 3.3.3 Testing the Activity ofAXS1 Using UDP-[U-’ 124 Catalytic Competence of UDP-4-ketoxylose 3.4 124 3.4.1 Preparation of IJDP-4-ketoxylose 127 3.4.2 Reaction of UDP-4-ketoxylose with AXS 1 132 Testing UDP-Xylose as a Potential Substrate 3.5 133 Attempted Synthesis of UDP-D-Apiose 3.6 139 Fluorinated Analogs of UDP-G1cA 3.7 139 3.7.1 Introduction 145 3.7.2 Synthesis of UDP-2-Deoxy-2-fluoro-D-glucuronic Acid 147 3.7.3 Testing UDP-2F-G1cA as a Substrate Analog 152 3.7.4 Synthesis of UDP-3-Deoxy-3 -fluoro-D-glucuronic Acid 153 3.7.5 Testing UDP-3F-G1cA as a Substrate Analog 160 Conclusions and Future Directions 3.8 162 Experimental 3.9 162 3.9.1 Materials and General Methods 162 3.9.2 Source and Cloning ofAXSJ 163 3.9.3 Over-expression and Purification ofAXS1 164 3.9.4 Source and Cloning of amA 164 3.9.5 Over-expression and Purification of AmA 164 3.9.6 Over-expression and Purification of UDP-Glucose Dehydrogenase 165 3.9.7 Monitoring Enzyme Incubation by NMR Spectroscopy 166 3.9.8 AXS1 Activity Test with UDP-G1cA 166 C]-Glucuronic Acid 3 3.9.9 Enzymatic synthesis of UDP-[U-’ 167 3.9.10 Studies with UDP-[U-13C]-GlcA 168 3.9.11 Enzymatic Synthesis and Isolation of UDP-4-ketoxylose 168 3.9.12 Test of UDP-4-ketoxylose as a Potential Intermediate for AXS1 169 3.9.13 AXS1 Activity Test with UDP-Xyl 169 3.9.14 Attempted Synthesis of UDP-Api 10 3.9.14.1. Methyl 3’ -benzyl-2,3-O-isopropylidene-f3-D-erythro-apifuranose 5 169 170 3.9.14.2. 3’ -Benzyl- 1,2,3 -O-triacetyl-D-erythro-apifuranose 6 171 3.9.14.3. 3’-Benzyl-D-emythro-apifuranose 12 171 3.9.14.4. 1,2,3,3’ -Tetrabenzyl-D-erythro-apifuranose 13 172 3.9.14.5. 2,3,3 ‘-Tribenzyl-D-erythro-apifuranose 14  3.2  vi  HA  061 LU 9L1 9L i cU 17L1 CLI EL I  X!pUddV £qdu.IoJjq  V°I9.JE-[DEtfl]dGfl pU V°ID-1E-aGfl I{TM uoiwqnolq TSXV 6U6 OUIO 81 6 E -n1-aun JO StStflU 1 vIo--[D 8Z V01D-.!I-dUf1J0 S!STW1(S 0T 1(Z10tU1D LT6E vojoiaun T4flM uorpqnouj isxv 9U6 U V0IO-Z-dGflJO sisqu( tWUIICZUOUJI3 cF6C .9 j7 I 6 C  i qdsoqd sou  pCzuqwj  List of Figures Figure 1.1 Generic sugar nucleotide  2  Figure 1.2 Biosynthesis of a sugar nucleotide from an unactivated sugar  3  Figure 1.3 A) Mechanism employed by epimerase catalyzing a reaction at an activated stereocenter. B) Reaction catalyzed by epimerase at an unactivated center. B and BR represent active site basic and acidic residues, respectively  5  Figure 1.4 Positions and roles of the catalytic residues in SDR enzymes  6  Figure 1.5 Inversion of configuration catalyzed by UDP-galactose 4-epimerase (GalE)  7  Figure 1.6 Proposed mechanism of the UDP-galactose 4-epimerase mechanism, showing the UDP-4-ketoglucose  8  Figure 1.7 Solvent isotope incorporation experiment with GalE showing an absence of incorporation  8  Figure 1.8 Tritium incorporation into the ketone intermediate analog of GalE  9  Figure 1.9 Accidental release of UDP-4-ketoglucose from the active site. Square brackets indicate species bound to GalE active site  10  Figure 1.10 Mechanism of the reaction catalyzed by GalE  11  Figure 1.11 Reactions catalyzed by GDP-mannose 3,5-epimerase (GME)  12  Figure 1.12 Mechanism of the reaction catalyzed by GME  13  Figure 1.13 Positioning of Cys 145 and Lys 217 in the active site of the GME  14  Figure 1.14 Biosynthesis of GDP-L-fucose from GDP-a-D-mannose  15  Figure 1.15 Proposed mechanism for the reaction catalyzed by GDP-L-fucose synthase  16  Figure 1.16 Deuterium washout experiment with Hisl79Gln GFS  17  Figure 1.17 The reaction catalyzed by Rm1C  18 viii  Figure 1.18 The reactions catalyzed by NovW and EvaD  19  Figure 1.19 Non-hydrolyzing and hydrolyzing UDP-G1cNAc 2-epimerases  20  Figure 1.20 Reaction and mechanism catalyzed by the bacterial “non-hydrolyzing” UDP GIcNAc 2-epimerase  21  Figure 1.21 Positional isotope exchange experiment with RffE  22  Figure 1.22 Mechanism of the hydrolyzing UDP-G1cNAc 2-epimerase  23  Figure 1.23 Proposed mechanism for the dTDP-Glc 4,6-dehydratase reaction  24  Figure 1.24 dTDP-glucose 4,6-dehydratase (Rm1B) catalyzes a stereospecfic hydride transfer.. 25 Figure 1.25 Mechanism of the reaction catalyzed by the PLP-dependent Ty1B enzyme  26  Figure 1.26 Formation of TDP-3-keto-6-deoxyglucose  27  Figure 1.27 Proposed mechanism of the UDP-glucose dehydrogenase reaction  28  Figure 1.28 Proposed mechanism for the reaction catalyzed by UDP-glucuronic acid decarboxylase  30  Figure 1.29 Sialic acid and sialic acid-related sugars  31  Figure 1.30 Biosynthesis of CMP-NeuAc in mammals and bacteria  33  Figure 1.31 Biosynthesis of CMP-pseudaminic acid  36  Figure 1.32 Structure of legionaminic acid derivatives and sialic acid  37  Figure 1.33 Structure of L. pneumophila LPS, the major component of the outer surface of the outer membrane of Gram-negative bacteria  38  Figure 1.34 Proposed biosynthesis of CMP-Leg5Ac7Ac  39  Figure 1.35 Biosynthesis of UDP-Bac2,4diNAc in Campylobacterfejuni  40  Figure 1.36 Structure of L-mycarose and proposed mechanism of the reaction catalyzed by Ty1C 3  42  ix  Figure 1.37 D-Apiose involvement in formation of borate cross-link between two different homogalacturonan chains  43  Figure 1.38 The UDP-D-apiose/UDP-D-xylose synthase reaction  44  Figure 1.39 Retro-aldol mechanism for the UDP-D-apiose/(JDP-D-xylose synthase reaction.  ...  46  Figure 1.40 Fate of the labeled C-3”0fUDP-[3”C]-glucuronic acid during the reaction 14 catalyzed by UDP-D-apiose/UDP-D-xylose synthase  47  Figure 1.41 Fate of the labeled 4”H of UDP-[U3 C, 414 H1-G1cA during the reaction catalyzed 3 by UDP-D-apiose/UDP-D-xylose synthase  47  Figure 1.42 Determination of stereochemistry in the conversion of UDP-[4”H]-G1cA to IJDP 3 H]-apiose 3 [3”-  48  Figure 1.43 Carbon migration mechanism for the UDP-D-apiose/UDP-D-xylose synthase reaction  50  Figure 2.1 Chemoenzymatic synthesis of UDP-Bac2,4diNAc  55  Figure 2.2 SDSPAGE gel of enzymes involved in the biosynthesis of UDP-Bac2,4diNAc  57  Figure 2.3 The 400 MHz ‘H NMR spectrum of UDP-Bac2,4diNAc in D 0 2  59  Figure 2.4 SDS-PAGE gel showing NeuC purification  61  Figure 2.5 3 ’P NMR spectra monitoring the reaction of UDP-Bac2,4diNAc with NeuC  62  Figure 2.6 Continuous coupled UDP assay used to determine kinetic parameters of the hydrolyzing epimerase  64  Figure 2.7 Kinetic plots of initial velocity (vo)I[E]o vs. UDP-Bac2,4diNAc  64  Figure 2.8 ‘H NMR spectra monitoring the reaction of UDP-Bac2,4diNAc with the NeuC in 0. A) Enzymatic conversion of UDP-Bac2,4diNAc to [22 D H]-6-deoxy-ct-MandiNAc, followed 2  x  by mutarotation to 2 [2H ]-6-deoxy--MandiNAc and B) ‘H NMR spectra monitoring the enzymatic reaction  66  Figure 2.9 HMQC experiment with 6-deoxyMandiNAc  68  Figure 2.10 C-O vs P-O bond cleavage experiment. A) Products from C-O bond cleavage mechanism. B) Products from P-O bond cleavage mechanism  70  Figure 2.11 ESI-mass spectra of C-O vs. P-O bond cleavage experiment with NeuC  71  Figure 2.12 Mechanism of the reaction catalyzed by the hydrolyzing UDP-Bac2,4diNAc 2epimerase  72  Figure 2.13 SDS-PAGE gel showing Ma1E-NeuB purification  74  Figure 2.14 Reaction catalyzed by N,N’-diacetyllegionaminic acid synthase (NeuB) and Fischer projections of 6-deoxyMandiNAc and N,N’-diacetyllegionaminic acid  76  Figure 2.15 31 P NMR spectra monitoring the reaction of PEP and 6-deoxyMandiNAc with MalE NeuB. A) Before the addition of Ma1E-NeuB and B) 18 h after the addition of Ma1E-NeuB  77  Figure 2.16 Partial ‘H NMR spectra monitoring the incubation of PEP and 6-deoxyMandiNAc with Ma1E-NeuB A) Before the addition of Ma1E-NeuB and B) after the addition of Ma1E-NeuB. 79 Figure 2.17 ‘H NMR spectrum of Leg5Ac7Ac in D 0 2  81  Figure 2.18 Proposed C-O vs. P-O bond cleavage mechanisms for Ma1E-NeuB. A) C-O bond cleavage mechanism and B) P-O bond cleavage mechanism  83  P NMR spectra monitoring the incubation of Ma1E-NeuB with [2-’ Figure 2.19 31 0]PEP and 68 deoxyMandiNAc A)Before the addition of Ma1E-NeuB and B) after the addition of Ma1E-NeuB. 84 Figure 2.20 SDS-PAGE gel showing NeuA purification  86  xi  Figure 2.21 Reaction catalyzed by CMP-Leg5Ac7Ac synthetase  87  Figure 2.22 Biosynthesis of CMP-Leg5Ac7Ac in Legionellapneumophila  89  Figure 2.23 Potential tetrahedral intermediate analogs for NeuB  92  Figure 3.1 Preparation of eDNA from plant mRNA  109  Figure 3.2 SDS-PAGE gel showing AXS1 purification  111  Figure 3.3 Partial UV spectrum of 147 iM AXS1 in 20 mM Tris-HC1 pH 8.0 before and after the addition of sodium borohydride (to give 1.2 mM sodium borohydride)  112  Figure 3.4 ‘H NMR spectra monitoring the reaction of UDP-G1cA with the AXS1 at different time intervals  116  Figure 3.5 Non-enzymatic decomposition of the UDP-Api to a-D-apio-D-furanosyl 1,2-cyclic phosphate and UMP  117  Figure 3.6 31 P NMR spectra monitoring the enzymatic conversion of UDP-G1cA to UDP-Api and UDP-Xyl. UDP-Api decomposes to UMP and 1,2 cyclic phosphate after over-night incubation.. 118 Figure 3.7 Preparation of UDP-[U-’ C]-D-glucuronic acid 3  120  Figure 3.8 ‘ C NMR spectra monitoring the enzymatic conversion of UDP-[U-’ 3 C]-G1cA to 3 C] -Api and UDP-[U-’ 3 UDP-[U-’ C1-Xyl 3  122  Figure 3.9 ‘ C]-Api and UDP-[U-’ 3 C NMR spectrum of UDP-{U-’ 3 C]-Xyl 3  123  Figure 3.10 Biosynthesis of UDP-4-amino-4-deoxy-L-arabinose in polymyxin-resistant E. coli and Salmonella typhimurium  125  Figure 3.11 SDS-PAGE gel showing AmA purification  126  Figure 3.12 Reaction of UDP-4-ketoxylose with H 0 and MeOH 2  127  xli  Figure 3.13 A) Proposed reaction between UDP-4-ketoxylose and AXS 1 [NADH]. B) Partial UV spectra of AXS 1-bound NADH. Spectra were collected in 10 mM phosphate buffer, pH 8.0, at 37 °C with [AXS1j  =  70 jiM and [UDP-4-ketoxylose]  =  81 jiM  129  Figure 3.14 Negative ESI-mass spectra monitoring the incubation ofAXS1 with UDP-4ketoxylose  131  Figure 3.15 Structures of UDP-Api and UDP-galactofuranose  134  Figure 3.16 Open chain form of D-apiose and possible cyclization products  135  Figure 3.17 Proposed synthetic route for the formation of UDP-Api  137  Figure 3.18 Alternative synthetic route for preparation of anomerically deprotected apiose.  ...  139  Figure 3.19 Proposed mechanism for the inactivation of CDP-D-glucose 4,6-dehydratase by CDP-6-deoxy-6,6-difluoro-glucose  141  Figure 3.20 A) Reaction catalyzed by 1-deoxy-D-xylulose-5-phosphate (DXP) reductoisomerase. B) cL-ketol rearrangement mechanism. C) Retro-aldol rearrangement mechanism. D) Fluorinated substrate analogs  143  Figure 3.21 Proposed results of the incubation of fluorinated analogs with AXS 1  145  Figure 3.22 Synthesis of UDP-2-deoxy-2-fluoro-D-glucuronic acid 22  147  Figure 3.23 Partial ‘ F NMR spectra monitoring the reaction of UDP-2F-G1cA with AXS1 over a 9 24hperiod  148  Figure 3.24 Partial 1 H NMR spectrum of isolated product form the reaction between UDP-2FG1cAandAXS1  150  Figure 3.25 Proposed reaction between UDP-2F-GIcA 22 and AXS 1  151  Figure 3.26 Synthesis of UDP-3-deoxy-3-fluoro-D-glucuronic acid 28  153  xlii  Figure 3.27 ‘ F NMR spectra monitoring the incubation of UDP-3F-G1cA 28 with AXS1 in 50 9 mM potassium phosphate pH 8.0 (282.4 MHz, D 0) 2  154  Figure 3.28 Partial ‘ C NMR spectra monitoring the reaction of UDP-[U-’ 3 C]-3F-G1cA with 3 AXS1 over a 24 h period  157  Figure 3.29 Proposed reaction between UDP-3F-G1cA 28 and AXS 1  159  Figure A.1 ‘H NMR spectrum of 2,4-diacetamido-2,4,6-trideoxymannose (D 0, 300 MHz)... 191 2 Figure A.2. ‘H NMR spectrum of UDP-4-ketoxylose (hydrate form, D 0, 400 MHz) 2  192  Figure A.3 ‘H NMR spectrum of 2,3,3’-tribenzyl-D-apiose 14. (CDC1 , 400 MHz) 3  193  Figure A.4 ‘H NMR spectrum of UDP-2-deoxy-2-fluoro-D-glucuronic acid as a triethyl ammonium salt 22. (D 0, 400 MHz) 2  194  Figure A.5 Partial ‘H COSY spectrum of UDP-2F-Xyl (D 0, 400 MHz) 2  195  Figure A.6 ‘H NMR spectrum of UDP-3-deoxy-3-fluoro-D-glucuronic acid as a triethy1-  196  Figure A.7 ‘ C]-3-deoxy-3-fluoro-D-glucuronic acid as a 3 C NMR spectrum of UDP-[U-’ 3 triethylammonium salt (D 0, 600 MHz) 2  197  xiv  List of Symbols and Abbreviations chemical shift (ppm) Ac  acetyl  0 2 Ac  acetic anhydride  AcOH  acetic acid  0 2 Ag  silver(I) oxide  A1cDH  alcohol dehydrogenase  A1dDH  aldehyde dehydrogenase  amA  gene encoding the protein AmA from E. coil  AmA  bifunctional enzyme involved in the biosynthesis of UDP-N-formyl-4amino-4-deoxy-L-arabinose.  ArnB  pyridoxal phosphate-dependent enzyme from E. coil  ATP  adenosine triphosphate  AXSJ  gene encoding AXS 1 in Arabidopsis thaliana  AXS 1  UDP-D-apiose/UDP-D-xylose synthase  Bn  benzyl  BSA  bovine serum albumin  CoA  coenzyme A  CKABP  2-carboxy-3 -keto-D-arabinitol 1 ,5-bisphosphate  CMP-Leg5Ac7Ac  cytidine 5 ‘-monophospho-N,N ‘-acetyllegionaminic acid  CMP-NeuAc  cytidine 5 ‘-monophospho-N-acetylneuraminic acid  COSY  correlation spectroscopy  CTP  cytosine triphosphate  d  doublet (NMR)  dt  doublet of triplets (NMR) xv  D  H) 2 deuterium (  Da  Dalton  dd  doublet of doublets (NMR)  DMF  N,N-dimethylformamide  DMSO  dimethyl sulfoxide  DNA  deoxyribonucleic acid  DTT  1 ,4-dithio-D,L-threitol  E. coil  Escherichia coil strains BL2 1 (DE3) or JM 109  ESI-MS  electrospray ionization mass spectrometry  EvaD  dTDP-3 -amino-2,3 ,6-trideoxy-3 -C-methyl-D-erythro-hexopyranos- ulose 5-epimerase  EtOH  ethanol  Gal  galactose  GalE  UDP-galactose 4-Epimerase  GDP-Fuc  guano sine 5’ -diphospho-L-fucose  GDP-Man  guanosine 5’ -diphospho-D-mannose  GDP-Gal  guanosine 5’ -diphospho-D-galactose  GFS  GDP-fucose synthase  GME  GDP-mannose 3,5-epimerase  GMER  GDP-4-keto-6-deoxy-mannose 3 ,5-epimerase/4-reductase  Glc  glucose  GlcNAc  N-acetylglucosamine  G1cNAc 6-P  N-acetylglucosamine 6-phosphate  H. pylon  Helicobacterpylori  HMQC  heteronuclear multiple-quantum coherence NMR spectroscopy  Hz  Hertz (s’) xvi  IPTG  isopropyl 3-D-thiogalactopyranoside  J  coupling constant (NMR); subscripts indicate coupling partners  kcat  catalytic rate constant  kcat/Km  specificity constant  KDN  2-keto-3-deoxy-D-glycero-D-galacto-2-nonulosonic acid  KDO  2-keto-3-deoxy-D-manno-2-octulosonic acid  Km  Michaelis constant  LB  Luria-Bertani medium  Leg  legionaminic acid  Leg5Am7Ac  5-N-acetimidoyl-7-N-acetyllegionaminic acid  Leg5Ac7Ac  N,N’-diacetyllegionaminic acid  LPS  lipopolysaccharide  m  multiplet (NMR)  Man  mannose  ManNAc  N-acetylmannosamine  ManNAc 6-P  N-acetylmannosamine 6-phosphate  MeOH  methanol  mRNA  messenger ribonucleic acid  MW  molecular weight  NAD  nicotinamide adenine dinucleotide, oxidized form  NADH  nicotinamide adenine dinucleotide, reduced form  NADP  nicotinamine adenine dinucleotide phosphate, oxidized form  NADPH  nicotinamine adenine dinucleotide phosphate, reduced form  NaOMe  sodium methoxide  neuA  gene encoding the protein NeuA in L. pneumophila xvii  NeuA  CMP-Leg5Ac7Ac synthetase from L. pneumophila  neuB  gene encoding the protein NeuB in L. pneumophila  NeuB  Leg5Ac7Ac synthase from L. pneumophila  neuC  gene encoding the protein NeuC in L. pneumophila  NeuC  hydrolyzing UDP-Bac2,4diNAc 2-epimerase from L. pneumophila  NeuAc  N-acetylneuraminic acid or sialic acid  N. meningitidis  Neisseria meningitidis  NMR  nuclear magnetic resonance  NovW  dTDP-6-deoxy-D-xylo-4-hexulose 3 -epimerase  600 0D  optical dispersion at 600 nm  PCR  polymerase chain reaction  PEP  phosphoenolpyruvate  1 P  inorganic phosphate  PIX  positional isotope exchange  pglD  gene encoding PglD in C. jejuni  Pg1D  UDP-6-deoxy-4-amino-G1cNAc acetyltransferase  pglE  gene encoding Pg1E in C. jejuni  Pg1E  UDP-6-deoxy-4-keto-G1cNAc PLP-dependent aminotransferase  pglF  gene encoding PglF in C. jejuni  PglF  UDP-N-acetyl-glucosamine 4,6-dehydratase  PMP  pyridoxamine phosphate  PLP  pyridoxal phosphate  ppm  parts per million  Pse  pseudaminic acid  PseB  UDP-N-acetylglucosamine 5-inverting 4,6-dehydratase xviii  PseC  UDP-2-acetyl-2,6-dideoxy-13-L-arabino-4-hexulose aminotransferase  PseF  CMP-pseudaminic acid synthase  PseG  UDP-2,4-diacetamido-2,4,6-trideoxy-L-altrose hydrolase  PseH  UDP-2-acetamido-2,4,6-trideoxy-L-altrose N-acetyltransferase  Psel  pseudaminic acid synthase  pyr  pyridine  RffE  non-hydrolyzing UDP-G1cNAc 2-epimerase from E. coli  Rm1B  dTDP-glucose 4,6-dehydratase  Rm1C  dTDP-4-dehydrorhamose 3 ,5-epimerase  RG-II  rhamnogalacturonan-Il  RNA  ribonucleic acid  rpm  revolutions per minute  RT  room temperature  s  singlet (NMR)  [SJ  substrate concentration  SAM  S-adenosyl methionine  SDR  short chain dehydrogenase/reductase  SDS-PAGE  sodium dodecylsulfate polyacrylamide gel electrophoresis  SelectfluorTM  1 -chloromethyl-4-fluoro- 1 ,4-diazoniabicyclo[2.2.2] octane bis(tetrafluoroborate)  t  time  TB  terrific broth  TDP  thymidine diphosphate  THF  tetrahydrofuran  Tris  2-amino-2-(hydroxymethyl}- 1,3 -propanediol  Ty1A2  TDP-4,6-glucose dehydratase xix  Ty1B  TDP-6-deoxy-4-keto-glucose isomerase  Ty1C3  TDP-2,6-dideoxy-4-ketoaltrose methyl transferase  U  enzyme unit  UDP  uridine diphosphate  UDP-Api  uridine 5’ -diphospho-D-apiose  UDP-Bac2,4diNac  uridine 5’ -diphospho-N, N’-diacetylbacillosamine  UDP-Gal  uridine 5’ -diphospho-D-galactose  UDP-Glc  uridine 5 ‘-diphospho-D-glucose  UDP-G1cA  uridine 5’ -diphospho-D-glucuronic acid  UDP-2F-G1cA  uridine 5’ -diphospho-2-deoxy-2-fluoro-D-glucuronic acid  UDP-3F-G1cA  uridine 5’ -diphospho-3-deoxy-3 -fluoro-D-glucuronic acid  UDPG1cDH  UDP-glucose dehydrogenase  UDP-G1cNAc  UDP-N-acetylglucosamine  UDP-ManNAc  UDP-N-acetylmannosamine  UMP  uridine 5’-monophophosphate  UTP  uridine 5’-triphosphate  UV  ultraviolet  Vmax  maximal reaction velocity (rate)  xx  Acknowledgements Firstly, I would like to thank my supervisor Martin Tanner for supporting me for the past several years. Martin is an excellent professor who has a great deal of patience. I appreciate his taking the time to help me and share his experiences with me. The ‘Life Lessons’ you have passed on to me will not be forgotten. I would also like to thank Dr. Wayne Chou and Dr. James Morrison for assisting me during my first few years in the lab. Also, without the help and support of the current group members I would not be able to complete my projects. Thank you to Timin Hadi for providing me with an excellent breakdown of both the chemistry and sports worlds. Thank you to Feng Liu for helping me with computer related issues, Louis .Luk for giving me lessons in microbiology, and Alaine “PD” Mayer for his help with this thesis. I would also like to thank all other group members: Stephen Lau, Xu Li, Jackie Bassiri, Jennifer Griffith, Niusha Mahmoodi and Yanjie Liu for providing me with an excellent environment to work in. I am grateful to Dr. Elena Polishchuk and Jie Chen for accommodating me in Biological Services and lending a helping hand when I needed it, and to Maria Ezhova of the NMR Facility for her assistance. Thank you also to the Withers group and Perrin group for allowing me to have use of the UV/Vis spectrophotometer. Finally, I would like to thank our collaborators Dr. Martin Young and David Watson of NRC for generously donating the recombinant plasmid encoding neuC, neuB and neuA, and thank you to Dr. Reinhard Jetter for donating the plant leaves ofArabidopsis thaliana.  xxi  Dedicated to My wfe Sarb, My parents and My sister Anna  xxii  Chapter One Sugar Nucleotide-Modifying Enzymes  1  1.1 Sugar Nucleotides Current progress in glycobiology and carbohydrate biochemistry has caused a re-evaluation of the importance of carbohydrates in biology. Generally, carbohydrates are known to be used as an energy source, as structural elements, and as solubilizing parts of glycoproteins.’ Recent studies have shown that carbohydrates are also involved in biochemical signaling and cell recognition. A sugar nucleotide is a nucleotide linked via a phosphate ester bridge to the anomeric position of a sugar (Figure 1.1). The base portion of the molecule can be either a purine (adenine or guanine) or a pyrimidine (cytosine, uracil or thymine).  Base  Figure 1.1 Generic sugar nucleotide. The nucleotide portion of a sugar nucleotide serves to activate the sugar since it acts as an excellent leaving group in a transferase reaction, where the sugar is transferred to a glycosyl acceptor molecule. Moreover, many transferases and sugar-modifying enzymes will use the nucleotide portion of the sugar to differentiate between various activated sugars. The biosynthesis of sugar nucleotides commonly occurs via one of two distinct pathways. 2 Firstly, the sugar nucleotide may be generated from an unactivated sugar in a process that involves several enzymatic steps (Figure 1 .2). A 6-phospho sugar is initially generated by a reaction between adenosine 5’-triphosphate (ATP) and an unactivated sugar, which is catalyzed by a kinase enzyme. The product is then isomerized to a higher energy 1-phosphosugar by the 2  action of a phosphomutase enzyme. The sugar nucleotide is finally generated by reaction with a nucleotide triphosphate as catalyzed by a pyrophosphorylase enzyme. Sugar nucleotides such as UDP-glucose (UDP-GIc), UDP- N-acetylglucosamine (UDP-G1cNAc) and GDP-mannose (GDP 4 Man) are all biosynthesized in this fashion.  =  kinase  OH  OH  2 3 opo  +ADP  unactivated  6-phosphosugar  oxtp +pp  1 -phosphosugar  nucleotide  Figure 1.2 Biosynthesis of a sugar nucleotide from an unactivated sugar. An alternative pathway to biosynthesize sugar-nucleotides involves the direct modification of pre-existing sugar nucleotides with the use of sugar nucleoticle-modifying enzymes. This is an efficient method as it avoids the requirement of additional enzymes to activate each sugar subsequent to its formation. TJDP-galactose (UDP-Gal) and GDP-fucose (GDP-Fuc), as well as a large number of less common sugar nucleotides which are produced by plants and microorganisms are produced in this manner. 5 This chapter will discuss the best understood sugar nucleotide-modifying enzymes, with an emphasis on the chemical mechanism and the role of key active site residues. Many of the enzymes that will be discussed belong to the short chain dehydrogenase/reductase (SDR) family. A brief introduction of this family will also be presented in this chapter. Finally, the research goals of this thesis will be introduced.  3  1.2 Epimerases and Racemases Epimerases and racemases are enzymes that catalyze an inversion of a carbon 6 Racemases act on substrates with only one asymmetric stereocenter in biological molecules. center, while epimerases catalyze a stereochemical inversion on molecules with multiple asymmetric centers. The inversion of a carbon stereocenter could conceivably occur by breaking and reforming any of the four bonds at that center, however, most racemases and epimerases 6 utilize C-H bond cleavage during catalysis. Epimerases can be divided into two general types. The first type of epimerase operates at what is generally called an “activated” center. The activated center is a carbon adjacent to a functional group (e.g. a carbonyl) that can stabilize a carbanion. The PKa of the proton at these activated stereocenters is relatively low (generally  <  30) allowing for a direct deprotonationlreprotonation mechanism to invert stereocenter (Figure 1.3 A). The second type of epimerase catalyzes the reaction at an “unactivated” stereocenter. The PKa of the proton at these unactivated stereocenters is generally above 30. As a result, the enzymes are not able to directly deprotonate the stereocenter and must use an alternative pathway.  4  A 0  R  R1LR  1HX BH  XH B: BH  BH  Jw  9  B  H X 2 RR  Mechanism  X H RR  Figure 1.3 A) Mechanism employed by epimerase catalyzing a reaction at an activated stereocenter. B) Reaction catalyzed by epimerase at an unactivated center. B and BR represent active site basic and acidic residues, respectively.  1.3 Short-Chain Dehydrogenase/Reductase (SDR) Family The short-chain dehydrogenase/ reductase family (SDR) is a large and evolutionarily old family of NAD(P)(H)-dependent enzymes. Of the approximately 3000 known members of the family, 63 have been identified in the human genome, a number similar to that of the ubiquitous P450 enzymes. 7 The functions of SDR enzymes span several categories including oxidoreductases, lyases, and isomerases, with oxidoreductases making up the majority of the 8 Despite having low sequence identity among different enzymes (typically 10-30%), all forms. available three-dimensional structures display a highly similar a/f3 folding pattern with a Rossman fold near the N-terminus of the peptide. Most SDR enzymes have a 250-300 residue core structure, with the binding and the active site regions being most conserved. 7 Binding of the NAD(P)(H) cofactor occurs in the N terminal part of the molecule, where a conserved Gly-X -Gly-X-Gly sequence is located. 3 8 SDR family enzymes are also identified by their active site catalytic triad, that is responsible for promoting a hydride transfer to and from the NAD(P)(H) cofactor and a carbonyl group of the 5  substrate. The triad consists of a Ser, Tyr and Lys residue, of which Tyr is the most highly conserved residue throughout the whole family. All previous research indicates that the Tyr functions as a catalytic acid that protonates the carbonyl during the reduction (Figure 1 4)•7 8 The serine is located in close proximity to the carbonyl of the substrate and the hydroxyl group of the Tyr. Ser is responsible for polarizing the carbonyl group and helping to adjust the pKa of the ’ 9 Tyr.  10  The Lys interacts with the nicotinamide ribose, and lowers the pKa of the Tyr-OH to  1 A recent discovery suggests that a fourth residue, Asn, could play promote the proton transfer.’ a role in the catalytic reaction. It is believed that Asn forms an interaction with a water molecule which in turn is involved in the proton relay system.’ ’ 2  13  However, considering the enormous  spread of SDR enzymes, it is important to note that this mechanism and the role of the active site residues do not have to be identical for all of the SDR enzymes. NADH :0 ADP’  substrate HHH /OH Asparagine  5 .  O Serine  Lysine  Figure 1.4 Positions and roles of the catalytic residues in SDR enzymes.  6  1.4 SDR Sugar Nucleotide Epimerases 1.4.1  UDP-Galactose 4-Epimerase (GalE) One of the most heavily studied SDR sugar-nucleotide modifying enzymes is UDP  galactose 4-epimerase (GalE), which catalyzes the interconversion of UDP-galactose and UDP glucose (Figure 1.5). GalE is found in the Leloir pathway that converts galactose into glucose 14 Deficiencies in any one of the enzymes that are involved in the Leloir pathway in phosphate.’ humans result in the diseased state known as galactosemia.lZ After its initial discovery, the GalE enzyme was referred to as a “Waldenase” enzyme.’ 6 At that time, a reaction at carbon that proceeds with an inversion of stereochemical configuration was referred to as a “Walden 17 Thus, it was thought that nucleophilic attack at C-4” by water would displace the Cinversion”. 4” hydroxyl and result in the inverted configuration. Mechanistic studies later disproved this potential mechanism but the name “Waldenase” remained with the enzyme. OH OH HO OUDP  UDP-galactose  UDP-galactose 4-epimerase  OH HO—R OUDP UDP-glucose  Figure 1.5 Inversion of configuration catalyzed by UDP-galactose 4-epimerase (GalE). The epimerase isolated from Escherichia coli has been the focus of numerous mechanistic and structural investigations. Initial research on GalE focused on the elucidation of the chemical mechanism. TheE. coli enzyme functions as a homodimer of subunits containing 338 amino acid residues and one tightly bound NAD 8 cofactor.’ 19 The NAD cofactor cannot be removed from ’ the active site without denaturating the enzyme and is regenerated upon the completion of each  7  reaction cycle. Based on this information, it was suggested that reaction occurs through an oxidation at C-4” to generate a ketone intermediate (Figure 1.6). NAD  NADH  OH ,OH  NAD  ,OH  0H 1  HOQ ‘OH OUDP  OHI  ‘OHI  OUDP UDP-4-ketoglucose  UDP-galactose  OUDP  UDP-glucose  Figure 1.6 Proposed mechanism of the UDP-galactose 4-epimerase mechanism, showing the UDP-4-ketoglucose. In order to confirm the proposed mechanism several expreriments were performed. One of 0 or 0 2 18 as a solvent for 2 H the earliest experiments performed on the GalE enzyme was using D 20 Results indicated that neither of the labeled atoms was incorporated the enzymatic reaction. into the substrate upon extended incubation (Figure 1.7). This showed that the mechanism of the reaction  does  involve  not  a  simple  deprotonationlreprotonation  at  C-4”  or  a  dehydration/rehydration sequence. H<,0H 9  bHl  OUDP UDP-galactose  UDP-galactose 4-epimerase 18 2 H 0 or 0 H 2  /JH HO---&--O HO-\  OH1  OUDP UDP-glucose  Figure 1.7 Solvent isotope incorporation experiment with GalE showing an absence of incorporation. The first direct evidence for the formation of the UDP-4-ketoglucose intermediate was reported in 1971. lii that experiment Kirkwood initially treated GalE with NaB 4 which H 3 H as the cofactor. 3 produced a reduced form of the enzyme containing NAD ’ The incubation of 2 8  the enzyme with previously prepared UDP-4-keto-6-deoxyglucose produced a mixture of UDP 6-deoxy- [43 HI-glucose and UDP-6-deoxy- [4H]-galactose (Figure 1.8). Further evidence 3 supporting the direct C-4” oxidation mechanism comes from the observation of a primary kinetic isotope effect. ’ 22  23  Using UDP-[4H]-hexoses, a primary kinetic isotope effect was observed, 3  indicating that the C-4” C-H bond is broken during catalysis.  O: + +  OIJDP  UDP-4-keto6-deoxyglucose  NADH -  H ouiw  UDP-6-deoxyH]-galactose 3 -[4-  ou  UDP-6-deoxyH]-glucose 3 -[4-  Figure 1.8 Tritium incorporation into the ketone intermediate analog of GalE. Lastly, it was reported that the extended incubation of UDP-glucose with GalE leads to the accumulation of UDP-4-ketoglucose and inactive enzyme. The inactivation occurred due to the occasional release of the 4-keto intermediate that leaves the free enzyme bearing a tightly bound NADH cofactor (Figure 1.9). The reduced form of the enzyme will then bind either UDP-glucose or UDP-galactose to form an abortive complex where epimerization cannot occur. ’ 24  25  The  intermediate was trapped by reduction with NaB 4 to generate UDP-[4H 3 H]glucose and UDP 3 H]galactose. 3 [426  9  GalE  NADH  NADOH OH  NAD  OH OO  4  OUDP UDP-galactose  OUDP UDP-4-ketoglucose I  I  UDP-glucose  accidental release of intermediate  OH 0  0  HO  +  OH OUDP UDP-4-ketoglucose  HO  GalE [NADH] \\UDP-hexose  OUDP +  GalE [NADH] UDP-hexose  OHOH 3O  abortive complex  OUDP  Figure 1.9 Accidental release of UDP-4-ketoglucose from the active site. Square brackets indicate species bound to GalE active site. Interesting aspects of the GalE reaction still remained unsolved even after the chemical mechanism of the reaction had been elucidated. Unlike other dehydrogenases in the SDR family, which catalyze stereospecific hydride transfers to and from one particular face of a molecule, 25 ’ 21 GalE reduces the ketone intermediate without a stereochemical preference. The hydride transfer is non-stereospecific with respect to the substrate. This is not the case with the NAD cofactor, since the hydride always transfers to the pro-S position on the B side of the nicotinamide ring. In order for enzymatic reaction to occur there are two possible scenarios: i) the ketone intermediate is released into the solution and is rebound with the other face of the carbonyl exposed to the hydride, or ii) the ketone intermediate is retained within the active site and undergoes a dramatic  10  conformational reorientation. Based on all previous research, it was hard to imagine that the active site of the enzyme could acconimodate full reorientation of the substrate. In order to address the phenomenon, an isotopic crossover experiment was performed. An incubation of GalE with UDP-glucose-d 7 and UDP-glucose-do was monitored by mass 27 The results showed that no isotopic scrambling that would form d spectrometry. 6 species 1 or d occurred during the reaction, indicating that the substrate remains in the active site of the enzyme throughout the catalytic cycle. This means that epimerization occurs within a single active site with one NAD cofactor controlling the reaction (Figure 1.10). It was proposed that a 1800 rotation of the keto intermediate occurred within the active site and enabled both faces of the carbonyl to be exposed to the static cofactor. 28 NADH  NAD  OH 0 HO  HO OH UDP-glucose rotation  NADH  NAD  UDP-galactose Figure 1.10 Mechanism of the reaction catalyzed by GalE. In the late 1990’s, X-ray crystallography was used to help identify key residues responsible for the enzymatic reaction and to verify the proposed conformational reorganization. The crystal structures of E. coil GalE revealed the location of the three conserved residues of 11  SDR enzymes: Lys 153, Tyr 149 and Ser 124. Lys 153 was found to be important for enhancing  the chemical activity of NAD upon the binding of uridine nucleotides in the active site. 29 The ’ 18 Tyr and Ser residues were responsible for deprotonating the C-4” hydroxyl group during the hydride transfer. Ultimately, the crystal structures of inactive, mutant forms of GalE in complexes with either UDP-Glc or UDP-Gal confirmed the proposed reorientation mechanism utilized by the enzyme. ° 3 1.4.2  GDP-mannose 3,5-Epimerase (GME) GDP-mannose 3,5-epimerase (GME) catalyzes the interconversion of GDP-a-D-mannose  with GDP-3-L-galactose and GDP-t3-L-gulose (Figure 1.1 1).’ The enzyme was first discovered  32 The enzyme is unique among sugar epimerases in the green alga Chiorella pyrenoidosa. because it is able to invert the configuration at two distinct “unactivated” stereocenters using a 33 Even though the term “epimerase” is commonly used to describe this single active site. enzyme, GME is not an epimerase in the strict sense because GDP-ct-D-mannose and GDP--L galactose are diastereomers and not epimers. The formation of GDP-3-L-galactose is believed to be the first committed step in the biosynthesis of L-ascorbic acid (vitamin C), whereas the fate of GDP--L-gulose is still unclear. However, it has been postulated that L-gulose and L-gulo-1,4lactone could also be involved in the biosynthesis of vitamin C. 34 OH OH  GME  OGDP GDP-c-D-mannose  HO  OH IjOGDP  GME  HO  JGDP  OH  OH  GDP-f3-L-gulose  GDP-3-L-ga1actose  Figure 1.11 Reactions catalyzed by GDP-mannose 3,5-epimerase (GME).  12  In order to elucidate the catalytic mechanism of GME, an enzymatic incubation was carried out in tritium-labeled water. It was found that solvent-derived tritium was incorporated at the C-3” and C-5” positions of the sugar skeleton. 35 Given that GME is a member of the SDR family and bears a tightly bound NAD cofactor, this fmding indicates that epimerization proceeds through ene(di)ol intermediates. The mechanism of the reaction is thought to occur via a transient oxidation at C-4” to give a ketone intermediate and tightly bound NADH (Figure 1.12). The formation of an “activated” center at C-4” lowers the pKa of the protons at C-3” and C-5” and, as a result, the reaction can proceed via a deprotonation!reprotonation mechanism at each carbon center. A final reduction at C-4” regenerates the alcohol and the NAD cofactor. 33 NAD  NADH  OH  NADH C-5” inversion  oxidation  OGDP  OGDP GDP-a-D-mannose  ring flip, then reduction  ring flip,  then reduction  OGDP  NAD  NAD HOJOGDP  GDP-3-L-galactose  GDP-3-L-gu1ose  Figure 1.12 Mechanism of the reaction catalyzed by GME.  13  The structures of complexes between GME from Arabidopsis thaliana and different GDP-hexoses bound in the active site were solved. 33 The results indicated that GME has the classical SDR family fold, with all three important catalytic residues present. The mutant GME Tyrl74Phe was not able to catalyze the epimerization, indicating that Tyr 174 is involved in the oxidation of C-4”. Since the formation of GDP-D-altrose (C-3” epimerization only) was not observed, it was proposed that GME epimerizes C-5” first and C-3” second. The two active site residues located in appropriate positions to act as acidlbase catalysts in the epimerization steps are Cys 145 and Lys 217 (Figure 1.13). In the forward direction, Cys 145 is located below the sugar and is thought to act as a base to abstract the proton at C-5”, while Lys 217 position above the sugar and delivers the new proton to C-5”. The sugar is then believed to undergo a slight movement in the active site to expose the C-3” center for the inversion reaction. Lys2l7 2 NH HOH ON OH -O HO H OGDP H  s— Cys145 Figure 1.13 Positioning of Cys 145 and Lys 217 in the active site of the GME. 1.4.3  GDP-fucose Synthase (GFS) GDP-L-fucose is biosynthesized in two enzymatic steps from the precursor GDP-a-D  mannose. The first step is catalyzed by GDP-mannose 4,6-dehydratase (GMD) which converts GDP-cc-D-mannose into GDP-6-deoxy-4-keto-mannose (Figure 1.14).36  The mechanism of 14  this class of enzymes will be discussed later in this chapter. The second step of the biosynthesis is catalyzed by the SDR enzyme, GDP-fucose synthase (GFS). GFS is able to catalyze three distinct reactions in a single active site. It inverts the configuration at both C-3” and C-5” of the GDP-6-deoxy-4-keto-mannose, and then carries out an NADPH-dependent reduction of a ketone at C-4” to give GDP-L-fucose. 2 L-Fucose is found in many important glycoconjugates in both prokaryotes and eukaryotes. 38  NADPH OGDP GDP-cc-D-mannose  OGDP GDP-6-deoxy-4keto-D-mannose  OH GDP-L-fucose  Figure 1.14 Biosynthesis of GDP-L-fucose from GDP-c-D-mannose.  The mechanism of GFS has recently been elucidated (Figure 1.1 5)•39 Due to the presence of an activated center at C-4”, the stereochemical inversions occur via a deprotonationl reproto nation mechanism. A subsequent reduction of the C-4” carbonyl by NADPH generates the product. Based on X-ray crystallography data obtained with E. coli GFS, Cys 109 and His 179 were identified as the potential acid/base residues required to perform the epimerizations. ° 4 ’ 36  15  His 179 H1+ HC OH Ot  C OH 3 H 0  OGDP  OGDP  OGDP  Cy 109  CyslO9 GDP-6-deoxy4-keto-D-altrose  GDP-6-deoxy4-keto-D-mannose  His 179 H1 COGDP 3 H  NADPH 31 H OGDP H  OHO GDP-L-fucose  GDP-6-deoxy4-keto-L-galactose  OH  Figure 1.15 Proposed mechanism for the reaction catalyzed by GDP-L-fucose synthase.  The CyslO9Ser mutant produced GDP-6-deoxy-D-altrose as the major product, indicating that C-3” epimerization occurs first. The premature reduction of the GDP-4-keto-6-deoxy-Daltrose intermediate led to the formation of this product. Moreover, the His 179Gm mutant was inactive towards normal catalysis, yet catalyzed the wash-out of deuterium label from the C-3” position of [3”H]-GDP-6-deoxy-4-keto-mannose (Figure 1.1 2  This result indicates that the  mutant is properly folded and able to carry out the first step of catalysis. The unmodified CyslO9 can act as a base and removes the deuterium from C-3” to generate the enol(ate) intermediate. The inversion of configuration does not take place because the acidic His 179 was mutated,  16  however, an exchange of the Cys 109 thiol deuterium with the bulk solvent can occur and results in the delivery of a proton back to the C-3” position. HCOH  OH  F  B  GFS [Hisl79Gln]  1OGDP  [3 “H1-GDP-6-deoxy2 4-keto-D-mannose  NADPH  OGDP GDP-6-deoxy4-keto-D-mannose  Figure 1.16 Deuterium washout experiment with His 179Gm GFS.  As a result, it is believed that GFS first performs an epimerization at C-3”, followed by epimerization at C-5”. This is unlike GME which carries out the inversions in the reverse order. 33 The reduction step catalyzed by GFS uses the same catalytic triad as the other SDR family enzymes, with Tyr 136 acting as acid to deliver the proton to the oxygen at C-4” (Figure 1.15).  1.5 Cofactor-Independent Sugar Nucleotide Epimerases 1.5.1  Epimerases Operating on Substituents Bearing Activated Stereocenters As discussed earlier, most SDR enzymes carry out epimerization by first creating an  activated center next to a carbon which will undergo an inversion. There are few examples of enzymes that do not rely on the presence of an NAD(P) cofactor. The classic example of this is dTDP-4-dehydrorhamnose 3,5-epimerase (RmlC). 42 This enzyme catalyzes the second step in ’ 41 the biosynthesis of dTDP-rhamnose and represents a potential drug target against bacterial infections. RmlC catalyzes the inversion of configuration at two stereocenters of dTDP-6-.deoxyD-xylo-4-hexulose to produce dTDP-6-deoxy-L-lyxo-4-hexulose (Figure 1.17). Despite the  17  seeming similarity between GME/GFS and Rm1C reactions, the enzymes share no sequence or structural similarity, as no redox chemistry is required in the latter case. Rm1C 3,5-epimerase  OdTDP dTDP-6-deoxy-D-xylo-4-hexulose  dTDP-6-deoxy-L-lyw-4-hexulose  Figure 1.17 The reaction catalyzed by Rm1C.  The incubation of Rm1C with dTDP-6-deoxy-D-xylo-4-hexulose in deuterated solvent 0) resulted in the incorporation of deuterium at C-3” and C-5”. 2 (D 43 This provides evidence for a direct deprotonation/reprotonation mechanism. This conclusion is not surprising, considering that both carbons are activated by their location next to a ketone at C-4”. Investigations of Rm1C from Streptococcus suis have identified a histidine residue that acts as a base and a Tyr that acts as an acid in both epimerization steps.’ There are two other enzymes that are closely related to RmlC (Figure 1.18). One is dTDP-6-deoxy-D-xylo-4-hexulose 3-epimerase (NovW) and the other is dTDP-3-amino-2,3 ,6trideoxy-3 -C-methyl-D-erythro-hexo-pyran-4-ulose  5-epimerase  (EvaD),  cofactor-independent epimerizations at C-3” and C-5”, respectively. ” 4  which 46  catalyze  Their chemical  mechanisms are believed to involve deprotonation! reprotonation steps, similar to RmIC.  18  NovW  HO OH OdTDP  3 -epimerase  dTDP-6-deoxy-D-ylo-4-hexu1ose  OH  0 OcITDP  dTDP-6-deoxy-D-ribo-4-hexulose  03  0  O  0  EvaD  C 3 H  C 3 H  2 NH  OdTDP  3 CH NH 2  5-epimerase  dTDP-3-amino-2,3 ,6-trideoxy-3 -C-methyl-D-eiythro-hexopyran-4-ulose  OdTDP  dTDP-3-amino-2,3,6-trideoxy-3-C-methyl-L-threo-hexopyran-4-ulose  Figure 1.18 The reactions catalyzed by NovW and EvaD.  1.5.2  UDP-N-Acetylglucosamine 2-Epimerases UDP-N-Acetylglucosamine 2-epimerases are interesting examples of enzymes that act at  an “unactivated” stereocenter. There are two known classes of UDP-G1cNAc 2-epimerases. An example of a “non-hydrolyzing” UDP-GlcNAc 2-epimerase (RffE in E. coli), which catalyzes the reversible epimerization of UDP-G1cNAc and UDP-ManNAc, exists only in bacteria (Figure 1.1 9)47 48 A “hydrolyzing” UDP-G1cNAc-2-epimerase is involved in the biosynthesis of sialic acid. This enzyme also hydrolyzes the glycosidic bond such that the release of UDP accompanies the inversion of configuration (Figure 1.1 9)49 The latter reaction is not a true epimerization as it is irreversible and the substrate and products are not epimers. Both “non-hydrolyzing” and hydrolyzing epimerases share sequence homology and both catalyze similar reactions on the same substrate, so it is reasonable to expect that they utilize similar catalytic mechanisms.  19  non-hydrolyzing UDP-GIcNAc 2-epimerase  HO  HO  Hc-I-R  OUDP HO HQ AcFIN OUDP  NHAc  OUDP hydrolyzing UDP-G1cNAc 2-epimerase HQR O 2 H  UDP  OH  Figure 1.19 Non-hydrolyzing and hydrolyzing UDP-G1cNAc 2-epimerases. The unusual mechanism of the “non-hydrolyzing” epimerase is believed to involve an antielimination that generates 2-acetamidoglucal and UDP, followed by a syn-addition of the UDP group to generate the final product (Figure 1.20). Initial evidence for this mechanism was O. It was observed that deuterium was 2 provided by performing the enzymatic incubation in D incorporated into the sugar at C-2”, forming [2”H]UDP-G1cNAc and 47 2 H]UDP-ManNAc. 2 [2”’ 48  Moreover, a primary kinetic isotope effect of 1.8 was observed when rates of reactions with  H]UDP-G1cNAc and unlabeled UDP-GlcNAc were compared. This supports the notion that 2 [2”the C-2” C-H bond is broken during the enzymatic reaction.  20  Enz—B:N anti-elimination  HO  HO  syn-addition  NHAc  OH )_  AcHN j Enz—BH  OIOUMP  OH  o..j  V  OUMP  II  H  0  0 UDP-ManNAc  UDP-GIcNAc  Figure 1.20 Reaction and mechanism catalyzed by the bacterial “non-hydrolyzing” UDP GlcNAc 2-epimerase.  A positional isotope exchange (PIX) experiment provided further evidence for this mechanism (Figure 1.21). UDP-G1cNAc was prepared with an ‘O label at the anomeric position. During the enzymatic incubation with RflE it was observed that the  180  label was scrambled  between the bridging and non-bridging positions of the UDP. This scrambling indicates that the anomeric C-0 bond is broken during catalysis and the 3-phosphate has time to rotate such that any of the three chemically equivalent oxygen atoms could reform the C-0 bond. Finally, it was found that, upon an extended incubation of RffE and UDP-GlcNAc, the intermediates 2acetamidoglucal and UDP are released into solution. It turns out that this process is thermodynamically favorable so the intermediates accumulate during the incubation and can be detected.  21  Hl0 Ac18Op_OU,il UDP-G1cNAc  O-P-OUMP 0 UDP-G1cNAc  -  HOO -  2-acetamidoglucal  \ AcFIN  9 OH  / HQO O-P-OUMP 0 UDP-ManNAc  UDP-ManNAc  Figure 1.21 Positional isotope exchange experiment with RffE. The “hydrolyzing” UDP-GlcNAc 2-epimerase is believed to use a very similar mechanism (Figure 1 .22).° Running the reaction in D 0 led to the incorporation of deuterium 2 into the C-2 position of ManNAc. In addition, incubation in 0 18 led to the incorporation of 180 2 H label at the anomeric carbon of ManNAc. Moreover, when [1-’ 0]UDP-G1cNAc with 8  180  in the  bridging position between the anomeric carbon and the n-phosphorus was incubated with the hydrolyzing UDP-GlcNAc 2-epimerase, the products were x-ManNAc and ‘ 0-labeled UDP. 8 The reaction mechanism is thought to be initiated by the anti-elimination of UDP, followed by the syn-hydration of 2-acetamidoglucal, forming x-ManNAc.  22  anti-elimination__HOQ\ AcHN  syn-addition  1 HO-H  UDP-G1cNAc  OH cL-ManNAc  Figure 1.22 Mechanism of the hydrolyzing UDP-G1cNAc 2-epimerase.  1.6 Sugar Nucleotide Deoxygenation 1.6.1  SDR Sugar Nucleotide 4,6-Dehydratases Sugar nucleotide 4,6-dehydratases catalyze the formation of 4-keto-6-deoxy sugar  nucleotides via an overall elimination of water (Figure 1.23). These dehydratases use a transient oxidation reaction that requires the use of an NAD cofactor. 5 This is reminiscent of the epimerases discussed earlier which create an “activated” center in the molecule through transient oxidation of a hydroxyl group. Sugar nucleotide 4,6-dehydratases are required in the biosynthesis of all 6-deoxyhexoses, which are the most abundant naturally occurring deoxy sugars after 2’ Three of the most intensely studied sugar nucleotide 4,6-dehydratases are 5 deoxy-D-ribose. dTDP-glucose 4,6-dehydratase (Rm1B), GDP-mannose 4,6-dehydratase (GMD), and CDP glucose 4,6-dehydratase (CGD). GMD was previously mentioned in the discussion of GDP fucose biosynthesis (Section 1.4.3). All of the enzymes are SDR family members and are believed to employ a similar mechanism that is exemplified by the Rm1B reaction.  23  The first step of the mechanism is believed to be an oxidation at C-4” to create a ketone ’ 52 intermediate and NADH (Figure.1.23).  The presence of a ketone at C-4” creates an  “activated” center at C-5”, which can then undergo a dehydration reaction across the C-5”/C-6” bond. This will generate an ct,3-unsaturated ketone, which may then undergo a reduction reaction. A hydride from NADH is delivered to C-6”, and a proton from an active site acid is delivered to C-5” to give the product. 53 NADH  NAD  NADH  OH  OTDP  OTDP  dTDP-glucose  NAD  Al, BH NAD çCH3  NADH  OTDP dTDP-4-keto-6-deoxy-glucose  a. f3 unsaturated ketone intermediate  Figure 1.23 Proposed mechanism for the dTDP-Glc 4,6-dehydratase reaction.  One of the first pieces of evidence in support of this mechanism came from running the reaction in either 2 0 or 3 H 0. It was discovered that solvent-derived isotope label was 2 H 5 More compelling evidence was provided ’ 54 incorporated at the C-5” position of the product. when dTDP-glucose 4,6-dehydratase was incubated with dTDP-(6S)-[4H, 56 2 H]-glucose. The 3 6final product was discovered to be dTDP-4-keto-6-deoxy-(6R)- [6H, 62 H] -glucose, indicating 3  24  that Rm1B catalyzes an intramolecular hydride transfer from C-4” to C-6” in a stereospecific fashion (Figure 1.24).  Rm1B  dTDP-(6S)- 3 H 2 [4”H ]-glucose ,6”-  H 2 dTDP-4-keto-6-deoxy-(6R)-[6”H 3 ]-glucose ,6”-  Figure 1.24 dTDP-glucose 4,6-dehydratase (Rm1B) catalyzes a stereospecfic hydride transfer. 1.6.2  Sugar Nucleotide Aminotransferases Many naturally occurring sugars contain nitrogen atoms as a substituent. Amino sugars  are found in glycoproteins, glycolipids and a variety of secondary metabolites. I have already discussed UDP-N-acetylglucosamine (Section 1.5.2), which is probably the most common example of a nitrogen-containing sugar nucleotide. The nitrogen functionality is introduced in early steps of the biosynthesis when fructose 6-phosphate is converted to glucosamine 6phosphate by glucosamine 6-phosphate synthase. 57 The introduction of nitrogen into sugar nucleotides is somewhat rare and generally occurs by a reaction between a corresponding keto sugar and ammonia derived from an amino acid. This reaction is catalyzed by pyridoxal phosphate (PLP)-dependent enzymes called aminotransferases. One such sugar nucleotide aminotransferase is found in the biosynthetic pathway of 3amino-3,6-dideoxyglucose which is a part of the antibiotic tylosin. 58 The reaction between TDP 3-keto-6-deoxyglucose and glutamate, as the source of ammonia, is catalyzed by TylB which is a pyridoxal phosphate (PLP)-dependent enzyme (Figure 1.25). TDP-3-keto-6-deoxyglucose is 25  generated from TDP-glucose by the sequential action of TDP 4,6-glucose dehydratase (Ty1A2) and an isomerase (Ty1M3) that isomerizes the ketone functionality between C-4” and C-3” (Figure 1.26). C 3 H Ho—R H3N  TyIB  OTDP TDP-3-amino-3,6dideoxyglucose  TDP-6-deoxy3-ketoglucose  +  PMP  PLP  c -o 1 2 -cO ---0  a-ketoglutarate  3 NH  L-glutamate  Figure 1.25 Mechanism of the reaction catalyzed by the PLP-dependent Ty1B enzyme. It was reported that, in the reaction catalyzed by TylB, an equivalent of pyridoxamine phosphate (PMP) could replace PLP and glutamate to give the final product. 59 The reaction mechanism is believed to involve first a reaction between PLP and glutamate to form PMP and ct-ketoglutarate. Afterwards, the second transamination can occur between PMP and the ketone of the sugar (Figure 1.25). Several other aminotransferase 60 enzymes, 61 such as ArnB ’ 62 which catalyzes the formation of UDP-4-amino-4-deoxy-13-L-arabinose, have been reported since the discovery of Ty1B.  26  TyIA2  3 H OTDP TDP-glucose  Ty1M3  TDP-6-deoxy-3-ketoglucose  Figure 1.26 Formation of TDP-3-keto-6-deoxyglucose.  1.7 Sugar Nucleotide Dehydrogenases Altering the oxidation state of the sugar nucleotide is an important transformation that is carried out by several enzymes. 2 The majority of NAD-dependent dehydrogenases catalyze reactions involving a single oxidation step. The introduction of a carboxylate into a substrate is commonly achieved via two sequential oxidations that require two different enzymes. An alcohol dehydrogenase (A1cDH) is normally required to oxidize an alcohol to an aldehyde, and an aldehyde dehydrogenase (A1dDH) to oxidize an aldehyde to a carboxylic acid. However, there is a small group of enzymes that can carry out the 2-fold oxidation in a single active 63 site. The most thoroughly investigated of these is UDP-glucose dehydrogenase. The enzyme catalyzes the conversion UDP-D-glucose to UDP-D-glucuronic acid. The reaction is believed to occur via a two-fold NAD-dependent oxidation with an aldehyde as an intermediate of the reaction (Figure 1 .27).64  27  ENZ  I  HO  NAD NADH  ENZ  I  S  HO-  -  OUDP a e y e intermediate  OUDP thiohemiacetal intermediate  NADH ENZ HO  HOR  OUDP UDP-D-glucuronic acid  thioester intermediate  Figure 1.27 Proposed mechanism of the IJDP-glucose dehydrogenase reaction.  The first oxidation occurs at C-6” to generate the aldehyde intermediate, which is rapidly attacked by an active site thiol to give a thiohemiacetal intermediate. A second molecule of NAD is then bound to the enzyme and oxidizes the thiohemiacetal to give a covalently bound thioester intermediate. Finally, hydrolysis of the thioester generates the product. A close examination of all known sequences of the enzyme has indicated the presence of a conserved Cys in the active site of homologs of this enzyme. The Cys26OSer mutant in the Streptococcus pyogenes enzyme led to an essentially inactive catalyst. Moreover, a prolonged incubation of the serine mutant resulted in the formation of the corresponding ester intermediate, which was detected by mass spectrometry. 65 It appears that the mutant can slowly catalyze both oxidation 28  steps, however the hydrolysis of the unnatural ester linkage is exceptionally slow, so the adduct accumulates.  1.8 Sugar Nucleotide Decarboxylases There are many reported examples of enzymes that catalyze decarboxylation reactions, however, there are only a few that are known to perform this reaction on sugar nucleotides. 2 One of these enzymes is UDP-glucuronic acid decarboxylase, which catalyzes the conversion of UDP-D-glucuronic acid (UDP-G1cA) to UDP-D-xylose (UDP-Xyl, Figure 1.28). UDP-Xyl is required for the biosynthesis of proteoglycans, molecules that are found in the extracellular matrix and on the cell surface of animal cells. It is also required in the biosynthesis of plant polysaccharides such as xyloglucan and xylan. 66 This enzyme is also a member of the SDR family and the reaction is initiated by an NAD dependent oxidation at the C-4” carbon to form a 13-ketoacid (Figure 1.28). Decarboxylation can readily occur to form an enol(ate) intermediate which is protonated at C-5” to give UDP-4-keto-xylose. UDP-Xyl is finally formed by the delivery of hydride from NADH back to the C-4” ketone. The incubation of UDP-[4H]-glucuronic acid with UDP-glucuronic 3 acid decarboxylase, isolated from Cryptococcus laurentii, demonstrated that tritium label was retained in the product during the decarboxylation reaction. 68 Later, a primary kinetic isotope ’ 67 effect was observed with UDP-[4”H]-glucuronic acid, but not the with C-3” or C-5” labeled 3 substrates, indicating that the oxidation at C-4” is a rate limiting step. 69 It was also found that incubation of UDP-[5”Hj-glucuronic acid with the UDP-glucuronic acid decarboxylases 3 derived from C. laurentii and wheat germ led to the inversion of configuration at C-5” after  29  69 The other known UDP-glucuronic acid decarboxylase is involved in decarboxylation. biosynthesis of UDP-D-apiose (UDP-Api) and will be discussed later in this chapter. NAD  E1Z  NADH  E  BH  z  NADH  EZ HB  HO  OUDP UDP-glucuronic acid  HO  OUDP 13-keto acid intermediate  NAD  OUDP  NADH  HOO  OUDP UDP-xylose  OUDP UDP-4-ketoxylose  Figure 1.28 Proposed mechanism for the reaction catalyzed by UDP-glucuronic acid decarboxylase.  1.9 Biosynthesis of Higher Order Sugars The term “higher order sugars” is commonly used to refer to carbohydrate structures bearing more than six linear carbon atoms. While many higher order sugars are biosynthesized from “unactivated” precursors, others, such as sialic acids, are generated from already existing sugar nucleotides.  1.9.1  Sialic Acid Biosynthesis Sialic acids are nine carbon cL-keto acids that were first discovered in the 1930s. 70 Since  then major efforts have focused on identifying the structures and the functional roles that these carbohydrates play. With over 50 different derivatives of N-acetylneuraminic acid (NeuAc) and  2-keto-3-deoxy-D-glycero-D-galacto-2-nonulosonic acid (KDN, Figure 1.29), sialic acids 30  represent a large portion of carbohydrates found on the surface of eukaryotic cells. To date the known derivatives of sialic acid include modifications at the C-4, C-7, C-8 and C-9 carbons with  lactate, phosphate, sulphate, methoxy ethers and acetate groups. Molecules such as 5,7diacetamido-3 ,5,7,9-tetradeoxy-L-glycero-L-manno-nonulosonic acid (pseudaminic acid, Pse) and 2-keto-3-deoxy-D-manno-2-octulosonic acid (KDO) are examples of molecules that are structurally similar to sialic acids yet differ in stereoconfiguration, substitution patterns or carbon chain length. OH OH HO!I  Sialic Acid (NeuAc) OH  OH  OH  AcHN Pseudaminic acid (Pse)  OH  KDN  KDO  Figure 1.29 Sialic acid and sialic acid-related sugars. In mammals, sialic acids are typically found capping the non-reducing ends of many cellsurface glycan chains. These serve important functions in regulating cellular events, such as cellular recognition, adhesion processes and cell 71 development. 72 An interesting use of sialic ’ acid derivatives was reported during the investigation of red blood cells (erythrocytes). During their production, the cell surfaces are initially polysialylated. Over the span of the erythrocyte life time (120 days) the sialic acids are removed by sialidases or spontaneously in solution. Near  31  the end of the life cycle host macrophages recognize, bind and phagocytose the exposed erythrocytes. This demonstrates that sialylation of mammalian cells is necessary for the host immune system to recognize its own tissues. 72 Many viruses use sialic acids to facilitate an attachment to the host cell. For example, influenza viruses demonstrated almost mandatory dependence on the host cell surface sialic acids for infection. 70 The presence of sialic acids in bacterial cells also has been reported. The display of sialic acid on the surface of pathogenic bacteria is believed to lead to evasion and protection from the host immune system through molecular mimicry of host sialylated cells. Neisseria meningitidis 73 and E. coil K i 74 are responsible for causing meningitis in humans. Closer inspection of their cell surfaces have shown the presence of an c-(2---*8)-linked N-acetylneuraminic acid polymer similar to the one found on the surfaces of mammalian cells. Some of the enzymes discussed earlier in this chapter are involved in the biosynthesis of NeuAc. Since NeuAc is found in both mammals and bacteria, both biosynthetic pathways will be presented (Figure 1  32  OH HO HO AcHN  OUDP UDP-G1CNAc  UDP-GIcNAc 2-Epimerase NeuC  ManNAc Kinase  NHAc HO_-(Q HO  HDp  NHAc HO—j  OH  ATP  ManNAc  BACTERIA PEP  ManNAc 6-P  MAIVIMALS  NeuNAc Synthase NeuB  PEP  NeuNAc 9-P Synthase NeuB  OH_ HOjj  OH  COOH CMP-NeuNAc Synthetase NeuA  NeuNAc 9-P Phosphatase NeuNAc  NeuNAc 9-P  CTP  HO  OH  OCMP  HOOH CMP-NeuNAc  Figure 1.30 Biosynthesis of CMP-NeuAc in mammals and bacteria. In mammals, the biosynthesis of NeuAc is initiated by the bifunctional enzyme, UDP-N acetylgiucosamine 2-epimerase/N-acetylmannosamine kinase (Figure 1.30, outer pathway). 49 This enzyme first acts as a hydrolyzing 2-epimerase that catalyzes an inversion at C-2” and the release of UDP from UDP-N-acetylglucosamine (UDP-G1cNAc) to form N-acetylmannosamine (ManNAc). It then phosphorylates ManNAc at C-6, producing N-acetylmannosamine 6phosphate (ManNAc 6-P). The following enzyme in the pathway, NeuAc 9-P synthase, introduces the three additional carbons by way of a reaction between phosphoenolpyruvate (PEP) and ManNAc 6-P. This forms N-acetylneuraminic acid 9-phosphate (NeuAc 9-P), the 9-carbon precursor of NeuAc. The mechanism of this reaction will be presented in Chapter 2. Free NeuAc  33  is formed by dephosphorylation at C-9 via N-acetylneuraminic acid phosphatase (NeuAc 9-P phosphatase). While many other monosaccharides are activated as uridine or guanine diphosphates (UDP-Glc, UDP-Gal and GDP-Man), sialic acids are activated as cytidine mononucleotides. This is accomplished by CMP-NeuAc synthetase that converts NeuAc into CMP-N-acetylneuraminic acid (CMP-NeuAc) using CTP. In bacteria, the biosynthetic pathway for sialic acid differs from the mammalian version (Figure 1.30, inner pathway). 49 UDP-G1cNAc is first converted to ManNAc by the action of a hydrolyzing 2-epimerase (NeuC). The formation of ManNAc 6-phosphate is not observed. Instead, the NeuAc synthase enzyme (NeuB) catalyzes the direct reaction between PEP and unphosphorylated ManNAc to form NeuAc. Lastly, the bacterial CMP-NeuAc synthetase (NeuA), catalyzes the coupling of NeuAc and CTP to form CMP-NeuAc. In some ways the biosynthesis of NeuAc in bacteria is more efficient, as it avoids the production of extra intermediates. 1.9.2  Pseudaminic Acid Biosynthesis Pseudaminic acid (Pse) was initially discovered in Pseudomonas aeruginosa as a  modified version of 2-keto-3-deoxy-nonulosonic acid. 75 The actual structure was determined to be 5,7-diacetamido-3 ,5 ,7,9-tetradeoxy-L-glycero-L-manno-nonulosonic acid (Figure 1.29). More recently this sugar was discovered on flagella of the pathogenic Gram-negative bacteria Campylobacter jejuni and Helicobacter pylon. C. jejuni, commonly found in animal feces, is a major cause of gastroenteritis in North America, whereas H pylon has been reported to cause peptic ulcers and gastric cancer in some cases. ’ 76  77  Pseudaminic acid is found as an 0-linked  post-translational modification of the flagellar proteins of the bacteria. Mutant strains unable to  34  biosynthesize Pse have been demonstrated to be aflagellate, non-motile and non-pathogenic. Therefore, the presence of the Pse is absolutely required for the proper assembly of flagella. 78 The biosynthetic pathway of Pse has now been fully elucidated (Figure 1.3 1)’  80  The  first step of the biosynthesis is catalyzed by the dehydratase PseB, which forms UDP-2diacetamido-2,6-dideoxy-3-L-arabino-hex-4-ulose (UDP-6-deoxy-4-keto-L-IdoNAc) from UDP G1cNAc. This enzyme is a sugar nucleotide 4,6-dehydratase, yet it is unique among this family of enzymes in that it also promotes an inversion of configuration at C-5”. In the second step, a PLP-dependent aminotransferase PseC introduces an amino group at C-4” to form UDP-4amino-2,4,6-trideoxy-AltNAc. An acetyl group is introduced in the next step by the acetyltransferase enzyme PseH to give UDP-6-deoxy-AltdiNAc. The acetyltransferase enzyme employs acetyl coenzyme A (acetyl-C0A) as the acetyl source. Acetyl-CoA contains a primary thiol group which is attached to the acetyl group being transferred via a thioester linkage. The next step in the biosynthesis of Pse is the hydrolysis of the UDP linkage by PseG, which forms 2,4-diacetamido-2,4-6-trideoxy-L-altrose (6-deoxy-AltdiNAc). The additional three carbons are finally introduced by a reaction between PEP and 6-deoxy-AltdiNAc, which is catalyzed by Pse synthase (Psel). Just like with sialic acid, the activated version of Pse is formed by the action of a CTP-dependent synthetase (PseF) which gives cytidine monophosphate pseudaminic acid (CMP Pse).  35  PLP-amino transferase PseC  UDP-GIcNAc dehydratase PseB  HO  2 NH HO  L-G1u a-ketoglutorate  OUDP  UDP-4-amino-6-deoxy A1tNAc  UDP-6-deoxy-4-keto IdoNAc  UDP-GIcNAc  H \ 3 AcFIN  AcCoA  QH  OH  -7-—CO 2 C 3 H 0  Pse Synthase Psel  NHAc  NHAc  HO-b. D 3 H ‘OH AcHN  Pseudaminic Acid  6-deoxy-AltdiNAc  AcFIN  PEP  PseG hydrolase  (N UDP  0 2 H  Ace1ansferase PseH  NHAc HO..SL C \ 3 H 0 AcHN UDP-6-deoxy-AltdiNAc  CMP-Pse synthetase \seF CTP  ocw  OH  L COj O 7 C 3 H  OH NHAc CMP-Pse  Figure 1.31 Biosynthesis of CMP-pseudaminic acid. 1.9.3  Legionaminic Acid Biosynthesis Legionaminic acid (5 ,7-diamino-3 ,5 ,7,9-tetradeoxy-D-glycero-D-galacto-non-2-ulosonic  acid or Leg) is a nine-carbon ct-keto acid that is related in structure to sialic acid (NeuAc) (Figure 1.32). It was first discovered in the lipopolysaccharide (LPS) of Legionella pneumophila, which  8183 The disease, is a facultative intracellular parasite that is the cause of Legionnaires’ disease. which can cause a fatal pneumonia, was first recognized in a large outbreak at an American Legion convention in Philadelphia in 1976. After the initial discovery of legionaminic acid, the 36  incorrect L-glycero-D-galacto configuration was assigned to it. It was only after legionaminic acid was prepared synthetically that the correct configuration of D-glycero-D-galacto, which is the same as sialic acid, was determined. 84  LCOOH H7 C 3 N 2 H  OH  OH  Legionaminic Acid (Leg)  XO, N,N’-diacetyllegionaminic acid  (Leg5Ac7Ac) X=NH, 5-N-acetimidoyl-7-N-acetyllegionaminic acid (Leg5Am7Ac) OH  OH  Sialic Acid N-acetylneuraminic acid (NeuNAc) Figure 1.32 Structure of legionaminic acid derivatives and sialic acid. LPS is an immunogenic glycolipid that makes up the outer surface of the outer membrane 85 LPS consists of the three covalently linked domains: lipid A, the of Gram-negative bacteria. core region and the 0-chain polysaccharide (Figure 1.33). The 0-chain polysaccharide in the LPS of L. pneumophila is composed of a repeating homopolymer of x-(2----*4)-linked 5-Nacetimidoyl-7-N-acetyllegionaminic acid (Leg5Am7Ac, Figure  1 .32).86  87  The 0-chain  polysaccharide is a repeating oligosaccharide polymer of 1-40 units and its primary role appears to be protective.  37  Figure 1.33 Structure of L. pneumophila LPS, the major component of the outer surface of the outer membrane of Gram-negative bacteria. Recently, Leg5Am7Ac was discovered in Campylobacter coil where it was found as a posttranslational modification on the flagellar proteins. 88 Campyiobacter species require the flagella for motility and pathogenicity. It was determined that the flagellin proteins are heavily glycosylated with derivatives of legionaminic acid and pseudaminic acid. Without these posttranslational modifications, the bacteria were not able to assemble functional flagellae. Legionaminic acid derivatives have also been found in the 0-antigens of Pseudomonas fluorescens  89  Vibrio aiginolyticus  90  and several other bacteria.  None of the enzymes involved in the biosynthesis of legionaminic acid have yet been identified. However, it is reasonable to assume that the biosynthesis of legionaminic acid would involve similar transformations as those in the biosynthesis of sialic and pseudaminic acids. At the onset of this thesis work we proposed that the biosynthesis would begin with the action of a hydrolyzing  2-epimerase  that  would  convert  UDP-N,N’-diacetylbacillosamine  (UDP  Bac2,4diNAc) to 2,4-diacetamido-2,4,6-trideoxymannose (6-deoxyMandiNAc, Figure 1.34).  38  This would be distinct from a pathway in which epimerization and hydrolysis occurs in two separate steps. The subsequent reaction between 6-deoxyMandiNAc and PEP would be catalyzed by legionaminic acid synthase to produce N,N’-diacetyllegionaminic acid (Leg5Ac7Ac). Finally the c*-keto acid would be activated as CMP-Leg5Ac7Ac by the corresponding synthetase. Recently, a 30 kb lipopolysaccharide locus was identified in the L. pneumophila genome which appears to contain the genes for the biosynthesis of N,N’-diacetyllegionaminic acid. ’ Based on 9 the homology of the encoded proteins to those involved in bacterial sialic acid biosynthesis (NeuA-C), tentative functions can be assigned to several of these genes. C 3 H AcHN  -  OUDP  H c NHAc AcHN—IO  hydrolyzing 2-epimerase 0 2 H  UDP 6-deoxy-MandiNAc  UDP-N,N ‘-diacetylbacillosamine (UDP-Bac2,4diNAc)  PEP  eg c synthetase -  Leg5Ac7Ac synthase  c  Cdp7CO 3 H 2 AcHN  OH  1 PP  CTP  CMP-Leg5Ac7Ac  AcHN  OH  N,N ‘-diacetyliegionaminic acid  Figure 1.34 Proposed biosynthesis of CMP-Leg5Ac7Ac.  