Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural investigation and bacteriophage degradation of bacterial polysaccharides Karunaratne, Desiree Nedra 1985

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

Item Metadata

Download

Media
831-UBC_1985_A1 K37.pdf [ 10.21MB ]
Metadata
JSON: 831-1.0060518.json
JSON-LD: 831-1.0060518-ld.json
RDF/XML (Pretty): 831-1.0060518-rdf.xml
RDF/JSON: 831-1.0060518-rdf.json
Turtle: 831-1.0060518-turtle.txt
N-Triples: 831-1.0060518-rdf-ntriples.txt
Original Record: 831-1.0060518-source.json
Full Text
831-1.0060518-fulltext.txt
Citation
831-1.0060518.ris

Full Text

STRUCTURAL INVESTIGATION AND BACTERIOPHAGE DEGRADATION OF BACTERIAL POLYSACCHARIDES BY DESIREE NEDRA KARUNARATNE B.Sc, The University of Colombo, Sri Lanka, 1979 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE"REQUIREMENTS FOR THE DEGREE OF : ' V DOCTOR OF PHILOSOPHY IN THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA MARCH 1985 © D.N. Karunaratne, 1985 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available f o r reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of CbemnsV*'^  The University of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 D a t e a s * f w i ms-ii ABSTRACT Seventy eight serologically distinct strains of Klebsiella bacteria are known to exist. The capsular polysaccharide surrounding the bacterial cell of these pathogenic Enterobacteria is of immunological significance. Structures of the capsular polysaccharides of nearly sixty seven strains of Klebsiella have been established, and each one found to be unique. The structures of the K antigens from Klebsiella, serotypes K67 and K80 are presented as a contribution to the continuing program of elucidation of the chemical structures of these antigens in an attempt to explain their immunological responses. Chemical methods of structural elucidation were employed and the following two structures were obtained. 3-Ma n-i-^Ma n J-^Gi c I - 1 R h a^. 0 G a l ^ G l c A 0 a a a 1 Rha Klebsiella K67 2 M a nL_2 M a nJ_3 G a lJ_ a GlcA 4 1 Rha A 3 \ / pyr Klebsiella K80 The polysaccharide from K67 was unique among the Klebsiella K antigens in having a four-plus-two-plus-one repeating-unit (indicating a branch on a side chain), while K80 was unique as it was the first instance that a pyruvic acetal was found on a terminal rhamnose unit. The importance of bacteriophage-borne enzymes in the generation of single iii repeating-units containing labile substituents is demonstrated. Klebsiella K44 polysaccharide was degraded using a crude solution of #44 bacteriophage. The oligosaccharides obtained were crucial in the determination of the position of the O-acetate group. In the case of the polysaccharide from Klebsiella K26, the degradation performed using a crude solution of #26 bacteriophage resulted in the isolation of a single repeating-unit containing a pyruvic acetal together with an oligosaccharide corresponding to a single repeating-unit devoid of its terminal pyruvate containing sugar. The structures of these compounds which are as follows, were useful in further confirmation of the structures of the original polysaccharides. GlcAi-^Rha^Rha^GlcJ-^Glc 0 « « 0 16 OR R = H or Ac Oligosaccharides from Klebsiella K44 Glc A J-^Ma n J-^Ma ni-^Ga 1 ^ a a a a 1 Glc 6 J 1 Glc 4 OR R = H or pyr *Gal I ' 0 Oligosaccharides from Klebsiella K26 iv Escherichia coli. another pathogenic Enterobacteria possessing immunologically significant K antigens, has been found to contain capsular polysaccharides bearing a strong resemblance to those of Klebsiella. Recently it was discovered that some strains of E. coli contained K antigens comprising amino sugars. A preliminary study on the chemical behaviour of amino sugars, and chemical methods of structure elucidation of such polysaccharides have been included in an appendix as this has been a new area of research in this laboratory. An appendix containing compilations of the cross-reactions, known structures, and chemotypes of the Klebsiella K antigens has also been included. V Table of Contents Abstract Table of Contents List of Appendices List of Tables List of Figures List of Schemes Acknowledgments I Introduction 1 II Methodology of structural analysis of polysaccharides 18 11.1 Isolation and purification 19 11.1.1 Klebsiella Polysaccharides 21 11.1.2 Escherichia coli Polysaccharides 22 11.2 Separation techniques 22 11.2.1 Paper chromatography 23 11.2.2 Paper electrophoresis 23 11.2.3 Gas-liquid chromatography 24 11.2.4 High performance liquid chromatography 26 11.2.5 Gel-permeation and ion-exchange chromatography 27 11.3 Sugar analysis 28 II.3.1 Total hydrolysis - 28 Page ii v x xi xiii xv xvii vi II.3.1.a Hydrolysis with acids 29 II.3.1.D Methanolysis 30 11.3.1. C Acetolysis 30 11.3.2 Characterization of the sugars 31 11.3.3 Determination of the absolute configuration of sugars 31 11.3.4 Modification of uronic acids for gas-liquid chromatographic detection 33 11.4 Establishment of linkage position 34 11.4.1 Methylation analysis 34 11.4.2 Characterization of methylated sugars 37 11.4.3 Applications of mass spectrometry 38 II.4.3.a Characterization of monosaccharide derivatives 40 II.4.3.b Identification of derivatized oligosaccharides 49 11.5 Sequencing of sugars 52 11.5.1 Partial hydrolysis 52 11.5.2 Periodate oxidation and Smith degradation 55 11.5.3 Base catalyzed /3-elimination from hexuronic acid residues 61 11.5.4 Deamination of aminosugars. 62 11.6 Determination of the anomeric configuration of linkages 65 11.6.1 Optical rotation 65 11.6.2 Nuclear magnetic resonance spectroscopy 66 11.6.2. a JH-n.m.r. spectroscopy 67 II.6.2.b ^C-n.m.r. spectroscopy 74 11.6.3 Other techniques 78 vii II.6.3.a Enzymatic hydrolysis 78 II.6.3.D Chromium trioxide oxidation 79 II. 7 Location of O-acetyl groups 80 III General Experimental Conditions 83 III. 1 Paper chromatography 84 111.2 Gas-liquid chromatography and g.l.c.-m.s. spectrometry 84 111.3 Gel-permeation and ion-exchange chromatography 86 111.4 Optical rotation and circular dichroism 87 111.5 Nuclear magnetic resonance spectroscopy 87 111.6 General conditions 87 111.7 Formation of alditol acetates and per-acetylated aldononitriles 88 111.8 N-deacetylation of polysaccharides 89 111.9 Deamination 89 111.10 Isolation and purification of the polysaccharides 90 III.10.1 Klebsiella polysaccharides 90 III. 10.2 E. coli polysaccharides 91 111.11 Bacteriophage isolation and propagation 92 111.11.1 Isolation of bacteriophages from sewage 92 111.11.2 Tube and flask lysis 93 viii IV Structural Investigation of Klebsiella capsular polysaccharides 96 IV.1 Structure elucidation of the capsular poly-saccharide of Klebsiella serotype K67 97 IV.1.1 Abstract 97 IV.1.2 Introduction 98 IV.1.3 Results and discussion 98 IV.1.4 Conclusion 109 IV.1.5 Experimental 109 IV.2 Structural investigation of Klebsiella serotype K80 capsular polysaccharide 115 IV.2.1 Abstract 115 IV.2.2 Introduction 116 IV.2.3 Results and discussion 116 IV.2.4 Conclusion 130 IV. 2.5 Experimental 131 V Bacteriophage degradation of the capsular polysaccharides from Klebsiella serotypes K26 and K44 137 V. l Introduction 138 V.2 Isolation and analysis of the oligosaccharides from the depolymerization of Klebsiella K44. 147 V.2.1 Abstract 147 V.2.2 Introduction 148 V.2.3 Results and discussion 148 V.2.4 Conclusion 162 V.2.5 Experimental 162 ix V.3 Isolation and characterization of the products from the bacteriophage depolymerization of Klebsiella K26 166 V.3.1 Abstract 166 V.3.2 Introduction 166 V.3.3 Results and discussion 168 V.3.4 Conclusion 177 V.3.5 Experimental 178 VI Bibliography 180 X List of Appendices Appendix I ]95 Preface J95 Preliminary investigation leading to structure elucidation of amino sugar containing capsular polysaccharides from Escherichia coli Introduction j^g Sugar analysis 197 Chemical Investigation 200 Chemical analysis of E. coli 08:K45:H9 (K45 antigen) 202 Structural analysis of E. coli 08:K44(A):H" (K44 antigen) , 209 Spectral analysis of E. coli 08:K43:H11 (K43 antigen) 218 Conclusion 221 Experimental 222 Bibliography 228 Appendix II 229 a) Qualitative analysis and chemotyping of Klebsiella capsular polysaccharides 230 b) Cross-reactions between Klebsiella K antigens 231 c) The known structures of the Klebsiella capsular polysaccharides 232 Appendix III 253 *H- and ^ C-n.m.r. spectra 254 xi List of Tables Nomenclature of the Family ENTEROBACTERIACEAE Diseases caused by encapsulated bacteria N.M.R. data for Klebsiella K67 polysaccharide and the derived oligosaccharides Methylation analysis of native, and degraded K67 polysaccharide Analysis of the oligosaccharides from partial hydrolysis of Klebsiella K67 polysaccharide N.M.R. data for Klebsiella K80 polysaccharide and the derived oligosaccharides Methylation analyses of K80 polysaccharide and derived products Sugar analysis of the Smith degradation products P1R - P6R Methylation analyses of selectively hydrolyzed K80 polysaccharide Methylation analysis of oligosaccharides 5 and 6 N.M.R. data for Klebsiella K44 polysaccharide and the derived oligosaccharides Sugar analysis of the methyl vinyl ether protected, methylated polysaccharide, and oligosaccharide 5 Ethylation analysis of native K44 polysaccharide and the 6-O-methyl polysaccharide and oligosaccharide Determination of the degree of polymerization and the reducing end of oligosaccharides Pla, PI and P2 from Klebsiella K26 Methylation analyses of oligosaccharides Pla, PI and P2 400 MHz 'H-n.m.r. data for Pla, PI and P2 oligo-saccahrides from the bacteriophage degradation of Klebsiella K26 xii V.8 100 MHz 1:>C-n.m.r. data for Pla, PI and P2 isolated from Klebsiella K26. 1 7 5 A.l Sugar analyses of E. coli K43, K44 and K45 polysaccharides 199 A.2 Methylation analysis of E. coli K45 polysaccharide 205 A.3 Methylation analysis of E. coli K44 polysaccharide 213 xiii List of Figures Figure 1.1 Schematic representation of the Gram-positive and Gram-negative cell wall of bacteria 3 1.2 Schematic structure of Salmonella lipopolysaccharides. The number of nonhydroxylated and hydroxylated fatty acids depicted is arbitrary 7 1.3 Chemical nature of the serological specificities expressed by O factors 122 a n c* 112 * n Salmonella typhi and S. tvphimurium and by O factor 34 in S. illinois 10 1.4 Illustration of the different antigenic determinants in the LPS of S. tvphimurium 10 1.5 Various regions on a tetrasaccharide (determinants) possibly reacting with different antibodies 11 1.6 The common immunodominant sugars in Pneumococcus type III polysaccharide and Klebsiella K5 polysaccharide involved in cross-reactions with anti-Pnlll serum 13 11.1 Comparison of the chromatograms produced by the alditol acetates of a hydrolyzate from E. coli K44 on (a) a packed column and (b) a capillary column 25 11.2 (a) Formation of hemiacetals resulting in under-oxidation of oxidizable sugar units and (b) their susceptibility to oxidation on reduction with sodium borohydride 56 11.3 Deamination sequences of some amino sugars 63 11.4 Relationship between dihedral angle (<t>) and coupling constants for a - and /3 - D-hexoses 69 11.5 Schematic representation of different regions in the H^-n.m.r. spectrum of polysaccharides 71 11.6 H^-n.m.r. spectra of (a) Native K80 polysaccharide and (b) after depyruvalation 73 11.7 The characteristic regions for resonances of carbon atoms belonging to different monosaccharide residues in polysaccharides 76 xiv IV.1 Mass spectrum of the methylated oligosaccharide A5 from Klebsiella K67 polysaccharide 106 IV.2 Chromatogram of the products isolated from the Smith degradation of the K80 polysaccharide 124 IV.3 Mass spectrum of the methylated oligosaccharide alditol P3R from Klebsiella K80 polysaccharide 126 V.l Relative sizes of a bacterium (E. coli) and an assortment of biological entities (Bdellovibrio, bacteriophages and a Colicin) which attack it. Colicin K is beyond the resolution of current electron microscopes 139 V.2 Basic morphological types of bacteriophages with the types of nucleic acid 141 V.3 (a) Schematic diagram demonstrating the structure of a T-even bacteriophage, (b) Electron micrograph of a T4 bacteriophage 141 V.4 A schematic diagram illustrating the steps in the infection of a bacterium by a T-even phage 143 V.5 Drawing depicting the adsorption of E. coli <£29.onto the encapsulated E. coli K29. The phage tail spikes recognize and bind the exopolysaccharide. The endoglycosidase of the tail spikes hydrolyze Glc-Ur-GlcA linkages in the glycan strand of the exopolysaccharide opening a path for the phage particle. Triggering and ejection do not occur until the phage has reached the cell wall; penetration of the phage nucleic acid apparently takes place at points of fusion between wall and cytoplasmic membrane 14 5 V.6 Acetate location: gas-liquid chromatogram of methylated alditol acetates derived from (a) K44 polysaccharide (b) pentasaccharide obtained by action of bacteriophage 157 V.7 The C^-n.m.r. spectra of (a) native K44 polysaccharide, (b) deacetylated polysaccharide, (c) oligosaccharide 5, (d) oligosaccharide 4 161 A.l Elution profile of the products formed on deamination of the N-deacetylated E. coli K44 polysaccharide 217 XV List of Schemes Schemes 11.1 Reduction of carboxylic acid in aqueous solution using carbodiimide reagent 32 11.2 Methylation analysis of Klebsiella K67 polysaccharide 36 11.3 Fragmentation pathways of a permethylated methyl glycoside 39 11.4 Mass spectra of (a) hexitol hexaacetate, (b) 1,5-di-O-acetyl-2,3,4,6-tetra-0-methylhexitol 41 11.5 Fragmentation patterns of alditol derivatives 42 11.6 Fragmentation patterns of (a) 2-acetamido-2-deoxy-hexitol pentaacetate and (b) 3-acetamido-3-deoxyhexitol pentaacetate 44 11.7 Mass spectra and fragmentation patterns of (a) 1,5-di-O-acetyl-3,6-dideoxy-3-N-methylacetamido-2,4-di-0-methyl-hexitol and (b) l,5-di-0-acetyl-2-deoxy-2-N-methyl-acetamido-3,4,6,tri-0-methylglucitol 45 11.8 Mass spectra of (a) l,4,5-tri-0-acetyl-2,3,6-tri-0-ethylglucitol and (b) l,4,5-tri-0-acetyl-2,3-di-0-ethyl-6-O-methylglucitol 47 11.9 Fragmentation by A and B series of a permethylated disaccharide derivative 48 11.10 Fragmentation pattern of a permethylated oligosaccharide alditol 50 11.11 Common products formed by terminal and mono substituted sugars on periodate oxidation followed by borohydride reduction and hydrolysis 54 11.12 Inter-residue hydrogen bonding between acetamido groups and hydroxyl groups 58 11.13 Periodate oxidation and Smith hydrolysis of the Klebsiella K80 polysaccharide 60 11.14 Location of O-acetyl substituents using the methyl vinyl ether protection method 81 IV.l Fragmentation pattern of the oligosaccharide A5 from Klebsiella K67 polysaccharide 107 xvi The glycolaldehydes formed from the product obtained by Smith degradation of the K80 polysaccharide. Fragmentation pattern of the oligosaccharide alditol P3R from Klebsiella K80 polysaccharide xvii Acknowledgments It has been a pleasure to work under Professor G.G.S. Dutton, and the outcome extremely rewarding. I wish to express my sincere gratitude to him for his encouragement, advice and guidance both in and out of the laboratory and extend a special thank you for his continued interest. The cooperation and assistance given by my colleagues during the course of this work is gratefully acknowledged. I wish to thank, in particular, Dr. E.H. Merrifield for his advice and guidance and Drs. Jose L. Di Fabio and Eleanora Altman for their helpful discussions. Thanks are also due to Dr. S.C. Churms (University of Cape Town, South Africa) for gel-permeation measurements, Dr. S.O. Chan and the staff of the N.M.R. Service and Dr. G. Eigendorf and the staff of the Mass Spectrometry Service for their invaluable assistance. I am grateful to Dr. D.A.I. Goring for the excellent editing and proof-reading so willingly performed. My thanks also go to Rani Theeparajah for the patience expressed and the care taken in typing this thesis. Finally, and most importantly, the encouragement, assistance and understanding expressed by my husband, Veranja, throughout the course of this work is greatly appreciated and acknowledged with thanks. Dedicated to my parents -1-I. INTRODUCTION - 2 -I. INTRODUCTION The abundance of carbohydrate-containing macromolecules in nature is manifest in almost all the life forms. Thus, plants are known to consist of wood polysaccharides, gums, and mucilages, while algae, fungi, and yeast, contain their own polysaccharide matrices. Most bacteria produce exclusively carbohydrate polymers. Glycoconjugates consisting of proteins linked to carbohydrates e.g. glycoproteins, proteoglycans, and peptidoglycans are widespread in animals and man, as are nucleic acids which are also carbohydrate containing biopolymers. Polysaccharide chains covalently attached to lipids (lipopolysaccharides, LPS) are a part of cell walls of many bacteria. The teichoic acid group of substances where phosphorodiester linkages coexist with glycosidic linkages also fall into this category on account of their carbohydrate content. Thus the term "polysaccharide" embraces all the above mentioned polymers and is not confined solely to O-glycosidically linked carbohydrates.* Commercially, polysaccharides are of importance in the food, cosmetic, textile, paper, and paint industries as well as in oil well drilling fluids, and in photographic and rubber processing chemicals. The thickening and gelling properties of the polysaccharides are the main features contributing to their usefulness, while their emulsifying, binding, coating and film forming abilities play a secondary role. More than 50% of the commercially available polysaccharides are used in the food industry to control the texture of foods as well as their flavour, appearance and colour.^ The functions of microbial extracellular polysaccharides as storage and energy reserves and in virulence (protection against phagocytosis) have been known for a long time. However their role in these areas was not clearly - 3 -C a p s u l a r P o l y s a c c h a r i d e C a p s u l a r P r o t e i n «— P e p t i d o g l y c a n W i t h Teichoic A c i d ^ . P o l y m e r s P h o s p h o l i p i d b i layer • w i t h v a r i o u s m e m b r a n e p ro te ins , e n z y m e s a n a p e r m e a s e s I n t e r i o r Of C e l l - ' ' ' " " ' C e l l envelope of the Gram-positive c e l l wall C a p s u l e C e l l W a l l C y t o p l a s m i c M e m D ' a n e • C a p s u l a r P o l y s a c c h a r i d e , - I I s o m a t i c , L i p o p o l y s a c c h a n d e l 0 a n t i g e n ' C a p s u l e P h o s p h o l i p i d B i l a y e r w i t h s t r u c t u r a l a n a e n z y m i c proteins Celt W a l l I L ipoprote in P e r i p l a s m s E n z y m e s P e p t i d o g l y c a n P h o s p h o l i p i d B i l a y e r \ | n n e r m e m b r a n e w i t h VOTKXJS m e m b r a n e > prote ins , e n z y m e s a n d j ( c y t o p l a s m i c p e r m e a s e s • • " m e m b r a n e ) I n t e r i o r Of C e l l ' -'.' :i\ '. C e l l envelope of the Gram-negative c e l l wall Fig 1.1 Schematic representation of the Gram-positive and Gram-negative cell wall of bacteria - 4 -understood. It is now well established that polysaccharides are important in biological recognition functions where they a) act as receptors for phage and bacteriocins, b) act as specific receptors in eukaryotes for viruses, bacteria, hormones, and toxins, c) are of antigenic specificity (capable of combining with specific antibodies), and d) are immunogenic (inducing the formation of antibodies).-^ In plants they function as regulators of growth, development, reproduction, and disease. The outermost cell component of the bacteria, which is the antigen being recognized, is the crucial participant in an immune response. The location of the microbial polysaccharide on the outer cell wall is therefore the reason for its antigenicity.'* These polysaccharides are either an integral part of the cell wall as in somatic lipopolysaccharides of Gram-negative Enterobacteriaceae (e.g. Salmonella) or form extracellular capsules as those of Pneumococcus. Klebsiella, and many Escherichia coli (see Fig. 1.1). Classification of the Enterobacteriaceae Enterobacteriaceae is a large family of Gram-negative bacteria. The family is subdivided into tribes, genera and species. A group of related species forms a genus, and a group of related genera constitutes a tribe. The recent classification of the family Enterobacteriaceae by Edwards and Ewing^ is shown in Table 1.1. This is an update of the famous classification by Kaufmann.^ Species are further divided into serotypes and bioserotypes, some of which have names while others are denoted by antigenic formulae. For example, S. enteritidis consists of many serotypes such as S. enteritidis ser. Tvphimurium (old name - S. tvphimurium) and bioserotypes like S. enteritidis bioser. Paratyphi A - 5 -Table 1.1: Nomenclature of the Family ENTEROBACTERIACEAE Tribes Genera Species I Escherichiaeae i Escherichia ii Shigella E. coli S. dysenteriae, S. sonnei, S. flexneri, S. boydii II Edwardsielleae i Edwardsiella E. tarda III Salmonelleae i Salmonella ii Arizona iii Citrobacter S. cholerae-suis S. typhi, S. enteritidis A. hinshawii C. freundii IV Klebsielleae i Klebsiella ii Enterobacter iii Pectobacterium iv Serratia K. pneumoniae, K. ozaenae, K. rhinoschleromatis E. cloacae, E. hafniae, E. aerogenes, E. liquefaciens P. carotovorum S. marcescens V Proteeae i Proteus ii Providencia P. vulgaris, P. mirabilis, P. morganii, P. rettgeri, P. alcalifaciens, P. stuartii - 6 -(old name - S. paratyphi A). The serotypes of Klebsiella and E. coli are denoted by their antigenic formulae e.g: Klebsiella K44, Klebsiella K67, and ]L coli 08:K44(A):H". Polysaccharide antigens of bacteria The antigens from the Salmonella species have been studied extensively, with the hope of understanding their pathogenicity in man. E. coli and Klebsiella which are also pathogenic are being investigated, with chemical structure elucida-tions of the nearly 80 different serotypes of Klebsiella being almost complete. Unlike the simple sugars and structures found in Klebsiella, the K antigens of E\ coli contain diverse sugars and are of varying degrees of complexity. The antigens of the Enterobacteriaceae are responsible for their classification into serotypes. The polysaccharide antigens are of two classes (the K and O antigens) with the third antigenic determinant (the H antigen) being composed of protein. The K antigen is in the form of a discrete capsule surrounding the cell. It may exist, also as a loose slime unattached to the cell surface (envelope). This somatic K antigen is capable of inhibiting agglutinations with O antisera if present in sufficient quantity.^ Thus the capsular K antigen completely characterizes the serotypes of Klebsiella. The K antigens of E. coli comprise the L, A and B antigens. The strains containing the thermostable A antigen are capsulated, whereas those with the heat labile L or B antigens lack a morphological capsule since they are envelope antigens. Other types of K antigens include the Vi antigen of S. typhi and certain serotypes of Citrobacter. the B antigens of Shigella, and the M antigen of some strains of Salmonella and Arizona and many others.^ - 7 -The heat stable, somatic, smooth O antigen is covalently linked to the phospholipid components in the outer membrane and is hence termed a lipopolysaccharide. The LPS plays an important role in bacteriophage typing and serological classifications of acapsular Gram-negative bacteria. Additionally, the LPS exerts endotoxic activities for which the lipid moiety is responsible.^ The LPS from the Salmonella species which has been studied extensively may be used as a model in comparison with those of other Enterobacteriaceae7 (see Fig. 1.2). ooopoqcM>o^$K»)<^ f O-Specific Chain Core • ><- Lipid A -j Polysaccharide Lipid o Monosaccharide, • Phosphate, -v> Ethanolamine Long Chain (Hydroxy) Fatty Acid. Fig. 1.2 Schematic structure of Salmonella lipopolysaccharides. The number of nonhydroxylated and hydroxylated fatty acids depicted is arbitrary There are three well defined regions consisting of (i) an O-specific polysaccharide, (ii) a core-oligosaccharide and (iii) lipid A. The lipid A is buried in the outer membrane of the bacterial cell by hydrophobic interactions with the cell wall lipoproteins. The core region is linked to the lipid A via 2-keto-3-deoxymannooctulosonic acid (KDO) in all Gram-negative bacteria. Whereas the lipid A is of constant composition, five different structures of core-oligosaccharides are known for the Enterobacteriaceae.^ The O antigens are subject to smooth (S) to rough (R) form variations. The R antigenic property - 8 -belonging to the core is present in all forms, but is masked in the smooth forms by the O-specific polysaccharide responsible for the O antigenicity. The rough R antigen varies serologically, and in many other ways, from the smooth O antigen. This is due mainly to the lack of the O-specific polysaccharide and its high lipid content. The absence of the O antigen results in a loss of pathogenicity and hence the R antigens are not immunologically important.^ H antigens are present in bacteria containing flagella. Klebsiella and Shigella, being non-motile, lack the flagella H antigens. E. coli is composed of both motile and non-motile bacteria, with the H antigenic property belonging solely to the motile organisms. The amino acid sequence of the flagella determines their antigenic specificities in a manner similar to the carbohydrate sequence determining the O and K antigenic properties of the polysaccharide antigens.^ Serological cross-reactions of antigens It was in the early 1920's that Heidelberger, the pioneer in the immunological field, made use of the observation by Dochez and Avery,^ that a certain "soluble substance" capable of precipitating specific anti-pneumococcal sera of a homologous type was present whenever Pneumococci were grown in fluid media.U This soluble substance was shown to be composed of carbohydrates and to be type specific^!3 (having the ability to precipitate antibodies of its own type).^ In the 80 serological types of Pneumococcus recognized, the structures of the type-specific (capsular) substances differ from one another.^ The immunological importance of pneumococcal bacterial polysaccharides was proved by their protective effect towards pneumonia infections. 16,17 ^he antibodies to - 9 -these capsular antigens are capable of protecting experimental animals against pneumococcal disease. The antibodies recognize certain immunodeterminant groups on the antigen. Therefore the shape of the antibody combining site must conform quite closely to the shape of the immunodeterminant group.^ The antigenic specificity or ability to react with antibodies depends on the binding of the immunodeterminant group with the specific antibody. Serological cross-reactions occur when different antigens give precipitin reactions with the same antiserum on account of the similar nature of their immunodominant group. Thus immunochemical analyses of polysaccharide antigens which combine serological and chemical studies are important in defining the chemistry of the oligosaccharide structure of immunological significance within the polysaccharide.^ Until cross-reactions of various polysaccharide antigens (mainly Salmonella LPS and Pneumococci). with several antisera were carried out in the 1960's by Heidelberger, their importance in immunology was not clearly understood.^ The antigenic specificity of the polysaccharide or the immunodeterminant group, is made up of several regions centering around a particular sugar unit. The immunodominant sugar is the major contributor to the serological specificity.^ These sugars may be terminal non-reducing or within the polysaccharide chain. Most branched polysaccharides possess immunodominant sugars in their side chains. The antigenic specificity is not confined to the immunodominant sugar but extends along the polysaccharide chain. Hence, the nature of the sugar, its anomeric configuration, position of linkage, and even its position of linkage to the adjacent sugar are contributory factors. This is best illustrated in Fig. 1.3 where the three sugars glucose, galactose, and mannose, despite having the same sequence, exhibit completely unrelated serological specificities owing to the differences in their linkages and anomeric configurations. - 10 -Serological specificity Glc (O factor) 1 a 4 Icall^Mani- 122 a 0 Glc 1 a S G a l l ^ M a n i -a a 12 Glc 1 a 3GaiJ_6Man^-/3 « 34 Fig. 1.3 Chemical nature of the serological specificities expressed by O factors 122 a n <* *12 ' n Salmonella typhi and S.tvphimurium and by O factor 34 in S. Illinois Studies on the O-antigens of Salmonella serotypes have shown that a polysaccharide may contain more than one antigenic determinant. Fig. 1.4 shows [-^ManJ-^RhaJ-^Gal^ManJ-i-Rhai-kJali-],, a 0 a a p a 11 Fig. 1.4 Illustration of the different antigenic determinants in the LPS of S. tvphimurium - 11 -the different antigenic determinants of Salmonella tvphimurium. the specificities being denoted by their antigenic numbers. Different and separable antibodies may combine with the same immuno-dominant sugar by reacting with it from different sites of the molecule. Hence sugar A in Fig. 1.5 can combine with antibody I as well as antibody II, even though the antibodies have different specificities.^ Fig. 1.5 Various regions on a tetrasaccharide (determinants) possibly reacting Thus cross-reactions are not always indicators of the presence of identical sugars or structural features in a polysaccharide as is illustrated by the above two examples. However, cross-reactions can be important indicators when used with antisera of known antigenic structural specificities. In the acidic capsular polysaccharides from Pneumococcus. Klebsiella, and R coli the charged component is usally immunodominant. Non-carbohydrate substituents (e.g. O-acetyl, pyruvic acetal, phosphate) especially those conferring a negative charge upon the polysaccharide are capable of functioning as the antigenic determinant. Thus, some O-acetyl groups of Salmonella LPS (e.g. 2-O-acetyl-abequose of S. tvphimurium - Fig. 1.4) are of antigenic specificity.^ Similarly, pyruvic acetals found on Pneumococcus. Rhizobium. Klebsiella, and E. coli polysaccharides are responsible for their antigenicity.^' Cross-reactions of with different antibodies. - 12 -certain pyruvate containing polysaccharides from Klebsiella. Pneumococci. and Rhizobia. with anti-pneumococcal sera have indicated that the presence of pyruvate is essential for cross-reaction in some cases, while several of the Rhizobium polysaccharides showed increased precipitin reactions on depyruvalation.2* Pyruvic acetals contain three structural features that affect immunological specificity: a) The configuration of the monosaccharide bearing the acetal; b) The positions of the two hydroxyl groups bridged by the acetal; and c) The configuration of the methyl and carboxyl groups at the chiral acetal carbon atom. The first two features are established by cross-reactions of various polysaccharides, and by the loss of antibody precipitating capacity on removal of the pyruvic acetal.^ The importance of the latter feature was demonstrated by , model studies where the diastereomers 1 and 2 were synthesized and used to precipitate anti-PnXXVII serum. The polysaccharide from Pneumococcus type XXVII contains a pyruvic acetal on position 4,6 of glucose with the same acetal configuration as 1. Only compound 1 was capable of showing a precipitin reaction with anti-PnXXVII serum; compound 2 did not give a precipitate. - 13 -Studies with Klebsiella polysaccharides have revealed that the O-acetyl group is antigenic only if it is positioned on the immunodominant sugar. Hence Klebsiella K5 which contains an O-acetate on glucose cross-reacts with anti-PnIII serum even on deacetylation because its antigenicity depends mainly on ' the glucuronic acid unit^ (see Fig. 1.6). Pneumococcus type III [^GiCA^G\c^n Klebsiella K5 [-^GlcAl^Glcl^Mani,] n OAc pyr Fig. 1.6 The common immunodominant sugars in Pneumococcus type III polysaccharide and Klebsiella K5 polysaccharide involved in cross-reactions with anti-PnIII serum A review on cross-reactions of 60 of the Klebsiella K antigens with a series of anti-pneumococcal sera has been published.^ The importance of cross-reactions in establishing partial chemical structures of unknown immunogenic polysaccharides by using antisera to antigens of known structure and vice versa^ is illustrated therein. The K antigens of the nearly 80 different Klebsiella serotypes have been analyzed chemically and classed into chemotypes^ based on the nature of the constituent sugars. No such investigation has been carried out with the approximately 100 'K\ 164 'O' or 56 'H' antigens recognized so far in the E. coli bacteria. It is therefore necessary to screen for all classes of sugars which may be present in these antigens. The structures of the known E. coli K and O antigens - 14 -have been compiled by Altman^ and therefore are not duplicated in this thesis. The classification of the Klebsiella K antigens into chemotypes, and the 67 known structures of the capsular antigens are given in Appendix II. The cross-reactions between the Klebsiella K antigens are also included therein. Immunological significance of antigens The structure elucidation of the polysaccharides has been embarked upon in order to clarify two important issues: what makes a polysaccharide immunogenic? and what is the chemical basis of its antigenicity (serological specificity)? Cross-reactions are useful in understanding the antigenic specificity. Knowing the chemistry of the structure alone is obviously not enough to solve such a highly complex biological process, but structural investigation must be carried out in the attempt to understand this phenomena. Immune responses produced by an immunogen fall into two classes, humoral and cell mediated. The former category is where the response can be transferred from one animal to another via the serum containing the antibody, while the latter class needs sensitized cells and not serum for transfer of the immume response. T lymphocytes (thymus derived cells) are found to be responsible for the cell mediated responses while B lymphocytes (bone marrow derived), responsible for antibody formation, are involved in humoral responses. The antibodies secreted by B cells are protein molecules (immunoglobins, Ig). When T cells are involved in activating B cells into production of antibodies (IgM and IgG) the immune response is termed T cell dependent and results in a retention of immunological memory. The antigens involved in T cell dependent responses are capable of producing an even larger immune response on subsequent interaction - 15 -due to the memory retained by the T cells following the first encounter with the antigens. T cell independent responses, where B cells secrete immunoglobulins (IgM) without T lymphocyte cooperation, lack the retention of memory. Polysaccharide antigens usually stimulate T cell independent responses while protein antigens are T cell dependent response stimulators.^ It should be noted however that this is a simplified description of the complex functions occurring in an immune response; T cells have many other functions in addition to those mentioned above. The virulence of bacteria depends on their ability to survive and propagate within the host by evading the host's immune system. The direct interaction of their surface antigens with the host immune system is important in bacterial pathogenesis and in the stimulation of an immune response. The capsular nature of the antigens is useful in protecting the bacteria from phagocytosis. The function of capsular polysaccharides as human vaccines is in generating immunity towards bacterial infections. The humoral response to a pure capsular polysaccharide differs from that to the same polysaccharide when it is an integral part of the bacterium.29 Thus, the immunity received on recovery from bacterial infections, differs from that produced by vaccination with pure capsular polysaccharides. Heidelberger and coworkers were successful in producing immunity towards a serotype of Streptococcus pneumoniae by administering a single injection of a hexavalent polysaccharide vaccine.^ The immunogenicity of a pure polysaccharide can be enhanced by conjugation to a protein carrier thereby converting it into a T cell dependent antigen. This is important mainly in young children who cannot develop high levels of antibody (IgM) to polysaccharide vaccines, but respond to the protein vaccines by production of high levels of IgM and IgG with additional retention of immunological memory.^ Vaccination with Pneumococcal and Meningococcal polysaccharides is widely used now and is - 16 -effective in preventing the diseases caused by these bacteria. Consideration of the various diseases caused by encapsulated bacteria^** (see Table 1.2), together with the emergence of their widespread multiple antibiotic resistance, has given rise to a renewed interest in the prevention of these diseases by immunization. Table 1.2 Diseases caused by encapsulated bacteria Bacterial Species Diseases Streptococcus pneumoniae Pneumococcal pneumonia, meningitis, otitis media Neisseria meningitidis Meningitis Hemophilus influenzae type b Meningitis, epiglottitis, septic arthritis, and pneumonitis with empyema. Salmonella typhosa Typhoid fever Group B Streptococcus Meningitis in newborns Escherichia coli Meningitis, neonatal septicemia, urinary tract infections Staphylococcus aureus Abscesses, septicemia Klebsiella pneumoniae Pneumonia, meningitis - 17 -Uses of bacteriophages in polysaccharide chemistry Depolymerization of polysaccharides using bacteriophages is important in generating oligosaccharide repeating units without removal of immunologically significant O-acetyl and pyruvic acetal groups. These oligosaccharides may be used to study their binding behaviour with immunoglobulins,^ or may be coupled to protein carriers for use as immunogens.