Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural studies of Klebsiella capsular polysaccharides Yang, Mo-Tai 1974

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

Item Metadata

Download

Media
831-UBC_1974_A1 Y35.pdf [ 8.55MB ]
Metadata
JSON: 831-1.0061058.json
JSON-LD: 831-1.0061058-ld.json
RDF/XML (Pretty): 831-1.0061058-rdf.xml
RDF/JSON: 831-1.0061058-rdf.json
Turtle: 831-1.0061058-turtle.txt
N-Triples: 831-1.0061058-rdf-ntriples.txt
Original Record: 831-1.0061058-source.json
Full Text
831-1.0061058-fulltext.txt
Citation
831-1.0061058.ris

Full Text

STRUCTURAL STUDIES OF KLEBSIELLA CAPSULAR POLYSACCHARIDES by MO-TAI YANG B.Sc, National Taiwan University, 1963 M.Sc, National Taiwan University, 1965 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department °* Chemistry •< We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1974 In p r e s e n t i n g t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f the r equ i r emen t s f o r an advanced degree at the U n i v e r s i t y o f B r i t i s h C o l u m b i a , I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s tudy . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y pu rposes may be g r a n t e d by the Head o f my Department o r by h i s r e p r e s e n t a t i v e s . It i s u n d e r s t o o d that c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l owed w i thout my w r i t t e n p e r m i s s i o n . Department o r The U n i v e r s i t y o f B r i t i s h Co lumbia Vancouver 8. Canada D a t e ^ j r , ABSTRACT The genus K l e b s i e l l a b e l o n g s t o t h e f a m i l y E n t e r o b a c t e r i a c e a e . E i g h t y t y p e s o f K l e b s i e l l a h a v e b e e n s e r o l o g i c a l l y c l a s s i f i e d u s i n g t h e c a p s u l a r p o l y s a c c h a r i d e s as a n t i g e n s . I n o r d e r t o u n d e r s t a n d t h e c h e m i c a l b a s i s o f s e r o l o g i c a l r e a c t -i o n s , s t r u c t u r a l s t u d i e s o f t h e c a p s u l a r p o l y s a c c h a r i d e s f r o m t h r e e s t r a i n s ° f K l e b s i e l l a ^ n a m e l y . K 5 , K 6 2 and K 1 8 were c a r r i e d o u t by means o f c o n v e n t -i o n a l m e t h o d s ( s u c h as h y d r o l y s i s , m e t h y l a t i o n and p e r i o d a t e o x i d a t i o n ) as w e l l a s m o d e r n m e t h o d s ( s u c h as c i r c u l a r d i c h r o i s m ( c . d . ) f o r t h e a s s i g n m e n t o f t h e £ o r L c o n f i g u r a t i o n o f m o n o s a c c h a r i d e s , g a s - l i q u i d c h r o m a t o g r a p h y ( g . l . c . ) and mass s p e c t r o m e t r y f o r i s o l a t i o n and i d e n t i f i -c a t i o n o f s u g a r m o i e t i e s ) . I n t h e c o u r s e o f t h e p r e s e n t i n v e s t i g a t i o n , p . m . r . s p e c t r o s c o p y a t 9 5 ° f o r t h e a s s i g n m e n t o f a n o m e r i c c o n f i g u r a t i o n s o f the c o n s t i t u e n t s u g a r s , and S e p h a d e x g e l f i l t r a t i o n f o r t h e s e p a r a t i o n o f a s e r i e s o f a c i d i c o l i g o s a c c h a r i d e s o b t a i n e d f r o m t h e p a r t i a l h y d r o l y s i s o f p o l y s a c c h a r i d e s f o r s e q u e n t i a l a n a l y s i s h a v e b e e n d e v e l o p e d . The r e p e a t i n g u n i t s o f K 5 , K 6 2 and K18 a r e as f o l l o w s . CH3 COOH V 4 A 6 K5 K 6 2 / \ D-GlcAp D-Glcp ~ D-Manp ~ — - - y6 ~ i ~ s$ - ~ s& 2 - 0 A c ~ D-GlcAp ~ D-Manp ~ ^ D-Galp ~~~ D-Glcp ~ -D-Manp K 1 8 ~ L-Rhap ~ D-Galp ~ ~ D-Glcp ~ D-GlcAp L-Rhap i ^ -cx 11 D - G l c p The structure of K5 capsular polysaccharide lacks any side chain and is also unusual in affording the f i r s t example of a 4,6-0-(l-carboxyethylidene)-D-raannose unit In a natural product. Capsular polysaccharide of K.62 belongs to the prevailing structural pattern so far published in this genus which is composed of a repeating unit involving a single sugar in the side chain with three to four mono-saccharides in the backbone. Capsular polysaccharide from K18 is unusua complex in that i t consists of two L-rhamnoses in the hexasaccharlde repeat unit which renders the polysaccharide less viscous and acid l a b i l e . Finally, an attempted immunological study on K5 is briefly discussed. i i i TABLE OF CONTENTS Page I. INTRODUCTION 1 II. METHODOLOGY OF STRUCTURAL STUDY ON POLYSACCHARIDES 7 1. General Structural Remarks of Polysaccharides .... 7 2. Isolation and Purification 8 3. Total Hydrolysis of Polysaccharides 9 4. Methanolysis 16 5. Methylation 17 6. Carboxy1 Reduction of Acidic Carbohydrates ....... 21 7. Partial Hydrolysis 22 8. Periodate Oxidation '.. 24 9. Gas-liquid Chromatography 28 10. Mass Spectrometry of Partially Methylated A l d i t o l Acetates 31 11. Determination of Anomeric Configuration 36 12. Determination of D- and.L-Configuration of a Sugar Residue 39 13. Immunochemical Methods 39 III. THE STRUCTURE OF THE CAPSULAR POLYSACCHARIDE FROM KLEBSIELLA K-TYPE 5 43 SUMMARY 44 DISCUSSION 44 EXPERIMENTAL 51 IV. THE STRUCTURE OF THE CAPSULAR POLYSACCHARIDE FROM KLEBSIELLA K-TYPE 62 77 i v SUMMARY 78 DISCUSSION 7.9 EXPERIMENTAL 84 V. THE STRUCTURE OF THE CAPSULAR POLYSACCHARIDE FROM KLEBSIELLA K-TYPE 18 101 SUMMARY 102 DISCUSSION 103 EXPERIMENTAL 108 APPENDIX A: IMMUNOCHEMICAL STUDIES ON KLEBSIELLA K-TYPE 5 POLYSACCHARIDE 126 APPENDIX B: P.M.R.SPECTRA 129 K5 CAPSULAR POLYSACCHARIDE 130 SODIUM HYDROXIDE TREATED K5 131 TWICE PER10DATE OXIDIZED AND SODIUM BOROHYDRIDE REDUCED K5. . 132 METHYLATED K5 133 ; METHYLATED AND REDUCED K5 134 CELLOBIOURONIC ACID ( A l ) FROM K5 135 ALDOTRIOURONIC ACID (A2) FROM K5 136 NEUTRAL DISACCHARIDE (N3 ) FROM K5 137 PYRUVIC ACID-MANNOSE-ERYTHRONIC ACID COMPLEX FROM K5 138 K62 CAPSULAR POLYSACCHARIDE 139 MILDLY DEGRADED (DILUTE ACID) K6.2 140 ALDOBIOURONIC ACID (A2) FROM K62 141 . ALDOTRIOURONIC ACID (A3;) FROM K62 1 142 PSEUDOTETRAOURONIC ACID (A4) FROM K62 143 V P a £ e DISACGHARIDE GLYCOSIDE FROM SMITH DEGRADATION OF K62 144 K18 CAPSULAR POLYSACCHARIDE 145 ALDOBIOURONIC ACID (A2) FROM K18...' 146 NEUTRAL DISACCHARIDE ( N l ) FROM K18 147 PSEUDOTRIOURONIC ACID (A3) FROM K18 148 TETRASACCHARIDE.(A4) FROM K18 149 A5a FROM K18 150 A6a FROM K18 151 MILD ACID TREATED POLY ALCOHOL FROM K18 ..' 152 APPENDIX C: EXPRESSIONS OF RAW SUGAR RATIO DATA FROM G.L.C.-ANALYSIS .153 EXAMPLES FROM K5 154 EXAMPLES FROM K5 156 EXAMPLES FROM K18 158 BIBLIOGRAPHY 161 v i LIST OF TABLES Table No. Page TABLE I RATE CONSTANTS FOR HYDROLYSIS OF THE GLYCOSIDIC BONDS IN THE AEROBACTER AEROGENES A3 (Sl) POLYSACCHARIDE 14 TABLE II RATE CONSTANTS FOR HYDROLYSIS OF THE GLYCOSIDIC BONDS IN SEROTYPE 2 CAPSULAR POLYSACCHARIDE OF AEROBACTER  AEROGENES 15 TABLE III PRIMARY FRAGMENTS IN THE MASS SPECTRA OF PARTIALLY METHYLATED SUGARS IN THE FORM OF THEIR ALDITOL ACETATES 37 TABLE IV OPTICAL ROTATIONS AND P.M.R. DATA ON KLEBSIELLA K 62 CAPSULAR POLYSACCHARIDE AND DERIVED OLIGOSACCHARIDES . 81 v i i LIST OF FIGURES Figure No. Page Figure 1 The repeating unit of the Aerobacter aerogenes A3 (Sl) polysaccharide 14 Figure 2 The structure of the serotype 2 capsular polysaccharide of Aerobacter aerogenes 16 Figure 3 Periodate degradation of Klebsiella K-type 62 polysaccharide 25 Figure 4 Periodate degradation of Klebsiella K-type 5 polysaccharide 27 v i i i ACKNOWLEDGEMENTS I am s i n c e r e l y g r a t e f u l to Professor G.G.S. Dutton f o r h i s patience, encouragement and advice throughout the e n t i r e research and during the pre p a r a t i o n of t h i s t h e s i s . I wish to thank Dr. K.B. Gibney, Dr. Y'.M. Choy, Dr. G'.M. Bebault, Dr. J.M. Berry and Roger Walker f o r t h e i r v a l u a b l e d i s c u s s i o n s and s t i m u l a t i o n s . I a l s o g r a t e f u l l y acknowledge the in v a l u a b l e a s s i s t a n c e extented by Dr. P.C. Simon i n the immunological study on K5, Dr. P.J. S a l i s b u r y f o r growing the b a c t e r i a , and Dr. P'.E. Reid f o r the e l e c t r o p h o r e t i c a n a l y s i s . F i n a l l y , I would l i k e to express my si n c e r e g r a t i t u d e to Miss O l i v e Cheung f o r her a s s i s t a n c e i n the pre p a r a t i o n of t h i s manuscript. I. INTRODUCTION As a l l diseases are fundamentally chemical phenomena of livi n g organisms, a chemical approach to understanding the disease at the molecular level is essential. The defensive mechanisms of li v i n g organisms consist, in part, of a mechanical barrier, such as the skin. After this barrier i s penetrated by an invasive foreign agent (antigen), e.g., bacterium, the hosts' immune system is stimulated to produce humoral antibodies and cellular immunity. Besides the phagocytosis by macrophages, the specific antibody fa c i l i t a t e s the mechanisms responsible for the destruction and removal of the foreign agents. If the immune system f a i l s to remove the antigen(s), the disease persists and the host requires medical treatment. The indispensable role of the immune system in the body's defence is thus evident, though many of i t s basic processes are s t i l l , unknown. However, since the discovery of vaccination by Edward Jenner in 1798, some major breakthroughs have been achieved; i.e., discovery of cellular events in the mechanism of immunity (in addition to humoral immunity), molecular basis of antigenicity, and structural elucidation of antibodies (1). The determination of the structure of the slime polysaccharides of Klebsiella-- also examining the molecular basis of their antigenicity--|is the principle concern of this thesis. The structures of three strains of Klebsiella capsular antigens (K-antigens), namely K5, K62 and K18, were determined. The "inoculation of rabbits, guinea pigs and mice with K5 has been carried out, and no demonstrable antibodies have been obtained. Nevertheless, a study of antigenic determinants of these K-antigens has 2 been conducted by Eriksen (2) and by Heidelberger (3). Their results are consistent with the structures proposed in this research, and w i l l be mentioned briefly later. Generally, polysaccharides are considered to be weakly or non-antigenic though they may have high molecular weights. Unlike proteins polysaccharides lack a definite three-dimensional structure to activate the immune system. Nevertheless the repeating nature of the capsular polysaccharides is somehow "impressive", and possesses antigenicity. The genus Klebsiella belongs to the family Enterobacteriac A l l c e l l s of Enterobacteriaceae have lipopolysaccharide (LPS) bound and external to the c e l l wall. LPS can be extracted by reagents such as phenol or ethylenediaminetetraacetic acid (EDTA), suggesting that the linkage between lipopolysaccharide and the c e l l wall may be ionic rather than covalent. The lipopolysaccharide layer may be further enveloped by an extracellular capsule in the case of encapsulated strains. Most of the Klebsiella strains possess heavy capsules (K+) whereas the noncapsulated variants are separable from K+ cultures. Most of the capsular antigens are negatively charged. The surface charge confers resistance to phagocytosis, thus imparting virulence to the bacteria. Generally, slime can be considered as excreted capsular polysaccharide. Wilkinson (4) has shown that capsular and slime poly-saccharides are identical in chemical composition. However; since M-antigen (5,6) and DNA released as a result of autolysis of the c e l l may also be present in the slime (7), their presence must be viewed with some caution. 