The proposed starting material UDP-Bac2,4diNAc has been identified in Campylobacter jejuni where it is used in the N-linked protein glycosylation system. 9294 It was found that the N39  linked glycans play a major role in host adherence, invasion and colonization. The biosynthesis of UDP-Bac2,4diNAc in C. jejuni has been fully characterized, and the pathway involves three enzymes that require UDP-G1cNAc as the starting material (Figure 1.35). The first enzyme is UDP-G1cNAc 4,6-dehydratase (Pg1F) which catalyzes the formation of UDP-6-deoxy-4-ketoG1cNAc. The product is then subjected to the reactions catalyzed by PLP-dependent aminotransferase (PglE) and acetyltransferase (Pg1D) in order to generate the product UDP Bac2,4diNAc. Genes encoding for homologs of the enzymes Pg1D-F also found in the genome of L. pneumophila and, as a result, it is reasonable to assume that UDP-Bac2,4diNAc is used as the precursor for legionaminic acid by this organism. 95 UDP-G1cNAc dehydratase  UDP-G1cNAc  UDP-6-deoxy-4-keto-G1cNAc  amino transferase Pg1E  Acetyl transferase Pg1D AcCoA UDP-Bac2,4diNAc  ,-  LG1u ct-ketoglutarate  C 3 H AcFIN  O1HP  UDP-6-deoxy4-amino-G1cNAc  Figure 1.35 Biosynthesis of UDP-Bac2,4diNAc in Campylobacterjejuni.  40  1.10 Branched-Chain Sugar Nucleotides The branched—chain sugars can be divided into two groups based on their biogenesis. Group I consists of sugars bearing methyl or two-carbon side chains, while Group II consists of hydroxymethyl-branched and formyl-branched sugars. 96 Group I sugars are generally formed by the coupling reaction between a nucleotide sugar and a one- or two-carbon unit from the appropriate donors. Both the hydroxymethyl-branched and formyl-branched sugars are generally formed by a rearrangement of the sugar chain so that one of the carbon atoms is reoriented to form the branch. 97 1.10.1  Methylation Reaction in the Biosynthesis of Mycarose. Streptomyces fradiae has been reported to produce the macrolide antibiotic tylosin. 98 L  Mycarose (Figure 1.36), as well as D-mycaminose and D-mycinose sugars, form an important part of the antibiotic structure. The genes responsible for the biosynthesis of the Group I sugar, L-mycarose, have recently been identified. One of the genes encodes the Ty1C3 enzyme which is believed to be responsible for the C-methyltransferase reaction. The enzyme uses S adenosylmethionine (SAM) as the source of the methyl carbon and delivers it onto the C-3” position of dTDP-2,6-dideoxy-4-ketoaltrose. The reaction is believed to be initiated by a deprotonation at the C-3” position. The enolate intermediate then attacks the methyl group of SAM. The overall reaction occurs with the inversion of configuration at C-3” (Figure 1 .36).98 99  41  Ty1C3 HO (OTDP dTDP-2,6-dideoxy4-ketoaltrose  3 CH  Enolate intermediate Ty1K TyIC2 Ty1C4  Met®Ado Ado=Adenosyl  Hc7N OR 3 CH L-mycarose  Figure 1.36 Structure of L-mycarose and proposed mechanism of the reaction catalyzed by Ty1C 3.  1.10.2  Biosynthesis of UDP-D-Apiose. UDP-D-apiose is the activated form of D-apiose and it belongs to Group II of the  branched—sugar nucleotides (Figure 1.37). Apiose was initially discovered in 1901 in the flavone glycoside apiin from parsley.’°° The actual structure of apiose as 3-C-hydroxymethyl-Derythrose was elucidated much later.’°’ D-Apiose is the only branched carbon sugar that is found 02 In higher plants, D-apiose is only present in the in the plant cell wall as a monosaccharide.’ pectic polysaccharide rhamnogalacturonan-Il (RG-II), which is a component of the primary cell wall. RG-II has a structural function, but nowadays is also considered to be a highly complex 3 RG-II is believed to be the only polysaccharide associated with the primary cell organelle.’° wall which contains boron, which is an important micronutrient for higher  The two  vicinal hydroxyl groups of the furanose ring of D-apiose are ideally positioned to form a cyclic  42  diester with borate (Figure 1.37). This linkage serves as an attachment point for two highly complex side chains to the homogalacturonan backbone. The D-apiose sugars are linked to the rest of the homogalacturonan backbone of RG-II via D-galacturonate residues.’° 5 Even small structural changes in the RG-II structure, such as replacing the L-fucose residue with structurally similar L-galactose, reduced the formation of borate cross-links. This led to development of dwarf plants and abnormal leaf growth in Arabidopsis thaliana, indicating that plant growth heavily depends on wall pectic polysaccharide formation.’° 6 Homogalacturonan Chain 1  D-Galacturonate  D-Apiose  D-Apiose  D-Galacturonate  Homogalâcturonan Chain 2 Figure 1.37 D-Apiose involvement in formation of borate cross-link between two different homogalacturonan chains. Biochemical studies on the synthesis of D-apiose initially focused on the enzyme from parsley (Petroselinum hortense).’° 7 It has been established that UDP-D-apiose is produced from UDP-D-glucuronic acid (UDP-GlcA) by an NAD-dependent UDP-D-apiose/UDP-D-xylose 43  synthase (AXS 1). 108 Initially it was believed that there were two separate enzymes responsible for production of each UDP-sugar. However, despite repeated improvements in purification procedures of the enzymes from the cell cultures of parsley, the enzymatic activity for apiose and xylose synthesis could not be separated.’° 9 This is a bifunctional enzyme that catalyzes the conversion of UDP-glucuronic acid into the mixture of UDP-D-apiose and UDP-D-xylose (Figure 1.38). UDP-D-apiose/UDP-D-xylose synthase belongs to the SDR superfamily of enzymes.  OUDP UDP-xylose  AXS1 +  HOHO.  NAD  UDP-glucuronic acid OHOH UDP-apiose Figure 1.38 The UDP-D-apiose!UDP-D-xylose synthase reaction. The detailed mechanism of this conversion still remains unclear, however previous investigations into the mechanism of UDP-D-apiose/UDP-D-xylose synthase from both duckweed (Lemna minor) and parsley (Petroselinum hortense) have established that UDP-D apiose is formed from UDP-D-glucuronic acid (UDP-GlcA).’° ’ 9  110  This occurs in a single  enzymatic reaction which leads to the decarboxylation of the substrate, followed by re arrangement of the carbon skeleton and ring contraction. The initial steps of the reaction are believed to be the same as those catalyzed by UDP-glucuronic acid decarboxylase (Section 1.8), 44  which begins with NAD-dependent oxidation of C-4” followed by decarboxylation (Figure 1.39). It is believed that the UDP-4-ketoxylose intermediate can either be converted to UDP-D xylose (UDP-Xyl) or UDP-D-apiose (UDP-Api). A reduction at C-4” by NADH would lead to the formation of IJDP-Xyl, whereas rearrangement followed by a reduction would give UDP Api.”  112  The mechanism for the rearrangement that produces the D-apiose skeleton is still  unknown, however, it could involve a retro-aldol cleavage between C-2” and C-3” to give an open chain enol(ate) intermediate, followed by aldol condensation between C-2” and C-4” to form a contracted ring structure (Figure 1.39))’ This transformation is very similar to the 3 mechanism proposed in the biosynthesis of streptose.” ’ 4  “  Finally, UDP-Api is formed by a  reduction of the aldehyde at C-3” by NADH.  45  NAD HO-\ NAD 6  4  ///  NADH  NADH  OUDP UDP-Xyl  -  COO 5O  HO’—Q  HO—. OIJDP UDP-glucuronic acid  OUDP UDP-4-ketoglucuronic acid  UDP-4-ketoxylose  H NADH  NAD  NADH 3  CO 7 HOH &JDP ‘r—( OHOH UDP-Api  HO  OHCO  4  B UDP-3-aldehydo-apiose intermediate  Open chain enol(ate) intermediate  Figure 1.39 Retro-aldol mechanism for the UDP-D-apiose/UDP-D-xylose synthase reaction.  Previous biochemical studies support this mechanism. When the shoots of young parsley C]-G1cA, it was discovered that the C-3” of apiose must 4 plants were fed with UDP-[3,4- ‘ originate from either C-3” or C-4” of glucose.” 6 Later, the incubation of UDP-[3”-’ C1-GlcA 4 with UDP-D-apiose/UDP-D-xylose synthase from cell-suspension cultures of parsley demonstrated that the branch hydroxymethyl carbon (C-3”) of D-apiose comes from the C-3” of the UDP-G1cA (Figure 1.40).107  46  COOHO-ç----\--Q\ c’ 14 HO— OUDP  AXS1  UDP-{3 “C]-G1cA 14  o  14 2 H0H C  C]-apiose 4 UDP-{3”’-’  Figure 1.40 Fate of the labeled C-3”0fUDP-[3”-’ C]-glucuronic acid during the reaction 4 catalyzed by UDP-D-apiose/UDP-D-xylose synthase. Subsequent studies focused on determining the fate of the C-4” hydrogen. UDP-[UC, 414 Hj-G1cA was incubated with UDP-D-apiose/UDP-D-xylose synthase isolated from Lemna minor 3 (Figure 1.41).” The result of the experiment demonstrated that the tritium label from C-4” of the starting material was exclusively found at the branch carbon of the UDP-Api. This result can be explained by a hydride transfer process involving an NAD cofactor. The tritium label is initially transferred to the enzyme-bound NAD cofactor to form NAD H and UDP-43 ketoglucuronic acid. The enzyme-bound cofactor would then transfer the label to the branch carbon after the carbon rearrangement had occurred.  H 3  I  coo  HOQ OUDP C, 44 UDP-[U-’ H]-G1cA 3  AXS1  HC 3 HOH  0  1?OUDP OH OH Hjapiose 3 C, 3’4 UDP-[U-’  H of UDP-{U-’ 3 C, 44 H]-G1cA during the reaction 3 Figure 1.41 Fate of the labeled 4”catalyzed by UDP-D-apiose/UDP-D-xylose synthase  In order to study the stereochemistry involved in the reduction of the C-3” aldehyde of the UDP-3-aldehydo-apiose, the enzyme was incubated with UDP-[4”H]-G1cA. The resulting 3 47  mixture was first treated with acid and then sodium borohydride to form apiitol (Figure 1.42). 112 The treatment of apiitol with sodium periodate gave glycolic acid. It was then oxidized by with glycolate oxidase, which is specific for the pro-R hydrogen.’ 8 The final analysis of the product from the enzymatic transformation demonstrated that the tritium label from UDP-[4”H]-G1cA 3 is transferred only to the pro-R position on the C-3” methylene of apiose. Similar experiments were performed with UDP-[5”H]-GIcA as a substrate for UDP-D-apiose/UDP-.D-xylose 3 synthase to show that the enzymatic transformation occurs with an inversion of configuration at C-4” of UDP-Api.” 2 HQ H 3 H—c— HO\_0\  4 2)NaBH  HD-OH  OUDP OH OH  HO— ‘-“I’ OUDP  C 2 HOH  H]-apiose 3 UDP_[3”L  H]-G1cA 3 UDP-[4-  -C—OH HHOH 3 C  H]apiitol 3 [1 / 4 NaIO  H 2 CO HO 3 ‘H  +  I  CO H  Glycolic oxidase .  H 2 O  /  H—OH ii  H]Glycolic acid 3 [2-  Figure 1.42 Determination of stereochemistry in the conversion of UDP-[4”H]-G1cA to 3 UDP-[3” H]-apiose. 3  Finally, to prove the existence of the UDP-4-keto-xylose intermediate, the enzymatic C]-G1cA and UDP-D-apiose/UDP-D-xylose synthase was quenched 4 reaction between UDP-[IJ-’  48  . It was shown that after extensive purification over 90% of the tritium 4 H 3 by the addition of NaB label was found at C-4” of UDP-Xyl.” 2 While the mechanism shown in Figure 1.39 is consistent with all of these observations, an alternative mechanism has also been proposed (Figure 1.43).” Following the initial oxidation at C-4” and decarboxylation at C-5”, a ring contraction occurs via deprotonation of the C-3” hydroxyl group and an alkyl migration to form a UDP-3-aldehydo-apiose intermediate. This is then reduced by the NADH cofactor to form UDP-Api. Similar rearrangements can be found in the non-mevalonate pathway for the biosynthesis of isoprene units 119 and branched chain amino ° 2 acids)  49  NAD HO0  NADH  NAD  ENZ  6  ]H 2 CO  -  0 C00  HO H0’  HO  OUDP UDP-glucuronic acid  //  NADH  OUDP UDP-xylose  OUDP DP4keto glucuronic acid  NAD C 2 HOH  H OH UDP-apiose  oui ENZ  UDP-4-ketoxylose  NADH 0HCO  OHOH TJDP-3-aldehydo-apiose  Figure 1.43 Carbon migration mechanism for the UDP-D-apiose/UDP-D-xylose synthase reaction.  50  1.11 Project Goals The aim of this thesis is to study two different biosynthetic pathways which are involved in the production of two unique sugar nucleotides. My studies elucidating the biosynthetic pathway of CMP-N,N’-diacetyllegionaminic acid are presented in Chapter 2. Mechanistic studies on the enzyme responsible for the formation of UDP-D-apiose are the focus of Chapter 3. Chapter 2 will focus on identifying the biosynthetic pathway of CMP-N,N’ diacetyllegionaminic acid for the first time. It is believed that the biosynthesis of legionaminic acid will involve similar transformations as those in the biosynthesis of sialic acid (Section 1.9). The first goal of Chapter 2 is to identif’ and study the first enzyme of the biosynthetic pathway. If the first enzyme is indeed a hydrolyzing 2-epimerase, then the mechanism of the enzyme can be investigated using isotopic incorporation experiments. The second enzyme is proposed to be N,N’-diacetyllegionaminic acid synthase (NeuB), a potential PEP-condensing synthase homologous to the NeuAc synthase. The second goal of Chapter 2 is to identifr the synthase and analyze its mechanism using isotopically labeled PEP. Finally, the last part of Chapter 2 will focus on identifying the synthetase that generates CMP-N,N’-diacetyllegionaminic acid. Since N,N’-diacetyllegionaminic acid is found on the flagellum of Campylobacter coli, 88 and the presence of the acid is essential for motility and pathogenicity of the organism, elucidation of the biosynthetic pathway could provide the basis for the development of new antibiotic drugs. UDP-Api biosynthesis is the main focus of Chapter 3. It is believed that UDP-Api is biosynthesized from UDP-glucuronic acid (Section 1.10). Just recently, the candidate gene for UDP-D-apiose/UDP-D-xylose synthase was identified in Arabidopsis thaliana and was functionally expressed in Escherichia coli.’ ’ The first goal of the project was to establish the 1  51  activity of the enzyme using NMR spectroscopy and to characterize the expected branched chain product. The catalytic competency of the potential UDP-4-ketoxylose intermediate to produce both UDP-Xyl and UDP-Api will be tested. The second goal of the chapter is to provide evidence for one of the two potential mechanisms that could be used by UDP-D-apiose/UDP-D xylose synthase to perform carbon skeleton rearrangements. Several substrate analogs, such as UDP-2-deoxy-2-fluoro-D-glucuronic acid, UDP-3-deoxy-3 -fluoro-D-glucuronic acid and UDP CJ-3-deoxy-3-fluoro-D-glucuronic acid, were prepared and tested with the enzyme in order 3 [U-’ to help elucidate the mechanism.  52  Chapter Two Biosynthesis of CMP-N,N’-Diacetyllegionaminic Acid  53  2.1 Introduction The focus of this chapter will be on the identification of the three enzymes involved in the biosynthesis of CMP-N,N’-diacetyllegionaminic acid (CMP-Leg5Ac7Ac) and the elucidation of their chemical mechanisms. The inability to produce the proposed starting material, UDP-N,N’ diacetylbacillosamine (UDP-Bac2,4diNAc), hampered any previous investigation on the biosynthetic pathway. As mentioned in Chapter 1, the recent discovery of the enzymes involved in the biosynthesis of UDP-Bac2,4diNAc in C. jejuni allows for the efficient preparation of the desired starting material from UDP-G1cNAc. 92 The preparation of UDP-Bac2,4diNAc will be discussed in the first part of this chapter. The chapter will then focus on the identification of, and mechanistic studies on, a hydrolyzing 2-epimerase similar to that involved in the biosynthesis of sialic acid. Experiments will be presented that show the enzyme converts UDP-Bac2,4diNAc into 2,4-diacetamido-2,4,6-trideoxymannose (6-deoxyMandiNAc) and will demonstrate that the elimination of UDP occurs through a C-O bond cleavage process. It will also be demonstrated that the epimerase reaction proceeds with a net retention of configuration at C-i. The second enzyme to be identified in this chapter is N,N’-diacetyllegionaminic acid synthase (NeuB) that catalyzes the formation of the N,N’-diacetyllegionaminic acid (Leg5Ac7Ac)  and phosphate  from 6-deoxyMandiNAc  and PEP.  The  isolation and  characterization of the enzymatically produced acid confirms its stereoconfiguration as being identical to previously chemically synthesized derivatives. It will also be shown that the reaction catalyzed by N,N’-diacetyllegionaminic acid synthase proceeds via a C-O bond cleavage mechanism.  The  final  section  of this  chapter  will  demonstrate  that  CMP-N,N’  diacetyllegionaminic acid is biosynthesized from Leg5Ac7Ac by the action of CMP-Leg5Ac7Ac synthetase. 54  2.2 Preparation of UDP-Bac2,4diNAc. In order to test the hypothesis that Leg5Ac7Ac was biosynthesized from UDP Bac2,4diNAc, it was first necessary to prepare this starting material. This was accomplished in a three-step chemoenzymatic synthesis (Figure 2.1) that utilizes two recombinant C. jejuni enzymes (Pg1F and Pg1E). The acetyltransferase (Pg1D) was also prepared but not used in the synthesis due to the efficient chemical acetylation of primary amines using acetic anhydride.  PgIF 1120 UDP-G1cNAc  OUDP  TJDP-6-deoxy-4-keto-G1cNAc  ,—  L-Glu  Pg1E  ci-ketogIutarate  HC  O, pyr 2 Ac  Ad11N&—0  HO AcFIN  HC H N 2 -&_-R HO  OUDP  UDP-Bac2,4diNAc  AcHN  OUDP  UDP-6-deoxy4-amino-G1cNAc  Figure 2.1 Chemoenzymatic synthesis of UDP-Bac2,4diNAc  55  2.2.1  Expression and Purification of Pg1F, Pg1E and PglD enzymes The plasmids used in the over-expression of Pg1F (Cj 11 20c), Pg1E (Cj 1121 c) and Pg1D  (Cj 1 123c) were donated by Dr. Ian C. Schoenhofen. E. coil cells, transformed with appropriate plasmids, were grown in the presence of ampicillin. Over-expression was induced by the addition of IPTG (isopropyl f3-D-1-thiogalactopyranoside) due to the presence of a iac operon upstream from pgiF, pgiF or pgiD in the corresponding plasmids. Normally, in the absence of lactose or IPTG, a repressor binds to the iac operon, preventing the T7 RNA polymerase from binding to a T7 promoter region on the plasmid. IPTG binds to the repressor, causing the repressor to release from the iac operon, and allowing T7 RNA polymerase to bind and express the gene. Cells were lysed and the soluble fraction was loaded onto an affinity chromatography column containing immobilized nickel (Ni ). The N-terminal hexahistidine tags on the enzymes 2 bind with Ni 2 allowing the cell lysate to be washed from the column. The tagged enzyme was eluted with 500 mM imidazole and the appropriate fractions were dialyzed against a phosphate buffer before being flash-frozen with liquid nitrogen in the presence of 10% glycerol. The enzymes can be stored for up to nine months without significant loss of activity. The purity of the resulting proteins was determined to be >90% by SDS-PAGE analysis (Figure 2.2).  56  1  2  3  4  66kDa  29kDa  •  Figure 2.2 SDS-PAGE gel of enzymes involved in the biosynthesis of UDP-Bac2,4diNAc. Lane 1: molecular weight standards; lane 2: purified PgIF (Cj 1 120c); lane 3: purified Pg1D (Cj 1 123c); lane 4: purified Pg1E (Cj 1 121c). Molecular weight standards BSA (29 kDa) and carbonic anhydrase (66 kDa).  2.2.2  Chemoenzymatic synthesis of UDP-Bac2,4diNAc The first step in the synthesis of UDP-Bac2,4diNAc involved treating UDP-GlcNAc with  Pg1F. The progress of the reaction was followed by monitoring the appearance of the UDP-6deoxy-4-keto-GlcNAc using ESI-mass spectrometry (m/z 587 {M-Hfl. Once the reaction was completed, the resulting mixture was used for the next transformation without further purification. Since Pg1E is a PLP-dependent aminotransferase enzyme, L-glutamate and PLP were added to the reaction mixture. The second reaction was initiated by adding Pg1E and after several hours of incubation it was determined that >95% of the UDP-6-deoxy-4-keto-G1cNAc was converted to UDP-6-deoxy-4-amino-G1cNAc (m/z 589 [M-Hfl. Both enzymes were removed by centrifugal ultrafiltration and the resulting sugar was purified by ion-exchange chromatography.  57  In the biosynthetic pathway, the transformation of UDP-Bac2NAc to UDP-Bac2,4diNAc is catalyzed by the acetyltransferase enzyme Pg1D in the presence of acetyl-CoA (Figure 1.35). This cofactor is used by a variety of acyltranferase enzymes to transfer an acetyl group to various acceptors including amines, alcohols, carbon nucleophiles and other thiol groups. During the preparation of UDP-Bac2,4diNAc the acetyltransferase enzyme Pg1D was available to me, however acetylation of the 4-amino group was performed chemically. Chemical acetylation is a high yielding process which does not require the use of expensive starting reagents like acetyl CoA. The acetylation reaction was initiated by the addition of acetic anhydride to a solution of UDP-6-deoxy-4-amino-GlcNAc and methanol. Analysis by negative ESI-masspectromety showed that after 24 hours all of the starting material was converted to UDP-Bac2,4diNAc (m/z 631 [M-Hfl. The UDP-Bac2,4diNAc was purified by ion exchange chromatography, and it was characterized by 1 H NMR spectroscopy (Figure 2.3). The observed NMR chemical shifts are in agreement with previously published results. 92  58  ’ 2 H  C 3 H AcHN —‘------  HO  4 H  3 H  R *  AcNN OUDP  4 H  4.50  4.00  ppm  *  NAc  6 H  I  I  8.0  I  I  I  7.0  I  I  I  6.0  I  I  I  5.0  I  ppm  I  I  4.0  I  I  I  I  3.0  I  I  2.0  L I  I  1.0  0, 25 °C).*= triethylammonium 2 Figure 2.3 1 H NMR spectrum of UDP-Bac2,4diNAc (400 MHz, D 59  2.3 Identification and Mechanistic Studies on the Hydrolyzing UDP-Bac2,4diNAc 2-Epimerase from Legionellapneumophila Previous studies on LPS biosynthesis in L. pneumophila had identified a neuC homolog that encodes an enzyme required in the biosynthesis of legionaminic acid. ’ The protein product 9 of the gene shares 30% and 28% sequence identity with the NeuC responsible for sialic acid biosynthesis in Neisseria meningitidis and E. coli K1, respectively. Previous work on the bacterial biosynthesis of sialic acid in Neisseria meningitidis and E. coli Ki indicated that NeuC is a hydrolyzing UDP-N-acetylglucosamine 2-epimerase. ’ 50  121  This result suggests that the L.  pneumophila NeuC could also be a hydrolyzing 2-epimerase that uses UDP-Bac2,4diNAc as a potential substrate. 2.3.1  Expression and Purification of the Hydrolyzing UDP-Bac2,4diNAc 2-epimerase An expression plasmid for the neuC gene of L. pneumophila (1pg0753) was prepared by  our collaborators in Dr. Martin Young’s laboratory at NRC. The recombinant plasmid was overexpressed by induction with 1.0 mM IPTG. The resulting cells were harvested by centrifugation and stored in pellet form at -80 °C. Following cell lysis, the crude protein extract was loaded onto an affinity chromatography column containing immobilized nickel (Ni ). The enzyme was 2 eluted using a 500 mM imidazole buffer. Following buffer exchange, the purified protein was either used directly or stored at -80 °C in a buffer containing 10% glycerol. The purity of the protein was determined to be greater than 90% by SDS-PAGE analysis (Figure 2.4).  60  1  2  3  66kDa  29kDa  Figure 2.4 SDS-PAGE gel showing NeuC purification. Lane 1: molecular weight standards; lane 2: crude cell lysate; lane 3: purified NeuC. Molecular weight standards BSA (29 kDa) and carbonic anhydrase (66 kDa). 2.3.2  Testing the Activity of the Hydrolyzing UDP-Bac2,4diNAc 2-epimerase To test the activity of the hydrolyzing UDP-Bac2,4diNAc 2-epimerase (NeuC), a sample  of UDP-Bac2,4diNAc was incubated with the enzyme and the reaction was monitored by 31 P NMR spectroscopy. Before the addition of NeuC, signals at -11.08 ppm and -12.89 ppm were observed that correspond to the diphosphate group of UDP-Bac2,4diNAc (Figure 2.5A). After the addition of enzyme and incubating for 15 minutes, a new set of signals appeared at -7.54 ppm and -10.32 ppm, corresponding to free UDP (Figure 2.5B). The sample was spiked with UDP to confirm the identity of the signals. This result indicated that a hydrolysis reaction was taking place. The progress of the reaction was also followed by both negative and positive ESI-mass spectrometry. It was found that all of the starting material UDP-Bac2,4diNAc (m/z 631 [M-Hf) was converted to UDP (m/z 403 [M-Hf) along with a compound whose mass corresponded to 2,4-diacetamido-2,4,6-trideoxymannose (6-deoxyMandiNAc, m/z 269 [M+Na]).  61  3 CH  AcHN -:--------Q HO91 9c AcHN o-P-o-P-oU I I 00  A  UDP-Bac2,4dlNAc  13-P  B  Free UDP  -5.0  -10.0  -15.0  ppm Figure 2.5 31 P NMR spectra monitoring the reaction of UDP-Bac2,4diNAc with NeuC. 0, 25 2 A) before the addition of NeuC and B) after the addition of NeuC (121.5 MHz, D °C). After the UDP was removed by ion exchange chromatography, the sugar product was determined to be a 1:1 mixture of anomers of 2,4-diacetamido-2,4,6-trideoxymannose (6deoxyMandiNAc) by 1 H NMR spectroscopy (Appendix Figure A.1). The small  JHI,H2  values  observed in both of the anomers (0.91 Hz and 1.47 Hz) indicated that the acetamido group at C-2 was in an axial position, and therefore an inversion of stereocenter had occurred at C-2. The ‘H 22 These NMR data obtained were in complete agreement with those reported in the literature.’  62  results confirm that the L. pneumophila NeuC homolog is a hydrolyzing UDP-Bac2,4diNAc 2epimerase. After demonstrating the activity of hydrolyzing UDP-Bac2,4diNAc 2—epimerase, the specificity of the enzyme was tested. The reaction catalyzed by the enzyme is very similar to that of the hydrolyzing UDP-G1cNAc 2-epimerase, so it is reasonable to assume that it is able to catalyze the conversion of UDP-G1cNAc to ManNAc and UDP. Therefore, a sample of UDP GlcNAc was incubated with hydrolyzing UDP-Bac2,4diNAc 2—epimerase at 37 °C for two days and the progress of the reaction was followed by 31 P NMR spectroscopy. No reaction was observed, indicating that hydrolyzing UDP-Bac2,4diNAc 2—epimerase is specific for UDP Bac2,4diNAc and is involved only in the biosynthesis of the Leg5Ac7Ac. 2.3.3  Kinetic Characterization of the Hydrolyzing UDP-Bac2,4diNAc 2-epimerase. A continuously coupled UDP assay was used to determine the kinetic parameters of the  UDP-Bac2,4diNAc 2-epimerase.’ 23 The assay involved the use of pyruvate kinase and lactate dehydrogenase enzymes (Figure 2.6). The first coupling enzyme, pyruvate kinase, takes phosphoenolpyruvate and UDP to form UTP and pyruvate. The second coupling enzyme reduces pyruvate to lactate while consuming one equivalent of NADH. This is monitored by UV spectroscopy as a decrease in absorbance at 340 nm  (  =  6220 M 1 cm’). The reaction followed  Michaelis—Menten kinetics and the kinetic parameters (Figure 2.7) obtained with UDP Bac2,4diNAc as a substrate were a kcat of 59.1  ±  1.6 1 s a KM of 36.5 ,  ±  3.2 jiM and kcat/KM of 1.6  x 10 M’ s (in 50 mM 4 PO pH 7.5, 10 mM MgC1 2 NaH , 2 mM PEP and 0.2 mM NADH). 2  63  hydrolyzing epimerase UDP-Bac2,4dlNAc  .  6-deoxy + MandiNAc UTP  cyoPo o 2 3  J). o2Cto  pyruvate kinase PEP  NADH  }  C yOR 2 _0  lactate CH dehydrogenase 3 pyruvate  CH 3 lactate  Figure 2.6 Continuous coupled UDP assay used to determine kinetic parameters of the hydrolyzing epimerase.  60  ::: 40 0  w 0  >  0 0  200 400 600 800 UDP-Bac2,4ctiNAc (jtM)  1000  Figure 2.7 Kinetic plots of initial velocity (vo)/[E]o vs. UDP-Bac2,4diNAc.  2.3.4  Stereochemical Analysis and Solvent Isotope Incorporation The first mechanistic study conducted on the NeuC enzyme was devised to determine the  first formed product of the enzymatic reaction. Since the produced sugar could readily undergo mutarotation, the enzymatic formation of one anomer of 6-deoxyMandiNAc would quickly lead 64  to a mixture of anomers. 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 NMR spectroscopy. A dilute buffer was employed to minimize the possibility of buffer-catalyzed mutarotation. Moreover, the above experiment will allow simultaneous monitoring for any incorporation of solvent-derived deuterium into a nonexchangeable position of the product. The enzymatic conversion of UDP-Bac2,4diNAc to UDP and 6-deoxyMandiNAc was monitored using ‘H NMR spectroscopy in a buffer prepared using D 0. The initial ‘H NMR 2 spectrum showed the anomeric proton signal of UDP-Bac2,4diNAc at 5.45 ppm in the absence of enzyme (Figure 2.8, t= 0 mm). This signal appears as a doublet of doublets due to coupling to both the 13-phosphorus atom and H-2”. A relatively large amount of enzyme was added to the sample and a new spectrum was recorded immediately to minimize the possibility of mutarotation. The NMR spectral time course revealed the appearance of signals corresponding to 6-deoxyMandiNAc and UDP, with full conversion to products after 120 minutes. The chemical shifts for the H-i” protons of the 6-deoxyMandiNAc anomers appeared at 5.06 ppm (downfield signal) and 4.90 ppm (upfield signal). After only two minutes, the spectrum showed a nonequilibrium mixture of anomers in 10:1 ratio favoring the anomer with the downfield signal (Figure 2.8, t= 2 mm). After twelve hours, the spectra showed full conversion to products and the non-enzymatic mutarotation had equilibrated the anomers at a 1:1 ratio (Figure 2.8, t 12 h). It is clear that the anomer displaying the downfield chemical shift (5.06 ppm) is the true product of the enzymatic reaction and the anomer displaying the upfield chemical shift (4.90 ppm) is formed by mutarotation.  65  A H C H AcHNI  Hydrolyzing Epimerase NeuC  AcHNI O-UDP  Non-enzymatic Mutarotati on  H NHAc C 3 AcHN1P  AcBN  0 2 D  DI 11  OH H]-a-6-deoxy2 [2MandiNAc  UDP-Bac2,4diNAc  H]-f3-6-deoxy2 [2MandiNAc  CH  B  AcHN  -  AcHN  t= 0 mill  O-UDP  H NHAc C 3 AcHNH  t 2 mm  NHAc AcHNl°QH  t= 12 h  H  5.50  5.40  5.30  I  I  I  5.20  5.10  5.00  4.90  ppm  Figure 2.8 1 H NMR spectra monitoring the reaction of UDP-Bac2,4diNAc with the NeuC in D 0. A) Enzymatic conversion of UDP-Bac2,4diNAc to [22 H]-6-deoxy-a-MandiNAc, 2 H]-6-deoxy-f3-MandiNAc and B) ‘H NMR spectra 2 followed by mutarotation to [2monitoring the enzymatic reaction (400 MHz, D 0, 25 °C). 2  In order to determine the identity of the first formed anomer, the anomeric signals of 6deoxyMandiNAc had to be assigned. For sugars with an axial proton at C-2, the H-i to H-2  66  coupling constant (4i, H2) values can serve this purpose  (JH1,  for the (x-anomer is larger than  the 13-anomer). The same principle cannot be applied to sugars with an equatorial proton at C-2 because the coupling constants are similar for both anomers. Bock and Pedersen reported that the C-i to H-i coupling constant (Jc,,  Hi)  values in a ‘H-coupled ‘ C NMR spectrum are 10 Hz 3  larger in the c-anomer than in the 13-anomer for a variety of D-monosaccharides. 124 I therefore ran a two-dimensional heteronuclear NMR experiment (HMQC) on a sample of 6deoxyMandiNAc and established that the first formed product has a C-i” ‘ C signal at 92.93 3 ppm and the second formed product had C-i” ‘ C signal at 92.86 ppm. Due to the small amounts 3 of product available, the  JCI,H1  could not be directly obtained from a ‘H-coupled ‘ C NMR 3  experiment. Instead, an equilibrated sample of 6-deoxyMandiNAc was used in an HMQC NMR experiment, with the ‘H-’ C coupling constant retained in the ‘H dimension (Figure 2.9). 3  67  ‘H spectrum HOD  AcRN1OH  HC NHAc AcH  f3-6-deoxyMandiNAc Ci,H1 163 Hz  ‘ H-coupled C 3 spectrum  H91.50 —92.00  92.86  92.93  Ij  L-93.00  ppm  -93.50 -94.00  c-6-deoxyMandiNAc CI,Hl 173 Hz  5.00  -94.50  4.50  ppm  Figure 2.9 HMQC experiment with 6-deoxyMandiNAc (400 MHz, D 0, 25 °C). 2  The results showed that the ‘ C NMR signal at 92.93 ppm exhibits a .JCI,H1 value of 173 Hz and 3 therefore is assigned as the ct-anomer, whereas the ‘ C NMR signal at 92.86 ppm exhibits a Jc 3 , 1 Hi  value of 163 Hz and is assigned to the 13-anomer. The observation that 6-deoxy-a-MandiNAc  is the first formed product indicates that the stereochemical course of the reaction proceeds with a net retention of configuration at C-i”.  68  This experiment also shows that the reaction proceeds with solvent isotope incorporation at C-2 and that the product formed is actually [2H]-6-deoxy-a-MandiNAc. The absence of an H-2 2 signal in the ‘H NMR spectrum of the product obtained from the reaction run in D 0 (not shown 2 in Figure 2.8) proved that deuterium incorporation occurred at C-2”. This was also shown by the appearance of the anomeric proton signals as singlets due to a negligible coupling constant to the C-2 deuterium atom and loss of coupling to the 3-phosphorus atom of UDP (Figure 2.8). Moreover, the isotope incorporation was further supported by positive ESI-mass spectrometric analysis of the produced 2 [2H ]-6-deoxyMandiNAc product. The sample prepared in deuterated water was one mass unit larger than the product produced in H 0. The incubation of UDP 2 Bac2,4diNAc or 6-deoxyMandiNAc in the same deuterated solvent without the presence of the NeuC enzyme did not lead to the incorporation of deuterium in either sugar. Earlier studies on other hydrolyzing 2-epimerases had also demonstrated the incorporation of solvent-derived deuterium label, indicating that deprotonation occurs at the C-2” position, followed by reprotonation with a solvent-derived deuterium atom. 50 2.3.5  Test for C-O vs. P-O Bond Cleavage The next mechanistic study was designed to test if the loss of UDP proceeds via a C-O or  P-O bond cleavage process. A C-O bond cleavage mechanism would involve the elimination of UDP followed by the attack of water at C-i” (Figure 2.10 A). A P-O bond cleavage mechanism would involve the nucleophilic attack of water at the phosphorus atom of the 3-phosphate and displacement of the sugar (Figure 2.10 B). A subsequent epimerization of the free sugar would give the product. If the incubation of the hydrolyzing 2-epimerase with UDP-Bac2,4diNAc in 18 led to the incorporation of the 18 2 H 0 0-label into 6-deoxyMandiNAc, the reaction would  69  involve the cleavage of the C-O bond. If the ‘ 0-label is found in the UDP, the reaction occurs 8 via a P-O bond cleavage mechanism (Figure 2.10).  