-^ The biological implications of the bacterial antigens discussed so far give a brief outline of some of the important roles they perform. It is with the aim of understanding these functions that the chemical analysis of the antigens has been embarked upon. This thesis covers the published work on the structural investigation of the capsular polysaccharides from Klebsiella K67 and K80. The bacteriophage depolymerizations carried out on the polysaccharides from Klebsiella K44 and K26 which have been accepted for publication and submitted for publication respectively are also included. Chemical analysis of the capsular K antigens from E. coli K43, K44 and K45 which were performed in an attempt in understanding the chemical behaviour of amino sugars is described in Appendix I. -18-II. METHODOLOGY OF STRUCTURAL ANALYSIS OF POLYSACCHARIDES - 19 -II. METHODOLOGY OF STRUCTURAL ANALYSIS OF POLYSACCHARIDES In the course of studies conducted on bacterial polysaccharides,-^ the existence of complex and immensely diverse patterns of polysaccharide chains has been revealed. For the elucidation of any one structure of such complexity necessitates the use of a combination of different chemical analyses. These include the qualitative and quantitative estimation of the component sugars and analysis for the presence of non-sugar substituents (O or N-acetyl, phosphate, pyruvic acetal etc.). Thereafter, the determination of the position of linkage, the glycosidic configuration and finally the sequencing of the sugars in the polymer have to be embarked upon. Some of the known chemical methods employed for these purposes are described in the following section. The examples used to illustrate some of the techniques are taken from the relevant applications on Klebsiella or E. coli polysaccharides. II.1 Isolation and Purification The presence of cell wall constituents, lipids, proteins and nucleic acids from the bacteria complicates the isolation of a polysaccharide in a pure form. Thus purification has to be routinely carried out in order to obtain the chemically pure, homogeneous polysaccharide. In the case of capsular polysaccharides from bacteria and exudate gums, the lyophilic nature of the polysaccharides is a factor in their ready isolation in homogeneous form. On the other hand, the cell wall components in plants and microorganisms are less soluble in aqueous media due to - 20 -their close association with the matrix or other macromolecules. Graded extraction is required to obtain selective solubility with such components. The use of water alone at various temperatures is sufficient in many extractions. Some polysaccharides may be solubilized by polar non-aqueous solvents (e.g. starch and glycogen in dimethyl sulfoxide,^ cellulose in N-methylmorpholine N-oxide^). Extractions under acidic conditions must be avoided for preservation of glycosidic linkages. The use of alkaline solutions for extractions can cause structural modifications or base catalyzed degradations (removal of O-acetyl groups, cleavage of the O-glycosidic bonds to proteins in proteoglycans) and is best avoided where possible. The addition of liquid phenol to extracts of bacterial polysaccharide permits the precipitation of lipopolysaccharides and nucleic acids from the aqueous phased7 The purification procedures used in fractionating the crude extractions into pure form are based on the nature of the polysaccharide being investigated. Isolation from aqueous solution by the addition of a water-miscible solvent-^ (e.g. ethanol, acetone) results in the precipitation of neutral and acidic polysaccha-rides. Fractional precipitation by selective salt or complex formation can be used in separating acidic polysaccharides from neutral polysaccharides. Acidic polysaccharides show salt formation with potassium chloride,-^ cupric acetate^ and cetyltrimethylammonium bromide (Cetavlon).^ Chondroitin sulfates form complexes with calcium or barium salts.^ Complex formation of galactomannans with borate,^ mannans with Fehlings solution^ and gluco- and galacto-mannans with barium hydroxide^ is also used in fractionations. Purification by chromatographic techniques is used when dealing with milligram quantities of polysaccharides. Gel filtration^ or molecular-sieve chromatography^7 on cross-linked dextrans (e.g. Sephadex) is the most commonly used chromatographic technique and is applied to separations of both poly- and oligo-saccharides. - 21 -Acidic polysaccharides containing carboxylic acid, as well as those with sulfate hemi ester or phosphoro- diester groups are fractionated by ion-exchange chromatography^^ (e.g. Streptococcus pneumoniae on DEAE Sephadex A-50 (acetate form),^^ Hemophilus influenzae type C phosphorodiester-linked capsular antigen on DEAE Sepharose with a sodium chloride gradient-^). Affinity chromatography^' using concanavalin A covalently bound to an inert support such as Sepharose has been used extensively for the fractionation of N-glycosidic glycopeptides containing a-D-mannopyranose residues.^ To establish the purity of the polysaccharide, the absence of heterogeneity, rather than the presence of homogeneity is demonstrated by as many independent criteria as possible. Sugar ratios, analysis for particular functional groups (e.g. hexuronic acid, amino sugars) and spectroscopic examination by nuclear magnetic resonance indicate constancy of chemical composition. Physical properties can be verified by optical rotation, electrophoresis^'^ and gel-permeation chromato-graphy.^ Hence it is important to specify the criteria used when dealing with new polysaccharides. II.1.1 Klebsiella Polysaccharides^,57 Cultures of Klebsiella K26, K44, K67 and K80 were obtained by courtesy of Dr. Ida 0rskov. A single colony of a strain was inoculated in beef-extract medium, incubated at 37° for 5 h. The liquid culture was then spread on a tray of sucrose-yeast extract-agar and incubated for three days. The lawn of bacteria produced was harvested and a solution of 1% phenol was used to destroy the bacteria. The dead cells were spun down by ultracentrifugation, and the polysaccharide was precipitated by the addition of Cetavlon (cetyl trimethyl - 22 -ammonium bromide). Centrifugation, dissolution of the precipitate in 4M sodium chloride, and reprecipitation with ethanol was carried out twice. The final precipitate was dissolved in water, dialyzed for two days and freeze dried to yield the purified capsular polysaccharide. II.1.2 Escherichia coli Polysaccharides Cultures of E. coli K43, K44 and K45 from Dr. Ida (Zirskov, were inoculated in Mueller-Hinton broth, and grown on sodium chloride enriched Mueller-Hinton agar at 37° for three days. The harvesting and purification were done as for Klebsiella polysaccharides (Section II.1.1). II.2 Separation Techniques Purity is an important factor in the analysis of any compound. The first step is to obtain the polysaccharide in the pure form, separated from contaminants. Column chromatography (gel-permeation, ion-exchange, affinity) was shown to be a useful tool for the purification of crude polysaccharides (Section II. 1). To obtain information on the constituent sugars of the polysaccharide and its degradation products, separation of the mixtures into individual components is essential. Paper chromatography, gas-liquid chromatography, paper electrophoresis, high pressure liquid chromatography, gel-permeation and ion-exchange chromatography are all useful techniques in separating the mixtures of sugars or their derivatives. Some applications of these techniques are described briefly in the following section. - 23 -II.2.1 Paper Chromatography (P.C.) 3 8 , 3* Basically, paper chromatography, is the separation and identification of mono- and oligo-saccharides by exploiting the relative mobility of the individual components in a particular solvent. On an analytical level, small quantities are used for identification of the free sugars generated from hydrolysis of the poly-saccharide, or modified polysaccharide (e.g. products from periodate oxidation, fractions isolated from partial hydrolysis or chromium trioxide oxidation etc.). Preparative p.c, used in separating mixtures of oligosaccharides generated from partial hydrolysis or deamination, is complementary to gel-permeation and shows good recovery of material. Sugars are detected*^ with AgN03/NaOH/Na2S203 or with rj-anisidine hydrochloride spray followed by heating at 110° for 5 min. Amino sugars are detected with ninhydrin upon heating at 110° for 5-10 min. II.2.2 Paper Electrophoresis^!'^ Electrophoresis involves the migration of charged species in a conducting solution under the influence of an applied electric field. Solutions are buffered at optimal pH values to ensure maximum separation. This technique is useful in separating protein contaminants from bacterial polysaccharides and for separation of acidic oligosaccharides resulting from partial hydrolysis of the polysaccharide. The sugars are detected^ as in paper chromatography. 24 II.2.3 Gas-liquid Chromatography (G.L.C.) 6 4" 7 3 Structural studies are based initially on the identification of the component sugars. Gas-liquid chromatography is instrumental in the separation of a wide range of sugars and its use in carbohydrate chemistry is extensive. This technique is based on the distribution of volatile components between a mobile gas phase and a stationary absorbent phase. The nature of the stationary phase affords selectivity and a variation in polarity can lead to better separations of the sugar derivatives.^ Since carbohydrates are non-volatile, early studies were carried out on the volatile methylated methyl glycosides.66 Conversion of sugars into their volatile trimethylsilyl (TMS) derivatives^? afforded new applications of this technique for identification of sugars. Although the TMS derivatives are easily formed, the existence of anomeric forms of sugars at equilibrium yields a complicated chromatograph of multiple peaks. In an attempt to overcome this problem, the acyclic sugar alditols were converted into volatile acetates, trifluoroacetates or trimethylsilyl ethers. Since the trifluproacetates show partial de-esterification on the column, and the TMS derivatives of the alditols showed poor separation,**** the readily formed alditol acetates^ have proved to be the derivatives of choice having good resolution and short retention times. In recent i times g.l.c. has been coupled to mass spectrometry.?^ For this application, alditol acetates are particularly useful since they yield simpler mass spectra than the other derivatives used for g.l.c. Conventional pack columns made of non-polar stationary phase (e.g. silicone gums SE-52, SE-30, XE-60 and carbowax) give good separations of TMS methyl glycosides, but not of alditol acetates.?* Base line separations of mixtures of many unalkylated and alkylated alditol acetates?^ are possible on columns made of ECNSS-M (ethylene succinate-cyanoethyl silicone copolymer), SP-2340 (75% - 25 -(a) Column: OV-225 Programmed at 220° for 8 min,then 8°/min to 250 Fig II.1 Comparison of the chromatograms produced by the alditol acetates of a hydrolyzate from E. coli K44 on (a) a packed column and (b) a capillary column - 26 -cyanopropyl silicone), and OV-225 (25% phenyl, 25% cyano propyl, methyl silicone), all of which are polar materials. Fused silica capillary columns are thermally stable at high temperatures (upto 350°), give good resolution, and, most importantly, can detect samples in the microgram range. Chemically bonded phases for fused silica columns have been obtained by covalent chemical bonding both in the form of cross-links within the polymer as well as bonds from the phase to the silica surface. Such phases have dramatically increased the capacity and dynamic range of capillary columns without a sacrifice in resolution. Studies on alditol acetates of amino sugars have revealed that shorter retention times are obtained with non-polar or medium-polar columns, and that chemically bonded capillary columns give better resolution than conventional columns (see Fig. II.1). Oligosaccharide alditol acetates,^ and peralkylated oligosaccharide alditols?^ Can also be separated by this technique. Therefore the use of g.l.c, alone or in conjunction with mass spectrometry (g.l.c.-m.s.), is a powerful tool in the analysis and sequencing of oligosaccharides. II.2.4 High Performance Liquid Chromatography (H.P.L.C.)?^ High performance liquid chromatography also known as high speed or high pressure liquid chromatography is a recent development of liquid chromatography. Unlike gas-liquid chromatography where selectivity is achieved by use of columns with a wide range of stationary phases, only a limited number of supports are available for h.p.l.c. (e.g. silica, alumina, polar bonded phases, cation or anion exchangers, and reverse phase material). Resolution is thus obtained by variation of column temperature or composition of the mobile phase (e.g. by use of more than one solvent, the addition of organic solvents such as methanol or acetonitrile - 27 -to buffer solutions, variation of the pH and ionic strength of buffer solutions, or the additions of small amounts of amines, acids, buffers, electrolytes or detergents to mobile phases).7'* This technique is advantageous for the direct analysis of underivatized sugars. It is also useful in preparative separations of peralkylated oligosaccharides7^ and peralkylated oligosaccharide alditols.7^ However its limitations in the field of acidic polysaccharide analysis (acidic sugars and oligosaccharides containing uronic acids cannot be analyzed) far outweigh its advantages in direct application unless derivatization of the acidic function is resorted to. II.2.5 Gel-permeation40'3-*''' and Ion-Exchange Chromatography48''8 Gel-permeation chromatography is also known as gel filtration, or molecular-sieve chromatography. As the names imply, the method is based on fractionation of molecules of different sizes on a column consisting of a gel of a three dimensional network. The smaller molecules penetrate further into the pores of the gel than do the larger, and are retained longer on the column. The larger molecules are thereby eluted first. The applications of this technique include separation of mixtures of oligosaccharides (particularly mixtures produced by degradation with bacteriophage or by partial hydrolysis), purification of polysacharides (Section II.1), and the determination of molecular weights.^ Separations on ion-exchange resins are dependent on a combination of ion exclusion and fractionation by molecular size. A column packed with the resin contains two physically distinguishable liquid environments: the liquid inside the matrix of the cross-linked resin and that occupying the interstices between the particles of porous resin. Non-ionic solutes are eluted solely by a molecular-sieve - 28 -mechanism, whereas ionic solutes have an additional retention force exerted by the ionic matrix. This technique is commonly used in separating acidic mono- or oligo-saccharides from neutral sugars or oligosaccharides. Amino sugars and amino acids can also be separated from neutral sugars or polysaccharides using ion-exchange resins. II.3 Sugar analysis II.3.1 Total hydrolysis The quantitative hydrolysis of a polysaccharide into individual mono-saccharides with minimum degradation is the most important step in their analysis.* Hydrolysis with acid, methanolysis, acetolysis, trifluoroacetolysis and mercaptolysis yield monosaccharides or derivatized monosaccharides as the case may be. The choice of reagent depends on the sugars, the types of glycosidic linkages present in the polysaccharide, and the stability of the generated mono-saccharide to acid. In the following section the applications of some of these reagents are reviewed. For an excellent review on structure determination, and chemical characterization of polysaccharides see G.O. Aspinall, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 1, Academic Press, New York, (1982), pp 35-131. - 29 -II.3.1.a Hydrolysis with a c i d s 6 4 ' 7 9 - 8 6 Depolymerizations of polysaccharides have been successfully carried out using sulfuric and hydrochloric acids. The latter may be removed by volatiliza-tion. However hydrochloric acid usually causes more degradation of sugars than sulfuric acid, 7 9 and, therefore, the more volatile and easily removable trifluoro-acetic acid 8^ i s being used increasingly instead of the mineral acids. Preliminary hydrolysis with acetic acid 8* has been helpful in avoiding N-deacetylation of amino sugars, and formic acid is sometimes used in solubilizing methylated polysaccharides.8^'8^ Anhydrous hydrofluoric acid, being mild, does not remove the N-acetyl group from acetamido sugars, and is thus able to hydrolyze amino sugar containing polysaccharides in quantitative yields. 8 4' 8^ The use of ion-exchange resins for hydrolysis is popular in the field of glycoproteins. The resin may be used alone or with mineral acid, the best results being obtained in conjunction with hydrochloric acid.7^ Some sugars are more acid labile than others; for example 2-amino-2-deoxy sugars and uronic acids are more resistant to hydrolysis than neutral sugars. Deoxysugars, ketoses, and sialic acid are extremely acid labile and are liberated by mild acid on short exposure. One of the problems in hydrolysis is in choosing the ideal conditions. In many cases, no one method will cleave all the glycosidic linkages quantitatively. The use of strong acids (e.g. 4M HC1) 8 6 to cleave 2-amino-2-deoxy sugars usually leads to degradation of deoxy and keto sugars. Hence analyses using differing acid strengths and exposure times are necessary in order to obtain information on all the sugars present. Appendix I deals with some of the problems encountered in the hydrolysis of the amino sugar containing polysaccharides from E. coli K43, K44 and K45. t - 30 -II.3.1.b Methanolysis 8 7 - 8 9 Whereas hydrolysis in acid is conducted in an aqueous medium, methanolysis, acetolysis, and trifluoro acetolysis are done under non-aqueous conditions. Methanolic hydrogen chloride has the advantage of being less destructive of deoxy sugars and sialic acids. 8 7' 8 8 On methanolysis the formation of methyl glycoside methyl esters and methyl glycosides imparts stability to the normally acid labile sialic acids and deoxy sugars respectively. A method developed in our laboratory 8 9 to overcome the resistance of glycuronic acid linkages to hydrolysis involves methanolysis of the glycosyl linkages with the simultaneous esterification of the carboxylic acid, which is then reduced to the corresponding alcohol. The resulting mixture of neutral glycosides is hydrolyzed with acid and converted into alditol acetates for analysis by gas-liquid chromatography. II.3.1.C Acetolysis 9 0' 9 1 Acetolysis is used in cases where the parent polysaccharide (e.g. cellulose) is insoluble in aqueous systems, but the acetylated derivative is readily soluble in acetic anhydride-acetic acid mixtures containing sulfuric acid as the acid catalyst. Generally, acetolysis is advantageous in generating oligosaccharides by cleavage between 1—*-6 linkages since these are more resistant than other linkages to acid hydrolysis.90 Sialic acid linkages are unusually stable enabling the isolation of sialic acid-containing oligosaccharides on acetolysis. A modification of this method has been reported recently 9 1 whereby acetolysis was used to cleave methylated 2-amino-2-deoxy sugar linkages which are resistant to acid hydrolysis - 31 -due to the conversion of the N-methyl acetamido hexosyl residue into a positively + charged methyl amido hexosyl (CH3-N-) residue in acidic solutions. II.3.2 Characterization of the sugars The techniques used for identification and characterization of the sugars formed on hydrolysis have advanced from the historical crystallization and melting point determinations of derivatives, through paper chromatography,58,59 and thin layer chromatography9^ to the more innovative applications of gas-liquid chromatography,64'9^ high performance liquid chromatography,74 and mass spectrometry (Section II.4.3). G.l.c. and g.l.c.-m.s. are also used for characteri-zation of the products formed on deamination of amino sugars 9 4 and for the analysis of aldononitriles produced for the determination of the degree of polymerization of oligosaccharides.9-* Colorimetric analysis 9 6' 9 7 permits the classification of sugars into broad classes (hexoses, pentoses, uronic acids, deoxy or amino sugars, and sialic acid) but has limited applications for individual characterization. II.3.3 Determination of the absolute configuration of sugars' 8" 1 0 4 The use in routine analyses of the techniques described previously cannot distinguish between enantiomers of the same sugar. Only when quantities of material large enough for polarimetric measurement (optical rotation [«*]r> circular dichroism measurements98 of optically active alditol acetates, partially methylated alditol acetates, and aldononitriles) or enzymic studies (D-glucose - 32 -RCOOH + RCOO'H H" pH4.75 RCH20h NoBH4 NHR' N0BH4 pH 5-7 O I NHRn R&H +0=6 TH NHR' E.D.C. = l-ethyl-3- (3-dinethylartdnopropyl)carbcdiimide C.M.C. = 1-cyclohexyl-3-(2-morpholinoethyl)carbodiimide itietiio-p-toluene sulphonate Scheme II.1 Reduction of carboxylic acid in aqueous solution using carbodiimide reagent - 33 -oxidase," D-galactose oxidase 1 0 0) are isolated can the absolute configuration of the sugar be designated. A convenient approach involves the conversion of enantiomers into diasteromers by use of optically active alcohols ((-)-2-butanol, 1 0 1 (+)-2-octanol10^) and their subsequent separation by g.l.c. Direct resolution on a chiral stationary phase10-^ is also possible, but both methods result in complicated chromatograms obtained from anomeric mixtures of glycosides. A practical method involves the formation of the acyclic diastereomeric dithio-acetals by using (+)-l-phenylethanethiol10^ followed by separation as the TMS or acetylated derivatives by g.l.c. This technique eliminates the appearance of multiple peaks for a single diastereomer. II.3.4 Modification of uronic acids for gas-liquid chromatographic detection The carboxylic acid function prevents the detection of uronic acid and its derivatives by h.p.l.c. and g.l.c. analysis. The resistance of the glycosiduronic linkage to acid hydrolysis can hinder quantitative analysis of component sugars in a polysaccharide. Electrostatic repulsions between periodate and the carboxylate ion have resulted in incomplete oxidation of sugars. Hence the reduction of the carboxylic group into an alcohol eliminates the problems cited above. A good procedure is to treat the acidic polysaccharide in aqueous solution with water-soluble carbodiimide reagent 1 0^ to form an O-acylisourea, which on reduction with sodium borohydride yields the alcohol. With this method, unlike on methanolysis and reduction,89 the glycosidic linkages are left intact (Scheme II.1). The powerful reducing agent lithium aluminum hydride can only be used to reduce compounds soluble in ether-type solvents.10** Hence it is not suitable for unsubstituted polysaccharides, but will reduce the carboxylic esters in a - 34 -permethylated polysaccharide. Aqueous sodium borohydride can be used to reduce esters formed by acidic polysaccharides with diazomethane 107 and diborane is used in the preferential reduction of carboxylic acid in fully acetylated polysaccharides.108 II.4 Establishment of linkage position II.4.1 Methylation a n a l y s i s 3 6 ' 7 0 ' 1 0 9 " 1 2 1 This technique is based on the protection of free hydroxyl groups of a polysaccharide by etherification, 1 0 9 - 1 1 1 which acts as a label in distinguishing the originally unlinked positions of a sugar on its release by hydrolysis. Separation as their partially methylated alditol acetate derivatives by gas-liquid chromatography and identification by g.l.c.-m.s. gives an insight into the types of linkages existing in the polymer, but not of the sequencing, or the anomeric nature of the linkages. In the early days, the use of dimethyl sulfate and sodium hydroxide by Haworth 1 1 2 led to the formation of methyl ethers. The unavailability of diethyl sulfate at that time prevented the formation of the ethyl ether derivatives. Therefore the use of methyl iodide and silver oxide according to the method ofPurdie and Irvine1 ^  was still preferred because the reaction proceeds in a homogeneous phase allowing more complete methylation unlike in the two phase reaction by Haworth. The method of Purdie and Irvine was further improved by K u h n 1 1 4 who used, in addition, N,N-dimethylformamide to increase the solubility of the polysaccharide. A more convenient method, usually resulting in complete methylation, was devised by Hakomori 1 1^ wherein the dimethyl - 35 -sulfoxide-solubilized polysaccharide is treated with the strong base sodium methylsulfinylmethanide (sodium dimsyl) with subsequent addition of iodomethane.116 If undermethylation is evident, the Purdie method should be carried out on this material, since a second Hakomori treatment results in base catalyzed degradation of uronic acid-containing polysaccharides. Even in recent times methylation is more prevalent than ethylation due to the greater reactivity of iodomethane over iodoethane and the resistance to hydrolysis of ethylated polysaccharides in contrast to the methylated polymer.1 1 7 Complete ethylation has been effected by using the stronger base, potassium methylsulfinylmethanide (potassium dimsyl). 1 1 8 , 1 1 9 The use of a strong base in methylation is advantageous only in instances where base-labile substituents are absent. Many bacterial polysaccharides contain O-acetyl groups and pyruvic acetals. The former are easily removed by base resulting in loss of information on acetate linkage position while the latter are stable under basic conditions. Methyl trifluoromethane sulfonate is a milder base and effects methylation without cleavage of acyl groups in the presence of 2,6-di-tertiarybutylpyridine and trimethyl phosphate as solvent. 1 2 0 Hakomori methylation leads to the esterification of uronic acids and, in amino sugars, the N-methylation of acetamido groups. Completeness of methylation can be verified by the absence of hydroxyl absorption at 3600 cm"1 in the i.r. spectrum, or by analysis of the methoxy content. Insolubility and high viscosity of the polysaccharide can lead to undermethylation. The discovery of N-methylmorpholine-N-oxide (MMNO) as a potential solvent for the DMSO insoluble cellulose made possible methylation by the Hakomori method in a MMNO-DMSO mixture.36 This may be applied to methylations of other less soluble polysaccharides. Hydrolysis of the methylated polymer is usually performed with 2M tri-- 36 -Klebsiella K67 polysaccharide o 11 - + i) base (CH,SCH_ Na ) ii) CH3I MeO OMe r-OMe pOMe M e o LiAlH4/THF OMe MeO OMe 1. 4,6-Ote2-Mann°se 2. 2,4,6-Ofe-j-Mannose 3. 2,4,6-<*fe3-Glucose 4. 2,4-Ctfe2-Rhamnose 5. 2-CMg-Glucose 6. 2,3,4,6-CMe4-Galactose 7. 2,3,4-Ot'le3-Rhaninose i) NaBH4 ii) Ac20/pyridine AcO-AcO OAc OAc UQf.fe rQAc MsO-AcO--ais -OAc rQAc a-ie AcO--OAc -aie - OAc AcO-OMe JteCH OAc AcO *-OMe CH., OAc -CMs AcO-OAc MeO-QAc OAc OAc •Ctte MeO--OAc AcO l-OAc OMe •otfe CH., Scheme II.2 Methylation analysis of Klebsiella K67 polysaccharide - 37 -fluoroacetic acid on a steam bath (95°) for 18 h. Uronic acid containing polysaccharides are carboxyl-reduced before (carbodiimide reduction) or after (lithium aluminum hydride) the permethylation step. Scheme II.2 illustrates a typical reaction sequence. Amino sugar containing polysaccharides show better results with 2M hydrochloric acid at 95° for 4-5 h, or on acetolysis (see Section II.3.1.C) with 0.25M sulfuric acid in 95% acetic acid at 80° for 6 h followed by dilution with water and further heating for 3 h. Prolonged hydrolysis for liberation of glycosamines leads to degradation of neutral sugars and demethyla-tion, 1 2 1 particularly of 3-linked permethylated galactose. This prevented quantitative results being obtained for either procedure (see Appendix I). II.4.2 Characterization of methylated sugars The techniques for separating and identifying methylated sugars have evolved to the point where analysis by gas-liquid chromatography dominates all others. Paper chromatography is still used for primary identification by the colours formed on spraying with rj-anisidine hydrochloride and heating, and also by their Rf values. The uses of g.l.c. were dealt with in Section II.2.3. Extensive reviews on the application of g.l.c. to carbohydrate analysis have been published.64'65>93 The alditol acetate derivatives of the methylated sugars are used far more , than any other derivative due to the simplicity of the chromatograms obtained and ease of quantitation without using response factors (except when analyzing glycosamines122). Papers on methylation analysis containing relative retention times on capillary g.l.c. columns are numerous,12-* and many mass spectra of the methylated alditol acetates,111 have been published. The use of partially - 38 -ethylated alditol acetates 1 1 7 in identification of sugars is useful in separating sugars that are unresolved as the partially methylated derivatives on g.l.c. as well as in distinguishing from already existing methyl ethers. There are very little published data on methylated amino sugars. However, retention times and mass spectral data for various methylated derivatives of the more common 2-amino-2-deoxyglucose have been reported. 9 1' 1 2 4 The amino sugar derivatives have longer retention times than the corresponding neutral sugars. The appearance of spurious peaks has been attributed to contamination by phthalates (used as plasticisers) which can be differentiated by their charac-teristic peak at m/z 149 on mass spectral analysis.12^ Identification of derivatives is possible by comparison of retention times with known values on various columns, or by coelution with authentic samples, followed by confirmation of the substitution pattern obtained on g.l.c.-m.s. (see following section). II.4.3 Applications of mass spectrometry'U,izo-i:>& A mass spectrum consists of a plot of the relative intensities of gaseous ions formed.by ionization and subsequent fragmentation of the volatile molecules, against their mass-to-charge ratio (m/z). Different types of instrumentation can be used to record mass spectra. The inlet systems can be either a hot reservoir inlet, a direct probe inlet, or a g.l.c. inlet (g.l.c.-m.s.). Ionization of the molecules can be effected by electron impact techniques (e.i.), chemical ionization ( c . i . ) 1 2 6 ' 1 2 7 , field desorption (f.d.), 1 2 8 field ionization (f.i.), 1 2 9 or fast atom bombardment (f.a.b.). 1 3 0' 1 3 1 The non-volatile carbohydrates are analyzed usually as their volatile - 39 -Scheme II.3 Fragmentation pathways of a permethylated methyl glycoside - 40 -derivatives. The fragmentation patterns depend on the stability of the fragments produced, and are characteristic of the derivatives. Stereoisomeric derivatives usually give near identical mass spectra with very small differences in intensity. Thus, this technique does not permit assignment of configuration. G.l.c.-m.s. is invaluable in analyzing mixtures of carbohydrate derivatives and its application in the characterization of methylated/ethylated alditol acetates is extensive.70 Sequencing of volatile derivatives of oligosaccharides (acetates, 1 3 2' 1 3 3 T.M.S. derivatives, 1 3 4 methyl ethers 1 3^" 1 3 7) is also possible from the nature of the fragmentation patterns obtained (see later). The electron-impact technique is the standard method used in the analyses conducted in this laboratory. The high intensity (70 eV) beam of electrons used leads to extensive fragmentation of the molecules and results in a weak or no molecular ion peak. Decreasing the energy of electrons used does not cause a significant increase in the population of high mass ions. Molecular ions and larger fragments can be obtained by c.i., f.i. and f.d. mass spectrometry which favour minimal fragmentation. F.a.b.-m.s. is becoming more attractive since it has the ability to give both molecular weight and fragment data and can be used in analysis of unmodified carbohydrates and glycolipids. 1 3 8 II.4.3.a Characterization of monosaccharide derivatives / U' y i' 1- J y" 1 4 y Direct derivatization of monosaccharides by permethylation, peracetylation, pertrimethylsilylation resulting in anomeric mixtures of glycosides has been studied extensively. 1 3 9' 1 4 0 The major fragmentation pathways of these derivatives follow the pattern depicted in Scheme II.3. Alditols 1 4 1 are of more importance than other derivatives due to the absence - 41 -I B B 9 0 8 0 7 * 6 8 5 0 4 0 3 0 2 0 1 0 0 ' l I !•• I' i ¥ 1 I 'T i - i " | I I I I I I I I I | I I I I I I I -3 0 0 3 S 0 (a) 100 9 0 0 0 7 0 C0 5 0 4 0 1#1 44 129 t-r 1 5 0 i r''l i n ' i ' i i i i . i i | i i I [ i r i i i | i i i i i i i i r-pi i i i i i , i T " | -B 2 S 0 3 0 0 3 5 0 4 0 0 (b) Scheme II.4 Mass spectra of (a) hexitol hexaacetate, (b) 1,5-di-O-acetyI-2,3,4,6-tetra-O-methylhexitol - 42 -Primary fragmentation: CH OAc i 2 117 1 HCOMe 1 161 1 MeOCH 1 205 1 HCOMe 1 161 1 HCOAc 1 i 45 CH2OMe Secondary fragmentation: HOQMe 0 HC=OMe HCOMe -AcOH COMe | > II CH HCOAc CH2OMe m/z 205 CH2OMe m/z 145 CH„OAc HCOMe HC=OMe m/z 161 CH, -AcOH CH. COMe I HC=OMe e m/z 101 * II© H C N Q / H Me m/z 71 HC^ OMe -AcOH HCOAC CH2OMe m/z 161 HC=OMe -CH2CO m/z 129 © HCOMe I c=o I CH3 m/z 87 Scheme II.5 Fragmentation patterns of alditol derivatives - 43 -of anomers and ease of separation on g.l.c., and their mass spectra will be described in detail. The alditol acetates of interest are the peracetylated, partially methylated, and partially methylated/ethylated derivatives. All these derivatives show no molecular ions, and those having the same substitution pattern give similar mass spectra that are typical of the substitution pattern. Therefore the mass spectrum of D-glucitol hexaacetate is representative of all peracetylated hexitols, and that of l,5-di-0-acetyl-2,3,4,6-tetra-0-methylglucitol is identical to any l,5-di-0-acetyl-2,3,4,6-tetra-0-methyl hexitol (see Scheme II.4). The fragmentation pattern simply consists of primary fragments resulting from o-cleavage of the carbon atoms in the alditol chain, 7 0' 1 4 2 the intensities of which decrease with increasing molecular weight. Secondary ions are obtained by loss of acetic acid (m/z 60), ketene (m/z 42), methanol (m/z 32), or acetamide (m/z 59) (see Scheme II.5). The fission between carbon atoms is governed by the stability of the resulting radical, the methoxylated radicals being more stable than acetoxylated radicals. Preference of bond cleavage decreases in the following order with no observable cleavage between two carbon atoms if either one is deoxygenated.143 Me Me H I I I I I I I I I C_N-Ac -C-N-Ac -C-O-Me -C-O-Me -C-N-Ac -C-O-Ac I > I > I > I > I > I C-O-Me -C-O-Ac -C-O-Ac -C-O-Ac -C-O-Ac -C-O-Ac I I I I I I The primary fragmentation of amino alditol acetates is largely governed by the acetamido group. 1 2 4' 1 4 4 The secondary fragments are formed as previously described (see Scheme II.6). Partially methylated amino alditol acetates show fission almost exclusively between the methylacetamido group and the adjacent methoxy or acetoxy group.91 The primary fragment (m/z 158) formed by - 44 -102. 84 C H 2 ° A C 300 -60 / J U U -42 144 HCNHAc 360 -60 AcOCH HCOAc HGOAc CH2OAc -4 -59 259 -60 318 199 i 139 -60 258 V--59 -60 -42 (a) -60 114 -42 CH2OAc HCOAc 174 V-216 /-60 156 j -60 96 AcHNCH HCOAc CHOAc CH2OAc -42 288 -60 202 T -42 246 -60 -42 186 144 -42 228 4, -60 168 >126 -42 -60 (b) Scheme II.6 Fragmentation patterns of (a) 2-acetamido-2-deoxyhexitoI pentaacetate and (b) 3-acetamido-3-deoxyhexitol pentaacetate - 45 -100 90 at 70 50 40 I f 66 i—n 1*0 130 142 131 160 117 216 -60 CH OAc 58 1 -42 t 100 * HCQMe -42 T 142 -60 HCNMeAc 130 -30 -60 160 -42 156 -42 174 '-60 202 131 114 4 -42 72 HCQMe HCOAc 116 I I I M i l | I I 1 60 20* CH, i r i i r | i I i i i i i i i | i i i i i i i i i | i i i i i i i i i | 250 300 350 400 (a) i0> 90 8 0 70 C f 5 0 4 0 30 -2 0 " -i 129 142 -42 116 < 158 145 • 205 -60 CH2OAc HCNMeAc MeOCH -60 202 >142 161 HCQMe HCOAc 45 CH^ OMe I 'I I'" I-'-I' i i 'I i r i " | i T 1 1 i r i i I i [ 2 " 250 300 I 1 1 1 1 6 0 (b) Scheme II.7 Mass spectra and fragmentation patterns of (a) 1,5-di-O-acetyl-3,6-dideoxy-3-N-methyIacetamido-di-2,4-0-methyl-hexitol and (b) l,5-di-0-acetyl-2-deoxy-2-N-methyI-acetamido-3,4,6,tri-0-methy!giucitol - 46 -2-acetamido-2-deoxy amino sugars is eclipsed by the secondary fragment (m/z 116) formed by loss of ketene, the latter being even stronger than the usual base peak 1 4 2 CH3C=0 (m/z 43). A similar situation arises with 3-amino-3-deoxy sugars 1 4^ ( s e e Scheme II.7). Thus g.l.c.-m.s. is an invaluable tool in the analysis of amino sugars and methylated derivatives. Another area of application is in the analysis of ethylated sugars. 1 4 6 In the location of the acetate group in Klebsiella K44 polysaccharide, the position of the acetate function was replaced by a methyl group, and the polysaccharide ethylated. The alditol containing the methyl group was detected by the shift of several masses by 14 units on comparison with the alditols from the fully ethylated polysaccharide (see Scheme II.8). Labelling studies with deuterium to determine the reducing end, as in the case of the oligosaccharide from bacteriophage degradation of Klebsiella K44, result in derivatives where the masses are shifted by one unit when compared with the undeuterated derivative as shown by g.l.c.-m.s. Reduction of uronic acids with deuterides followed by g.l.c.-m.s. analysis is useful in distinguishing the acidic sugar from the existing neutral sugars. G.l.c.-m.s. of the partially methylated alditol acetates can be accomplished only if the relative retention times (T values) of the components in a mixture are sufficiently different. The existence of various columns, however, permits good separations in most cases, thus overcoming this problem. For materials containing one sugar of any one class (such as hexose, amino hexose, 6-deoxyhexose), the identification by m.s. is unambiguous.143 When two or more of the same class exist (e.g. glucose and galactose), the relative retention time on g.l.c. is also important in their assignment. The identification of aldononitriles and methylated aldononitriles by g.l.c.