3 In the course of the structural study of Klebsiella capsular antigen type 5, a problem of irreversible mutation from K5+ to K5- (8) (prskov's notation, less encapsulated substrain) was encountered. This resulted in greatly reduced slime production and greater contamination in the slime of the neutral polysaccharides which were cleaved from 0-antigen. The heavy contamination was very hard to remove completely. S e r o l o g i c a l l y K5+ and K5- are the same strain, but K5- gives better immunological response, possibly due to a much thinner capsule which results in less incidence of immunoparalysis in the immunized animal (8). That K5- is the better antigen and that i t s antiserum reacts solely against the capsular polysaccharide is of great importance. This property i s also similar to the blood group antigens whose end group monosaccharides are the antigenic determinants (9). These two examples are evidence for the principle that the exposed components are the determinants of the whole antigen. Additionally, allergenic constituents of pollen which induce hypersensitivity are shown to be located in the outer c e l l wall (exine) of the pollen grain (10). In the present study of Klebsiella capsular polysaccharides classic and modern methods and sometimes, new approaches have been employed. The chemistry of Klebsiella has been reviewed by Luderitz, Jann and Wheat (11). Now more than eighty strains have been serologically identified. Nimmich (12,13) has reported the qualitative analysis of the constituent sugars of K-types 1 to 80. Most of them contain D-glucuronic acid in combination with D-mannose, D-glucose, D-galactose and, to a lesser extent, with L-rhamnose and L-fucose. About half of the polysaccharides are found to incorporate pyruvic acid. Uronic acid and/or 4 pyruvic acid confer the acidity on a l l of the Klebsiella capsular polysaccharides. Wheat, Dorsch and Godoy were the f i r s t to report the presence of pyruvic acid in K. rhinoscleromatis polysaccharide (K-type 3) (14), and the f i r s t complete structural study was carried out on K54 by Sanford and Conrad (15,16) using improved methylation procedures devised by Hakomori (17). The occurrence of pyruvic acid as a covalently linked ketal unit in polysaccharides was f i r s t reported by Hirase (18) in a study on agar which contains 4,6-0_-(l-carboxyethylidene)-D-galactose units. The presence of 4,6-0-(l-carboxyethylidene)-D-glucose units was subsequently demonstrated by Sloneker and Orentas (19) in the exocellular polysaccharide from Xanthomonas campestrls and the occurrence of 3,4-0-(l-carboxyethylidene)-L-rhamnose was found in the capsular polysaccharide of Klebsiella K-type 72 (20). The presence of the D-galactose ketal has also been shown in the polysaccharides from Corynebacterium insidiosum (21) and Klebsiella K-type 21 (22). Pyruvic acid has been found as well in the M-antigens °f Salmonella and in E.coli linked as 3,4-0-(l-carboxyethylidene)-D-galact-ose (23,24). In addition to these ketals of known structure the existence of pyruvic acid has also been demonstrated in other bacteria (14,25,26) and in certain cases this acid may be an immunodominant group (26). When our structural study on Klebsiella capsular polysaccharides was started, the only known structures were K-types 2 (27), 8 (28), and 54 (15,16). Now the structures of K-types 9 (29), 20 (31), 21 (22), 24 (32), 38 (33), 47 (30), 52 (34), 56 (35) and 72 (20) have also been reported. A detailed structural study of K-types 5 (36,37), 62 and 18 w i l l be the contribution of this thesis. The structures of K5, 62 and 18 can be represented as follows. CH3 COOH K5 V 1 4 1 3 x 1 D - GlcAp ~g D - Glcp D - Manp ~g 2-0Ac K62 - 2 — | - GlcAp ^ D - Manp ^  D - Galp ^ D - Glcp ~ ^ 1 D - Manp K18 — L - Rhap ~ ~ D - Galp — - O - Glcp ~ 1 D - GlcAp—-L-Rhap D -Glcp K5 slime polysaccharide comprises a repeating unit of only three monosaccharides. This polysaccharide lacks any carbohydrate side chain and is also unusual in affording the f i r s t example of a 4,6-0-(l-carboxy-ethylidene)-D-mannose unit in a natural product. K62 slime polysaccharide belongs to the prevailing structural pattern of this genus, i.e., a repeating unit composed of one side unit to a backbone of t r i - or tetra-saccharide. K18 capsular polysaccharide is composed of hexasaccharide repeating units with two L-rhamnoses in each repeating unit and one D -glucose as the side chain. This polysaccharide is comparatively complicated due to the length of the repeating block and the presence of L-rhamnoses which are acid l a b i l e . These structures are proposed on the basis of methylation analysis, periodate oxidation and partial hydrolysis studies. The general methods of investigation are outlined in section II, and the specific evidence on which the structures of K5, 62 and 18 are based is discussed in sections III, IV and V, respectively. Finally, an attempted immunological study on K5 is briefly discussed. II. METHODOLOGY OF STRUCTURAL STUDY ON POLYSACCHARIDES 1. General Structural Remarks on Polysaccharides Polysaccharides, with or without obvious repeating units, can be cl a s s i f i e d into one of the following groups: A. Linear horaoglycans. Such as amylose, cellulose and pectin. B. Branched horaoglycans. Such as amylopectin, glycogen and dextran. C. Linear heteroglycans. Such as Pneumococcus type III and Klebsiella K-type 5 polysaccharides. D. Branched heteroglycans. Such as wood hemicelluloses and plant gums. In the case of a polysaccharide consisting of repeating units, a structural study should characterize the nature of the sugar residues, the position of linkage, the sequence, the anomeric configurations, and i f possible, the molecular weight. In other words, the exact structure of the repeating unit and an approximate degree of polymerization have to be demonstrated. On the other hand, the exact overall sequence can not be determined in the case of polysaccharides lacking repeating units. The investigations can only provide information on the "average structure" including the different monosaccharide units present, the type of end groups,th p o s i t i o n of the-branch points and the nature of linkages. Some polysaccharides may have a main chain (such as hemicelluloses) while others may consist of a multiple-branched system in which no main chain can be defined (e.g., amylopectin and gum exudates). The following discussions w i l l only be concerned with poly-saccharides with a repeating unit. Of course, most of the techniques are also applicable to the polysaccharides of irregular structure. 8 2. Isolation and Purification Each strain of Klebsiella was inoculated in a sucrose-yeast-extract broth culture in an Erlenmyer flask which was shaken for 24h unti l a definite turbidity due to the growth of bacteria was observed. The culture was then spread on trays on sucrose-yeast-extract agar for 3 or 4 days to reach a maximum growth of slime. Slime was harvested, diluted with aqueous phenol and centrifuged. The acidic polysaccharide in the supernatant was precipitated with 5% hexa-decyltrimethyl ammonium bromide (Cetavlon). The precipitate was dissolved in brine again precipitated with ethanol, then dialyzed against tap water. The purified, deionized acidic polysaccharide was subjected to micro-analysis for nitrogen and ash, as well as ultraviolet spectrophoto-metry analysis for protein and nucle i c acids content. The constituent sugars were qualitatively examined by paper chromatography following hydrolysis of the polysaccharide, while quantitative analysis was performed by gas-liquid chromatography (g.l.c.) on the alditol acetates derived from the sugar residues. Normally, i t is not d i f f i c u l t to obtain a purified acidic polysaccharide. Most of the contamination is by neutral polysaccharides which are not precipitated by Cetavlon and as a result remain in the supernatant. In the case of heavy contamination by neutral polysaccharides, Cetavlon precipitation has to be repeated until a constant sugar ratio of the monosaccharides i n the a c i d i c polysaccharide i s obtained. However, i f the contamination is overwhelming, there is d i f f i c u l t y of singling out the desired fraction and an examination of the slime-producing bacteria is necessary, as the Klebsiella bacterial culture may be contaminated or 9 undergoing mutation to a substrain which synthesizes only a minor fraction of the acidic polysaccharide, as in the case of Klebsiella K-type 5. However, the minute yet persistant contamination (e.g., a galactan in K5 acidic polysaccharide) can be ignored, for i t s presence is too l i t t l e to be considered part of the repeating structure. Cetavlon precipitation has been proven sufficient for the purpose of purification of Klebsiella capsular polysaccharides, provided that the i n i t i a l separation of the cells by centrifugation is complete and gives a clear yellowish supernatant. Separation, by anion exchange column using diethylaminoethyl (DEAE) Sephadex is only suitable for analytical purpose on a milligram scale due to the extraordinarily high viscosity of the acidic polysaccharide. Purification by precipitation with the specific antiserum to the acidic polysaccharide is not feasible, since the isolated acidic polysaccharide is a very weak immunogen and e l i c i t s antibody rather poorly. 3. Total Hydrolysis of Polysaccharides The f i r s t step in the investigation of a polysaccharide structure is total hydrolysis followed by analysis of the hydrolyzed products qualitatively and quantitatively. Hydrolysis using hydrochloric acid (IM) is known to be more destructive on sugars than sulfuric acid (0.5M) (38), and 6 to 8 h of heating at 100° in either acid is sufficient to break down completely a neutral polysaccharide into i t s monomers. Hydrochloric acid can be removed at diminished pressures at 30-40° while sulfuric acid has to be neutralized with barium carbonate, and after removal of the precipitate (barium sulfate plus excess barium carbonate) the supernatant is deionized by cation then anion exchange resins. 10 Two molar trifluoroacetic acid (TFA) (39) has approximately the same hydrolytic strength, as compared to IM hydrochloric acid and 0.5M sulfuric acid, and i t does not significantly degrade sugars in the conditions normally used for hydrolysis ( i . e . 100°, 6-8h). Moreover, trifluoroacetic acid is easily removed from the reaction mixture by evaporation, and has been frequently used in this study. As to acidic polysaccharides, the conditions employed in the hydrolysis of neutral polysaccharides are not strong enough to ensure a complete cleavage of the glycuronosyl-bonds, and a considerable amount of aldobiouronic acid remains intact. The incomplete release of the uronic acid and the neutral sugar attached to i t give rise to a discrepancy of the sugar ratio in the hydrolyzate and the sugar ratio in the original polymer. Nevertheless, incomplete hydrolysis i t s e l f is a useful indication of the possible combination of the aldobiouronic acid. In order to achieve complete hydrolysis, an acidic polysaccharide can be converted to the neutral alditois by methanolysis to form methyl glycosides and the methyl ester methyl glycoside which is reduced in methanol with sodium borohydride to the methyl glycoside. The mixture of methyl glycosides is then hydrolyzed with 2M trifluoroacetic acid and reduced to alditois. The sugar ratio is obtained by g.l.c* analysis of the alditoi acetates. Alternatively, a direct total hydrolysis without transforming the uronic acid to a neutral polyol can be achieved by i n i t i a l hydrolysis and reduction in the usual manner, followed by more stringent conditions to further hydrolyze the remaining aldobiouronic acid in the alditoi form which is more resistant to acidic degradation than reducing sugars. By removing the uronic acid in i t s acidic form from the hydrolyzate with anion-exchange resin, the ratio of the constituent 11 neutral monosaccharides can be obtained by g.l.c. analysis of the alditol acetates. By comparing the ratio of only neutral sugars with the ratio of neutral and the transformed uronic acid, the type of the uronic acid as well as the molar ratio of individual sugars can be obtained. The preceding design was applied to K5. acidic polysaccharide because of i t s particular relevancy and partially applied to K62 and K18. F i r s t l y , K5 was methanolyzed (4% methanolic hydrogen c h l o r i d e , 100°C 10 h) to methyl glycosides which