A  ACHj  AcHtR  +  2 H  AcHN  AcHN0  +  +  up 0 18L  OH  o  o—P-oup  7’0  AcHN_jE0  -ii 0-P—OUMP 8 ‘  18 2 H 0  Figure 2.10 C-0 vs P-0 bond cleavage experiment. A) Products from C-0 bond cleavage mechanism. B) Products from P-0 bond cleavage mechanism.  The NeuC enzyme and UDP-Bac2,4diNAc were incubated in a phosphate buffer (pH 7.5) containing  50% 0. 18 The reaction was run until completion and the products were investigated 2 H  using negative (UDP) and positive (6-deoxyMandiNAc) ESI-mass spectrometry. The mass spectral analysis of the isolated UDP demonstrated the absence of the  180  label. However, the  resulting 6-deoxyMandiNAc showed a 1:1 ratio of m/z 269 [M+Na] and m/z 271 [M+Na+2j’, 0-label had been incorporated into the sugar (Figure 2.11). Since the 8 indicating that the ‘  180..  label would be incorporated at the anomeric position of 6-deoxyMandiNAc, it is possible that this label could non-enzymatically exchange with solvent, giving a false positive result. In order to confirm that the label was incorporated during the enzymatic reaction and not through non70  enzymatic wash in, a sample of unlabeled 6-deoxyMandiNAc was incubated in buffer prepared with ‘ 0-labeled water. The progress of the experiment was followed by positive ESI-mass 8 spectrometry and, while a slow incorporation of  180  into the sugar was observed, the rate was  orders of magnitude slower than observed in the enzymatic reaction. WCH 3’. I AcHN HO AdHN -  O-UDP  A  610  63L3  615  B  625 269.1  620  630  640  63$  645 m/  HC NHAc  269.1271.1  +  NHAc  C  AcNN+0\  HO  i 1 rTrrTr,  260  265  270  H  +  ,  275  +  Na  ,  280  285  mi:  Figure 2.11 ESI-mass spectra of C-0 vs. P-0 bond cleavage experiment with NeuC. A) Negative ESI-mass spectrum of UDP-Bac2,4diNAc before addition of NeuC. B) Positive ESIMS spectrum of product of the enzymatic reaction carried out in 100% H 0. C) 2 Positive ESIMS spectrum of products of the enzymatic reaction carried out in 50% H 0 2 and 50% 112180. Based on the fact that UDP-Bac2,4diNAc 2-epimerase catalyzes the conversion of UDP Bac2,4diNAc to 6-deoxyMandiNAc through the cleavage of both the C-2”—H bond and the C 1 “—0 bond, it is reasonable to assume that the reaction proceeds via a 6-deoxy-2,4-  71  diacetamidoglucal intermediate (Figure 2.12). This is very similar to the UDP-G1cNAc 2epimerase enzyme which is involved in the biosynthesis of sialic acid. ’ 49  125  It has also been  demonstrated that this hydrolyzing epimerase reaction proceeds with net retention of configuration at C-i. This means then that the enzyme mechanism is initiated by the antielimination of UDP, followed by the syn-hydration of 6-deoxy-2,4-diacetamidoglucal, forming 6-deoxy-ct-MandiNAc (Figure 2.12). The hydrolyzing UDP-GlcNAc 2-epimerase also catalyzes a net-retention of configuration at C-i and the second step involves the syn-addition of water. :B-ENZ anti-elimination  AcHN  syn-addition AcHN BH UDP+ ENZ  UDP-Bac2,4diNAc  2 OH  6-deoxy-2,4-diacetamidoglucal intermediate  H NAc C 3 ACHN1H OH  cL-6-deoxyMandiNAc  Figure 2.12 Mechanism of the reaction catalyzed by the hydrolyzing UDP-Bac2,4diNAc 2epimerase.  2.4 Identification and Mechanistic Studies on N,N’ Diacetyllegionaminic Acid Synthase (NeuB) Previous studies have identified a neuB homolog (lpg0752) in L. pneumophila that is clustered with the neuC and neuA homologs and was believed to encode an N,N’ ’ The protein is 61% identical in sequence to sialic acid 9 diacetyllegionaminic acid synthase. synthase from N. meningitidis.  72  2.4.1  Expression and Purification of N,N’-Diacetyllegionaminic Acid Synthase. The neuB gene was cloned by our collaborators in the laboratory of Dr. Martin Young at  NRC. Recombinant plasmids containing the neuB gene were transformed into E. coil and overexpressed to give both N-terminal and C-terminal His-tagged proteins. The resulting proteins were purified using a nickel column, and then subjected to a buffer exchange using centrifugal protein filters. In both cases, high levels of over-expression were achieved; however most of the resulting protein was insoluble and remained in the cell pellet during purification. As a result, only a very small amount of soluble protein (4 tg/L) could be prepared in this way. It is likely that during over-expression most of the protein misfolded creating inclusion bodies. Factors such as the temperature of induction, the pH of the cultivation medium or the changes in amino acid sequence due to addition of His-tag region all could have had an effect on formation of inclusion 26 Varying the temperature of induction and pH of the medium reduced the amount of the bodies.’ protein produced but did not improve its solubility. As a result, our collaborators also expressed neuB as a maltose-binding protein fusion,’ 27 MalE-NeuB. Maltose binding protein is a useful affinity tag that has been shown to increase the expression levels and solubility of the resulting tagged proteins in many instances. The recombinant plasmid was transformed into E. coil and over-expressed by induction with IPTG. The resulting protein was purified using an amylase column, eluting with buffer containing maltose. The purified protein was used directly in further studies or stored in a buffer containing 10% glycerol at -80 °C. The protein was determined to be greater than 90% pure by SDS-PAGE analysis (Figure 2.13).  73  1  2  3  66kDa  29kDa  1  Figure 2.13 SD S-PAGE gel showing MalE-NeuB purification. Lane 1: molecular weight standards; lane 2: crude cell lysate; lane 3: purified NeuB-Ma1E. Molecular weight standards BSA (29 kDa) and carbonic anhydrase (66 kDa).  Very high amounts of soluble protein (54 mg/L) could be prepared in this way. Unfortunately, it was later discovered that Ma1E-NeuB possesses only very low levels of synthase activity. Since Ma1E-NeuB was designed to contain a protease cleavage site between the two fusion partners, it was subjected to a thrombin treatment. Thrombin is a serine protease that converts soluble fibrinogen into insoluble strands of fibrin.’ ’ 129 The thrombin cleavage site 28 (Leu-Val-Pro-Arg-Gly-Ser) had been incorporated into the Ma1E-NeuB sequence and NeuB produced after cleavage was designed to possess an N-terminal hexahistidine tag. The MalE NeuB enzyme was incubated with thrombin at 25 °C, and the reaction was followed by SDS PAGE (not shown). The results indicated that full cleavage was reached after twelve hours of incubation. The resulting mixture was loaded onto a nickel column and NeuB was eluted with 500 mM imidazole buffer. SDS-PAGE analysis showed that NeuB protein was still intact and did not degrade during the thrombin cleavage. However, no synthase activity was detected either before or after the His-tag purification column. As a result, for all remaining studies the MalE NeuB version of the N,N’-diacetyllegionaminic acid synthase was used. 74  2.4.2  Test for N,N’-Diacetyllegionaminic Acid Synthase (NeuB) Activity Once the Ma1E-NeuB enzyme was prepared, its activity was tested by incubation with 6-  deoxyMandiNAc, PEP, and MgC1 2 in Tris-DC1 buffer prepared using D 0. Due to the extremely 2 low activity of the fusion protein, very high concentrations were used (typically 50 mg/mL) and it was difficult to force the reaction to completion. The reaction catalyzed by the N,N’ diacetyllegionaminic acid synthase is shown in Figure 2.14. The progress of the reaction containing a 2:1 mixture of PEP to 6-deoxyMandiNAc was monitored using 31 P NMR spectroscopy (Figure 2.15). Before the addition of MalE-NeuB, the 31 P NMR spectrum showed a single phosphorus signal at -3.19 ppm belonging to the phosphate group of PEP (Figure 2.1 5A). After the addition of MalE-NeuB and incubation for 18 hours, a new signal appeared at 0.03 ppm corresponding to inorganic phosphate (Figure 2.1 5B). Control reactions lacking enzyme or 6deoxyMandiNAc did not produce any phosphate under similar conditions, indicating that the reaction is not due to a phosphatase impurity. Moreover, the incubation of MalE-NeuB with ManNAc and PEP did not produce any phosphate, indicating that NeuB does not possess sialic acid synthase activity.  75  +  opo 1 öc 2 3  NeuB  OH 6-deoxyMandiNAc  CHO AcHN—j—H HO+H HfNHAc H—j—OH 3 CR  6-deoxyMandiNAc  (open chain)  AcHN PEP  2 4 HP0  OH  N,JV’-diacetyllegionaminic acid (Leg5Ac7Ac) O\ COO 2 3CH H—j-OH AcHN+H HO—frH H—fr NHAc H—frOH CH N,N’-diacetyllegrnaminic acid (open chain)  Figure 2.14 Reaction catalyzed by N,N’-diacetyllegionaminic acid synthase (NeuB) and Fischer projections of 6-deoxyMandiNAc and N,N’-diacetyllegionaminic acid  76  A  PEP  B HPO  0.0  -1.0  ppm  -2.0  -3.0  Figure 2.15 31 P NMR spectra monitoring the reaction of PEP and 6-deoxyMandiNAc with Ma1E-NeuB. A) Before the addition of Ma1E-NeuB and B) 18 h after the addition of MalE NeuB (121.5 MHz, D 0, 25 °C). 2  The Ma1E-NeuB enzymatic reaction was also followed by ‘H NMR spectroscopy. The most efficient way to follow the enzymatic conversion was to focus on the 1.7 to 2.4 ppm region of the spectrum (Figure 2.16). Before the addition of MalE-NeuB, the ‘H NMR spectrum shows peaks corresponding to the acetamido protons for the f3-anomer and cL-anomer of 6deoxyMandiNAc at 2.03 and 2.12 ppm, and 2.04 and 2.08 ppm, respectively (Figure 2.16A). After the addition of MalE-NeuB, the signals corresponding to 6-deoxyMandiNAc were found to be replaced by signals consistent with those expected for Leg5Ac7Ac (Figure 2.16B). 84 A single anomer of Leg5Ac7Ac is the main product formed in the reaction as indicated by the appearance of only one new pair of acetamido methyl signals at 1.99 and 2.01 ppm. This is assigned to the f3anomer which is expected to be more stable due to the equatorial carboxylate group. A similar preference is seen with sialic acid.’ ° The C-3” methylene protons of Leg5Ac7Ac appear at 1.83 3 77  ppm and 2.23 ppm. The doublet of doublets at 2.23 ppm corresponds to the H-3 equatorial proton (H-3eq) and is characterized by a large J3eq,3 value (13.1 Hz) and small J 4 value (4.7 , 3 Hz). The doublet of doublets that appears as a triplet at 1.83 ppm corresponds to the H-3 axial proton (H-3ax). This signal is split by a strong geminal coupling to the H-3eq proton 13.1 Hz) and a strong coupling to the H-4 proton (J 4 , 3  =  (J3ax,3eq  =  12.2 Hz). The large J 4 value , 3  indicates that H-3ax and H-4 have a trans-diaxial relationship and that a newly formed hydroxyl group occupies the equatorial position. This demonstrates that Ma1E-NeuB catalyzes the addition of PEP to the si-face of the open chain aldehyde of 6-deoxyMandiNAc to form the (5)configuration at C-4. This is the same stereospecificity as displayed by sialic acid synthase and pseudaminic acid 49 synthase. 131 ’  78  -NAc  c-NAc iI;C N1[\c 1 I\ ‘‘j\  -f  A  p  OH  B ,\c11  I ——  —  ---r-’-  2.30  2.20  210  2.00  1.90-  1.80  ppm Figure 2.16 Partial ‘H NMR spectra monitoring the incubation of PEP and 6deoxyMandiNAc with Ma1E-NeuB A) Before the addition of Ma1E-NeuB and B) after the addition of Ma1E-NeuB (400 MHz, D 0, 25 °C). 2 Since Ma1E-NeuB exhibits very low activity of 4.4 x i0 tmo1 min’mg’, no attempt was made to kinetically evaluate the reaction. Similar activity studies were performed on both Nand C- terminal His-tagged versions of the NeuB enzyme. It was found that the C-terminal Histagged enzyme was about 10-fold more active and the N-terminal His-tagged was about 50-fold more active than Ma1E-NeuB. These low levels of activity likely do not represent the true activity of NeuB in vivo. It is possible that modification of either terminus results in an enzyme with dramatically reduced activity. Perhaps more likely is the possibility that NeuB requires the presence of another unidentified protein or cofactor for full activity. Similar to sialic acid 79  synthase, Leg5Ac7Ac synthase requires the presence of a divalent cation in order for the reaction to occur. Performing the enzymatic incubation in the presence of 1.0 mM EDTA instead of 2 did not lead to formation of inorganic phosphate or Leg5Ac7Ac. This will be discussed MgC1 further in Section 2.4.4. 2.4.3  Isolation and Characterization of N,N’-Diacetyllegionaminic Acid Previous  synthetic  studies  assigned  the  D-glycero-D-galacto  configuration  to  22 In order to confirm that Ma1E-NeuB produces Leg5Ac7Ac and not an epimer, it Leg5Ac7Ac.’ was necessary to fully characterize the product. Enzymatically produced Leg5Ac7Ac was isolated by first removing Ma1E-NeuB using centrifugal protein filters and then passing the solution through a single anion-exchange chromatographic column. The isolated product was then characterized using ESI-mass spectrometry and ‘H NMR spectroscopy. The ESI-mass spectral analysis of Leg5Ac7Ac showed one major peak at m/z 333 [M-Hf. The ‘H NMR spectrum of Leg5Ac7Ac was in complete agreement with that of the synthetic material previously reported in the literature (Figure 2.1 7)•84  80  OH eq  3 H  AcHN  ax  /  N-Ac  9 H 74 H  5 H  6 H  I  1  I  I  4.00  I  I  I  I  I  3.50  I  I  ax 3 H  H3eq  *  I  I  3.00  I  I  I  j  •  •  I  I  2.00  2.50  I  I  I  150  I  I  I  I  I  1.00  p1)In 0, 25 °C). 2 Figure 2.17 ‘H NMR spectrum of Leg5Ac7Ac (400 MHz, D  =  triethylammonium  81  2.4.4  Potential Synthase Mechanisms and Test for C-O vs P-O Bond Cleavage Enzymes that catalyze the condensation between a sugar carbonyl and PEP have been  proposed to follow one of two mechanisms: a C-O bond cleavage mechanism or a P-O bond cleavage mechanism. 132 The first step of the C-O bond cleavage mechanism is the attack of C-3 of the PEP at the carbonyl carbon of the open chain aldehyde of 6-deoxyMandiNAc (Figure 2.1 8A). This attack would be facilitated by the presence of the divalent cation activating the aldehyde carbon. An oxocarbenium ion intermediate would be formed as a result of the PEP attack. A tetrahedral intermediate is then formed by the attack of a water molecule onto the oxocarbenium ion. Upon the collapse of the tetrahedral intermediate and release of phosphate, the open chain of the N,N’-diacetyllegionaminic acid is generated, which subsequently cyclizes to the pyranose form in solution. The C-O bond cleavage mechanism is employed by both sialic acid and pseudaminic acid synthase. It has also been observed with 2-keto-3-deoxy-D-manno-2octulosonic acid 8-phosphate synthase and 2-keto-3-deoxy-D-arabino-2-octulosonic acid 7phosphate 33 synthase.’ 134 The P-O bond cleavage mechanism is initiated by the attack of water ’ on to the phosphate group of the PEP (Figure 2.1 8B). This attack would result in the formation of the enolate anion of pyruvate and the release of phosphate. The formed enolate anion would then attack the carbonyl carbon on the open chain form of 6-deoxyMandiNAc to form the open chain of the Leg5Ac7Ac, which again would cyclize to the pyranose form in solution. The P-O bond cleavage mechanism is employed by pyruvate kinase 135 and PEP carboxykinase’ 36 where catalysis is thought to proceed with a nucleophilic attack at the phosphate group. In order to distinguish between a C-O and a P-O bond cleavage mechanism, an incubation of [2-’ 0]-PEP with PEP-condensing synthases is generally used (see ‘ 8 0-labeled atoms in 8 Figure 2.l8).132  134  MalE-NeuB operates via a C-O bond cleavage mechanism, then the 82  reaction with [2-’ 0-]PEP would lead to the formation of ‘ 8 0-labeled phosphate. On the other 8 hand, if the enzyme employs the P-O bond cleavage mechanism, then ‘ O-labeled Leg5Ac7Ac 8 should be formed. Similar labeling studies on both sialic acid synthase and pseudaminic acid synthase had shown that a C-O bond cleavage mechanism is employed by both of these enzymes. 131 ’ 49  —  oc 2  A  18 = o—o  H( >=O% R  -  B  2  2 M  c 2 o  18  H 18  -  .n r  =  I R  tetrahedral intermediate  oxocarbenium ion intermediate  3  O2O  “OH R  2”OH R  Leg5Ac7Ac (open chain)  AcNH”  ç ) OH  AcNH  18-j = c_O-PO 2 O 3  H  -  (  —  6-deoxyMandiNAc (open chain) + PEP  —  18+ = cg—PO 2 o 3  —  18— c._Y_O 2 O -  2 OH  H  4  —  H 4 PO  18 0  c 2 O  C 3 H  )  .111  OH  I “‘OH  R  R  6-deoxyMandiNAc (open chain) + enolate  6-deoxyMandiNAc (open chain) + PEP  Leg5Ac7Ac (open chain)  Figure 2.18 Proposed C-O vs. P-O bond cleavage mechanisms for MalE-NeuB. A) C-O bond cleavage mechanism and B) P-O bond cleavage mechanism. 0]-PEP was generously donated by a former group member Dr. Wayne Chou.’ 18 [2’ 3 The purity of the compound was determined by an ESI- mass spectral analysis. It was found that the [20]-PEP had 54% 18  180  label incorporation. The substitution of 160 for 180 in a singly  P NMR signal of the bonded position to phosphorus results in a small upfield shift in the 31 137 therefore, 31 P NMR spectroscopy was used to monitor the enzymatic labeled phosphorus atom, [2-’ 0 J-PEP and 12 mM 6-deoxy incubation. MalE-NeuB was incubated with 20 mM 8 MandiNAc in a buffer containing 1.0 mM MgCl . The initial 31 2 P NMR spectrum of the 54% 83  labeled [2-’ 0]-PEP showed two phosphorus signals at -3.01 ppm and -3.03 ppm corresponding 8 0-labeled PEP, respectively (Figure 2.1 9A). After several hours of 8 to unlabeled PEP and ‘ incubation, two new phosphorus signals appeared. The signal at 0.16 ppm is attributed to the unlabeled phosphate and the signal at 0.14 ppm to the ‘ 0-labeled phosphate (Figure 2.19B). The 8 ratio of  160  to  180  was similar to that in the [2-’ 0j-PEP, indicating that the 8  180  label was fully  retained in the phosphate produced. The results of this experiment show that the reaction proceeds through the C-0 bond cleavage mechanism.  0-PEP 8 ‘ A  0-PEP 6 ‘  —  ji  i1 ()-P O 12 H 3  I  B  0.2  0.1  0  -2.8  -3.0  -3.2  pp11’  Figure 2.19 31 P NMR spectra monitoring the incubation of Ma1E-NeuB with {2-’ 0]PEP 8 and 6-deoxyMandiNAc A)Before the addition of Ma1E-NeuB and B) after the addition of Ma1E-NeuB (121.5 MHz, D 0, 25 °C). 2  84  2.5 CMP-N,N’-Diacetyllegionaminic Acid Synthetase  Previous studies had identified a neuA homolog (1pg075 1) in L. pneumophila that is believed to encode a CMP-Leg5Ac7Ac synthetase. 91 The protein is 54% and 50% identical in sequence to the CMP-NeuAc synthetases from E. coli and N meningitidis, respectively. 2.5.1  Identification, Expression and Purification of CMP-N,N’-Diacetyllegionaminic Acid Synthetase The neuA gene was cloned by our collaborators in the laboratory of Dr. Martin Young at  NRC. This recombinant plasmid was over-expressed in E. coli by induction with IPTG. The resulting cells were harvested, and following lysis, the crude cell lysate was loaded onto an affinity chromatography column containing immobilized nickel (Ni ) and the enzyme of interest 2 was eluted using 500 mM imidazole buffer. Following buffer exchange, the protein was either used directly or stored at -80 °C in a buffer containing 10% glycerol. The protein was determined to be greater than 90% pure by SDS-PAGE analysis (Figure 2.20).  85  1  2  4  66kDa  29kDa  Figure 2.20 SDS-PAGE gel showing NeuA purification. Lane 1: molecular weight standards; lane 2: crude cell lysate before over-expression; lane 3: crude cell lysate after incubating with ITPG for 5 h; lane 4: purified NeuA. Molecular weight standards BSA (29 kDa) and carbonic anhydrase (66 kDa).  2.5.2  Activity of CMP-N,N’-Diacetyllegionaminic Acid Synthetase Once the NeuA enzyme was prepared, its activity was tested by incubation with  Leg5Ac7Ac and cytidine-5’-triphosphate (CTP) in Tris-HC1 buffer containing MgC1 . The 2 reaction catalyzed by CMP-Leg5Ac7Ac synthetase is shown in Figure 2.21. Due to the low activity of Leg5Ac7Ac synthase (MalE-NeuB) only small amounts of pure Leg5Ac7Ac were prepared. As a result, the progress of the NeuA reaction was only followed by negative ES I-mass spectrometry. Before the addition of NeuA, the ESI-mass spectrum showed a single peak at m/z 333 [M-Hf belonging to the starting material Leg5Ac7Ac. After the addition of NeuA and incubation for two hours, a new signal appeared at m/z 660 [M+Na-2Hf belonging to CMP Leg5Ac7Ac. This observation was used to confirm that the L. pneumophila NeuA possessed CMP-Leg5Ac7Ac synthetase activity. No further characterizations of the product or studies were performed on the NeuA enzyme.  86  OH  OH  OH C 3 H  7N 1  AcHN  OH  CTP  PPi  Leg5Ac7Ac  OCMP  HC Ad-IN  OH  CMP-Leg5Ac7Ac  Figure 2.21 Reaction catalyzed by CMP-Leg5Ac7Ac synthetase.  2.6 Conclusions Three candidate genes coding for the proteins involved in the biosynthesis of CMP  Leg5Ac7Ac were cloned by our collaborators in Dr. Martin Young’s laboratory. I have characterized the corresponding enzymes for the first time and thus have elucidated the biosynthetic pathway of CMP-Leg5Ac7Ac. Several experiments were carried out with the enzymes in order to elucidate their mechanisms of action. I first demonstrated that the NeuC homolog was a hydrolyzing 2-epimerase that catalyzed the conversion of hydrolyzing UDP-Bac2,4diNAc into 6-deoxyMandiNAc and UDP. The reaction is analogous to the UDP-GlcNAc 2-epimerase of sialic acid biosynthesis. 49 The incubation of UDP-Bac2,4diNAc in a deuterated buffer with the hydrolyzing 2-epimerase led to the formation of 6-deoxy-c-[2H]MandiNAc. The fact that the c-anomer was formed first 2 showed that the reaction proceeds with the retention of configuration at C-i. The observation that the reaction proceeds with solvent-derived isotope incorporation at C-2” indicates that C-2” is deprotonated and reprotonated with a solvent-derived isotope during the course of the reaction. The reaction carried out in 0 18 buffer demonstrated that the loss of UDP occurs through a C-O 2 H bond cleavage mechanism. Together, these results indicate that the first step of the reaction consists of an anti-elimination of UDP. The second step of the reaction involves a syn-hydration 87  of the 6-deoxy-2,4-diacetamidoglucal intermediate. This is in agreement with the mechanism proposed for the UDP-G1cNAc 2-epimerase enzyme that is involved in the biosynthesis of sialic acid. The second enzyme in the biosynthesis of CMP-Leg5Ac7Ac is a PEP-dependent Leg5Ac7Ac synthase. The incubation of 6-deoxyMandiNAc with PEP in the presence of MalE NeuB led to the formation of Leg5Ac7Ac and phosphate. The resulting Leg5Ac7Ac had a D glycero-D-galacto configuration which is in agreement with previously published results.’ 22 The 0]-PEP in the Ma1E-NeuB reaction resulted in the formation of 18 8 use of [2-’ 0-labeled phosphate, indicating that MalE-NeuB utilizes a C-O bond cleavage mechanism. The mechanism is believed to involve an initial attack of the C-3 of PEP onto the open chain aldehyde of 6deoxyMandiNAc forming an oxocarbenium ion intermediate, followed by the attack of water and formation of a tetrahedral intermediate (Figure 2.1 8A). The tetrahedral intermediate then collapses, releasing phosphate and forming the open chain form of Leg5Ac7Ac which cyclizes in the solution to form predominantly the f3-anomer. Finally, the last enzyme in the biosynthetic pathway, NeuA, was identified as CMP-N,N’ diacetyllegionaminic acid synthetase and converted Leg5Ac7Ac into CMP-Leg5Ac7Ac in the presence of CTP and MgC1 . The biosynthesis of CMP-Leg5Ac7Ac from UDP-Bac2,4diNAc in 2 Legionellapneumophila is shown in Figure 2.22.  88  HC AcHN\—Q HO—..-1 AcF’ OUDP  C NHAc 3 H  NeuC  AcHN HO  0 2 H  OH cL-6-deoxyMandiNAc  UDP-N,N ‘-diacetylbacillosamine (UDP-Bac2,4diNAc)  NeuB PEP  2 C 3 H CO AcFIN  OH  OCMP  OH  OH  CMP-Leg5Ac7Ac  NeuA -  P[  CTP  OH  H C 3 COj AcHN  OH  N,N’-diacetyllegionaminic acid (Leg5Ac7Ac)  Figure 2.22 Biosynthesis of CMP-Leg5Ac7Ac in Legionella pneumophila.  89  2.7 Future Directions The three enzymes involved in the biosynthesis of CMP-Leg5Ac7Ac have been identified and investigated. Further studies could focus on a more detailed investigation of their chemical mechanisms and on understanding the reasons for the low activity exhibited by Leg5Ac7Ac. In order to strengthen the proposal of an elimination-hydration mechanism for the hydrolyzing 2-epimerase reaction, it would be useful to incubate synthetically prepared 6-deoxy2,4-diacetamidoglucal intermediate with the enzyme. This would test whether the intermediate is catalytically competent to serve as a substrate for the second step of the reaction (hydration). A similar experiment was performed on the hydrolyzing UDP-G1cNAc 2-epimerase. ° Obtaining an 5 X-ray crystal structure of the UDP-Bac2,4diNAc hydrolyzing 2-epimerase would help to identify key residues in the active site of the epimerase and give insight into their role during the enzymatic reaction. In case of the bacterial UDP-GlcNAc hydrolyzing 2-epimerase several mutants were prepared and investigated. The Asp 131 Asn mutant catalyzed the formation and a release of 2-acetamidoglucal intermediate into solution. ° It is believed that Aspi 31 is involved 5 in the hydration of the glycal intermediate. Similar studies can be performed on the UDP Bac2,4diNAc hydrolyzing 2-epimerase with the corresponding Asp 136 residue to determine if it plays the same role in the catalytic cycle. The Leg5Ac7Ac synthase requires full kinetic characterization. It would be necessary to prepare a more catalytically active version of the enzyme. If the low activity of NeuB is due to N- or C-terminal modification, then a native enzyme must be obtained. One way to prepare a native version of the enzyme is to make a version of the N- or C-terminal His-tag which can be cleaved during the purification process. Another approach would involve the preparation of the 90  NeuB enzyme without any modification to the amino acid sequence and the purification of the enzyme using several chromatographic steps. However, the low activity of NeuB could be due to a requirement for another protein or cofactor, and to test for this the protein could be co incubated with proteins encoded by nearby genes in the operon. The activity could also be tested in the presence of crude cell lysate of Legionella pneumophila. The stereochemical course of the addition of PEP to 6-deoxyMandiNAc during the reaction needs to be addressed. The separate incubations of NeuB with Z-[3Hj-PEP and E-[32 H]-PEP in the presence of 6-deoxyMandiNAc 2 would help to determine the stereochemistry of the addition.’ ’ 139 38  have demonstrated that the  reaction occurs through a C-O bond cleavage mechanism, however, the stereoconfiguration of the tetrahedral intermediate is still unknown (Figure 2.23). Several tetrahedral intermediate analogs can be prepared to help determine the stereoconfiguration at C-2. These compounds would also be expected to serve as potent inhibitors of the enzyme. They could be prepared as a mixture of epimers with different stereoconfiguration at C-2. Upon incubation of the mixture with synthase, it would be expected that the enzyme would bind most tightly to the epimer that bears the same stereoconfiguration as the normal intermediate. The stereoconfiguration of the tetrahedral intermediate could be elucidated by analyzing the resulting enzyme-inhibitor complexes with X-ray crystallography. Finally, the UDP-Bac2,4diNAc that was used in these experiments was prepared with Campylobacterfejuni enzymes. It is necessary to locate the genes encoding these three enzymes in the L. pneumophila genome in order to demonstrate that the bacterium is capable of generating the substrate for the pathway described in this thesis.  91  unknown stereoconfiguration  •PO=3 HO  AcI-INII’  ‘OH  AcHN  AcHN  •‘IIOH  •‘IIOH  C 3 H Tetrahedral intermediate  C 3 H Analog 1  Analog 2  Figure 2.23 Potential tetrahedral intermediate analogs for NeuB.  92  2.8 Experimental 2.8.1  Materials and General Methods All chemicals and enzymes, unless otherwise noted, were purchased from Sigma-Aldrich  and were used without further refinement. ‘ 0-enriched H 8 0 (95%) was purchased from 2 Cambridge Isotope Laboratories. All the buffer solutions were prepared using distilled water. ‘H NMR spectra were obtained on Bruker AV300/AV400 NMR spectrometers. ‘ C NMR spectra 3 were obtained on Bruker AV300/AV400 NMR spectrometers at a field strength of 75 MHz or 100 MHz, respectively. Proton-decoupled 31 P NMR spectra were recorded either on a spectrometer at a field strength of 121.5 MHz or 162 MHz, respectively. ESI-mass spectrometry was performed on a Bruker Esquire LC mass spectrometer. Chelex® 100 resin (200-400 mesh, Na form), AG® 1-X8 resin (100-200 mesh, formate form), and Bio-Gel® P-2 resin were purchased from Bio-Rad Laboratories. DE-52 resin (DEAE Cellulose) was purchased from Whatman. Amicon Ultra-4 centrifugal protein filters (4 mL, 10 000 MWCO) were purchased from Millipore. All proteins were handled at 4 °C unless otherwise stated. Protein concentrations were determined according to the method of Bradford’ ° with bovine serum albumin as the 4 standard. Protein purity was determined using SDS-PAGE gel electrophoresis and visualized using coomassie blue stain according to the method of Laemmli.’ ’ Protein molecular masses 4 were determined using BSA (66 kDa) and carbonic anhydrase (29 kDa) as mass standards.  93  2.8.2  Cloning of Legionellapneumophilia neuA, neuB and neuC.  The cloning of neuA, neuB and neuC was performed in the laboratory of Dr. Martin Young at NRC. The neuA, 1pg0751, neuB, 1pg0752, neuC, 1pg0753, genes were obtained from the genomic DTM by PCR amplification with DNA of L. pneumophilia subsp, pneumophilia ATCC 331 52  Phusion DNA polymerase (New England Biolabs Inc.) according to the manufacturer’s instructions. The primers used incorporated either Ndel or Sal] cloning sites (underlined) and a 6X His tag in either the 5P or 3P PCR primer. The primer pairs used were: (neuA5PHis) 5’CTAGCTAGCTAGCATATGCATCACCATCACCATCACAGAATATTGGCAGTAATCC CGGC-3 ‘(forward)  and  5’ -CTAGCTAGCTAGGTCGACTTATTATACTAGAGCCTCTT  GGTTTAATTCC-3 ‘(reverse), (neuA3PHis) 5 ‘-CTAGCTAGCTAGCATATGAGAATATTGGCAGTAATCCCGGC-3’ (forward) and 5’ -CTAGCTAGCTAGGTCGACTTATTAGTGATGG TGATGGTGATGTACTAGA-GCCTCTTGGTTTAATTCC-3‘(reverse),  (neuB5PHis)  5’-  CTAGCTAGCTAGCATATGCATCACCATCACCATCACACTTGTTTTATTATTGCTGAAG CAGG-3’ (forward) and 5’ -CTAGCTAGCTAGGTCGACTTATTAATATGTTCCCATAACA AAGTTAGTACCCGC-3 ‘(reverse), (neuBPHis)5 ‘-CTAGCTAGCTAGCATATGACTTGTTTT ATTATTGCTGAAGCAGG-3 ‘(forward) and 5’ -CTAGCTAGCTAGGTCGACTTATTAGTGA TGGTGATGGTGATGATATGTTCCCATAACAAAGTTAGTACCCGC-3‘(reverse), (neuC5PHis) 5’ -CTAGCTAGCTAGCATATGCATCACCATCACCATCACATCAGAAAAA TAATTTATGTTACAGGTACTCG-3’ (forward) and 5’ -CTAGCTAGCTAGGTCGACTTATT AGTATGCATTGCATTTATTCAATATTTGTGAG-3’ (reverse), (neuC3PHis) 5’ -CTAGCTA GCTAGCATATGATCAGAAAAATAATTTATGTTACAGGTAC-TCG-3’ (forward) and 5’CTAGCTAGCTAGGTCGACTTATTAGTGATGGTGATGGTGATGGTATGCATTGCATT TATTCAATATTTGTGAG-3’ (reverse). The PCR products were gel purified and cloning sites  94  were generated by double digestion with Ndel and Sail restriction enzymes according to the manufacturer’s suggested protocol (New England Biolabs Inc.). The genes were cloned into Nde 1/Sail digested plasmid pCWoriand the constructs maintained in Escherichia coil AD202. 2.8.3  Over-expression and Purification of L. pneumophila NeuA, NeuB and NeuC. The recombinant neuC plasmid was transformed into E.coii BL21 (DE3) competent cells  which were incubated in 10 mL Luria-Bertani (LB) medium containing 50 mg/L ampicillin at 37 °C/225 rpm for 10 h. The overnight culture was then poured into 500 mL of LB medium containing 50 mg/L ampicillin and shaken at 37 °C/225 rpm until an 0D 600 of 0.6  —  1.0 had been  reached. Cultures were induced with 1 mM isopropyl /3-D-galactopyranoside (IPTG). After incubation for 5 h at 37 °C, the cells were harvested by centrifugation and stored as a pellet at  —  80°C. The pellets were resuspended in 10 mL of a phosphate buffer (20 mM, pH 8.0) containing 2 mM dithiothreitol (DTT), 1 mg/L of aprotinin, and 1 mg/L pepstatin A at 4 °C. The cells were subsequently lysed by passage through a French Pressure cell at 20 000 psi. The lysate was centrifuged at 6 000xg for 1 h, passed through 0.45 jim and 0.22 jim filters, and loaded onto a column containing 10 mL of Chelating Sepharose Fast Flow resin (Pharmacia Biotech), which 4 and washed with sodium phosphate buffer (20 mM, was previously charged with 100 mM NiSO pH 8.0, containing 0.5 M NaCl and 5 mM of imidazole). The purification process was monitored by a Flow Thru UV monitor Spectrometer at 280 nm. Nonspecifically bound proteins were washed away by applying buffers containing first 5 mlvi and then 125 mlvi imidazole. Finally, bound enzyme was eluted using a 500 mM imidazole buffer.  The fractions containing the  desired enzyme were combined and concentrated using Amicon Ultra Centricons (Millipore) before flash freezing with liquid N 2 in the presence of 10% glycerol. Similar procedures were used for NeuB and NeuA. For NeuA, the fractions eluted with 500 mM imidazole were dialyzed 95  against 20mM Tris-HC1 buffer (pH 7.5) containing 200 mM NaC1 before being concentrated and 2 in the presence of 10% glycerol. flash-frozen with liquid N 2.8.4  Sub-cloning of L pneumophila neuB for Fusion Protein, Overexpression and .  