-m.s.147 is useful in determinations of the degree of polymerization of - 47 -1 0 0 SB 8 0 7 0 6 0 SB IB 3B 2B 1 5 9 7 3 I I •I, Hll.l l-l 2 2 5 (a) IBB SB B 0 7 0 8 0 5 0 4 0 3 0 2 0 1 0 J I i1 Ms (159-14) (Z01-H) I I56 l l r i I ' I 1 " I I • F 4 7 ( 2 6 1 - 14) i | i— i—r 2 5 0 (b) Scheme II.8 Mass spectra of (a) l,4,5-tri-0-acetyI-2,3,6-tri-0-ethylglucitoI and (b) ] ,4,5-tri-0-acetyl-2,3-di-0-ethyI-6-0-methylglucitol - 48 -OMe Scheme II.9 Fragmentation by A and B series of a permethylated disaccharide derivative - 49 -oligosaccharides. The 2-amino-2-deoxyhexoses having the gluco configuration at C-2 (e.g. glucosamine and galactosamine) form 2,5-anhydrohexitols on deamination.148 In the characterization of these amino sugars g.l.c.-m.s. is a valuable tool as it also is in the analysis of E. coli K44 which consists of both glucosamine and galactosamine. G.l.c.-m.s. of the methylated anhydro alditols 1 4 9 is useful in further confirmation of the linkage position of the original amino sugar. II.4.3.b Identification of derivatized oligosaccharides76'132'137'139,150-152 The ability to analyze intact oligosaccharides is severely limited by the thermal instability and very low volatility of these compounds. Nevertheless derivatization into the volatile acetates, 1 3 2' 1 3 3 methyl ethers 1 3^" 1 3 7 and trimethylsilyl ethers 1 3 4 enables their analysis by mass spectrometry. Although the use of mass spectrometry in sequencing of derivatized oligosaccharides has increased considerably over the years, e.i.-m.s. frequently fails to show their molecular ions. However c.i.-m.s.1-*0 has shown greater potential by its ability to provide molecular ions and large fragments. The mass spectrometric behaviour of permethylated oligosaccharides 1 3^"1 3 7 has been investigated, and studies on the disaccharide derivatives 1 3 9 showed that the fragmentation proceeds in a manner similar to that of the monosaccharide derivatives. Scheme II.3 depicts the nomenclature used by Chizhov and Kochetkov which was later modified by Kovacik and coworkers. 1^ 1' 1^ -p n e example given in Scheme II.9 is for the degradation of a disaccharide methyl glycoside. Thus baA] denotes that the ion has been formed by cleavage of ring b following pathway A and is substituted with ring a. - 50 -CH,OMe CH,OMe MeO CH,OMe CHjOMe CH,OMe MeO CH,OMe I HCOMe I HCOMe O-^—CH I HCOMe I CHjOMe 0 CH,OMe I HCOMe I HCOMe © I CH HCOMe CHjOMe Scheme 11.10 Fragmentation pattern of a permethylated oligosaccharide alditol - 51 -The A-series of fragments establishes the molecular weights of the component sugar residues. The B-series of fragments obtained by degradation of ring b can be used to establish the nature of the linkage between the two sugars, since 0-6 and 0-3 linked disaccharides fail to follow this pattern of fragmentation. Ions of the D-series arising from scissions of both C1-C2 and C4-C5 bonds are diagnostic of 6-linked disaccharides, the characteristic aJ] ions forming on further degradation of the Dj ion. Derivatives of permethylated oligosaccharide alditols have the added advantage of being susceptible to analysis by g.l.c.-m.s. In contrast to the permethylated reducing oligosaccharides they give single peaks on g.l.c, and their mass spectra are simpler. The sugar units have fragmentation patterns typical of the monosaccharide and the alditol portions. The non-reducing sugars are identified by the A-series of fragments and the alditol part by a fragment 3 produced by the D-sequence of fragmentation14^'1-*1 (Scheme 11.10). The B-series serves the purpose of establishing linkages as previously described. This fact was used to confirm the linkage of the permethylated oligosaccharide alditol obtained from the Smith degradation of Klebsiella K80 polysaccharide. Analysis of peralkylated oligosaccharides by g.l.c.-m.s. and h.p.l.c.-m.s., and their importance in determining the sugar sequences have been elaborately described by Albersheim and coworkers.7^ The rapid expansion of the application of m.s. in structure elucidation ranging from simple monosaccharides to the biologically important glycoconjugates is proving its importance in the field of carbohydrate chemistry. - 52 -II.5 Sequencing of sugars II.5.1 Partial hydrolysis 1 5 3 - 1 6 1 Isolation of oligosaccharides generated by partial hydrolysis is a major key to the elucidation of the sequence of sugars in a polysaccharide. The method exploits the acid lability of some glycosidic linkages over others which are more resistant to hydrolysis. Rate constants and kinetic parameters for the acid-catalyzed hydrolysis of various glycosides have been reviewed. 1 5 3' 1 5 4 The rate of hydrolysis is affected by several factors which include the ring size, configuration, conformation, position of linkage, the polarity of the sugar as well as the size and polarity of the aglycone. 1 5 3 Hence it is difficult to single out one factor in order to explain observed differences in hydrolysis rates. The following generalizations can be made for broad classes of sugars: (i) Furanosides are more labile than pyranosides. (ii) Deoxy sugars and pentopyranosides are more easily hydrolyzed than hexopyranosides. (iii) a-Glycosides are generally more labile than ^-glycosides. (iv) 1—>6 glycosidic linkages are more resistant to acid hydrolysis than are. 1—>4 and 1—*-2, with 1—*~3 linkages being the most easily hydrolyzed. Terminal non-reducing sugars and those in side chains are more easily cleaved than in-chain glycosidic bonds. (v) Uronic acids and 2-amino-2-deoxy glycosides need strong conditions to effect hydrolysis. (vi) 2-Acetamido-2-deoxy glycosides are easily hydrolyzed if the hydrolytic conditions prevent formation of the acid resistant 2-amino-2-deoxy - 53 -derivative by N-deacetylation. In some instances, hydrolysis under non-aqueous conditions is necessary to preserve certain linkages or to avoid degradation of more labile sugars. Normal conditions of hydrolysis degrade 3,6-anhydrohexoses, sialic acid and KDO. In such cases methanolysis, acetolysis 1^ (see Section II.3), trifluoroacetolysis, or mercaptolysis is favoured. N-acetyl hexosamine-containing oligosaccharides and glycoconjugates are degraded by transamidation during trifluoroacetolysis. 1^ The degradation occurs by a process in which the trifluoroacetyl groups stabilize the glycosidic linkages of the sugar residues, and subsequently cleave the peptide (amide) bonds.1^7 Anhydrous hydrogen fluoride is capable of cleaving glycosidic linkages of amino, neutral and acidic sugars without degradation of the sugars.84 This reagent is useful in quantitative analysis of polysaccharides and in cleavage of sugars-from glycoproteins leaving the peptide moiety intact. 8 4 The temperature dependance of the rate of cleavage of glycosidic linkages 8 4' 1^ 8 by HF can be used in effecting differential cleavage of neutral and acidic sugars.1 ^ 9 This has been utilized in the generation of oligosaccharides by preferential cleavage of the glycosylamino linkage over the glycosylaminouronic linkage,1**0 and in the generation of acetate and methoxy containing oligosaccharides.1^9 Studies carried out by Jennings and Lugowski using aqueous HF indicate its ability to cleave phosphate esters and alditol aglycones without affecting other glycosidic linkages.1**1 Hydrolysis by acid is also used in removal of acid labile pyruvic acetals and labile terminal non-reducing sugars in side chains without cleavage of the in-chain glycosidic linkages. Klebsiella K67 and K80 both contain a terminal rhamnosyl unit in the side chain that can be removed while the rest of the - 54 -PYRANOSES (end groups) and 6-linked ,-0-(H) HO )—0 2 10, -0-HO OH CHO 0-(H) 0. BH./H 4 y g l y c e r o l 2 - l i n k e d HO OK 1 10, HO 0— 3-linked. no r e a c t i o n ;> HO OH FURANOSES |_|Q end groups (~)H-0 0-(H) 5- or 6-linked #8 HO OH l io. 2 10, 1 10, 0 -H -0 OH -OH H HO CHO OHC i io" HH - 0 > (-)H-0 0-(H) -0-OHC^  B I V H + i <* ) g l y c e r o l BH4/H ,Q_ ^ uno x i d i z e d sugar BH~/H+ _^ e r y t h r i t o l or t h r e i t o l BH 4/H + ^ g l y c e r o l BHT/H + pentose e r y t h r i t o l or t h r e i t o l Scheme 11.11 Common products formed by terminal and mono substituted sugars on periodate oxidation followed by borohydride reduction and hydrolysis - 55 -polymer is left intact. Partial hydrolysis of K67 enabled the isolation of di, tri, tetra and penta-saccharides all of which contained glucuronic acid at the non-reducing terminal. II.5.2 Periodate oxidation*0^"*7** and Smith degradation Oxidative cleavage of 1,2 diol groups1**^ by sodium metaperiodate1*^ or lead tetraacetate1*'4 is of analytical importance in structure determinations of polysaccharides. Oxidations are usually carried out in aqueous media with the water soluble metaperiodate ion, lead tetraacetate in acetic acid being used only in instances where the polysaccharide is insoluble in water. Periodate oxidation is a quantitative reaction; each 1,2 diol consumes one mole of periodate and is oxidized to an aldehyde by cleavage of the carbon bonds. 1,2,3-Triols liberate formic acid 1*^ b y double cleavage of the carbon chain, and exocyclic diols produce formaldehyde,1*'*' the products being analyzed by titration and colorimetry respectively. The reduction of the periodate ion to iodate can be monitored by titration or spectrophotometrically.1*^ Substituted rings are oxidized in various ways. The "polyaldehyde" produced may then be reduced with sodium borohydride into the polyol and hydrolyzed. The products of such a series of reactions can lead to information on the original substitution pattern (see Scheme 11.11). The rate of oxidation varies according to the configuration of the glycols. It is generally observed that open chain glycols are oxidized at a faster rate than cyclic cis glycols. Cyclic trans glycols are oxidized more slowly or not at all if fixed in an unfavourable conformation as in some bicyclic anhydrohexoses.1*5^ Kinetic studies conducted on some polysaccharides revealed that oxidation of - 56 -COOH Fig II.2 (a) Formation of hemiacetals resulting in under-oxidation of oxidizable sugar units and (b) their susceptibility to oxidation on reduction with sodium borohydride - 57 -pyranoid rings followed rather complex second order type rate constants.'*'7 In accordance with the observation that cis glycols were oxidized faster than trans, Ebisu et.al.1*'8 were able to selectively oxidize terminal /3-D-galactopyranosyl residues in the Pneumococcus S-14 polysaccharide leaving the 1—M linked 0-D-glucose units in the main chain intact. In a similar experiment, the selective oxidation of the terminal rhamnopyranosyl over the galactopyranosyl residue in Klebsiella K67 failed due to the presence of trans 1,2 diols in both residues; the two sugars were oxidized at the same rate. Non-ideal behaviour of polysaccharides during periodate oxidation results from both under- and over-oxidation. Over-oxidation is usually encountered with glycuronic acids, where it is caused by dehydrogenation at C-5 followed by hydroxylation.1**9 Such non-specific oxidations may be minimized at low pH (2.2-4.0), but will lead to exposure of previously protected diol groups. The pyruvic acetal at position 2,3 on the terminal rhamnosyl unit in Klebsiella K80 polysaccharide was removed during the oxidation, and probably may have survived if a buffer at pH 6.0-7.0 had been used. Incomplete oxidation (under-oxidation) is a common problem leading to errors in assignment of linkage positions. A major cause is the formation of hemiacetals between unoxidized hydroxyl groups and aldehyde functions of adjacent oxidized sugar units. 1 7 0 Reduction with sodium borohydride disrupts the hemiacetals reexposing the unoxidized hydroxyl groups to further oxidation (see Fig. II.2). Electrostatic repulsions between periodate ions and weakly acidic 1 7 1 groups (carboxylic acids) in acidic polysaccharides result in under-oxidation. This problem can be overcome by the addition of a salt (0.2M sodium perchlorate) to minimize the repulsion of periodate ion from the negativity charged domain of the uronic acid. Sugars containing free amino groups are oxidized if a 1,2-relationship - 58 -a) P r o p o s e d scheme f o r i n t e r - r e s i d u e h y d r o g e n b o n d i n g between a c e t a m i d o g r o u p s and o x i d i z a b l e h y d r o x y l g r o u p s p r e v e n t i n g t h e o x i d a t i o n of the 4 - l i n k e d 8 - g l u c u r o n i c a c i d . — 0 \ b) P r o b a b l e h y d r o g e n b o n d i n g between g a l a c t o s a m i n e and t h e 4 - l i n k e d g l u c u r o n i c a c i d i n E.c o 1 i K44 p o l y s a c c h a r i d e . Scheme 11.12 Inter-residue hydrogen bonding between acetamido groups and hydroxyl groups - 59 -between the free amine and a free hydroxyl groups exists. An acetyl group on the amine or hydroxyl function prevents its oxidation. Hence a 2-acetamido-2-deoxy glucopyranosyl unit is oxidized only if it is terminal in a side chain, or linked at position 6; in both cases the 3,4 diol is the active participant. Under-oxidations involving amino sugars result from hydrogen bonding between the acetamido group and an oxidizable hydroxyl group on the neighbouring sugar residue. 172,173 The resistance to periodate of the 4-linked glucuronic acid residues in E. coli K44 polysaccharide was attributed to inter-residue hydrogen bonding between a susceptible hydroxyl group and the acetamido group of the sugar flanking it (see Scheme 11.12). The most important application of the periodate oxidation is in the generation of oligosaccharide fragments which can lead to conclusive structural evidence in a polysaccharide. The Smith degradation 1 7 4 involves mild acid hydrolysis of the polyol at room temperature wherein the acyclic acetals are cleaved in preference to the more resistant glycosidic linkages. A complication arising from the Smith degradation is the formation of a glycolaldehyde acetal from the reaction between the glycol aldehyde of an oxidized sugar with its alditol fragments. 1 7 4' 1 7^ In the Smith degradation of Klebsiella K80, such glycolaldehyde acetal formation resulted in several compounds consisting of the same monosaccharide sequence showing diffent chromatographic mobilities (see Section IV.2.3). A second problem arises from the presence of acid labile glycosidic linkages (e.g. deoxy sugars) which prevent the use of strong acids to effect cleavage of the more resistant acetal fragments formed by the oxidative cleavage of hexuronic acids. These problems can be avoided by using the modification proposed by Lindberg and coworkers,17*' where the polyol is methylated prior to mild acid hydrolysis. The "aglyconic" hydroxyl groups liberated after hydrolysis yield - 60 -valuable structural information on realkylation with ethyl iodide or trideuteriomethyl iodide. A typical sequence of periodate oxidation and Smith hydrolysis is shown in Scheme 11.13 for Klebsiella K80 capsular polysaccharide. K 0 H •o) \?1 HON— 1) NalO. 2) NaBH 4 OH HOH 2C C H3 / HOH2C COOH -OH -OH CH20H erythronic acid H2C0H \?1 H0H2C— y 1) Hydrolysis 2) Reduction 3) Acetylation g a l a c t i t o l , mannitol, hexaacetates, glyc e r o l t r i a c e t a t e e r y t h r i t o l tetraacetate Scheme 11.13 Periodate oxidation and Smith hydrolysis of the Klebsiella K80 polysaccharide - 61 -II.5.3 Base catalyzed /3-eIimination from hexuronic acid residues 1 4 9' 1 7 7* 1 8^ On esterification the carboxyl groups in uronic acid residues acquire an electron withdrawing character and thus favour (3-elimination on treatment with a suitable base. 1 7 7 The substituent eliminated is a better leaving group when it is etherified or is an aglycone. Application of this method to polysaccharide a n a l y s i s 1 4 9 ' 1 7 8 ' 1 7 9 involves reaction of the permethylated polysaccharide with sodium dimsyl to eliminate the 4-0-(glycosyl or methyl) substituent with formation of a hex-4-enopyranosiduronate residue. On treatment with mild acid, the extremely acid labile hex-4-enopyranosiduronic linkage is cleaved 1 8 0 to expose the "aglyconic" hydroxyl group which formerly carried the hexuronic acid residue. Alkylation with ethyl iodide or trideuterio- methyl iodide followed by g.l.c.-m.s. then reveals the position of the uronic acid linkage. The main steps involved in this reaction are outlined as follows: C O O C H 3 b a s e ( D i m s y l ) J  0 C O O C H , F u r t h e r d e g r a d a t i o n by b a s e f r o m l i b e r a t e d r e d u c i n g g r o u p s d e p e n d i n g on n a t u r e and l o c a t i o n o f s u b s t i t u e n t s. e x t e n d e d b a s e t r e a t m e n t o r i m i l d a c i d h y d r o l y s i s H O O -I f u r t h e r a 1 k y l a t i o n R O ' ( i d e n t i f i c a t i o n o f s i t e o f e x p o s e d h y d r o x y l g r o u p ) - 62 -It was discovered that complete cleavage of the hex-4-enopyranosiduronate was possible under normal conditions of base degradation, and further alkylation may be performed in a one-pot procedure by avoiding the mild acid hydrolysis step. 1 8 1 One of the problems encountered in this reaction scheme is the base catalyzed degradation of the reducing sugar liberated by ^ -elimination. The carbonyl group of the reducing sugar is electron withdrawing and effects elimination of the substituent at position 3, with the formation of 3-deoxy hex-2-enopyranoses. When position 3 contains a glycosyl residue, the elimination results in the exposure of its reducing end which can further degrade causing a "peeling" reaction along a polysaccahride chain. When the base 1,5-diazabicyclo-[5,4,0]undec-5-ene (DBU) is used to effect the /3-elimination the loss of structural information may be avoided if the exposed reducing groups are immediately protected by acylation with acetic anhydride. 1 8 2 II.5.4 Deamination of aminosugars 8' °- w° Free amino groups in sugars can be deaminated by nitrous acid to form various derivatives. Since most polysaccharides contain N-acetyl groups, N-deacetylation is essential prior to deamination. Hydrazinolysis with anhydrous reagent in the presence of a catalytic amount of hydrazine sulfate, 1 8 3 or treatment with sodium hydroxide in aqueous DMSO with thiophenol as oxygen scavenger 1 8 4 results in removal of the acetamido function. Depending on the position of the amino group and its configuration, the deamination process follows a basic pattern. Equatorial amino functions at position 2 of pyranoses (e.g. glucosamine) result in the formation of 2,5- anhydro hexoses. Thus glucosamine is converted into 2,5-anhydromannose148 and 3-0-glycosyl s u b s t i t u e n t 4-0-glycosyl s u B s t i t u e n t Fig II.3 Deamination sequences of some amino sugars - 64 -galactosamine into 2,5-anhydrotalose.185 These occur as the major products resulting from the ring contraction which involves the O-Cl bond migration, inversion of configuration at C-2, and subsequent cleavage of the glycosidic linkage. A small amount of 2-C-formyI pentofuranoside is formed by an alternate ring contraction involving the participation of the C3-C4 bond as shown in Fig. II.3. Axial amino functions at position 2 (e.g. mannosamine) deaminate by a simple mechanism with replacement by OH. This results in inversion at C-2 without a change in ring size or cleavage of glycosidic linkages. 1 8 5 Hexoses containing equatorial amino functions at position 3 deaminate to form two C-formyl pentofuranosides without rupture of any glycosidic linkages (see Fig. II.3). The major product is formed by the migration of the C4-C5 bond. The furanoside from the involvement of the C1-C2 bond is produced in small yield due to the lower nucleophilicity of C-l which is bonded to two oxygen atoms.185 Dutcher and coworkers made several unsuccessful attempts to deaminate 3-amino-3,6-dideoxy-D-mannose.186 A related problem was probably the reason for the failure to observe the products on deamination of the 3-amino-3,6-dideoxyhexose present in E. coli K45 (see Appendix I). The presence of an axial 4-amino group in 2-acetamido-4-amino-2,4,6-trideoxy-D-galactose (from Streptococcus pneumoniae type l 1 8 7 and other bacterial sources 1 8 8' 1 8 9) facilitated the study of its behaviour on deamination. The products obtained lead to the conclusion that the three pathways followed in the deamination result from: (1) the displacement by OH with inversion at C-4; (2) a 5—*-4 hydride shift with subsequent attack at C-5 by OH; and (3) a 3—*-4 hydride shift with elimination of the 3-linked substituent to yield a 3-keto derivative 1 9 0 (see Fig. II.3). - 65 -When the reaction involves glycosidic cleavage, the deamination process is useful in obtaining oligosaccharides from the polysaccharide 1 8 7 since the amino sugar is usually resistant to acid hydrolysis. II.6 Determination of the anomeric configuration of linkages II.6.1 Optical r o t a t i o n 1 9 1 The existence of multiple chiral centres in carbohydrates renders them optically active. The determination of the anomeric configuration is based on the rotational contribution made by the asymmetric anomeric centre. Van't Hoff's Principle of Optical Superposition,192 which proposed the addition of rotational contributions of different asymmetric centres in a complex molecule, was used by Hudson 1 9 3 to determine the configuration at the anomeric centre of free sugars and glycosides. Application of the famous Hudson's rules of isorotation gives information on the overall molecular rotation of a poly- or oligo-saccharide, but neglects the contribution of O-acyl and pyruvate groups to the total molecular rotation. The specific rotation of a compound is given by: « x 100 •> where t temperature 1 . c X wavelength (usually the Na D-line) length of sample tube (dm) c concentration of the solution (g/100 mL) the observed rotation - 66 -The molecular rotation M is defined by: [a] x M.W. M = 100 Hudson's isorotation rules can be applied to predict specific rotations of oligosaccharides and polysaccharides 1 9 4 by comparison of the individual glycosidic linkages involved with model compounds (methyl glycosides) of known molecular rotations. Thus M =y ;^M; where Mj is the molecular rotation of the component sugars. This method for anomeric classification has now been overshadowed by the more informative technique of nuclear magnetic resonance spectroscopy. However it is still useful for checking the anomeric configurations of oligo- and poly-saccharides. II.6.2 Nuclear magnetic resonance spectroscopy The use of proton and carbon-13 n.m.r. spectroscopy is a valuable tool in the determination of the configuration of unknown carbohydrates and for ascertaining the conformations of known sugars in solutions. This technique can be used in polysaccharide analysis only if regular repeating-units exist, making a uniform, homogeneous polymer. Random structures result in broad lines with less information and poor resolution. Thus the sharp signals in the spectra obtained for bacterial polysaccharides serve as proof of the regular repeating-units present - 67 -within them. II.6.2.a H^-n.m.r. spectroscopy 1 9 5" 2 0 7 Proton magnetic resonance (p.m.r.) spectroscopy is a firmly established, widely used technique for the structural, configurational, and conformational analysis of carbohydrates and their derivatives. Since the first application of p.m.r. to carbohydrates 1 9 5 over two decades ago, rapid progress has been made in instrumentation and techniques used. The introduction of magnets based on superconducting solenoids 1 9 6 has increased the field strength available, thereby enhancing the resolution of the spectra (notably of polysaccharides). Fourier-transform techniques are important in enhancement of sensitivity, 1 9 7 especially in averaging signals from dilute samples or from nuclei of low natural abundance (e.g. 1 3C, 1 5N). Other recently evolved methods include the nuclear Overhauser effect (n.O.e.), internuclear double- and triple- resonance techniques,196 and many others. Many of these methods have limited application in polysaccharide analysis due to the complexity of the spectra. High field ^-n.m.r. spectroscopy is advantageous in conformational analysis of both polysaccharides 1 9 9 and glyco-proteins. 2 0 0 The use of two dimensional homo- and hetero-nuclear n.m.r. gives enhancement of resolution. 2 0 1 These methods are valuable in obtaining coupling information and assigning chemical shift values. 2 0 2 Interpretation of the n.m.r. spectrum requires measurement of various parameters. - 68 -(i) The relative areas or integrals of individual signals. The number of protons resonating at each particular frequency is proportional to the area of the signal produced.20-^ Thus the integrals are indicative of the relative number of anomeric linkages, 6-deoxy sugars, N- and O-acetyl groups, and 1-carboxyethylidene substituents present in a sample; they also indicate the quantitative amounts of reducing sugars including pyranose and furanose forms in mono- and oligo- saccharides. (ii) Coupling constants Nuclear spin-spin coupling constants over two chemical bonds are designated as 2J(N1,N2)> a n c* o v e r three bonds as ^J(NI N2)- Couplings for vicinal protons are hence expressed as ^ JHI,H2- These are particularly useful in determining the configuration of the protons (mainly anomeric) because of the existing relationship between the vicinal coupling constant (J) and the dihedral angle (<J>) between the protons. An approximation of this relationship is given by the Karplus equation. 2 0 4 3 j(Hl,H2) = 8.5 Cos 2* - 0.28, 0° < <i> < 90° 9.5 Cos 2* - 0.28, 90° < <*> < 180° In a first order spectrum, the magnitude of the coupling constant can be determined directly from the spectrum. Large vicinal couplings (-8-10 Hz) indicate antiparallel protons (diaxial, J A A ) , whereas small values (-2-4 Hz) are typical of protons in the gauche form 2 0^ (diequatorial, J E E and axial-equatorial, - 69 -ft = H,OH Rj - H Fig II.4 Relationship between dihedral angle (4>) and coupling constants for a- and /3- D-hexoses - 70 -J a e). The coupling constants predicted by use of the Karplus equation are usually in good agreement with observed values (see Fig. II.4). Deviations from this equation occur when substituents are of different electronegativity, as in peracetylated derivatives, 2 0 3 or N-glycosidically linked glycoproteins.200 The coupling constants of anomeric protons are thus useful in establishing the configuration (a,/3) as well as the overall conformation (pyranose, furanose, chair/boat forms) of the sugars. Geminal coupling of the C-6 protons (^JH6A,H6B) i n pyranose forms are usually not measured because the signal is hidden in the ring proton region. The presence of an acetate at 0-6 in Klebsiella K44 enabled measurement of this coupling constant by causing a downfield shift of the signals. The coupling constants for geminal protons are usually much larger (>8 Hz) than for vicinal protons and sometimes are of negative sign. 2 0 3 They are more useful in determining conformations of pentoses in the pyranose form. (iii) Chemical shift The chemical shifts of protons depend on many factors. Substitution, orientation of the molecule, electronegativity effects of neighbouring and distant groupings, and the nature of the solvent can induce protons to resonate at different field strengths. A polysaccharide spectrum contains three main regions: a) the anomeric region (5 = 4.5-5.5); b) the ring proton region ( 5 = 3.0-4.5); and c) the high field region ( 8 = 1.0-2.5) (see Fig. II.5). The direct shielding effect caused by the ring-oxygen atom in carbohydrates accounts for the characteristic low-field shift of the anomeric protons. 2 0 6 As a general rule, the axial ring protons resonate at a higher field than their equatorial counterparts. Hence for a pair of anomers in the pyranose form, the a anomer CH 3 o f 6-deoxy 6(p.p.a.) Fig II.5 Schematic representation of different regions in the 'H-n.m.r. spectrum of polysaccharides - 72 -has a lower chemical shift than the /3 anomer (axial proton). The substituents at C-2 and C-3 have a large effect on the anomeric protons. Equatorial substituents at C-2 and C-3 yield an anomeric pair with a difference of 0.6 p.p.m. in their chemical shifts (e.g. a -Glc 5 = 5.23, (3-Glc 6=4.64). Axial substituent at C-2 yields an anomeric pair having a chemical shift difference of 0.2-0.3 p.p.m.20-^  (e.g. a-Man <5 = 5.10, /3-Man 8 = 4.90). The ring protons are usually unassigned except when specific protons resonate at lower field (H-5 of glucuronic or galacturonic acids, H-5 of fucose, H-2 of mannose, rhamnose etc) or at a higher field (H-2 of (3-glucose and ^-glucuronic acid). The chemical shift for the methyl group of the pyruvic acid acetal changes significantly with its orientation (axial or equatorial), and thus permits the determination of the configuration of the acetal carbon present on certain hexoses in some extracellular bacterial polysaccharides.207 Line broadening of signals, and interference by exchangable protons (O-H and N-H) affect the quality of the high resolution ^ H-n.m.r. spectrum.1" The latter is minimized by the prior exchange of these protons with deuterium oxide (D2O) and by the use of D2O (preferably 99.95 atom %) as solvent. Nevertheless, a strong peak due to residual water (HOD signal) is often obtained. The chemical shift of the HOD signal at room temperature ( 8 -4.8) interferes with the anomeric signals in this region. Elevating the temperature results in an upfield shift of the HOD signal ( 8 -4.3 at 90°) thereby exposing the valuable anomeric region. F.T. techniques (e.g. saturation decoupling where the short relaxation time of the polymer proton is used to advantage over the larger relaxation time of the solvent proton) can also be used to minimize interference by the HOD signal. Signal broadening is caused by the relatively short relaxation times of the polymer protons; i.e. line width is proportional to the rate of spin-spin relaxation - 73 -(a) T 1 1 r 5.0 4.0 3.0 2.0 6 (p.p.m.) Fig II.6 1H-n.m.r. spectra of (a) Native K80 polysaccharide and (b) after depyruvalation - 74 -( T 2 ) .1" This problem may be minimized by recording the spectrum at a higher field which also results in increased resolution, or at an elevated temperature (70-90°). Reducing the viscosity of the polysaccharide by autohydrolysis or mild hydrolysis also helps in reducing line broadening. However these procedures are not recommended when extremely labile linkages or substituents are present. Irregular distribution of substituents in a polysaccharide results in twinned signals. Klebsiella K80 polysaccharide contains pyruvic acetals at positions 3 and 4 on some of its terminal rhamnosyl units and thus gives two anomeric signals. A similar observation is made in Klebsiella K44 which contains partial amounts of acetate which produces twinning of the anomeric signal of the acetate containing sugar. Such twinnings disappear on removal of the substituent responsible for the effect. This is illustrated in Fig. II.6 which shows the 1H-n.m.r. spectra of Klebsiella K80 polysaccharide, both in the native form and after depyruvalation. II.6.2.b l3C-n.m.r. spectroscopy 2 0 7" 2 1 7 The advent of Fourier-transform techniques has resulted in a steady increase in the use of 13C-n.m.r. spectroscopy as a tool in the structure elucidation of polysaccharides.208 13C-n.m.r. is complementary to ^ H-n.m.r. spectroscopy, and has the ability to give better signal separation and thereby more information owing to the wider range of chemical shifts involved. 2 0 9 This technique, which is rapid and non-destructive, has great potential in the study of polysaccharides of biological origin where only small amounts of material are available for analysis. The 13C-n.m.r. spectrum also employs many of the same parameters used in the interpretation of p.m.r. spectra, a notable exception being the unsatisfactory - 75 -integration caused by saturation phenomena and n.O.e. effects, which result in loss of information of relative numbers of nuclei resonating at each chemical shift value. 2 1 0 Nevertheless, if the spectra are measured under suitable conditions, comparison of integrals of carbon atoms carrying the same number of hydrogen atoms often yields accurate information about the relative amounts of components in a mixture. 2 0 8 An important condition for obtaining correct integrals is a good signal to noise ratio (s/n). High field instruments, large sample tubes and increased concentrations will increase the s/n; but too high a concentration can lead to line broadening, which, in turn, has an adverse effect on the s/n. 2 0 8 Simple but well defined C^-n.m.r. spectra, which can be easily interpreted to yield a wealth of information, are obtained using proton decoupled conditions. 2 1 0 However single bond ^C-^H coupling [*Jci Hi] x s useful i n differentiating anomeric pairs in the pyranose form, since the J values differ by -10 Hz. 2 1 1' 2 1 2 Identification of furanose forms by this procedure is unreliable due to the small difference (-2-3 Hz) between the coupling constants of the 213 anomers. 1 3 The chemical shifts of individual, unsubstituted carbon atoms of poly- and oligo-saccharides show reasonable agreement with those obtained for previously assigned monosaccharides.214 In certain instances the sensitivity of chemical shifts to effects of substitution renders this technique very useful in the determination of structures of unknown compounds and their linkages. Substi-tution by a glycosyl unit causes a large increase in the chemical shift of the carbon atom directly involved in the linkage, with small changes (usually a decrease) in the chemical shifts of the neighbouring /3-carbon atoms. 2 0 9' 2 1 4 A similar large shift occurs in the a-carbon atom on O-alkylation, with a much smaller shift occurring if the oxygen is acylated. Replacement of an oxygen by C 0 C - l 1 75 1 1 0 C - l p a HCOR HCOH CH 2OH 0 HCN CH 3 C I ^ N O-C-0 CH^CO 1 0 0 70 6 0 5 0 30 r-2 0 CH 3C — i p . p . m , Fig II.7 The characteristic regions for resonances of carbon atoms belonging to different monosaccharide residues in polysaccharides - 77 -nitrogen (amino sugars) usually results in a large decrease (upfield shift) in the chemical shift of the carbon atom bonded to the nitrogen, 2 0 8 thereby enabling its easy identification (see Fig. II.7). The enormous upfield shift in the a-carbon atom caused by substitution of an oxygen by a hydrogen atom reflects the influence of the changes in electronegativity brought about by substitution. The potential of studies in conformational analysis depends on the sensitivity of the chemical shift towards conformational changes. Although many pairs of anomers give quite different signals for the anomeric carbon atoms, no general relationship has been discovered between the anomeric configuration and the chemical s h i f t 2 0 8 of pyranoses. However, furanoses with trans oriented substituents at C-l and C-2 (e.g. /3-galactofuranoside) show higher anomeric chemical shifts than the corresponding cis isomer (a-furanoside).21^ A change in the ring size from the six membered pyranose form to a five membered furanose ring results in a downfield shift of the anomeric signal. Glycosylation at C-l amounts to O-alkylation, and hence the anomeric carbon atom shows a corresponding substitutional shift to lower fields (usually -3-8 p.p.m.). Thus, in a polysaccharide spectrum, the anomeric region (100 ± 8 p.p.m.) is well separated from the ring carbon atom region (75 ± 5 p.p.m.), with the primary hydroxy-methyl carbon atoms (at C-6) resonating away from the ring region at a higher field (65 — 5 p.p.m.) Other characteristic chemical shift values are: carbonyl groups from uronic acid, pyruvic acid, N and O acetyl substituents at 175 - 6 p.p.m.; nitrogen containing (amino groups) ring carbon atoms at 48-55 p.p.m.; methyl groups from O and N acetyl substituents at 20-28 p.p.m.; methyl groups from pyruvate at 18-26 p.p.m.; methyl groups of 6-deoxy sugars at 15-17 p.p.m. (see Fig. II.7). The chemical shift values are also useful in the determination of the stereochemistry of acetal carbons in pyruvates since axial methyl groups resonate - 78 -at -18 p.p.m. and equatorial groups at -26 p.p.m.207 The carboxylic acid at C-l of K D O 2 1 6 and sialic a c i d 2 1 7 is sensitive to the configuration at the anomeric centre (C-2) and resonates at characteristic values corresponding to the a or /3 anomer. Thus the use of 13C-n.m.r. in the structure elucidation of polysaccha-rides leads to valuable information on the nature of linkages, and anomeric centres, the presence of functional groups, and the stereochemistry of certain substituents. II.6.3 Other techniques II.6.3.a Enzymatic hydrolysis 2 1 6 The use of enzymes to hydrolyze a specific sugar with a defined anomeric configuration is made possible by the existence of glycanases and glycosidases. Exoglycosidases are used to remove non-reducing terminal sugar units and are unaffected by the position of linkage to the adjacent residue. On the other hand, exoglycanases which act in a similar manner are highly sensitive to the linkage type, and a different linkage type or branch point will terminate its activity towards the specified sugar. 2 1 8 Thus their use is usually restricted to oligosaccharides generated from modifications of the polysaccharides. In contrast, endo enzymes are capable of cleaving both internal and external chains of polysaccharides, but only the unbranched regions are acted upon. Branch points and other structural constituents are immune to action by endoglycanases. Bacteriophages are known to have highly specific endoglycanase activity with the ability to cleave branch points, thus yielding oligosaccharides from some capsular polysaccharides (see Section V). - 79 -II.6.3.b Chromium trioxide oxidation 2 1 9" 2 2^ Chromium trioxide in acetic acid can be used to oxidize ethers and acetals into esters without affecting acetates 2 1 9 or other existing ester groups. Thus a fully acetylated glycopyranoside is oxidized into a 5-hexulosonate.220 The rate of oxidation of a gycosides is considerably slower than that of the )3 anomer, thus permitting the estimation of the anomeric configuration by comparison with the original unoxidized sugars.221 The a anomers would therefore survive, but the oxidized /3 anomers are conspicuous by their absence on g.l.c. analysis. OR This supposition is valid for the a pyranosides of the common hexoses, 6-deoxy hexoses, 2-acetamido-2-deoxy hexoses, and xylose. However substitution can change the conformational equilibrium of a-rhamnosyl and a-fucosyl residues, making them vulnerable to oxidation. Methylation by the Hakomori procedure cleaves all ester linkages and methylates the exposed hydroxyl groups, thereby yielding structural information valuable for sequence analysis. 