were then reduced in methanol by sodium borohydride, then hydrolyzed, reduced, and the polyol mixture was analyzed by g.l.c. as alditol acetates. The ratio of D-glucitol hexaacetate (mp. 95°) to D-mannitol hexaacetate (mp. 123°) was 2:1. Secondly, in order to achieve a total hydrolysis followed by the removal of the acidic sugars, K5 was i n i t i a l l y hydrolyzed in the usual way (2M trifluoroacetic acid, 10 h, 97°). The hydrolysate was evaporated to dryness then made alkaline with 0.5M sodium hydroxide and reduced with sodium borohydride at room temperature for lh to a mixture of mono- and disaccharide alditols. Hydrochloric acid was added to a concentration of 2M and the solution was heated at 100° for 4h to achieve further hydrolysis. After sodium borohydride reduction, the hydrolyzate was acidified and passed through anion-exchange resin to completely remove L-gulonic acid. The neutral effuent was analyzed and found to contain D-glucose and D-mannose in a ratio of 1:1. Thus one can conclude that the uronic acid in K5 is D-glucuronic acid and the repeating unit of the polysaccharide consists of D-glucuronic acid, D-glucose and D-mannose in a 1:1:1 ratio. On the other hand, i f knowledge of the type of the acidic sugar 12 rather than the exact ratio of sugars is desired, a modified procedure can be used. The acidic polysaccharide is hydrolyzed with 2M hydrochloric acid at 100° for 7h. A portion of the hydrolyzate is treated with anion-exchange resin to remove a l l the acidic components and the neutral sugars are analyzed. A separate portion of the hydrolyzate i s evaporated to dryness. The residue is reduced with sodium borohydride in methanol to convert neutral sugars as well as glucuronolactone into alditois, which are then analyzed. These two analyses of sugars can provide the information as to the type of the uronic acid and an approximate sugar ratio. A milder reagent, 2M trifluoroacetic acid, can be used for hydrolysis instead of 2M hydrochloric acid with lesser depolymerization and degradation of sugars. After examination of various hydrolysis conditions applied to the depolymerization of polysaccharides, i t i s desirable to have an overall survey of the acid l a b i l i t y of glycosidic linkages to provide a basis for the choice of a certain condition of hydrolysis. Capon (40) has reviewed comprehensively the first-order rate constants for the acid-catalyzed hydrolysis of the glycosides of mono-saccharides. Those data can be extrapolated to provide information as to the l a b i l i t y of the glycosidic linkages in a polysaccharide. To demonstrate the difference in rate of hydrolysis with respect to different types of glycosidic linkage, examples from poly-saccharides w i l l be given for the purpose of easy comparison. For simplification, only the major factors governing the rate of hydrolysis are discussed, i.e. the ring size, the type of the sugars, the position of sugar units in the polymer, and the anomeric configuration of the glycosidic bonds. Several generalizations (41) can be made: 13 a. Furanosides are more labile than pyranosldes. b. Deoxysugars are more readily hydrolyzed than hexoses (27). c. Uronic acids are more resistant to hydrolysis and as a result aldobiouronic acids always persist at the end of hydrolysis. d. Amino sugars are more acid resistant than common hexoses. e. Pentopyranosides are more acid labile than hexopyranosides. f. Alpha glycosidic bonds are generally more labile than beta. g. Residues present as side chains are often more easily hydrolyzed than when present in the main chains. Examples demonstrating these main points are discussed below. A pentosan isolated from the water-insoluble portion of Durum wheat endosperm (42) was found to have a general structure consisting of D-xylopyranosyl units linked 1—»4 with single-unit L-arabinofuranosyl side chains attached to the 2 and 3 positions of the xylose units. On mild hydrolysis of the pentosan (0.025M sulfuric acid, 75-77° for 4.5h) and followed by precipitation with ethanol, at a fina l concentration of 85%, the supernatant was shown to contain 60% of the L-arabinose and trace of D-xylose; while the precipitate contained over 90% of the D-xylose. The observation of the extreme acid l a b i l i t y of the L-arabinose indicated that i t existed in the furanose form. The water-soluble wheat flour pentosan with a similar structure possessed the same acid-labile property (43)* Klebsiella 0 group 9 lipopolysaccharide (44) is composed of D-galactofuranose and D-galactopyranose units. When the polysaccharide was methylated and treated with 90% formic acid at 70° for 45 min., a methylated disaccharide was released as a result of preferential cleavage of positions corresponding to the furanosyl linkages. 14 From the study of the hydrolysis of Aerobacter aerogenes A3 (Sl) polysaccharide (Fig. 1), Conrad et a l . (16) have reported some useful rate constants for hydrolysis of glycosidic bonds by measuring the release and: subsequent/cleavage of oligosaccharides as well as the progressive release of monosaccharides during the hydrolysis. Figure 1: The repeating unit of the Aerobacter aerogenes A3 (Sl) polysaccharide. J L D-Glcp ^ ~ D-GlcAp ~ L-Fucp - — — — / O — _ o < _ — D-Glcp TABLE I RATE CONSTANTS FOR HYDROLYSIS OF THE GLYCOSIDIC BONDS IN THE AEROBACTER AEROGENES A3 (Sl) POLYSACCHARIDE. Rate Constant Bond cleaved (sec _ 1) X 10 3 L-fucose D-glucose 1.3 D-glucose D-glucose 0.13 D-glucose i ~ D-glucuronic acid 0.05 D-glucuronic acid i - 3 - L-fucose 0.004 In the Aerobacter aerogenes A3 (Sl) polysaccharide the l a b i l i t y of the L-fucosidic linkage i s about 300 times greater than that of D-glucuronosidic linkage and the st a b i l i t y of other glycosidic bonds l i e s in between. Although identical in type of linkage, D-glucose as a side 15 chain is 2.6 times as labile as compared with i t s l a b i l i t y when in the backbone. The marked l a b i l i t y of L-fucosyl bonds explains the observation that oligosaccharides werefridt found containing _L-fucosyl linkages. Thus the anomeric configuration of L-fucose in the polysaccharide could not be determined by p.m.r. spectroscopy of oligosaccharides. This kind of d i f f i c u l t y was circumvented in this laboratory by using intact poly-saccharide for p.m.r. measurements. A study of rate constants for hydrolysis of the glycosidic bonds in the serotype 2 capsular polysaccharide of Aerobacter aerogenes (Fig. 2) (27) gave the data lis t e d in Table II. Figure 2: The structure of the serotype 2 capsular polysaccharide of Aerobacter aerogenes. D-Glcp i - ^ D-Manp Kr D-Glcp - — — — /? = I —- = — Oi D-Glcp TABLE II RATE CONSTANTS FOR HYDROLYSIS OF THE GLYCOSIDIC BONDS IN SEROTYPE 2 CAPSULAR POLYSACCHARIDE OF AEROBACTER AEROGENES. Rate constant Bond cleaved (sec" 1) X 10 D-glucose D-mannose D-glucose D-mannose 0.119 0.285 0.011 0.362 16 One can notice the increased rate for hydrolysis of the D-glucuronic acid as the branch unit in the serotype 2 capsular polysaccharide of Aerobacter aero genes (0.011 sec "*•), as compared to the D-glucuronic acid (in main chain) in Aerobacter aerogenes A3 (Sl) polysaccharide (0.004 sec~\ Table I ) . Fi f t y percent of D-mannose in Klebsiella K62, f i f t y percent of D-glucose in Klebsiella K18, and aldobiouronic acid in Klebsiella K20 (31) are a l l branching units and preferentially released in the course of hydrolysis. 4. Methanolysis Bishop et a l . (45-47) have conducted systematic studies of the methanolysis of pentoses and hexoses, and reported in detail on the formation, isomerization, distribution, and stability of various anomeric and ring isomers in terms of conformational analysis. Entlicher and BeMiller (48!) reported that methanolysis of the glycosidic bonds of amylose was more facile than acid hydrolysis and that methanolysis gave predominantly, or perhaps complete, inversion. Sandford et a l . (49) resorted to methanolysis for the sugar analysis of black yeast NRRL Y-6272 polysaccharide which was composed of N-acetylglucosamine and N-acetylglucosaminuronic acid, because the acidic polysaccharide underwent extensive degradation to form humin upon acid hydrolysis. Methanolysis was employed in the complete depolymerization of acidic polysaccharides as mentioned in section II 3. F.m.r. spectroscopy is a useful physical method for the detection of acidic components such as formyl, acetyl and pyruvic acid ketal units (50,51) in a polysaccharide. Alternatively, a chemical method 17 consisting of methanolysis followed by alkaline hydroxylamine treatment (52) on the polysaccharide has been devised to convert the acidic components in the polysaccharide into their -hydroxamateswhich are subsequently analyzed by thin layer chromatography ( t . l . c ) . Polysaccharide from K5 was methanolyzed and treated with alkaline hydroxylamine. Hydroxamates of acetic, pyruvic and D-glucuronic acid were detected on t . l . c . by comparison with authentic standards. It was found in the course of methanolysis of K5 that the pyruvic acid ketal, in contrast to glycosidic bonds and ester linkages, was quite resistant to methanolysis. This discovery was applied to the partial hydrolysis of periodate-oxidized and sodium borohydride-reduced K5 to obtain an oligosaccharide containing pyruvic acid. An attempt was made to break down K5 by methanolysis to isolate oligomers carrying pyruvic acid. The product of methanolysis was separated by paper chromatography and fafractions with an p.m.r. signal-of the- pyruvate methyl group were obtained. But no further purification and identification was done. This method merits further investigation for i t s general applicability for the isolation of the complex of sugar-pyruvate ketal. 5. Methylation In the technique of methylation analysis,the free hydroxyl groups in a polysaccharide are f i r s t exhaustively etherified and then the glycosidic bonds are hydrolyzed to obtain partially methylated monosaccharides which are subsequently separated and identified. Structural information such as branchings and types of sugar can thus be derived. The methods of methylation generally used individually or in combination are: a. Hakomori methylation (17), b. Kuhn methylation (53), 18 and c. Purdie methylation (54). a. Hakomori methylation: The methylsulfinyl anion prepared by heating dimethyl sulfoxide (DMSO) with sodium hydride at 60-70° (15) was used to generate alkoxide ions from free hydroxyl groups in a polysaccharide. Methyl iodide was then added to the polysaccharide alkoxide to produce a methylated polysaccharide which may be purified by dialysis against tap water. A polysaccharide containing uronic acids is not suitable for more than one Hakomori methylation, because the methyl ester of the uronic acid can react with methylsulfinyl anion to form methylsulfinylmethy1 ketone; and the methyl ester can undergo ^-elimination in a strongly basic medium. The jB -elimination is particularly pronounced when the uronic acid is substituted at C4. These side reactions can be minimized i f the incompletely methylated polysaccharide is treated with dilute aqueous alkaline acetone solution prior to the second methylation. b. Kuhn methylation: When the i n i t i a l degree of methylation is low following the Hakomori method, the Kuhn methylation may be used next. The partially methylated polysaccharide is dissolved in a mixture of N, N-dimethyl-formamide (DMF) and methyl iodide. Silver oxide is added in small portions at various time intervals with st i r r i n g . The reaction is allowed to proceed for 16 h, then the mixture is centrifuged and the solid substance is washed with chloroform. The combined centrifugate and chloroform washings are washed with water then evaporated to dryness. c. Purdie methylation: . This method is generally used to ensure maximum methylation of the polysaccharide after Hakomori, or Hakomori and Kuhn methylations. 