Purification. The cloning of neuB was performed in the laboratory of Dr. Martin Young. The NeuB5PHis clone, verified by sequence analysis, was double digested with Ndel and Sail restriction enzymes. The liberated insert was gel purified and sub-cloned into pCWOri containing the E. coli malE gene with a downstream thrombin cleavage recognition sequence that was cloned as a BamHl to Ndel fragment. The constructs were maintained in E. coli AD202, and cells bearing positive clones were identified by colony PCR using specific malE and neuB primers and by restriction mapping. The resulting malE-neuB plasmid was transformed into E. coli BL21 (DE3) competent cells, which were then grown at 37 °C in a 2YT media supplemented with 50 mg/L ampicillin. Over-expression of the fusion protein was induced by the addition of IPTG to a final concentration of 1 mM at an A 600 of 0.5, and growth continued for a further 6 h. Cells were harvested by centrifugation at 10 000 x g for 15 mm, resuspended in a 20 mM Tns-HC1 buffer pH 7.5 containing 1 mgIL of aprotinin, and 1 mg/L pepstatin A, and lysed by passage through a French Pressure cell at 20 000 psi. The cell lysate was clarified by centrifugation at 27 000 x g for 30 mm and the cell debris was discarded. The total membrane and soluble protein fractions were obtained from clarified cell extracts by centrifugation at 10 000 x g for 60 mm. Following the adjustment to 200 mM NaC1 and 1 mM EDTA the soluble protein fraction was passed through a 20 mL amylose resin (New England Biolabs Inc.) column previously equilibrated with 200 mM NaCl, 20 mM Tris-HC1 pH 7.5 and 1 mM EDTA. The column was washed with 3 column volumes of an equilibration buffer and bound protein was 96  eluted with an equilibration buffer containing 10 mM maltose. Fractions containing protein of interest, as judged by SDS-PAGE, were pooled and dialyzed against 200 mM NaC1, 20 mM Tris HC1 buffer pH 7.5. Glycerol (10%) was added to the solution and aliquots were flash frozen in liquid nitrogen and stored at -80 °C. 2.8.5  Over-expression and Purification of His-Tagged Pg1F and Pg1E The plasmids pNRC4O. 1 and pNRC4 1.3 were used in the over-expression of pglF  94 In each (Cj 11 20c) and pglE (Cj 1121 c) to give His-tagged proteins, as described previously. case the appropriate plasmid was transformed into E. coli BL21 (DE3) competent cells which were incubated in 10 mL Lucia-Bertani (LB) medium containing 50 mg/L ampicillin at 37 °C/225 rpm for 10 h. The culture was then added to 500 mL of LB medium containing 50 mg/L 600 of 0.6 had been reached. The culture was ampicillin and shaken at 37 °C/225 rpm until an 0D allowed to continue to grow for 5 h after 70 mg!L of IPTG was added. Cells were harvested by centrifugation and the resulting pellet was stored at -80 °C. The pellet was then resuspended in 10 mL of a phosphate buffer (10 mM, pH 7.0) containing 2 mM dithiothreitol (DTT), 1 mg/L of aprotinin, and 1 mg/L pepstatin A. The cells were lysed by passage through a French Pressure cell at 20 000 psi. The lysate was centrifuged at 10 000 x g for 1 h and passed both through a 0.45 jim and a 0.22 jim filters. A column containing 10 mL of Chelating Sepharose Fast Flow , washed with 20 mL of 4 resin (Pharmacia Biotech) was charged with 20 mL of 100 mM NiSO distilled H 0 and 30 mL of a sodium phosphate buffer (10 mM, pH 7.0, containing 0.5 M NaC1 2 and 5 mM of imidazole). The lysate was loaded onto the column and eluted with the same buffer containing increasing amounts of imidazole in a step-wise fashion (5 mM, 125 mM and 500 mM). Fractions containing protein of interest, as judged by SDS-PAGE, were pooled and 2 in dialyzed against a 20 mM phosphate buffer (pH 7.0) before being flash frozen with liquid N 97  the presence of 10% glycerol. 2.8.6  Chemo-enzymatic Synthesis of UDP-Bac2,4diNAc UDP-Bac2,4diNAc was prepared under conditions slightly modified from those  92 A purified sample of Pg1F (5 mg) was added to 50 mL of a phosphate described previously. buffer (10 mM, pH 7.0) containing 500 mg of UDP-N-acetylglucosamine disodium salt and 200 ,iM NAD. The solution was incubated for 6 h at 37 °C and the reaction progress was followed by a negative ESI-mass spectrometry. Almost all of the starting material was converted to a UDP-4-keto-sugar.  To synthesize the UDP-4-amino sugar, the following components were  added to the mixture: 5 mg of purified Pg1E, L-glutamate (15 mM final concentration) and pyridoxal 5’-phosphate (PLP) (100 jiM final concentration). The solution was incubated for 4 h at 37 °C and the reaction progress was monitored by a negative ESI-mass spectroscopy. It was determined that  >  95% of the UDP-4-keto sugar (m/z 587, [M-Hf) was converted to UDP  Bac2Ac (mlz 589, [M-Hf) during this time. Enzymes were then removed by centrifugal ultrafiltration and the resultant filtrate was loaded onto a 220 mL column of DEAE cellulose (DE-52, Whatman Inc.) and eluted with a linear gradient of a 0 to 0.5 M triethylammonium bicarbonate buffer.  The A 254 of the eluant was monitored, and UV-active fractions were  analyzed by a negative ESI-mass spectroscopy.  Those containing UDP-Bac2Ac were  lyophilized to dryness. The lyophilized sugar (295 mg) was stirred with 1.2 mL acetic anhydride in 25 mL methanol at room temperature for 24 h. Negative ESI-MS showed that the starting material was completely converted to UDP-Bac2,4diNAc (m/z 631, EM-Hf) during this time. After removal of the solvent under reduced pressure, the product was loaded onto a DE-52 anion exchange column and subjected to linear gradient elution as described above. After 0 and lyophilized again. This procedure 2 lyophilization, the product was dissolved in 10 mL H 98  was repeated twice more to yield 151 mg (28%) of the UDP-Bac2,4diNAc as its 92 ‘H NMR triethylammonium salt. ‘H and 31 P NMR spectra matched those in the literature. 0): 2 (D  7.97 (d, 1H, .15,6  (dd, iH, 2 ”, 1 J ”  3.3 Hz,  8.1 Hz, H-6); 5.97 (d, 2H, J 8.i, 6 , 5 Ji”,p  =  Hz  ,  H-5, H-i’); 5.48  7.0 Hz, H-i”); 4.38. (m, 2H, H-2’,H-3’); 4.28. (m, 1H, H-4’);  4.18-4.23 (m, 2H, H-5’); 4.05 (m, iH, H-5”), 4.02 (m, iH, H-2”), 3.79 (t, iH, 3 ”, 2 J ” Hz, H-3”); 3.69 (t, 1H, 5 ”, 4 J ”  2.8.7  p  =10.2  10.2 Hz, H-4”); 2.06 (s, 3H, H CCONH); 2.03 (s, 3H, 3  P NMR (D 0): 2 CCONH); 1.20 (d, 1H, 6 3 H ”, = 6.1 Hz, H-6”). 31 5 J ” —12.67 (d, JPa  =  —10.84 (d,  JPa, pp  20.1 Hz),  20.3 Hz). ESIMS: m/z 631 [M-Hf  Characterization of Hydrolyzing 2-Epimerase Activity 2.8.7.1.  NeuC Homolog Activity Assay  A glycerol stock solution of the NeuC homolog (70 tg) was subjected to a bufferexchange with a 25 mM phosphate buffer (JH 7.5, 150 iL final volume) using centrifugal ultrafiltration. This was added to a solution of UDP- Bac2,4diNAc (3.0 mg) dissolved in 850 iL 1120 (1.0 mL final volume) and “P NMR and positive ion ESI-mass spectra were acquired at  timed intervals. Once the reaction was complete, the enzyme was removed by centrifugal ultrafiltration and the resultant filtrate was passed through a column (15 mL) of AG-iX8 resin (100-200 mesh, formate form) and eluted with water to remove the UDP. The flow through was 0 for spectral analysis. Material prepared in this 2 lyophilized to dryness and redissolved in D fashion was also analyzed by a 2-D heteronuclear NMR experiment (HMQC) both with, and without, the ‘H-’ C coupling constant retained in the ‘H dimension. This was done in order to 3 0): c 2 establish the identity of the H-i” signals for each of the anomers (vide infra). ‘H NMR (D anomer,  5.10 (d, 1H, J,”, ” 2  1.2 Hz, H-i); 4.30 (dd, 1H, 3 ”, = 4.5 Hz, H-2”); 4.06 (dd, 1H, 2 J ” 99  ”, 3 J ” 4  =  10.5 Hz, H-3”); 3.97 (m, 1H, H-5”); 3.80 (t, 1H, 5 ”, = 10.4 Hz, H-4”); 2.08 (s, 311, 4 J ”  CCONH); 1.19 (d, 3H, 3 CCONH); 2.04 (s, 3H, H 3 H  6.4 Hz, H-6”);. 13-anomer, 6 4.96 (d,  1H, Ji”, ” = 1.3 Hz, H-i”); 4.47 (dd, 1H, 3 2 ”, = 4.3 Hz, H-2”); 4.47 (dd, 1H, 3 2 J ” ”, = 4.3 Hz, H2 J ” 2”); 3.67 (t, 111, J ”,s” = 10.2 Hz ,H-4”); 3.67 (m, 1H, H-5”); 2.12 (s, 3H, H 4 CCONH); 2.03 (s, 3 CCONH); 1.23 (d, 3H, Js”, 3 ” = 6.3 Hz, H-6”). ESIMS: m/z 269 [M+Na]. 6 3H, H 2.8.7.2.  NeuC Kinetic Studies  Enzyme kinetics were measured using a continuous coupled assay for UDP formation.’ 23 PO buffer (pH 7.5), 10 mM MgC1 2 NaH Each cuvette contained a 50 mM 4 , 2 mM PEP, 0.2 mM 2 NADH, 20 units of lactate dehydrogenase, 18 units of pyruvate kinase, and UDP-Bac2,4diNAc (varying from 25 to 1000 tiM) at a total volume of 800 .tL. The concentrations of stock UDP sugar solutions were determined by measuring A 260  (  =  9 890 M’ cm’). Enzymatic reactions  were initiated by the addition of a 20 L of 0.05 mglmL enzyme solution (final concentration 2.0 nM). Rates were measured by monitoring the decrease in A 340 at 37 °C. Kinetic parameters were determined by fitting initial velocities to the Michaelis-Menten equation using GraFit 4.0. No detectable background release of UDP was observed in the absence of the added NeuC homolog 2.8.7.3.  Stereochemistry and Solvent Deuterium Isotope Incorporation Studies  A glycerol stock solution of NeuC (70ig) was subjected to a buffer-exchange with a 25 0 buffer (pD 7.4, 200 p1 final volume) using centrifugal ultrafiltration. This 2 mM phosphate /D was added to a solution of UDP- Bac2,4diNAc (3 mg) dissolved in 800 iL D 0 (1 mL final 2 volume) and  111  NMR spectra were acquired at timed intervals during the incubation at 25 °C.  100  The isotope incorporation also was monitored by positive ESI-mass spectrometry as a function of time. 2.8.7.4.  Metal Dependency of NeuC  Two aliquots of a solution containing UDP-Bac2,4diNAc (3 mg per aliquot) in a 25 mM 2 and the other EDTA phosphate buffer (pH 7.5) were prepared. One aliquot received MgC1 tetrasodium salt, each at a 10 mM final conc. at a total volume of 990 j.tL. The NeuC homolog (70 jig) was added to each sample and the mixtures were incubated for 2 h at room temperature. P NMR spectroscopy with integration of the The progress of the reactions was monitored by a 31 diphosphate signals. 2.8.8  Test for C-O vs. P-O Bond Cleavage Mechanism  A solution of a 25 mM phosphate buffer (pH 7.5, 1.60 mL) was prepared from 50% 16 and 50% 0 2 H 0 18 (95% isotopic enrichment) and divided into two aliquots. To one aliquot 2 H UDP-Bac2,4diNAc (2.0 mg) was added, while 6-deoxyMandiNAc (1.0 mg) was added to the other. The NeuC homolog (70 jig) was added to each sample and the mixtures were incubated at room temperature. The isotope incorporation was monitored by both positive (sugar detection) and negative (IJDP detection) ESI-MS as a function of time. The extent of incorporation into 2,4,6-trideoxy-diacetamidomannose was calculated from the ratio of peaks at m/z 269 [M+Na]) and m/z 271  (180,  (160,  [M+2+Nafl.  101  2.8.9  Characterization of N,N’-Diacetyllegionaminic Acid Synthase  A solution containing 6-deoxyMandiNAc (12 mM) and PEP (20 mM) in a deuterated Tris/DC1 buffer (700 j.iL, 10 mM, prepared in using D 0, pD 7.4) was placed in an NMR tube. 2 Initial ‘H and proton-decoupled 31 P NMR spectra were taken. The solution was removed from 2 in the same deuterated the tube and mixed with 5 mg of the NeuB homolog and 1 mM MgC1 buffer (1 mL total volume). After incubation of the reaction mixture for 20 h at 25 °C, Chelex 0) was added, and the solution was incubated for 2 100 resin (20 mg, previously rinsed with D an additional hour at room temperature. The resulting mixture was analyzed directly by  111  and  p NMR spectroscopy. In order to determine the activity of the NeuB homolog under initial 31 velocity conditions, a solution containing 2,4-diacetamido-2,4,6-trideoxymannose (6.4 mM), 2 (1 mM), and the NeuB homolog (6.5 mg) in Tris/DC1 buffer prepared PEP (20 mM), MgCl using D 0 (10 mM, pD 7.4, 1.0 mL total volume) was placed into NMR tube and immediately 2 H NMR spectroscopy. Spectra were taken every 5 minutes for a period of one monitored by 1 hour while incubating at 25 °C. The conversion rate was calculated by comparing the integrals of the signals due to the acetamido methyl protons of both 2,4-diacetamido-2,4,6trideoxymannose anomers (2.03, 2.04, 2.08 and 2.12 ppm) to those of the product N,N’ diacetyllegionaminic acid (1.99 and 2.01 ppm). The rate of the reaction was determined by using the data that was accumulated during the first 15% of the reaction. 2.8.10 Isolation and Characterization of the N,N’-Diacetyllegionaminic Acid  The NeuB homolog was removed from the enzymatic activity test reactions by centrifugal ultrafiltration and the resulting filtrate was loaded onto a 15 mL column of Dowex  102  AG1 X8 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 0.1 M increments (50 mL per increment) was used to elute the product. N,N’-Diacetyllegionaminic acid eluted from the column in the 0.2 M and 0.4  M  fractions  which  concentrated  were  in  vacuo  and  then  lyophilized.  N,N’  Diacetylegionaminic acid was characterized using a ‘H NMR and negative ESI-MS mass spectrometry and the spectroscopic data was identical to that reported for the synthetically 0) f3-anomer 6 1.17 (d, 3H, 9 2 ’ 122 ‘H NMR (D 84 ”, 6.2 Hz, H-9”), 1.84 (dd, 8 J ” produced material. 1H,  ax”, 3 J ” 4  =11.9 Hz,  J3ax”,3eq”  CCONET), 2.26 (dd, 1H, 3 H ”, 5 J ” 6 =  CCONH); 2.01 (s, 3H, 3 13.1 Hz, H-3ax”), 1.99 (s, 3H, H  J3eq”,3ax”  13.1,  7 . 4 J3eq”,4”  Hz, H-3eq”), 3.64 (dd, 1H,  10.5 Hz, H-5”), 3.75-3.8 (m, 2H, W7”, H8”), 3.93 (ddd, 1H, J3eq,4 4.8, 4 ax, 3 J  10.3 Hz 11.9, J 5 , 4  10.3 Hz, H-4”), 4.23 (dd, 1H, 6 ”, = 10.5, 7 5 J ” ”= 6 J ” , 1.9 Hz, H-6”). ESIMS: m/z 333 [M-Hj.  2.8.11  Test for C-O vs. P-O Bond Cleavage Mechanism ’ 143 and found to have a 42 0]-PEP disodium salt was prepared as described previously’ 8 [2-’  P NMR and mass spectral analysis. A 54% incorporation of the isotopic label as indicated by 31 0]-PEP (20 mM) in Tris/DC1 buffer 18 solution containing 6-deoxyMandiNAc (12 mM) and [2(700 j.iL prepared using D 0, 10 miVi, pH 7.4) was placed in an NIVIR tube. Chelex-100 resin 2 P NMR spectrum was obtained using a (20 mg) was added and an initial proton-decoupled 31 previously reported procedure.’ ’ The Chelex resin was removed by decanting the solution, and 3 the solution was mixed with 5 mg of the NeuB homolog and 1.0 mM MgC1 2 in 250 jtL of the same TrisIDCl buffer. The reaction was incubated for 20 h at 25 °C and a solution of EDTA tetrasodium salt in D 0 (10 mM final concentration) was added to the reaction. Another proton2  103  decoupled 31 P NMR spectrum was acquired with the same parameters: 31 P NMR 0.16 (s, Pi-’ 60) 0, PEP), -3.03 (s, P-’ 6 0, PEP). 8 0.14 (s, Pi-’ 0), -3.01 (s, P-’ 8 2.8.12  Characterization of CMP-N,N’-Diacetyllegionaminic Acid Synthetase Activity A glycerol stock solution of the NeuA homolog (containing 1.25 mg of enzyme) was  subjected to a buffer-exchange with a 20 mM Tris-HC1 buffer, pH 7.5. The enzyme was added to a solution containing 500 iM N,N-diacetyllegionaminic acid, 1.5 mM cytidine 5’-triphosphate disodium salt and 1.0 mM MgC1 2 in the same buffer (final volume 1.0 mL). The reaction was incubated at 25 °C and the progress was monitored using a negative ion ESI-mass spectrometry. After 2 h, all of the starting material with m/z 333 ([M-Hf) was converted to a CMP-N,N’ diacetyllegionaminic acid product with m/z 660 ([M+Na-2Hfl.  104  Chapter Three Mechanistic Studies on UDP-D-Apiose Synthase  105  3.1 Introduction The focus of this chapter will be on the elucidation of the chemical mechanism of UDP-D apiose/UDP-D-xylose synthase (AXS1), the enzyme that converts UDP-D-glucuronic acid (UDP G1cA) into UDP-D-apiose (UDP-Api) in plants (Figure 1.38). As discussed in Section 1.10.2, D apiose is a critical component of the rhamnogalacturonan-JI (RG-II) that, in plant cell walls crosslinks the homogalacturonan backbone via the formation of a borate ester.’° 6 The lack of UDP-D-apiose/IJDP-D-xylose synthase resulted in RG-II deficiency and cell death.’ 44 The work in this chapter was prompted by a report that a candidate gene AXSJ from Arabidopsis thaliana, encoding a potential UDP-D-apiose/UDP-D-xylose synthase, had been identified by comparing the sequence to that of known UDP-D-glucuronate decarboxylases. It was also reported that expression of this gene in E. coil led to production of an active UDP-D-apiose/UDP-D-xylose synthase.” The first part of this chapter describes the over-expression of the AXS] in E. coil in order to generate a hexa-histidine tagged version of UDP-D-apiose/UDP-D-xylose synthase. The purified enzyme was then incubated with UDP-G1cA and the progress of the enzymatic incubation was followed by NMR spectroscopy and ESI-mass spectrometry to fully characterize the reaction products. Further studies were designed to mechanistically investigate the UDP-D.-apiose/UDP D-xylose synthase. A potential intermediate in the UDP-D-apiose/UDP-D-xylose synthase catalyzed reaction, namely UDP-4-ketoxylose, was prepared enzymatically. The reaction between AXS 1 and UDP-4-ketoxylose supported the notion that it is the intermediate of the enzymatic transformation, however the rate of the reaction was extremely low. Finally, UDP-2deoxy-2-fluoro-D-glucuronic acid, UDP-3 -deoxy-3-fluoro-D-glucuronic acid and UDP- [U-’ C] 3  -  3-deoxy-3-fluoro-D-glucuronic acid were prepared. Incubations with these three substrate 106  analogs provided further evidence for the retro-aldol mechanism of the carbon skeleton rearrangement carried out by UDP-D-apiose/UDP-D-xylose synthase.  3.2 Cloning, Expression and Purification of UDP-D-Apiose/UDP-D Xylose Synthase A search of the Arabidopsis thaliana genome sequence revealed two genes, AXS1 (At2g27860) and AXS2 (AtglgO8200) which could potentially encode UDP-D-apiose/UDP-D xylose synthase.” Both of the gene products are SDR enzymes that contain the conserved Gly -Gly-X-Gly sequence responsible for the binding of NAD and the conserved catalytic triad 3 X responsible for promoting hydride transfer. It was demonstrated that AXS1 and AXS2 share 96% sequence identity and that AXS2 encodes a functional UDP-D-apiose/UDP-D-xylose synthase that is identical to the enzyme encoded by AXSJ. As a result, we chose to focus our studies on the enzyme that is encoded by AXSJ. 3.2.1  Preparation of eDNA from Arabidopsis thaliana  Compared to prokaryotic and other eukaryotic genomes, plant genomes tend to be larger and more complex. Generally, they are 10 to 100 times larger than those of other eukaryotes. The major part of the plant genome is composed of introns, which are the non-coding intervening sequences of DNA. The minor part is composed of exons which are amino acid coding regions that are eventually translated to give proteins. Both introns and exons can be transcribed into 45 The transcribed niR.NA undergoes splicing which niRNA by action of an RNA polymerase.’ removes the introns. This is carried out by small nuclear ribonuclear proteins and produces mature mRNA. Since mature mRNA serves as a template for protein synthesis, it is isolated from plant tissues to obtain complementary DNA (eDNA). To obtain mRNA from A. thaliana, several 107  leaves of the plant were flash frozen, ground and then treated with a lysis buffer. The plant leaves were generously donated by Dr. Reinhard Jetter, and Ortwin Guhling provided invaluable assistance in performing mRNA and cDNA preparation. The RNeasy® kit (Qiagen) was used to 46 To obtain cDNA, the isolated mRNA was subjected to the isolate the mRNA from the leaves.’ action of a reverse transcriptase enzyme, which is a DNA polymerase that will use either an RNA or DNA strand as a template (Figure 3.1).’ A short oligonucleotide complementary to the poly-A tail at the 3’ end of the mRNA is first hybridized to the RNA to act as a primer for the reverse transcriptase, which then copies RNA into a complementary DNA chain. The mRNA is degraded by the action of Ribonuclease H (RNaseH) that specifically degrades the RNA in the RNA:DNA hybrid, without affecting DNA. The single strand of eDNA prepared in this way is devoid of both upstream and downstream regulatory sequences, and of introns. Therefore, cDNA from eukaryotes can be translated into functional proteins when expressed in bacteria.  108  Plant leaves lyse cells, purification 3’  5’  WAAA  mRNA  Reverse Transcriptase  5,  AAAAAA TTTTTT  3  mRNA cDNA  RNaseH  TTTTTT  3  5’  single strand cDNA copy of the original mRNA Figure 3.1 Preparation of cDNA from plant mRNA 3.2.2  Expression and Purification of AXS1 The next step in the cloning process was to amplifr the AXS1 sequence from the eDNA  mixture by PCR and insert it into an expression vector. The preparation of an expression plasmid bearing an AXSJ that encodes a His-tagged enzyme was completed using the Novagen Xa ligation-independent cloning (LIC) kit. The AXS] sequence was PCR-amplified from Arabidopsis thaliana eDNA to create a double-stranded piece of DNA containin AXS1 sequence sandwiched between two specifically designed oligomers 15 and 17 nucleotides long. The PCR product was then incubated with T4 DNA polymerase and dGTP. Under these conditions, T4 polymerase hydrolyzes nucleotides from the 3 ‘-ends of the PCR product until the first guanosine 109  is reached. As a result of the polymerase action, the DNA strands are no longer perfectly matched. They contain 12- or 15-base single stranded overhangs that are generally referred to as “sticky” ends. These “sticky” ends are complementary to the overhangs in a commercially available pET-30 vector, such that the target insert anneals with the vector to give a doubly nicked plasmid. The annealed mixture was transformed into competent E. coli cells where ligation occurs to form a plasmid. At this point the plasmid was purified, and the AXS] insert was sequenced at the Nucleic Acid Protein Service (NAPS) unit to verify that no errors were introduced during the PCR and DNA polymerase treatments. The plasmid was then transformed into E. coli competent cells for protein overexpression. The cells were grown at 37 °C in LB medium containing kanamycin. The prepared plasmid contains a gene encoding for kanamycin resistance, so only those cells bearing the plasmid would grow in the above medium. Following induction with IPTO, the cells were harvested, lysed, and the soluble fraction was loaded onto an affinity chromatography column ). The hexahistidine tag on AXS1 binds with Ni 2 , and the 2 containing immobilized nickel (Ni remainder of the cell lysate can be washed from the column. AXS1 was eluted with 500 mM imidazole, and the AXS1 containing fractions were dialyzed into a buffer at pH 8.0. As previously reported, AXS 1 was found to be unstable and a significant loss of activity was observed within 48 h, or upon storage at 4 °C, -20 °C or -80  OC.W  The presence of up to 2 mM  of 1 ,4-dithiothreitol did not have an effect on the stability or the activity of AXS 1. As a result, the enzyme was prepared freshly prior to use in all experiments. AXS 1 prepared in this fashion typically yielded 250 mg per liter of culture. SDS-PAGE analysis (Figure 3.2) of the purified protein revealed a single band at -46 kDa, consistent with the predicted molecular weight and indicating a purity of greater than 90 %. 110  1  2  3  4  66kDa  29kDa Figure 3.2 SDS-PAGE gel showing AXS1 purification. Lane 1: molecular weight standards; lane 2: Cell lysate at 0D 600 = 0.6 (before IPTG); lane 3: Crude cell lysate after 20 h, 0D 600 = 1.6; Lane 4: Purified AXS1. Molecular weight standards BSA (29 kDa) and carbonic anhydrase (66 kDa).  Proper folding of the recombinant protein was assessed by the observation of a tightly bound NADH cofactor using UV/Vis spectroscopy. It has been reported that recombinant AXS 1 from A. thaliana was isolated with NAD tightly bound, and that it showed detectable levels of activity without the addition of exogenous cofactor.” The UV spectrum of the recombinant AXS1 prepared in this study showed a strong absorbance band at 355 nm indicative of bound NADH with a chromophore that is slightly red-shifted from that of free NADH (Figure  3•3)•147  Sodium borohydride was added to convert any bound NAD into NADH. However, only a 10% increase in absorbance was observed, indicating that the bound cofactor was present as a 9:1 ratio of NADH:NAD.  111  0.8 0.7 0.6  0  with NaBH 4 adde44  CA 03 0.2  AXS1 only  0:1  0 290  rrr 440 340 390 490  WàveIéngth(ñiii) Figure 3.3 Partial UV spectra of 147 iM AXS1 in 20 mM Tris-HC1 pH 8.0 before and after the addition of sodium borohydride (to give 1.2 mM sodium borohydride).  For further studies with the enzyme, the catalytically relevant oxidized form of the tightly bound cofactor was desired. Several attempts were made to generate a sample of the enzyme containing an NAD cofactor. As reported previously, an extended incubation of AXS1 with NAD did not lead to the exchange of the cofactor bound to the active site.” 1 Most of the SDR family enzymes that employ transient oxidation mechanisms bind tightly to the cofactor and only release it upon protein denaturation. In order to facilitate exchange of the bound cofactor, the enzyme was incubated in a 100 mM Tris-HC1 buffer containing 8 M urea for 30 minutes. It was anticipated that this would result in partial denaturation of the protein and in cofactor release. The resulting mixture was then dialyzed against a buffer solution containing 20 mM Tris-HC1 and 1 mM NAD over a period of 112  24 hours.’ 48 Following buffer exchange and concentration, the UV spectrum of reconstituted AXS 1 was measured and showed no significant absorbance at wavelengths above 310 nm. Upon the addition of sodium borohydride, absorbance bands at 354 nm and 425 nm were observed (spectra not shown), which were consistent with the generation of tightly bound NADH. The 354 nm band corresponds to the biologically relevant 1,4-reduction product, whereas the 425 nm band corresponds to the 1,2- and  1,6-reduction products, as previously reported.’ 49  Unfortunately, the reconstituted AXS 1 showed no enzymatic activity when incubated with UDP D-glucuronic acid despite repeated attempts. The reason for the lack of activity in an apparently folded protein is not clear, but the loss of the activity observed upon storage of AXS 1 is consistent with the loss that occurred upon extended incubationldialysis in these studies. As a result, the initially isolated AXS 1 having 90% of the cofactor in the wrong oxidation state was used in all experiments described below.  3.3 Testing the Activity of AXSI 3.3.1  Monitoring the Activity of AXS1 Using ‘H and 31 P NMR Spectroscopy In the previous studies, only limited amounts of AXS 1 were available, and the product  UDP-Api had never been directly characterized (only degradation products were characterized to support apiose formation). With the recombinant AXS 1 enzyme in hand, I could test its activity 0 and containing 1 2 by incubating it with UDP-G1cA, in phosphate buffer at pH 8.0 prepared in D mM NAD. Even though the enzyme already has a tightly bound cofactor, it was reported that addition of NAD to the incubation mixture helps to improve enzyme stability. ” The progress 1 of the reaction was monitored using ‘H NMR spectroscopy (Figure 3.4). The initial ‘H NMR spectrum showed the anomeric proton signal of UDP-GlcA at 5.49 ppm in the absence of  113  enzyme (Figure 3.4, t= 0 mm). This signal appears as a doublet of doublets due to coupling to the f3-phosphorus atom and H-2”. Initial attempts to observe activity by adding 2-5 mg of enzyme showed only partial reaction even after extended incubation times, indicating that the recombinant enzyme had very low levels of activity. Low activity levels of recombinant AXS 1 were also reported in the literature.” The low activity of AXS 1 could be due to N-terminal modification introduced during the preparation of the recombinant His-tagged protein. It is also conceivable that the true substrate is not UDP-GlcA, but a structurally related sugar nucleotide. However this is unlikely, since no other branched chain sugar nucleotides are known. A more likely scenario is that AXS 1 prepared using E. coli lacks the post-translational modifications necessary for optimal activity. Of course, the fact that only 10% of the protein contains cofactor in the correct oxidation state exacerbates this problem. A further difficulty is that UDP-Api is known to decompose upon extended incubation,’ 50 and therefore it was not possible to detect it in this manner. In order to circumvent these problems, a relatively large amount of enzyme (150 mg) was added to the sample and a new spectrum was recorded immediately. The NMR spectral time course revealed the appearance of signals corresponding to UDP-Api and UDP-Xyl, with full conversion to products after 15 minutes (Figure 3.4). A new signal at 5.58 ppm is tentatively assigned to the UDP-Api anomeric proton. The signal is a doublet of doublets, but shows up as a triplet due to the similar coupling constants between  JH-1”,H-2”  and  JH-l”,P.  The other doublet of  doublets at 5.42 ppm is assigned to UDP-Xyl by its comparison to a sample of commercially available material. This demonstrates that the same enzyme catalyzes the formation of both UDP-Api and UDP-Xyl and allowed for the first spectral characterization of UDP-Api. Control reactions lacking either the enzyme or UDP-GlcA did not produce any products under similar conditions. The ratio of the two products was found to be approximately 1:1. It has been reported  114  that the ratio of UDP-Api to UDP-Xyl varies with the ions present in the solution, with phosphate buffer having the highest ratio of UDP-Api to UDP-Xyl.’° 2 As a result, all of the experiments were performed in phosphate buffer with the exception of those monitored by 31 P NMR spectroscopy. After ten hours, the spectrum showed full decomposition of UDP-Api, while no degradation was observed for UDP-Xyl (Figure 3.4, t  10 h). Since the recombinant AXS1  shows extremely low activity (‘-6.7 x i0 jimol min’mg’) no kinetic studies were performed.  115  AXS1  c? 2 HOH  OUDP  ° 0 H OUDP  OH OH  UDP-D-glucuronic acid  UDP-D-xylose  UDP-D-apiose COOHO\..-’  H  \\  OUDP  t  0 mm  mm  CO 2 HOH  H  5.650  5.600  5.550  5.500  5.450  5.400  5.350  ppm  Figure 3.4 1 H NMR spectra monitoring the reaction of UDP-G1cA with the AXS1 at different time intervals (400 MHz, D 0, 25 °C). 2 Previous studies had shown that UDP-Api is not stable at a weakly alkaline pH and has a half life of about 97 minutes)’ ’ 0  150  The sugar decomposes non-enzymatically to a-D-apio-D  furanosyl 1,2-cyclic phosphate and uridine 5’-monophosphate (UMP, Figure 3.5). The formation ’ and TDP-65 of 1,2 cyclic phosphates at alkaline pH has also been observed with UDP-glucose’ ; however, the rates of cyclization were considerably slower. The newly formed 152 deoxy-L-talose  116  five-membered phosphorus-containing ring prefers to adopt a planar conformation. In both furanose and pyranose rings the C-O bonds at C-i” and C-2” must be in an eclipsing conformation to accommodate the formation of the phosphorus-containing ring. In certain envelope conformations of furanosides the eclipsing of these bonds is already established and, as a result, furanosides can readily undergo such a cyclization. In the case of a pyranoses, the ring must distort, partially or completely, from the more stable “chair” conformation to the less stable “half-chair” conformation, and the cyclization is slow.” 0 In the case of UDP-Api the energy barrier for formation of a 1,2 cyclic phosphate compound is low, therefore rapid decomposition of the sugar is observed.  HOHC  \ 2 HOH +  OH  °L)  -  OH O\ -  ulviP  /  0 /P 0  UDP-D-apiose  c-D-apio-D-furanosyI 1,2-cyclic phosphate  Figure 3.5 Non-enzymatic decomposition of the UDP-Api to a-D-apio-D-furanosyl 1,2cyclic phosphate and UMP. The reaction catalyzed by AXS1 was also monitored by 31 P NMR spectroscopy (Figure 3.6). Before the addition of AXS 1, the 31 P NMR spectrum showed two doublets belonging to the phosphate groups of UDP-GlcA. After the addition of the enzyme and incubation for 15 minutes, all of the starting material was converted to UDP-Api and UDP-Xyl, which displayed 31 P NMR signals that were similar in chemical shift to each other and to those of UDP-G1cA. After extended incubation (10 h), the UDP-Api non-enzymatically decomposed to cL-D-apio-D furanosyl 1,2-cyclic phosphate (19 ppm) and UMP (3 ppm). UMP and UDP-Xyl standards were 117  spiked into the reaction mixture to confirm the identity of these signals. A control reaction lacking AXS1 did not lead to the formation of the products. C 2 H0H  COO AXS1 H0—R H0-\- 0 NAD 0 0-P-01-0U -(5 0 UDP-D-glucuronic acid  +  H0  Th—( OHOH  OUDP UDP-D-xylose  JJDP-D-apiose  UDP-G1cA t  13-P  0 mm  UDP-G1cA cL-P  t= 15 mm  \ 2 H0H OH 0\ t= l0h  /  /PO 0 _  UMP UDP-xyl 13—P  20.0  15.0  10.0  5.0 ppm  0.0  -5.0  UDP-xyl cL—P  -10.0  Figure 3.6 31 P NMR spectra monitoring the enzymatic conversion of UDP-GlcA to UDP Api and UDP-Xyl. UDP-Api decomposes to UMP and 1,2 cyclic phosphate after over night incubation (121.5 MHz, D 0, 25 °C). * Signal corresponds to NAD. 2  ESI-mass spectrometry was also used to monitor the enzymatic reaction and showed the conversion  of UDP-GlcA to UDP-Api and UDP-Xyl. The initial ESI-mass spectrum, taken  before the addition of AXS1, showed a signal at m/z 579 EM-Hf. After the addition of AXS1 a 118  new peak appeared at m/z 535 [M-Hf, corresponding to both UDP-Api and UDP-Xyl. Upon extended incubation the decomposition product UMP (m/z 323 {M-Hf) could be detected. 3.3.2  Preparation of UDP-[U-’ C]-D-Glucuronic Acid 3  In order to provide more direct evidence for the formation of carbon dioxide and a branched-chain sugar in the reaction catalyzed by AXS 1, a fully ‘ C-labeled hexose skeleton was 3 used in the reaction (UDP-[U-’ C1-glucuronic acid). The advantage of this substrate is that the 3 reaction can be directly monitored by ‘ C NMR spectroscopy, and only the hexose skeleton (or 3 products thereof) will be observed. From the changes in the 3 C-’ coupling patterns, the ‘ C C]g1ucuronic acid 3 rearrangement or cleavage of the skeleton will become evident. UDP[Ui was prepared using enzymatic synthesis, starting with commercially available and relatively inexpensive [U-’ C1-D-glucose (Figure 3.7). [U-’ 3 C]-D-glucose was first converted to UDP-[U 3 Cj-D-glucose by the actions of hexokinase, phosphoglucomutase and UDP-glucose 3 ‘ pyrophosphorylase. This enzymatic synthesis has been used previously with unlabeled glucose’ 53 and the activity of all three enzymes was described earlier in this thesis (Section 1.1) as one of the biosynthetic pathways leading to sugar-nucleotides. The transformation begins with phosphorylation  at  C-6  of  C]-D-glucose 3 [U-’  by  hexokinase  and  ATP.  