2 2 2 This was an asset in the analysis of the K80 capsular polysaccharide, where the /3 galactose unit was oxidized and its linkage position detected on methylation. More importantly, the configuration of the rhamnosyl unit was established as /3 linked by partial oxidation; its linkage position was confirmed by the presence of the tri- and tetra- methyl glucose obtained in fractional molar proportions. - 80 -This method is also useful in generating oligosaccharides where possible from polysaccharides for structural analysis.22^ II.7 Location of O-acetyl groups 2 1 0' 2 1 1' 2 2 4" 2 2 8 Some polysaccharides contain naturally occuring acetyl groups, either as esterified hydroxyl groups or as amides from amino sugars, which may be of immunological significance. The extreme base lability of O-acetyl groups create difficulties in the unambiguous location of their linkage positions. Nuclear magnetic resonance is useful in detecting the presence of acetyl groups; p.m.r. spectroscopy shows a singlet at 8 = 2.15-2.28 for the methyl protons, and ^C-n.m.r. a signal at 21-23 p.p.m. for the methyl carbon atom. The presence of the acetyl group causes a shift in the neighbouring protons, 2 2 4 which may not be apparent in a complex spectrum of a polysaccharide. However, the presence of acetate at position 6 of the glucose residue in the oligosaccharide from the bacteriophage degradation of Klebsiella K44 was easily detected by the large downfield shift of the two protons at that position. More information on the linkage position may be obtained from a ^ C-n.m.r. spectrum due to the larger shift of the signal on acetylation. 2 1 0 Comparison of the spectra of the acetylated and deacetylated polysaccharides can often provide useful information on the location of the O-acyl group. 2 1 1 The presence of acetate on C-6 of hexoses may be assigned unambiguously by such comparison22^ e.g. Klebsiella K44 polysaccharide (see Section V.2). Quantitation of the O-acetyl group is possible by p.m.r. spectroscopy and by g.l.c.,22^ with precise information being obtained by spectrophotometric determination.227 - 81 -Chemical methods used for location of acetate groups are: (a) periodate oxidation (see Section II.5.2) which provides information on acetate containing sugars only if the acetate substituent and other glycosidic linkages are suitably located; (b) Prehm methylation 1 2 0 (see Section II.4.1); and, (c) the method of De Belder and Norrman 2 2 8 where the free hydroxyl groups are protected with methyl vinyl ether, followed by base catalyzed O-deacetylation which leaves the base stable methoxyethyl acetals intact. The resulting free hydroxyl groups are then methylated. I C-OH I C-OAc CH 2=CHOCH 3 > TsOH I OCH--C-O-CH 1) base | \ CH 3 > -C-OAc 2) CH 3I I I / OC H 3 1 C-O-CH H 3 0 + -C-OH I \ C H 3 > I C-OCH3 -C-OCH3 I I Scheme 11.14. Location of O-acetyl substituents using the methyl vinyl ether protection method Hydrolysis of the modified polysaccharide removes the acetal protection, and analysis of the sugars produced demonstrates the positions of the methyl groups (which were originally the positions occupied by the acetate substituents). However, "underprotection" sometimes causes false estimation of acetate groups not previously present. Such underprotection was detected when treating the polysaccharide from Klebsiella K44; but its oligosaccharide (from bacteriophage - 82 -degradation) gave very good results. When multiple sugars of the same kind are present, a problem arises as to which sugar residue contains the acetate group. The oligosaccharide from Klebsiella K44 showed that it contained acetate at C-6 on one of its glucose units. This was resolved by mild acid treatment of the modified oligosaccharide (which removed the acetal protection) followed by ethylation and then hydrolysis into individual sugars. The fact that the two sugars had different linkage positions enabled the assignment of the methyl containing ethylated sugar derivative as the original acetylated residue on g.l.c.-m.s. analysis (see Section V.2.3). -83-III. GENERAL EXPERIMENTAL CONDITIONS - 84 -III. GENERAL EXPERIMENTAL CONDITIONS 111.1 Paper Chromatography Paper chromatography was performed by the descending method using Whatman No. 1 paper and the following solvent systems: A) 1-butanol - acetic acid - water (2:1:1) B) ethyl acetate - acetic acid - formic acid - water (18:3:1:4) C) ethyl acetate - pyridine - water (8:2:1) D) 1-butanol - ethanol - water (4:1:5, upper phase) Preparative paper chromatography was carried out by the descending method using Whatman No. 1 paper and solvents A and/or B. The relevant strips were cut out and eluted with water overnight. The aqueous solutions were filtered, concentrated and freeze dried. Chromatograms were developed with alkaline silver nitrate, or by heating at 110° for 5-10 min after spraying with p.-anisidine hydrochloride in aqueous 1-butanol, or by dipping in an acidified aqueous acetone solution containing 0.25% ninhydrin in 1-butanol followed by heating at 110° for 5 min. 111.2 Gas-liquid chromatography and g.l.c.-m.s. spectrometry Analytical g.l.c. separations were performed using: (i) a Hewlett-Packard 5700 instrument fitted with dual flame ionization - 85 -detectors. Stainless steel columns (1.8 m x 3 mm) were used with a carrier-gas nitrogen flow-rate of 20 mL/min, or (ii) a Hewlett-Packard 5890A capillary gas chromatograph fitted with flame ionization detectors. A fused silica capillary col umn (15 m x 0.256 mm) was used with a carrier gas helium flow rate of 1.1 mL/min. The following packing materials and programs were used (unless otherwise stated): (A) 3% of SP-2340 on Supelcoport (100-120 mesh), programmed from 195° for 4 min, and then at 2°/min to 260°; (B) 5% of ECNSS-M on Gas Chrom Q (100-120 mesh) at 170°, isothermal; (C) 3% of OV-225 on Gas Chrom Q (100-120 mesh), isothermal at 220°; and (D) 5%/min of SP-1000 on Gas Chrom Q (100- 120 mesh) at 220°, isothermal. The capillary column was coated with DB-17 and programmed at either a) 180° for 2 min and then 5°/min to 220° or b) 180° for 1 min and then 2°/min to 250°. All injections were checked for reproducibility. Preparative g.l.c. was conducted with an F & M model 720 dual-column instrument fitted with thermal conductivity detectors. Stainless-steel columns (1.8 m x 6.3 mm) were used with a carrier-gas (helium) flow rate of 60 mL/min. Unles otherwise stated the packings and conditions used were: a) 5% of SP-2340 on Supelcoport (100-120 mesh) at 190°, and 4°/min to 260°; and b) 5% of OV-225 on Supelcoport (100-120 mesh) at 200°, and 4°/min to 260°. G.l.c.-m.s. analyses were performed with either: a) a V.G. Micromass 12 instrument fitted with a Watson-Biemann separator using columns A, B, C, or D; or b) a Nermag R 10-10 mass spectrometer with capillary columns coated with - 86 -DB-17, DB-225, or SE-30. The spectra were recorded at 70 eV with an ionization current of 100 MA and an ion source at 200°. Mass spectra of per-O-methylated oligosaccharides were recorded with a Kratos MS-50 model mass spectrometer using a 70 eV beam of electrons. III.3 Gel-permeation and ion-exchange chromatography Preparative gel-permeation chromatography was performed using a column (2.5 x 70 cm) of Bio-Gel P-4 (400 mesh). The concentration of the sample applied to the column was 200 mg/mL. The column was irrigated with water-pyridine-acetic acid (500:5:2) at a flow rate of 6.2 mL/h. Fractions of 4 mL were collected, freeze dried, weighed, an elution profile obtained, and then chromatographed on paper. Ion-exchange chromatography for separation of acidic and neutral oligomers was performed on a column (2.0 x 22 cm) of Bio-Rad AG 1-X2 (200-400 mesh, formate form). The neutral fraction was eluted with water, and the acidic with 10% formic acid. Deionizations of neutral and acidic sugars were carried out with Amberlite IR-120 (H +) resin by elution with water. Amino sugars were separated from hydrolyzates by applying them to a column (1.0 x 15 cm) of Amberlite IR-120 (H +) and eluting the neutral and acidic sugars with water. Amino sugars were eluted with 1-5% hydrochloric acid. Deionizations of methylated acetolysis products containing amino sugars were performed on a column (1.0 x 15 cm) of Bio-Rad AG 3-X4A (20-50 mesh, acetate form) with methanol as the eluant. - 87 -111.4 Optical rotation and circular dichroism Optical rotations were measured on aqueous solutions at 21°- 4 on a Perkin-Elmer model 141 polarimeter with a 1 dm cell (5 mL). Circular dichroism spectra (c.d.) were recorded on a Jasco J-20 automatic recording spectropolarimeter with a quartz cell of 0.3 mL capacity and a path length of 0.1 cm. Compounds were dissolved in spectral grade acetonitrile and spectra were recorded in the range 230-300 nm. 111.5 Nuclear magnetic resonance spectroscopy Proton magnetic resonance spectra were recorded on Bruker WP-80, Nicolet-Oxford H-270, or Bruker WH-400 instruments. Spectra were obtained at ambient temperature or at 90° - 5 with acetone as the internal standard. All values are given relative to that of internal sodium 4,4-dimethyl-4-silapentane-sulfonate taken as 0. Samples (10-20 mg) were prepared by dissolving in D2O and freeze drying 2-3 times from D2O solutions. Tubes with a diameter of 5 mm were used. ^C-n.m.r. spectra were recorded on a Bruker WH-400 spectrometer at ambient temperature. Samples (30-50 mg) were dissolved in D2O and acetone added as an internal standard. Tubes of 5 or 10 mm in diameter were used. 111.6 General conditions The i.r. spectra of methylated derivatives were recorded on a Perkin-Elmer - 88 -model 457 spectrophotometer. The solvent used was carbon tetrachloride (spectroscopic grade). All solutions were concentrated on a rotary evaporator in vacuo at a bath temperature of 40°. Spectrophotometric determinations were carried out on a Pye Unicam PU 8800 uv/vis spectrophotometer. III.7 Formation of alditol acetates and peracetylated aldononitriles a) Alditol acetates: The hydrolyzate (5-10 mg) containing monosaccharides was dissolved in water (1-2 mL). NaBH4 (50-100 mg) was added to the solution which was stirred for 2 h or overnight. The excess borohydride was destroyed with a few drops of 50% acetic acid, and the hydrolyzates that did not contain amino sugars were deionized on a column of IR-120 (H +) resin. The eluate was evaporated to dryness, and coevaporated with methanol (3x5 mL) to remove borate. The product obtained was dried in vacuo (2-3 h), dissolved in pyridine - acetic anhydride - 1:1 (2 mL), and heated on a steam bath for 45-60 min. The excess pyridine and acetic anhydride were removed by successive evaporations (5x5 mL) with water. b) Aldononitriles (p.a.a.n. derivatives) The hydrolyzate containing the monosaccharides was evaporated to dryness and to it was added 5% hydroxylaminehydrochloride in pyridine (0.2 - 89 -mL/mg aldose). The solution was heated on a steam bath for 15-20 min, and cooled. Acetic anhydride (0.2 mL/mg aldose) was added and heated a further 1 The solution was diluted with water (3-5 mL) and extracted with C H C I 3 . 111.8 N-deacetylation of polysaccharides The polysaccharide (15-20 mg) was dissolved in water (0.5 mL) and DMSO (2.5 mL) and NaOH (200 mg) and a drop of thiophenol were added. The mixture was stirred in a stoppered vial for 15 h at 80°. The excess base was neutralized with 2 M H C 1 and the solution was dialyzed overnight. The resulting solution was centrifuged and the supernatant was freeze dried. The extent of N-deacetylation was checked by p.m.r. spectroscopy. 111.9 Deamination Hydrolyzates of polysaccharides (5 mg) and N-deacetylated polysaccharid (5-10 mg) were dissolved in water (0.5 mL) and to them were added 3 3 % acetic acid (1.0 mL) and 5% NaNC>2 0-0 mL). After stirring at room temperature for 1 h, the mixture was diluted with water (3.0 mL) and freeze dried. The product obtained was dissolved in water (2.0 mL) and reduced with sodium borodeuteride (50 mg) for 2 h. The excess NaBD4 was destroyed with 50% acetic acid, and the solution was deionized with IR-120 (H +) resin. The borate was removed by successive coevaporations with methanol. - 90 -III.10 Isolation and purification of the polysaccharides III.10.1 Klebsiella polysaccharides Stab cultures of Klebsiella serotypes K26, K44, K67 and K80 were received from Dr. I. 0rskov (Copenhagen). The following media were used in the growth of the bacteria: (i) Beef extract medium ("nutrient broth"): Bactopeptone (5 g), bacto beef extract (3 g), NaCl (2 g), H 20 (1 L). (ii) Nutrient agar: Agar (15 g)/nutrient broth (1 L). (iii) Sucrose-yeast extract-agar: Sucrose (75 g), bacto yeast extract (5 g), agar (37.5 g), NaCl (5 g), K H 2 P 0 4 (2.5 g), MgS04.7H20 (0.625 g), CaS0 4 (1.25 g), H 20 (2.5 L). The bacteria were streaked on nutrient agar plates and incubated at 37° (20-24 h). Individual colonies were inoculated in beef-extract medium (100 mL) and cultured at 37° for 5 h with vigorous shaking. The actively growing culture was spread on a tray (86 cm x 46 cm) of sucrose-yeast extract-agar and left in a reasonably sterile atmosphere for 3 d. The slime produced together with the bacteria was scraped from the agar surface and added to an equal volume of 2% phenol solution to destroy the live organisms. The polysaccharide was separated from the cells by ultracentrifugation (30,000 r.p.m.) for 3 h. The viscous supernatant was decanted, and the polysaccharide was precipitated with 3 volumes of ethanol. The precipitate was dissolved in a minimum quantity of water and reprecipitated with a 10% cetyltrimethylammonium bromide (Cetavlon) solution. The precipitated acidic polysaccharide was isolated by centrifugation - 91 -and dissolved in 4M NaCl. The polysaccharide was precipitated with 3 volumes of ethanol, dissolved in 2M NaCl, and reprecipitated with ethanol. This precipitate was dissolved in water and dialyzed against running tap water for 2-3 d. The dialyzed solution was ultracentrifuged, and the supernatant was freeze dried. III.10.2 Escherichia coli polysaccharide Stab cultures of Escherichia coli K43, K44, and K45 were obtained from Dr I. 0rskov (Copenhagen). The following media were used to grow the bacteria: (i) Mueller-Hinton broth (MHB): Bacto Mueller Hinton broth (Difco 0757-01) (21 g), NaCl (2 g), water (1 L). (ii) Mueller-Hinton agar: (a) Mueller Hinton agar (BBL 11438) (38 g), NaCl (2 g ), water (1 L), or (b) agar (15 g)/MHB (1 L). Bacteria were streaked on Mueller Hinton agar plates and incubated at 37° overnight. Cultures were made as for Klebsiella, but using MHB (50 mL). The cultures were spread on trays (30 x 50 cm) of Mueller Hinton agar (1.5 L/tray) and left in a sterile atmosphere for 3-4 d. The slime was collected and purified according to the procedure outlined for Klebsiella polysaccharide. - 92 -III.11 Bacteriophage isolation and propagation III.11.1 Isolation of bacteriophages from sewage The bacterial media used were as follows: (i) Tenfold broth: (a) Nutrient broth - using the same ingredients but 100 mL of water instead of 1 L (for Klebsiella) (b) Mueller Hinton broth - same quantities as for normal broth but made up in 100 mL of water (for E. coli). (ii) Raw sewage effluent: (from four sewage processing plants, G.V.R.D. Board, B.C.). (iii) Bacterial culture: One colony of bacteria from a freshly grown plate was inoculated into normal broth (30 mL) (Klebsiella in nutrient broth, and E. coli in MHB) and grown with vigorous shaking for 5 h at 37°. Raw sewage effluent (900 mL), tenfold broth (100 mL), and bacterial culture (30 mL) were combined in an Erlenmeyer flask (1 L) and incubated at 37° for 20-24 h. 50 mL of this solution was transferred into a small flask; C H C I 3 (10 mL) was added and the flask shaken vigorously. The suspension was centrifuged and the clear supernatant was drawn off. A few drops of C H C I 3 were added to it to prevent bacterial growth. The crude solution was analyzed by the "plaque assay technique" for the presence of bacteriophage (see Section III.11.2a). Purification was carried out by repicking single plaque forming units in three consecutive assays. - 93 -III.11.2 Tube and flask lysis Bacteriophages 4>44 and 4>26 for Klebsiella polysaccharides and bacteriophages <£33, 4>43, and <t>44 for E. coli polysaccharides were isolated from sewage (Klebsiella <t> 44, courtesy of Dr. S. Stirm, Freiburg, Germany and Klebsiella 4>26, Dr. J.L. Di Fabio, The University of British Columbia, Vancouver, B.C., Canada). <t>44 was propagated on its host Klebsiella strain in nutrient broth and 026 on its host strain in P-Medium by both tube and flask lysis. The E. coli bacteriophages were propagated on their respective host strains in MHB by tube lysis. The dialyzable P-medium used instead of nutrient broth is made up as follows: Solution I: D-Glucose 25 g/ H 20 (1 L). Solution II: Difco Casamino acid (6.25 g), L-tryptophan (0.40 g), L-cystein (0.30 g), K H 2 P 0 4 (2.50 g), Na 2HP0 4.12H 20 (15.60 g), NH4C1 (1.30 g), Gelatine (1.01 g), H 20 (1 L). Solution III: MgS04.7H20 (50.0 g), FeS0 4 (0.1 g), H 20 (1 L), few drops of hydrochloric acid. Solution IV: CaCl 2 50 g/H2Q (1 L). All solutions were sterilized and with solutions I, II, III, IV mixed in a 7.0-7.2 by the addition of a few drops stored at 4°. P-Medium was reconstituted ratio of 20:80:1:0.1 and the pH adjusted to of NaOH. a) Tube lysis: A fresh single colony of bacteria derived from an actively growing culture - 94 -was obtained by successive replating on agar plates, and inoculated in broth (3 mL). The culture was incubated at 37° with vigorous shaking for 3 h, and the actively growing culture (0.5 mL) was transferred into 6 x 5 mL of sterile broth. After incubation of the tubes at 37° for 1 h, 0.5 mL of a bacterio-phage-containing solution was added to each test tube at 30 min intervals. Continued incubation resulted in gradual clearing of the cloudy solutions due to cell lysis. After 4-5 h of incubation a few drops of chloroform were added to each tube to prevent bacterial growth. The last two tubes to clear were combined and centrifuged to remove bacterial debris. The solution was then analyzed by the "plaque assay technique". This technique was based on making tenfold dilutions (10"1, 10"2, 10"3 ...) by successive transferring of 0.3 mL portions of phage into 2.7 mL portions of diluent (broth). One drop from each dilution was then placed on a bacterial "lawn". The "lawn" was prepared by inoculating 2 mL of broth with an actively growing bacterial colony and incubating for 3 h. The actively growing culture was spread on a previously dried agar plate, and the excess liquid removed after 1-2 min. The plate was left to dry with the lid partly open (15-20 min), and then incubated for a further 1 h to give the stable "lawn". The plate was then incubated at 37° overnight. At high phage concentrations (usually 10"* - 10"5 dilutions) individual phage particles could not be distinguished, but at suitable dilutions (10"6 - 10'8) individual plaques, sometimes surrounded by halos, could be easily counted. The titer of the bacteriophage solutions in plaque forming units (P.F.U.) per mL was then calculated: number of plaques x dilution Bacteriophage titre = drop size (mL) - 95 -b) Flask lysis: This technique is similar to that described for the tube lysis, except that larger volumes of bacteriophage are produced. The actively growing bacterial culture (3 mL) was inoculated into 6 x 50 mL of broth and incubated for 1.5 h after which 3 mL of bacteriophage solution was added to each flask at 30 min intervals. The procedure was then continued as described for the tube lysis. A yield of 300 mL of phage solution with a titre of 10 1 0 P.F.U./mL was usually obtained. -96-IV. STRUCTURAL INVESTIGATION OF Klebsiella CAPSULAR POLYSACCHARIDES - 97 -IV. STRUCTURAL INVESTIGATION OF Klebsiella CAPSULAR POLYSACCHARIDES IV.1 Structure elucidation of the capsular polysaccharide of Klebsiella serotype K67 IV.1.1 Abstract The determination of the structure of the capsular polysaccharide from Klebsiella K67 involved the use of methylation, periodate oxidation, partial hydrolysis, and /3-elimination. The nature of the anomeric linkages was established by using ^H- and C^-n.m.r. spectroscopy, and was further confirmed by chromic acid oxidation of the fully acetylated polysaccharide. The polysaccharide was found to have the heptasaccharide repeating-unit shown. A structure having a branched side chain is unique in this series of capsular polysaccharides. 3)-a-L-Rha-(l->3)-a-D-Man-(l —3)-a-D-Man-(l-»3)-/3-D-Glc-(l 2 T l 0-D-Gal-(l — 3)-/3-D-GlcA 4 T 1 a-L-Rha - 98 -IV.1.2 Introduction Klebsiella serotype K67 is one of the two strains whose capsular polysaccharides are composed of D-glucuronic acid, D-galactose, D-glucose, D-mannose, and L-rhamnose. The other strain, Klebsiella K14, has a 1-carboxy-ethylidene disubstituent (see Appendix II). The heptasaccharide repeating-unit of K67 is unique in the series of Klebsiella capsular polysaccharides in the sense that the side chain D-glucuronic acid is branched. Thus, the pattern may be considered to be a "four-plus-two-plus-one", instead of "four-plus-three" (which would imply a linear branch of three sugar residues). IV.1.3 Results and discussion Composition and n.m.r. spectra - Previously described methods were used in order to isolate and purify the polysaccharide.56'57 The purified product obtained after Cetavlon precipitation was shown to be homogeneous by gel-permeation chromatography ( M w = 1.8 x 106 daltons). The optical rotation of the polysaccharide [<*]D -17.3° was found to be in reasonable agreement with the value of -3.8° calculated using Hudson's rules of isorotation.l 9 3 The presence of galactose, glucose, mannose, rhamnose, and glucuronic acid in the acid hydrolyzate of the polysaccharide was observed by paper chromato-graphy. Determination of the neutral sugars as the alditol acetates, and as the peracetylated aldononitrile (PAAN) derivatives,95 gave rhamnose, mannose, glucose, and galactose in the ratios of 2.4:1.6:1.2:1.0. The carboxyl-reduced polysaccharide89 gave rhamnose, mannose, glucose, and galactose in the ratios of - 99 -Table IV.1 N.M.R. data for Klebsiella K67 polysaccharide and the derived oligosaccharides ComDOund3- 'H-N.m.r. data 13C-N.m.r. data & , CSzl Inteeral orotons Assignment2- P.p.m.i Assienment^ -GlcA-L-^ Man-OH 5.30 s 0.8 2-Man-OH 6 4.97 2 0.2 CI 2-Man-OH B 4.55 8 1.0 GlcA-6 GlcA-LlMan-LlMan-OH 5.29 s 1.0 2-Man— .a ^Man-OH 102.6 GlcA-B 6 a A3 5.17 s 0.6 100.6 2-Man-4.91 2 0.4 ^Man^OH 97.8 a ^Man-OH 4.55 8 1.0 8 GlcA-B 94.5 8 3-Man-OH a GlcAi-^ManJ-lMani-^Glc-OH B a a A4 5.27 2 1.4 2- Man— 3- Glc-i)H a 102.5 GlcA-8 5.23 s 1.0 2-Man— a 101.6 101.5 2-Man— a 4.66 8 0.6 3-Glc-OH ft 100.8 2-Man— a 4.57 8 1.0 R GlcA-8 96.7 93.0 2G1C-OH 8 2G1C-OH a GlcA-L-^ManLlMan-LJkjlc-L^Rha-g a a B •OH 5.26 s 1.0 2-Man-a 102.3 GlcA-P A5 5.22 s 1.0 3-Man-a 3-Rha-OH a 101.5 3-Glc-B 5.03 2 0.4 101.45 101.42 2-Man— a 2-Man-4.90 2 0.4 3-Rha-OH 6 ct 4.68 8 4.55 8 1.28 s 1.0 1.0 3.0 2Glc-8 GlcA-6 C-6 of Rha 100.9 95.9 19.9 3-Rha-OH 8 3-Rha-OH a C-6 of Rha - 100 -Table IV.1 .. continued • Compound3- H^-N.m.r. data  lj 2 Integral Assignment^ (HZ) protons l3C-N.m.r. data  P,D.m.& Assignment6-^Rha-LlManLlManJ-^Glci-2 a 1 GlcA PI 5.33 1 1.0 2-Man-a 5.27 1 1.0 3-Man— a 5.10 1 1.0 3-Rha-a 4.73 8 1.0 3-Glc-B 4.53 8 1.0 GlcA-- 6 1.30 4 3.0 C-6 of Rha iRha-LiMan-L-^ Man-L-^ Glc-L. 5.29 1 GlcA 3 B 1 Gal SH 3 R h a J-^ Ma n-L^ Ma n L i e 1 c-1 a 2 <* a t Rha-LlGlcA a 3 Gal K67 1 1.0 2-Man- 104.7 3-GlcA-a e 5.26 1 1.0 3-Man - 103.7 2G1C-cc e 5.04 2 1.0 3-Rha-a 103.1 2Gal-s 4.86 8 1.0 Gal-g 101.5 3-Man— 4.72 6 1.0 101.1 2-Man— a 4.54 8 1.0 3-GlcA-g 99.6 3-Rha-a 1.30 7 2.0 C-6 of Rha 17.5 C-6 of Rha 5.29 1 1.0 2-Man— Cl 105.1 3-GlcA-6 5.26 1 1.0 3-Man- 105.0 3-Glc-O 5.03 , 1.0 Ct 3-Rha- 103.3 P 3-Gal-a B 4.90 1 1.0 Rha- 101.6 3-Man— Ct a 4.84 8 1.0 Gal-D 101.2 2-Man— 4.72 8 1.0 P 2GIC-c 99.5 Cl 3-Rha- ' 4.50 8 1.0 p 3-GlcAg- 96.2 Rha-17.4 C-6 of Rha 3- For the sources of A 2 , A3, A4, A5, PI, and SH, see text, k Chemical shift relative to internal acetone; 2.23 downfield from sodium 4,4-dimethyl-4-silapentane- 1-sulfonate (DSS). £ The numerical prefix indicates the position at which the sugar is substituted; a or 8, the configuration of the glycosidic bond, or the anomer of a (terminal) reducing, sugar residue. Thus -3-Glc-^  refers to the anomeric proton of a 3-linked glucosyl residue in the 0-anomeric configuration. The absence of a numerical prefix indicates a (terminal) non reducing group. Chemical shift in p.p.m. downfield from Me4Si, relative to internal acetone; 31.07 p.p.m. downfield from DSS. £- As in £ but for nuclei. - 101 -2.1:2.0:1.6:1.0, indicating that the uronic acid is glucuronic acid. The glucose and mannose were proved to be of the D, and rhamnose to be of the L, configuration by circular dichroism measurements98 made on the alditol acetates. Galactose was assigned the D configuration from circular dichroism measurements made on a partially methylated derivative. The 'H-n.m.r. spectrum of the polysaccharide (see Appendix III, Spectrum No. 1) indicated the presence of seven anomeric protons, corresponding to 3 a [ 5 5.29 (1H), 6 5.26 (1H), and 5 5.03 (1H)], 3/3[5 4.84 ( J 1 > 2 8 Hz, 1H), 5 4.72 ( J 1 2 8 Hz, 1H), and 6 4.50 ( J 1 2 8 Hz, 1H)], and one borderline signal [ 5 4.90 (Jj 2 1 H z> The latter was proved by chromium trioxide oxidation 2 2 1' 2 2 2 to be an a linkage, resulting in 4 a and 3 /3 linkages. The assignments were based on chemical shift values and coupling constants, with further confirmation by the chromium trioxide oxidation results. There were no signals observed to indicate the presence of O-acetyl or pyruvic acetal substituents. The 13C-n.m.r. data (see Appendix III, Spectrum No. 2) were in agreement with a heptasaccharide repeating-unit, with the 0 signals resonating at 105.1 p.p.m., 105.0 p.p.m., 103.3 p.p.m. and the a signals resonating at 101.6 p.p.m., 101.2 p.p.m., 99.5 p.p.m. and 96.2 p.p.m. The signals were assigned on comparison of the 1H- and 13C-n.m.r. spectra of the polysaccharide with the corresponding spectra of oligosaccharides (obtained from partial hydrolysis) and modified polysaccharides, obtained from periodate oxidation and selective hydrolysis (see Table IV.1). Methylation analysis - Methylation of the K67 polysaccharide, followed by hydrolysis, conversion of the neutral sugars into alditol acetates, and g.l.c.-m.s. analysis thereof, gave the values shown in Table IV.2, column I. Reduction of the - 102 -Table IV.2: Methylation analysis of native, and degraded K67 polysaccharide Methylated sugar-S. (as alditol acetates) Mole %£ Column (ECNSS-M) Column C^ (OV-225) Column DS (SP1000) i i II III 2,3,4-Rha 0.44 0.51 0.52 23.4 12.8 16.1 2,4-Rha 0.99 0.93 1.00 17.7 14.8 11.6 2,3,4,6-Gal 1.20 1.13 1.14 21.9 15.6 18.8 2,3,4,6-Glc 1.92 1.63 1.71 15.0 15.0 12.2 2,4,6-Man 1.97 1.70 1.82 15.0 15.0 12.2 4,6-Man 3.20 2.53 2.58 7.0 14.4 14.4 2,6-Glc 3.52 2.94 2.69 - - 14.4 2-Glc 8.58 5.04 4.92 _ 13.1 _ •= 2,3,4-Rha = l,5-di-0-acetyl-2,3,4-tri-0-methylrhamnitol, etc. Relative retention-time referred to 2,3,4,6-Glc as 1.00. - Values are corrected by use of the effective, carbon-response factors given by Albersheim et a l . 1 2 2 — Isothermal; 170°. - Isothermal; 220°. - I, Original capsular polysaccharide; II, compounds from LiAlH4 reduction of methylated K67; III, compounds from remethylation of the reduced, methylated polymer. - 103 -methylated polysaccharide with lithium aluminum hydride, and analysis of the products revealed a new peak (2-O-methylgIucose) arising from the reduced uronic acid; also found was an increase in the proportion of 4,6-di-O-methyl mannose (see column II). These results indicate that the branch point mannose is linked to glucuronic acid which in turn contains a branch, and that the two terminal sugars are rhamnose and galactose. Remethylation of the lithium aluminum hydride-reduced polysaccharide produced 2,6-di-O-methylglucose at the expense of the monomethyl glucose in confirmation of its origin from glucuronic acid (see column III). ff-Elimination (uronic acid degradation) 1 8 0' 1 8 1 - A sample of methylated K67 polysaccharide was subjected to base catalyzed /3-elimination, and the product was methylated. Analysis by g.l.c. as the partially methylated alditol acetates showed the presence of three principal sugars: 2,4,6-tri-O-methylmannose, 2,4,6-tri-O-methylglucose, and 2,4-di-O-methylrhamnose in the ratios of 1.9:1.0:1.0. Small amounts of a tri-O-methylrhamnose and a tetra-O-methylgalactose were also observed. This result indicates that the uronic acid is in the side chain, and that both a (terminal) rhamnosyl and a (terminal) galactosyl unit are attached to the uronic acid. It also proves that the uronic acid is linked to 0-2 of the branched mannose residue. Partial hydrolysis - Partial hydrolysis of the native polysaccharide with acid was followed by separation of the acidic and neutral fractions by using ion-exchange chromatography. The neutral fraction contained monosaccharides only. The acidic fraction yielded four oligosaccharides (A 2, A3, A4, and A5), - 104 -Table IV.3: Analysis of the oligosaccharides from partial hydrolysis of Klebsiella K67 polysaccharide Oligosaccharide Sugar analysis As alditol acetates (molar proportions) Methylation analysis As alditol acetates (mola proportions) A 2 Man 13 2,3,4-Glc 1 Glc(GlcA) 1 3,4,6-Man 1 A3 Man 23 -Glc(GlcA) 1 A4 Man 1.5* 2,4,6-Man 1 Glc 1.0 3,4,6-Man 1 2,4,6-Glc 1 2,3,4-Glc 1 A5 Man 1.5* 2,4,6-Man 1 Glc 1.0 3,4,6-Man 1 Rha 0.6 2,4,6-Glc 1 2,4-Rha 0.8 2,3,4-Glc 1 -= Including GlcA as Glc. * Neutral sugars only. - 105 -which were separated by paper chromatography. The n.m.r.-spectral data for these oligosaccharides (see Table IV.1), and methylation analysis (g.l.c.-m.s.) (see Table IV.3) proved the structures to be as follows: A 2 /3-GlcA-(l-»2)-Man A 3 /3-GlcA-(l->2)-a-Man-(l->3)-Man A 4 j3-GlcA-(l->2)-a-Man-(l — 3)-<*-Man-(l-*3)-Glc A 5 0-Glc A-( 1 -»2)- a-Man-( 1 — 3)- a-Man-( 1 -• 3)- /3-Glc-( 1 — 3)-Rha The aldotetraouronic acid (A4) obtained from partial hydrolysis is unique for Klebsiella K67. The closest structure to it /3-GlcA-(l -»2)-a-Man-(1 —>2)-a-Man-(l —»3)-Glc, where the linkage to a mannose differs, is obtained from the polysaccharides from Klebsiella K7 and K28 (see Appendix II). In an attempt to sequence the sugars, the aldopentaouronic acid (A5) was methylated and analyzed by m.s. as the permethylated methyl ester methyl glycoside. The following analysis of the mass spectrum using the method of Kovacik and coworkers 1^ 1' 1^ 2 permitted the assignment of the mass numbers to the relevant fragments (see Fig. IV.1 and Scheme IV.l). Periodate oxidation - The native polymer consumed four mol of periodate per repeating-unit, which is consistent with the concept that only the two (terminal) rhamnosyl and galactosyl groups are oxidized. Hydrolysis under mild conditions (Smith degradation) yielded a polymeric material (PI) whose n.m.r. spectra and analysis demonstrated the loss of one rhamnosyl and one galactosyl unit. Polymer PI was methylated, and the product was divided into three portions, (i) Hydrolysis, and estimation of the resulting neutral sugars, showed 100 8 0 6 0 4 0 2 0 0 1 0 0 8 0 6PI 4 0 2 0 0 a - • • 5 0 9 7 6 9 0 6 . 9 1 9 t~i—r~r~r~|~~r 9 0 0 9 5 0 U S * 1 0 2 0 1000 0 8 0 ~i—i—r—|—r~i—r~r~r'T r~i~r~~|—r _ i—rn~ 1 0 5 0 1 1 0 0 i ~ r r i — | — i — r ~ i — I ~ T ~ r 1 1 5 0 1 1 5 0 6 5 5 , 6 0 9 6 2 3 £ 4 1 I 7 1 1 r - r - J s - T - 1 11 1 1 I I 1 1 T - r L T - r 6 0 0 6 5 0 7 0 0 7 5 0 800 8 2 7 , 8 4 6 , 8 6 0 "T—rL"j—1^—F" r "T"r~r- • 8 5 0 1 0 0 8 0 6 0 4 0 2 0 1 0 0 8 0 6 0 4 0 2 0 0 O O N 3 4 9 T ~ r n - - T - T - r - r ~ H p ' l ~ ' i ' - - T - ' ? 6 i 9 ' i i I'V" 3 5 0 4 0 0 4 0 5 " n — r ^ r 4 ~ i ' i • i f — i — f — i — T— i — r - 1— i i ' i — i — m — r , 4 5 1 4 5 0 5 0 0 " p r r r r r r r 5 5 0 6 0 0 1 0 1 , 4 5 7 1 8 8 1 6 9 i i 12 1 9 r—r Jli.i U , 2 3 5 *~\ I T I | T ~ T ~ T ~ f ~ r T T • t ~ \ - V ~ T - ' r +* - r J T"T" 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 Fig IV.1 Mass spectrum of the methylated oligosaccharide A5 from Klebsiella K67 polysaccharide - 107 -m/z 991 I O—bcde -CO c»o )—nO-bcde MeO>—f OMe CHzCH-CH OCH3 O C H 3 m/z 10) m/z 990 eabcdCi CH3 abcd-0 OMe m/z 959 -OMe abed—0 -HCOOMe MeO OMe abcd_o OMe c 437!613 6414Q9 8 4 5b05 ,019^-COOMe J V \ J 1 O1 \ I , 1 x — X 1801 ' '597 1393 1189 !l5 a 2491 b 453i c 657' d 861! e 1035 A a A i COOMe i 1 r baA] cabAi dabcA] eabcdA) to »bc<J -O m/z 233 aA2 COOMe 1 + m/z 437 m/z 64) m/z 845 I I I b a A 2 c a b A 2 d a b c A 2 -x rx m/z 1019 I eabcdA^ m/z 201 a A 3 COOMe 2 + m/z 169 x> s ix J m/z 405 m/z 609 m/z 813 I I i b a A 3 c a b A 3 d a b c A 3 a-O ab-o ate-m/z 373 m/z 577 m/z 781 m/z 967 eabidAj CH3 P abcd-0 m/z 955 H m/z 101 OCH3 0 C H 3 6j m/z 88 . H' CH—CH 1 II 0 C H 3 OCH3 CH CH !! 'I 0 OCHo m/z 73 X= OMe Scheme IV.l Fragmentation pattern of the oligosaccharide A5 from Klebsiella K67 polysaccharide - 108 -the absence of terminal units of rhamnose and galactose, in comparison to the native polysaccharide. The remaining sugars were unchanged, and present in approximately equimolar proportions, (ii) Reduction of methylated PI before hydrolysis gave an additional peak, identified as that of 2,3,4-tri-O-methylglucose, and (iii) reduction and remethylation gave 2,3,4,6-tetra-O-methylglucose (on columns C and D). These results indicated that both of the terminal sugars are linked to glucuronic acid, and they enabled a partial structure to be drawn, as follows: /3-Gal-(l-»3 or 4)-GlcA 4 (or 3) T l a-Rha Two attempts to oxidize either the galactosyl or the rhamnosyl terminal group selectively ^ 7 were unsuccessful. In the first experiment, no oxidation was observed at the end of 20 min, and, in the second, both sugars had been decomposed at the end of 30 min (with no further change up to 120 min). Selective hydrolysis with acid - Treatment of the polysaccharide with 0.1M trifluoroacetic acid (TFA) for 25 min at 95°, and dialysis against distilled water, afforded a nondialyzable, polymeric material and a dialyzate. The dialyzable material contained only rhamnose. The nondialyzable material had an optical rotation of [«]£> -6.1°C (c, 0.328, water) which compares well with the value of [a]j) -34.2°C calculated using Hudson's rules of isorptation on the basis of 50% removal of the terminal rhamnosyl unit. The terminal rhamnosyl unit must be a linked in order to obtain good agreement with the observed optical - 109 -rotation, and is therefore consistent with the result from the chromium trioxide oxidation. Analysis of the polymeric material showed that 60% of the terminal rhamnosyl group had been removed, and methylation analysis yielded 2,4-di-O-methylglucose (from GlcA) or, after remethylation, 2,4,6-tri-O-methylglucose. These results established 0-4 as the position of linkage of the rhamnose. N.m.r. data for the polymer (SH) are shown in Table IV.1. IV.1.4 Conclusions The structure of the capsular polysaccharide from Klebsiella serotype K67 is thus based on the heptasaccharide repeating-unit shown. It is consistent with the qualitative analysis reported in the chemotyping by Nimmich2^ (see Appendix II). The structure is unique in this series, but bears a superficial resemblance to the pattern of K46, where a 1-carboxyethylidene group is attached to a lateral, but non-terminal, mannosyl residue.57 IV.1.5 Experimental General methods - The instrumentation used for n.m.r., g.l.c, g.l.c.-m.s., m.s., infrared, c.d., and measurements of optical rotation has been described in Section III. Paper chromatography and ion-exchange chromatography were performed as described in Section III. -110-Preparation and properties of K67 capsular polysaccharide - A culture of Klebsiella K67, obtained from Dr. Ida 0rskov, Copenhagen, was grown as previously described,5^557 and the polysaccharide was purified by one precipitation with Cetavlon. The isolated polysaccharide (4.7 g) had [«]£>25 -17.3° (c 0.196, water). The average molecular weight was determined by gel chromatography (courtesy of Dr. S.C. Churms, University of Cape Town, South Africa) to be 1.8 x 10^ daltons. N.m.r. spectroscopy (^ H and ^ C) was performed on the original K67 polysaccharide, but an improved p.m.r. spectrum was obtained at high, rather than at ambient temperature. The principal signals and their assignments, for both the *H- and ^ C-n.m.r. spectra, are recorded in Table IV.1. Hydrolysis of the polysaccharide - Hydrolysis of a sample (10 mg) of K67 polysaccharide with 2M trif luoroacetic acid (TFA) for 18 h at 95°, removal of the acid by successive evaporations with water, followed by paper chromatography (solvents B and C, see Section III), showed mannose, galactose, glucose, glucuronic acid, and rhamnose. Neutral sugars were quantitatively determined by g.l.c. as their alditol acetates or PAAN derivatives. The uronic acid was reduced by refluxing overnight a sample (20 mg) of K67 polysaccharide with 3% HC1 in methanol (10 mL), neutralizing the HC1 with PbC03, removing PbCl2, treating the dried product with NaBH4 (50 mg) in anhydrous methanol (10 mL), and stirring overnight. The excess of NaBH4 was neutralized with Amberlite IR-120 (H +) resin, and the boric acid, as methyl borate, was evaporated with methanol. The sample was then hydrolyzed, the alditol acetates were prepared, and those from rhamnose, mannose, galactose, and glucose were identified (in both cases) by g.l.c. (column A, see Section III). Preparative g.l.c. (column a, Section III), followed by measurements of the circular dichroism spectra, showed98 the mannitol - I l l -hexaacetate and glucitol hexaacetate to be of the D configuration, and the rhamnitol pentaacetate to be of the L configuration. Methylation analysis - The polysaccharide (235 mg), converted into the free acid form by passing the sodium salt through a column of Amberlite IR-120 (H +) resin, was dissolved in dry dimethyl sulfoxide (25 mL) and methylated 1 1 5 by treatment with dimethylsulfinyl anion (9 mL) for 4 h, followed by methyl iodide (4 mL) for 1 h. The product (180 mg), recovered after dialysis against tap water, showed complete methylation (no hydroxyl absorption in the i.r. spectrum). A portion of this product (15 mg) was hydrolyzed with 2M trifluoroacetic acid, the sugars were reduced with sodium borohydride, and the alditols acetyla-ted with 1:1 acetic anhydride-pyridine, and analyzed by g.l.c. in columns B, C, and D (see Table IV.2, column I). G.l.c.