19 The partially methylated polysaccharide is dissolved in refiluxing methyl iodide and silver oxide is added in small portions at several time intervals over 16 h. The reaction mixture i s centrifuged and the solid material is extracted with boiling chloroform three times. The combined centrifugate and washings are evaporated to dryness and the product is examined by IR spectrophotometry to determine the extent of raethylation. Occasionally colloidal silver salts persist as a suspension in the chloroform layer. Evaporation to dryness followed by dissolving in carbon tetrachloride can sometimes break down the colloidal silver salt. If this treatment does not work, one can resort to complexing the silver ion with cyanide and removing Ag(CN>2 i o n b y extraction with water. The chloroform solution is washed several times with water, and the cyanide solution is discarded. Oligo- and monosaccharides are easily methylated by the Hakomori and Kuhn methods. The-procedures-used are s i m i l a r to those described in the preceding paragraph except for the dialysis step due to the fact that oligo- and monosaccharides are dialyzable. Therefore, various procedures for isolation and purification of the products of Hakomori or Kuhn methylation of low molecular weight carbohydrates are now discussed. For the isolation of methyl glycosides of tetramethyl monosaccharides, the methylation mixture is poured into ice water and the glycosides are extracted with petroleum ether (65-110°) which is then washed with small volumes of ice water. The petroleum ether layer is withdrawn, dried over anhydrous sodium sulfate and evaporated to dryness to collect the product. This method can be extended to 20 trimethyl glycosides but not to dimethyl glycosides, because of the higher solubility of the latter in water, and chloroform has to be used instead. Methylation mixtures of d i - and higher oligosaccharides can be poured into ice water then extracted with chloroform which is in turn washed with cold water. Generally i t is very d i f f i c u l t to remove completely dimethyl sulfoxide from the chloroform extract. For less volatile disaccharide (e.g. aldobiouronic acids) and higher oligomers, i t is more convenient to d i s t i l dimethyl sulfoxide under vacuum directly from the methylation mixture, after excess chloroform is removed. The solid residue is dissolved in water which is in turn extracted with chloroform. The organic layer is dehydrated over anhydrous sodium sulfate and evaporated to dryness. Traces of dimethyl sulfoxide remaining can be easily removed by d i s t i l l a t i o n under vacuum for a short period. If the disaccharides are very vo l a t i l e (e.g. sucrose and neutral disaccharides containing deoxysugars) the duration of vacuum d i s t i l l a t i o n should be as short as possible to avoid the loss of methylated sugars. When an oligosaccharide is methylated by the Kuhn method, the reaction mixture is centrifuged and the solid is washed with chloroform. The combined centrifugate and washings are evaporated to remove chloroform and further high vacuum d i s t i l l a t i o n is used to remove N, N-dimethyIformamide. In the course of methylation the sugar unit at the reducing end may undergo equilibration between furanoside and pyranoside. The methylation of the furanoside form may be substantial when the sugar has a stable furanoside isomer (e.g. galactose), and reaction at a low temperature is required to avoid the side reaction. 21 6. Carboxyl reduction of acidic carbohydrates. Reduction of carboxyl groups in acidic carbohydrates can be carried out at four different stages during a structural investigation: a. Direct reduction on native acidic polysaccharides (55). The polysaccharide is f i r s t propionylated with propionic anhydride in pyridine (or pyridine and formamide), followed by treatment with lithium borohydride in anhydrous tetrahydrofuran (THF). The reduction mixture is dialyzed against tap water for 24h and the carboxyl-reduced polysaccharide remaining in the dialysis tubing is freeze-dried. For the reduction of the carboxylic groups in K5 this method was used. b. Lithium aluminum hydride reduction of the methylated acidic polysaccharide. The reduction is carried out in anhydrous tetrahydrofuran or ether. This is one of the approaches to obtaining the exact sugar ratio as well as the nature of the uronic acid, since the neutral polysaccharide is readily hydrolyzable to i t s constituent units, and the neutral sugar derived from the acidic unit is easier to identify both qualitatively and quantitatively than the original uronic acid. c. Carboxyl reduction of acidic oligosaccharides.' The oligosaccharides are esterified either by diazomethane (CH^2) o r b v methylation. The methyl ester of the original oligosaccharide is reduced with sodium borohydride in methanol while the permethylated acidic oligosaccharide is reduced under the same conditions or in anhydrous tetrahydrofuran by lithium aluminum hydride. d. Conversion of uronic acid to the corresponding al d i t o l through the uronolactone (56). D-Glucurono- and D-mannuronolactone can be formed by d i s t i l l i n g concentrated hydrochloric acid or trifluoroacetic acid from the corresponding acids. The lactone is then reduced to the alditol by sodium borohydride in methanol. 22 7. Partial hydrolysis. A series of oligosaccharides from the partial hydrolysis of a polysaccharide is the key to the determination of the sequential arrangement of the constituent monosaccharides in that polymer, for the oligosaccharides reveal the relative position of one building block to another in an additive manner. To achieve the structural elucidation of oligosaccharides methods such as total hydrolysis, reducing end group analysis, progressive partial hydrolysis, methylation, p.m.r. spectroscopy and enzymatic hydrolysis are employed. From the discussion on rates of hydrolysis of glycosidic linkages in section II 3, i t is clear that appropriate conditions (concentration of acid, temperature and length of hydrolysis) can be chosen to effect preferential cleavage of certain weaker linkage(s) and thereby rel e a s i n g preferentially certain subunits of the polymer. For example, K5 was autohydrolyzed at 97° and pH 2.7 (1.2% aqueous solution) to produce a series of acidic oligosaccharides (see page 63, A^.g) and a neutral disaccharide. Hydrolysis of K62 in IM t r i f luoroacetic acid at 97° for Ih yielded-an a-ldotriouronicaacid as the major oligosaccharide plus tetra- and penta-saccharides with very l i t t l e aldobiouronic acid; while hydrolysis of K62 in IM trifluoroacetic acid at 97° for 5h gave an aldobiouronic acid as .the..major, product. When K18 o was heated at 97 in 0.5M trifluoroacetic acid for 0.5h aldotetrao-, aldopentaouronic acid and higher oligomers were produced, while heating K18 under the same conditions for 4h afforded aldobiouronic acid as the major product plus a trace of higher oligosaccharides. At the end of partial hydrolysis the solution was evaporated to dryness, and the residue was twice dissolved in water and again 23 evaporated to dryness to remove most of the trifluoroacetic acid. The partially hydrolyzed mixture can be separated directly by preparative paper chromatography, or applied to a charcoal-Celite (1:1) column using f i r s t water to elute the monosaccharides, which are.the major components of the mixture and interfere, with'the^sep^a'ration of oligosaccharides by 'gel- ' chromatography (57),and then eluting with aqueous ethanol and aqueous isopropanol to recover the oligo-saccharides. The fractions containing oligosaccharides are concentrated and fractionated on a column of Sephadex G 15 using water as eluent. Partially hydrolyzed K5 was separated on f i l t e r paper to obtain two disaccharides, one trisaccharide and higher oligomers. Partially hydrolyzed K62 was separated f i r s t on a charcoal-Gelite column to remove monosaccharides then successfully fractionated on Sephadex G 15 column to afford an aldobiouronic acid, an aldotriouronic acid, an aldotetraouronic acid which was further purified by paper chromatography, and aldopentaouronic acids. The partially hydrolyzed K18 was processed on a charcoal-Celite column to remove the monosaccharides. The oligo-saccharides of K18 were not separable into pure fractions by Sephadex G 15 but were successfully fractionated by preparative paper chromatography to obtain a series of oligosaccharides. To obtain an oligosacchride sufficiently pure for p.m.r. spectroscopic analysis of the anomeric protons, Sephadex (G 15. or G 10) gel f i l t r a t i o n was frequently employed to purify the oligosaccharide which had already been separated by paper chromatography. 24 8. Periodate oxidation (58). Vicinal diols are oxidatively cleaved by periodate resulting in the formation of dialdehydes . In the case of contiguous t r i o l s , formic acid w i l l be derived from the central carbon. The product of periodate oxidation, polyaldehyde, is reduced to a polyol which can be totally hydrolyzed or partially hydrolyzed with acids. Total hydrolysis of the polyol can provide considerable information of the structure of the polymer. Partial hydrolysis of the polyol from the oxidized and reduced polysaccharide was developed by F. Smith and coworkers. The Smith degradation is a powerful tool which can be used to cleave the poly-saccharide specifically at the sugar unit containing vicinal diols to provide fragments which aid in the structural analysis. The cause of the specific cleavage can be explained as follows. When the sugar unit in a polysaccharide is cleaved by periodate ions and subsequently reduced, the resultant polyalcohol is a true acetal and is more susceptible to acid than the glycosidic linkages of the intact sugar unit which is immune to periodate oxidation. (The relative rate (59) for hydrolysis is 10 3 to 10 4 :1). The usefulness of Smith degradation of K62 and total hydrolysis of the polyol is demonstrated in Figure 3. (See page 25) Periodate oxidation is complicated both by over and under oxidation (60-63). Over oxidation can be minimized by using dilute solutions of periodate buffered at a proper pH (e.g. 4), and by keeping the reaction mixture in . the dark at low temperature (4°). Incomplete oxidation can occur due to the formation of formyl esters of the reducing sugars, the formation of intramolecular hemiacetal through the 25 Figure 3. Periodate degradation of Klebsiella K-type 62 polysaccharide (annular protons are not shown). CH2OH CH2OH HCOOH + C HO OH C C°2H HOHoC 9 9 „ , CH20H 2 | 0 / HO J-LQ CH2OH HOH2C COOH H0H2p HOH2C CH20H C H20H HOH2&) total hydrolysis CH2OH partial hydrolysis CH9OH I z CHo0H HO-C-H * 1 CHO COOH CHO °\j-0 - C-H +•• H-C-QH +. H-C - OH t CH2OH 0 y 1 1 CH20H CH9OH CH20H OH Z CH2OH HO CH20H H , O H -+ H O CH 2 0 H •f H-C-OH I .CH2OH C H 2 O H H-fe-OH- C H 2 O H C H O " COOH' j H , O H + H - C - O H H C - O W + H C - Q H . + H C ^ - O H + ( ! ; H 2 O H C H 2 O H C H ^ H C H 2 0 H ' 2o dialdehyde generated by i n i t i a l periodate cleavage, as well as steric hindrance that limits the accessibility of the vicinal diols to periodate ions. As demonstrated in the periodate oxidation of K.62, the side chain unit, as a rule i s oxidized more rapidly than the backbone unit due to the steric factor. While carrying out the periodate oxidation of K5 (Figure 4), a persistent underoxidation was observed, even after the 0-acetyl group on C 2 of the D-glucose unit was removed by saponification. Hydrolysis of the oxidized and reduced K5 gave rise to D-glucose as well as O-mannose, demonstrating that D-glucose was protected through inter-residual hemiacetal by the condensation of the C 3 hydroxyl on D-glucose to the C 2 aldehyde of the degraded D-glucuronic acid. On the other hand, the carboxyl reduced K5 underwent complete reaction with periodate, both the D-glucose and the D-glucuronic acid being oxidized, and thus hydrolysis of the polyol gave rise only to D-mannose. The anomalous underoxidation of the native K5 is similar to the incomplete periodate oxidation of alginate, as well as the extremely slow uptake of periodate by polymer having contiguous I—>4 linked pyranosyl units, such as 1—>k linked galactan of lupin seeds, maize-cob xylan, and amylose. Painter and colleagues have reported the study of the under-oxidation of alginate (61). Three repetitions of periodate oxidation followed by reduction were required to reach the theoretical oxidation, li m i t . The cleavage of the C2-C3 diol to dialdehyde in the 1—>4 linked polymer 27 Figure 4. Periodate degradation of Klebsiella K-type 5 polysaccharide (annular protons not shown). 