Then  phosphoglucomutase reversibly transfers the phosphate group from C-6 to C-i, generating the a anomer of [U-’ C]-D-glucose 1-phosphate. The second enzyme requires the presence of glucose 3 1 ,6-diphosphate in order to stay active. Finally, the action of UDP-glucose pyrophosphorylase and UTP couples UMP to the resultant [U-’ C]-glucose 1-phosphate to generate the product 3 C]-D-glucose. An alternative method for preparing sugar nucleotide diphosphates 3 UDP-[(J-’ involves chemical synthesis using phosphoromorpholidates, however, the one-step enzymatic preparation is much simpler and less expensive. 154 119  OH HOOH OH  -  hexokinase ATP  ADP  [U- 13 C]-D-glucose  2 3 0P0  1 6 phospho glucomutase  OH glucose-1,6OH diphosphate Cj-glucose3 [U-’ 6-phosphate  1-phosphate UTP  UDP-glucose pyrophosphorylase  OOC HO°  UDP- lucose dehydroge:ase  H  OIJDP  UDP-[U- 13 C]-D-glucuronic acid  C]-D-glucose UDP-[U- 13  Figure 3.7 Preparation of UDP- [U- ‘ C] -D-glucuronic acid. 3  Upon completion of the reaction, all of the enzymes were removed using centrifugal ultrafiltration and the resulting sugar was purified by ion-exchange and size exclusion chromatography. The purified UDP-[U-’ C1-D-glucose was then oxidized with UDP-glucose 3 dehydrogenase (UDPG1cDH) and NAD. Unlike the first three enzymes which were purchased from Sigma, the UDPG1cDH was prepared in the laboratory. The plasmid pGAC 147, which was used previously in this laboratory to express UDPGIcDH from Group A Streptococci, was transformed into E. coil. 64 The cells were grown in the presence of chloramphenicol, and expression was induced by IPTG. After 3 hours of growth, the cells were harvested by centrifugation and stored in the pellet form at -80 °C. When the enzyme was needed, the cells were lysed and the insoluble cell debris was removed by ultracentrifugation. The resulting crude UDPG1cDH was used in the synthesis without further purification. The reaction of UDPG1cDH 120  with UDP-[U-’ C]-glucose and NAD was monitored by negative ESI-mass spectrometry to 3 ensure that all of the starting material was converted to UDP-[U-’ C]-glucuronic acid (m/z 585 3 [M-Hfl. The resulting sugar was purified by ion-exchange chromatography to give the triethylammonium salt of UDP-[U-’ C]-D-glucuronate in 74% overall yield. 3 3.3.3  Testing the Activity of AXS1 Using 13 UDP-IU- Acid C]-Gluduronic The purified UDP-[U-’ C]-glucuronic acid was incubated with AXS1 in a sealed NMR 3  tube and the progress of the reaction was followed by ‘ C NMR spectroscopy (Figure 3.8). The 3 C NMR signals in the region of the spectrum between 85 and 180 ppm were particularly 3 ‘ informative for observing this conversion. The initial ‘ C NMR spectrum, recorded before the 3 addition of AXS 1, presented a signal at 176 ppm corresponding to the carboxylate group. This signal appeared as a doublet due to a coupling between C-5” and C-6” of the sugar. A doublet corresponding to C-i” of UDP-[UC1-glucuronic acid was found at 94.5 ppm (Figure 3.8, t 0 13 mm). After incubation with AXS1, a new signal appeared as a singlet at 160 ppm corresponding to carbon dioxide (Figure 3.8, t 40 mi. If the NMR tube was left open overnight, the signal at 160 ppm disappeared. The spectrum also showed the appearance of two new doublets at 95.2 and 98.2 ppm belonging to C-i” of UDP-Api and C-i” of UDP-Xyl, respectively. This further confirms that AXS1 is able to catalyze the formation of both UDP-Api and UDP-Xyl.  121  Coo t 0 mm 01_lop  0  C-i,,  CO 2 H0H ‘r—r’ouDP  Co 2  t40min  OHOH  C-i,, 4  170  I  I  160  150  140  130  120  110  100  90  ppm Figure 3.8 13 C NMR spectra monitoring the enzymatic conversion of UDP-[U-’ C]-G1cA 3 to UDP-[U-’ C]-Api and UDP-[U-’ 3 C]-Xyl (150 MHz, D 3 0, 25 °C). 2 In order to confirm that the signals attributed to UDP-[UC]-Api were associated with a 13 sugar containing a branched carbon chain, the ‘ C]-Api and UDP 3 C NMR spectra of UDP-[U-’ 3 C]-Xyl were fully assigned. The first step was to assign the signals due to UDP-[U-’ 3 [U-’ C]3 Xyl. This was done by allowing a reaction to proceed to completion and to incubate long enough for the complete conversion of UDP-[U-’ C]-Api into the cyclic phosphate breakdown product. 3 The UMP and the a-D-apio-D-furanosyl 1,2-cyclic phosphate were removed by ion exchange chromatography, and the isolated UDP-[U-’ Cj-Xyl was characterized by ‘ 3 C NMR 3 spectroscopy (spectra not shown). Once the signals that belong to UDP-[U-’ C]-Xyl were 3 assigned, the signals due to UDP-[UC]-Api in the mixture of sugar nucleotides could also be 13 assigned (Figure 3.9). The signal for the key tertiary C-3” carbon of UDP-[U-’ C]-Api appears 3 122  downfield at 76 ppm as a doublet of doublets of doublets due to the coupling constants between 3 ” 2 Jc” ,c, JC-3”,C4” and Jc3”,c5”. 31??  HOH C  HO---R  HO-  “  “ _-  OUDP  UDP  OH OH  C]-Xyl 3 IJDP-[U-’  C]-Api 3 UDP-[U-’  C-i” Xyl  C-3”C-2” Api Api C i” Api C-3”  L,  Xvi  C-3” C-4” Xi  C-4” Api  LJ I  I  90  80  70  ppm  C]-Api and UDP-{U-’ 3 Figure 3.9 13 C]-Xyl (150 MHz, 3 C NMR spectrum of UDP-[U-’ 25 °C). O, 2 D The enzymatic reaction was also monitored by ESI-mass spectrometry and showed the C]-GlcA to UDP-[U-’ 13 C]-Api and UDP-[U-’ 3 conversion of UDP-[UC]-Xyl. The initial ESI 3 mass spectrum, taken before the addition of AXS 1, showed a peak at m/z 585 [M-H]  -.  After the  addition of AXS 1 a new peak appeared at m/z 540 [M-Hf, corresponding to both UDP-[U-’ C]3 C]-Xyl. 3 Api and the UDP-[U-’  123  3.4 Catalytic Competence of UDP-4-ketoxylose 3.4.1  Preparation of UDP-4-ketoxylose The formation of both UDP-Api and UDP-Xyl is thought to occur through a common  intermediate, UDP-4-ketoxylose (Figure 1 •39)•96 Several studies had already been performed on UDP-Api synthase from Petroselinum hortense to provide indirect evidence for the existence of the 4-keto intermediate, however, this compound has never been isolated from the enzymatic reaction. 112 In order to test whether the UDP-4-ketoxylose is catalytically competent to act as ’ 108 an intermediate in the AXS 1 reaction, I wished to incubate it with the enzyme and demonstrate that it could be converted into a mixture of UDP-Xyl and UDP-Api. This process would result in an overall reduction of the UDP-sugar, and therefore would require a stoichiometric amount of the enzyme in the NADH-containing form. Since 90% of our recombinant AXS1 contained bound NADH, it could be used directly in these studies. This would also provide us with a method for generating the NAD-containing form of the enzyme, without the need for denaturation. The chemical synthesis of UDP-4-ketoxylose would certainly pose a difficult challenge, given the variety of sensitive groups in the molecule. Fortunately, this compound is known in nature, and thus an enzymatic synthesis could be employed. UDP-4-ketoxylose is found as an intermediate in the biosynthesis of UDP-N-formyl-4amino-4-deoxy-L-arabinose (Figure 3.10). This sugar-nucleotide is used in the modification of normal lipopolysaccharide structures in E. coli and Salmonella typhimurium.’ ’ It was found that 5 the addition of N-formyl-4-amino-4-deoxy-L-arabinose to lipid A caused the bacteria to be resistant to polymoxin and cationic antimicrobial peptides.’ ’ 56  157  The biosynthesis begins with  the conversion of UDP-glucose to UDP-G1cA by the action of UDP-glucose dehydrogenase. 124  UDP-G1cA is then oxidatively decarboxylated by the bifunctional enzyme AmA to produce the UDP-4-ketoxylose. This activity of AmA is very similar to that of UDP-GlcA decarboxylase (Section 1.8) with the exception that it consumes NAD and does not reduce the C-4” carbonyl following decarboxylation. This 4-keto sugar is subsequently converted to a 4-amino sugar by the pyridoxal phosphate-dependent enzyme ArnB. The biosynthesis is completed by the second activity of AmA that transfers a formyl group to give UDP-N-formyl-4-amino-4-deoxy-Larabinose. Ultimately, this sugar is incorporated into lipid A by the action of several more enzymes. 158  OH  UDP-glucose  HQdehydrogenase  o  -  HO  2+ +co UDP-glucose  UDP-G1cA  NADH  UDP-4-ketoxylose  glutamate  —s  ArnB  a—ketoglutarate AmC  ‘  0 NH  ArnD  Lipid A HO\ OUDP UDP-N-fonnyl-4-amino4-deoxy-L-arabinose  N-b formyltetrahydrofolate  HO—\  ou UDP-4-amino-4-deoxyL-arabinose  Figure 3.10 Biosynthesis of UDP-4-amino-4-deoxy-L-arabinose in polymyxin-resistant E. coli and Salmonella typhimurium.  125  The amA gene was amplified from genomic E. coil K- 12 W3 110 DNA by PCR and cloned using a procedure similar to the one described in Section 3.2.2. The resulting plasmid encoding an N-terminal hexahistidine-tagged AmA was then transformed into competent E. coil. The cells were grown in kanamycin-containing media and protein expression was induced by the addition of IPTG. The cells were harvested, lysed, and the soluble fraction was purified using nickel affinity column chromatography. SDS-PAGE analysis of the purified protein revealed a single band at 77 kDa. That is consistent with the expected, previously reported molecular weight, and indicated a purity of greater than 80 % (Figure 3.1 1).’ 1  2  3  4  66kDA  29kDA  ji  Figure 3.11 SDS-PAGE gel showing AmA purification. Lane 1: molecular weight standards; lane 2: Cell lysate (before IPTG); lane 3: Crude cell lysate 5 h after IPTG induction; Lane 4: Purified AmA. Molecular weight standards BSA (29 kDa) and carbonic anhydrase (66 kDa). UDP-GlcA was incubated with NAD and purified AmA in phosphate buffer (pH 7.5). The progress of the reaction was monitored by negative ESI-mass spectrometry, or more specifically, by monitoring the appearance of the unhydrated UDP-4-ketoxylose sugar (m/z 533 [M-Hf) as well as the hydrated UDP-4-ketoxylose (m/z 551 [M-Hf) that are formed in a 15 to 1 ratio as estimated by the intensity of the ESI-mass spectra signals. In solution, UDP-4-ketoxylose would  126  readily and reversibly react with water to form a hydrated version of the ketone (Figure 3.12 top). Similarly, in the presence of a high concentration of MeOH, a hemiacetal can be reversibly formed (Figure 3.12 bottom), as will be apparent in the later experiments. Once the reaction was complete, the enzyme was removed by centrifugal ultrafiltration and the resulting sugar was purified by ion exchange chromatography. The UDP-4-ketoxylose was characterized by ‘H NMR spectroscopy (Appendix Figure A.2) and the observed values are in agreement with previously published results.’ 5 According to the integration of the ‘H NMR spectra signals, the UDP-4-ketoxylose is found predominantly (50:1) as the hydrated compound in D 0. 2  0 2 H  OuDP hydrated UDP-4-ketoxylose  unhydrated UDP-4-ketoxylose  MeOH OUDP unhydrated UDP-4-ketoxylose  MeO out’p hemiacetal form of UDP-4-ketoxylose  Figure 3.12 Reaction of UDP-4-ketoxylose with H 0 and MeOH. 2 3.4.2  Reaction of UDP-4-ketoxylose with AXS1 The proposed reaction between the reduced form of AXS 1 and UDP-4-ketoxylose is  shown in Figure 3.1 3A. In the first attempt at monitoring this reaction, a large excess (100-fold) of UDP-4-ketoxylose was incubated with AXS 1 and changes in the oxidation state of the enzyme bound cofactor (90% NADH at t 0) were followed by UV spectroscopy (Figure 3.13B). Before addition of the substrate, there was a large absorption peak at 355 nm belonging to the tightly 127  bound NADH. Having most of the AXS 1 enzyme in the NADH form is beneficial for this experiment because the proposed reaction is a net reduction and requires a stoichiometric amount of the reducing agent. No changes were observed immediately following the addition of ketose, indicating that, if the reaction does occur, it is extremely slow. However, after incubating the mixture for 20 hours at 37 °C, there was a significant decrease in the NADH band (-70%) due to reduction of the 4-keto sugar that generates NAD. The excess UDP-4-ketoxylose was removed using centrifugal ultrafiltration, and NaBH 4 was added to the enzyme solution. Immediately after the addition of NaBH , the intensity of the band at 355 nm increased, and a new band at 425 nm 4 appeared. The 425 mm shoulder is due to 1,2- and 1,6-reduction products, which are known to form upon reduction of NAD with NaBH . 4  149  As shown by the U’V spectrum, not all of the  NADH cofactor bound to AXS 1 was oxidized to NAD even though a large excess of substrate was added. As a control reaction, incubation of AXS 1 in phosphate buffer alone was also monitored by UV spectrometry. No decrease in the intensity of the NADH band was observed. The fact that the reaction did not proceed to completion was attributed to a very slow rate of the reaction, combined with a loss of enzyme activity upon prolonged incubation. This kinetic argument is favoured over the thermodynamic argument, since it is expected that reduction of a cyclohexanone functionality by NADH should be a favourable process. This is because the presence of an sp center in a cyclohexyl ring introduces torsional strain into the molecule.’ 2 59 Similar arguments explain why the cyclohexanone is largely hydrated in solution.  128  A HOHCO HO  AXS1 H]  +  +  OUDP  0  UDP-4-ketoxylose  (S1 ADj  OH OH  UDP-Api  UDP-Xyl  B  0.4 AXS1 only  0.35 V  0.3  AXSI afterreaction with UDP-4-ketoxylose for 20 h  ci 025 0.2  AXSI after addition of NaBH 4  015  0 290  340  390  440  490  Wavelength (rim) Figure 3.13 A) Proposed reaction between UDP-4-ketoxylose and AXS 1 [NADH] B) Partial UV spectra of AXS 1 -bound NADH. Spectra were collected in 10 mM phosphate buffer, pH 8.0, at 37 °C with [AXS1] = 70 1 tM and [UDP-4-ketoxylose] = 81 1 iM.  A second experiment was designed in order to detect the sugar nucleotide products of the reaction. In this case a 1:1 molar ratio of enzyme to substrate was employed in order to maximize the percentage of ketone converted to product. The progress of the reaction was followed by diluting aliquots into a 9:1 mixture of H 0:MeOH and analyzing the products by negative ESI 2 mass spectrometry (Figure 3.14). Before the addition of AXS 1, there were three major peaks that 129  belonged to the starting material UDP-4-ketoxylose. The signal at m/z 533 corresponds to the unhydrated version of starting material, the signal at m/z 551 can be assigned to the hydrated version and the signal at m/z 565 can be assigned to the MeOH adduct of the ketone. After AXS1 was added, the mixture was incubated for 20 hours at 37 °C before an ESI-mass spectrum was recorded. A new signal had appeared at m/z 535 [M-Hf that could correspond to either UDP-Xyl or UDP-Api. Since UDP-Api is not stable under prolonged incubation times, the signal was tentatively assigned to UDP-Xyl. The signal at m/z 540 belongs to an NAD fragment. A weak signal corresponding to UMP was observed at m/z 323 [M-Hf, however, no signal corresponding to the cyclic 1,2 phosphate species was detected. Thus, it was not possible to definitively conclude that UDP-Api had been formed during the reaction. Given that only a fraction of the ketone had been converted to products (10-20%), and that the sample was inherently diluted due to the stoichiometric nature of the reaction, the absence of the signal could easily be attributed to a poor signal-to-noise ratio.  130  A  hydrated UDP-44cetoxylose  unhydrated UDP-4-ketoxylose  MeOHadductof TTThD A l. ‘E —T-.el.oy OSC  ‘1’  I  558  533.1  B UDP-Xyl  NAD fragment ‘1’  i’’  it  525  a  a  a  iii  530  a  535  550.9  •  it  540  ‘a  i,  545  ;  550  a  a  iii  555  a  ii  560  a  iiij  a  a  at  565  Figure 3.14 Negative ESI-mass spectra monitoring the incubation of AXS 1 with UDP-4ketoxylose. A) Before addition of NAD and AXS1 enzyme. B) After incubation for 20 hours. The observation that incubation of the enzyme with UDP-4-ketoxylose results in production of NAD and UDP-Xyl supports the notion that the UDP-4-ketoxylose is a true intermediate in the enzymatic reaction. It is somewhat surprising that the rate of the reaction between the intermediate and the enzyme is extremely slow; however, this is not unprecedented. There are several possible reasons for such low turnover of the intermediate. The conformation and/or protonation state of the free enzyme in solution may differ considerably from that which normally binds the intermediate during the course of the enzymatic reaction. This could introduce a significant kinetic barrier towards formation of the productive enzyme-intermediate complex. Alternatively the intermediate may adopt a different form (such as the hydrate of a ketone) when free in solution. The latter case was observed when testing the intermediate of the  131  enzyme ribulose bisphosphate carboxylase that catalyzes the formation of two molecules of 3phosphoglycerate from ribulose 1,5-bisphosphate and CO . The intermediate in the reaction is 22 carboxy-3-keto-D-arabinitol 1,5-bisphosphate (CKABP).’ ° The intermediate was found to exist 6 as a hydrate in the active site of the enzyme, whereas the keto form of CKABP is predominant in solution. It was reported that feeding of CKABP to the activated enzyme resulted in a significantly lower turnover rate as compared to the natural substrate.’ ° It is thought that the 6 keto-CKABP is not a kinetically competent intermediate, because the keto version is not the enzyme-bound intermediate formed during the reaction. The rate of conversion of exogenously added intermediate into product is thought to depend on the rate of the hydration of the ketone in the enzyme active site.’ ’ In the case of AXS 1 it is believed that the keto version of the UDP-46 ketoxylose is bound to the  However, in solution the UDP-4-ketoxylose is  predominantly in the hydrated form.’ 55 Therefore, the rate of conversion of UDP-4-ketoxylose by AXS 1 may depend on the rate of dehydration that generates the keto form of the sugar (either in the active site or in solution). Moreover, the free enzyme would normally contain NAD prior to binding substrate for catalysis. In this case, the free enzyme responsible for reducing UDP-4ketoxylose must contain an NADH cofactor, and this may alter the normal protonation state/conformation of the active site. It is possible that, when the enzyme is bound to the NADH form of the cofactor, the active site is “clamped down” so that UDP-4-ketoxylose cannot enter it. The barrier towards adopting a productive conformation/protonation state may slow the reaction considerably.  3.5 Testing UDP-Xylose as a Potential Substrate Previous studies had indicated that UDP-Xyl could not be converted into UDP-Api by UDP-D-apiose/UDP-D-xylose synthase.  1, 12, 162  This seems counterintuitive since all the steps 132  past the decarboxylation in either of the proposed mechanisms (Figure 1.39 and Figure 1.43) would be expected to be reversible. In order to try to resolve this discrepancy, I decided to perform a reaction between UDP-Xyl and AXS 1. The progress of the reaction was monitored by P and ‘H NMR spectroscopy. After 24 hours of incubation, no UDP-Api or its decomposition 31 products were detected. One possible explanation for the observed results is that the normal product ratio of UDP-Xyl to UDP-Api (approx. 1:1) represents a kinetic ratio of products and does not reflect the relative thermodynamic stabilities of the sugar-nucleotides. If UDP-Xyl is thermodynamically much more stable than UDP-Api, than the conversion of UDP-Xyl to UDP Api would occur more slowly than the conversion of UDP-Api to UDP-Xyl, and the product would not be expected to accumulate. The extremely low activity of recombinant AXS 1 is likely the reason why the interconversion of UDP-Xyl to UDP-Api cannot be detected.  3.6 Attempted Synthesis of UDP-D-Apiose The inability to observe the AXS 1-catalyzed conversion of UDP-Xyl into UDP-Api may be due to the lower thermodynamic stability of the latter compound. If this were the case, the conversion of UDP-Api into UDP-Xyl should be much more readily detected. Therefore, I attempted to undertake the first chemical synthesis of UDP-Api in order to obtain a pure sample of this substrate. It had been reported that UDP-Api has a relatively short half life at 37 °C and pH 8, however, only 6% of a sample of UDP-Api degraded upon storage for 120 days at -20 °C and pH  6.4.110  Moreover, the total synthesis of UDP-galactofuranose, a sugar-nucleotide that has  a similar stereochemical orientation at C-i” and C-2” as UDP-Api (Figure 3.15), had been 63 Both sugars are susceptible to the formation of a 1,2—cyclic phosphate reported.’ decomposition product, yet a 35% yield of UDP-galactofuranose was reported for the coupling  133  reaction of the corresponding sugar 1-phosphate and activated 5’-UMP. This suggests that the synthesis of UDP-Api is possible.  HOHC  HO  o  1—?OUDP  0 HOJSL.\  OHOH  UDP-Api  OH  OH  UDP-galactofuranose  Figure 3.15 Structures of UDP-Api and UDP-galactofuranose. A key aspect in preparing UDP-Api is that the open chain form of D-apiose can cyclize into four possible closed-chain configurations (Figure 3.16).  102  There are three potential ways to  keep the sugar in the erythro form during the synthesis. The first approach is to protect the anomeric hydroxyl group so the open-chain form cannot be produced. The second method involves protecting the C-3’ hydroxyl group so that it cannot be involved in cyclization. Finally the third method is to protect the C-2 and C-3 hydroxyl groups as a cis-fused 5-membered acetal. All of these methods will avoid the formation of the threo products in which the C-2 and C-3’ hydroxyl groups bear a trans relationship, and are used at some point in the following synthetic scheme.  134  C3’ 2 HOH  C 2 HOH  OH +  CHO H  2  C 2 HOH  OH  OH  OH D-eiythro-furanoses  OH  3’ 4  OH 2 CH HOH  D-apiose +  C 2 HOH  OH  C 2 HOH  L-threo-furanoses  Figure 3.16 Open chain form of D-apiose and possible cyclization products. Commercially available D-xylose was converted to the known compound 4 in five steps (Figure  3.l7).164 165  An initial protection with 2-methoxypropene gave aldehyde 1 which, upon  treatment with formaldehyde and base, yielded diol 2 in an aldol-Canizzaro type reaction. Treatment of 2 with dilute acetic acid selectively removed the cyclic acetal from C-4 and C-5, and the addition of sodium meta-periodate in the presence of base led to the formation of the aldose, which rapidly cyclized to the designed erythro-furanose form 3. The anomeric position was then protected as a methyl glycoside by treatment with trimethyl orthoformate in the presence of pyridinium p-toluenesulfonate to give 4 as the mixture of two anomers with the  13-  anomer being the major product. Benzylation of the mixture gave the fully protected compound 5 and the removal of the methoxy group followed by acetylation led to the formation of known compound 6 as a mixture of two anomers) 66 Unfortunately, the selective anomeric deacetylation proved to be problematic. It was found, that once the acetyl group at C-i had been removed, a  135  migration of the acetyl group from C-2 to C-i readily occurred. As a result, the isolated material was a mixture of deacetylated products and 7 could not be obtained in the pure form.  136  OMe Dowex H+  2 CH O, NaOH 2  OH 2 ,OCCH MeC 0-C  I I  9 H 2 Me C± 0-C-H  I I  52%  D-xylose  OH 2 CH  CHO  \  HCO  HC0 2 /CMe H-C—0  I  2 )CMe  CO 2 H 2  H 1  1) HOAc 2) NaTO 4  66%  Bn0  OH 2 CH  OH 2 CH 3 HC(OMe)  1) NaH, DMF  OMe  2) BnBr 70%  O\ /0 C 2 Me 5  OMe  OH PPTS  0\ 0 C 2 Me  0\ 0 C 2 Me  75%  3  4  1) Dowex H 2) Ac 0, pyr 2 61% BnO  hydrazineacetate OAc  BnO_  -  BnO PN(iPr) 2 1)(BnO) 2) H 0 2  OAc OAc  0—P OBn OAc OAc I OBn 8  OAc OAc  6  —  7  1) 2 H Pd/C , 2) NaOMe  H0 1) UMP-N-methylimidazolide 2) N}{ OAc 4  —  0P OH OH 10  HO  0—P OH I OH OH —  OH 9  UDP-Api  Figure 3.17 Proposed synthetic route for the formation of UDP-Api. 137  In order to avoid the deacetylation step, an alternative synthetic route was designed (Figure 3.18). The major modification consisted in the use of benzyl protecting groups for all of the hydroxyl groups in order to eliminate the chance of migration. Compound 3 was treated with sodium hydride followed by benzyl bromide to form 11, which was subsequently deprotected with formic acid to give a previously reported compound 12.166 The remaining benzyl group at C-3’ prevented the formation of any L-threo-furanose structures. All of the hydroxyl groups were then protected using a treatment with benzyl alcohol/HC1, followed by benzyl bromide/NaH to form the fully protected apiose 13 as a mixture of anomers. Generally, benzyl protecting groups are cleaved by hydrogenolysis, however, anomeric benzyl groups can be selectively removed using acid.’ 66 Treatment of 13 with formic acid led to formation of 14 as an a/ mixture of anomers. The ‘H NMR spectrum of the anomerically deprotected 14 is shown in the Appendix as Figure A.3. Sugar 14 was phosphorylated using a phosphoramidite reagent followed by an oxidation to generate 15. The progress of the reaction was monitored by positive ESI-mass spectrometry. All of the starting material was converted to 15 (m/z 703 [M+Na]), however, all attempts to purify the resulting sugar using flash chromatography only led to its decomposition. The use of base-treated silica gel or alumina was also unsuccessful and resulted in the partial or full decomposition of 15 into a variety of different products. It was consequently decided that 1phosphoapiose derivatives are too unstable to synthesize and no further attempts towards the synthesis of UDP-Api were made.  138  OH 2 CH  C 2 BnOH OH  C Me 2  ‘sZOBn 85%  3  BnOH C formic acid  JV’POH  C 2 Me  OHOH  11  12  65%  1) HC1, BnOH 2)NaH,DMF 3)BnBr  BnOH,  BnO•_••o  formi: acid  OBn OBn 15  14  OBn OBn 13  Figure 3.18 Alternative synthetic route for preparation of anomerically deprotected apiose.  3.7 Fluorinated Analogs of UDP-G1cA 3.7.1  Introduction Fluorinated substrate analogs are widely used as mechanistic probes and inhibitors in order  to gather information about the catalytic mechanisms of enzymes.’ 67 They have been employed to investigate various types of reactions including anionic, cationic, oxidation-reduction, and radical-based processes. 168 One advantage of fluorine lies in the fact that it can mimic a hydroxyl 69 It can also act as a leaving group in group due to its similar size and electronegativity.’ reactions involving carbanionic intermediates and as a destabilizing group in reactions involving carbocationic intermediates. One example of the use of fluorinated analogs was reported in the case of CDP-D-glucose 4,6-dehydratase from Yersinia pseudotuberculosis.’ ° The generally 7 accepted mechanism for the reaction catalyzed by 4,6-dehydratase is discussed in Section 1.6.1. 139  Researchers prepared CDP-6-deoxy-6,6-difluoroglucose’ ° and found it to be an irreversible 7 inhibitor, which bound covalently to the active site of the enzyme. It is thought that CDP-6deoxy-6,6-difluoroglucose is a suicide substrate operating by the mechanism shown in Figure 3.19. After the initial oxidation, a deprotonation at C-5” would lead to the elimination of HF, giving a fluoro-enone intermediate. Subsequently, NADH would deliver a hydride to C-6”, generating an enolate ion that would eliminate the second fluoride to give an cz,13 unsaturated ketone. Since the active site now bears NAD, this intermediate could not be reduced further. Instead, it would react with an active site nucleophile, covalently modifying the active site. This proposed mechanism is supported by the facts that fluoride ions were released in solution during the reaction and that mass spectral analysis indicated the formation of a covalent adduct between the enzyme and the inhibitor.’ 70  140  NADH—.  NADH  NAD  F  B:  fl  6Th  HO-j  HO-  HO—  OCDP CDP-6-deoxy-6,6-difluoroglucose  fluoro-enone mtermediate  NAD  NAD  +  H 0  ( \  4  0  -‘a(  —  HO  HO  OH OCDP  B: Jwv  NAD + H’ H  H  s)  0  ,H OH 13 OCDP  wI  a, 13-unsaturated ketone intermediate  Covalent adduct  Figure 3.19 Proposed mechanism for the inactivation of CDP-D-glucose 4,6-dehydratase by CDP-6-deoxy-6,6-difluoro-glucose.’ ° 7 Another example of the use of fluorinated substrate analogs to investigate an enzyme mechanism was reported with 1-deoxy-D-xylulose-5-phosphate (DXP) reductoisomerase, which is a NADPH-dependent enzyme that catalyzes the conversion of DXP to methyl-D-erythritol 4phosphate (MEP, Figure 3.20A).’ ’ There are two proposed mechanisms for the conversion of 7 DXP to MEP and they closely mirror the suggested mechanisms for the rearrangement catalyzed by AXS 1. The first mechanism involves an a-ketol rearrangement (alkyl migration) to produce the methylerythrose phosphate as an intermediate, which is then reduced by NADPH to produce the desired product (Figure 3.20B).  171  The other proposed mechanism involves a retro-aldol  rearrangement to form the common methylerythrose phosphate, which is then reduced to form 141  DXP (Figure 3.20C).  171  In order to try to provide evidence for one of the mechanisms three  fluorinated substrate analogs were prepared (Figure 3 .20D). It was found that the 1 -fluoro-DXP substrate analog acted as a poor substrate, while the 3-fluoro- and 4-fluoro-DXPs were noncompetitive inhibitors. The results suggest that both the C-3 and C-4 hydroxyl groups of DXP were crucial for the catalysis.  142  A reductoisomerase P—OH  C-I 3 H NADPH  NADP  1 -deoxy-D-xylulose-5-phosphate (DXP)  methyl-D-erythritol-4-phosphate (MEP)  B o 3 Hc 3 S  2  0  NADPH  H  4  0—P—OH OH  MEP  OH  methylerythrose phosphate intermediate  C  I  B:  o”)  (o 2 C 3 H  5  1  OHO  0  I  .  II  H OH  OH  oH  0  H B: I  D  oQH FJOf OH  1 -deoxy- 1 -fluoro-D-xylulose-5phosphate  1 ,3-dideoxy-3-fluoro-D-xylulose-5phosphate  1 ,4-dideoxy-4-fluoro-D-xylulose-5phosphate  Figure 3.20 A) Reaction catalyzed by 1-deoxy-D-xylulose-5-phosphate (DXP) reductoisomerase. B) x-ketol rearrangement mechanism. C) Retro-aldol rearrangement mechanism. D) Fluorinated substrate analogs.  The reaction catalyzed by UDP-D-apiose/UDP-D-xylose synthase is proposed to occur either via a retro-aldol mechanism (Figure 1.39) or a carbon migration mechanism (Figure 1.43). 143  In an attempt to gain more insight into the mechanism of AXS1, I prepared UDP-2-deoxy-2fluoro-D-glucuronic acid (UDP-2F-G1cA) and UDP-3-deoxy-3-fluoro-D-glucuronic acid (UDP 3F-G1cA, Figure 3.21) and examined their competence as substrates. For both fluorinated substrates it is expected that the corresponding fluorinated UDP-xylose products could be formed regardless of the mechanism employed, because neither the C-2” hydroxyl nor the C-3” hydroxyl plays a key role in UDP-xylose formation. A productive reaction will indicate that the fluorine substitution did not affect binding or the ability of the enzyme to catalyze the hydride transfer and decarboxylation steps. In the case of UDP-2F-GlcA, the formation of the UDP-2-deoxy-2fluoroapiose would only be observed if the carbon migration mechanism was operative. This is because deprotonation of the C-2 hydroxyl is not required in this mechanism, yet is required for the retro-aldol process. In the case of UDP-3F-GIcA, no UDP-3-deoxy-3-fluoroapiose would be expected to form if either mechanism were at play. However, it may be possible to detect alternative ring-opened by-products formed by a retro-aldol ring opening that leads to the formation of unnatural intermediates.  144  retro-aldol mechanism COO-  HO OUDP UDP-2-deoxy-2-fluoroxylose  OUDP UDP-2F-G1cA carbon migration mechanism  C 2 HOH HO H +  OUDP  OH  IJDP-2-deoxy-2-fluoroxylose  retro-aldol mechanism  -o  UDP-2-deoxy-2-fluoroapiose  HO +  ring opened side products  OUDP lose y-3-fluoroxy UDP-3-deox  HO OHI OUDP  UDP-3F-GIcA  HO carbon migration mechanism  OUDP UDP-3-deoxy-3-fluoroxylose  Figure 3.21 Proposed results of the incubation of fluorinated analogs with AXS 1.  3.7.2  Synthesis of UDP-2-Deoxy-2-flnoro-D-glncuronic acid UDP-2-deoxy-2-fluoro-D-glucuronic acid (UDP-2F-G1cA, 22) was prepared following  the synthetic route shown in Figure 3.22. Compound 20 was prepared according to a previously s.’ The key transformation is the introduction of 72 published procedure with minor modification  145  the fluorine group at C-2” which was accomplished with SelectfluorTM (1-chloromethyl-4-fluoro1 ,4-diazoniabicyclo{2.2.2] octane bis(tetrafluoroborate)), a commonly used fluorinating reagent in carbohydrate chemistry.’ 