-m.s. was conducted with column D. 1,5-Di-O-acetyl-2,3,4,6-tetra-Q-methylgalactitol was isolated by preparative g.l.c. (column b), and found to give a positive c.d. curve, indicating that the galactose had the D configuration.98 Another portion of the fully methylated polysaccharide (30 mg) was subjected to carboxyl reduction with lithium aluminum hydride in anhydrous oxolane. Half of the product was hydrolyzed with 2M TFA, and converted into the alditol acetates, and these were analyzed by g.l.c. in columns B, C and D (see Table IV.2, column II); column D was used for g.l.c.-m.s. The other half was remethylated by the Hakomori method,115 and the product converted into the alditol acetates as before. G.l.c. analysis was conducted in columns B, C and D (see Table IV.2, column III), and g.l.c.-m.s., in column D. - 112 -Partial hydrolysis - A solution of the K67 polysaccharide (997 mg) in 0.5M trifluoroacetic acid (50 mL) was heated for 3 h on a steam bath. After removal of the acid by successive evaporations with water, the acidic and neutral fractions were separated on a column of Bio-Rad AG1-X2 ion-exchange resin. The acidic fraction (434 mg) was separated by preparative chromatography, using solvent B. The fractions A3, A4, and A5 were repurified by the same procedure, but solvent A was used, in order to give better separation and to remove contaminants present. The subsequent yields of pure fractions were 26 mg of aldobiouronic acid (A 2), 45 mg of aldotriouronic acid (A3), 63 mg of aldotetraouronic acid (A4), and 80 mg of an aldopentaourinic acid (A5). Paper chromatography of the neutral fraction showed only galactose, glucose, mannose, and rhamnose, and there was no indication of the presence of oligosaccharides. This material was not examined further. The analyses performed on the oligosaccharides were as follows: (a) Sugar analysis, (i) Acidic oligosaccharides (A 2 and A3) were treated with 3% HC1 in anhydrous methanol for 8 h on a steam bath. The methyl ester obtained was reduced with sodium borohydride in anhydrous methanol, followed by hydrolysis of the product with 2M TFA, reduction to the alditols, and acetylation with 1:1 acetic anhydride-pyridine. The alditol acetates obtained were analyzed by g.l.c. in column A. (ii) Acidic oligosaccharides (A4 and A5) were hydrolyzed with 2M TFA, the products reduced, the alditols acetylated, and the alditol acetates analyzed by g.l.c. in column A. (b) Methylation analysis: All of the methylations were conducted by the method of Hakomori. 1^ Portions of the methylated oligosaccharides were hydrolyzed with 2M TFA, the products converted into partially methylated alditol acetates, and these analyzed by g.l.c. and g.l.c.-m.s. in column D. The remainder - 113 -was reduced with LiAlH4 in anhydrous oxolane, hydrolyzed with 2M TFA, the products converted into alditol acetates, and these analyzed by g.l.c. and g.l.c.-m.s. in column D. A portion of the methylated oligosaccharide (A5) was purified on a column of Sephadex LH-20 with MeOH-CHCl3 (1:9) as eluant and subjected to mass-spectral analysis (see Fig. IV.l, and Scheme IV.l). The results obtained for each oligosaccharide are given in Table IV.3, and the n.m.r. data in Table IV.l. Periodate oxidation - A solution of K67 polysaccharide (21.9 mg) in water (5.0 mL) was mixed with 0.03M NaI04 (5.0 mL), and stirred in the dark at room temperature (23°). The periodate consumption was monitored spectrophotometri-cally. 2 2 9 After 95 h (consumption of periodate stabilized at 3.7 mol per mol of K67), ethylene glycol (0.2 mL) was added. The polyaldehyde was dialyzed over-night, reduced to the polyalcohol with NaBH4 (200 mg), and the excess hydride was neutralized with 50% acetic acid, and the solution dialyzed, and freeze-dried. A fraction of the product (2 mg) was analyzed for the sugars present by treatment with 2M TFA overnight on a steam bath, and conversion into alditol acetates. Analysis by g.l.c. in column A showed rhamnitol, mannitol, and glucitol in the ratios of 1:2:1. There was no galactose present. The rest of the material was treated with 0.5M TFA for 24 h at room temperature, and the acid was removed by repeated addition and evaporation of water. The n.m.r. data for this product (PI) are given in Table IV.l. Permethylated PI was (i) analyzed for neutral sugars; (ii) reduced with LiAlH4, and analyzed; and (iii) reduced, remethylated, and analyzed (columns C and D). - 114 -Selective periodate oxidation1**? . Two series of experiments were performed. In the first, K67 polysaccharide (216 mg) was oxidized with 0.02M NaIC>4 (20 mL), and aliquots (5 mL) were withdrawn at 5, 10, 15, and 20 min. In the second, polysaccharide (263 mg) was oxidized with 0.02M NaI04 (20 mL), and aliquots were withdrawn at 30, 60, 90, and 120 min. Each aliquot was added to ethylene glycol (0.4 mL), and analyzed as the alditol acetates, following dialysis, borohydride reduction, and hydrolysis. Uronic acid degradation 1 8 0' 1 5 1 - A sample (40 mg) of methylated K67 polysaccharide was dried, and dissolved in 19:1 dimethyl sulfoxide-2,2-dimethoxy-propane (15 mL), together with a trace of p_-toluenesulfonic acid. The flask was sealed under N 2, and to its contents was added dimethylsulfinyl anion (8 mL); the solution was kept overnight at room temperature, cooled, and methyl iodide (6 mL) was added. The excess of base was neutralized with 50% acetic acid, and, after dialyzing for 2 d, the methylated, degraded product was isolated as the nondialyzable fraction. G.l.c. analysis as the alditol acetates, after hydrolysis, was conducted in column D. Chromic acid oxidation of the fully acetylated polysaccharide 2 2 1' 2 2 2 - To a solution of the original K67 polysaccharide (10 mg) in formamide (5.0 mL) were added acetic anhydride (1.5 mL) and pyridine (1.5 mL), and the mixture was stirred for 20 h. Following dialysis for 2 d, the nondialyzable material was freeze-dried, and the product dissolved in acetic anhydride (2.0 mL); CrC>3 (150 mg) was added, the solution was stirred for 2 h at 50°, and the product was isolated by partition between CHCI3 and water. Hydrolysis and analysis (as -115-alditol acetates) by g.l.c. in column A gave rhamnose and mannose in the ratio of 1:1. Glucose and galactose were both absent. IV.2. Structural Investigation of Klebsiella serotype K80 capsular polysaccharide. IV.2.1 Abstract The use of methylation, periodate oxidation, /3 -elimination and selective hydrolysis enabled the structure of the K80 polysaccharide to be determined. The nature of the anomeric linkages was established using n.m.r. spectroscopy and further confirmed by the results of oxidation of the fully acetylated polysaccha-ride with chromic acid. The K80 polysaccharide is comprised of repeating-units of the pentasaccharide shown and contains a pyruvic acid acetal on each repeating-unit. This pattern represents the first instance, in this series of polysaccharides, of a pyruvate acetal attached to a side chain rhamnose residue. 3)- 0-D-Gal-( 1 -»2)- a-D-Man-( 1 ->2)- a-D-Man-( 1 -» 3 T l a-D-GlcA 4 T l /3-L-Rha 4 3 \ / C / \ CH 3 C 0 2 H - 116 -IV.2.2 Introduction Serotype K80, according to the chemotyping by Nimmich,2^ is one of three strains of Klebsiella whose capsular polysaccharides are composed of D-glucuronic acid, D-galactose, D-mannose and L-rhamnose. The other two strains, Klebsiella K40 (ref. 26) and K53 (ref. 230), do not contain a 1-carboxyethylidene substituent. The only other known Klebsiella polysaccharides containing a 1-carboxyethylidene substituent on a rhamnose residue are Klebsiella K32 (ref. 231), K70 (ref. 232) and K72 (ref. 233). The pentasaccharide repeating-unit of Klebsiella K80 was found to be of the "three-plus-two" type where three sugars are in the main chain with two sugars in a side chain. IV.2.3 Results and discussion Composition and n.m.r. spectra - Isolation and purification of the polysaccharide were achieved as previously described.5^>57 The purified product obtained after one Cetavlon precipitation was shown by gel-permeation chromato-graphy to be homogeneous ( M w = 0.9 x 10^ ). The H^-n.m.r. spectrum (see Appendix III, spectrum, No. 13) indicated the presence of five anomeric protons at 8 5.33 (1H), 5 5.23 (1H), 8 5.17 (1H), 8 4.85, 4.80 (0.4H, 0.6H), and 8 4.53 (1H). The three former signals correspond to a linkages, while the latter two protons arise from /3 linked sugars. One )3 signal showed twinning, which was also reflected in the doubling of the signal due to the protons on C-6 of the rhamnose unit [ 8 1.33 (1.2H), 8 1.27 (1.8H)], and was shown to have been caused by the removal of -30% of the 1-carboxyethylidene - 117 -Table IV.4 N.M.R. data for Klebsiella K80 polysaccharide and the derived oligosaccharides Compound3- 'H-N,m,r. data ^C-N.m.r. data  A^ S i j 2 Inteeral Assignment^ P.p.m,& Assignment^ (#11 protons Gal-L-2.ManL2Gly P3R 5.21 s 4.46 8 -Man— G a l -103.1 98.4 G a l -^ B 2-Man— a [iGalUManJ-^Man-Li a GlcA 4 ? 1 Rha K80D [^Gali-iMan-LiiMani-],, 3 3 a a J n 1 GlcA 4 6 1 Rha 4 3 \ / pyr K80 5.35 s 5.23 s 5.18 s 4.80 s 4.53 8 1.26 7 5.33 br 5.23 br 5.17 br 4.85 4.80 4.53 br 1.57 s 1.33 d 1 ) 0.4 ' 0.6 ' 1 1.2 0.4 ^GIcA-2-Man-3 I -Man— Rha-p-i G a l ^ C H 3 of Rha ^ G l c A --Man— 3 a I 2-Man— a Rha— ,Rha— 4 3 6 6 \ / pyr SGalg C H 3 of pyr C H 3 of Rha 4 3 \ / pyr 102,9 102.0 100.8 95.8 17.3 102.9 102.2 101.9 100.9 100.6 96.0 30.8 17.4 3-Gal-B R h a -4-GlcA—,-Man-a 3 a *Man— a C H 3 of Rha 3-Gal-6 Rha—, Rha-6 4 3 \ / pyr ^GlcA-•^Man— 3 a I 2-Man— a C H 3 of pyr C H 3 of Rha. C H 3 of Rha 4 3 \ / pyr 1.27 0.6 C H 3 of Rha - 118 -Table IV.4 .. continued Compound3- iH-N.m.r. data 2] 2 Integral Assignment^ CHz) protons 1 3C-N.m.r. data P.p.m,— Assignment-[2GalI-^ManI^Mani-GlcA 4 1 Rha 50% K80SH 5.35 s 1 5.21 s 1 5.18 s 1 4.80 s 0.2 4.53 br 1.27 br 1 0.9 * G l c A -a 2-Man— 3 a 2-Man— a R h a -102.9 102.1 100.9 100.7 S-Gal- 95.8 B C H 3 of Rha 17.3 2 G a l F R h a -^GlcA-a -Man— 3 a •^Man— C H 3 of Rha 3- For the sources of P2, K80D, and K80SH, see text. & Chemical shift relative to internal acetone; 2.23 downfield from sodium 4,4-dimethyl-4-silapentane-1-sulfonate (DSS). °- The numerical prefix indicates the position at which the sugar is substituted; a or 0, the configuration of the glycosidic bond, or the anomer of a (terminal) reducing, sugar residue. Thus -2-Man— refers to the anomeric proton of a 2-linked mannosyl residue in the o-anomeric configuration. The absence of a numerical prefix indicates a (terminal) nonreducing group. — Chemical shift in p.p.m. downfield from Me4Si, relative to internal acetone; 31.07 p.p.m. downfield from DSS. £ As in £ but for I 3 C nuclei. - 119 -groups during the harvesting of the native polysaccharide. Both the anomeric and the high field signals collapsed to single resonances [ 6 4.80 (IH), 5 1.26 (3H)] when the native polymer was treated with mild acid, which removed all of the 1-carboxyethylidene acetal groups (see spectrum No. 15, Appendix III). This indicated that the rhamnose residue carries the acetal substituent. The C^-n.m.r. spectrum of the native polysaccharide (see Appendix III, spectrum No. 14) also exhibited the same type of twinning of the anomeric carbon atom of rhamnose (102.2 p.p.m. and 101.9 p.p.m.), which was reduced to one signal at 102.0 p.p.m. in the depyruvalated polymer (see Appendix III, spectrum No. 16). Paper chromatography of an acid hydrolyzate of the polysaccharide showed galactose, glucuronic acid, mannose and rhamnose. Determination of the neutral sugars as the alditol acetates showed rhamnose, mannose and galactose in the ratios of 1.00:1.45:1.11. The carboxyl-reduced polysaccharide89 gave rhamnose, mannose, galactose and glucose in the ratios of 0.90:1.80:1.11:1.00 indicating that the uronic acid is glucuronic acid. By circular dichroism measurements98 made on the alditol acetates, the glucose, (and hence, glucuronic acid) and mannose were proved to be of the D configuration, and rhamnose of the L configuration. Galactose was assigned the D configuration from circular dichroism measurements made on the partially methylated alditol acetate derivative. Methylation analysis - Methylation of the K80 polysaccharide, followed by hydrolysis, conversion of the neutral sugars into alditol acetates, and g.l.c.-m.s. analysis thereof gave the values shown in Table IV.5 column I. These results indicate that the polysaccharide consists of a pentasaccharide repeating-unit with a branch on mannose and a terminal rhamnose unit. Part of the 1-carboxy-ethylidene groups had been removed under the conditions of methylation - 120 -Table IV.5: Methylation analyses of K80 polysaccharide and derived products T^ Mole % c Methylated sugarS (as alditol acetates) Column C d Column Column Ii II III IV (OV-225) (ECNSS-M) (SP1000) 2,3,4-Rha 0.50 0.47 0.54 21.15 14.84 13.92 9.40 2-Rha 1.38 1.53 1.37 6.34 4.56 4.18 16.97 3,4,6-Man 1.80 1.91 1.60 30.21 22.83 23.20 22.93 2,4,6-Gal 1.98 2.20 1.87 27.19 22.37 22.04 18.58 2,3,6-Glc 2.24 2.37 1.87 - - 13.92 15.37 4,6-Man 2.85 3.18 2.54 15.10 21.69 22.73 16.74 2,3-Glc 4.26 5.75 3.33 - 13.70 ^ 2,3,4-Rha = l,5-di-0-acetyl-2,3,4-tri-0-methylrhamnitol, etc. ^ Relative retention-time, referred to 2,3,4,6-Glc as 1.00. c Values are corrected by use of the effective, carbon-response factors given by Albersheim et a l . 1 2 2 — Isothermal; 180°; £ Isothermal 175°; ^ Isothermal 220°; & I, Original capsular polysaccharide; II, compounds from LiAlH4 reduction of methylated K80; III, compounds from remethylation of the reduced, methylated polymer; IV, compounds from carboxyl-reduced, methylated polysaccharide. - 121 -(exchanging with IR-120 (H +) resin), but from those that survived it could be deduced that this group is linked to 0-3 and 0-4 of the terminal rhamnosyl group. On reduction of the methylated polysaccharide with LiAlH4 (see Table IV.5, column II), the proportion of l,2,3,5-tetra-0-acetyl-4,6-di-0-methylmannitol was increased, and l,4,5,6-tetra-0-acetyl-2,3-di-0-methylglucitol was formed, demonstrating that glucuronic acid is linked at 0-4 and that it is joined to the branch point mannose. Remethylation of the lithium aluminum hydride-reduced polymer, and analysis of the product, gave the values shown in Table IV.5, column III. These results imply the presence of one branch point on the D-mannosyl residue and a 1-carboxyethylidene group linked to the terminal rhamnosyl group. Methylation of the carboxyl-reduced polysaccharide (see Table IV.5, column IV) showed an increase in the proportion of 2-O-methylrhamnose, as this sample was not deionized by treatment with a cation exchange resin. This observation confirmed the presence of a 3,4 linked 1-carboxyethylidene group on the terminal rhamnosyl group. The appearance of l,4,5-tri-0-acetyl-2,3,6-tri-0-methylglucitol provided further proof of the presence of a 4-linked glucuronic acid residue. Base-catalvzed alduronic acid degradation - A sample of the methylated K80 polysaccharide was treated overnight with dimethylsulfinyl anion. The product was treated with methyl iodide and extracted from water with chloro-form. Analysis as the partially methylated alditol acetate derivatives showed the presence of l,2,5-tri-0-acetyl-3,4,6-tri-0-methylmannitol and 1,3,5-tri-O-acetyl-2,4,6-tri-O-methylgalactitol in the ratio 2.19:1.00. Small proportions of terminal rhamnose and the branch point di-Q-methylmannose were also obtained. This proved that the uronic acid is linked at 0-3 of the branch point mannose. - 122 -Oxidation of the carboxvl-reducedlu:> polysaccharide with chromium  t r i o x i d e 2 2 1 ' 2 2 2 - The carboxyl-reduced, fully acetylated polysaccharide was subjected to oxidation with chromic acid. The product was analyzed for the presence of neutral sugars (as their alditol acetates). Galactose was completely degraded, but only a part of the rhamnosyl residues was oxidized. The ratios of rhamnose, mannose and glucose obtained were 0.42:2.00:0.71. The survival of the glucosyl and both mannosyl residues is proof of their a linkages. The galactose is /3-linked, and so undergoes oxidation. Rhamnose was assigned a /3 linkage as part of the sugar was oxidized and the anomeric signal in the 'H-n.m.r. spectrum had a low value ( 8 4.80). On methylation, and analysis as the partially methylated alditol acetates, 2,3,4-tri-O-methylrhamnose, 2,3,4,6-tetra-O-methylglucose, 3,4,6-tri-O-methylmannose, 2,4,6-tri-O-methylmannose, and 2,3,6-tri-O-methylglucose were found in the ratios of 0.25:0.80:1.00: 1.05:0.50. The methylated glucose units arise from the reduced alduronic acid substituted at 0-4 by rhamnose, and from that which had part of the rhamnose degraded, thus proving that rhamnose is linked to 0-4 of glucuronic acid. The 2,4,6-tri-O-methylmannose is obtained from the original branch-point mannose on degradation of the /3-linked galactose, thus showing that galactose is linked to 0-2 of the mannose residue. The results obtained thus far lead to the partial structure and showed in addition, the presence of a 2-linked a-mannosyl residue. - 123 -3)- /3-D-Gal-(l —2)-a-D-Man-(l-3 T l a-D-GlcA 4 T l 0-L-Rha 4 3 \ / C / \ CH 3 CQ 2H Periodate oxidation - Periodate oxidation of the native polymer and analysis, as the peracetylated alditols, of the sugars after hydrolysis showed the presence of glycerol, erythritol, mannose, galactose and glucose in the ratios of 0.44:0.71:1.00:1.00:0.25. These results indicated that the 1-carboxyethylidene groups do not survive under the reaction conditions employed, and that one mannosyl, the rhamnosyl, and part of the glucosyluronic acid residues were oxidized. As expected, the 3-linked galactosyl and the branched mannosyl residues survived oxidation. Smith degradation of the polyol revealed the presence of six compounds when chromatographed (see Fig. 1V.2). These compounds, PI - P6 were isolated by preparative paper chromatography using solvent A. The compounds were reduced with NaBH4 to P1R - P6R. Analysis by g.l.c. as the alditol acetates showed P1R and P2R to be the same, and P3R, P4R, P5R and P6R to be similar (see Table IV.6). Methylation analysis of P3R and P4R indicated the presence of 2,3,4,6-tetra-O-mcthylgalactosc, and 3,4,6-tri-O-methylmannose in the ratio of 1.0:1.0. The compounds P3 - P6 are thus identical, with PI and P2 consisting mainly of erythritol. The different chromatographic mobilities of P3 - P6 were - 124 -0 % 5 0 Gal 0 G l c A fc>4^n 0 Rha •P2 •PI tF6 rj-0 JP5 GlcA Gal P4 o >an P3 a) Solvent A (42 h) • b) Solvent C (24 h) Fig IV.2 Chromatogram of the products isolated from the Smith degradation of the K80 polysaccharide Table IV.6 Sugar analysis of the Smith degradation products P1R-P6R Alditol acetates -j-a Mole Ratio P1R P2R P3R P4R P5R P6R Glycerol 0.07 0.25 0.20 0.17 2.9 trace 0.2 Erythritol 0.19 1.00 1.00 trace 0.6 trace trace Rhamnose 0.31 trace trace trace - trace -Mannose 0.88 - - 1.0 1.0 1.0 1.0 Galactose 0.94 - - 1.2 1.3 1.2 1.3 a Retention times relative to glucitol hexaacetate as 1.00, on a column of SP-2340 programmed at 195° for 8 min and then 4°/min to 260°. - 125 -attributed to the formation of glycolaldehyde acetals 1 7 4' 1 7^ (see Scheme IV.2). The p.m.r. spectra of PI - P6 and P1R - P6R indicate that PI, P2, P1R, and P2R are identical, and P3 - P6 are similar with no significant changes ' P o l y o l ' + OH Scheme IV.2 The glycolaldehydes formed from the product obtained by Smith degradation of the K80 polysaccharide in the spectra of P3R - P6R. The signal at 6 5.21 (IH) in the spectra of P3R - P6R was assigned to the a mannose residue, while the signal at 5 4.46 (J\2 8 Hz, IH) was assigned to (3 galactose. The ^ C-n.m.r. spectrum of P3R also showed one 13 signal (103.1 p.p.m.) and one a signal (98.4 p.p.m.). Thus the general structure given to these fragments (P3R - P6R) is shown by P3R. 1 0 0 8 0 6 0 4 0 2 0 0 1 0 0 8 0 6 0 4 0 2 0 0 1 0 0 8 0 6 0 4 0 2 0 0 —I I" 1 I t 4 3 9 4 5 3 i I, u < 6 8 _ J _ 7 . r r T T T T T - n ~ p ~ [ - r r m i i | i i i i i i ~ r r i | i i i i r n i i | i i r T T - r r n - p r r i i I i i i i | i I I i ' f r r 1 ! i | i i i r r r r i i | i i . i i M i i i | 4 4 0 4 6 0 4 8 0 5 0 0 5 2 0 5 4 0 5 6 0 6 8 0 6 0 0 6 2 0 * 2 0 , 2 7 5 T 2 4 0 WiTrn * rp ^ -ff n -f h 2 6 0 7 1 3 6 7 3 4 7 -rYn 2 8 0 3 0 0 3 2 0 3 4 0 3 6 0 , 3 9 1 1 4 1 1 1 4 2 3 I'l I I I 1^^~'~'~ri'1'T JT-[ ' I I I ' l l ' r i I p i I I 5 7 1 0 1 8 8 7 5 6 0 If (V].L|i|ltl 8 0 iVl't't' I'lVt 1 1 1 1 1 2 7 I + Lt I »ii ^  • '••'T'T JT"rLrtLt 1 0 0 WiYtVm'h^rYtVt1 2 0 1 4 0 1 4 9 3 8 0 1 8 7 4 2 0 'n't't'p 1 6 7 1 6 0 T~l'I I ' I ' I 180 , 2 1 9 V ^ r r r V r - i V r r i n i ' l ' i i i i n ' r r r ON 2 0 0 220 240 Fig IV.3 Mass spectrum of the methylated oligosaccharide alditol P3R from Klebsiella K80 polysaccharide - 127 -m/z 305 m/z 155 m/z 359 Scheme IV.3 Fragmentation pattern of the oligosaccharide alditol P3R from Klebsiella K80 polysaccharide - 128 -CH 2OH I /3-Gal-(l -* 2)-a-Man-(l — OCH I CH 2OH P3R This fragment can only arise if the original 2-linked mannose was linked at 0-2 to the branch point mannose, and was itself linked at 0-3 to galactose. It is therefore in accord with the partial structure previously predicted. The mass spectrum of the per-O-methylated oligosaccharide P3R provided further proof of its structure (see Fig. IV.3 Scheme IV.3). Because ions of the B-fragmentation series are obtained only for (1—*-2) or (1—*-4) linked oligosaccharides,151 it is evident that the branch point mannosyl residue is linked at 0-2 (not 0-3) to the galactosyl residue. Selective partial hydrolysis - Treatment of the K80 polysaccharide with very dilute acid, followed by dialysis of the products against distilled water, afforded a nondialyzable, polymeric material and a dialyzate. Paper chromato-graphy, and g.l.c. analysis (as the alditol acetates) showed that the dialyzate contained rhamnose. The ^ -n.m.r. spectrum of the nondialyzable material (K80SH) showed the depression of one signal (5 4.80), and the ^ C-n.m.r. spectrum, of the signal at 102.0 p.p.m. Only 50% of the rhamnose had been removed as was indicated both by ^ -n.m.r. spectroscopy and sugar analysis. The per-O-acetylated alditols from rhamnose, mannose and galactose were in the ratios of 0.50:1.50:1.15. Carboxyl reduction 1 0 5 of the nondialyzable material and analysis as the per-O-acetylated alditols, gave rhamnose, mannose, galactose and glucose in - 129 -Table IV.7: Methylation analyses of selectively hydrolyzed K80 polysaccharide Methylated sugars3 ~pb_ Mole Column C-d-(OV-225) Column (ECNSS-M) if II 2,3,4-Rha 0.55 0.46 24.05 3.07 3,4,6-Man 1.47 1.92 25.31 27.93 2,3,6-Gal 1.56 2.21 25.31 24.58 2,3,4-Glc 1.64 - - 14.52 4,6-Man 1.91 3.19 25.31 24.86 2,3-Glc 2.22 - - 5.02 a,b,£ As in Table IV.5. ^ Programmed for 4 min at 180° and then at 2°/min to 260°. £ Isothermal; 170°. ^ I, Selectively hydrolyzed K80 polysaccharide (K80SH); II, compounds from L i A l H 4 reduction of the methylated K80SH. - 130 -the ratio 0.22:2.07:1.00:0.77. Methylation analysis of the nondialyzable fraction as the partially methylated alditol acetates, and analysis after reduction of the permethylated polysaccharide by LiAlH4 (Table IV.7 columns I and II) showed that the terminal rhamnosyl group is linked to position 0-4 of the glucosyluronic acid residue. IV.2.4 Conclusion These experiments demonstrate that the structure of the capsular polysaccharide of Klebsiella K80 is as shown. This pattern is novel, in that the 1-carboxyethylidene group is attached to 0-3 and 0-4 of a terminal L-rhamno-pyranosyl group whereas, in Klebsiella serotypes K32 (ref. 231), K70 (ref. 232) and K72 (ref. 233) the 0-(l-carboxyethylidene)-L-rhamnose unit is a constituent of the main chain. The similarity of the structure of K74 polysaccharide2-^ to that of K80 is noteworthy. [ —3)- /3-Gal-( 1 — 2)-a-Man-( 1 -»2)- a-Man-( 1 — ] 3 l T l a-GlcA 4 T l |3-Gal 4 6 \ / C / \ CH 3 CQ 2H K74 - 131 -[ —3)- /3-Gal-( 1 — 2)- a-Man-( 1 — 2)- o-Man-( 1 — ] 3 n T l a-GlcA 4 T l /3-Rha 4 3 \ / C / \ CH 3 CQ 2H K80 IV.2.5 Experimental General methods - The instrumentation used for n.m.r., g.l.c., g.l.c.-m.s., m.s., i.r., c.d., and measurement of optical rotation has been described in Section III. Paper chromatography was performed according to the description in Section III. Preparation and properties of K80 capsular polysaccharide - A culture of Klebsiella K80, obtained from Dr. Ida 0rskov, Copenhagen, was grown as previously described,-^57 and the polysaccharide was purified by one precipitation with Cetavlon. The isolated polysaccharide (1.5 g) was shown by gel chromatography on a Sephadex 4B column (courtesy of Dr. S.C. Churms, University of Cape Town, South Africa) to have an average molecular weight of 0.9 x 10^ . N.m.r. spectroscopy (*H and ^ C) was performed on the original K80 polysaccharide. All ^H-n.m.r. spectra were recorded at high temperature in order - 132 -to obtain better resolution. The principal signals and their assignments, for both the and C^-n.m.r. spectra, are recorded in Table IV.4. Hydrolysis of the polysaccharide - Hydrolysis of a sample (10 mg) of K80 polysaccharide with 2M trifluoroacetic acid (TFA) for 24 h at 95°, removal of the acid by repeated coevaporation with water, followed by paper chromatography (solvents B and C, see Section III) showed rhamnose, mannose, galactose, glucuronic acid and an aldobiouronic acid. The neutral sugars were quantitatively determined by g.l.c. (column A, see Section III) as their alditol acetates and the presence of rhamnose, mannose and galactose was confirmed. The quantitative, sugar analysis of the carboxyl-reduced polysaccharide 1 0 5 revealed, in addition, the presence of the alditol acetate of glucose. Circular dichroism measurements - Glucitol hexaacetate, mannitol hexaacetate and rhamnitol pentaacetate were separated on column a. A positive curve was obtained for the first two compounds and, hence, the D configuration was assigned to mannose and glucuronic acid; a negative curve for the last-mentioned compound was proof of its L configuration. Column b programmed from 195° at 4°/min to 260° was used to separate l,5-di-0-acetyl-2,3,4,6-tetra-0-methylgalactitol. A positive curve was obtained for its circular dichroism spectrum, and this permitted assignment of the D configuration to galactose. Methylation analysis - The capsular polysaccharide (140 mg) in the free-acid form, obtained by passing the sodium salt through a column of - 133 -Amberlite IR-120 (H +) resin, was dissolved in dry dimethyl sulfoxide (12 mL) and methylated 1 1 5 by treatment with dimethylsulfinyl anion (5 mL) for 4 h, followed by methyl iodide (4 mL) for 1 h. The product (150.3 mg), recovered after dialysis against running tap water, showed complete methylation (no hydroxyl absorption in its i.r. spectrum). A portion (15 mg) of this product was hydrolyzed with 2M TFA, and the sugars reduced with sodium borohydride. The alditols were acetylated with 1:1 acetic anhydride-pyridine, and analyzed by g.l.c. in columns B, C and D [see Table IV.5, column I]. G.l.c.-m.s. was conducted on column B. Another portion (30 mg) of the fully methylated polysaccharide was subjected to carboxylic ester reduction with lithium aluminum hydride in anhydrous oxolane. Half of the product was hydrolyzed with 2M TFA, and converted into the alditol acetates, and these were analyzed by g.l.c. in columns B, C and D [see Table IV.5. column II)]; column B was used for g.l.c.-m.s. The other half was remethylated by one Hakomori methylation, 1 1 5 and the product converted into the alditol acetates as before. G.l.c. analysis was conducted in columns B, C and D (see Table IV.5 column III) and g.l.c.-m.s. in column B. Carboxyl-reduced K80 polysaccharide (22.5 mg), obtained by using the carbodiimide-reduction method,105 was dissolved in dimethyl sulfoxide (3 mL) and methylated by the Hakomori procedure.115 The product showed complete methylation (no hydroxyl absorptions in the i.r. spectrum). The partially methylated alditol acetates were prepared in the usual way, and analyzed by g.l.c. in column B (see Table IV.5 column IV). G.l.c.-m.s. was conducted with column B. All g.l.c: analyses on columns B, C and D were performed isothermally, at 175, 180 and 220° respectively. Uronic acid degradation 1 8 0' 1 5 1 - A sample (20 mg) of methylated K80 - 134 -polysaccharide was dried, and then, together with a trace of fi-toluenesulfonic acid, was dissolved in 19:1 dimethyl sulfoxide-2,2-dimethoxypropane (12 mL), and the flask sealed under nitrogen. Dimethylsulfinyl anion (5 mL) was added, and allowed to react for 18 h at room temperature. The solution was cooled, methyl iodide (3 mL) added, and the mixture stirred for 1 h. The methylated, degraded product was isolated by partition between chloroform and water. The product was hydrolyzed with 2M TFA for 4 h at 95°, and the partially methylated alditol acetates were prepared as described earlier. G.l.c analysis and g.l.c.-m.s. were conducted in Column B at 175° (isothermal). Chromic acid oxidation 2 2 1' 2 2 2 To a solution of the carboxyl-reduced K80 polysaccharide 1 0 5 (20 mg) in formamide (5 mL) were added acetic anhydride (2 mL) and pyridine (2 mL). The mixture was stirred overnight at room temperature, dialyzed for 2 d and freeze-dried. The product was dissolved in acetic acid (5 mL), CrC>3 (150 mg) was added, and the solution was heated for 1 h at 50°. The product was isolated by partition between chloroform and water. A third of the product was hydrolyzed with 2M TFA, and the sugars converted into the alditol acetates, which were then analyzed by g.l.c. in column A. The rest was methylated by the Hakomori procedure,115 and the product converted into the partially methylated alditol acetates. G.l.c. analysis and g.l.c.-m.s. were conducted in column B. Periodate oxidation - To the K80 polysaccharide (42 mg) in water (10.0 mL) was added 0.03M NaI04 (10.0 mL), and the solution was kept for 120 h at room temperature. Following addition of ethylene glycol (0.2 mL) and, after 1 h, - 135 -sodium borohydride (100 mg), the product was isolated by dialysis and lyophilization. G.l.c. analysis was performed in column A. Smith degradation - A solution of K80 polysaccharide (428.1 mg) in water (88.8 mL) was mixed with a solution of 0.2M NaC10 4 in 0.03M NaI0 4 (88.8 mL), and kept in the dark for 120 h at room temperature. Ethylene glycol (0.2 mL) was added, the mixture stirred for 1 h, and the polyaldehyde formed dialyzed overnight, and reduced with NaBH 4 (1 g). The base was neutralized with 50% HOAc, and the solution was dialyzed and freeze-dried. Smith hydrolysis was effected by treating with 0.5M TFA (100 mL) and stirring for 48 h at room temperature. Six compounds were isolated by preparative paper chromatography using solvent A for 30 h. The six compounds PI (60 mg), P2 (29 mg), P3 (37 mg), P4 (10 mg), P5 (29 mg) and P6 (35 mg) were individually reduced with aqueous NaBH 4 to yield P1R, P2R, P3R, P4R, P5R and P6R. These compounds were analyzed as alditol acetates on column A programmed for 8 min at 195°, and then 4°/min to 260°. The results indicated that P1R and P2R comprised mainly of erythritol, while P3R-P6R were disaccharides of galactose and mannose (see Table IV.6). Methylations of P3R and P4R were performed according to the method of Hakomori. 1^ Analyses as the partially methylated alditol acetates were carried out on columns B (at 175°, isothermal) and C (programmed for 4 min at 180° and then 2°/min to 230°). A part of the methylated oligosaccharide (P3R) was purified on a Sephadex LH-20 column, with 1:1 methanol-chloroform as the eluant, and examined by mass spectrometry (see Fig. IV.3, Scheme IV.3). Selective partial hydrolysis - Treatment of K80 polysaccharide (91 mg) - 136 -with 0.1M TFA (25 mL) for 20 min at 95°, and dialysis for 2 d, gave a non-dialyzable material that had all of the sugars intact; only the 1-carboxy-ethylidene group was removed [see Table IV.4, K80D]. Analysis as the alditol acetates by g.l.c. in column A, showed the same molar ratios as for the native polysaccharide. K80 polysaccharide (103 mg) and 0.1M TFA (25 mL) were heated for 90 min at 95°, and dialyzed, after removal of TFA by coevaporation with water. The dialyzate was shown by paper chromatography (solvent B) to contain rhamnose. Analysis by g.l.c. revealed the presence of rhamnose only. The nondialyzable fraction, and the carboxyl-reduced105 nondialyzable fraction, were analyzed as the alditol acetates on column A. The nondialyzable fraction (17 mg) was methylated by one Hakomori methy-lation. 1 1 5 One third was hydrolyzed with 2M TFA, and the sugars converted into the alditol acetates (see Table IV.7 column I). G.l.c. analysis was performed in column B, 170° (isothermal). The remaining two-thirds was reduced with L i A l H 4 in anhydrous oxolane, and the product hydrolyzed with 2M TFA. The compounds in the hydrolyzate were converted into alditol acetates and analyzed by g.l.c. on column C, programmed for 4 min at 180°, and then 2°/min to 260° (see Table IV.7 column II). The n.m.r. data for K80SH are given in Table IV.4. -137-BACTERIOPHAGE DEGRADATION OF THE CAPSULAR POLYSACCHARIDES FROM Klebsiella SEROTYPES K26 AND K44 - 138 -V. BACTERIOPHAGE DEGRADATION OF THE CAPSULAR POLYSACCHARIDES FROM Klebsiella SEROTYPES K26 AND K44 Y.l. Introduction In the late nineteenth century, scientists discovered that certain diseases in plants and animals were caused by a "filterable substance" which Pasteur termed a "virus". 2 3 5 Twort (in 1915) 2 3 6 and D'Herelle (in 1917) 2 3 7 independently discovered that such viruses were capable of infecting and destroying bacteria. These viruses were termed bacteriophages or phages (</>) and are designated by the number of its host strain. Bacteriophages are different from bacteriocins which are protein substances produced by one strain of bacteria and are lethal to another strain of the same species. Thus bacteriocins function in a manner similar to antibiotics, the major variation being their narrow spectrum of activity in contrast to the diversity of antibiotic functions. Since the early part of the twentieth century, hundreds of different types of viruses have been recognized and described. Bacteriophages are the best characterized and studied group of viruses today, despite being the last major group to be recognized. This is attributed to the technically easier propagation and manipulation of phages than other types of viruses. Phages can be easily isolated from any bacterial environment (e.g. sewage, gastrointestinal tract) consisting of many closely related bacterial strains. If a sample of sewage filtrate is inoculated with a growing bacterial culture and spread on an agar plate, the bacterial growth on it will be seen interrupted by small clear zones which are termed plaques. Each plaque corresponds to the propagation of a single phage - 139 -Bdtllonbrro bact»riovoru% Fig V.l Relative sizes of a bacterium (E. coli) and an assortment of biological entities (Bdellovibrio, bacteriophages and a Colicin) which attack it. Colicin K is beyond the resolution of current electron microscopes - 140 -particle and is analogous to a single bacterial colony. The development of the electron microscope permitted the "first glimpses" into the morphological characteristics of viruses which were elusive to detection by even the most powerful light microscope. Bacteriophages are morphologically quite distinct from other virus types, owing to their structural complexity and greater diversity. An indication of the relative size of a bacterium in comparison to that of various bacteriophages238 is given in Fig. V.l. Bdellovibrio bacteriovorus is a cell (bacterium) able to move, respire, increase in size and reproduce by fission. It is not a true phage despite its ability to produce clear zones in bacterial lawns, and cause a decrease in turbidity of bacterial cultures. Colicin K which also shows the phenomena described above is a protein molecule unable to reproduce itself. Bacteriophages are intermediate in size between these two, and are of intermediate complexity. Morphological classification of bacteriophages by Bradley 2 3 9 enabled the characterization of six types (see Fig. V.2). Types A-C contain two strands of DNA, while Type D has one. These four types are unique to bacteriophages. Types E and F contain a strand of RNA and a strand of DNA, respectively, and are similar to many common plant and animal viruses. The bacteriophage consists of a head of basically icosahedral shape composed of repeating identical protein monomers (see Fig. V.3) and contains the nucleic acid within it. The tail is a helical arrangement of protein molecules forming a tube. This tube is encased by a sheath made of a different protein monomer extending from the collar to the end plate. The contractibility of the sheath is important in phage replication and infection; it forces the nucleic acid through the hollow tail into the host bacterium. The plate contains small pins (or spikes) to which are attached six long tail fibres. These fibres are important in attaching the phages to the host cells. These structural properties are basically of - 141 -2-DNA 2-DNA 2-DNA 1-DNA 1-RNA 1-DNA Fig V.2 Basic morphological types of bacteriophages with the types of nucleic acid Head — Coat p r o t e i n •Phage DNA C o l l a r C o n t r a c t i l e sheath T a i l f i b e r s !4> *, (a) (b) Fig V.3 (a) Schematic diagram demonstrating the structure of a T-even bacteriophage, (b) Electron micrograph of a T 4 bacteriophage - 142 -the Bradley Type A phages. Type B is similar to type A, but lacks the contractile sheath. Type C has a base plate and spikes at the end of a very short tail, but has neither a sheath nor tail fibres. Type D contains a capsomere (or knob) at each apex, and lacks a tail. The presence of only a head, with no tail or capsomeres, is categorized as Type E, with Type F being quite different to all these, having a filamentous nature. Variations within any one type can occur due to differences in the size and shape of the head, or the number of tail spikes. E. coli viruses are known to belong to all six Bradley types, while 55 bacteriophages isolated for Klebsiella showed that the majority belonged to Bradley group C, with 12 phages being of type B and 3 of type A. 2 4 0 Bacterio-phages of E. coli (or coliphages) have been studied in great detail. There are seven "T" phages, T i - T 7. The T-even phages T 2, T 4 and T 6 belong to Bradley group A. However one of the T-even phages has an octahedral head instead of the usual icosahedral type. T i and T5 belong to group B, while T3 and T7 are C. Phages exist in three distinguishable forms. 