2 8 followed by the formation of a 6-membered dioxane-type stable inter-residue hemiacetal ring (see Figure 4) provide complete protection against the oxidation of the immediate neighboring sugar units. No intra-residue 6-membered hemiacetal involving the C6 group i s possible at C6, since this is a carboxyl group and not a primary alcohol. It should be noted that an acidic group may, through i t s inductive effect, possibly contribute to the pronounced st a b i l i t y of the protective inter-residue hemiacetal- a property less remarkable in a neutral polysaccharide. In the cases of 1—*4 linked D-glucan and D-galactan, circumstantial evidence points to an equilibrium favouring inter-residue instead of intra-residue hemiacetal, judging by the sluggish periodate uptake of those polysaccharides. The intra-residue hemiacetal leaves the neighboring sugar units free and susceptible to periodate oxidation. In the case of 1—»4 linked neutral polysaccharides, the intr i n s i c instability of the inter-residue hemiacetal and the existance of the competitive intra-residue hemiacetal in the equilibrium result in the ultimate theoretical periodate uptake. 9. Gas-liquid chromatography The application of gas-liquid chromatography (g.l.c.) to the analysis of hydrolysis products of a polysaccharide has provided a qualitative as well as a rapid, accurate and quantitative analysis of the constituent sugars in the polymer. In order to make g.l.c. applicable to carbohydrates, efforts have long been made to render the sugars volatile by suitable derivatization. Sweeley and colleagues (64) introduced O-trimethylsilyi groups to the sugar molecule using a very convenient reaction to provide derivatives of suitable v o l a t i l i t y . Each 29 monosaccharide may give rise to a maximum of four peaks through anomeric and ring isomerizations thus yielding a complex chromatogram. An elegant method was devised by Gunner et a l . (65) by converting a sugar into i t s alditoi acetate which has suitable v o l a t i l i t y and gives only a single peak on g . l . c , corresponding to the parent sugar. The isolation of the alditoi acetates and part i a l l y methylated alditoi acetates has been improved by Sawardeker and co-workers (67) using a copolymer of ethylene glycol succinate polyester and a n i t r i l e silicone polymer (ECNSS-M) as the liquid phase. Lindberg and colleagues have published a very informative review (68) on gas-liquid chromatography and mass spectrometry in methylation analysis of polysaccharides. An extensive review of g.l.c. analysis of methylated methyl glycosides was reported by Aspinall (69) using butane -1,4-diol succinate polyester (BDS) and the less polar polyphenyl ether (m-bis- (m-phenoxyphenoxy)-benzene) as liquid phases. In summary, g.l.c. analysis of sugar derivatives can be stated as follows: various derivatizations of the reducing sugars and their alditois by methylation, trimethylsilylation, or acetylation, or, i f necessary, a combination of these methods and subsequent separation of the derivatives using various liquid phases or a combination of liquid phases to achieve the best separation. Applications of these principles w i l l be discussed. In the analysis of the partially methylated a l d i t o i acetates of K18, the separation of the acetates of 3,4 -di-O-methyl-L-rhamnitol, 2,4-di-0-methy1-L-rhamnitol, 2,3,4,6-1e tra-0-me thy1-D-glucitol and 2,4,6-tri-O-methyl-D-galactitol (mixed with 2,6-di-O-methyl-D-glucitol) was achieved by an Apiezon L column. Further separation of the mixture 30 of 2,4,6-tri-CMnethyl-D-galactitol triacetate and 2,6-di-0-methyl-D-glucitol tetraacetate was effected on a 3% ECNSS-M column. In the case of K24 (32), the acetates of 2,4,6-tri-0-methyl- and 3,4,6-tri-0-roethyl-D-hexitols were not resolvable on columns of ECNSS-M and BDS, so that the methylated sugars were pertrimethylsilylated. Separation was then effectively achieved on ECNSS-M and BDS columns. These tri-O-raethyl-D-hexoses can also be readily separated as their acetates. The effluent^of each component separated by g.l.c. can be collected from the outlet into a capillary tube. As a rule, alditol acetates can be identified by the melting points, while the identification of part i a l l y methylated alditol acetates can be accomplished by mass spectrometry, or melting points: j f they are crystallizable. Generally, sufficient quantities of partially methylated alditol acetate can be collected from g.l.c. for mass spectrometric analysis by raethylating ca. 2mg of an oligosaccharide. While trimethyl-silylated disaccharides (70) are readily separable by g . l . c , a direct application of g.l.c. - mass spectrometry on a permethylated sodium borodeuteride reduced aldotriouronic acid has been demonstrated (29). G.l.c. and g.l.c.-mass spectrometry which are capable of analyzing a sample on a microgram scale have a great potential for the structural analysis of minute amounts of biological material containing carbohydrate, such as glycoproteins (71,72)and glycolipids. Glycoproteins are widely believed to be the receptor site on the c e l l membrane for various biological functions of the c e l l . Understanding of the structure of c e l l membrane and i t s biological function are of primary importance in the f i e l d of l i f e science. 31 10. Mass spectrometry of partially methylated alditoi acetates Of the three variations of mass spectrometer available today, i.e., electron impact (E.I.) (73), chemical ionization (G.I.) (74) and f i e l d ionization (F.I.) (75) mass spectrometers, only electron impact spectrometry which gives rise to desirable fragmentations of the sugar derivatives was employed in the present study. C.I. and F.I. mass spectrometry can produce large fragments and molecular peaks, thus fa c i l i t a t i n g the molecular weight determination. F.I. is particularly suitable for nonvolatile compounds such as free sugars. Heyns et a l . (76) reported the results of mass spectrometry of fu l l y methylated glycosides, while mass spectrometry of trimethylsilylated ethers of partially methylated glycosides was investigated by Kochetkov et a l . (77), and by Samuelson et a l . (78). Chizhov and colleagues (79) have reported their study of mass spectra of alditoi acetates and determined that primary fragments from al d i t o i acetates arise by fission between two adjacent carbon atoms in the molecule, and either fragment can carry the p o s i t i v e charge. Secondary fragments are derived from the primary fragments by elimination of acetic acid (60) and/or ketene (42). Intensity of primary fragments increased with the decreased size of the fragment. These investigators have also shown that the intensities of the primary fragmentations from the aldi t o i acetates are much lower than the intensities of the corresponding peaks from the partially methylated alditoi acetates. Therefore, the fission of the carbon-carbon bond in structures I and II takes place preferentially as compared to that in structure III. 32 I I l H—C — OCH, H-C-OCH, H-G-OCOCH, I 3 I 3 I 3 H—C—OCH, H-C-OCOCH. H-C—OCOCH. I 3 j 3 , 3 I II III Moreover, no fragments containing vicinal methoxy are observed unless there are three pairs of vicinal methoxy groups in the same molecule and two of them are cleaved. This indicates the preferred fission of structure I to II. Through systematic study of the fission patterns of partially methylated ald i t o l acetates, Lindberg (68) has made the following generalizations: 1. Derivatives with the same substitution pattern (e.g. 2,3, 4,6-tetra-O-methyl derivatives of hexitols) give very similar mass spectra, typical of that substitution pattern. 2. The base peak in almost a l l determined spectra is m/e 43 + \ (CH 3C""0). 3. Primary fragments are formed by fission between carbon atoms in the chain. Fission between a methoxylated and an acetoxylated carbon is preferred over fission between two acetoxylated carbons. The fragments with the methoxyl groups carry the positive charge because the positive ion is stabilized by the methoxyl group. The alditol acetate derived from 2-0-methyl-D-mannose therefore only gives two main primary fragments. ™ 2 _ _ _ 0 A C „ 33 333 117 CH3o — G • I-- H A c0 -1 C — H H — C 1 - 0 A c H — 1 G | - 0 A C 1 C H2GAC 4. When the molecule contains two adjacent methoxylated carbons, the fission between them is the preferred reaction and both fragments are found as positive ions. CH 0— OA-I 2 HC 0A_ 1 HC OCH3 189 131 CH3O— CH A.O-CH *» 1 CH3 5. The secondary fragments are derived from the primary fragments by single or consecutive elimination of acetic acid (60), ketene (42), methanol (32), or formaldehyde (30). The important fragments according to Lindberg's designation are li s t e d below: A. m/e 45: The lowest molecular weight primary fragment which is formed by a molecule with a methoxyl group at C I. B. m/e 59: Formed from 6-deoxyhexitols when position 5 is methylated. C. m/e 89: Obtained from the ald i t o i with positions 1 and 2 methylated. This is the exception mentioned above. This species w i l l 34 give rise to (B) after elimination of formaldehyde. D. m/e 117: Derived from an alditol when C 1 is acetylated and C 2 methylated. E. m/e 131: Obtained from a 6-deoxyhexitol when C 4 is methylated and C 5 acetylated. F l . m/e 161: Obtained in low intensities from alditols methylated at positions 2 and 3. It may become prominent when C 4 is also methylated. F2. m/e 161: Always prominent when C 1 and C 3 are methylated. The secondary fragments having m/e 129 and m/e 101 are obtained from Fl and F2 by loss of methanol (32) and acetic acid (60), respectively. Further loss of ketene (42) from m/e 129 gives m/e 87, and loss of formaldehyde (30) from m/e 101 might give m/e 71. G. m/e 175: Obtained from a 6-deoxyhexitol methylated at positions 2,3 and 4. H. m/e 189: From alditols methylated at C 3 but not C 1 and C 2. The secondary fragment m/e 129 may be formed by elimination of acetic acid, and the fragments m/e 99 and m/e 87 by further loss of formaldehyde and ketene, respectively. I. m/e 203: From a 6-deoxyhexitol, methylated at C 3 but not at positions 4 or 5. KI. m/e 205: Observed in low intensities when positions 2, 3, and 4 are methylated. When C 5 is also methylated i t w i l l be obtained in high intensities. K2. m/e 205: From alditols methylated at positions 1,2,4 and 5. K3. m/e 205: From alditols methylated at positions 1,3,4 and 5. An important peak, m/e 145, is obtained from K3 by loss of acetic acid. 35 L. m/e 233: Obtained from alditois methylated at positions 1 and 4. A prominent secondary peak, m/e 113, is obtained from (H) by loss of two molecules of acetic acid. M. m/e 261: Derived from alditois methylated at C 3 and acetylated at C 4, 5 and 6. CH 2 = 0 CH, HC = 0 - CH, + HC = 0 CH„ + HC = 0 CH, CH, H2C 0 - CH, H2C - OCOCH3 A 45 B 59 C 89 D 117 HC + 0 CH, HC = 0 CH, HC i - CH. HC - OCOCH- HC - OCH„ HC - OCOCH, CH, H2C - OCOCH3 HC - OCOCH. CH, E 131 Fl 161 I 203 + HC = 0 - CH. HC - OCH, HC - OCH, I 3 H2C - 0C0CH3 Kl 205 + 0 HC I HC - 0C0CH I HC - OCH. I 3 H2C CH, OCH, K2 205 HC = 0 - CH, I HC - OCH„ HC - OCOCH I 3 H2C - 0 - CH3 K3 205 + HC = 0 - CH, HC - OCOCH, + HC = 0 - CH, I HC - OCOCH. HC - OCOCH, I ' H2C - 0CH3 L 233 HC - OCOCH, H0C - OCOCH, M 261 On reduction some pairs of methylated sugars, e.g., a 3-0-methyl- and a 4-0-methyl-hexose, give rise