73 Treating the commercially available glycal 16 with SelectfluorTM produced a mixture of the desired compound 17a and its C-2” epimer 17b. Both sugars were acetylated and separated by flash chromatography to give 18 in 25% yield over two steps. The inversion of configuration at C-i” of 18 was accomplished in two steps by first generating the a glycosyl bromide using HBr, followed by a reaction with silver(I) acetate to produce the desired 3-glycosylacetate 19 in 95% yield. In order to introduce the a-phosphate, a modification’ 74 of the method of MacDonald’ 75 was used. This procedure allowed the stereospecific preparation of the fully acetylated a-phosphate that was immediately deacetylated with LiOH to give the free sugar 1-phosphate 20 in 51% yield. The remaining steps were carried out enzymatically. Sugar 20 was dissolved in Tris-HC1 buffer pH 7.8 containing UTP and UDP-glucose pyrophosphorylase. After completion of the reaction, 21 was purified using ion-exchange chromatography. The purified sugar nucleotide was then oxidized with UDP-glucose dehydrogenase and NAD to yield UDP 2F-G1cA 22 in 67% yield. The final product was fully characterized by ‘H NMR spectroscopy (Appendix Figure A.4) and negative ESI-mass spectrometry (m/z 581 [M-Hfl.  146  OAc  OAc  AcO Selectfluor  AcOO AcO—s  0, DMF 2 H  x  AcOI AcO  16  0, 12 2 Ac  AcO Ac  25% over 2 steps 18  17a X=H,Y=F 17b X=F, Y=H  1) r, AcOH 2) AgOAc, ACN  OH  UDP-glucose pyrophosphorylase F  21  NAD 67%  OUDP  PPi  95%  OH HO H  HO  OAc  UTP  20  1) 4 P0 3 H 2) 2M L1OH 3) NEt 3  2 3 OPO 51%  AcOOAc AcO F 19  UDP-glucose dehydrogenase  COO HO OUDP UDP-2F-G1cA 22  Figure 3.22 Synthesis of UDP-2-deoxy-2-fluoro-D-glucuronic acid 22. 3.7.3  Testing UDP-2F-G1cA as a Substrate Analog Once the 2-fluoro analog 22 was prepared, it was incubated with AXS 1 in phosphate  buffer at pH 8.0 in the presence of 1 mM NAD. The progress of the reaction was monitored by F NMR spectroscopy (Figure 3.23) and negative ESI-mass spectrometry. The initial ‘ 9 ‘ F NMR 9 spectrum, measured before the addition of AXS 1, showed a single fluorine signal at -201.16 ppm assigned to the fluorine atom of UDP-2F-G1cA (Figure 3.23, t= 0). After incubation with AXS1,  147  a new signal appeared at -201.37 ppm corresponding to either UDP-2-deoxy-2-fluoroxylose (IJDP-2F-Xyl, Figure 3.23, t= 12 h) or UDP-2-deoxy-2-fluoroapiose (UDP-2F-Api). After further incubation, almost all of the starting material was converted to the new product. These results were confirmed by the ESI-mass spectrometric analysis. Before the addition of enzyme, the mass spectrum showed a signal at m/z 581 [M-Hf which belonged to the starting material UDP-2F-GJcA. After 24 hours of incubation with AXS 1, all of the starting material was converted to a product with an MS signal at m/z 537 [M-Hf, which corresponded to either UDP 2F-Xyl or UDP-2F-Api. It should be noted that, unlike UDP-Api, UDP-2F-Api would not be expected to decompose under extended incubation due to the lack of the C-2” hydroxyl. COO =  OUDP  t=12h  tL’+fl  CO 2 HOH  HO  or  OH F  OIJDP  I  I  I  I  -199.0  I  I  I  -200.0  I  I  I  I  -201.0  -202.0  I  N  I  -203.0  I  I  I  I  I  -204.0  ppm  F NMR spectra monitoring the reaction of UDP-2F-G1cA with AXS1 9 Figure 3.23 Partial ‘ 0, 25 °C). 2 over a 24 h period (282.4 MHz, D 148  In order to decide whether the newly formed product was UDP-2F-Xyl or UDP-.2F-Api a sample was purified by ion exchange chromatography and analyzed by ‘H NMR spectroscopy (Figure 3.24). If the isolated product was UDP-2F-Api then we would not expect to see J coupling between H-2” and H-3” because the tertiary C-3” of UDP-2F-Api does not bear a hydrogen. Analysis by 2D COSY NMR spectroscopy (correlated spectroscopy) indicated that there is coupling between H-2” and H-3” (Appendix Figure A.5) in the isolated product, and the analysis of the J values was consistent with a sugar in the xylose configuration. This means that the carbon skeleton remained intact and the isolated product is indeed the UDP-2F-Xyl. No presence of UDP-2F-Api was detected either by 31 F NMR spectroscopies or ESI-mass 9 P!’ spectrometry. The activity of the enzyme towards the substrate analog 22 was estimated to be 100-fold lower than with the natural substrate.  149  4l  H  1 H9O  0UDP  *  H-2”  H-i”  I  I  I  5.50  H-4”  H-3”  I  ‘  5.00  F  I  4.50  I  4.00  I  I  I  3.50  ppm Figure 3.24 Partial ‘H NMR spectrum of isolated product form the reaction between UDP 0, 25 °C). 2 impurity 2F-GJcA and AXS1 (400 MHz, D ‘  A control reaction run without AXS 1 did not lead to the formation of the same mixture. Moreover, a control reaction with the natural substrate UDP-G1cA led to the formation of both UDP-Api (or its cyclic decomposition product) and UDP-Xyl. The fact that UDP-2F-Xyl but not UDP-2F-Api was detected during the reaction between UDP-2F-GlcA and AXS1 (Figure 3.25) supports the retro-aldol mechanism for the carbon skeleton rearrangement (Figure 3.21). The presence of the 2-fluoro group does not prevent C-4” oxidation and decarboxylation, and therefore permits the formation of the UDP-2-deoxy-2-fluoro-4-ketoxylose intermediate. Once the intermediate is formed, it can be reduced by the enzyme to generate the UDP-2F-Xyl. 150  However the carbon skeleton rearrangement cannot be initiated because it requires the deprotonation of the hydroxyl group at C-2” to promote the retro-aldol ring opening process. NAD H0—R H0 NAD  /  NADH  NADH 50:  HO\—°\  22  P-2F-XyI  OUDP  OUDP UDP-2F-G1cA  OUDP  UDP-2-deoxy-2-fluoro4-ketoglucuronic acid  UDP-2-deoxy-2-fluoro4-ketoxylose  IN CO 2 H0H  —(  UDP OHF UDP-2-deoxy-2-fluoro-apiose Not Observed  Figure 3.25 Proposed reaction between UDP-2F-G1cA 22 and AXS1. If the alkyl migration mechanism were at play in this enzyme (Figure 1.43), one might expect that UDP-2F-Api could be still generated since the C-2” is not intimately involved in the rearrangement process. However, the results from this experiment do not completely rule out the alkyl migration mechanism, since it is possible that the presence of the fluorine at C-2” significantly reduces the rate of the alkyl migration relative to the rate of C-4” reduction. Therefore, the next set of experiments focused on UDP-3F-GlcA as a substrate analog.  151  3.7.4  Synthesis of UDP-3-Deoxy-3-fluoro-D-glnduronic acid UDP-3-deoxy-3-fluoro-D-glucuronic acid (UDP-3F-G1cA, 28) was synthesized in seven  steps from D-glucose (Figure 3.26). The synthesis of compound 27 had been previously reported.’ 177 D-Glucose was first protected as the bis-acetonide 23 and then the C-3 hydroxyl ’ 76 was oxidized to give the ketone 24. Next, a stereospecific reduction of 24 gave 1 ,2:5,6-di-Oisopropylidene-c-D-allofuranose (25). It is thought that the access by reducing agent to the bottom face of the ring is hindered by the presence of the 1 ,2-isopropylidene group, therefore 178 The introduction of the fluorine exclusive reduction from the top face of the ring is observed. atom at C-3 to give 26 was accomplished with the use of the fluorinating reagent, DAST (diethylaminosulfur trifluoride), which directly replaces a hydroxyl group with fluorine. DAST is relatively mild and can be used with acid-sensitive compounds. The reaction is believed to occur ’ 79 through a SN2-like displacement with an inversion of stereoconfiguration.’  180  3-Deoxy-3-  fluoro-D-glucose (27) was obtained after removal of both isopropylidene groups under acidic conditions. Compound 27 was then incubated with hexokinase, phosphoglucomutase, and UDP glucose pyrophosphorylase followed by UDP-glucose dehydrogenase, to form sugar-nucleotide 28 in a 35% yield. The final product was purified by ion exchange chromatography and stored as F 9 P and ‘ its triethylammonium salt. It was fully characterized by ‘H (Appendix Figure A.6), 31 NMR spectroscopy. The negative ESI-mass spectrum of UDP-3F-G1cA displayed a signal at m/z U-’ acid 3 UDP-[ ]-3-deoxy-3-fluoro-D-glucuronic C-labeled compound C 581 [M-Hf. The 13 ]-3F-GlcA) was prepared using exactly the same procedure, but starting with [U (UDP-[U-’ C 3 C]-3F-GIcA is shown in Appendix Figure 3 C NMR spectrum of UDP-[U-’ 3 Cj-D-glucose. The ‘ 13 A.7.  152  C><O—l 3 H H C 3 O  o OH  Acetone S0 2 H 4  HOQ HO—\OH  3 CH OH 23 3  55%  D-glucose  o  DMSO 0 2 Ac \  0 24  0  3 CH 3 CH  70% over 2 steps NaBH 4  CO 3 H H C 3 O  C><O 3 H  CO 3 H 0 1 HO  Dowex-50 (Hj EtOH, H 0 2 quant  OH 27  o  DAST C1 2 CH °CH 3 3 CH 26  7-48%  OH 25  -CH 3 3 CH  1) hexokinase, phosphoglucomutase, UDP-glucose pyrophosphorylase, ATP, UTP 2) UDP-glucose dehydrogenase, 35%  COO HO’—Q F-\OUDP UDP-3F-G1cA 28  Figure 3.26 Synthesis of UDP-3 -deoxy-3 -fluoro-D-glucuronic acid 28. 3.7.5  Testing UDP-3F-G1cA as a Substrate Analog Initial studies showed that the reaction of compound 28 catalyzed by AXS 1 was  extremely slow and required near stoichiometric amounts of enzyme to reach completion. However, when a sample of pure compound 28 (3 mg) was incubated with AXS1 (125 mg) in phosphate buffer at pH 8.0, the enzymatic reaction could be monitored by ‘ F and 31 9 P NMR  153  spectroscopies. The initial ‘ F NMR spectrum, taken before the addition of AXS 1, showed a 9 single signal at -200.9 ppm corresponding to the fluorine group of UDP-3F-GlcA (Figure 3.27, t=  0). After incubation with AXS1, a new signal appeared at -122.6 ppm (Figure 3.27, t= 12 h).  Addition of sodium fluoride enhanced this signal, indicating that fluoride was released from 28. After further incubation, all of the starting material was consumed and only one peak corresponding to free fluoride ion was observed (Figure 3.27, t= 24 h). No spectral changes were seen in a control sample of 28 in absence of AXS 1, indicating that the production of fluoride was enzyme- catalyzed. In the 31 P NMR spectra (spectra not shown), the release of free UDP was observed at the same rate as the formation of the free fluoride in the 19 F NMR spectra. UDP-3F-G1cA 28  t=  12 h P  ---— t=  24 h  -120  -130  -140  -150  -160  -170  -180  -190  -200  ppm Figure 3.27 ‘ F NMR spectra monitoring the incubation of UDP-3F-G1cA 28 with AXS1 9 0, 25 °C). 2 in 50 mM potassium phosphate pH 8.0 (282.4 MHz, D  154  The observation of the loss of both UDP and fluoride from compound 28 indicates that a serious structural perturbation of the sugar nucleotide has occurred. It is likely that the enzymatic reaction generated an unstable intermediate which underwent further decomposition in solution. In each experiment the ratio of UDP-3F-GlcA and catalytically active enzyme was around 10 to 1. This means that the enzyme must regenerate the NAD with every catalytic cycle. The result of the incubation with UDP-3F-GlcA is somewhat surprising and it is difficult to explain the release of free fluoride and UDP in the solution with either of the proposed mechanisms. Since only small amounts of product can be obtained in this fashion, and since it will likely be composed of an uncharged pentose (or fragments thereof), analysis of the fate of the glucuronic skeleton poses a serious problem. For this reason, I decided to incubate UDP-[U-’ C]3 3F-GlcA with AXS1 and monitor the reaction by ‘ C NMR spectroscopy. This would have two 3 advantages. Firstly, it would be possible to deduce whether decarboxylation had occurred from the appearance of CO . Secondly, it might be possible to determine whether the carbon skeleton 2 had fragmented into smaller pieces from analysis of 3 C-’ coupling patterns. In the retro-aldol ‘ C mechanism, a similar cleavage occurs and such an observation would support the mechanism. Therefore, AXS1 was incubated with UDP-[U-’ C]-3F-GlcA and the reaction was followed by 3 C NMR spectroscopy (Figure 3.28). Before addition of the enzyme, there was a doublet at 3 ‘ 175.1 ppm that can be assigned to the carboxylate group. After incubation for 10 hours, two new signals appeared as a doublet at 180 ppm and a singlet at 159.7 ppm. The singlet at 159.7 ppm was attributed to carbon dioxide, because it has been observed previously with the normal substrate and it disappeared when the reaction tube was left open overnight. The signal at 180 ppm must belong to a carbon that is still attached to the sugar ring since it appears as a doublet due to J coupling between adjacent isotopically labeled carbons. This is likely due to a C-4”155  oxidized intermediate that has not yet been decarboxylated. Ultimately, all of the substrate underwent decarboxylation as the CO 2 signal at 159.7 ppm was the only peak in the downfield region of the spectrum (Figure 3.28 24 h). Unfortunately, the upfield region of the ‘ C NMR 3 spectra displayed a complex mixture of signals in the 45-120 ppm region and did not provide any further information about the structure of the species that was created during the release of fluoride and UDP. It appears that more than one product is formed during the decomposition of the enzyme-generated intermediate. The progress of the reaction was also followed by ESI-mass spectrometry. Before the addition of enzyme, the mass spectrum showed a signal at m/z 587 C]-3F-GlcA. After 24 hours of 3 which can be assigned to the starting material UDP-[U-’ incubation with AXS 1, all of the starting material was consumed and only a peak at m/z 403 [M Hf that is assigned to UDP could be detected. No other information could be obtained from the mass spectrum. The activity of the enzyme towards 28 was estimated to be 100—fold less than in the case of the unfluorinated substrate. In summary, the treatment of UDP-3F-GlcA with AXS 1 resulted in the catalytic formation of UDP, fluoride, CO 2 and unknown carbon species.  156  C]-3F-G1cA 13 UDP-[U-  t= 0 h  COOj—  t=lOh 2 Co 9  t=24h  I  I  2 CO  I  185.0  I  I  180.0  I  175.0  I  170.0  I  165.0  I  160.0  I  I  I  I  155.0  ppm  Figure 3.28 Partial ‘ C NMR spectra monitoring the reaction of UDP-[U-’ 3 C]-3F-G1cA 3 with AXS1 over a 24 h period (150 MHz, D 0, 25 °C). 2 One possible explanation accounting for the formation of the observed products is shown in Figure 3.29. The reaction is initiated by the oxidation at C-4” to form UDP-3-deoxy-3-fluoro4-ketoglucuronic acid. A fragmentation of the carbon skeleton then occurs before a decarboxylation reaction takes place. This would mimic the ring-opening step of the normal retro-aldol mechanism (Figure 1.39). The open chain enol(ate) intermediate would not be able to close in a manner that generates the apiose skeleton because the fluoride at C-3” does not allow this aldol reaction to occur. Following a rapid enolization, the C-2” aldehyde could then be reduced to regenerate the NAD cofactor and a reduced open chain intermediate that is released 157  into solution. This reduction may mimic the reduction of the UDP-3-aldehydo-apiose intermediate that occurs during normal catalysis (Figure 1.39 and Figure 1.43). The proposed reduction at C-2” is favoured over a potential reduction at C-4” since it leaves the fluoromethyl ketone functionality intact and this could later be hydrolyzed non-enzymatically to generate fluoride. Once in solution, the reduced open-chain intermediate would rapidly undergo decarboxylation and decomposition of the carbon skeleton to form free fluoride ion, UDP, and an unknown carbon skeleton. If this decomposition pathway could be confirmed, it would support the retro-aldol mechanism for the carbon skeleton rearrangement, since that mechanism occurs via an open-chain intermediate. It is interesting to note that, unlike the case of the UDP-2F-GlcA, no evidence for the formation of the UDP-3-deoxy-3-fluoroxylose was obtained. It is expected that, if this product were formed, it would be stable and readily identified. The free sugar, 3deoxy-3-fluoroxylose, is known to be stable in solution,’ ’ 81  182  so even if the hydrolysis of the  glycosidic bond and UDP release occurred, no fluoride ion would be released. It is clear from the production of CO 2 that C-4” oxidation is occurring, however, an alternative step (such as retro aldol ring opening) must take place before decarboxylation and reduction can generate UDP-3deoxy-3 -fluoroxylose.  158  NAD HO  OUDP UDP-3-deoxy-3-fluoroxylose  NAD  NADH  NADH  coo-  HO  3N coo0  Axsl  HO--O\ F—\-  F  OUDP  F UDP-3-deoxy-3-fluoro4-ketoglucuronic acid  UDP-3F-GIcA 28  0  OUDP  Open chain enol(ate) intermediate Enolizatio\\\  NADH  NAD 0  Released from the active site F  HO  OUDP Decomposition  0  COO-  aldehyde reduction  -  F  HO  ourp  Reduced open chain intermediate  F  0  OUDP  Open chain intermediate  UDP + P + carbon skeleton + CO 2  Figure 3.29 Proposed reaction between UDP-3F-G1cA 28 and AXS1.  159  3.8 Conclusions and Future Directions The gene encoding the protein involved in the biosynthesis of UDP-D-apiose was cloned from the cDNA of Arabidopsis thaliana and expressed in E. coli. It was found that 90% of the isolated AXS1 enzyme had the wrong form of the cofactor (NADH) bound in the active site. The incubation of UDP-D-glucuronic acid with AXS 1 led to the formation of a mixture of UDP-D apiose and UDP-D-xylose roughly in a 1:1 ratio and to the release of CO . The progress of the 2 reaction was monitored by ‘H, 31 P and ‘ C NMR spectroscopy. The specific activity of AXS1 3 was very low (—6.7 x 1 0 imol min’mg’) possibly because of absence of any post-translational modifications. As previously reported, UDP-D-apiose decomposed under the incubation conditions to x-D-apio-D-furanosyl 1,2-cyclic phosphate and UMP.”° UDP-4-ketoxylose was prepared using the AmA enzyme that is involved in the biosynthesis of UDP-N-formyl-4-amino-4-deoxy-L-arabinose. The competence of UDP-4ketoxylose to act as a potential intermediate in the AXS 1 reaction was tested. It was found that about 10-20% of UDP-4-ketoxylose was converted to either UDP-Xyl or UDP-Api during the incubation with AXS 1. The slow rate of conversion could be attributed to the fact that UDP-4ketoxylose exists preferentially in the hydrated form in solution whereas in the active site the keto form serves as the true intermediate. Attempts to generate UDP-Api from UDP-Xyl using AXS 1 were unsuccessful, presumably due to the instability of the former compound and the low enzyme activity. Three substrate-based analogs were prepared and tested with AXS 1. It was found that the reaction of UDP-2F-G1cA with AXS1 led to the formation of only UDP-2F-Xyl. This result is consistent with the retro-aldol mechanism, because it requires the presence of the C-2” hydroxyl  160  group to initiate the ring opening. The only products that were identified from the incubation of UDP-3F-G1cA and AXS1 were UDP, free fluoride ion and CO . The structure of the resulting 2 carbon skeleton could not be conclusively assigned based on the obtained data. It is postulated that the enzyme releases a reduced open-chain intermediate into solution, which then decomposes to UDP, free fluoride and the unidentified carbon species (Figure 3.29). This is interpreted as being consistent with a retro-aldol mechanism for the AXS 1 reaction. In the future, the conditions required to obtain a more active version of AXS 1 need to be found. As mentioned in Chapter 2, one way is to try to prepare the AXS 1 enzyme without any modification to the amino acid sequence (such as a His-tag) and to purify the enzyme using traditional chromatographic methods. Another approach would involve the use of cDNA from other plant sources such as Lemna minor and Petroselinum hortense to see if a more active version of the enzyme can be prepared. Finally, the preparation of recombinant AXS 1 using a yeast or insect expression system would be attractive since it may result in the incorporation of post-translational modifications necessary for full activity. Obtaining an X-ray crystal structure of AXS 1 would help to identify key residues in the active site of the enzyme and give insight into their role during the enzymatic reaction. A structure could guide site-directed mutagenesis studies that would help to elucidate the individual roles of the residues. For example, the mutations of a residue involved in the retro-aldol rearrangement should eliminate the ability of AXS 1 to form UDP-Api but would not affect its ability to form UDP-Xyl.  161  3.9 Experimental 3.9.1  Materials and General Methods The previous “Materials and General Methods” from Chapter 2 also applies in this Chapter  with the following additions. The isotopically labeled glucose was purchased from Cambridge Isotopes Laboratories. Dry solvents were distilled fresh, using CaH 22 C1 pyridine, DMSO) (CH , or Nalbenzophenone (THF) as drying agent. ‘ F NMR spectra were obtained on the Bruker 9 AV300 spectrometer at 282.3 MHz. ‘ C NMR spectra were obtained on the Bruker AV600 3 spectrometer at 150 MHz. 3.9.2  Source and Cloning ofAXS1 Approximately 50 mg of leaf material of Arabidopsis thaliana was kindly provided by  Dr. Reinhard Jetter. It was placed in a mortar chilled with liquid nitrogen and ground for several minutes. The tissue powder was transferred into a microcentrifuge tube and the liquid nitrogen was allowed to evaporate. The mRNA of the plant was then isolated using the RNeasy kit 146 The eDNA was prepared by Ortwin Guhling from Dr. Reinhard Jetter’s laboratory. (Qiagen). The AXS] gene (GenelD: 817332) was amplified by PCR using Arabidopsis thaliana cDNA as a template. Oligonucleotide primers, synthesized by the NAPS Unit at UBC, included overhangs for  ligation-independent  GGCGAATGGAGCTAATAGAG-3’  cloning: (forward  sequence,  5 ‘-GGTATTGAGGGTCGCAT AXSJ)  and  5’-AGAGGAGA-  GTTAGAGCCGGAGAGTTTAGGAAGCC-3’ (reverse sequence, AXSJ). A general mixture in the PCR tube included: 5.0 j.iL of lox PCR buffer (Invitrogen), 1.0 iL of 10 mM dNTP mix, 1.5 jiL of 50 mM MgCl , —0.5 jiL of Arabidopsis thaliana cDNA, 25 pmol of each primer, 0.25 pL 2 of 5 U/iL Taq polymerase, and distilled H 0 to a total volume of 50 iL. DNA was amplified 2 162  using an iCycler Thermal Cycler (Bio-Rad) according to the following cycles: one cycle of 3 mm at 94 °C; thirty cycles of 60 s at 94 °C, 60 s at 55 °C, and 90 s at 72 °C; and followed by cooling to 4 °C. The PCR product was cloned into the pET-30 Xa!LIC vector (Novagen) using the ligation-independent cloning method according to the manufacturer’s directions. The resulting plasmid was transformed into NovaBlue Gigasingles chemically competent E. coli cells (Novagen). The presence of the gene was confirmed by DNA sequencing. Over-expression and Purification of AXS1  3.9.3  The recombinant AXSJ plasmid was transformed into E. coli BL21 (DE3) competent cells which were incubated in 10 mL terrific broth (TB) medium containing 12 g/L of bacto tryptone, 24 gIL yeast extract, 4 mL/L glycerol, 2.31 g/L of 4 PO 12.54 gIL of 4 2 KH HPO and 2 K 35 mg/L kanamycin at 37 °C/225 rpm for 10 h. The overnight culture was then poured into 500 600 mL of LB medium containing 35 mg/L kanamycin and shaken at 37 °C/225 rpm until an 0D of 0.6  —  1.0 had been reached. Over-expression was induced with 1mM IPTG. After incubation  for 20 h at 22 °C, the cells were harvested by centrifugation and stored as a pellet at -80 °C. The pellets were re-suspended in 10 mL of a phosphate buffer (20 mM, pH 8.0) containing 1 mM dithiothreitol (DTT), 1 mM NAD, 1 mg/L of aprotinin, and 1 mg/L pepstatin A at 4 °C. The cells were subsequently lysed by passage through a French Pressure cell at 20 000 psi. The lysate was centrifuged at 6 000xg for 1 h, passed through 0.45 tm and 0.22 j.Lm filters, and loaded onto a column containing 10 mL of Chelating Sepharose Fast Flow resin (Pharmacia Biotech), which was previously charged with 100 mM NiSO 4 and washed with sodium phosphate buffer (20 mM, pH 8.0, containing 0.5 M NaCl and 5 mM of imidazole). The purification process was monitored by a Flow Thru UV monitor Spectrometer at 280 nm and was carried out at 4 °C. Nonspecifically bound proteins were washed away by applying buffers containing first 5 mM, and then 125 mM 163  imidazole. Finally, bound enzyme was eluted using a 500 mM imidazole buffer. The fractions containing the desired enzyme were combined and concentrated using Amicon Ultra Centricons (Millipore) and stored on ice until required for use. Typical yields of purified protein were —250 mg/L of cell culture. 3.9.4  Source and Cloning of amA The amA gene (EcoGene Accession Number: EG14091) was amplified by PCR using  Escherichia coli K-12 W3 110 genomic DNA as a template. Oligonucleotide primers, synthesized by the NAPS Unit at UBC, included overhangs for ligation-independent cloning: 5’GGTATTGAGGGTCGCATGAAAACCGTCGTTTTTGCCT-3’ (forward sequence, amA) and 5 ‘-AGAGGAGAGTTAGAGCCTCATGATGGTTTATCCGTAAG-3’ (reverse sequence, amA). The plasmid containing amA was prepared according to procedure described in Section 3.9.2. 3.9.5  Over-expression and Purification of AmA Over-expression and purification of AmA was carried out as described for Pg1F (Section  2.8.5). Typical yields of purified protein were 30 mg/L of cell culture. AmA could be stored frozen in the storage buffer (25 mM Tris-HC1, pH 8.0, containing 10% glycerol) at —80 °C for several months without loss in activity. 3.9.6  Over-expression and Purification of UDP-Glucose Dehydrogenase The pGAC147 plasmid was transformed into E, coli JM1O9 (DE3) cells and inoculated  ’ 64 onto LB-agar plates containing 25 tg/mL chloramphenicol.  183  After overnight incubation at  37 °C, a single colony was used to inoculate 500 mL of TYPG media (8 g bacto-tryptone, 8 g yeast extract, 2.5 g sodium chloride, 1.25 g dibasic potassium phosphate, and 2.5 g D-glucose) containing 25 g/mL chioramphenicol. The cell culture was grown at 37 °C with vigorous 164  600 of 0.7-1.0 IPTG (48 mg) was added at this point to a final shaking (280 rpm) until an 0D concentration of 0.4 mM to induce the over-expression of UDP-glucose dehydrogenase. After 3 h of further growth, the cells were harvested by centrifugation for 15 mm at 5000 rpm and the cell pellet was stored at —80 °C. The cell pellet was later thawed rapidly with warm water and re-suspended in 10 mL cold 50 mM triethanolamine-HC1 (Trien-HC1) buffer, pH 8.7, containing 2 mM DTT, 10% (v/v) glycerol, 1.5 mM phenylmethanesulfonyl fluoride, 1 mg/L pepstatin, and 1 mg/L aprotinin. The cells were subsequently lysed by passage through a French Pressure cell at 20 000 psi. The lysate was centrifuged at 6 000xg for 1 h, passed through 0.45 im and 0.22 jim filters. The crude enzyme was used in reactions without further purifications. 3.9.7  Monitoring Enzyme Incubation by NMR Spectroscopy The general procedure to monitor enzyme incubations by NMR spectroscopy is as follows.  The freshly prepared AXS 1 was exchanged into deuterated buffer (pD 7.4 potassium phosphate, 0) by successive centrifugal filtration steps to affect a 1000-fold dilution of the storage buffer. 2 D The enzyme solution was stored at 4 °C or on ice and used within a few hours. The substrate was dissolved with the same deuterated phosphate buffer and concentrations were measured by UV spectroscopy at 262 nm (molar absorptivity  =  9890 M’ cm 1 in 10 mM potassium phosphate,  pH 7.0). Before the addition of enzyme, a spectrum of substrate only, which served as the zero time point, was measured. Incubations were initiated by addition of the enzyme to the NMR tube. Spectra were taken, initially every 5 mm, then after progressively longer time intervals.  165  3.9.8  AXSI Activity Test with UDP-G1cA In order to determine the activity of AXS 1 under initial velocity conditions, a solution  containing UDP-G1cA (5.2 mlvi) and AXS1 (85 mg) in 10 mM phosphate buffer prepared using 0 containing 1 mM NAD (pD 8.0, 1.0 mL total volume) was placed into an NMR tube and 2 D immediately monitored by ‘H NMR spectroscopy. Spectra were taken every two mm for a period of one hour while incubating at 25°C. The conversion rate was calculated by comparing the integrals of the signals due to the anomeric proton of UDP-GlcA (5.49 ppm) to those of the products UDP-Api and UDP-Xyl (5.76 and 5.42 ppm). The rate of the reaction was determined from data accumulated during the first 15% of the reaction. 3.9.9  Enzymatic synthesis of UDP-[UC]-Glucuronic Acid 13 The enzymatic coupling of UDP to [U-’ C]-glucose was carried out in an incubation 3  mixture of the following composition: [U-’ C]-glucose (25 mg, 134 imol), ATP (227 mg, 412 3 4 (72 mg, tmo1), UTP (200 mg, 412 tmol), glucose-1,6-diphosphate (2.2 fig, 0.165 j.imol), MgSO 292 imol), 70 mM Tris-HC1, pH 7.8, (40 mL), hexokinase (100 units), phosphoglucomutase (210 units), UDP-glucose pyrophosphorylase (25.5 units), and inorganic pyrophosphatase (33 units). The incubation was carried out at 30 °C for 5 h. The progress of the reaction was monitored by 31 P NMR spectroscopy and ESI-mass spectrometry. Once the reaction was complete, the enzymes were removed by centrifugal ultrafiltration and the remaining solution was loaded onto a DE-52 column (220 mL) and eluted with a linear gradient of 0  -  400 mM  triethylammonium bicarbonate buffer (800 mL total). The triethylammonium bicarbonate buffer 2 into 400 mM triethylamine solution over 10 h. The separation was prepared by bubbling CO was monitored by a UV detector (Spectra/Chrom Flow Thru UV monitor controller, Spectrum) Cj-glucose were lyophilized to dryness. 3 with a 254 nm filter. The fractions containing UDP-{U-’ 166  The sugar was then dissolved in 50 mM phosphate buffer (10 mL final volume), pH 8.0, containing 5 mM NAD, 2 mM DTT, and a solution (5 mL) of freshly prepared UDP-glucose dehydrogenase, and incubated at 37 °C for 3 h. The progress of the reaction was followed by ESI-mass spectrometry. The sugar was once again purified by a DE-52 column (220mL) and lyophilized to dryness. ‘ C NMR (D 3 0): 6 176 (d, J 2 ” 1 Jc ’ =46 ,c ”, c-6’ 57 Hz, C-6”); 94.86 (d, 2 5 Hz, C-i”); 70.5-73 (m, C-2”, C-3”, C-4”, C-5”). 31 P NMR (D 0): 6 —10.51 (d, 2  g 1 .JPa, p  20.3 Hz),  —12.51 (m). ESIMS: m/z 585 [M-Hf.  C]-G1cA 13 3.9.10 Studies with UDP-[UC]-G1cA (3.0 mg) 3 The prepared AXS1 (85 mg) was added to a solution of UDP-[U-’ dissolved in 850 jiL phosphate buffer solution pH 8.0 and 1 mM NAD (1.0 mL final volume). C NMR and negative ESI-mass spectra were acquired at 3 The reaction was incubated at 25 °C. ‘ timed intervals. Once the reaction was complete, the mixture was allowed to incubate for another several hours before the enzyme was removed by centrifugal ultrafiltration. The resulting solution was loaded onto a DE-52 column (220 mL) and eluted with a linear gradient of 0 400 -  mM triethylammoniurn bicarbonate buffer (800 mL total). The triethylammonium bicarbonate buffer was prepared by bubbling CO 2 into 400 mM triethylamine solution over 10 h. The separation was monitored by an UV detector (Spectra/Chrom Flow Thru UV monitor controller, C]-Xyl were lyophilized to 3 Spectrum) with a 254 nm filter. Fractions containing UDP-[U-’ 0): 6 96.7 (d, 2 2 dryness and analyzed by ‘ C NMR. ‘ 3 C NMR (D 3 Jc.i”,c 43 Hz, C-i”); 73.9 (m, ” C-3”); 72.6 (m, C-2”); 70.15 (t,J 4 ” 3  ”42.9Hz, 5 Jc”,c-5”43 Hz, C-4”); 61.3 (d,Jc”,c..  P NMR (D 0): 6 —10.71 (d, JPa, Pfl= 20.7 Hz), —12.35 (m). ESIMS m/z 540 [M-Hf. 2 C-5”). 31  167  3.9.11  Enzymatic Synthesis and Isolation of UDP-4-ketoxylose UDP-4-ketoxylose was prepared under conditions slightly modified from those described  55 A purified sample of AmA (7 mg) was added to 50 mL of a phosphate buffer (10 previously.’ mM, pH 7.0) containing 50 mg of UDP-G1cA disodiuin salt and 150 mg of NAD. The solution was incubated for 4 h at 37 °C and the reaction progress was followed by negative ESI-mass spectrometry. It was determined that> 95% of the UDP-GlcA (m/z 581, [M-Hf) was converted to UDP-4-ketoxylose (m/z 551, [M+H 0-Hf) during this time. The enzyme was then removed 2 by centrifugal ultrafiltration and the resultant filtrate was loaded onto a DE-52 column (220 mL) and eluted with a linear gradient of a 0 to 0.5 M triethylammonium bicarbonate buffer. The A 254 of the eluant was monitored, and UV-active fractions were analyzed by a negative ESI-mass spectrometry. Those containing UDP-4-ketoxylose were lyophilized to dryness. The lyophilized sugar (29 mg, 65% yield) was dissolved in 10 mL H 0 and lyophilized again. This procedure 2 was repeated twice more to yield the UDP-4-ketoxylose as its triethylammonium salt. ‘H and 31 P NMR spectra matched those in the literature.’ 55 ‘H NMR (D 0): 6 7.85 (d, IH, J 2 56 ” 2 6); 5.90 (d, 2H, H-5, H-i’); 5.48 (dd, 1H, J,”,  3.3 Hz,  J,”p  3.84 (d, 1H,  =  10.1 Hz, H-3”); 3.58 (dt, 1H, J,” ” 2  5a”,Sb” 3  8.1 Hz, H-  7.0 Hz, H-i”); 4.28 (rn, 2H, H-  2’, H-3’); 4.20 (m, 1H, H-4’); 4.10-4.18 (m, 2H, H-5’); 3.84 (d, 1H, (d, iH, 3 ”, 2 J ”  =  a” 5b” 5 J 12  3.3 Hz, J”,y’  10.1,  12 Hz, H-5a”). 31 P NMR (D 0): 6 —10.24 (d, 2  JPa, pfl=  Hz, H-5b”), 3.72  ‘2”,P  3.1 Hz, H-2”);  20.7 Hz), —12.10 (d,  0-Hf. 2 Jp p 20.7 Hz). ESIMS: mlz 533 [M-Hf and m/z 551 [M+H 3.9.12 Test of UDP-4-ketoxylose as a Potential Intermediate for AXS1 AXS1 (3 mg) was added to a solution of a sample of UDP-4-ketoxylose (4.43 mg, 8.3 imol) dissolved in a solution of sodium phosphate buffer pH 8.0 (1.0 mL final volume). The reaction was incubated at 25 °C and monitored by UV/Vis spectroscopy. The mixture was 168  allowed to incubate for 24 h before the keto substrate was removed by centrifugal ultrafiltration. The resulting enzyme solution was dissolved back to 1 mL and another UV/Vis spectrum was acquired. NaBH 4 (0.03 mg, 750 nmol) in sodium phosphate buffer pH 8.0 (20 tl) was added to the enzyme solution. The UV/Vis spectrum was acquired immediately. In another experiment, AXS1 (202 mg) was added to a solution of UDP-4-ketoxylose (3 mg, 5.6 .tmol) in solution of phosphate buffer pH 8.0. The reaction was incubated at 25 °C and was monitored by negative ESI-mass spectrometry. 3.9.13 AXS1 Activity Test with UDP-Xyl A sample of AXS1 (100 mg) was added to a solution containing UDP-Xyl (3.0 mg, 5.6 tmol) and 1 mM NAD in 10 mM sodium phosphate/D 0 buffer (pD 7.4, 1.0 mL final volume). 2 The reaction was incubated at 25 °C. ‘H, and 31 P NMR spectra were acquired at timed intervals. The mixture was allowed to incubate for another several hours before the enzyme was removed by centrifugal ultrafiltration. A final ‘H NMR spectrum was acquired. 3.9.14 Attempted Synthesis of UDP-Api 10 3.9.14.1.  Methyl 3’-benzyl-2,3-O-isopropylidene--D-erythro-apifuranose 5  Methyl 2,3-O-isopropylidene-J3-D-erythro-apiofuranose 4 was synthesized as a mixture of anomers according to procedure described in literature.  