2 3 5 (i) The free particle, where it is virtually inert in a state of suspended animation (no metabolic processes occur). (ii) The prophage or lysogenic form, where the host bacterium carrying the phage is immune to superinfection by the phage and is not destroyed by it. In this form, the phage can replicate within the host under certain conditions. These are called "temperate" phages. (iii) Vegetative or lytic phages, also termed "virulent" which are responsible for total destruction of the host in the replication process resulting in the formation of new phages at the expense of the host DNA. The lytic cycle of a virulent phage involves the following phases:241 (see Fig. V.4). - 143 -1 h v . » -Phage DNA Bacterial chromosome; 1 Mr f Late protein ' » 5vW. **** ^ . . . • . . . Fig V.4 A schematic diagram illustrating the steps in the infection of a bacterium by a T-even phage - 144 -(i) Adsorption of the phage particle to the susceptible host (Fig. V.4.a,b,c). (ii) Contraction of the sheath and penetration of the cell by the tail followed by injection of viral DNA (Fig. V.4.d). (iii) Replication of the phage nucleic acid and synthesis of phage protein (Fig. V.4.e,f). (iv) Phage maturation and release (Fig. V.4.g,h,i). The absorption of the phage to its receptor is highly specific. 2^ 2 Flagella, pili, capsules, LPS, teichoic acid-peptidoglycan complexes, and surface protein can act as receptors for specific phages limiting the range of bacterial strains viable to adsorption by a single phage type. The expolysaccharide acts as a receptor for phage binding. Hence non-capsular or non-slime producing mutants have been found to be phage resistant. In the next stage the exopolysaccharide is penetrated (depolymerized) by the adsorbed phage, and this is evident by the formation of a halo around the plaque. The halo is due to an enzyme that can decapsulate the bacteria (hydrolyze the capsule) without killing i t . 2 4 3 The adsorption and penetration of the capsular polysaccharide of E. coli K29 by 029, the attachment of the phage to the outer membrane of the cell, and the release of viral DNA into the host has been studied by electron microscopy 2 4 4 (see Fig. V.5). Bacteriophages infecting capsular host cells carry exopolysaccharide degrading enzymatic activity. 2 4 0 These enzymes are classified according to the type of reaction they catalyze, and the genus of the bacterial host. The commonly found enzymes are the hydrolases which are of two types, glycanases (glycoside hydrolases) and deacetylases (carboxyl ester or amide hydrolases). The less commonly found lyases (carbon-oxygen lyase) act on exopolysaccharides to produce oligosaccharides terminating in unsaturated glycuronic acids. 3 3 These types of viral penetrases have been studied using phages for E. coli. Klebsiella. - 145 -Lipopolysaccharide Fig V.5 Drawing depicting the adsorption of E. coli 029 onto the encapsulated E. coli K29. The phage tail spikes recognize and bind the exopolysaccharide. The endoglucosidase of the tail spikes hydrolyze Glc^^-GlcA linkages in the glycan strand of the exopolysaccharide opening a path for the phage particle. Triggering and ejection do not occur until the phage has reached the cell wall; penetration of the phage nucleic acid apparently takes place at points of fusion between wall and cytoplasmic membrane - 146 -Salmonella. Shigella. Proteus. Pseudomonas. Rhizobium. and Streptococcus. These studies indicated that the enzymatic activity was associated with the tail spikes of the phages.245 A comparative study on the depolymerizing ability of 55 Klebsiella bacteriophages on the host polysaccharides 2 4 0 revealed the following: (i) Their enzymatic activity (depolymerizing ability) was very specific. Thirty three phages cross reacted with none, eighteen with one, two phages with two, and one each with three or four of the 73 heterologous polysaccharides. (ii) Most often the reducing end sugar formed is linked at position three. (iii) In the majority of cases, /3-glycosidic linkages are hydrolyzed. (iv) In most polysaccharides that are acted upon by several phage enzymes, the same glycosidic bonds are cleaved by the different phages. (v) In most cases cleavage occurred on either side of the sugar unit carrying the negative charge (e.g. uronic acid, pyruvic acetal). However reducing glucuronic acids were not produced. This observation is consistent with the evidence that reciprocal charges on the phage tail and the receptor sites of cells are involved in the formation of electrostatic bonds during attachment.235 All these Klebsiella phages have tails since they belong to Bradley groups A, B or C. 2 4 0 With some E. coli phages, it has been shown that such bonds are formed between the amino groups on the phage tail and carboxyl groups on the cell surface, thus making the adsorption of the phage a pH dependent reaction. 2 3 5 The use of bacteriophage-associated glycanases permits the isolation of oligosaccharide fragments usually corresponding to single repeating-units of the polysaccharides. Importantly, they enable the isolation of oligomers containing acid-labile substituents (acetyl, pyruvic acetal). Such oligosaccharides are useful - 147 -for examination by n.m.r. spectroscopy,2^ for studies of conformation in solution, for coupling as haptens to immunoglobulins^ for immunological studies, and for verification of the structures of the original polysaccharides. The bacteriophage degradations of Klebsiella K26 and K44, using <t>26 and </>44 respectively, are presented here. The oligosaccharides obtained were characterized by chemical and spectroscopic methods. V.2. Isolation and analysis of the oligosaccharides from the depolymerization of Klebsiella K44. V.2.1 Abstract Bacteriophages (<t>) have been used to degrade polysaccharides into oligo-saccharides containing one or more of their repeating-units. The capsular polysaccharide from Klebsiella K44 contains an acetate group, and n.m.r. spectroscopy and chemical methods have been employed to prove its linkage to 0-6 of the 0-4 linked glucose residue. 4>44 was shown to be an a-glucosidase, incapable of recognizing the acetate moiety, and thus was able to depolymerize the polysaccharide into pentasaccharide repeating-units, some of which contained acetate on 0-6 of the reducing glucose residue. The two oligosaccharides were studied by ^H- and C^-n.m.r. spectroscopy, and their spectra were compared with those of the native and deacetylated polysaccharide. ^C-N.m.r. was a useful tool for locating the 0-6 linked acetate, the position of which was confirmed by the methyl vinyl ether method. The importance of using bacteriophages to produce oligosaccharides is highlighted by the better results obtained with the - 148 -oligosaccharide in comparison to the polysaccharide, both in n.m.r. spectroscopy and the methyl vinyl ether protection method. V.2.2 Introduction In recent years, bacteriophages have been used in our laboratories as a means of generating simple oligosaccharide repeating-units from Klebsiella exopolysaccharides.2^-252 The enzymic activity of the bacteriophages on several Klebsiella host capsules has been studied2"*0 and 044 was reported to be an endo glucanase of unknown position of cleavage. The isolation of a single repeating-unit enabled us to determine the nature of the reducing glucose residue by chemical methods and by *H- and C^-n.m.r. spectroscopy. We now report the isolation of the pentasaccharide repeating-unit from Klebsiella K44 polysaccharide of known structure 2^ by using 4>44 and the location of the position, hitherto unknown, of the acetate group. V.2.3 Results and Discussion Preparation of polysaccharide - Capsular polysaccharide from Klebsiella K44 was grown and purified as previously described. 2^ Due to its very low degree of polymerization [M w 2 x 105], the polysaccharide is extremely lyophilic; on removal of the O-acetate with mild base shows no observable change in viscosity. The acetate group is relatively stable in mild base (0.01M NaOH, room temperature, 24 h), but is easily removed with a stronger concentration of NaOH (0.3M) under the same conditions. - 149 -Generation and isolation of oligosaccharides - Bacteriophage 044 was propagated on the host strain Klebsiella K44 by using a nutrient broth medium.2**? Depolymerization of the polysaccharide 2^ with 044 generated two pentasaccharides 4 and 5 corresponding to the repeating-unit of the polysaccha-ride. These pentasaccharides were isolated by preparative paper chromatography. 0 -D-GlcA-( 1 -> 2)- a-L-Rha-( 1 — 3)-a-L-Rha-( 1 — 3)- /3-D-Glc-( 1 -»4)-D-Glc~OH L I6 OR 4 R = H 5 R = Ac 6 Alditol of 4 7 Alditol of 5 K44 Oligosaccharides Quantitation of the acetate groups22? - Spectrophotometric determination of the hydroxamic acid formed by the acetate using acidic ferric chloride to generate a coloured compound indicated that the polysaccharide contained 64 mole percent acetate. Oligosaccharide 5 was shown to contain one mole of acetate per mole (100 mole percent acetate). This was in accordance with the yields of compounds 4 (34%) and 5 (66%) isolated. Characterization of oligosaccharides 4 and 5 - The degree of polymeriza-tion of the oligosaccharides was determined by reduction with sodium borohydride and conversion of the oligosaccharide alditol 6 formed into the aldonontrile derivative.95'2**? The ratios of rhamnononitrile, glucononitrile and glucitol obtained were 2:1:1. On reduction of the uronic acid, the ratio was changed to - 150 -Table V.l Methylation analysis of oligosaccharides 5 and 6 Methylated sugar- Relative retention Mole fas alditol acetate) time on OV-225^ II HI 1,2,3,5,6-Glc 0.44 18.7 12.8 -3,4-Rha 0.87 28.0 21.3 21.8 2,4-Rha 0.94 30.8 24.2 23.3 2,4,6-Glc 1.64 22.5 23.0 27.3 2,3,4-Glc 1.95 - 18.7 -2,3,6-Glc 2.00 - - 27.6 3 1,2,3,5,6-Glc = 4-0-acetyl-l,2,3,5,6-penta-0-methyl glucitol etc. * Retention time of partially methylated alditol acetates,relative to that of l,5-di-0-acetyl,2,3,4,6-tetra-0-methyl-D-glucitol on a column of 3% of OV-225 on Gas Chrom Q (100-120 mesh) at 200° isothermal. - Values corrected using e.c.r. correction factors. 1 2 2 — I, methylated oligosaccharide 6; II, methylated, reduced oligosaccharide 6; III, methylated oligosaccharide 5. - 151 -Table V.2 N.M.R. data for Klebsiella K44 polysaccharide and the derived oligosaccharides Compound3- H^-N.m.r. data  1] 2 Integral Assignment0-protons 1 3C-N m.r. data  P.p.m.- Assignment-5.29 5.21 5.13 4.67 4.65 4.53 1.32 1.28 5.29 5.20 5.12 4.66 4.53 4.47 4.33 4.29 s 8 8 8 8 8 s 8 12,2 8 12,4 12,4 0.4 1 1 0.6 1 3 3 1 1.6 1 1 0.5 0.5 i-Rha-a 4Glc~OH a ^Rha-ot GlcA-3 4Glc~OH 3 ^Glc-3 CH 3 of Rha CH 3 of Rha 2-Rhj 0.4 4G\C~OH ^Rha-a GlcA-, Glc-OH 3 3 H 6 S of ^ Glc-OH 105.03 103.17 103.13 101.65 101.58 96.55 92.61 61.51 61.04 60.92 17.40 17.33 105.04 103.39 103.33 101.67 101.60 ^Glc-a , 3 GlcA-3 ^Glc-2-Rha-a ^Rha-^Glc-OH 3 ^Glc-OH a C-6 ^ Glc-3 C-6 ^Glc-OH 3 C-6 ^Glc-OH a CH 3 of Rha GlcA-3 ^Glc-3 2-Rha-a ^Rha-96.66 ^Glc-OH } H 6 R of ^ Glc-OH 92.66 ' a,3 ^Glc-OH a - 152 -Table V.2 .. continued Compound- H-N.m.r, data ^- 1] 2 Integral Assignment0-CHz) protons 13C-N.m.r. data P.p.m.- Assignment^ 2.14 1 1.32 8 1.28 8 5.28 5.12 4.68 8 4.59 8 1.31 8 1.28 8 2.6 C H 3 of acetate 3 C H 3 of Rha 3 C H 3 of Rha 1 1 1 1 3 3 2- Rha-a 3- Rha-a GlcA-3 2GIC-C H 3 of Rha C H 3 of Rha 63.87 C-6 4-Glc OH a,6 61.53 C-6 2G1C— B 21.05 C H 3 of acetate 17.41 17.35 } C H 3 of Rha 5.33 5.12 4.62 8 4.55 8 4.46 12 4.25 b 2.15 s 1.28 b 3 6 2- Rha-a S-Rha-a GlcA-B 3- Glc-B H6S of Glucitol 16 OAc **6R °f Glucitol 6| OAc C H 3 of acetate CH 3 of Rha - 153 -Table V.2 .. continued Compound 3- H^-N.m.r. data  1] 2 Integral Assignment0-fHz) protons 13C-N.m.r. data P.p.m.£ Assignment6-Native K44 [^lcA^-RhaJ-^Rha^Glc^Glc^n 3 a a 3 a 6 64% OAc 5.41 0.6 OAc 46' 4Glc- 105.09 ^GlcA-5.38 5.29 5.15 4.64 2.16 1.32 1.28 s s s 8 4.50 8 4.40 8 8 0.4 1 1 0.9 0.4 1.2 1.5 3 3 4Glc-a 2-Rha-3-Rha-a ^GlcA-^Glci-icic-3 a OAc 3<JIC!_4Q1C_ 3 a OAc 61 H 6 S of Glc _ CH3 of acetate CH 3 of Rha CH 3 of Rha 103.30 lGlcl-k} OAc .61 lc-103.08 2G 1 C1_4 GJ £ 101.76 2 R H A _ 6 101.64 3 R h J OAc 6| 99.61 4 ^ . a 99.51 4Glc-OAc 63.44 C-6 4GIC-a 61.60 C-6 2GIC-3 60.39 C-6 ^ Glc-a 21.10 CH 3 of acetate 17.44 C H 3 o f R h a 17.37 CH 3 of Rha - 154 -Table V.2 .. continued Compound3- ^-N.m.r. data 13C-N.m.r. data  £* 1] 2 Integral Assignment0- P.p.m.3- Assignment6-(Hz) protons Deacetvlated K44 [^lcAl^Rha^Rhai^Glc-LikjlcI-],, 3 a a 3 a 5- 3 7 s 1 4GIC- 105.08 ^GlcA-a 3 5-29 s 1 =Rha- 103.15 3-Glc-a g 515 s 1 3-Rha- 101.82 2-Rha-a a 4.63 b 1 4G1CA- 101.57 3Rha-3 a 4.50 b 1 2GI C- 0 99.16 4G1C-B a 1.32 b 3 CH, of Rha 61.62 C-6 ^Glc-3 1.28 b 3 CH 3 of Rha 60.60 C-6 ^Glc-a 17.42 CH 3 of Rha 17.32 CH 3 of Rha - For the sources of 4,5,6, and 7, see text. & Chemical shift relative to internal acetone; 2.23 downfield from sodium 4,4-dimethyl-4-silapentane-l-sulfonate (DSS). - The numerical prefix indicates the position at which the sugar is substituted; aor/3 , the configuration of the glycosidic bond, or the anomer in the case of a (terminal) reducing sugar residue. Thus -2-Rha— refers to the anomeric proton of a 2-linked glycosyl residue in the a-anomeric configuration. The absence of a numerical prefix indicates a terminal nonreducing group. — Chemical shift in p.p.m. downfield from Me4Si, relative to internal acetone; 31.07 p.p.m. downfield from DSS. £ As in £ but for 1 3 C nuclei. - 155 -2:2:1. Hence compounds 4 and 5 are two pentasaccharides. By reducing 5 with NaBH4 for 30 min under mild conditions, the acetate was left intact (compound 7). Methylation of compound 6 by the Hakomori method,115 reduction of a portion with lithium aluminum hydride, and analysis of both as the partially methylated alditol acetates revealed the compounds listed in Table V.l, columns I and II. The compounds obtained by methylation of oligosaccharide 5 (see Table V.l. column III), when compared with those from the methylated oligosaccharide 6 indicate the presence of a terminal glucuronic acid and an 0-4 linked, reducing, glucose residue. The p.m.r. spectra of 4 and 5 showed the disappearance of an asignal in the polysaccharide spectrum and its replacement with a pair correspond-ing to a reducing a:/3 ratio of 0.4:0.6 H. These reducing signals were missing in the spectra of 6 and 7 [see Table V.2]. This proves that bacteriophage 044 is an endo a-glucosidase. Location of the position of the acetate function - The native polysaccha-ride and oligosaccharide 5 were both treated with methyl vinyl ether.22** The protected products were then methylated by the Hakomori procedure 1 1 5 which, due to the basicity of the reaction conditions, removes the acetate function and permits methylation of the free hydroxyl position generated. On analysis as the alditol acetates, the compounds in Table V.3 columns I and II were obtained. In addition to 6-O-methylglucose small amounts of 2-O-methylrhamnose and 2-0-methyl-glucose were present in the sample from the polysaccharide. This may be attributed to the viscosity of the polysaccharide in comparison to that of the oligosaccharide 5 resulting in under protection of the polysaccharide (Fig. V.6). The predominance of 6-O-methylglucose in the product proves that the O-acetate is attached to 0-6 of one of the glucose residues. - 156 -Table V.3 Sugar analysis of the methyl vinyl ether protected, methylated, polysaccharide, and oligosaccharide 5 Alditol acetate- Relative retention Mole %  tirne^ I £ II UI 2-Rha 0.22 2.2 Rha 0.28 45.8 42.3 44.4 6-Glc (l-d)=- 0.54 - - 27.8 6-Glc 0.58 17.7 20.3 2-Glc 0.70 9.3 Glc 1.00 25.0 37.3 27.8 2-Rha = 2-0-methyl,l,3,4,5-tetra-0-acetyl rhamnitol etc. * Retention time relative to that of glucitol hexaacetate on a column of 3% of OV-225 on Gas Chrom Q (100-120 mesh) at 220° isothermal. I, polysaccharide; II, oligosaccharide 5; III, deuterium labelled oligosaccharide 5. 6-O-methyl-l,2,3,4,5-penta-0-acetylglucitol-[l-2H]. - 157 -Rha 2-Me Rha \ hi 6'Me Glc Glc 2-Me Glc Liu (a) Fig V.6 Acetate location: gas-liquid chromatogram of methylated alditol acetates derived from (a) K44 polysaccharide (b) penta-saccharide obtained by action of bacteriophage - 158 -In order to determine which of the two glucose residues is acetylated, two separate methods were employed. One method involved the removal of the protecting methyl vinyl ether under mild, acidic conditions leaving the methyl group and the glycosidic linkages intact, and then ethylating the deprotected hydroxyl groups. Since standards were not available, the native polysaccharide was ethylated, 1 1 5 and converted into alditol acetates (see Table V.4 column I). Thus, the vinyl ether-protected, methylated, polysaccharide was treated with 1% H2SO4 ~ acetone (3:2); the resulting deprotected polymer was ethylated by the Hakomori procedure. The alditol acetates obtained (Table V.4 column II) show that the 4-linked glucose contained a 6-Q-methyl group but the 3-linked glucose was fully ethylated (proved by g.l.c.-m.s.14^). The vinyl ether-protected, methylated, oligosaccharide was deprotected in the same manner and divided into two portions. One was reduced with sodium borohydride to its oligosaccharide alditol and then ethylated as before. The alditol acetates obtained (Table V.4 column III) showed the 3-linked glucose to be present as its ethylated derivative, but the 4-linked glucose, at the reducing end, had a methyl group at position 6. The second method involved the reduction of the remaining portion of the deprotected oligosaccharide with sodium borodeuteride in order to label the reducing glucose residue. Hydrolysis with 2M TFA and g.l.c.-m.s. of the alditol acetates (see Table V.3 column III) proved that only the 6-O-methylglucitol contained deuterium, while glucitol was unlabelled. These experiments confirmed that the methyl group is on O-6 of the reducing 4-linked glucose. The results therefore, prove the structure of oligosaccharide 5. ^-N.m.r. - The p.m.r. spectra of the native polysaccharide (Na + salt) and the deacetylated polysaccharide were not as well resolved as those of - 159 -Table V.4 Ethylation analysis of native K44 polysaccharide, and the 6-O-methyl polysaccharide and oligosaccharide Ethylated sugar3- Relative retention MoJe_%c. (as alditol acetate) time on OV-225^ f2 II HI 3,4-Et-Rha 0.85 22.2 56.9 43.9 2,4-Et-Rha 0.89 31.0 2,4,6-Et-Glc 1.58 12.3 25.4 32.3 2,3,6-Et-Glc 1.88 30.8 2,3-Et-6-Me-Glc 0.81 - 30.7 l,2,3,5-Et-6-Me-Glc 0.46 - - 14.5 3- 3,4-Et-Rha = 3,4-di-0-ethyl-l,2,5-tri-0-acetylrhamnitol. k Retention time of partially alkylated alditol acetates relative to that of l,5-di-0-acetyl-2,3,4,6-tetra-O-methylglucitol on a column of 3% of OV-225 on Gas Chrom Q (100-120 mesh) at 200° isothermal. c Values corrected using e.c.r. correction factors 1 2 2. - I, native K44 polysaccharide; II, methyl vinyl ether protected, methylated polysaccharide; III, methyl vinyl ether protected, methylated, deprotected, NaBH4 reduced oligosaccharide 5. - 160 -oligosaccharides 4 and 5 (see Table V.2). The twinning of H-l of a-and /3-glucose in the spectrum of the native polysaccharide is assumed to be due to the partial presence of acetate, since it is not evident in the spectrum of the deacetylated polysaccharide. There is no such observation in the spectra of 4, 5, 6 or 7 because they are either totally devoid of acetate or have 100% acetate intact. Two doublets of doublets in the range 5 = 4.27-4.35 (see Table II) are observed in the spectrum of oligosaccharide 5 and was assigned to H^R of the reducing glucose which contains the 6-OAc. 2 2 4' 2 2 5 The doublet of doublets at 5 = 4.53 was assigned to Hg§ of the reducing glucose. These signals are not observed in 4, 6 or in the deacetylated polysaccharide; but a broad doublet at 5 = 4.46 and a broader signal at 5 = 4.25 in the spectrum of 7 arise from Hgg and Hgj^ of the glucitol residue acetylated at 0-6. A similar pattern is also seen in the spectrum of the native polysaccharide. The twinning in the acetate signal, observed only for compound 5, is probably due to the fact that this group is located on the reducing residue and is, thus, influenced by the ct//3 mutarotational equilibrium. '•'C-N.m.r, - Since the C-6 signal of hexoses appears upfield from the ring carbon atom region, the shift towards lower field caused by the 6-0- acetate is clearly observed (see Fig. V.7). Again, the partial presence of acetate is obvious from the two signals obtained for C-6 of the a-glucosyl residue at 60.39 and 63.44 p.p.m. (see Fig. V.7 (a)), which collapse to one signal at 60.60 p.p.m. on deacetylation (see Fig. V.7 (b)). The anomeric signals of /3-glucose and a-glucose both show twinning which disappears on deacetylation and hence may be attributed to the presence of acetate. 2 1 0 The spectra of the oligosaccharides 4 and 5 show the shift of the C-6 signal of the reducing glucose residue (a) 60.92 and (/3) 61.04 p.p.m. to 63.87 p.p.m., - 161 b ) 90 60 30 p.p.m. Fig V.7 The 13C-n.m.r. spectra of (a) native K44 polysaccharide, (b) deacetylated polysaccharide, (c) oligosaccharide 5, (d) oligosaccharide 4 - 162 -whereas the 0 -glucose (non-reducing) signal remains virtually unchanged (61.51 and 61.53 p.p.m.) when C-6 of the reducing glucose is fully acetylated. The anomeric signal of /3-glucose exhibits a twinning which does not disappear on deacetylation and hence was thought to be caused by its close proximity to the reducing end. Comparison of the anomeric signals of 4 and 5 gave downfield shifts of 0.3 p.p.m. for /3-glucose, 0.11 p.p.m. for /3-glucose~OH, and 0.05 p.p.m. for a-glucose-OH. These shifts are caused by the presence of acetate, and are similar to those of the twinned anomeric signals in the native polysaccharide. Twinning due to the presence of acetate is also observed to a small degree in the ring carbon atom region of the native polysaccharide. V.2.4 Conclusion The generation of simple oligosaccharides by using bacteriophage was useful in two ways. Firstly, sharper and clearer spectra were obtained by n.m.r. spectros-copy; secondly, and more importantly, two oligosaccharides were produced which differed in acetate content at C-6 of glucose. From the yield of the oligo-saccharides obtained,, it is reasonable to conclude that 044 is an a-glucosidase insensitive to the presence of acetate at Q-6. V.2.5 Experimental General Methods - All instrumentation used has been described in Section III. Descending paper chromatography was performed on Whatman No. 1 paper, with the solvent systems listed in Section III. Analytical g.l.c. separations were - 163 -performed in columns A and C (see Section III). G.l.c.-m.s. was carried out on a Nermag R 10-10 mass spectrometer using capillary columns coated with (i) DB-225 or (ii) SE-30. Isolation of the polysaccharide - A culture of Klebsiella K44 (courtesy of Dr. I 0rskov, Copenhagen, Denmark) was grown and harvested by the usual procedure.213 The acidic polysaccharide was isolated by one precipitation with Cetavlon. The molecular weight was found to be 2 x 105 by gel-permeation chromatography on a column of Sephadex 4B (courtesy of Dr. S.C. Churms, University of Cape Town, South Africa). Propagation of bacteriophage and depolvmerization - Bacteriophage 044 (courtesy of Dr. S. Stirm, Freiburg, W. Germany) was propagated on the host Klebsiella serotype K44 in the usual way.247'2-*4 T n e ^ 4 4 capsular polysaccha-ride (1.0 g) was degraded with the bacteriophage 044 according to the method previously described.24** Oligosaccharides 4 and 5 were isolated by preparative paper chromatography in solvent A for 57 h. The yields of 4 and 5 obtained were 83.2 mg and 176.8 mg. Determination of acetate content 2 2 7 - To the acetate containing solution in sodium acetate buffer, a solution of 2M hydroxylamine hydrochloride (1.0 mL) and 3.5M NaOH (1.0 mL) were added, swirled on a vortex mixer, and left standing at room temperature for 2 min. Then one part of cone. HC1 in two parts of water (1.0 mL) was added, swirled, and 0.37M FeCl3-6H20 in 0.1N HC1 (1.0 mL) was - 164 -added and swirled again. The purple-brown colour was measured immediately against a reagent blank at 540 nm. Deacetvlation of the native polysaccharide - K44 polysaccharide (54 mg) was treated with 0.3M NaOH (20 mL), stirred at room temperature for 2 d, dialyzed, and freeze dried. Removal of the acetate group was confirmed by both 1 3C- and JH-n.m.r. Treatment with methyl vinyl ether22** - A dried sample of polysaccharide or oligosaccharide was dissolved in DMSO, together with rj-toluenesulfonic acid, sealed and flushed with N2 and left stirring overnight. Methyl vinyl ether was introduced in two aliquots (3 mL each) at -60°, with stirring for 4 h each time. The product was purified on a column of Sephadex LH-20 by elution with acetone. Methylation of the methyl vinyl ether-protected derivatives - The protected sample was dissolved in DMSO and treated with dimsyl anion for 2 h. Iodomethane was added to the cooled solution which was stirred for 1 h and extracted into CHCI3. The product was purified on Sephadex LH-20 with methanol as the eluent. Removal of the vinyl ether protection - The methylated, protected product was dissolved in acetone--l% H 2S0 4 (2:3) and stirred at room temperature - 165 -for 24 h. The acid was neutralized with 2N NaOH, and the solution was evaporated to dryness, dissolved in DMSO, filtered, and evaporated to yield the product. Ethylation of the methylated deprotected polysaccharide - The ethylation of the native polysaccharide and the methylated, deprotected polysaccharide was performed as follows: The sample was dissolved in DMSO under nitrogen, dimsyl anion was added, and the solution was stirred for 3.5 h. Iodoethane was added to the cooled solution which was stirred a further 1 h. The product was dialyzed overnight, extracted into chloroform, hydrolyzed with 2M TFA, and converted into alditol acetates in the usual way. Treatment of the methylated deprotected oligosaccharide - (a) One portion was reduced with sodium borodeuteride, deionized with IR-120 (H +) resin, and coevaporated with methanol. The deuterium-labelled oligosaccharide alditol was hydrolyzed with 2M TFA, and transformed into the peracetylated alditols. G.l.c. was carried out on OV-225 at 220° isothermally. A DB-225 capillary column programmed at 150° for 1 min and then 10°/min to 220° was used for g.l.c.-m.s. (b) The remainder was reduced with NaBH4, ethylated as before and extracted into chloroform prior to dialysis. The product was hydrolyzed with 2M TFA and converted into the perethylated alditol acetates. These were analyzed by g.l.c. on OV-225 at 200° isothermally, and by g.l.c.-m.s. on column (ii), programmed at 80° for 1 min, and then 5°/min to 250°. - 166 -V.3. Isolation and characterization of the products from the bacteriophage depolymerization of Klebsiella K26. V.3.1 Abstract Bacteriophage 026 was used to depolymerize the polysaccharide of Klebsiella K26. Three oligosaccharides were obtained. The major product was the heptasaccharide repeating-unit, with one of the minor products being the fourteen sugar oligosaccharide corresponding to two repeating-units. The third oligosaccharide was unusual, since it was a hexasaccharide devoid of the terminal pyruvate-containing galactose unit present in the side chain of the normal repeating unit. 026 was shown to be a /3-galactosidase and hence may have the ability to remove the terminal /3-galactose residue in the side chain. V.3.2. Introduction A bacteriophage capable of infecting its host bacterial cell often has an enzymatic degrading activity towards the capsular exopolysaccharide in order to penetrate the cell wall. The need to gain access to the interior of the bacterial cell is a logical reason for a phage to exhibit such a depolymerizing ability rather than acting as a hydrolase, removing only side chains and leaving the capsule intact, even though such instances are known33. Oligosaccharides containing 1-carboxyethylidene substituents (pyruvate acetals) can be obtained only by bacteriophage depolymerization of polysaccha-rides. As a continuation of our studies on oligosaccharide production by this - 167 -Table V.5: Determination of the degree of polymerization and the reducing end of oligosaccharides Pla, PI and P2 from Klebsiella K26 Peracetvlated derivative of T3- Mole %  Pla PI P2 Mannonitrile 1.00 32.6 30.3 31.4 Glucononitrile 1.05 44.3 31.9 35.3 Galactononitrile 1.09 - 17.4 23.2 Galactitol 1.41 23.1 20.2 10.0 - Determined on a column of DB17 programmed from 180° for 2 min and then 5°/min to 220°. - 168 -method, we now report on the nature of the products obtained by depolymeriza-tion of the capsular polysaccharide of Klebsiella K26. V.3.3. Results and discussion Klebsiella 0 26 was isolated from sewage, purified in the usual manner, and propagated on the host strain in P-medium2^ instead of the usual nutrient broth2****. The depolymerization was carried out at 37° for 2 d. Dialysis of the generated oligomers and gel filtration of the dialyzate yielded Pla (8.2%), PI (54.1%) and P2 (11.9%); Pla was a hexasaccharide, PI the heptasaccharide repeating-unit, and P2 the dimer of PI. No higher oligomers were observed. Analysis of Pla revealed it to be a hexasaccharide and PI a heptasaccha-ride, each with galactose at the reducing end. The difference between the two was the absence of the terminal pyruvate-containing galactose in Pla (see Table V.5). P2 also contained a reducing galactose unit, thus confirming 0 26 to have a /3-galactosidase activity. Methylation analysis of Pla confirmed the absence of the terminal galactose unit, and showed that the 4-linked glucose was replaced by a tetramethyl (terminal) glucose residue. PI contained the terminal galactose bearing the pyruvic acetal at positions 4 and 6 (see Table V.6). Comparison of the methylation data of Pla with those of the original polysaccharide 2^ showed that the 3-O-methylglucose (from glucuronic acid) was replaced by 2,3-di-O-methyl-glucose in Pla. Interestingly, the predominance of 2,5,6-tri-O-methylgalactose was attributed to the existence of 60% galactofuranose at the reducing end. In a similar comparison of the methylation data of P2, the reducing galactose unit of one repeat-unit was shown to be joined to 0-2 of the glucuronic acid of the second unit. Using the results from the original structure elucidation of the K26 Table V.6: - 169 -Methylation analyses of oligosaccharides Pla, PI and P2 Partially methylated Mol %£ (as alditol acetate)- -pb Pla Ad B E i A C P2 B C 1,2,4,5,6-Gal 0.76 - - - 8.8 - 5.3 2,3,4,6-Glc 1.00 25.8 20.4 - - - -3,4,6-Man 1.38 30.1 17.5 20.8 23.1 17.5 28.6 2,5,6-Gal 1.46 15.0 5.3 11.3 - 3.9 -2,3,6-Glc 2,3,4-Glc 1.51 21.8 36.3 43.2 48.8 46.1 45.7 2,4,6-Man 1.53 2,4,6-Gal 1.58 7.3 2.9 4.0 - 10.1 10.4 2,3-Glc 2.13 - 17.3 - - 6.6 -2,3-Gal 2.18 - - 20.7 19.3 9.6 19.6 3-Glc 2.73 _ _ _ _ 6.2 _ 3- 1,2,4,5,6-Gal = 3-0-acetyl-l,2,4,5,6-penta-0-methylgalactitol, etc. * Determined on a column of DB17 programmed from 180° for 1 min and then 2°/min to 250°. - Values corrected using e.c.r. factors given by Albersheim et.al. 1 2 2 — A, methyla-ted oligosaccharides; B, methylated LiAlH4 reduced oligosaccharides; C, NaBH4 reduced and methylated oligosaccharides. - 170 -polysaccharide 2^ together with the present data, the structures of Pla, PI and P2 can be given as follows: a cAi-^Mani-^Manl-S-Gal 4 1 G c Glc a a GlcA^Mani-^ManJ-^Gal a a a " l l Glc »ll Glc "I. Gal 6\ 14 pyr Pla PI Glc Ai-^Ma ni-^Ma ni-2-Ga l i - ^ l c A^ -3-Ma n^Ma ni-3-Ga 1 a 4 1 Glc M 1 Glc 4 1 Gal pyr a a a 0 0 0 4 1 Glc 6 1 Glc »|. Gal pyr a N.m.r. data P 2 The anomeric signals in the ]H- and 13C-n.m.r. spectra of Pla indicate the presence of six sugar residues, while those of PI indicate seven. Assignment of - 171 -Table V.7 400 MHz'H-N.M.R. data for Pla, PI and P2 oligosaccharides from the bacteriophage degradation of Klebsiella K26 Compound- A* Jj 2 Integral Assignment-(HzJ~ Proton Pla 5.43 s 1 S-Glc— a 4GICA_ 3  5.33 s 1 5.29 1 0.3 5.25 s 1 5.16 6 0.25 5.06 s 1 4.62 8 0.4 4.48 8 1 4.41 8 1 5.43 s 1 5.33 s 1 5.30 bs 0.3 5.25 s 1 5.17 6 0.25 5.07 s 1 4.63 8 0.4 4.52 8 2 4.42 8 1 1.57 s 3 3-Galp~OH, 3-Galf~OH a B 3-Man— a 3-Galf-OH a 2- Man— a 3- Galp-OH B G l c " B H-5 of ^ l c A — a PI   6G1C_ a 4 *GlcA_ a 3-Galp-OH, 3-Galf-OH a B 3-Man— a 3-Galf~OH a -Man— a 3-Gal-OH , e ^Glc_, pyrCGal— B . B H-5 of ^GlcA_ a CH3 of pyr - 172 -Table V.7 continued Compound-( H t r Integral  Proton Assignment5-P2 alditol of Pla alditol of PI 5.50 3 1 5.46 4 1 5.43 4 1 5.33 4 1 5.29 4 0.3 5.25 s 2 5.17 8 0.2 5.07 s 2 4.66 8 1 4.63 8 0.5 4.52 8 4 4.45 8 1 4.43 8 1 1.58 s 6 5.49 4 1 5.29 b 2 5.10 s 1 4.53 8 1 5.43 4 1 5.34 4 1 5.24 s 1 ^^GIcA — a $Glc— a S-Glc— 4G1CA_ 3-Galp-OH, ^ Galf-OH a 6 3-Man— a ^Galf-OH a -Man— a S-Gal— 6 ^Galp-OH 8 ^Glc—, pyr=TGal_ 6 6 H-5 of 2JGlc_ a H-5 of 4Glc— a C H 3 of pyr S-Glc— a ^-GlcA—, ^ Man_ a a 2-Man— a G l c — S-Glc-^GlcA— a ^Man— a - 173 -Table V.7 » continued Compound3- A* J j 2 Integral Assignment0-(Hz]~ Proton 5.09 s 1 2-Man— a 4.53 8 1 ^ ' c - B 4.52 8 1 pyr=rGal_ 6 4.45 8 1 H-5 of ^ GlcA — a alditol of P2 5.47 b 2 6G1C_, ZAG\CA— a a 5.44 b 1 fiGlc. a 5.25 s 3 ^GlcA-, 3-Man— a a 5.07 s 1 2Mani-3-Gal_ a 8 5.05 s 1 2ManJ-3-Galactitol a 4.66 8 1 3-Gal_ B 4.53 8 2 4cic_ 8 4.50 8 2 pyrcTGal— - For the source of Pla, PI and P2 see text. * Chemical shift relative to internal acetone; 2.23 downfield from sodium 4,4-dimethyl-4-silapentane-l-sulfonate (D.S.S). Spectra were recorded at 95°^5. £ The numerical prefix indicates the position in which the sugar is substituted; the aor/3, the configuration of the glycosidic bond, or the anomer in the case of a (terminal) reducing sugar residue. Thus ^Xjlc-refers to the anomeric proton of a 6-linked glucosyl residue in the a-anomeric configuration. The absence of a numerical prefix indicates a (terminal) nonreducing group. - 174 -these signals was made possible by comparison of the spectral data of Pla, PI, P2 and the original K26 polysaccharide (see Table V.7). The presence of reducing galactofuranose was seen in the high temperature p.m.r. spectra of each of Pla, PI, and P2, but not in the 13C-n.m.r. spectra run at ambient temperature. This is consistent with the interesting observation that furanose signals were absent in their p.m.r. spectra recorded at ambient temperature. The chemical shifts of H-l and H-5 of the 4-linked glucuronic acid at 5 = 5.33 and 6 = 4.42 showed an upfield shift to 5 = 5.25 and 5 = 4.17 respectively on conversion to the sodium salt. The value of 5 = 4.17 was observed only in the spectra run at ambient temperature; at high temperature the signal overlapped with the HOD peak. All signals showed slight changes in chemical shifts depending on the temperature at which the spectra were recorded 2 5 1. Assignment of the signals for a-galacto-furanose and /3-galactofuranose was based on their J j 2 values, the/3-anomer usually having J i 2 -1 Hz as compared to 4-5 Hz for the a-anomer.256>257 -r-},e 13C-n.m.r. spectra were also interpreted by comparison and were easily assigned (see Table V.8). The 2-linked mannose adjacent to the reducing galactose showed a twinning 2 5 2 caused by the a and /3 -anomers.255 A small upfield shift in the terminal /3-glucose signal of Pla on substitution by the pyruvate-containing galactose was observed in the 13C-n.m.r. spectrum, whereas the p.m.r. spectrum showed a shift towards lower fields. The p.m.r. and l^ C-n.m.r. spectra of P2 indicated that the 2,4-linked glucuronic acid unit resonated downfield of the mono substituted uronic acid. The sugars adjacent to glucuronic acid, namely 3Man and 6Glc , both gave two signals each due to the dual environment a a caused by the linkage of one uronic acid unit at position 2. In contrast, the p.m.r. spectrum showed only one signal for the two mannose residues, but gave a signal each for the two 6-linked a-glucose units. This observation is an indication of the versatility of l^ C-n.m.r. in detecting small changes in environment. The signal - 175 -Table V.8 100 MHz 13C-N.M.R. data for Pla, PI and P2 isolated from Klebsiella K26 Compound- Chemical shift^ (p.p.m.) Assignment^ Pla 173.78 103.42 102.95 101.15 99.90 97.18 95.49 95.22 93.08 C-6 of GlcA Glc-B 2-Man— a G-Glc-a ^GlcA-a S-Gal-OH 6 ^Mani-^al-OH a 6 ^Mani-^Gal-OH a a 3- Gal-OH PI 175.05 174.12 103.64 103.26 102.99 101.18 99.82 97.20 95.50 95.22 93.10 25.85 C-l of pyr C-6 of GlcA pyr d G a l — 6 4Glc-B 3-Man— Q S-GlC-a 4GlcA-a B ^-Mani-^al-OH a 6 ^Man-L-S-Gal-OH a a 3-Gal-OH a CH3 of pyr - 176 -Table V.8 .. continued Compound3- Chemical shift^ Assignment^ (p.p.m.) P2 174.44 173.72 105.14 103.64 103.24 103.07 102.97 101.19 100.69 99.95 97.18 95.49 95.44 95.16 93.08 25.77 C-l of pyr C-6 of GlcA 2- G a l -B pyrCTGal- (x 2) 6 3- Mani-, ^ Glc- (x 2) a 6 2-Man— a ^Glc-a ^Glc-2-^GlcA-a 4 G I C A -3-Gal-OH 8 2-ManJ-^Gali-2-a 8 ^Mani^Gal-OH a ' a ^Mani-^-Gal-OH Q. a 3- Gal-OH a. CH 3 of pyr ^ For the source of Pla, PI and P2, see text, fi Chemical shift, in p.p.m., downfield from Me4Si, relative to internal acetone; 31.07 p.p.m. downfield from D.S.S. £ As in c, Table V.7, but for 1 3 C nuclei. - 177 -at 105.14 p.p.m. in the 13C-n.m.r. spectrum of P2 indicates that the inchain 3-linked galactose is /}, in accordance with the signal in the p.m.r. spectrum at 8 = 4.66 (Jjt2 = 8 Hz). The presence of this signal in the spectrum of P2 only, permits its assignment to the inchain 3-linked /3-galactose unit. V.3.4. Conclusion Unlike the previous cases where bacteriophages were used to generate oligosaccharides corresponding to single or multiple repeat-units, the presence of the hexasaccharide, in addition to the heptasaccharide repeating-unit may be due to one or both of the following reasons. Firstly, it is possible that 026 has a poor 6 . A recognition of the pyrdGal^-=Glc linkage due to its close resemblance to the 4 a phage labile ^-Gal^^GlcA linkage. Secondly, there is a possibility that when 026 was isolated, it also contained a second bacteriophage which has a weak enzymatic action towards the formerly mentioned linkage. Bacteriophage degradation of K l l with 011 resulted in cleavage of the Glc-L^GlcA linkage to produce P2; but it was found that some of the Gal-L-3-Glc linkages were also a. cleaved. 