to aldit o i acetates with the same substitution pattern. This can be overcome i f the reduction is carried out with sodium deuterioborate. CHDOA I HC -. OA. CHDOA HC - OA. 261 CH30 - CH 190 A„0 - CH HC - OA HC - OA. 262 HC - OCH, 189 CH 2 - OAc HC - 0A„ I H2C - 0AC This same labelling technique can be used in the reduction of carboxylic group of the uronic. acid moiety. The origin of the reduced neutral sugar can readily be distinguished by mass spectrometry. The prominent primary peaks of mass spectrometry of various methylated aldi t o i acetates are compiled in Table III. 11. Determination of anomeric configurations Optical rotations have long been employed to provide an indication of the overall anomeric configurations of a polysaccharide, although i t provides no information concerning the individual sugar residues in the polymer. However, the optical rotation of a series of oligosaccharides obtained by partial hydrolysis of a polysaccharide together with the result of specific enzymatic hydrolysis, may reveal the anomeric configuration relating to each sugar unit. Recently, use has been made of the fact that oligosaccharide aldi t o i acetates containing j8 - linkages can be oxidized by chromium trioxide in acetic anhydride 37 TABLE III. PRIMARY FRAGMENTS (a) IN THE MASS SPECTRA OF PARTIALLY METHYLATED SUGARS IN THE FORM OF THEIR ALDITOL ACETATES. i m/e Position of CH3 45 59 89 117' 131 161 Pentoses 2 (4) X 3 2,3 (3,4) X 2,4 X 2,5 X X 3,5 X X 2,3,4 X X 2,3,5 X X X Hexoses 2 (5) X 3 (4) 6 X 2,3 X 2,4 (3,5) X 2,5 X 2,6 X X 3,4 3,6 X 4,6 X X 5,6 X X 2,3,4 X X 2,3,5 X 2,3,6 X X 2,4,6 X X X 2,5,6 X X X 3,4,6 X X 2,3,4,6 X X X 2,3,5,6 X X X 6-Deoxyhexoses 2 X 3 4 X 2,3 X 2,4 X X 3,4 X 2,3,4 X X X 2,3,5 X X 3,6-Dideoxyhexo ses 2 X 2,4 X X X X X X X X X X X X X X X X (a) Primary fragments having mass numbers higher than m/e 261, which are formed by fission between a methoxyl and an acetoxyl group, are of low intensity and are not included in the Table. 38 much faster than those containing cv - linkages (80), providing a means for differentiatingoC - from jS - configurations. The non-destructive technique of p.m.r. spectroscopy (27,50) has been applied to oligo- as well as polysaccharides to obtain definitive information on the anomeric configuration of each sugar residue, and i t was extensively used in this study. Generally, the proton at the axial position of C. 1 in an aldppyranoside^ (equivalent toj8 -D-1inkage) resonates at higher f i e l d c than the proton at equatorial position (50,51) (equivalent totx -D-linkage). Both types of anomeric proton have larger chemical shifts (S*) than the rest of annular protons. Furthermore, spin-spin coupling between the protons at C 1 and C 2 can provide useful information. Thus, in a trans-diaxial interaction as in the case of ft - D-glucosyl and >5-D-galactosyl linkages, ^ i s about 7-9 Hz, while J± 2 i s about 1-3 Hz for CX -D-glucosyl and <x -D-galactosyl linkages. The chemical shifts of anomeric protons are 4,6- 4.8 T* for CX -D-linkages and 5.0- 5.5 T for j8 -D-linkages. The carbohydrate hydroxyl protons have to be exchanged in deuterium oxide to eliminate the interference by hydroxyl protons. The deuterium exchange is never complete and.the DOH peak covers partially the anomeric protons of ^tf-D-glycosidic linkages in the region of 5-6 T . The DOH peak can be shifted upfield by heating, or downfield by cooling or by the addition of trifluoroacetic acid. The DOH peak can also be eliminated by means of Fourier transform p.m.r. spectroscopy (81) making use of the longer spin-lattice relaxation time (T^) of DOH as compared to other sugar protons. In the p.m.r. spectral analysis of oligo- and polysaccharides, the sample was heated to ca. 95° to shift the DOH peak upfield. Heating 39 has another advantage by reducing the viscosity of the slime poly-saccharides to give sharper resonance bands. 12. Determination of D- and L-configuration of a sugar residue The D-or L- configuration of individual sugars can be determined by their specific oxidases (e.g., D-glucosidase and D-galactosidase) or by the sign of their circular dichroism curves, measured on suitable derivatives were performed at 213 mu on alditol acetates or partially methylated alditol acetates which were routinely collected from g.l.c. (82) on a milligram scale. Compounds of known configuration have to be used for comparison. Uronic acid can be converted to i t s lactone for c d . measurement at 219 mu. 13. Immunochemical methods (97,98) The cross-reactions occuring among microbial antigens have been extensively employed for both classification of microorganisms and for diagnostic purposes, but only some of these cross-reactions have lent themselves to exploration of the chemical nature of cross-reactions. This largely depends on the availability of the purified antigenic material, such as capsular polysaccharide, lipopolysaccharide etc. and has been widely studied in Salmonella. Klebsiella. Escherichia and P n e u m oCOCCUS . Heidelberger (100) has developed methods consisting of the precipitation reactions of antibody (precipitin) and antigen. This immunochemical reaction is called the precipitin test. When a purified polysaccharide is used as the antigen and cross-reactions are carried out against antisera induced by polysaccharides of known and related structure.information on the nature of the sugar moieties, types of 40 linkage and configuration in the unknown polysaccharide may be obtained. The precipitin test involving polysaccharides as antigens consists in the e l i c i t i n g in appropriate animals of antisera to Pneumococcus and Salmonella capsular polysaccharides of known structure. Thus a spectrum of antisera exists each with different s p e c i f i c i t i e s for studies of the antigenic cross-reactions of polysaccharides of unknown structure on submicro scale, i.e. amounts which are too small for chemical investigations. The degree of cross-reaction measured by the amount of precipitation is an indication of the degree of similarity of the structure of the unknown to the structure of the polysaccharide used for e l i c i t i n g the immune serum for the test. A series of precipitin tests has to be performed to deduce the fine partial structure of the unknown. Heidelberger and Dutton (3) have reported immunochemical studies on Klebsiella capsular K - type 5, 18, 62, 7, 20, 21 and 24, using anti-Pneumococci and anti-Salmonella sera. The structure of the polysaccharides used for e l i c i t i n g antisera (3) and cellobiouronic acid, the important antigenic determinant, are l i s t e d as follows. Pneumococcal type II (Pn 'II) — 3 L-Rhap — L-Rhap — D-GlcApi~^-Glcp-L-Rhap_ - — = - = - = " D-GlcAp Pneumococcal type III (Pn III) 41 Pneumococcal type V (Pn V) — - D-GlcAp i - 1 L-FucNAcp i - 2 - D-GlcAp i - 3 - L_-FucNAcp - — ~ ~ ~ 4 ~ 14 i / 1 1 A L-PneNAcp Ut D - Glc L-PneNAcpA-it D-Glcp Pneumococcal type VIII (Pn VIII) * D-GlcAp i ~ D_-Glcp ~ p D-Glcp ~ g-Galp Cellobiouronic acid D-GlcAp — D-Glc = - ^ = The repeating unit of K5 consists of C H 3 ^OOH 4 1 4 L_3 * A 6 1 D-GlcAp D-Glcp g D-Manp „ ~ ~ „ I ~ 5 with pyruvic acid linked 2 '0AC ' 4,6- to D-mannose, and -0 A c at position 2 on D-glucose. The fact that anti-P n III precipitates K5 to a greater extent than by using anti-P n VTII confirms that the cellobiouronic acid unit is l-*3 linked to the next sugar as in the capsular polysaccharide of Pneumococcal type III (P n III) rather than l->4 linked as in the capsular polysaccharide of Pneumococcal type VIII (P n VIII), The antibody precipitated from anti-P n III was demonstrated to be part of that reactive with P n VIII, and the cross-reaction is known to be due to the repeating units of cellobiouronic acid in the antigenic determinants of both types. Polysaccharide from K5 before and after deacetylation precipitate the same fraction and the same amount of antibody from anti-P n III. This confirms that the -0AC group is not on the D-glucuronic acid, which is the predominant antigenic determinant, but is attached to the units of D-glucose. The cross-reactions also verify the identification of the reactive portions of the glucuronic 42 acid and glucose as the D-isomers. The polysaccharide from K18 has a repeating unit consisting of 4 1 2 1 3 1 3 1 4 1 — ~ D-GlcAp L-Rhap — L-Rhap ~ D-Glap ~~ D-Glcp ~^T~ ~ ^ ~ ~ ~ - ~ ~ & ~ 13-D-Glcp with D-glucose as the branch unit at position 3 on D-glucose. An antiserum to P n III does not precipitate K18, and this mixture s t i l l gives rise to f u l l precipitation by K5, confirming that the constituent D-glucuronic acid and D-glucose in K18 are not combined as cellobiouronic acid as in P n III and K5. The capsular material from K62 consists of a repeating unit of — - D-Glcp ~r D-GlcAp — D-Man L-1 D-Galp i 3 l l D-Manp and precipitates heavily in anti-P n V, li g h t l y in anti-P n II and not at a l l in anti-P n VIII. This demonstrates that D-glucuronic acid in K62 is linked l-*2 as in P n V and not in the form of cellobiouronic acid as in P n VIII nor in the form of non-reducing end groups, because poly-saccharides with multiple non-reducing end groups of D-glucuronic acid usually react much more strongly in anti-P n II than in anti-P n V. 43 III. THE STRUCTURE OF THE CAPSULAR POLYSACCHARIDE FROM KLEBSIELLA K- TYPE 5 44 SUMMARY Methylation, periodate oxidation and partial hydrolysis studies on the capsular polysaccharide, and on the carboxyl reduced polymer of Klebsiella K5 show the structure to consist of a repeating unit (36,37). CH3 ,COOH 4 1 4 1 3 1 D-GlcAp D-Glcp -yr D-Manp -jg-2-0Ac The polysaccharide has a molecular weight (by gel f i l t r a t i o n ) of 9 X 10"* and is the f i r s t Klebsiella capsular polysaccharide to be; found lacking a carbohydrate side chain. The proportion of 0-acetyl and pyruvate groups was determined by p.m.r. spectroscopy. DISCUSSION The capsular polysaccharide of Klebsiella K5 was shown by Nimmich (12) to contain D-glucose, D-mannose, D-glucuronic acid, and D-galactose, and also pyruvic acid in a subsequent personal letter. D-Galactose was proved to be the contamination from the lipopolysaccharide. A culture of K5 (E5051) was obtained from Dr. I. 0rskov, Copenhagen, as an agar slant and was grown for 24 h in a 3% sucrose-yeast-extract broth culture which was then evenly spread on four large trays of thin and solid 3% sucrose-yeast-extract agar (30g sucrose, 2g Bacto yeast-extract, 15g agar, 1 1 water, 2g sodium chloride, lg potassium dihydrogen phosphate, 0.25g magnesium sulfate 'TR^ O, and 0.5g calcium sulfate) and incubated at room temperature for 3 days to reach a maximum growth of slime. Ceils were harvested, and diluted with 5 volumes of water containing 17» phenol and centrifuged at 65,000g for l h . The crude polysaccharide was extremely viscous and i t was convenient to purify the material in small batches by centrifugation and precipitation with Cetavlon. The 45 precipitated acidic polysaccharide was collected and the supernatant was dialyzed for 48h and freeze-dried, giving neutral polysaccharides which contain mainly D-galactose linked at C 3. The neutral poly-saccharides were not further investigated. The Cetavlon precipitate was dissolved in IM sodium chloride and precipitated with ethanol by slowly decanting the viscous solution into 5 volumes of ethanol with vigorous st i r r i n g . The stringy precipitate was dissolved in water and dialyzed against tap water for 24h. The nondialyzable material was deionized with cation exchange resin Amberlite 1R-120 (H +), and subsequently dialyzed against d i s t i l l e d water for 24h, then freeze-dried to provide a fl u f f y white material. The acidic polysaccharide purified by Cetavlon, deionization and freeze-drying migrated as a single band on electrophoresis on cellulose acetate and had (o()D(23) -45.2° (c, 0.4, water). The molecular 2 5 weight by gel f i l t r a t i o n on Sagavac 6F was about 9 X 10 . The poly-saccharide was found by titration