165  Compound 4 (100 mg, 0.489  mmol) dissolved in dry DMF (5 mL) was added to a flask containing sodium hydride (32 mg) in DMF (10 mL). After the addition, the mixture was allowed to cool on ice for 10 mm. The solution of benzyl bromide (140 mg, 100 tL) was added slowly over the course of 10 mm. The reaction was allowed to stir for 5 mm before being tightly sealed. After stirring for 16 h, the mixture was poured into 10 mL of ice cold water and extracted with 4 x 20 mL of ethyl acetate. 169  The solvent was evaporated under reduced pressure and the resulting syrup was redissolved in ether and washed with brine to remove excess DMF. The ether was evaporated under reduced pressure to obtain a crude syrup. Purification by silica gel column chromatography (9:1 petroleum ether/ethyl acetate) gave the compound 5 as a 20:1 mixture of a:j3 anomers (101 mg, 0.0343 mmol) in 70% yield. ‘H NMR (CDC1 ) 5-f3 : 3 (d, 1H,  JPhCH.1, PhCH-2  7.30 (m, 5H, Ph); 4.91 (s, 1H, H-i); 4.63  12.4 Hz, PhCH ); 4.57 (d, 1H, JphcHI 2  PhCH-2  12.4 Hz, PhCH ); 4.27 (s, 2  1H, 11-2); 3.99 (d, 1H, J=10.i Hz, H-4 or H-3’); 3.85 (d, 1H, J=10.1 Hz, 11-4 or 11-3’); 3.67 (d, ); 1.47 (s, 3 1H, J=iO.i Hz, H-3’ or 11-4); 3.62 (d, 1H, J=10.1 Hz, H-3’ or H-4); 3.29 (s, 3H, OCH )). ESIMS: m/z 317 [M+Na]. 3 3H, C(CH )); 1.38 (s, 311, C(CH 3 3.9.14.2.  3’-Benzyl-1,2,3-O-triacetyl-D-erythro-apifuranose 6  Compound 5 (100 mg, 0.034 mmol) was added to a mixture of water and Dowex 50 H (300 mg, wet). The resulting mixture was heated at reflux for 16 h. The resin was removed by filtration and water evaporated under reduced pressure. The residue was dissolved in a mixture of pyridine (10 mL) and 4-dimethylaminopyridine (20 mg) and cooled on ice for 20 mm. Acetic anhydride (0.3 56 mL, 3.78 mmol) was added, and the resulting mixture was stirred for 16h at rt. The solvent was evaporated under reduced pressure, and the resulting syrup was dissolved in 3 (20 mL) and washed with 0.1 N HC1, NaHCO CHC1 3 and brine. The chloroform was evaporated under reduced pressure to obtain a crude syrup. Purification by silica gel colunm chromatography (3:7 petroleum ether/ethyl acetate) gave the compound 6 (76 mg, 0.021 mmol) in a 61% yield over 2 steps. Compound 6 was isolated a mixture of two anomers with 13 being the 66 ESIMS: m/z 389 [M+Na]. dominant form. ‘H NMR spectrum matched that in the literature.’  170  3.9.14.3.  3’-Benzyl-D-erythro-apifuranose 12  The known compound  3164  was converted to known compound 11 with a modification of  66 Compound 3 (100 mg, 0.526 mmol) dissolved in dry DMF (5 a previously reported procedure.’ mL) was added to a flask containing sodium hydride (40 mg) in DMF (10 mL). After the addition, the mixture was allowed to cool on ice for 10 mm. A solution of benzyl bromide (287 mg, 194 jL) was added slowly over the course of 10 mm. The reaction mixture was allowed to stir for 5 mm before being tightly sealed. After stirring for 16 h, the mixture was poured into 10 mL of ice cold water and extracted with 4 x 20 mL of ethyl acetate. The solvent was evaporated under reduced pressure and the resulting syrup was redissolved in ether and washed with brine to remove excess DMF. The ether was evaporated under reduced pressure to obtain a crude syrup. The syrup was purified by silica gel column chromatography (9:1 Petroleum ether/ethyl acetate). Compound 11 (100 mg, 0.027 mmol) was dissolved in 80% formic acid (25 mL), and heated at 60 °C for 2 h. The solvent was evaporated under reduced pressure. Traces of acid were removed by co-evaporating with toluene under reduced pressure, Compound 12 (53.8 mg, 0.024 mmol, ) of 12-cL: 6 7.32 (m, 5H, 3 83% yield) was isolated as 8:1 mixture of c:f3 anomers. ‘H NMR (CHC1 ); 4.01 (d, 1H, J10.0 Hz, H-4 or H 2 Ph); 5.24 (d, 1H, J , 2= 4.56 Hz, H-i); 4.56 (s, 2H, PhCH 1 3’); 3.93 (d, 1H, J,, = 4.56 Hz, H-2); 3.86 (d, iH, Ji0.0 Hz, H-4 or H-3’); 3.53 (d, iH, J 9.5 2 Hz, H-3’ or H-4); 3.50 (d, 1H, J= 9.5Hz, H-3’ or H-4); 1.6 (s, OH). ESIMS: m/z 263 [M+Na]. 3.9.14.4.  1,2,3,3’-Tetrabenzyl-D-erythro-apifuranose 13  Compound 12 (100 mg, 0.042 mrnol) was dissolved in benzyl alcohol (25 mL) and a catalytic amount of HC1 (300 iL). The mixture was heated to 70 °C and stirred for 16 h, at which time benzyl alcohol was removed by distillation at reduced pressure. The resulting syrup 171  dissolved in dry DMF (5 mL) was added to a flask containing sodium hydride (40 mg) in DMF (10 mL). After the addition, the mixture was allowed to cool on ice for 10 mm. The solution of benzyl bromide (287 mg, 194 p.L) was added slowly over the course of 10 mm. The reaction was allowed to stir for 5 mm before being tightly sealed. After stirring for 16 h, the mixture was poured into 10 mL of ice cold water and extracted with 4 x 20 mL of ethyl acetate. The solvent was evaporated under reduced pressure and the resulting syrup was redissolved in ether and washed with brine to remove excess DMF. The ether was evaporated under reduced pressure to obtain a crude syrup. Purification by silica gel column chromatography (7:1 petroleum ether/ethyl acetate) gave the compound 13 as a 12:1 mixture of a:f3 anomers (139 mg, 0.027 mmol) in 65% yield over three steps. ‘H NMR (CHC1 ) of l3-CL: 6 7.37-7.2 1 (m, 20H, Ph); 5.25 3 (d, 1H, J,, = 2.41 Hz, H-i); 4.76-4.44 (m, 8H, PhCH 2 ); 4.17 (d, 1H, J= 10.3 Hz, H-4 or H-3’); 2 4.09 (d, 111, J= 10.3 Hz, H-4 or H-3’); 4.01 (d, 111, J,, = 2.41 Hz, H-2); 3.69 (d, 1H, J 10.2 Hz, 2 H-3’ or H-4); 3.61 (d, 1H, J 10.2 Hz, 11-3’ or H-4). ESIMS: m/z 533 [M+Naj. 3.9.14.5.  2,3,3 ‘-Tribenzyl-D-erythro-apifuranose 14  Compound 13 (100 mg, 0.020 minol) was dissolved in 80% formic acid (25 mL), and heated at 60 °C for 2 h. The solvent was evaporated under reduced pressure. Traces of acid were removed by co-evaporating with toluene under reduced pressure. Compound 14 was isolated as 10:1 mixture of a:13 anomers (75 mg, 0.018 mmol, 91% yield). ‘H NMR (CHC1 ) of 14-a: 6 7.353 7.20 (m, 1511, Ph); 5.25 (d, 1H,J 2.39 Hz, H-i); 4.71-4.41 (m, 6H, PhCH 12 ); 4.17 (d, 1H, 2 J=10.3 Hz, H-4 or H-3’); 4.08 (d, 111, J, =i0.3 Hz, H-4 or H-3’); 4.01 (d, iH, J=2.4 Hz, H-2); 2 3.68 (d, iH, J=10.i Hz, H-3’ or H-4); 3.60 (d, 1H, J=iO.2 Hz, H-3’ or H-4). ESIMS: m/z 443 [M+Na].  172  3.9.14.6.  Dibenzyl 2,3,3’-Tribenzyl-D-erythro-apifuranose phosphate 15  84 and slightly modified. Dibenzyl N,N This procedure is taken from the literature’ diethyiphosphoramidite (2.50 mmol, 88OgiL) was added to a solution of compound 14 (100 mg, C1 The progress of the 2 0.023 mmol) and 1,2,4-triazole (72 mg, 4 mmol) in 5.0 mL dry CH reaction was followed by ESI-mass spectrometry. After stirring at rt for 2 h all of the starting material was converted to product and 30 mL diethyl ether was added. The organic solution was washed with saturated sodium bicarbonate solution (3 x 15 mL) and NaCl brine (3 x 10 mL). The organic layer was dried over sodium sulphate and evaporated to give an oil under reduced 0 (1.9 mL) 2 pressure. The oil was dissolved in 9 mL T}{F and cooled to —78 °C, then 30% H was added and the solution was allowed to warm to ft over 2 h. Diethyl ether (30 mL) was added and the organic solution was washed with saturated sodium bicarbonate solution (3 x 15 mL) and NaC1 brine (3 x 10 mL). The organic layer was dried over sodium sulphate and evaporated to give an oil under reduced pressure. Attempts to purify compound 15 by column chromatography were unsuccessful and no desired product was isolated from this procedure. ESIMS: m/z 703 [M+Na]. 3.9.15  Chemoenzymatic Synthesis of UDP-2F-G1cA 22  a-i -Phospho-2-deoxy-2-fluoroglucose 20 was synthesized according to literature ’ 175 A variation from these procedures was in the phosphorylation of 19, 72 described procedures.’ forming compound 20 (Figure 3.22). The neat mixture was heated to 65 °C to ensure full conversion to the desired product. Compound 20 (25 mg, 0.096 mmol) was dissolved in 70 mM Tris-HC1, pH 7.8, (40 mL) containing UDP-glucose pyrophosphorylase (25.5 units), UTP (200 4 (72 mg, 292 pmol). The incubation was carried out at 30 °C for 12 h. The mg) and MgSO 173  progress of the reaction was monitored by 31 P NMR spectroscopy and ESI-mass spectrometry. Once the reaction was complete, the enzymes were removed by centrifugal ultrafiltration and the remaining solution was loaded onto a DE-52 column (220 mL) and eluted with a linear gradient of 0  -  400 mM triethylammonium bicarbonate buffer (800 mL total). The triethylammonium  bicarbonate buffer was prepared by bubbling CO 2 into 400 mM triethylamine solution over 10 h. The separation was monitored by an UV detector (Spectra/Chrom Flow Thru UV monitor controller, Spectrum) with a 254 nm filter, The fractions containing UDP-2-deoxy-2-fluoro-Dglucose 21 were lyophilized to dryness. Compound 21 was then dissolved in 50 mM phosphate •buffer (10 mL), pH 8.0, containing 5 mM NAD, 2 mM DTT, and freshly prepared UDP-glucose dehydrogenase (700 mg of crude protein extract/mL of lysis buffer), and incubated at 37 °C for 3 h. The progress of the reaction was followed by ESI-mass spectrometry. The sugar was purified by a DE-52 column as described above, and fractions containing compound 22 were lyophilized O): 6 7.87 (d, 1H, J 2 6 , 5 to dryness. ‘H-NMR (D (dd, 1H,  ”,p 1 J  =  7.98 Hz, 2 ”, 1 J ”  2’,3’,4’,5’); 4.06. (d, 1H, J ”, 4 =  =  =  =  8.1 Hz, H-6); 5.92-5.87 (m, 2H, Hi’, H5), 5.70  3.51 Hz, Hi”); 4.45, 4.32 (m, 1H, H-2”); 4.28-4.07 (m, 5H, H 10.2, H-5”), 3.94 (m, 1H, H-3”); 3.45 (t, 1H, J 4 ” 3  10.2 Hz, H-4”). 31 P NMR (D 0): 6 —9.17 (d, 2  JPa, pfi  19.7 Hz), —10.97 (d,  JPa P/3  =  9.7, 5 ”, 4 J ” 19.7 Hz).  F NMR (D 9 ‘ O): 6 -201.18 (s). ESIMS: m/z 581.1 [M—Hf. 2 3.9.16 AXS1 Incubation with UDP-2F-G1cA A sample ofAXS1 (75 mg) was added to a solution of UDP-2F-.G1cA (3.0 mg, 5.0 .tmol) dissolved in solution of sodium phosphate buffer pH 8.0 and 1 mM NAD (1.0 mL final volume). The reaction was incubated at 25 °C. 31 P and ‘ F NMR and negative ESI-mass spectra 9 were acquired at timed intervals. After 24 h, all the starting material was converted to UDP-2FXyl (m/z 537 [M-Hf) and the enzyme was removed by centrifugal ultrafiltration. The resultant 174  filtrate was loaded onto a DE-52 column (220 mL) and eluted with a linear gradient of a 0 to 0.5 M triethylammonium bicarbonate buffer. The A 254 of the eluant was monitored, and UV-active fractions were analyzed by negative ESI-mass spectrometry. Those containing UDP-4ketoxylose were lyophilized to dryness. The lyophilized sugar was dissolved in 10 mL H 0 and 2 lyophilized again. This procedure was repeated twice more to yield the UDP-2F-Xyl as its triethylammonium salt. ‘H-NMR (D 0): (5 7.86 (d, 1H, J 2 ,6 5 Hi’, H5), 5.61 (dd, 1H,  ”,p 1 J  =  7.38 Hz, Ji, ’ 2  8.1 Hz, H-6); 5.89-5.82 (m, 2H,  3.51 Hz, Hi”); 4.38, 4.21 (m, IH, H-2”); 4.28-  4.07 (m, 51-I, H-2’,3’,4’,5’); 3.86 (m, 1H, H-3”); 3.63-3.51 (m, 2H, H-4”, H-5”); 3.43 (m, iH, H5”). 31 P NMR (D 0): (5—9.27 (d, 2  JPa, P/3  =  19.6 Hz), —10.87 (d,  pj  =  19.6 Hz). 19 FNMR  0): (5 -201.36 (s) ESIMS: m/z 537.1 [M—Hf. 2 (D 3.9.17 Chemoenzymatic Synthesis of UDP-3F-G1cA 28 3-Deoxy-3-fluoro-D-glucose was prepared according to a previously published synthetic ’ 176 route.  177  A variation in this procedure was in installation of the fluorine group to form  compound 26 (Figure 3.26). Compound 25 (400 mg, 1.53 mmol) was dissolved in 6 mL dry C1 in a flame-dried flask and the solution was chilled under argon to —40 °C in a dry CH 2 ice/acetonitrile bath. DAST (0.19 mL, 1.4 mmol) was added slowly via syringe and the solution was stirred at —40 °C for 1 h, and then allowed to warm to rt over 5 h. The solution was then cooled to —10 °C and 2.5 mL of methanol was added. The solvent was then removed under reduced pressure and the crude mixture was separated by silica gel chromatography. Silica gel column chromatography (9:1 petroleum ether/ethyl acetate) yielded 195 mg (48% yield) of i,2:S,6-di-O-isopropylidene-3-deoxy-3-fluoro-D-glucose 26. Compound 28 was prepared from compound 27 by the conditions described in Section 3.9.15. ‘H-NMR (D 0): (57.87 (d, ill, J 2 6 , 5 =  8.1 Hz, H-6); 5.92-5.87 (m, 2H, H-i’, H-5), 5.70 (m, 1H, H-i”); 4.52 (dt, 1H, J ”3 2 ”=J 4 ”  =  175  9.22 Hz, J ”, F 3  =  54.1 Hz, H-3”); 4.28-4.07 (m, 5H, H-2’,3’,4’,5’); 4.06 (d, 1H, 5 ”, 4 J  =  10.34 Hz,  H-5”), 3.80-3.64 (m, 2H, H-2”, H-4”). P NMR (D 31 0): 6 —10.75 (d, Jp p= 20.4 Hz), —12.67 (d, 2 Jpa, Pfl=  20.4 Hz). ‘ F NMR (D 9 0): 6 -200.993 (s) ESIMS: m/z 581.1 [M—Hf. 2  3.9.18 Chemoenzymatic Synthesis of UDP- [UC]-3F-G1cA 13 C]-3F-GlcA was prepared from [U-’ 3 UDP-[U-’ C]-D-glucose using the same procedure 3 that was used to prepare UDP-3F-G1cA 28. ‘ C NMR (D 3 0): 6 175.1 (d, 6 2 ”= 5 Jc ’ c 60.3 Hz, C6”); 94.8 (d, Jci”, c-2’  51.39 Hz, C-i”); 93.07 (dt, Jc.2”, C-3” Jc” ,C-4” 43.2 Hz, 3  Hz); 71.81 (m, C-2”); 69.54 (m, C-4”,C-5”). ESIMS: m/z 587.1 [M  —  JC-3”, F”  83.2  Hf.  3.9.19 AXS1 Incubation with UDP-3F-G1cA and UDP-[UC]-3F-G1cA 13 A sample of AXS1 (125 mg) was added to a solution of UDP-3F-G1cA (3.0 mg, 5.0 jtmol) dissolved in sodium phosphate buffer pH 8.0 and 1 mM NAD (1.0 mL final volume). The reaction was incubated at 25 °C. 31 P and ‘ F NMR and negative ESI-mass spectra were acquired 9 at timed intervals. Once all the starting material was converted to UDP and free fluoride, the enzyme was removed by centrifugal ultrafiltration and 31 F NMR spectra were acquired again. 9 P/’ C]-3F-GIcA and the reaction was monitored 3 A similar procedure was followed with UDP-[U-’ by ‘ C NMR spectroscopy. 3  176  Bibliography  1.  Dwek, R. A., Chem. Rev. 1996, 96, 683-720.  2.  Tanner, M. E., Curr. Org. Chem. 2001, 5, 169-192.  3.  Heidlas, J. E., Williams, K. W., and Whitesides, G. M., Acc. Chem. Res. 1992, 25, 307314.  4.  Gabriel, 0., and van Lenten, L., Biochemistry of Carbohydrates. University Park Press: Baltimore: 1978; Vol. 16, p 1-36.  5.  Liu, H.-w., and Thorson, 3. S., Annu. Rev. Microbiol. 1994, 48, 223-256.  6.  Tanner, M. E., Ace. Chem. Res. 2002, 35, 237-246.  7.  Oppermann, U., Filling, C., Hult, M., Shafqat, N., Wu, X., Lindh, M., Shafqat, J., Nordling, E., Kallberg, Y., Persson, B., and Jornvall, H., Chem. -Biol. Interact 2003, 143144, 247-253.  8.  Jomvall, H., Persson, B., Krook, M., Atrian, S., Gonzalez-Duarte, R., Jeffery, 3., and Ghosh, D., Biochemistry 1995, 34, 6003-6013.  9.  Jörnvall, H., Danielsson, 0., Hjelmqvist, L., Persson, B., and Shafqat, 3., Adv. Exp. Med. Biol. 1997, 372, 28 1-294.  10.  Oppermann, U., Filling, C., Berndt, K. D., Persson, B., Benach, 3., Ladenstein, R., and Jömvall, H., Biochemistry 1997, 36, 34-40.  11.  Benach, 3., Atrian, S., Gonzales-Duarte, R., and Ladenstein, R., J Mol. Biol. 1999, 289, 335-355.  12.  Filling, C., Berndt, K. D., Benach, J., Knapp, S., Prozoroviski, T., Nordling, E., Ladenstein, R., Jörnvall, H,, and Oppermann, U., J Biol. Chem. 2002, 277, 25677-25684.  177  13.  Kavanagh, K. L., JOrnvall, H., Persson, B., and Oppermann, U., Cell. Mol. Life Sci. 2008, 65, 3895-3906.  14.  Leoir, L. F., Arch. Biochem. Biophys. 1951, 33, 186-190.  15.  Frey, P. A., FASEBI 1996, 10, 461-470.  16.  Maxwell, E. S.,J Biol. Chem. 1957, 229, 139-15 1.  17.  Walden, P., Ber. 1895, 28, 2766-2773.  18.  Thoden, J. B., Frey P. A., and Holden, H. M., Biochemistry 1996, 35, 5137-5144.  19.  Wilson, D. B., and Hogness, D. S., I Biol. Chem. 1969, 244, 2132-2 136.  20.  Anderson, L., Landel, A. M., and Diedrich, D. F., Biochim. Biophys. Acta 1956, 22, 573575.  21.  Nelsestuen, G. L., and Kirkwood, S., I Biol. Chem. 1971, 246, 7533-7543.  22.  Nelsestuen, G. L., and Kirkwood, S., Biochim. Biophys. Acta 1970, 220, 633-635.  23.  Adair, W. L., Gabriel, 0., Ulfrey, D., and Kaickar, H. M., I Biol. Chem. 1973,248, 4635-4639.  24.  Wee, T. G., Davis, J., and Frey, P. A., I Biol. Chem. 1972, 247, 1339-1342.  25.  Wee, T. G., and Frey, P. A., I Biol. Chem. 1973, 248, 3 3-40.  26.  Maitra, U. S., and Ankel, H., Proc. Natl. Acad. Sci. 1971, 68, 2660-2663.  27.  Glaser, L., and Ward, L., Biochim. Biophys. Acta 1970, 198, 6 13-615.  28.  Kang, U. G., Nolan, L. D., and Frey, P. A., I Biol. Chem. 1975, 250, 7099-7 105.  29.  Thoden, J. B., Frey, P. A., and Holden, H. M., Biochemistry 1996, 35, 2557-2566.  30.  Thoden, J. B., and Holden, H. M., Biochemistry 1998, 37, 11469-11477. 178  31.  Wolucka, B. A., Persiau, G., Van Doorsselaere, J. V., Davey, M. W., Demo!, H., Vandekerckhove, J., Van Montagu, M., Zabeau, M., and Boerjan, W., Proc. Nati. Acad. Sc US.A. 2001, 98, 14843-14848.  32.  Barber, G. A., Arch. Biochem. Biophys. 1971, 147, 6 19-623.  33.  Major, L. L., Wolucka, B. A., and Naismith, 3. H., I Am. Chem. Soc. 2005, 127(5 1), 18309-18320.  34.  Wolucka, B. A., and Van Montagu, M., I Biol. Chem. 2003, 278, 47483-47490.  35.  Barber, G. A., .J Biol. Chem. 1979, 254, 7600-7603.  36.  Somers, W. S., Stahl, M. L. and Sullivan, F. X., Structure 1998, 6, 1601-1612.  37.  Sullivan, F. X., Stahl, M., Kumar, R., Kriz, R., Stahl, M., Xu, G. •Y., Rouse, J., Chang, X., Boodhoo, A., Potvin, B., and Cumming, D. A., I Biol. Chem, 1998, 273, 8193.  38.  Ma, B., Simala-Grant, 3. L., and Taylor, D. E., Glycobiology 2006, 16, 158R-184R.  39.  Lau, S. T. B., and Tanner, M. E., I Am. Chem. Soc. 2008, 130, 17593-17602.  40.  Rosano, C., Bisso, A., Izzo, G., Tonetti, M., Sturla, L., De Flora, A., and Bolognesi, M., I Mol. Biol. 2000, 303, 77-91.  41.  Giraud, M.-F., Leonard, G.A., Field, R.A, Berlind, C., and Naismith, J. H., Nat. Struct. Biol. 2000, 7 (5), 398-402.  42.  Graninger, M., Nidetzky, B., Heinrichs, D. E., Whitfield, C., and Messner, P., 1 Biol. Chem. 1999, 274, 25069-25077.  43.  Dong, C., MajOr, L. L., Srikannathasan, V., Errey, 3. C., Giraud, M.-F., Lam, J. S., Graninger, M., Messner, P., McNeil, M. R., Field, R. A., Whitfield, C., and Naismith, 3. H., I Mol. Biol. 2007, 365, 146-159  179  44.  Dong, C., Major, L. L., Allen, A., Blankenfeldt, W., Maskell, D. J., and Naismith, J. H., Structure 2003, 11, 715-723.  45.  Jakimowicz, P., Tello, M., Freel Meyers, C. L., Walsh, C. T., Buttner, M. J., Field, R. A., and Lawson, D. M., Proteins: Struct. Funct. Bioinf 2006, 63, 26 1-265.  46.  Tello, M., Jakimowicz, P., Errey, J. C., Freel-Meyers, C. L., Walsh, C. T., Buttner, M. J., Lawson, D. M., and Field, R. A., Chem. Commun. 2006, 1079-1081.  47.  Sala, R. F., Morgan, P. M., and Tanner, M. E., I Am. Chem. Soc. 1996, 118, 3033-3034.  48.  Morgan, P. M., Sala, R. F., and Tanner, M. E., I Am. Chem. Soc. 1997, 119, 1026910277.  49.  Tanner, M. E., Bioorg. Chem. 2005, 2 16-228.  50.  Murkin, A. S., Chou, W. K., Wakarchuk, W. W., Tanner M. E., Biochemistry 2004, 43, 14290- 14298.  51.  Stamford, N. P. J., Biosynthesis and Degradation. In Glycoscience: Chemistry & Chemical Biology, Fraser-Reid, B. 0., Tatsuta, K., Thiem, J., Coté, G. L., and Flitsch, S., Eds. Springer-Verlag: Berlin, 2001; Vol. 2, pp 1291-1306  52.  Frey, P. A., Complex pyridine nucleotide-dependent transformations. In Pyridine Nucleotide Coenzymes: Chemical, Biochemical and Medical Aspects, 1St ed.; Dolphin, D., Poulson, R., and .Avramovic, 0., Eds. John Wiley & Sons: New York, 1987; Vol. 2B, pp 461-511.  53.  He, X. M., and Liu, H. -w., Annu. Rev. Biochem. 2002, 71, 70 1-754.  54.  Melo, A., Elliott, W. H., and Glaser, L., I Biol. Chem. 1968, 243, 1467-1474.  55.  Glaser, L., and Zarkowsky, H., The Enzymes 1972, 5, 465-480.  56.  Snipes, C. E., Brillinger, G.-U., Sellers, L., Mascaro, L., and Floss, H. G., I. Biol. Chem. 1977, 252, 8113-8117. 180  57.  Golinelli-Pimpaneau, B., Le Goffic, F., and Badet, B., J Am. Chem. Soc. 1989, 111, 3029-3034.  58.  Chen, H., Yeung, S.-M., Que, N. L. S., Muller, T., Schmidt, R. R., and Liu, H.-w., J Am. Chem. Soc. 1999, 121, 7 166-7167.  59.  John, R. A., Comprehensive Biological Catalysis. Academic Press: San Diego, 1998; Vol. II, pp. 173-200.  60.  Obhi, R. K., and Creuzenet, C., I Biol. Chem. 2005, 280, 20902-20906.  61.  Schoenhofen, I. C., Lunin, V. V., Julien, J.-P., Li, Y., Ajamian, E., Matte, A., Cygler, M., Brisson, J,-R., Aurby, A., Logan, S. M., Bhatia, S., Wakarchuk, W. W., and Young N. M., I Biol. Chem. 2006, 281, 8907-89 16.  62.  Noland, B. W., Newman, J. M., Hendle,  J.,  Badger, J., Christopher, J. A., Tresser, J.,  Buchanan, M. D., Wright, T. A., Rutter, M. E., Sanderson, W. E., Muller-Dieckmann, H.-J., Gajiwala, K. S., and Buchanan, S. G., Structure 2002, 10, 1569-1580. 63.  Feingold, D. S., and Franzen, J. S., Trends Biochem. Sd. 1981, 6, 103-105.  64.  Campbell, R. E., Sala, R. F., van de Rijn I., and Tanner M. E., I Biol. Chem. 1997, 272, 34 16-3422.  65.  Ge, X., Campbell, R. E., van de Rijn, I., and Tanner M. E., I Am. Chem. Soc. 1997, 120, 6613-6614.  66.  Bakker, H., Oka, T., Ashikov, A., Yadav, A., Berger, M., Rana, N. A., Bai, X., Jigarni, Y., Haltiwanger, R. S., Esko, J. D., and Gerardy-Schahn, R., I Biol. Chem. 2009, 284 (4), 2576-2583.  67.  Ankel, H., and Feingold, D. S., Biochemistry 1965, 4, 2468-2475.  68.  Ankel, H., and Feingold, D. S., Biochemistry 1966, 5, 182-189.  69.  Schutzbach, J. S., and Feingold, D. S., I Biol. Chem. 1970, 245, 2476-2482. 181  70.  Angata, T., and Varki, A., Chem. Rev. 2002, 102, 43 9-469.  71.  Varki, A., FASEBJ 1997,11,248-255.  72.  Traving, C., and Schauer, R., Cell. Mo?. Life Sci. 1998, 54, 1330-1349.  73.  Edwards, U., Muller, A., Hammerschmidt, S., Gerardy-Schahn, R., and Frosch, M., Mo?. Microbiol. 1994, 14, 141-149.  74.  Annunziato, P. W., Wright, L. F.,Vann, W. F., and Silver, R. P., J. Bacterio?. 1995, 177, 3 12-3 19.  75.  Knirel, Y. A., Vinogradov, E.V., L’vov, V. L., Kocharova,. N.A., Shashkov, A. S., Dmitriev, B. A. and Kochetkov, N. K., Carbohydr. Res. 1984, 133, C5-C8.  76.  Yuki, N., Curr. Opin. Immuno?. 2005, 17, 577-582.  77.  van Amsterdam, K., van Vliet, A. H. M., Kusters, J. G., and van der Ende, A., FEMS Microbio?. Rev. 2006, 30, 131-156.  78.  Goon, S., Kelly, J. F., Logan, S. M., Ewing, C. P., and Guerry, P., Mo?. Microbiol. 2003, 50, 659-671.  79.  Schoenhofen, I. C., McNally, D. J., Brisson, J.-R., and Logan, S. M. , G?ycobio?ogy 2006, 16, 8C-14C.  80.  Schoenhofen, I. C., McNally, D. J., Vinogradov, E., Whitfield, D,, Youn.g N. M., Dick, S., Wakarchuk, W. W., Brisson, J.-R., and Logan, S. M., JBio?. Chem. 2006, 281, 723732.  81.  Fields, B. S., Benson, R. F., and Besser, R. E., C?in. Microbiol. Rev. 2002, 15, 506-526.  82.  Swanson, M. S., and Hammer, B. K., Annu. Rev. Microbio?. 2000, 54, 567-6 13.  83.  Albert-Weissenberger, C., Cazalet, C., and Buchrieser, C., Ce??. Mo?. Life Sci. 2007, 64, 432-448. 182  84.  Tsvetkov, Y. E., Shashkov, A. S., Knirel, Y. A., and Zahringer, U., Carbohydr. Res. 2001, 331, 233-237.  85.  Raetz, C. R. H., and Whitfield, C., Annu. Rev. Biochem. 2002, 71, 635-700.  86.  Knirel, Y. A., Rietschel, E. T., Marre, R., and Zahringer, U., Eur. J. Biochem. 1994, 221, 239-245.  87.  Kooistra, 0., Luneberg, E., Lindner, B., Knirel, Y. A., Frosch, M., and Zahringer, U., Eur. I Biochem. 2002, 269, 560-572.  88.  McNally, D. J., Aubry, A. J., Hui, J. P. M., Khieu, N. H., Whitfield, D., Ewing, C. P., Guerry, P., Brisson, J.-R., Logan, S. M., and Soo, E. C., I Biol. Chem. 2007, 282, 1446314475.  89.  Knirel, Y. A., Grosskurth, H., Helbig, J. H., and Zahringer, U., Carbohydr. Res. 1995, 279, 215-226.  90.  Nazarenko, E. L., Shasbkov, A. S., Knirel, Y. A., Ivanova, E. P., and Ovodov, Y. S., Bioorg. Khim. 1990, 16, 1426-1429.  91.  Luneberg, E., Zetzmann, N., Alber, D., Knirel, Y. A., Kooistra, 0., Zahringer, U., and Frosch, M.,  92.  mt. I Med. Microbiol. 2000, 290, 37-39.  Olivier, N. B., Chen, M. M., Behr, J. R., and Imperiali, B., Biochemistry 2006, 45, 1365913669.  93.  Szymanski, C. M., and Wren, B. W., Nat. Rev. Microbiol. 2005, 3, 225-23 7.  94.  Schoenhofen, I. C., McNally, D. J., Vinogradov, E., Whitfield, D., Young, N. M., Dick, S., Wakarchuk, W. W., Brisson, J.-R., and Logan, S. M., I Biol. Chem. 2006, 281, 723732.  95.  Chien, M., et. al., Science 2004, 305, 1966-1968.  96.  Grisebach, H., and Schmid, R., Angew. Chem. Intl. Ed. 1972, 11, 159-173. 183  97.  Grisebach, H., Adv. Carbohydr. Chem. Biochem. 1978, 35, 81-126.  98.  Chen, H., Zhao, Z., Hallis, T. M., Guo, Z., and Liu, H.-w., Angew. Chem. Intl. Ed. 2001, 40, 607-610.  99.  Takahashi, H., Liu, Y.-n., and Liu, H.-w., J. Am. Chem. Soc. 2006, 128 (5), 1432-1433.  100.  Vongerichten, E., Liebigs Ann. Chem. 1901, 318, 12 1-136.  101.  Schmidt, 0. T., LiebigsAnn. Chem. 1930, 483, 115-123.  102.  Grisebach, H., Branched-chain sugars: occurence and biosynthesis. In The Biochemistry ofPlants: a Comprehensive Treatise, Stumpf, P. K., and Conn, E. E., Eds. Academic Press: New York, 1980; pp 17 1-197.  103.  Rose, J. K. C., O’Neil, M. A., Albersheim, P., Darwill, A. The Primary Cell Wall of ,  Higher Plants. In Carbohydrates in Chemistry and Biology, Ernst, B., Hart, G. W., and Sinay, P., Eds. Wiley-VCH: New York, 2000; Vol. 4, pp 783-808. 104.  Blevins, D. G., and Lukaszewski, K. M., Annu. Rev. Plant Physiol. Plant Mol.Biol. 1998, 49, 48 1-500.  105.  O’Neill, M. A., Warrenfeltz, D., Kates, K., Pellerin, P., Doco, T., Darvill, A.G., and Albersheim, P., 1 Biol. Chem. 1996, 271, 22923-22930.  106.  O’Neill, M. A., Eberhard, S., Albersheim, P., and Darvill, A. G., Science 2001, 294, 846849.  107.  Kelleher, W. J., Baron, D., Ortmann, R., and Grisebach, H., FEBS Letters 1972, 22, 203204.  108.  Matern, U., and Grisebach, H., Eur. I Biochem. 1977, 74, 303-3 12.  109.  Baron, D., Streitberger, U., and Grisebach, H., Biochem. Biophys. Acta 1973, 293, 526533.  184  110.  Kindel, P. K., and Watson, R. R., Biochem. J 1973, 133, 227-241.  111.  Mølhoj, M., Verma, R., and Reiter, W.-D., PlantJ 2003, 35, 693-703.  112.  Baron, D., and Grisebach, H., Eur. J Biochem. 1973, 38, 153-159.  113.  Mendicino, J., and Picken, J. M. , J Biol. Chem. 1967, 242, 1629-1634,  114.  Candy, D. J., and Baddiley, J., Biochem. J 1965, 96, 526-529.  115.  Bruton, J., and Homer, W. H., J Biol. Chem. 1966, 241, 3142-3146.  116.  Grisebach, H., and Döbereiner, U. Z., Biochem. Biophys. Res. Commun. 1964, 17, 737741.  117.  Kelleher, W. J., and Grisebach, H., Eur. J Biochem. 1971, 23, 136-142.  118.  Rose, 1. A., J Am. Chem. Soc. 1958, 80, 5835-5836.  119.  Kuzuyama, T., and Seto, H., Nat. Prod. Rep. 2003, 20, 171-183.  120.  Calvo, J. M., Leucine biosynthesis in prokaryotes. In Amino acids, biosynthesis and genetic regulation., Herrman, K. M., and Somerville R.L., Eds. Addison-Wesley 1983; pp 267-284.  121.  Vann, W. F., Dais, D. A., Murkin, A. S., Tanner, M. E., Chaffin, D. 0., Rubens, C. E., Vionnet, J., and Silver, R. P., J Bacteriol. 2004, 186, 706-7 12.  122.  Tsvetkov, Y. E., Shashkov, A. S., Knirel, Y. A., and Zahringer, U., Carbohydr. Res. 2001, 335, 22 1-243.  123.  Gosselin, S., Aihussaini, M.,Streiff, M. B., Takabayashi, K., and Palcic, M. M., Anal. Biochem. 1994, 220, 92-97.  124.  Bock, K., and Pedersen, C. J., J Chem. Soc., Perkin Trans. 1974, 2, 293 -297.  185  125.  Chou, W. K., Hinderlich, S., Reutter, W., and Tanner, M. E., I Am. Chem. Soc. 2003, 125, 2455-2461.  126.  Strandberg, L., and Enfors, S.-O., Appi. Environ. Microbiol. 1991, 57, 1669-1674.  127.  Guan, C., Li, P., Riggs, P. D. and Inouye, H., Gene 1987, 67, 2 1-30.  128.  Chang, J. Y., Eur. I Biochem. 1985, 151, 2 17-224.  129.  Machovich, R., The Thrombin 1984, 1, 63-64.  130.  Vliegenthart, J. F. G., Dorland, L., van Halbeek, H., and Haverkamp, J., NMR Spectroscopy of Sialic Acids. In Sialic Acids Chemistry, Metabolism and Function, Schauer, R., Ed. Springer-Verlag: New York, 1982; pp 128-173.  131.  Chou, W. K., Dick, S., Wakarchuk, W. W., and Tanner, M. E., I Biol. Chem. 2005, 280, 35922-35928.  132.  Gunawan, J., Simard, D., Gilbert, M., Lovering, A. L., Wakarchuk, W. W., Tanner, M. E., and Strynadka, N. C., I Biol. Chem. 2005, 280, 3555-3563.  133.  DeLeo, A. B., and Sprinson, D. B., Biochem. Biophys. Res. Commun. 1968, 32, 873-877.  134.  Hedstrom, L., and Abeles, R., Biochem. Biophys. Res. Commun. 1988, 157, 816-820.  135.  Seehoizer, S. H., Jaworowski, A. and Rose, I. A., Biochemistry 1991, 30, 727-732.  136.  Matte, A., Tan, L. W., Goldie, H. and Delbaere, L. T., I Biol. Chem. 1997, 272, 8 105-  8 108. 137.  Cohn, M., and Hu, A., Proc. Natl. Acad. Sci. 1978, 75, 200-203.  138.  Sundaram, A. K., Pitts, L., Muhammad, K., Wu, J., Betenbaugh, M., Woodard, R. W. and Vann, W. F. Biochem 1 2004, 383, 83-89. ,  139.  .  Dotson, G. D., Nanjappan, P., Reily, M. D. and Woodard, R.W., Biochemistry 1993, 32 (46), 12392-12397. 186  140.  Bradford, M. M., Anal. Biochem. 1976, 72, 248-254.  141.  Laemmli, U. K., Nature 1970, 227, 680-685.  142.  Bondinell, W. E., Vnek, J., Knowles, P. F., Sprecher, M., and Sprinson, D. B., J Biol. Chem. 1971, 246, 6191-6196.  143.  Vialletelle, V., Rabiller, C., Heisler, A., and Levayer, F., Tetrahedron-Asymmetry 1992, 3, 673-676.  144.  Ahn, J.-W., Verma, R., Kim, M., Lee, J.-Y., Kim, Y.-K., Bang, J.-W., Reiter, W.-D., and Pai, H.-S., J. Biol. Chem. 2006, 281, 13708-13716.  145.  Alberts, B., Bray, D., Hopkin, K., Johnson, A., Lewis, J., Raff, M., Roberts, K., and Walter, P., Essential Cell Biology. 2 ed.; Garland Science: New York, 2004.  146.  Qiagen Instruction Manual: RNeasy Mini Handbook. 2006.  147.  Ding, L., Seto, B. L., Ahmed, S. A., and Coleman, W. G. Jr. J Biol. Chem. 1994,269, ,  24384-24390. 148.  Greighton, T. E., Protein Function: Practical Approach. 2 ed.; Oxford University Press: Oxford, 1997.  149.  Chaykin, S., King, L., and Watson, J. G. Biochim. Biophys. Acta. 1966, 124, 13-25.  150.  Kindel, P. K., Gustine, D. L., and Watson, R. R., Plant Physiol. 1970, 46(Suppi.), 27.  151.  Paladini, A. C., and Leloir, L. F., Biochem J. 1951, 51, 426-430.  152.  Gaugler, R. W., and Gabriel, 0., J Biol. Chem. 1973, 248 (17), 604 1-6049.  153.  Snetkova, E. V., Akulov, G. P., Gordeeva, L. S., and Kaminskii, Yu. L., Khimiya  ,  .  Prirodnykh Soedinenii 1987, 1, 125-128. 154.  Roseman, S., Distler, J. 3., Moffatt, J. G., and Khorana, H. G., I Am. Chem. Soc. 1961, 84, 633-675. 187  155.  Breazeale, S. D., Ribeiro, A. A., and Raetz, C. R. H., J Biol. Chem. 2002, 277, 28862896.  156.  Helander, 1. M., Kilpelainen, I., and Vaara M., Mo!. Microbiol. 1994, 11, 481-487.  157.  Hancock, R. E., Falla, T., and Brown, M., Adv. Microb. Physiol. 1995, 37, 135-175.  158.  Williams, J. G., Breazeale, S. D., Raetz, C. R. H., and Naismith J. H., .1 Biol. Chem. 2005, 280 (24), 23000-23008.  159.  Brown, H. C., and Ichikawa K., Tetrahedron 1957, 1, 22 1-230.  160.  Pierce, J., Andrews, T. J., and Lorimer, G. H., J Biol. Chem. 1986, 261.  161.  Cleland, W. W., Biochemistry 1990, 29, 3194-3197.  162.  Baron, D., Wellmann, E., and Grisebach, H., Biochim. Biophys. Acta 1972, 258, 31 03 18.  163.  Marlow, A. L., and Kiessling, L. L., Org. Lett. 2001, 3 (16), 25 17-2519.  164.  Koós, M., and Mosher, H. S., Carbohydr. Res. 1986, 146, 335-34 1.  165.  Nachman, R. J., Hoenel, M., Williams, T. M., Halaska, R. C., and Mosher, H. S., .1 Org. Chem. 1986, 51(25), 4802-4806.  166.  MbaYraroua, 0., Thang, T.-T., and Tapiero, C., Carbohydr. Res. 1994, 253, 79-99  167.  Walsh, C., Tetrahedron: Asymmetry 1982, 38, 871-909.  168.  Pongdee, R., and Liu, H.-w., Bioorg. Chem. 2004, 32, 393-437.  169.  O’Hagan, D., and Rzepa, H. S., J Chem. Soc., Chem. Commun. 1997, 645-652.  170.  Chang, C.-w. T., Chen, X .H., and Liu, H.-w., I Am. Chern. Soc. 1998, 120, 9698-9699.  171.  Wong, A., Munos, J. W., Devasthali, V., Johnson, K.A., and Liu, H.-w., Org. Lett, 2004, 6, 3625-3628. 188  172.  Stick, R. V., and Watts, A.G., Monatshefte fur Chemie 2002, 133, 541-554.  173.  Vincent, S. P., Burkart, M. D., Tsai, C.-Y., Zhang, Z., and Wong, C.-H., .1 Org. Chem. 1999, 64, 5264-5279.  174.  Withers, S. G., Maclennan, D. J., and Street, I. P., Carbohydr. Res. 1986, 154, 127-144.  175.  MacDonald, D. L., Methods Carbohydr. Chem. 1972, 6, 389-392.  176.  Stevens, J. D., in Methods in Carbohydrate Chemistry. Academic Press: New York, 1972; Vol. VI, pp 123-128.  177.  Weigel, T. M., Liu, L.-d., and Liu, H.-w., Biochemistry 1992, 31, 2129-2139.  178.  Theander, 0., Acta Chem. Scand. 1964, 18, 2209-22 16.  179.  Card, P. J., I Org. Chem. 1983, 48, 393-3 95.  180.  Card, P. J., Carbohydr. Chem. 1985, 4, 45 1-487.  181.  Wright, J. A., and Taylor, N. F., Carbohydr. Res. 1966, 3, 3 33-339.  182.  Wright, J. A., Methods in Carbohydrate Chemistry. New York, 1972; Vol. 6, pp 201-205.  183.  Campbell, R. E., and Tanner M. E., Angew. Chem. Intl. Ed. 1997, 36, 1520-1522.  184.  Sim, M. M., Kondo, H., and Wong, C.-H., J Am. Chem. Soc. 1993, 115, 2260-2267,  189  061  xipuddV  Figure A.1 ‘H NMR spectrum of 2,4-diacetamido-2,4,6-trideoxymannose (300 MHz, D 0, 2 25 °C). 191  1  C  C  In  e  C Ca In  Figure A.2. ‘H NMR spectrum of UDP-4-ketoxylose (400 MHz, D 0, 25 °C). 2  192  =  In  C  0  C  In  C  C  0  Figure A.3 ‘H NMR spectrum of 2,3,3’-tribenzyl-D-apiose 14. (400 MHz, CDCI , 25 °C). 3  193  ±  z  4-4  Figure A.4 ‘H NMR spectrum of UDP-2-deoxy-2-fluoro-D-glucuronic acid as a triethyl ammonium salt 22 (400 MHz, D 0, 25 °C). 2  0  •0  0  = I14  0  •0  •0  194  J 0  n  H H  0 I  5.50  5.00  4.50  I  4.00  3.50  ppm  Figure A.5 Partial 1 H COSY spectrum of UDP-2F-Xyl (400 MHz, D 0, 25 °C). 2  195  9  vL =  Figure A.6 ‘H NMR spectrum of UDP-3-deoxy-3-fluoro-D-glucuronic acid as a triethyl ammonium salt 28 (400 MHz, D 0, 25 °C). 2  196  -  = z  Figure A.7 ‘ C NMR spectrum of UDP-[U-’ 3 C]-3-deoxy-3-fluoro-D-glucuronic acid as a 3 triethylammonium salt (150 MHz, D 0, 25 °C). 2 197  

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}]}"
                            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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0061207/manifest

Comment

Related Items