2 5 4 K74 (ref. 248) which has a similar phage labile galactose unit does not show such activity. ^Man^ManL^-Gal-L-a a p 3 a 1 GlcA 4 18 1 Gal 6 4 \ / pyr t 074 iGicl^GicAiJGali-t 011 4 a a a 1 \ Gal 0 1 1 ? 6 4 \ / pyr K74 K l l - 178 -The n.m.r. spectroscopic studies on these three oligosaccharides enabled the reassignment of some of the 0 anomeric linkages in the *H- and 13C-n.m.r. data, and are shown in Tables V.7 and V.8. The resolution of the spectra is enhanced in the oligosaccharides in comparison to the polysaccharide, and the fine details were easily noted. The presence of the furanose form of galactose is seen in aqueous solution at elevated temperatures, and it predominates over the pyranose form in DMSO as seen from the methylation data (Table V.6). V.3.5 Experimental General methods - All instrumentation used has been described in Section III. Paper chromatography was performed according to the description in Section III. Propagation of bacteriophage and depolymerization - Bacteriophage 0 26 was isolated from sewage, and purified by a series of propagations of single plaques on the host Klebsiella K26 (ref. 248). The purified 026 was propagated on its host in a dialyzable P-medium254 (see Section III.11.2) to yield 7 x IO 1 2 p.f.u. After dialysis of the medium and concentration to 200 mL, the solution was added to the polysaccharide (1 g/100 mL water) and incubated at 37° for 48 h. This solution was then concentrated to 80 mL and dialyzed against distilled water (3 x 1L). The dialyzates were combined, freeze dried (yield = 780 mg), and applied onto a column of Bio-Gel P4 (2.5 x 70 cm) to yield the three oligosaccharides, Pla (64 mg), PI (422 mg) and P2 (93 mg), on elution with water—pyridine—acetic acid (250:2.5:1). - 179 -Analysis of the oligosaccharides - Aqueous solutions of the oligosaccharides were reduced with sodium borohydride. After workup they were hydrolyzed with 2M TFA, and converted into the peracetylated aldononitriles^5 (see Table V.5). Methylations of the oligosaccharides, and the sodium borohydride reduced oligosaccharides were performed according to the method of Hakomori. 1 1 5 Methylated oligosaccharides Pla and P2 were reacted with lithium aluminum hydride in anhydrous oxolane to reduce the uronic esters. These compounds were all hydrolyzed with 2M TFA for 18 h, and converted into the permethylated alditol acetates. The results obtained are presented in Table V.6. -180-VI. BIBLIOGRAPHY - 181 -VI. BIBLIOGRAPHY 1. G.O. Aspinall, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 1, Academic Press, New York, (1982), pp. 1-18. 2. P.A. Sandford and J. Baird, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 2, Academic Press, New York, (1983), pp. 411-490. 3. W.F. Dudman, in I. Sutherland (Ed.), "Surface Carbohydrates of the Prokaryotic Cell", Academic Press, London, (1977), pp. 357-414. 4. K. Jann and O. Westphal, in M. Sela (Ed.), "The Antigens", Vol. Ill, Academic Press, New York, (1975), pp. 1-125. 5. P.R. Edwards and W.H. Ewing, "Identification of Enterobacteriaceae". 3rd edn., Burgess Publishing Company, Minneapolis, (1972). 6. F. Kaufmann, "The Bacteriology of Enterobacteriaceae". Munksgaard, Copenhagen, (1966). 7. O. Luderitz, M.A. Freudenberg, C. Galanos, V. Lehmann, E. Th. Rietschel and D.H. Shaw, Current Topics in Membranes and Transport, 17 (1982) 79-151. 8. I. 0rskov, F. 0rskov, B. Jann and K. Jann, Bacteriol. Rev., 41 (1977) 667-710. 9. O. Westphal, K. Jann and K. Himmelspach, Prog. Allergy, 33 (1983) 9-39. 10. A.R. Dochez and O.T. Avery, J. Exp. Med., 26 (1917) 477-493. 11. M. Heidelberger and O.T. Avery, J. Exp. Med., 38 (1923) 73-79. 12. M. Heidelberger and O.T. Avery, J. Exp. Med., 38 (1923) 81-85. 13. M. Heidelberger and O.T. Avery, J. Exp. Med., 40 (1924) 301-316. 14. M. Heidelberger and F.E. Kendall, J. Exp. Med., 50 (1929) 809-823. 15. M. Heidelberger, Res. Immunochem. Immunobiol., 3 (1973) 1-40. 16. M. Heidelberger, CM. MacLeod, S.J. Kaiser and B. Robinson, J. Exp. Med., 83 (1946) 303-320. 17. M. Heidelberger, CM. MacLeod, R.G. Hodges, W.G. Bernhard and M.M. di Lapi, J. Exp. Med., 85 (1947) 227-230. 18. E.A. Kabat, J. Immunol., 84 (1960) 82-85. 19. K. Jann and B. Jann, in I. Sutherland (Ed.), "Surface Carbohydrates of the Prokaryotic Cell", Academic Press, London, (1977), pp. 247-287. - 182 -20. M. Heidelberger, Prog. Chem. Org. Nat. Prod., 18 (1960) 503-536. 21. M. Heidelberger, W.F. Dudman and W. Nimmich, J. Immunol., 104 (1970) 1321-1328. 22. L.G. Bennett and C.T. Bishop, Immunochemistry, 14 (1977) 693-696. 23. M. Heidelberger and G.G.S. Dutton, J. Immunol., I l l (1973) 857-859. 24. M. Heidelberger and W. Nimmich, Immunochemistry, 13 (1976) 67-80. 25. a) M. Heidelberger, W. Nimmich, J. Eriksen and S. Stirm, Acta Path. Microbiol. Scand. Sect. B, 86 (1978) 313-320. b) M. Heidelberger, W. Nimmich, J. Eriksen, G.G.S. Dutton, S. Stirm and C.T. Fang, Acta Path. Microbiol. Scand. Sect. C, 83 (1975) 397-405. 26. W. Nimmich, Z. Allg. Microbiol., 19 (5) (1979) 343-347. 27. E. Altman, Ph.D. Thesis, University of British Columbia, (1984). 28. C.T. Bishop and H.J. Jennings, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 1, Academic Press, New York, (1982), pp. 291-330. 29. H.J. Jennings, Adv. Carbohydr. Chem. Biochem., 41 (1983) 155-208. 30. M. Heidelberger, CM. MacLeod and M.M. di Lapi, J. Exp. Med., 88 (1948) 369-372. 31. P.J. Baker and P.W. Stashak, J. Immunol., 103 (1969) 1342-1348. 32. C.P.J. Glaudemans, Adv. Carbohydr. Chem. Biochem., 31 (1975) 313-346. 33. H. Geyer, K. Himmelspach, B. Kwiatkowski, S. Schlecht and S. Stirm, Pure Appl. Chem., 55 (1983) 637-653. 34. L. Kenne and B. Lindberg, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 2, Academic Press, New York, (1983), pp. 287-363. 35. H.W. Leach and T.J. Schoch, Cereal Chem., 39 (1962) 318-327. 36. J.-P. Joseleau, G. Chambat and B. Chumpitazi-Hermoza, Carbohydr. Res., 90 (1981) 339-344. 37. O. Westphal and K. Jann, Methods Carbohydr. Chem., 5 (1965) 83-91. 38. R.L. Whistler and J.L. Sannella, Methods Carbohydr. Chem., 5 (1965) 34-36. 39. T.J. Painter, Methods Carbohydr. Chem., 5 (1965) 98-100. 40. G.O. Aspinall, J.A. Molloy and J.W.T. Craig, Can. J. Biochem., 47 (1969) 1063-1070. - 183 -41. J.E. Scott, Methods Carbohydr. Chem., 5 (1965) 38-44. 42. R.W. Jeanloz, Methods Carbohydr. Chem., 5 (1965) 110-114. 43. A.O. Pittet, Methods Carbohydr. Chem., 5 (1965) 3-5. 44. T.E. Edwards, Methods Carbohydr. Chem., 5 (1965) 176-179. 45. H. Meier, Methods Carbohydr. Chem., 5 (1965) 45-46. 46. K.A. Granath, Methods Carbohydr. Chem., 5 (1965) 20-28. 47. Y.C. Lee and R. Montgomery, Methods Carbohydr. Chem., 5 (1965) 28-34. 48. H. Neukon and W. Kuendig, Methods Carbohydr. Chem., 5 (1965) 14-17. 49. R.C.E. Guy, M.J. How, M. Stacey and M. Heidelberger, J. Biol. Chem., 242 (1967) 5.106-5111. 50. P. Branefors-Helander, B. Classon, L. Kenne and B. Lindberg, Carbohydr. Res., 76 (1979) 197-202. 51. I.J. Goldstein and C.E. Hayes, Adv. Carbohydr. Chem. Biochem., 35 (1978) 127-130. 52. S. Narasimhan, J.R. Wilson, E. Martin and H. Schachter, Can. J. Biochem. 57 (1979) 83-96. 53. D.H. Northcote, Methods Carbohydr. Chem., 5 (1965) 49-53. 54. M. Breen, H.G. Weinstein, L.J. Black, M.S. Borcherding and R.A. Sittig, Methods Carbohydr. Chem., 7 (1966) 101-115. 55. S.C. Churms, Adv. Carbohydr. Chem. Biochem., 25 (1970) 13-51. 56. W. Nimmich, Z. Med. Mikrobiol. Immunol., 154 (1968) 117-131. 57. K. Okutani and G.G.S. Dutton, Carbohydr. Res., 86 (1980) 259-271. 58. K. Macek, in E. Heftman (Ed.), "Chromatography", 2nd edn., Reinhold Publishing Corporation, New York, (1967), pp. 139-164. 59. I. Smith and J.G. Feinberg, "Paper and Thin Layer Chromatography and Electrophoresis", 2nd edn., Longman, (1972). 60. F. Percheron, in E. Heftman (Ed.), "Chromatography", 2nd edn., Reinhold Publishing Corporation, New York (1967), pp. 573-605. 61. A.B. Foster, Adv. Carbohydr. Chem., 12 (1957) 81-115. 62. H. Weigel, Adv. Carbohydr. Chem., 18 (1963) 61-97. - 184 -63. R.J. Bock, E.L. Durrum and G. Zweig, "A Manual of Paper Chromatography and Paper Electrophoresis", Academic Press Inc., New York, N.Y., (1955). 64. G.G.S. Dutton, Adv. Carbohydr. Chem. Biochem., 28 (1973) 11-160. 65. C.T. Bishop, Adv. Carbohydr. Chem. 19 (1964) 95-147. 66. A.G. Mclnnes, D.H. Ball, F.P. Cooper and C.T. Bishop, J. Chromatogr., 1 (1958) 556-557. 67. C.C. Sweely, R. Bentley, M. Makita and W.W. Wells, J. Am. Chem. Soc, 85 (1963) 2497-2507. 68. J.S. Sawardeker, J.H. Sloneker and A.R. Jeanes, Anal. Chem., 37 (1965) 1602-1604. 69. S.W. Gunner, J.K.N. Jones and M.B. Perry, Can. J. Chem., 39 (1961) 1892-1899. 70. H. Bjorndal, C.G. Hellerqvist, B. Lindberg and S. Svensson, Angew. Chem. Int. Ed. Engl., 9 (1970) 610-619. 71. J.S. Sawardeker and J.H. Sloneker, Anal. Chem., 37 (1965) 945-947. 72. W.F. Lehnhardt and R.J. Winzler, J. Chromatogr., 34 (1968) 471-479. 73. V.N. Reinhold, Methods Enzymol., 25 (1972) 244-249. 74. A. Pryde and M.T. Gilbert, "Applications of High Performance Liquid Chromatography", Chapman and Hall Ltd., London, (1979). 75. B.S. Valent, A.G. Darvill, M. McNeil, B.K. Robertsen and P. Albersheim, Carbohydr. Res., 79 (1980) 165-192. o 76. M. McNeil, A.G. Darvill, P. Aman, L.-E. Franzen and P. Albersheim, Methods Enzymol., 83 (1982) 3-45. 77. K.H. Altgelt, Adv. Chromatogr., 7 (1968) 3-46. 78. S.A. Barker, B.W. Hatt, J.F. Kennedy and P.J. Somers, Carbohydr. Res., 9 (1969) 327-334. 79. P.J. Garegg and B. Lindberg, Acta Chem. Scand., 14 (1960) 871-876. 80. P. Albersheim, D.J. Nevins, P.D. English and A. Karr, Carbohydr. Res., 5 (1967) 340-345. 81. C.G. Hellerqvist and A.A. Lindberg, Carbohydr. Res., 16 (1971) 39-48. 82. C.G. Hellerqvist, B. Lindberg, S. Svennson, T. Holme and A.A. Lindberg, Carbohydr. Res., 8 (1968) 43-55. - 185 -83. H.J. Jennings, K.-G. Rosell, E. Katzenellenbogen and D.L. Kasper, J. Biol. Chem., 258 (1983) 1793-1798. 84. A.J. Mort and D.T.A. Lamport, Anal. Biochem., 82 (1977) 289-309. 85. H.J. Jennings and C. Lugowski, Can. J. Chem., 58 (1980) 2610-2612. 86. R.G. Spiro, Methods Enzymol., 28 (1973) 3-43. 87. J.F. Codington, K.B. Linsley and C. Silber, Methods Carbohydr. Chem., 7 (1976) 226-232. 88. S. Inoue, G. Matsumura and Y. Inoue, Anal. Biochem., 125 (1982) 118-124. 89. G.G.S. Dutton and M.-T. Yang, Can. J. Chem., 51 (1973) 1826-1832. 90. Y.C. Lee and C.E. Ballou, Biochemistry, 4 (1965) 287-298. 91. K. Stellner, H. Saito and S.-I. Hakomori, Arch. Biochem. Biophys., 155 (1973) 464-472. 92. R.E. Wing and J.N. BeMiller, Methods Carbohydr. Chem., 6 (1972) 42-59. 93. G.G.S. Dutton, Adv. Carbohydr. Chem. Biochem., 30 (1974) 9-110. 94. B.A. Dmitriev, L.V. Backinowsky, V.L. Lvov, N.K. Kochetkov and I.L. Hofman, Eur. J. Biochem., 50 (1975) 539-547. 95. LA. Morrison, J. Chromatogr., 108 (1975) 361-364. 96. Z. Dische, Methods Carbohydr. Chem., 1 (1962) 477-514. 97. D. Aminoff, W.W. Binkley, R. Schaffer and R.W. Mowry, in W. Pigman and D. Horton (Eds.), "The Carbohydrates", 2nd edn., Vol. 2B, Academic Press, New York, (1970), pp. 739-807. 98. G.M. Bebault, J.M. Berry, Y.-M. Choy, G.G.S. Dutton, N. Funnel, L.D. Hayward and A.M. Stephen, Can. J. Chem., 51 (1973) 324-326. 99. J.H. Pazur, Methods Enzymol., 9 (1966) 82-87. 100. D. Amaral, F. Kelly-Falcoz and B.L. Horecker, Methods. Enzymol., 9 (1966) 87-92. 101. G.J. Gerwig, J.P. Kamerling and J.F.G. Vliegenthart, Carbohydr. Res., 77 (1979) 1-7. 102. K. Leontin, B. Lindberg and J. Lb'nngren, Carbohydr. Res., 62 (1978) 359-362. 103. W.A. Konig, I. Benecke and H. Bretting, Angew. Chem., Int. Ed. Engl., 20 (1981) 693-694. - 186 -104. M.R. Little, Carbohydr. Res., 105 (1982) 1-8. 105. R.L. Taylor and H.E. Conrad, Biochemistry, 11 (1972) 1382-1388. 106. M. Abdel-Akher and F. Smith, Nature (London), 166 (1950) 1037-1038. 107. J.K.N. Jones and M.B. Perry, J. Am. Chem. Soc, 79 (1957) 2787-2793. 108. F. Smith and A.M. Stephen, Tetrahedron Lett. 7 (1960) 17-23. 109. B. Lindberg, Methods Enzymol., 28 (1972) 178-195. 110. E.L. Hirst and E. Percival, Methods Carbohydr. Chem., 5 (1965) 287-296. 111. P.-E. Jannsson, L. Kenne, H. Liedgren, B. Lindberg and J. Lonngren, Chem. Commun., Univ. Stockholm, 8 (1976). 112. W.N. Haworth, J. Chem. Soc, 107 (1915) 8-16. 113. T. Purdie and J.C. Irvine, J. Chem. Soc, 83 (1903) 1021-1037. 114. R. Kuhn, H. Trischmann and I. Low, Angew. Chem., 67 (1955) 32. 115. S.-I. Hakomori, J. Biochem. (Tokyo), 55 (1964) 205-208. 116. P.A. Sandford and H.E. Conrad, Biochemistry, 5 (1966) 1508-1516. 117. D.P. Sweet, P. Albersheim and R.H. Shapiro, Carbohydr. Res., 40 (1975) 199-216. 118. B.K. Robertsen, P. Aman, A.G. Darvill, M. McNeil and P. Albersheim, Plant Physiol., 67 (1981) 389-400. 119. L.R. Phillips and B.A. Fraser, Carbohydr. Res., 90 (1981) 149-152. 120. P. Prehm, Carbohydr. Res., 78 (1980) 372-374. 121. G.G.S. Dutton and R.H. Walker, Cellulose Chem. Technol., 6 (1972) 295-305. 122. D.P. Sweet, R.H, Shapiro and P. Albersheim, Carbohydr. Res., 40 (1975) 217-225. 123. R. Geyer, H. Geyer, S. Kiihnhardt, W. Mink and S. Stirm, Anal. Biochem., 121 (1982) 263-274. 124. C.G. Wong, S.-S.J. Sung and C.C. Sweeley, Methods Carbohydr. Chem., 8 (1980) 55-66. 125. W.F. Dudman and CP. Whittle, Carbohydr. Res., 46 (1976) 267-272. 126. D. Horton, J.D. Wander and R.L. Foltz, Carbohydr. Res., 36 (1974) 75-96. - 187 -127. O.S. Chizhov, V.I. Kadentsev, A.A. Solov'yov, P.F. Levonowich and R.C. Dougherty, J. Org. Chem., 41 (1976) 3425-3428. 128. H.D. Beckey and H.R. Schulten, Angew. Chem. Int. Ed. Engl., 14 (1975) 403-415. 129. H. Krone and H.D. Beckey, Org. Mass. Spectrom., 2 (1969) 427-429. 130. A. Dell, H.R. Morris, H. Egge and H. von Strecker, Carbohydr. Res., 115 (1983) 41-52. 131. A. Dell and C.E. Ballou, Carbohydr. Res., 120 (1983) 95-111. 132. O.S. Chizhov, N.K. Kochetkov, N.N. Malysheva, A.J. Shiyonok and Y.L. Chashchin, Org. Mass Spectrom., 5 (1971) 1157-1167. 133. K. Biemann, D.C. De Jongh and H.K. Schnoes, J. Am. Chem. Soc, 85 (1963) 1763-1771. 134. J.P. Kamerling, J.F.G. Vliegenthart, J. Vink and J.J. De Ridder, Tetrahedron, 27 (1971) 4749-4757. 135. J. Karkkainen, Carbohydr. Res., 17 (1971) 1-10. 136. J. Karkkainen, Carbohydr. Res., 17 (1971) 11-18. 137. J. Karkkainen, Carbohydr. Res., 14 (1970) 27-33. 138. L.S. Forsberg, A. Dell, D.J. Walton and C.E. Ballou, J. Biol. Chem., 257 (1982) 3355-3363. 139. N.K. Kochetkov and O.S. Chizhov, Adv. Carbohydr. Chem. Biochem., 21 (1966) 39-93. 140. D.C. De Jongh, in W. Pigman and D. Horton (Eds.), "The Carbohydrates", 2nd edn., Vol. IB, Academic Press, New York, (1981), pp. 1327-1353. 141. N.K. Kochetkov and O.S. Chizhov, Methods Carbohydr. Chem., 6 (1972) 540-554. 142. H. Bjorndal, B. Lindberg and S. Svensson, Carbohydr. Res., 5 (1967) 433-440. 143. J. Lonngren and S. Svensson, Adv. Carbohydr. Chem. Biochem., 29 (1974) 41-106. 144. K. Miyano, M. Ishibashi, N. Kunita, Y. Takeda and T. Miwatani, F.E.M.S. Microbiol. Lett., 14 (1982) 145-148. 145. J.H. Banoub, Carbohydr. Res., 100 (1982) C17-C23. 146. D.P. Sweet, R.H. Shapiro and P. Albersheim, Biomed. Mass Spectrom., 1 (1974) 263-268. - 188 -147. B.A. Dmitriev, L.V. Backinowsky, O.S. Chizhov, B.M. Zoltarev and N.K. Kochetkov, Carbohydr. Res., 19 (1971) 432-435. 148. C. Erbing, B. Lindberg and S. Svensson, Acta Chem. Scand., 27 (1973) 3699-3704. 149. B. Lindberg and J. Lbnngren, Methods Enzymol., 50 (1978) 3-33. 150. K.-I. Harada, S. Ito, N. Takeda, M. Suzuki and A. Takematsu, Biomed. Mass Spectrom., 10 (1983) 5-12. 151. V. KovaCik, S. Bauer, J. Rosik, and P. Kovlc, Carbohydr. Res., 8 (1968) 282-290. 152. V. Kovacik, S. Bauer and J. Rosik, Carbohydr. Res., 8 (1968) 291-294. 153. J.N. BeMiller, Adv. Carbohydr. Chem. Biochem., 22 (1967) 25-108. 154. B. Capon, Chem. Rev., 69 (1969) 407-498. 155. R.D. Guthrie and J.F. McCarthy, Adv. Carbohydr. Chem. Biochem., 22 (1967) 11-23. 156. B. Nilsson and S. Svensson, Carbohydr. Res., 65 (1978) 169-171. 157. B. Nilsson and S. Svensson, Carbohydr. Res., 72 (1979) 183-190. 158. A.J. Mort and W.D. Bauer, J. Biol. Chem., 257 (1982) 1870-1875. 159. A.J. Mort, J.-P. Utille, G. Torri and A.S. Perlin, Carbohydr. Res., 121 (1983) 221-232. 160. Y.A. Knirel, E.V. Vinogradov, A.S. Shashkov, B.A. Dmitriev and N.K. Kochetkov, Carbohydr. Res., 112 (1983) C4-C6. 161. H.J. Jennings and C. Lugowski, Can. J. Chem., 58 (1980) 2610-2612. 162. H.O. Bouveng and B. Lindberg, Adv. Carbohydr. Chem., 15 (1960) 53-89. 163. J.M. Bobbin, Adv. Carbohydr. Chem., 11 (1956) 1-41. 164. A.S. Perlin, Adv. Carbohydr. Chem., 14 (1959) 9-61. 165. G.W. Hay, B.A. Lewis and F. Smith, Methods Carbohydr. Chem. 5 (1965) 357-361. 166. G.W. Hay, B.A. Lewis and F. Smith, Methods Carbohydr. Chem. 5 (1965) 377-380. 167. K.M. Aalmo, M.F. Ishak and T.J. Painter, Carbohydr. Res., 63 (1978) C3-C7. - 189 -168. S. Ebisu, J. Lonngren and I.J. Goldstein, Carbohydr. Res., 58 (1977) 187-191. 169. A.S. Perlin, in W. Pigman and D. Horton (Eds.), "The Carbohydrates", 2nd edn., Vol. IB, Academic Press, New York, (1981), pp. 1167-1215. 170. T.J. Painter and B. Larsen, Acta Chem. Scand., 24 (1970) 813-833. 171. R.J. Harbinson and J.E. Scott, Histochemie, 14 (1968) 215-220. 172. J.E. Scott and M.J. Tigwell, Biochem. J., 173 (1978) 103-114. 173. L.-A. Fransson, Carbohydr. Res., 62 (1978) 235-244. 174. I.J. Goldstein, G.W. Hay, B.A. Lewis and F. Smith, Methods Carbohydr. Chem., 5 (1965) 361-370. 175. P.A.J. Gorin and J.F.T. Spencer, Can. J. Chem., 43 (1965) 2978-2984. 176. B. Lindberg, J. Lonngren, W. Nimmich and U. Ruden, Acta Chem. Scand., 27 (1973) 3787-3790. 177. J. Kiss, Adv. Carbohydr. Chem. Biochem., 29 (1974) 229-303. 178. B. Lindberg and J. Lonngren, Methods Carbohydr. Chem., 7 (1976) 142-148. 179. B. Lindberg, J. Lonngren and S. Svensson, Adv. Carbohydr. Chem. Biochem., 31 (1975) 185-240. ' 180. B. Lindberg, J. Lonngren and J.L. Thompson, Carbohydr. Res., 28 (1973) 351-357. 181. G.O. Aspinall and K.-G. Rosell, Carbohydr. Res., 57 (1977) C23-C26. 182. G.O. Aspinall and A.S. Chaudhari, Can. J. Chem., 53 (1975) 2189-2193. 183. Z. Yosizawa, T. Sato and K. Schmid, Biochim. Biophys. Acta, 121 (1966) 417-420. 184. L. Kenne and B. Lindberg, Methods Carbohydr. Chem., 8 (1980) 295-296. 185. J.M. Williams, Adv. Carbohydr. Chem. Biochem., 31 (1975) 9-79. 186. J.D. Dutcher, D.R. Walters and O. Wintersteiner, J. Org. Chem., 28 (1963) 995-999. 187. B. Lindberg, B. Lindqvist, J. Lonngren and D.A. Powell, Carbohydr. Res., 78 (1980) 111-117. 188. L. Kenne, B. Lindberg, K. Petersson, E. Katzenellenbogen and E. Romanowska, Carbohydr. Res., 78 (1980) 119-126. - 190 -189. H.J. Jennings, C. Lugowski and N.M. Young, Biochemistry, 19 (1980) 4712-4719. 190. B. Lindberg, Chem. Soc. Rev., 10 (1981) 409-434. 191. R.J. Ferrier, in W. Pigman and D. Horton (Eds.), "The Carbohydrates", 2nd edn., Vol. IB, Academic Press, New York, (1980), pp. 1354-1375. 192. J.H. Van't Hoff, "The Arrangement of Atoms in Space", Longmans, Green and Co., London, (1898). 193. C.S. Hudson, J. Am. Chem. Soc, 31 (1909) 66-86. 194. E.H. Merrifield, Ph.D. Thesis, University of Cape Town, (1978). 195. R.U. Lemieux, R.K. Kullnig, H.J. Bernstein and W.G. Schneider, J. Am. Chem. Soc, 80 (1958) 6098-6105. 196. B. Coxon, Adv. Carbohydr. Chem. Biochem., 27 (1972) 7-83. 197. R.R. Ernst, Adv. Magn. Resonance, 2 (1966) 1-135. 198. L.D. Hall, Adv. Carbohydr. Chem. Biochem., 29 (1974) 11-40. 199. A.S. Perlin and B. Casu, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 1, Academic Press, New York, (1982), pp. 133-193. 200. J.F.G. Vliegenthart, L. Dorland and H. van Halbeek, Adv. Carbohydr. Chem. Biochem., 41 (1983) 209-374. 201. R.R. Ernst and K. Wuetrich, Biochem. Biophys. Res! Commun., 96 (1980) 1156-1163. 202. L.D. Hall, G.A. Morris and S. Sukumar, J. Am. Chem. Soc, 102 (1980) 1745-1747. 203. G. Kotowycz and R.U. Lemieux, Chem. Rev., 73 (1973) 669-698. 204. M. Karplus, J. Chem. Phys., 30 (1959) 11-15. 205. L.D. Hall, in W. Pigman and D. Horton (Eds.), "The Carbohydrates", 2nd edn., Vol. IB, Academic Press, New York, (1980), pp. 1299-1326. 206. L.D. Hall, Adv. Carbohydr. Chem., 19 (1964) 51-93. 207. P.J. Garegg, P.-E. Jansson, B. Lindberg, F. Lindh, J. Lonngren, I. Kvarnstrom, and W. Nimmich, Carbohydr. Res., 78 (1980) 127-132. 208. K. Bock and C. Pedersen, Adv. Carbohydr. Chem. Biochem., 41 (1983) 27-66. 209. H.J. Jennings and I.C.P. Smith, Methods Enzymol., 50 (1978) 39-50. - 191 -210. P.A.J. Gorin, Adv. Carbohydr. Chem. Biochem., 38 (1981) 13-104. 211. J.-P. Joseleau, F. Michon and M. Vignon, Carbohydr. Res., 101 (1982) 172-185. 212. K. Bock and C. Pedersen, J. Chem. Soc. Perkin Trans., 2 (1974) 293-297, and Acta Chem. Scand., Ser. B, 29 (1975) 258-264. 213. J.-P. Joseleau, M. Lapeyre, M. Vignon and G.G.S. Dutton, Carbohydr. Res., 67 (1978) 197-212. 214. P.A.J. Gorin and M. Mazurek, Can. J. Chem., 53 (1975) 1212-1223. 215. R.G.S. Ritchie, N. Cyr, B. Korsch, H.J. Koch and A.S. Perlin, Can. J. Chem., 53 (1975) 1424-1433. 216. A.K. Bhattacharjee, H.J. Jennings and CP. Kenny, Biochemistry, 17 (1978) 645-651. 217. H.J. Jennings and A.K. Bhattacharjee, Carbohydr. Res., 55 (1977) 105-112. 218. G.O. Aspinall, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 1, Academic Press, New York, (1982), pp. 35-131. 219. S.J. Angyal and K. James, Carbohydr. Res., 12 (1970) 147-149. 220. S.J. Angyal and K. James, Aust. J. Chem., 23 (1970) 1209-1221. 221. J. Hoffman, B. Lindberg and S. Svensson, Acta Chem. Scand., 26 (1972) 661-666. 222. J. Hoffman and B. Lindberg, Methods Carbohydr. Chem., 8 (1980) 117-122. 223. B. Lindberg, B. Lindqvist, J. Lonngren and W. Nimmich, Carbohydr. Res., 58 (1977) 443-451. 224. V.S. Rao and A.S. Perlin, Can. J. Chem., 61 (1983) 2688-2694. 225. F.-P. Tsui, R.A. Boykins and W. Egan, Carbohydr. Res., 102 (1982) 263-271. 226. I. Fromme and H. Beilhartz, Anal. Biochem., 84 (1978) 347-353. 227. E.A. Kabat and M.M. Mayer, "Experimental Immunochemistry", 2nd edn., C.C. Thomas, Springfield, II., (1961), pp. 493-495. 228. A.N. De Belder and B. Norrman, Carbohydr. Res., 8 (1968) 1-6. 229. G.O. Aspinall and R.J. Ferrier, Chem. and Ind. (London), (1957), 1216. 230. G.G.S. Dutton and M. Paulin, Carbohydr. Res., 87 (1980) 107-117. 231. G.M. Bebault, G.G.S. Dutton, N. Funnell and K.L. Mackie, Carbohydr. Res., 63 (1978) 183-192. - 192 -232. G.G.S. Dutton and K.L. Mackie, Carbohydr. Res., 62 (1978) 321-335. 233. Y.-M. Choy and G.G.S. Dutton, Can. J. Chem., 52 (1974) 684-687. 234. G.G.S. Dutton and M. Paulin, Carbohydr. Res., 87 (1980) 119-127. 235. J. Levy, J.J.R. Campbell and T.H. Blackburn, "Introductory Microbiology", John Wiley and Sons, Inc., N.Y., (1973). 236. F.W. Twort, Lancet, 2 (1915) 1241-1243. 237. F. D'Herelle, Compt. Rend. 165 (1917) 373-375. 238. J. Douglas, "Bacteriophages", Chapman and Hall, London, (1975). 239. D.E. Bradley, Bacteriol. Rev., 31 (1967) 230-314. 240. D. Rieger-Hug, S. Stirm, Virology, 113 (1981) 363-378. 241. C.K. Mathews, "Bacteriophage Biochemistry", Van Nostrand Reinhold Co., N.Y., (1971). 242. A.A. Lindberg, in I. Sutherland (Ed.), "Surface Carbohydrates of the Prokaryotic Cell", Academic Press, London, (1977), pp. 289-356. 243. B.H. Park, Virology, 2 (1956) 711-718. 244. M.E. Bayer, H. Thurow and M.H. Bayer, Virology, 94 (1979) 95-118. 245. W. Bessler, E. Freiind-Molbert, H. Knufermann, C. Rudolph, H. Thurow and S. Stirm, Virology, 56 (1973) 134-151. 246. G.G.S. Dutton, A.V. Savage and M. Vignon, Carbohydr. Res., 58 (1980) 2588-2591. 247. G.G.S. Dutton, K.L. Mackie, A.V. Savage, D. Rieger-Hug and S. Stirm, Carbohydr. Res., 84 (1980) 161-170. 248. G.G.S. Dutton, J.L. Di Fabio, D.M. Leek, E.H. Merrifield, J.R. Nunn and A.M. Stephen, Carbohydr. Res., 97 (1981) 127-138. 249. G.G.S. Dutton and E.H. Merrifield, Carbohydr. Res., 103 (1982) 107-128. 250. G.G.S. Dutton and E.H. Merrifield, Carbohydr. Res., 105 (1982) 189-203. 251. J.L. Di Fabio, G.G.S. Dutton and H. Parolis, Carbohydr. Res., 126 (1984) 261-269. 252. J.L. Di Fabio, G.G.S. Dutton and H. Parolis, Carbohydr. Res., 133 (1984) 125-133. 253. G.G.S. Dutton and T.E. Folkman, Carbohydr. Res., 78 (1980) 305-315. - 193 -254. H. Thurow, H. Niemann and S. Stirm, Carbohydr. Res., 41 (1975) 257-271. 255. J. Di Fabio and G.G.S. Dutton, Carbohydr. Res., 92 (1981) 287-298. 256. R.U. Lemieux and J.D. Stevens, Can. J. Chem., 44 (1966) 249-262. 257. G. Chambat, J.-P. Joseleau, M. Lapeyre and A. Lefebvre, Carbohydr. Res., 63 (1978) 323-326. -19 4-APPENDICES - 195 -APPENDIX I PREFACE Previous research in this laboratory has been concerned with the capsular polysaccharides of Klebsiella, but now that all of the known serotypes have been studied, the group is embarking on a parallel study of E. coli antigens. The polysaccharides of the former genus are composed of only neutral sugars and uronic acids. By contrast, the capsules of some of the strains of the latter genus are being found to include amino sugars. The main thrust of the work described in this appendix is to become familiar with the chemical behaviour of amino sugars and, in particular, to develop methods of liberating them quantitatively from polysaccharides. Publications on structural investigations carried out by other groups on some amino sugar containing polysaccharides were generally found to give inadequate information on the experimental conditions employed. Those giving the necessary information have been useful in this study. It must be emphasized, however, that the limited methodology published to-date cannot be randomly applied, and needs modifications which depend on the nature of the amino sugars in the poly-saccharide. We hope that this work will be beneficial to the others in our laboratory who will be embarking on structural studies of amino sugar-containing polysaccharides. - 196 -PRELIMINARY INVESTIGATION LEADING TO STRUCTURE ELUCIDATION OF AMINO SUGAR CONTAINING CAPSULAR POLYSACCHARIDES FROM ESCHERICHIA COLI Introduction The occurrence of various amino sugars in bacterial antigens has been observed in many Enterobacteriaceae. The most commonly found sugar is 2-acetamido-2-deoxyglucose, with the corresponding galacto- derivative being a close second. Other derivatives, such as 3-acetamido-3-deoxyhexoses and 4-acetamido-4-deoxyhexoses, have also been found as the gluco- and galacto-derivatives in LPS of certain Enterobacteriaceae. The 3-amino-3,6-dideoxyhexose found in the E. coli K45 polysaccharide has been known to occur in the O antigens of some E. coli as the gluco- and galacto-dcrivatives.1'2 Recently, aminohexuronic acid derivatives (e.g. 2,3-diacetamido-2,3-dideoxy-L-glucuronic acid from Pseudomonas aeruginosa LPS,3 2-acetamido-2-deoxy-D-mannuronic acid in Shigella sonnei4) have also been discovered in these antigens. Bacterial aminoglycans are very rarely found to contain free amino groups, the majority existing as the N-acyl (usually acetyl) derivatives. The free amine group most frequently encountered has been in 2-acetamido-4-amino-2,4,6-trideoxy-D- galac-tose. This sugar is found in Streptococcus pneumoniae Type 1 capsular poly-saccharide,^ the species specific cell wall teichoic acid (or C-substance) from S. pneumoniae.** and the O-specific side chains of the Shigella sonnei phase I LPS.7 The detection and quantitation of amino sugars would be expected to follow the same procedures as for neutral or acidic hexoses. In general qualitative analysis by acid hydrolysis is similar in both cases except when 2-acetamido-2-- 197 -deoxy-hexoses, or 3 or 4-acetamido-3 or 4,6-dideoxyhexoses are present. In the former case the 2-acetamido group is N-deacetylated by acid, thus making the glycosidic linkage of the free amino sugar resistant to hydrolysis. The latter type behaves in a manner similar to deoxy sugars, but are even more acid labile. Because they can be degraded easily they evade detection, as was in the case with the 4-acetamido-4,6-dideoxy galactose in Shigella sonnei phase I enterobacterial common antigen.4 Therefore each capsular polysaccharide should be treated individually in order to find the conditions that suit each particular case. The following section deals with the individual hydrolytic conditions of E. coli K antigens from E. coli K43, K44, and K45. Sugar analysis Samples of polysaccharide (5-10 mg) were hydrolyzed with different acids of varying concentrations for specific lengths of time. A part of the hydrolyzates was chromatographed in duplicate on Whatman No. 1 paper in each of solvents B and C (see Section III) for 16-18 h. The chromatograms were developed with AgN03/NaOH for the detection of all sugars, and with ninhydrin (see Section III) for the detection of amino sugars only. The other part of the hydrolyzate was reduced with NaBH^ neutralized with 50%acetic acid and evaporated to dryness [except in one case where the neutral and acidic sugars were separated from the amino sugars by eluting with water and then 1% HC1 on a column of IR-120 (H+)]. The dried (in vacuo) hydrolyzates were coevaporated with methanol and the resulting mixture of sugars was acetylated and extracted into chloroform. These alditol acetates were analyzed by capillary g.l.c. using program (a) (see Section III). From the results obtained (see Table A.l) it was deduced - 198 -that hydrochloric acid was more capable of hydrolyzing amino sugar linkages than was TFA. The amino sugars present in E. coli K44 are thus more resistant to TFA than HC1. There is no definite pattern of hydrolysis with 2M HC1, except that the best molar ratios were obtained for a reaction time of 2 h. The reason for this is not known. Treatment with 2M HC1 for elongated periods (48 h) led to degradation of the other neutral sugars (see Table A.]). Heating at 95° with 2M HC1 for 2-18 h was considered to be the optimum condition for hydrolysis of the K44 polysaccharide. The polysaccharide was found to consist of equimolar proportions of rhamnose, galactose, glucosamine, galactosamine, and glucuronic acid. E. coli K45 was found to contain galactose, glucosamine and a 3-amino-3,6-dideoxyhexose (confirmed by g.l.c.-m.s.) having RcicNH2 = m s o l v e n t Their respective molar proportions were 2:2:1 (confirmed from the methylation analysis - see following section). Mild conditions (0.1M HC1 for 2 h) were used to release the deoxy amino sugar. Glucosamine and galactose (probably linked to glucosamine) were not totally hydrolyzed by this treatment. The use of stronger conditions was necessary to release the glucosamine units which was achieved without considerable degradation of 3-amino-3,6-dideoxyhexose. In this case, TFA was not much different to hydrochloric acid; the use of either acid yielded approximately similar information about the ratios of galactose and 3-acetamido-3,6-dideoxyhexose. The glucosamine linkages were more resistant to TFA, while the 3-acetamido-3,6-dideoxyhexose was completely hydrolyzed by both acids. A similar situation exists in the hydrolysis of the core oligosaccharide of Aeromonas  hvdrophila (Chemotype III) LPS which consists of galactose, glucose and L-glycero-D-mannoheptose with 3-acetamido-3,6-dideoxy-L-glucose as the sole amino sugar; here quantitative molar proportions are obtained on hydrolysis - 199 -Table A.l Sugar analyses of E. coli K43, K44 and K45 polysaccharides E. coli K44 Mole Ratio polysaccharide 2M HC1 2M HC1 2M HC1 2M HC1 2M HC1 0.1M HC1 2M TFA 48 h 18 h 5h 3h 2h 3h 18 h Rhamnitol 0.21 1.00 1.00 1.00 1.00 0.12 0.89 Galactitol 0.32 0.70 0.86 0.68 1.14 1.00 1.00 Glucosaminitol 1.00 0.32 0.44 0.34 1.10 0.40 0.20 Galactosaminitol 0.91 0.31 0.44 0.33 0.97 0.44 0.20 E. coli K45 Mole Ratio polysaccharide 2M HC1 2M HC1 1M HC1 0.1M HC1 2M TFA 18 h 3 h 24 h 2 h 18 h 6-deoxy aminitol 1.00 0.65 1.00 1.00 1.00 galactitol 1.52 2.00 1.66 1.01 2.17 glucosaminitol 1.40 0.60 0.78 0.31 1.03 E. coli K43 Mole Ratio polysaccharide 2M HC1 10 h 4M HC1 4 h 2M TFA 18 h mannitol 1.90 1.20 1.38 galactitol 1.00 0.75 1.00 glucosaminitol 0.95 1.00 0.60 - 200 -with TFA. The use of HO, acetolysis and other conditions failed to give reasonable results.** The polysaccharide from E. coli K43 behaved differently in that the release of the amino sugar was relatively easier. This is probably due to the fact that glucosamine is terminal in a side chain,^ whereas in E. coli K44 the amino sugars are inchain (see later). The resistance of one of the glucosamine units of K45 to acid hydrolysis is attributed to it being a branch point sugar (see following section). The use of strong acid (4M HC1 for 4 h) resulted in the usual degradation of the neutral sugars, with good results being obtained on hydrolysis with 2M HC1 for 10 h. The use of TFA resulted in low levels of mannose being obtained, but the molar proportion of glucosamine released was considerably higher than in the other two polysaccharides due to the reason described earlier. This polysaccharide was found to contain galactose, mannose and glucosamine in the molar ratios of 1:2:1. Chemical Investigation The following two sections deal with the chemical methods used and results obtained in an attempt to elucidate the structures of E. coli K45 and K44 antigens. The data collected on the E. coli K43 polysaccharide is also included. - 202 -Chemical analysis of E. coli 08:K45:H9 (K45 antigen) Composition and n.m.r. spectra The isolation and purification were achieved as described in Section III.10.2. The purified product obtained from one Cetavlon precipitation had [«]rj>25 -1.7°. The presence of galactose, glucosamine and 3-amino-3,6-dideoxyhexose was observed by paper chromatography. Determination as the alditol acetates showed these sugars in the respective molar ratios of 2:2:1 (see preceding Section). The carboxyl-reduced polysaccharide1^ and the carboxyl-reduced methyl glycosides11 both failed to show either an increase in the existing molar ratios, or the presence of a new sugar. Hence E. coli K45 polysaccharide consists of a pentasaccharide repeating-unit of two moles of galactose, two moles of glucosamine and one mole of a 3-amino-3,6-dideoxyhexose, and does not possess an acidic sugar. Further confirmation by deamination of some hydrolyzates indicated the conversion of glucosamine to 2,5-anhydrbmannitol. The ^-n.m.r. spectrum of the polysaccharide showed five anomeric protons corresponding to 4 0 [ <5 4.64 b, (IH), 8 4.58 b, (IH), 8 4.48 b, (2H)] and one border line signal [ 8 4.98 b, (IH)]. There were no signals corresponding to pyruvic acetals, but a signal at 8 2.07 (9H) indicated the presence of three N-acetyl groups. The C-6 protons of the 6-deoxyamino sugar resonated at 8 1.27 (3H) (see spectrum No. Al). The anomeric signal in the 13C-n.m.r. spectrum at 104.54 p.p.m., 104.29 p.p.m. (2C), and 102.38 p.p.m. were assigned as /3 while the signal at 99.93 p.p.m. was probably due to an a linked sugar. The signal at 55.61 p.p.m. was due to the carbon atoms bonded to nitrogen, with the CH3 groups of the N-acetyl substituents showing signals at 23.33 p.p.m. and 22.86 p.p.m. The C-6 carbon of Spectrum No. A2 E. coli K45 polysaccharide 1JC-n.m.r. 100.6 MHz, amb. temp. - 204 -the deoxyamino sugar appeared at 16.23 p.p.m. (see Spectrum No. A2). These n.m.r. signals were not assigned to specific sugars as the data collected were inadequate. Methylation analysis Methylation of the E. coli K45 polysaccharide was accomplished according to the method of Hakomori.12 Hydrolysis of the methylated polymer into individual sugars resulted in degradation of some sugars and incomplete hydrolysis of others. Comparison of the data from different hydrolytic conditions enabled the detection of 2,3,6-tri-O-methylgalactose, 2,4,6-tri-O-methylgalactose, 3-N-methylacetamido-3,6-dideoxy-2,4-di-0-methylhexose, 2-N-methylacetamido-2-deoxy-2-N-methyl-4,6-di-O- methylglucose and 2-N-methylacetamido-4-0-methylglucose (see Table A.2). Thus E. coli K45 polysaccharide consists of 3- and 4-linked linear galactose units, a 3-linked glucosamine residue, and a 3,6-linked branched glucosamine, with the 3-amino-3,6-dideoxy sugar being terminal in the side chain. Periodate oxidation The native polymer was treated with NaIC>4 and left in the dark for 5 d at 4°. The reaction was quenched with ethylene glycol, reduced with NaBH4 into the polyol and dialyzed for 2 d. The sugar analysis of the polyol (2M HC1 for 18 h) as alditol acetates indicated the presence of galactose, glucosamine and 3-amino-3,6-dideoxyhexitol in the ratios of 1.15:2.03:1.00. These results are in agreement with the presence of one oxidizable 4-linked galactose residue. The - 205 -Table A.2 Methylation analysis of E. coli K45 polysaccharide Partially methylated alditol acetates Mole Ratio 2M TFA 18 h 2M HC1 2 h Acet I Acet II 2,3,6-Gal 0.13 - - 0.74 2,4,6-Gal 0.65 - - 0.66 2,4-3Amino 6-deoxyhex. 