to have a neutralization equivalent of 233. Two carboxylic acids and one acetate group to three hexose units require a neutralization equivalent of 203. o The p.m.r. spectrum of the polysaccharide in D£0 run at 90 showed signals at T 8.50 (3H) characteristic of pyruvate, andT7.93 (3H) characteristic of 0-acetyl (50,51), along with the signals of anomeric protons a t T 5.15 (IH) and 5.50 (2H), suggesting that the repeating unit of the polysaccharide contains three monosaccharides, a l l in j8 -linkages, with one pyruvate and one 0-acetyl attached to them. The presence of pyruvic acid was demonstrated by paper chromatography (83) of the acid hydrolyzate of the polysaccharide and by preparation of the 2,4-dinitro-46 phenylhydrazone (84). When the polysaccharide was subjected to methanolysis both acetate and pyruvate were detected on paper by the ferr i c chloride-hydroxamic ester test (52). Acid hydrolysis of the polysaccharide gave D-glucose, D-raannose, D-glucurone, D-glucuronic acid and an aldobiouronic acid. The f i r s t two compounds were identified as their crystalline alditoi acetates and found to be in a ratio of 3:5, suggesting that D-glucose might be the neutral sugar at the reducing end of the aldobiouronic acid and thus incompletely released. When a portion of the hydrolyzate was saponified to hydrolyze any lactone, reduced, rehydrolyzed, and processed to avoid contamination of the neutral sugars by lactones, D-glucose and D-mannose were found in a ratio of 1:1. D-Glucuronolactone was saponified in 0.5M sodium hydroxide and reduction with sodium borohydride was carried out in the same solution. Half an hour to one hour is sufficient for complete reduction unless the sugar is substituted at C3 (56). The treatment of D-glucuronolactone in 0.5M sodium hydroxide should not exceed I h otherwise the extent of epimerization w i l l be noticeable. Another sample of K5 polysaccharide was methanoiyzed, the uronic ester was reduced and the product was hydrolyzed to give D-glucose and D-mannose in a ratio of 2:1. The glucose and mannose were shown to be of D-configuration by the circular dichroism curve of the aldito i acetates; D-glucuronic acid was similarly authenticated by the c.d. of glucurone and of 2,3-di-0-methyl-D-glucitol acetate obtained in the methylation study (82). The p.m.r. spectrum of the fu l l y methylated polysaccharide showed the absence of acetate but confirmed the presence of pyruvate in the ratio of one mole per three sugars. The neutral crystalline sugars 47 obtained on hydrolysis were identified as 2,3,6-tri-O-methyl-D-glucose (99) and 2-0-methyl-D-mannose (85) and were present in approximately equal amounts. The acidic component on reduction and hydrolysis gave 2,3-di-and 2,3,6-tri-O-methyl-D-glucose. These results were confirmed by the analysis of the methylated carboxyl-reduced polysaccharide which gave 2,3,6-tri-O-methyl-D-glucose, 2-0-methyl-D-mannose, and 2,3-di-0-methyl-D-glucose in a ratio of 1:1:1. Autohydrolysis of the original capsular polysaccharide removed the pyruvic acid ketal and methylation of the recovered polymeric material followed by hydrolysis gave 2,4,6-tri-0-methyl-D-mannose in place of the 2-methyl ether thus demonstrating the linkage of pyruvic acid to positions 4 and 6 of D-mannose. The identity of 2,4,6-tri-0-methyl-D-mannose was confirmed by i t s crystalline al d i t o i acetate. Partial hydrolysis of the original polysaccharide yielded a neutral oligosaccharide (N3) and a series of acidic oligosaccharides (A^.g). The compound crystallized and was identified as S-O-^-D-glucopyranosyl)-D-mannose by hydrolysis of the corresponding alditoi and by methylation. The compound A^ was found by ti t r a t i o n to be an aldobiouronic acid and was shown to be cellobiouronic acid by hydrolysis, reducing end analysis, methylation, and conversion to the crystalline hepta-O-acetyl-Go'-D-glucuronopyranosyl methyl ester)-c* -D-glucopyranose (m.p. 248°) (86). Compound analyzed as an aldotriouronic acid which was characterized as 3-0-(,5 -D-cellobiosyluronic acid)-D-mannose. The higher acidic oligosaccharides were not examined but were assumed to be aldotetra-, penta-, and hexauronic acids. Periodate oxidation of the original capsular polysaccharide resulted in the consumption of 0.35 mol of periodate per sugar residue 48 and the derived polyalcohol gave D-glucose in addition to D-mannose, on hydrolysis. Oxidation of de-0-acetylated capsular polysaccharide resulted in the i n i t i a l uptake of 0.35 mol of periodate per sugar residue but when the polyaldehyde was reduced and the product was again treated with periodate further reaction took place to give a total consumption of 0.65 mol of periodate per sugar residue. Oxidation of carboxyl-reduced and de-O-acetylated polysaccharide prepared as described for sapote gum using lithium borohydride in tetrahydrofuran (55) consumed 0.60 mol of periodate per sugar unit. Hydrolysis of the last two polyalcohols failed to reveal the presence of D-glucose. The results are readily interpreted in the light of work by Painter and colleagues (61,62). These authors have shown that in polysaccharides containing 1,4-linkages inter-residue hemiacetal linkages may be formed readily. Such linkages inhibit the consumption of the theoretical amount of periodate until they are destroyed by reduction. In a l l oxidized polysaccharides an equilibrium exists between the acyclic aldehyde and the cyclic hemiacetal (61). In the present example the presence of the uronic acid appears to favor the cyclic structure whereas in the reduced polysaccharide oxidation is not inhibited. Sloneker and co-workers (87) have also commented on similar anomalous oxidations. The polyaldehyde arising from the complete oxidation of de-O-acetylated polysaccharide was reduced and hydrolyzed. Paper chromatography in solvent A indicated the presence of components having the mobilities of D-mannose, erythritol, L-erythronolactone, and pyruvic acid. The material having the mobility of erythritol was eluted (fraction 3) and appeared homogeneous in solvent A. However, when this 49 fraction was chromatographed in solvent B prior to development in solvent A two components were detected one of which moved slower than erythritol and gave a positive response for lactone (52) while the other appeared to be erythritol. Hydrolysis of fraction 3 gave D-mannose, erythritol, L-erythronolactone, and thus this fraction contains 3-0- (4,6-0-(l-carboxyethylidene)-D-mannopyranosyl } -L-erythronolactone which, coincidentally, has the same mobility as erythritol in solvent A. The formation of this lactone is to be expected on periodate oxidation of a polysaccharide having the structure proposed below. CH3 COOH 4/\6 * D-GlcAp — D-Glcp ~ D-Manp ~ -The results of methylation, partial hydrolysis, and periodate oxidation are consistent with the proposed repeating unit and the only feature not yet accounted for is the position of the 0-acetyl group. This was located using the methyl vinyl ether method of de Belder and Norrman (88). Methylation and hydrolysis of the substituted poly-saccharide showed on paper chromatography a small spot corresponding to mono-methyl hexose in addition to a large quantity of D-glucose and D-mannose. G.l.c. of the hydrolyzate in the form of alditoi acetates gave a mono-methyl fraction which crystallized and which was identified as 2-0-methyl-D-glucito1 pentaacetate by mixed m.p. with an authentic sample (82). Low yields of methylated sugars have been obtained using this method on other capsular polysaccharides (32), presumably on account of their high molecular weight and steric hindrance. The structure of the capsular polysaccharide of Klebsiella K5 may therefore be represented by the repeat unit: CH3 COOH ^ </ 50 4 1 4 1 3 V * 1 D-GlcAp —g- D-Glcp ~jf~ P_-Manp 2-OAc The characteristic p.m.r. signal at about 7 8.5 of the CH\j — C of the pyruvate ketal and i t s derivatives persists with only slight change in chemical shift through the typical transformations of methylation, carboxyl reduction, and periodate oxidation and is thus a convenient means of checking that no inadvertent loss of this group occurs at any stage during the structural investigation of the polysaccharide. The presence of a cellobiouronic acid residue in the structure of K5 polysaccharide suggests that there should be a cross reaction with Pneumococcal III and VIII antisera (98 ) and such reactions have been found (3). The reaction is stronger with anti-P n III than with anti?P n VIII consistent with the cellobiouronic acid being linked to position 3 of the next residue in Pneumococcus type III rather than to position four as in Pneumococcus type VIII (9 8 ) . 51 EXPERIMENTAL A l l evaporations were c a r r i e d out under reduced pressure and at a bath temperature not exceeding 40° (except as noted). O p t i c a l r o t a t i o n s were measure d at 23* 1 on a Perkin Elmer model 141 pol a r i m e t e r . M e l t i n g p o i n t s were taken on a Fisher-Johns apparatus and are uncorrected. A n a l y t i c a l paper chromatography was c a r r i e d out on Whatman No. 1 f i l t e r paper. Chromatograms were developed using s i l v e r n i t r a t e - s o d i u m hydroxide f o r reducing and non-reducing sugars and _p_-anisidine i n 1-butanol f o r reducing and methylated sugars. Solvent.A, e t h y l a c e t a t e - a c e t i c a c i d - f r o m i c acid-water (18:3:1:4); B, e t h y l a c e t a t e - p y r i d i n e - w a t e r (8:2:2.); 6, 1-butanol-acetic acid-water (2:1:1); D, methyl e t h y l ketone-water azeotrope; E, phenol-water (4:1); F, 1-butanol-ethanol-water (4:1:5). Preparative paper chromatography was c a r r i e d out on Whatman No. 3 f i l t e r paper. Thin l a y e r chromatography was c a r r i e d . o u t using s i l i c a g e l as the absorbent. G a s - l i q u i d chromatography was c a r r i e d out on a F and M 720 instrument using columns: a, 4 ' x W 37„ ECNSS-M on 100/120 mesh Gas Chrom Q, b, 8' x V 5% buta n e d i o l succinate on 80/100 mesh Diatoport S, c, 8' x V 207, SF 96 on 80/100 mesh Diatoport S, and d, 8'x.^" 207. Apiezon L on 60/80 mesh Diatoport S. Peak areas were determined w i t h an I n f o t r o n i c s CRS-100 e l e c t r o n i c i n t e g r a t o r . P.m.r. spectra were recorded on Varian T-60 and HA-100 instruments using t e t r a m e t h y l s i l a n e as e x t e r n a l standard except as noted. I n f r a r e d spectra were recorded on a Perkin-Elmer model 457- i n f r a r e d spectrophotometer. Preparation and P r o p e r t i e s of K5 Capsular Polysaccharide A c u l t u r e of K l e b s i e l l a K5 (E5051) was obtained from D r . I . 