1.00 - 0.12 0.96 4,6-GlcNAc 0.70 - 1.00 ' 1.00 4-GlcNAc 0.22 - 0.44 0.26 Acet I = Acetolysis with 0.9M H 2S0 4 in 95% HOAc at 80° for 15h and water 5h. Acet II = Acetolysis with 0.25M H 2S0 4 in 95% HOAc at 80° for 6 h and water 3.5 h. other sugars due, to their linkages at position 3, and the terminal sugar which contains an N-acetyl group at position 3 are all periodate resistant. The n.m.r. spectra (^ H and 1 3C) also show a change in the structure by shifts of some anomeric signals (see Spectra Nos. A3, A4). Smith hydrolysis The polyol was treated with 0.5M TFA at room temperature for 24 h. After removal of TFA, the product was methylated.112 Acetolysis and conversion into partially methylated alditol acetates showed the absence of 2,3,6-tri-O-methyl-Spectrum No. A3 5 4 3 2 P-P-nn. Spectrum No. A4 Periodate oxidized E. coli K45 polysaccharide l3C-n.m.r. 100.6 MHz, amb. temp. - 208 -galactose, and the formation of 2-N-methylacetamido-3,4,6-tri-0-methylglucose, in addition to the original 2,4,6-tri-O-methylgalactose, 3-N-methylacetamido-3,6-di-deoxy-2,4-di-0- methylhexose, 2-N-methylacetamido-4-0-methylglucose, and 2-N-methylacetamido- 4,6-di-O-methylglucose in the ratios of 0.5:0.7:0.9:0.5:1.0. Such an observation may be explained by two possible structures. The first case where some of the terminal 6-deoxyamino sugars are cleaved on Smith hydrolysis, while some of the degraded 4-linked galactose units are not cleaved, would yield 3,4,6-tri-O-methyl-, and 4-O-methyl- 2-N-methylacetamidoglucose. If this was a 4<}al6J3-GlcNAcJ-^Gali-3-GlcNAc 13/6 I 6-deoxy-3-aminohexose probable structure, the formation of 2-N-methylacetamido-4,6-di-0-methyl- and 3,4-di-O-methyl- glucose would also be possible, with the former being indistinguishable from the original 3-linked glucosamine in the polymer. The absence of the latter may be proof of the invalidity of this proposed structure. In the second possibility, the 3-linked glucosamine is in the side chain (A) or main chain (B) linked at position 3 to the branch point sugar. By the same argument this would yield the terminal amino sugar by cleavage of the deoxy amino sugar (A), or by hydrolysis of the oxidized galactose residue (B). However only 2-N-methylacetamido-4,6-di-Q_-methylglucose is obtained by incomplete hydrolysis of the oxidized galactose unit (B), or cleavage of the deoxy sugar (B), or hydrolysis of the oxidized galactose residue (A). Thus one of A or B may be the probable structure of the E. coli K45 polysaccharide. - 209 -^ l c N A c i - ^ a l i - ^ a l -13 or ^lcNAc-i-^Gali-^ali-^-GlcNAc 16 GlcNAc 13 6-deoxy-3-aminohexose 6-deoxy-3-aminohexose A B Structural analysis of E. coli 08:K44 (A): HT (K44 antigen). Composition and n.m.r. spectra The polysaccharide from E. coli K44 was isolated and purified according to the procedure in Section III.10.2. The polysaccharide obtained from one Cetavlon precipitation had [a]r_>25 +50.4° which is indicative of predominantly 0 linkages. Acid hydrolyzates of the polysaccharide indicated the presence of galactose, rhamnose, glucuronic acid, galactosamine and/or glucosamine. Determination of the neutral and amino sugars as the alditol acetates gave equimolar proportions of galactose, rhamnose, glucosamine and galactosamine. The formation of equimolar proportions of 2,5-anhydromannitol and 2,5-anhydrotalitol on deamination of some hydrolyzates proved the presence of glucosamine and galactosamine in the polysaccharide. Both the carboxyl-reduced polysaccharide1^ and the carboxyl-reduced methyl glycosides,11 on analysis as the alditol acetates showed the presence of glucose (0.50 and 0.82 moles respectively). Thus E. coli K44 consists of a pentasaccharide repeating-unit of equimolar proportions of rhamnose, galactose, glucuronic acid, glucosamine and galactosamine. The ^-n.m.r. spectrum of the native polysaccharide (see Spectrum No. A5) Spectrum No. AS E. coli K44 polysaccharide H^-n.m.r. 400 MHz, 95°C Spectrum No. A6 Ei_coH K44 polysaccharide ^C-n.m.r. 100.6 MHz, amb. temp. - 212 -showed four signals corresponding to 4 /3 protons at 6 4.89 (1H), 8 4.72 (Jj 2 6 Hz, 1H) and 8 4.60 (b, 2H), and a borderline signal at 5 4.95 (1H). The C-6 protons of rhamnose appeared at 8 1.34 (3H). Two N-acetyl signals were observed at 8 2.09 (3H) and 8 2.06 (3H). There were no signals corresponding to pyruvic acetals. The 13C-n.m.r. spectrum showed three intense signals at 104.65 p.p.m., 101.56 p.p.m. and 98.93 p.p.m., and two weak signals at 102.23 p.p.m. and 102.16 p.p.m. (see Spectrum No. A6). The signals at 54.84 p.p.m. and 53.08 p.p.m. arose from C-2 of glucos- and galactos- amine. The CH3 carbon atoms of the N-acetyl groups were observed at 23.31 p.p.m. and 22.6 p.p.m., and that of C-6 of rhamnose at 17.54 p.p.m. Assignment of these signals to their respective sugars was not possible from the data collected. Methylation analysis E. coli K44 polysaccharide was methylated according to the Hakomori procedure.12 Hydrolysis of the methylated sugars was carried out under different conditions. Good proportions of the amino sugars were obtained on acetolysis (0.9M H2SO4 in 9 5 % HOAc, 80° for 15 h, dilution with water and further heating for 5 h) and reasonable ratios of the neutral sugars were found after treatment with 2M HC1 for 6 h. In many cases, complete demethylation of 2,4,6-tri-O-methyl-galactose into galactose occurred,13 the presence of which was detected by both g.l.c. and g.l.c.-m.s. as the alditol acetate. The derivatives detected by g.l.c. and identified by g.l.c.-m.s. were 2,4-di-O-methylrhamnose, 2,4,6-tri-O-methylgalactose, 2-N-methylacetamido-3,6-di-0-methyl- glucose and -galactose. - 213 -Analysis of the carboxyl-reduced polymer1^ showed in addition, the presence of 2,3,6-tri-O- methylglucose. This indicated that the uronic acid and the amino sugars were 4-linked while both rhamnose and galactose were 3-linked. There were no branch points or terminal sugars, which implied a linear pentasaccharide repeating-unit (see Table A.3). Table A.3 Methylation analysis of E. coli K44 polysaccharide Partially methylated alditol acetates Mole Ratio I II Ill 2,4-Rha 0.14 1.00 1.00 2,3,6-Glc - - 0.68 2,4,6-Gal .06 0.29 0.08 2,3,6-GlcNAc 1.00 0.39 0.76 2,3,6-GalNAc 0.75 0.26 0.35 Galactose - - 0.42 I Acetolysis with .9 M H 2S0 4 in 95% HOAc at 80° for 15 h and water 5 h. II 2M HC1 for 5 h. III Carbodiimide reduced polymer, 2M HC1 6 h. Periodate oxidation The acidic polysaccharide together with NaI0 4 was left stirring in the dark for 5 d at 4°. The product recovered was analyzed for neutral and amino sugars Spectrum No. A7 Periodate oxidized E. coli K44 polysaccharide 1 H-n.m.r. 400 MHz, 95°C Spectrum No. A 8 Periodate oxidized E. coli K 4 4 polysaccharide 13C-n.m.r. 100.6 MHz, amb. temp. - 216 -as the alditol acetates which revealed that those sugars were not consumed by periodate. The p.m.r. spectrum and 13c-n.m.r. spectrum of this material (see Spectra Nos. A7 and A8) were identical with those of the native polymer. Thus the 4-linked glucuronic acid unit, the only residue susceptible to oxidation, survived the conditions employed. This can be explained by the formation of hydrogen bonds between the N-acetyl group and a hydroxyl group of the uronic acid, thus rendering it periodate resistant (see Section II.5.2, Scheme 11.12). Deamination of the polysaccharide The polysaccharide was N-deacetylated with DMSO/H2O and NaOH, 1 4 and the removal of the N-acetyl substituent was verified by p.m.r. spectroscopy. The resulting polymer was deaminated with NaNG^/HOAc and the product was separated into oligomers on a column (2.5 x 25 cm) of Bio-Gel P2 irrigated with pyridine-acetic acid-water - 5:2:500. The elution profile obtained (see Fig. A.l) indicated the presence of two compounds. On analysis of the two compounds after hydrolysis and conversion into alditol acetates, negative results (no sugars present) were obtained, probably due to the very small amounts of material recovered. Base catalyzed uronic acid degradation The methylated native polymer was treated with dimethylsulfinyl anion for 18 h, and the product was methylated. Analysis of the sugars as the partially methylated alditol acetates revealed the presence of 2,4,6-tri-O-methylgalactose Fig A.l Elution profile of the products formed on deamination of the N-deacetylated E. coli K44 polysaccharide - 218 -and 2-N-methylacetamido-3,6-di-0-methylglucose. Replacement of 2,4-di-O-methylrhamnose by 2,3,4-tri-O-methylrhamnose indicated that rhamnose was linked at position 3 to glucuronic acid. The absence of the 2-N-methylacetamido-3,6-di-O-methylgalactose present in the native methylated polymer was probably due to it being the reducing sugar (liberated from the /3-elimination of the uronic acid), thus undergoing degradation. These results permit the assignment of the following sequence: ^GalNAc-LlGlcAi-^-Rha-Spectral analysis of E. coli 08:K43:H11 (K43 antigen) The polysaccharide from E. coli K43 was isolated as described in Section III.10.2. The *H- and 13C-n.m.r. spectra of the native polysaccharide were obtained and analyzed. *H-n.m.r. spectrum This spectrum showed many broad signals in the anomeric region (see Spectrum No. A9). There were no signals due to O-acetate or 6-deoxy sugars. The signal at 8 2.11 (3H) was assigned to the N-acetyl group of glucosamine. The signals at 6 5.41 (.6H) and 8 5.39 (.4H) were assigned to the branched mannose residue carrying the terminal glucosamine residue. The twinning of this signal was not explained. Signals at 8 4.79 (1H), 4.77 (1H) and 4.73 (1H) were due to glucosamine, galactose and mannose. These signals were not assigned. Spectrum No. A9 E. coli K43 polysacchai 1H-n.m.r. 400 MHz, 95°C Spectrum No. A10 E. coli K43 polysaccharide ^C-n.m.r. 100.6 MHz, amb. temp. - 221 -13C-n.m.r. spectrum The 13C-n.m.r. spectrum obtained, had four anomeric signals at 101.3, 101.5, 101.8 and 102.1 p.p.m. (see Spectrum No. A10). All the signals were obscured by a high noise level which was attributed to contamination of the polysaccharide with proteins during harvesting and purification. Signals were also seen in the ring region 69-80 p.p.m. and the primary hydroxyl region 60-62 p.p.m. The signal in the region 50-55 p.p.m. arising from C-2 of glucosamine was overshadowed by background noise signals, but the signal at 23.7 p.p.m. was assigned to the C H 3 group of N-acetylglucosamine. There were no signals due to C-6 of 6-deoxyhexoses. In conclusion, these spectra were useful in confirming the data obtained by sugar analysis of the K43 antigen, but, due to the poor quality of the spectra, were not useful in assignment of anomeric configurations. Conclusion Studies performed on these E. coli polysaccharides demonstrated that the presence of 2-acetamido-2-deoxyhexoses generally leads to complications in analysis of molar ratios of sugars present. It was concluded that TFA was useful for hydrolysis of glycosidic linkages of neutral and acidic sugars, but was not as effective toward 2-acetamido-2-deoxy sugars. The use of hydrochloric acid to effect hydrolysis of the latter sugars resulted in better molar ratios being obtained. However, there was evidence of degradation of some neutral sugars under certain conditions. This study indicated that each polysaccharide must be treated individually depending on the nature, position, and linkage of the amino-- 222 -containing sugar. Therefore general conditions for hydrolysis cannot be used for such polysaccharides as opposed to non-amino sugar-containing polymers. A similar situation was encountered in the hydrolysis of the methylated polymer, where the use of acetolysis or 2M HC1 for a short period was necessary in order to effect complete hydrolysis, and to prevent demethylation of neutral sugars. The preliminary investigations on the polysaccharides from E. coli K45 and K44 enabled partial structures to be assigned. Experimental General methods The instrumentation used for n.m.r., g.l.c, g.l.c.-m.s. and measurement of optical rotation has been described in Section III. Paper chromatography and ion-exchange chromatography were performed according to the procedure in Section III. Preparation and properties of E. coli K43, K44 and K45 capsular polysaccharides Cultures of Escherichia coli K43, K44 and K45 were obtained from Dr. I. 0rskov, Copenhagen and grown as described in Section III. 10.2. Purification of the polysaccharides was achieved by a single precipitation with Cetavlon. The yields obtained were E. coli K43 (478 mg from 3 trays), E. coli K44 (459 mg from 2 trays), and E. coli K45 (252 mg from 2 trays). The optical rotation of the - 223 -polysaccharide from E. coli K44 had |>]D^ 50.4° (c, 0.180, water), and that from E. coli K45 had [ « ] r j ) 2 5 -1.65° (c, .060, water). N.m.r. spectroscopy ( ]H and 1 3C) was performed on the original polysaccharides. All ^H-n.m.r. spectra were recorded at high temperature in order to obtain better resolution as well as to shift the HOD signal for observation of the /3 signals which were otherwise obscured. Hydrolysis of the polysaccharides Samples (5-10 mg) of the polysaccharides were hydrolyzed at 95° with 2M TFA (for 18 h), 4M HC1 (for 4 h), 2M HC1 (for 48 h, 18 h, 10 h, 5 h, 3 h, and 2 h), 1M HC1 (for 24 h), and 0.1M HC1 (for 3 h, and 2 h). The acid was removed by repeated coevaporation with water. Paper chromatography was performed on the hydrolyzates. Conversion of the hydrolyzates into alditol acetates was achieved by reduction with NaBH4 (1-2 h), acidification with 50% HOAc, evaporation to dryness, and repeated coevaporation with methanol followed by treatment with Ac20-pyridine (1:1) for 1 h at 95°. The mixture was extracted into chloroform, and traces of pyridine were removed by coevaporation with water (5 ml x 3). All g.l.c. analyses were performed on a capillary column using program (a) (see Section III.2). The polysaccharides from E. coli K44 and K45 were subjected to carboxyl reduction 1 1 by refluxing dried samples (11 and 12 mg respectively) with 3% HC1 in methanol (10 mL) overnight, neutralizing the HC1 with PbC0 3, removing PbCl2 by centrifugation and stirring the dried product in anhydrous methanol (10 mL) and NaBH4 (50 mg) for 18 h. The excess NaBH 4 was destroyed with 50% HOAc, and the dried product coevaporated with methanol (5 mL x 3). The methylglycosides obtained were hydrolyzed with 2M HC1 for 6 h at 95°. A - 224 -part of the hydrolyzate was reduced with NaBH4 (50mg for 2 h) and converted into alditol acetates. The rest was deaminated with 33% HOAc (1 mL) and 5% aqueous NaN02 (1 mL) for 1 h at 25°, diluted with water, and freeze-dried. The product was reduced with NaBD4 (50 mg) for 2 h, acidified with HOAc, and evaporated to dryness. Borate was removed by codistillation with methanol (7 mL x 2) and the resulting sugars were acetylated (see Section III.9). There was no evidence of the presence of a uronic acid residue in E. coli K45, but the presence of glucose in addition to the original sugars indicated that K44 contained glucuronic acid. The deaminated products from E. coli K45 contained only equimolar ratios of galactose and 2,5-anhydromannitol. The 3-amino-3,6-dideoxy-hexose was not detected on deamination (see Section II.5.4). The presence of rhamnose, galactose, glucose, 2,5-anhydromannitol, and 2,5-anhydrotalitol in the ratios 1.00:0.92:0.78:0.46:0.40 is further proof of a pentasaccharide repeating-unit in K.44. These results were confirmed by sugar analysis of the carboxyl-reduced polysaccharide obtained by carbodiimide reduction.^ Methylation analysis The dried polysaccharides in the native form (9-12 mg), and in the free acid form (15-20 mg) obtained by passing the sodium salt through a column of Amberlite IR-120 (H +) resin, were dissolved in dry dimethyl sulfoxide (2-3 mL), and methylated 1 2 by treatment with dimethylsulfinyl anion (2-3 mL) for 4 h under a nitrogen atmosphere, followed by methyl iodide (2 mL) for 1 h. The products, recovered after dialysis against tap water, showed complete methylation (no hydroxyl absorption in the i.r. spectra). The same procedure was performed on the carbodiimide-reduced polysaccharides. - 225 -Hydrolysis of the methylated polymers were performed with 2M HC1 for 2 h, 5 h, 6 h, and 2M TFA for 18 h at 95°. The conversion of the sugars into partially methylated alditol acetates (p.m.a.a.) was achieved by the same procedure described for unmethylated sugars. Acetolysis was carried out by adding 0.9M H 2S0 4, or 0.25M H 2S0 4, in 95% HOAc (1 mL) to the dried methylated product and heating at 80° for 15 h or 6 h respectively. The reaction mixture was diluted with H 20 (1 mL) and heated for a further 5 h or 3 h respectively. The hydrolyzate was passed through a column of AG3-X4A (acetate form) and washed with methanol. The eluate was evaporated to dryness under N 2, and the residue was dissolved in H 20 (1 mL) and reduced with NaBH 4 (50 mg) for 2 h. The alditols were converted into p.m.a.a.'s in the usual manner (Section III.7). All g.l.c. analyses were performed on a capillary column using program (b) (see Section III.2). G.l.c.-m.s. was performed as described in Section III.2. Uronic acid degradation A dried sample (20 mg) of methylated K44 polysaccharide, together with a trace of p_-toluenesulfonic acid, was dissolved in 19:1 dimethyl sulfoxide—2,2-dimethoxypropane (10 mL) and the flask was sealed under nitrogen. Dimethyl-sulfinyl anion (5 mL) was added and the reaction allowed to proceed at room temperature for 18 h. The solution was cooled and iodomethane (4 mL) was added to it. After stirring for 1 h, the methylated degraded product was isolated by partition between dichloromethane and water. The product was hydrolyzed with 2M HC1 for 5 h at 95°, and the p.m.a.a.'s were prepared as in Section III.7. G.l.c. and g.l.c.-m.s. analyses were performed as in Section III.2 using capillary columns and program (b). - 226 -N-Deacetylation14 of the polysaccharide was performed according to the procedure described in Section III.8. Deamination was achieved by following the procedure outlined in Section III.9. Periodate oxidation To each of the polysaccharides of E. coli K44 (48 mg) and E. coli K45 (41 mg) in water (10.0 mL) was added 0.5M NaIC«4 (10.0 mL) and the solutions were stirred in the dark for 5 d at 4°. Ethylene glycol (0.5 mL) was added to each, and after 4 h, NaBH4 (150 mg) was added and the solutions were left stirring overnight. The excess NaBH4 was neutralized with 50% HOAc and the solutions were dialyzed overnight. Following freeze-drying of the polyol, the yields of product isolated were 40 mg from E. coli K44 and 31 mg from K45. Sugar analysis of the polyols (2M HC1 for 18 h at 95°) revealed that E. coli K45 contained 3-acetamido-3,6-di-deoxyhexose, galactose, and glucosamine in the ratios of 1.00:1.15:2.03, while E. coli K44 contained equimolar proportions of its five original sugars. All g.l.c. analyses were conducted on a capillary column using program (a) (see Section III.2). Smith degradation The polyol from E. coli K45 was heated with 0.5M TFA (10 mL) at room temperature for 24 h. After removal of acid by repeated coevaporations with water, the dried product was dissolved in dimethyl sulfoxide (2 mL) and methylated 1 2 by treating with dimsyl anion (1 mL) for 2 h, followed by addition - 227 -of iodomethane (1 mL) to the cooled solution. After stirring for 1 h, the methylated product was extracted into chloroform. The absence of hydroxyl absorption in the i.r. spectrum indicated completeness of the reaction. This product was subjected to acetolysis by treating with 0.25 M H 2 S O 4 in 95% HOAc (1 mL) for 4 h at 90°, followed by the addition of water (1 mL) and further heating for 2 h at 90°. The hydrolyzate was deionized on a column of AG3-X4A (acetate form) with methanol as eluant. Evaporation of the eluate to dryness yielded a residue which was converted to the p.m.a.a.'s by reduction and acetylation (see Section III.7). The compounds were analyzed by g.l.c. (program b), and g.l.c.-m.s. using capillary columns as described in Section III.2. - 228 -BIBLIOGRAPHY 1. L. Kenne and B. Lindberg, in G.O. Aspinall (Ed.), "The Polysaccharides", Vol. 2, Academic Press, New York, (1983), pp. 287-263. 2. V.L. L'vov, N.V. Tochtamysheva, A.S. Shashkov, B.A. Dmitriev and K. Capek, Carbohydr. Res., 112 (1983) 233-239. 3. Y.A. Knirel, E.V. Vinogradov, A.S. Shashkov, B.A. Dmitriev and N.K. Kochetkov, Carbohydr. Res., 112 (1983) C4-C6. 4. C. Lugowski, E. Romanowska, L. Kenne and B. Lindberg, Carbohydr. Res., 118 (1983) 173-181. 5. B. Lindberg, B. Lindqvist, J. Lonngren and D.A. Powell, Carbohydr. Res., 78 (1980) 111-117. 6. H.J. Jennings, C. Lugowski and N.M. Young, Biochemistry, 19 (1980) 4712-4719. 7. L. Kenne, B. Lindberg, K. Petersson, E. Katzenellenbogen and E. Romanowska, Carbohydr. Res., 78 (1980) 119-126. 8. J.H. Banoub and D.H. Shaw, Carbohydr. Res., 98 (1981) 93-103. 9. Y.-M. Choy, personal communication. 10. R.L. Taylor and H.E. Conrad, Biochemistry, 11 (1972) 1382-1388. 11. G.G.S. Dutton and M.-T. Yang, Can. J. Chem., 51 (1973) 1826-1832. 12. S.-I. Hakomori, J. Biochem. (Tokyo), 55 (1964) 205-208. 13. I. Croon, G. Herrstrom, G. Kull and B. Lindberg, Acta Chem. Scand., 14 (1960) 1338-1342. 14. L. Kenne and B. Lindberg, Methods Carbohydr. Chem., 8 (1980) 295-296. - 229 -APPENDIX II Qualitative analysis and chemotyping of Klebsie capsular polysaccharides. Cross-reactions between Klebsiella K antigens. The known structures of the Klebsiella capsular polysaccharides as of Feb. 1985. - 230 -(a) Klebsiella caDsular Dolvsaccharides (K1-K83). Qualitative analysis and chemotype grouping GlcA Gal Glc 8P, l l p , 15, 25, 27p, 51 GlcA Gal Man 20, 21p, 29 P 42 p 43, 66, 74 p GlcA Gal Rha 9, 9*, 47, 52, 81, 83 GlcA Glc Man 2, 4, 5P, 24 GlcA Glc Rha 17,23,44,45,71 GlcA Glc Fuc 1, 54 GlcA Gal Glc Man 7 P 10, 13P, 26p, 28, 30p, 31p, 46p, 50, 59, 60, 61, 62, 69 p GlcA Gal Glc Fuc 16, 58 p GlcA Gal Glc Rha 12P, 18, 19, 36P, 41, 55P, 70P, GlcA Gal Man Rha 40, 53, 80 P GlcA Glc Man Fuc 6 p GlcA Glc Man Rha 64 p 65 p GlcA Gal Glc Man Fuc 68 p GlcA Gal Glc Man Rha 14p, 67 GalA Gal Man 3P, 49, 57 GalA Glc Rha 34, 48 GalA Gal Fuc 63 PyrA Glc Rha 72 PyrA Gal Rha 32 PyrA Gal Glc Rha 56 KetoA Gal Glc 22, 37, 38 GlcA glucuronic acid Gal galactose GalA galacturonic acid Man mannose PyrA pyruvic acid Rha rhamnose KetoA rare uronic acid Fuc fucose Glc glucose P pyruvic acid present in addition Note: K9 and K9*- see Appendix 11(c). K68 -.Later found to contain GalA,Gal, Man and P only - see Appendix 11(c). - 231 -(b) Cross-reactions between Klebsiella K antigens Cross-reactions among Klebsiella antigens are numerous. However, some strong reactions which have been reported more than once are the following. K l and K6 K2 and K69 K2, K13 and K30 K3 and K68 K7 and K10 K l l and K21 K12, K29 and K42 K14 and K64 K18 and K44 K22 and K37 K24 and K43 K27 and K46 K70 and K72 - 232 -(c) The known structures of the Klebsiella capsular polysaccharides K-tvpe a <t>v Structure 0 K l 4 G i c A l _ 4 F u c j _ 3 G l c L 3 2 \ / P a 8 K2 3-Glci-4-ManJ-4<Jlci-" 3 8 £ GlcA K3 2ca 1 A i ^ M a n i ^ M a n i ^ G a l i -a 3 1 Man 6 4 \ / pyr K4 3 < j i c I 2 G l c A l _ 3 M a n J _ 3 G i c l _ a a a 8 K5 P d / \ 4 6 IcicAi-lGlci-^Mani-" 2 3 OAc K6 P (D / \ 4 6 iGici-iMani-^lcAi-^-Fuci-a a - 233 -Gal 1 1 3-G1 c Ai-2-Ma nJ-2-Ma n ±-h 8 a a P / \ 4 6 3Glcl 3 B GlcA 1 3 G i c l _ 3 G a i l _ 3 G a i J L 3 34 a GlcA 1 iGali^-Rhai-^Rhai-^Rha^-a a, a , a ' ? ? iGlcAi-^RhaJ-^Gal-L^Rhai-^Rhai-3 a a a a GlcA,Gal,Glc,Man (c) f N 4 Gal 1 IcicLlGlcAJ-^Gali-Gal 1 4 GlcA 1 6<3i C H R h al-3-Ga l ^ G a 1 f i -- 234 -3<ji c I _ i M a n JLIGI C J _ 1 GlcA 4 .4 1 Gal -3 Rha 1 3 -GlcA-i-^-Galf^Glc^Manl-B B B 2 B 1 Glc 6 4 \ / P GlcA,Gal,Glc (S.S)f Gal 1 3Gi cJ_4Gi CAl-4Fuci-Rha 1 a 3 ^icAJ-iRhai-iGlci-lRhai-a B i B a - 235 -K18 B 3 1 Rha 2 1 GlcA 4 a 1 Glc K19 GlcA,Gal,Glc,Rha (J) f K20 GlcA 1 3 Gal 1 + OAc 2-Manl^GalJ-® Gal K21 3 G a i J _ 3 G i c A l _ 3 Ma n-J—2-Man— ® 1 6 Glc 1 COOH K22 iGallAcici-- 236 -^-Rhai-^-Glc^-2 a 1 Glc 6 n 1 3 GlcA Man 11 iGicAi^Mani-^Mani-^Glcl-Glc 1 2| GlcA 1 ^ G a l i - ^ l c J -6 Gal 4 H 41 Glc 1 I 6 Glc 1 4 I ^GicAi-^ManJ-^Mani-^GalJ-- 237 -K27 Glc P 1 / \ 4 6 4 6_GI ci-iGl c i-3-Ga 1 i-^-Ga l i -3 3 a/3? 6 a/S? 1 GlcA K28 2^ ai]_3 M a ni_2ManiJGlc-L-a 2 1 3 GlcA 3 1 Glc K29 GlcA,Gal,Man,pyr (N) f K30' Gal OAc 4QI c i _ 4 M a n J-^Ma ni-3 3 1 GlcA K31 pr G I C ^4 ! 2 Man 1 3 ^ i c l _ 3 G i c A i - l G a l i -- 238 -K32 P / \ 3 4 iGaiLlRhai-lRha-L-^RhaJ-K33 Gal 1 OAc 6 6 i M a n J-lMa nl-^Glci-3 1 GlcA Rha 1 K34 iRhai^Rhai-^-Glc^GalAi-^Rhai^ K35 ^-Gali-^-Mani-iMani-^-GlcL 4 6 a a 2. \ / ® P 1 GlcA K36 iGail^Rhai-iRhai-^RhaJ-1 GlcA 4 ^6 1 PCf Glc ^4 - 239 -K37 K38 ® 1 6 Glc 1 3G aiI_4Gi cl_ (X) 2 ^lcJ-^GalJ-^Gall 1 Glc CH3-C-O COOH OH (x) = COOH K39 GlcA,Gal,Glc,Man (D) f K40 GlcA,Gal,Man,Rha,Pyr (C) f K41 K42 G 1 6 Glc 1 a 4 GlcA 1 ^lci-^Rhai-^Gali-^Galfi-a a a GlcA,Gal,Man,pyr (N) f K43 GlcA,Gal,Man (N)1 - 240 -4GI CI_4GI cA i _ 2 R h aJ_iR ha -L-3-G1 c k |a 8 a a OAc (66%) GlcA 1 iGicJLlRhai^Rha^RhaJ-a' a' a' 7 3^ ail_343 ail_3^1cAl^ManI_ a p i a a Glc 1 D 3 P Man 1 O 4 P 3 G a i l _ 4 R h a l _ 3 a 1 B GlcA 4 a 1 Rha 3 Gi cl_^ Rh ai_4Gi c1^2RhaJ-2 a a j a 1 GalA OAc 2 o r 4GalA 1 3^ a li_2 M a nl_3_G aiJ_ a. a a - 241 -^ a i U c i c i - ^ l c A ^ M a n i - ^ M a n J -3 6 1 1 Glc 6 1 Gal GlcA 1 a 6 Glc 1 SGali-^Gal 1-a i 3<jail_2Rhal-4-GlcAi-3-Gali-4-RhaJ-2 1 Gal Rha 1 a 3 ^ j c A l ^ M a n L l M a n l l G a i L l R h a i -B a a 3 a Glc 1 + OAc i G i c I ^ i c A i - l F u c i -a a - 242 -OAc 2 3 a 1 Gal 3 a 1 GlcA / \ 4 6 SoicUGal^Gali-^alJ-2 a1 ? a 7 ' 1 Rha Man 1 a 4 IcaiUGaiAl^ManJ-a a OAc iG i c l ^ G i c A i - l F u c i -3 2 \ / P a 1 Gal GlcA 1 OAc OAc 3<3i c l_3_Gai i-lMa n J-3-Ma n-L-- 243 -K60 Glc 1 3^i cI_3 G l c Al_3_GalI- 3-Mani-6 2 a 2 1 8 1 Glc Glc K61 Gal 1 ^ l c Ai-2-Ma n J-^GlcJ-2GlcL-8 a 8 a 1 3r.ul 6r^i „1 K62 iGicAl^ManJ-^-GalJ-^lcJ-a 8 a a 1 Man K63 K64 3G ai Al^FucJ-^Gall-a a Rha 1 a + OAc 4-G1 c Ai-3-Ma n J-^Glci-^Ma ni-a a 8 2 a 6 1 P Glc 4 K65 K66 GlcA,Glc,Man,Rha,pyr (AMS)f GlcA,Gal,Man (N) f - 244 -K67 3_ M a ni-3-Ma ni-3-Glc I^-R h a i-a a 3i OI 1 G a l i - ^ l c A B 4 1 Rha K68J ^-Mani-^Gal^^GalAJ-4/2 a P. Man x 4 K69 GlcA,Glc,Gal,Man,pyr (I.S)f K70 K711 P 50% / \ 3 4 ^lcAi-^Rhai-^-Rhai-^GlcJ-^-Gal^Rha^-3 a a a "3 a -Rha—Rha—Rha—Rha—3-Glc-GlcA HI' t' ' Glc 1 1 K72 P / \ 3 4 IcicJLiRhai-lRhai-^RhaJ-8 a a a K731 i R h a i ^ G a l l ^ G l c J -8 3 P 8 GlcA - 245 -'^4 Gal 1 3 4 GlcA 1 a 3 2-Ga li^2-Ma ni-^Ma ni-a a E K68 E K46 E K39 = K15 -GicAi-^Rhai-^-Rhai-lRhai-^-GalJ-a a a 8 4 a 1 Glc 6 a 1 Glc - 246 -K80 3_Ga i I_2_Ma n J-2-Ma n i -3 3 1 GlcA 4 3 1 Rha 4 3 \ / P K81 iRharaRhal^QicAl^Rhal^Rhai^Gali-a a a a a K82 iGicIJoal-L^Gal-L P 4 P 0 1 1 GlcA K83 ^-Gal^Rhai-8 3 a 1 Gal 3 1 GlcA - 247 -Footnotes a Serotyping by J0rskov. Structure Ref. (a). Bacteriophage degradation Ref. (b). ' All sugars are D except for rhamnose and fucose which are L. Bacteriophage attack site. D. Rieger-Hug and S. Stirm, Virology, 113 (1981) 363-378. This serotype has been investigated by two laboratories, and two different structures have been proposed; denoted K9, K9. The other possible structure of K.9 is: ^lcAJ-^RhaJ-^-RhaJ-^-Gali-S-Rha-L /3 a a . a a Under investigation by AMS = A.M. Stephen C A.J. Chakraborty D G.G.S. Dutton IS I. Sutherland J J.P. Joseleau L B. Lindberg N W. Nimmich SS S. Stirm According to (d) above, the phage attack site is at glucose. Ref. K17(b) denotes rhamnose as the phage labile sugar. This structure is like K33 except OAc is present every other repeating-unit. Acetate on position 6 of Mannose, but not on all residues. According to the chemotyping, K68 was thought to consist of GlcA,Gal,Glc,Man and Fuc. Tentative structure. 1 K73 has been found to belong to Enterobacter aerogenes and is deleted from Klebsiella serotypes. I. £5rskov and M.A. Fife-Asbury, Int. J. Systematic Bacterid., 27 (1977) 386-387. - 248 -BIBLIOGRAPHY K l a) C. Erbing, L. Kenne, B. Lindberg, J. Lonngren and I.W. Sutherland, Carbohydr. Res., 50 (1976) 115-120. K2 a) L.C. Gahan, P.A. Sandford and H.E. Conrad, Biochemistry, 6, (1967) 2755-2767. b) H. Geyer, S. Stirm and K. Himmelspach, Med. Microbiol. Immunol., 165 (1979) 271-288. K3 a),b) G.G.S. Dutton, H. Parolis, J.-P. Joseleau and M.-F. Marais, personal communication. K4 a) E.H. Merrifield and A.M. Stephen, Carbohydr. Res., 96 (1981) 13-120. K5 a) i. G.G.S. Dutton and M.-T. Yang, Can. J. Chem., 50 (1972) 2382-2384. ii. G.G.S. Dutton and M.-T. Yang, Can. J. Chem., 51 (1973) 1826-1832. b) J.E.G. Van Dam, H. Snippe, M. Janze, J.M.N. Willers, H. van Halbeek, J.P. Kamerling and J.F.G. Vliegenthart, Xllth Int. Carbohydr. Symposium, Abstracts, (1984), Utrecht, The Netherlands. K6 a) U. ElsSsser-Beile, H. Friebolin and S. Stirm, Carbohydr. Res., 65 (1978) 245-249. K7 a) G.G.S. Dutton, A.M. Stephen and S.C. Churms, Carbohydr. Res., 38 (1974) 225-237. K8 a) I.W. Sutherland, Biochemistry, 9 (1970) 2180-2185. b) I.W.Sutherland, J. Gen. Microbiol., 94 (1976) 211-216. K9 a) B. Lindberg, J. LOnngren, J.L. Thompson and W. Nimmich, Carbohydr. Res., 25 (1972) 49-57. K9* a) S.C. Churms, E.H. Merrifield and A.M. Stephen, S. Afr. J. Sci., 76 (1980) 233-234. K l l a) H. Thurow, Y.-M. Choy, N. Frank, H. Niemann and S. Stirm, Carbohydr. Res., 41 (1975) 241-255. b) i. W. Bessler, E. Freiind-MMbert, J. Knufermann, C. Rudolph, H. Thurow and S. Stirm, Virology, 56 (1973) 134-151. ii. H. Thurow, H. Niemann and S. Stirm, Carbohydr. Res., 41 (1975) 257-271. K12 a) G.G.S. Dutton and A.V. Savage, Carbohydr. Res., 83 (1980) 351-362. b) G.G.S. Dutton and A.V. Savage, unpublished results. - 249 -H. Niemann, N. Frank and S. Stirm, Carbohydr. Res., 59 (1977) 165-177. H. Niemann, H. Beilharz and S. Stirm, Carbohydr. Res., 60 (1978) 353-366. G.G.S. Dutton, H. Parolis and L.A.S. Parolis, Carbohydr. Res., in press. G.G.S. Dutton and L.A.S. Parolis, unpublished results. A.J. Chakraborty, H. Friebolin, H. Niemann and S. Stirm, Carbohydr. Res., 59 (1977) 525-530. G.G.S. Dutton and T.E. Folkman, Carbohydr. Res., 80 (1980) 147-161. G.G.S. Dutton, J.L. Di Fabio, D.M. Leek, E.H. Merrifield, J.R. Nunn and A.M. Stephen, Carbohydr. Res., 97 (1981) 127-138. G.G.S. Dutton, K.L. Mackie and M.-T. Yang, Carbohydr. Res., 65 (1978) 251-263. G.G.S. Dutton, A.V. Savage and M. Vignon, Can. J. Chem., 58 (1980) 2588-2591. Y.-M. Choy and G.G.S. Dutton, Can. J. Chem., 51 (1973) 3015-3020. i. Y.-M. Choy and G.G.S. Dutton, Can. J. Chem., 51 (1973) 198-207. ii. Y.-M. Choy and G.G.S. Dutton, Carbohydr. Res., 21 (1972) 169-172. G. G.S. Dutton, K.L. Mackie, A.V. Savage, D. Rieger-Hug and S. Stirm, Carbohydr. Res., 84 (1980) 161-170. H. Niemann, H. Friebolin and S. Stirm, unpublished results. G. G.S. Dutton, K.L. Mackie, A.V. Savage and M.D. Stephenson, Carbohydr. Res., 66 (1978) 125-131. Y.-M. Choy, G.G.S. Dutton and A.M. Zanlungo, Can. J. Chem., 51 (1973) 1819-1825. H. Thurow, H. Niemann, C. Rudolph and S. Stirm, Virology, 58 (1974) 306-309. ,b) H. Niemann, B. Kwiatkowski, U. Westphal and S. Stirm, J. Bacterid., 130 (1977) 366-374. J.L. Di Fabio and G.G.S. Dutton, Carbohydr. Res., 92 (1981) 287-298. J.L. Di Fabio, D.N. Karunaratne and G.G.S. Dutton, submitted to Carbohydr. Res. - 250 -K27 a) S.C. Churms, E.H. Merrifield and A.M. Stephen, Carbohydr. Res., 81 (1980) 49-58. K28 a) M. Curvall, B. Lindberg, J. Lonngren and W. Nimmich, Carbohydr. Res., 42 (1975) 95-105. K30 a) B. Lindberg, F. Lindh, J. Lonngren and I.W. Sutherland, Carbohydr. Res., 76 (1979) 281-284. K31 a) C.-C. Cheng, S.-L. Wong and Y.-M. Choy, Carbohydr. Res., 73 (1979) 169-174. K32 a) G.M. Bebault, G.G.S. Dutton, N. Funnel and K.L. Mackie, Carbohydr. Res., 63 (1978) 183-192. G.G.S. Dutton, K.L. Mackie, A.V. Savage, D. Rieger-Hug and S. Stirm, Carbohydr. Res., 84 (1980) 161-170. K33 a) B. Lindberg, F. Lindh, J. Lonngren and W. Nimmich, Carbohydr. Res., 70 (1970) 135-144. K34 a) J.-P. Joseleau, F. Michon and M. Vignon, Carbohydr. Res., 101 (1982) 175-185. K35 a) G.G.S. Dutton and A.V.S. Lim, submitted to Carbohydr. Res. K36 a) G.G.S. Dutton and K.L. Mackie, Carbohydr. Res., 55 (1977) 49-63. K37 a) B. Lindberg, B. Lindqvist, J. L6nngren and W. Nimmich, Carbohydr. Res., 58 (1977) 443-451. K38 a) B. Lindberg, B. Samuelson and W. Nimmich, Carbohydr. Res., 30 (1973) 63-70. K41 a) J.-P. Joseleau, M. Lapeyre, M. Vignon and G.G.S. Dutton, Carbohydr. Res., 67 (1978) 197-212. b) J.-P. Joseleau and A.V. Savage, unpublished results. K44 a) G.G.S. Dutton and T.E. Folkman, Carbohydr. Res., 78 (1980) 305-315. b) G.G.S. Dutton and D.N. Karunaratne, Carbohydr. Res., in press. K45 a) G.G.S. Dutton, J.L. Di Fabio and A. Zanlungo, Carbohydr. Res., 106 (1982) 93-100. K46 a) K. Okutani and G.G.S. Dutton, Carbohydr. Res., 86 (1980) 259-271. b) J.L. Di Fabio, G.G.S. Dutton and H. Parolis, Carbohydr. Res., 133 (1984) 125-133. K47 a) H. Bjorndal, B. Lindberg, J. Lonngren K.-G. Rosell and W. Nimmich, Carbohydr. Res., 27 (1973) 373-378. - 251 -K48 a) J.-P. Joseleau, personal communication. K49 a) J.-P. Joseleau, personal communication. K50 a) E. Altman and G.G.S. Dutton, Carbohydr. Res., 118 (1983) 183-194. K51 a) A.K. Chakraborty, U. Dabrowski, H. Geyer, R. Geyer and S. Stirm, Carbohydr. Res., 103 (1982) 101-105. K52 a) H. Bjfirndal, B. Lindberg, J. LOnngren, M. Mezaros, J.L. Thompson and W. Nimmich, Carbohydr. Res., 31 (1973) 93-100. K53 a) G.G.S. Dutton and M. Paulin, Carbohydr. Res., 87 (1980) 107-117. K54 a) i. P.A. Sandford, J.R. Bamburg, J.D. Epley and T.J. Kindt, Biochemistry, 5 (1966) 2808-2817. ii. P.A. Sandford and H.E. Conrad, Biochemistry, 5 (1966) 1508-1517. b) i. I.W. Sutherland, Biochem. J., 104 (1967) 278-285. ii. G.G.S. Dutton and E.H. Merrifield, Carbohydr. Res., 105 (1982) 189-203. o iii. L.-E. Franze'n, P. Aman, A.G. Darvill, M. McNeil and P. Albersheim, Carbohydr. Res., 108 (1982) 129-138. K55 a) G.M. Bebault and G.G.S. Dutton, Carbohydr. Res., 64 (1978) 199-213. K56 a) Y.-M. Choy and G.G.S. Dutton, Can. J. Chem., 51 (1973) 3021-3026. K57 a) J.P. Kamerling, B. Lindberg, J. LOnngren and W. Nimmich, Acta Chem. Scand. Ser. B, 29 (1975) 593-598. K58 a) G.G.S. Dutton and A.V. Savage, Carbohydr. Res., 84 (1980) 297-305. K59 a) B. Lindberg, J. Lfinngren, U. Rude'n and W. Nimmich, Carbohydr. Res., 42 (1975) 83-93. K60 a) G.G.S. Dutton and J.L. Di Fabio, Carbohydr. Res., 87 (1980) 129-139. b) J.L. Di Fabio, G.G.S. Dutton and H. Parolis, Carbohydr. Res., 126 (1984) 261-269. K61 a) i. A.S. Rao, N. Roy and W. Nimmich, Carbohydr. Res., 67 (1978) 449-456. ii. A.S. Rao and N. Roy, Carbohydr. Res., 76 (1979) 215-224. K62 a) G.G.S. Dutton and M.-T. Yang, Carbohydr. Res., 59 (1977) 179-192. K63 a) J.-P. Joseleau and M.-F. Marais, Carbohydr. Res., 77 (1979) 183-190. - 252 -G.G.S. Dutton and E.H. Merrifield, 103 (1982) 107-128. E.H. Merrifield and A.M. Stephen, Carbohydr. Res., 74 (1979) 241-257. G.G.S. Dutton and D.N. Karunaratne, Carbohydr. Res., 119 (1983) 157-169. L.A.S. Parolis, personal communication. G.G.S. Dutton and K.L. Mackie, Carbohydr. Res., 62 (1978) 321-335. G.G.S. Dutton and E.H. Merrifield, unpublished results. E.H. Merrifield and A.M. Stephen, unpublished results. Y.-M. Choy and G.G.S. Dutton, Can. J. Chem., 52 (1974) 684-687. L. Batavyal and N. Roy, Carbohydr. Res., 98 (1981) 105-113. G.G.S. Dutton and M. Paulin, Carbohydr. Res., 87 (1980) 119-127. G.G.S. Dutton, J.L. Di Fabio, D.M. Leek, E.H. Merrifield, J.R. Nunn and A.M. Stephen, Carbohydr. Res., 97 (1981) 127-138. G.G.S. Dutton and A.V.S. Lim, submitted to Carbohydr. Res. G.G.S. Dutton and D.N. Karunaratne, Carbohydr. Res., 134 (1984) 103-114. M. Curvall, B. Lindberg, J. Lonngren and W. Nimmich, Carbohydr. Res., 42 (1975) 73-82. G.G.S. Dutton and A.V.S. Lim, Carbohydr. Res., 123 (1983) 247-257. B. Lindberg and W. Nimmich, Carbohydr. Res., 48 (1976) 81-84. -253-APPENDIX III Spectrum No. 1 K67 polysaccharide *H-n.m.r. 400 MHz, 90°C Spectrum No. 2 K67 polysaccharide 13C-n.m.r. 100.6 MHz, amb. temp Spectrum No. 3 K67 SH polysaccharide H^-n.m.r. 400 MHz, 90°C Spectrum No. 4 K67 SH polysaccharide 13C-n.m.r. 100.6 MHz, amb. temp —1 Spectrum No. 5 K67 PI polysaccharide H^-n.m.r. 400 MHz, 95°C t j l j A J\lU — r 5 Spectrum No. 6 K67 compound Aj GlcA^Man JH-n.rn.r. 400 MHz, 80°C Spectrum No. 8 K67 compound A 3 GlcAl^Manl^Man 13 C - X A^Ma  ^ -n/m.r. 100.6 MHz, amb temp. to 0\ Spectrum No. 9 K67 compound A4 GlcA^Mani-S-Mani-S-GIc lH-n.Xr. ° 400 MHz, 80°C Spectrum No. 10 K67 compound A4 GIcAi^Manl^ManyGlc 13C-nfm.r. 100.6 MHz, amb. temp. Spectrum No. 11 K67 compound A§ GlcA^Mani^Manl^Glc ^-n.ft.r. 400 MHz, 90°C Spectrum No. 12 K67 compound Ag GlcA^ManJ-^-Mani-^-Glc^Rha 100.6 MHz, amb. temp. TOO Spectrum No. 13 K80 polysaccharide H^-n.m.r. 400 MHz, 95°C Spectrum No. 14 K80 polysaccharide 13C-n.m.r. 100.6 MHz, amb. temp. Spectrum No. IS K80 D polysaccharide H^-n.m.r. 400 MHz, 95°C Spectrum No. 16 K80 D polysaccharide 13C-n.m.r. 100.6 MHz, amb. temp. 1 H 175 100 . p . m . Spectrum No. 17 K80 SH polysaccharide H^-n.m.r. 400 MHz, 95°C 5 4 Spectrum No. 18 K80 SH polysaccharide 13C-n.m.r. 100.6 MHz, amb. temp. 175 100 p . p . m . Spectrum No. 19 K80 compound P3R Gal^Manl/gly ^-iHm.r. 400 MHz, amb. temp. Spectrum No. 20 K80 compound P3R GalUManl/gly 1JC-fi.m.r. 100.6 MHz, amb. temp. 50 p . p . m . -J p . p . m . Spectrum No. 22 K80 compound P5R Gall^Manirgly 400 MHz, amb. temp. Spectrum No. 23 K80 compound P6R G a l ^ M a n y g l y iH-riTm.r. 400 MHz, amb. temp. Spectrum No. 24 K44 polysaccharide (acidic form) ^H-n.m.r. 400 MHz, 95°C 5 4 2 1 p . p . m -Spectrum No. 25 K 4 4 polysaccharide (acidic form) 13C-n.m.r. 100.6 MHz, amb. temp. Spectrum No. 28 K44 polysaccharide (deacetylated) 13C-n.m.r. 100.6 MHz, amb. temp. K ) OO KxT 75 50 25 p.p.m. Spectrum No. 29 K44 compound 4 G I c A ^ R h a A ^ R h a l ^ l c I ^ G l c H^-n.m.r. P 400 MHz, 95°C 5 4 Spectrum No. 30 K44 compound 4 GlcAi^Rhai-^-Rha^Glci^lc lCC-„?m.r. P 100.6 MHz, amb. temp. JJ 100 OO 50 0 p . p . m . Spectrum No. 31 K44 compound 5 GlcAi^Rhai-3-Rhai-^lci^Clc * ° * |6 ^-n.m.r. OAc 400 MHz, 95°C Spectrum No. 32 K44 compound 5 GIcAi-^Rhai-^Rhai-^GIci-^GIc 0 ° ° 0 13 C-n.m.r. 100.6 MHz, amb. temp. 16 OAc 170 4 I — JJ u 100 60 Spectrum No. 33 K44 compound 6 G I c A i ^ R h a l J R h a l ^ I c l ^ I u c i t o l 1H-n.m.r. " 400 MHz, 95°C Spectrum No. 34 K44 compound 7 GIcAi^Rhai-^Rhal-^-Glcl^Glucitol 0 0 |6 OAc *H-n.m.r. 400 MHz, 95°C 5 Spectrum No. 35 K26 compound Pla H^-n.m.r. 400 MHz, 95°C T Spectrum No. 36 K26 compound Pla 13C-n.m.r. 100.6 MHz, amb. temp. Spectrum No. 37 K26 compound PI *H-n.m.r. 400 MHz, 95°C Spectrum No. 38 K26 compound PI 13C-n.m.r. 100.6 MHz, amb. temp. Spectrum No. 39 K26 compound P2 H-n.m.r. 400 MHz, 95°C c 5 4 3 2 ppm Spectrum No. 40 K26 compound P2 l3C-n.m.r. 100.6 MHz, amb. temp. V 100 — I 1 1 1 I 50 10 p.p.m. Spectrum No. 42 K26 alditol of PI H^-n.m.r. 400 MHz, 95°C Spectrum No. 43 K 2 6 alditol of P 2 H^-n.m.r. 400 MHz, 95°C 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0060518/manifest

Comment

Related Items