0rskov, Copenhagen, and grown on 37, sucrose-yeast-extract agar (30 g 52 sucrose, 2g Bacto yeast-extract, I5g agar, 1 1 water, 2g sodium chloride, lg potassium dihydrogen phosphate, 0.25g magnesium sulfate ^h^O, and 0.5g calcium sulfate)at room temperature for 3-4 days. The growth was scraped off the plates to give 600 ml of slime from 1 1 of agar medium. The slime was worked-up in portions in the following manner: each 50ml portion of the slime was diluted six-fold with water containing 1% phenol, then centrifuged at 65,000 X gravity for lh to obtain a clear, slightly yellow solution which was freeze-dried for storage or to which was added 45 ml of 3% hexadecyltrimethy1ammonium bromide (Cetavlon). The acidic polysaccharide which precipitated was dissolved in sodium chloride (35ml, 2M) and precipitated with five volumes of ethanol. The precipitate collected by centrifugation was dissolved in water , treated with Amberlite resin 1R-120 (H*), dialyzed against d i s t i l l e d water for 24h, then freeze-dried to give capsular polysaccharide ( 0.63 ) per 50ml of slime. This material had (o<)n -45.2° (c 0.44, water). The equivalent weight was found by titration with 0.01M sodium hydroxide (phenolphthalein) to be 233. The molecular weight was determined by applying the polysaccharide ( i mg in 1 ml IM 2 sodium chloride) on a column of Sagavac 6F (60 X 0.9 cm, eluent: IM sodium chloride, flow rate: 1.5 ml/h). Electrophoresis of the poly-saccharide was carried out on cellulose polyacetate at pH 8.8 and 300V for 30 min, to show a migration of 3.8 cm. Ash content n i l ; nitrogen n i l . The supernatant from the precipitation of the acidic poly-saccharide with hexadecyltrimethy1ammonium bromide was dialyzed against tap water for 48lf and freeze-dried, yield O.lg. This material was 53 hydrolyzed to give D-galactose as the major component, along with D-glucose and D-mannose. Methylation of the neutral polysaccharides showed the presence of 2,4,6-tri-0-methyl-D-galactose as the major sugar unit. The neutral polysaccharides were not further investigated. To a portion of the Cetavlon purified acidic polysaccharide (0.89g) in water (50ml) was added sodium hydroxide (50ml, 2M) and the solution was dialyzed against tap water for 24h, passed through Amberlite resin 1R-120 (H4"), dialyzed against d i s t i l l e d water, and freeze-dried, to give deacetylated polysaccharide (0.8g). Total Hydrolysis of the Acidic Polysaccharide The polysaccharide (8mg) was heated in sulfuric acid (0.5M) at 98° for 8h (or with trifluoroacetic acid, 2M, 98°, lOh). The cooled solution was neutralized (BaCOj), centrifuged, and decationized with Amberlite 1R-120 (H +). Paper chromatography using solvent B(8h), then solvent A(10h) revealed, by the silver nitrate-sodium hydroxide method, two major spots for D-mannose and D-glucose, along with a small fast moving spot with the same mobility as D-glucuronolactone and two small slow moving spots corresponding to D-glucuronic acid and an aldobiouronic acid. Acidic components were removed by passage through a column of Duolite A-4(0H") anion exchange resin and the f i l t r a t e was concentrated and reduced with sodium borohydride. Borate was removed as the volatile methyl borate by passage through Amberlite 1R-120 (H +) and d i s t i l l a t i o n thrice with methanol. The alditols were dissolved in a 1:1 mixture of acetic anhydride and pyridine and heated at 100° for 20 min. The acetylation reagents 54 were evaporated with the addition of several portions of ethanol. The alditoi acetates, injected in chloroform onto column a programmed from 185-210° at l°/min, with a helium flow rate of 60 ml/min, gave peaks identical to authentic standards of the hexacetates of D-mannitol (20min) and D-glucitol (24min) in a ratio of 5:3. Samples were collected and the m.p, and mixed m.p. of D-mannitol hexaacetate was 121-123°, and that of D-glucitol hexaacetate was 93-95°. The c.d. curves of both hexaacetates were positive and identical to that of standard samples. The glucuronic acid from several hydrolyzates was separated on paper and evaporated with the addition of hydrochloric acid. The c.d. curve of authentic D-glucurone and the isolated material were both positive. In a second experiment the hydrolyzate was processed to obtain complete hydrolysis of the aldobiouronic acid and to avoid contamination of the neutral sugars by D-glucurone. The hydrolyzate was treated with sodium hydroxide (0.5M) then with sodium borohydride. The reaction mixture was allowed to stand at-room temperature for lh then hydrochloric acid was added to a concentration of 2M and the solution was heated at 100° for 4h. The hydrolyzate was evaporated to dryness, d i s t i l l e d three times with methanol, made alkaline with sodium hydroxide (0.5M), and reduced with sodium borohydride. The solution was passed successively through columns of Amberlite 1R-120 (H+) and Duolite A-4(0H~) to remove the sodium ion and the L-gulonic acid. The neutral effluent was evaporated to dryness, acidified with acetic acid, and d i s t i l l e d with methanol. The ratio of D-mannitol to D-glucitol as hexaacetates was found to be 1:1. 55 K5 polysaccharide (lOmg) was heated in a sealed tube with methanolic hydrogen chloride (4%, 1ml) for lOh. After neutralization with silver carbonate and centrifugation, the clear supernatant was treated with sodium borohydride (lOrag) for 24h. Borate was removed and the residue was hydrolyzed with trifluoroacetic acid (2M) at 100° for 8h. Analysis gave the ratio of D-mannitol hexaacetate (m.p. and mixed m.p. 120-122°) to D-glucitol hexaacetate (m.p. and mixed m.p. 93-95°) as 1:2. Characterization of Pyruvic and Acetic Acids The acidic polysaccharide (30mg).was hydrolyzed with hydrochloric acid (IM, 2ml) in a sealed tube at 100° for 3h. Paper chromatography (solvent A) showed a fast moving spot with the same mobility as pyruvic acid (R.£ 0.7) and having the characteristic fluorescence of pyruvic acid under UV when the chromatograra was sprayed with a solution consisting of equal volumes of 0.2% ethanolic o-phenylenedi-amine and 207. aqueous trichloroacetic acid. The hydrolyzate was extracted with ethyl ether (3 X 15ml). The extract was neutralized with dilute sodium hydroxide solution and evaporated to dryness. The residue was dissolved in a small volume of water and treated with a slight excess of 2,4-dinitrophenylhydrazine (in 2M hydrochloric acid). The solid formed was collected by centrifugation and dissolved in benzene. T.l.c. was carried out in toluene-acetic acid-water (4:3:1) on s i l i c a gel. The major spot (Rf 0.41) was identical in color and mobility with that of authentic pyruvic acid 2,4-dinitropheny1hydrazone. The polysaccharide (20mg) was heated in methanolic hydrochloride 56 (47., 1ml) in a sealed tube in a steam bath for 2 h. To the hydrolyzate methanolic hydroxylamine hydrochloride (IM, 1.5 ml) and methanolic sodium hydroxide (IM, 1.5 ml) were added in that order. After centrifugation, the supernatant was concentrated to ca. 0.5ml which was examined chromatographically using the solvent system 1-butanol-acetic acid-water (4:1:5). The chromatogram was dried and sprayed with an aqueous solution of 1% hydrochloric acid and 1% fe r r i c chloride, giving characteristic spots corresponding to the hydroxamates of the methanolyzed products of D-glucuronolactone (Rf 0.16 and 0.22), pyruvic acid (R^ Q ^ a n ( j Q j^)f ar>d acetate from D-glucose pentaacetate (Rf 0.56). Faint yet distinct spots from pyruvic acid as compared to intensive acetic acid spot indicated the pyruvic acid in the poly-saccharide was mostly intact under the methanolysis conditions. Proton Magnetic Resonance Spectra of Original arid De-O-acetylated Polysaccharides. " Capsular polysaccharide (ca. 20mg) was dissolved in D20 and freeze-dried. After two further treatments, the 100 MHz spectrum was run at 90° and showed sharp singlets at T^8.50 (pyruvate) and 7.93 (acetate) together with signals at7T5.50 (2H) and 5.15 (IH). The spectrum of deacetylated material was similar with the absence of the peak a t t 7.93. The 60 MHz spectra were run on deacetylated polysaccharide at pH 1 and 11, showing the chemical shifts of pyruvate CH^ at?" 8.50 and 8.55, respectively. Methylation of Original Capsular Polysaccharide To vacuum dried polysaccharide (400mg) in anhydrous dimethyl 57 sulfoxide (20ml) was added an excess of methylsulfinyl carbanion prepared from l g sodium hydride (55% o i l dispersion) according to the method of Conrad and Sandford (15). The flask was sealed with a serum cap and shaken for 12h. An excess of methyl iodide (4ml) was added slowly to the flask while sti r r i n g and keeping the temperature below 20°. The reaction mixture was dialyzed against tap water for 24h and extracted with chloroform (2 X 30ml). The chloroform layer was dried over anhydrous sodium sulfate and evaporated to dryness, giving 100 mg of methylated polysaccharide which showed no hydroxyl absorption in the IR spectrum. The aqueous layer was freeze-dried (300mg) and showed no hydroxyl absorption in IR after three Purdie methylations. The p.m.r. spectrum (chloroform-d) of the methylated polysaccharide showed ^8.5 (3 H, pyruvate CH 3 ) 6-7(40H, calcd. 43H for ring protons and 0CH 3»s), and T~5.55 and 5.44 (3 H, anomeric protons). The methylated polysaccharide (300mg) was digested in sulfuric acid (72%, 3ml) with ice cooling for 2h then diluted with water (24ml) and refluxed for 4h. The cooled hydrolyzate was neutralized with barium carbonate and the supernatant was collected after centrifugation. The precipitate was washed twice with a small amount of water. The combined supernatant and washings were concentrated to a small volume, filter e d , passed through successive columns of Amberlite 1R-120 (IT1") resin and Duolite A-4(0H") resin. The solution was evaporated to dryness, giving 178mg of neutral sugars. Acidic sugars were recovered by eluting the Duolite A-4(0H") resin column with 10% formic acid which was then evaporated to dryness, giving 60mg of an acidic mixture. 58 The neutral sugars were chromatographed using solvent D. The chromatogram showed two spots: the fast moving spot (R f 0.55) was identical to 2,3,6-tri-O-methyl-D-glucose in mobility and characteristic color response, and the slower running spot (R^ 0.07) was in the region of monomethylated sugars. To the water solution of neutral sugars (lOmg) sodium borohydride (15mg) was added. The reduction was allowed to proceed for 24h, then worked up as usual. The methylated alditois were acetylated and the methylated alditoi acetates were dissolved in a small volume of chloroform and injected onto column a at 180° (helium flow rate 88 ml/min) giving peaks corresponding to the alditoi acetates of 2,3,6-tri-0-methyl-D-glucose (6.4 min) and a monomethylated sugar (18 min) in 1:1.2 ratio. Samples were collected and mass spectra were taken. The component at R^  6.4 min gave a fragmentation pattern identical to that from the alditoi acetate of 2,3,6-tri-O-methyl-D-glucose while the second component (Rj, 18 min) gave a fragmentation pattern similar to the a l d i t o i acetate of 2-0-methylaldoses. The neutral sugars (170mg) was separated on f i l t e r paper in solvent D. Guide strips cut from the chromatogram were developed with p-anisidine trichloroacetate and the corresponding sections of the two components were cut out from the paper chromatogram, eluted with water, and the water solutions were concentrated to dryness. The component at R^  0.55 (58rag) was crystallized twice from ethyl ether, giving m.p. 119-121°, mixed m.p. 117-119° with authentic 2,3,6-tri-O-methyl-D-glucose. The second component (R^ 0.07) was a syrup (40mg) which was seeded with authentic 2-0-methyl-D-mannose. The crystals were collected 59 after removal of the surrounding syrup by cold ethanol. The ethanol solution was concentrated, seeded and cooled to give clusters of o platelets. The crystals were recrystallized from ethanol, m.p. 139-140 , mixed m.p. 138-140° with the authentic 2-0-methyl-D-mannose. The acidic mixture (5mg), eluted from Duolite A-4(0H~) resin, was refluxed .for ,6h with methanolic hydrogen chl o r i d e ( 3%, 3ml ). neutralization (silver carbonate) and centrifugation, the methanol solution was concentrated to dryness. Injection in chloroform, onto column b at 170° and using a helium flow rate of 75 ml/min, gave characteristic peaks corresponding to methyl 2,3,6-tri-0-methyl-D-gluco-pyranoside (4.4 min (m), 6.0 min (s) ) and methyl 2,3-di-0-methyl-D-glucuronic acid methyl ester (11.5 min (m), and 13.5 min (s) ). A. portion of the acidic mixture (5mg) was refluxed with methanolic hydrochloride (3%, 5ml) for 4h. The cooled reaction mixture was neutralized with silver carbonate and centrifuged. The combined centrifugate and washings were evaporated to dryness and treated with sodium borohydride (lOmg in 10 ml water). The reduction was allowed to proceed for 24h. The excess sodium borohydride was destroyed by adding dilute acetic acid and the solution was passed through a column of Amberlite IR-120 (H +) resin, then evaporated to dryness. The residue was dissolved in trifluoroacetic acid (2M, 1ml) and the solution was heated on a steam bath for 4h. The reaction mixture was evaporated to dryness. The hydrolyzate was examined by paper chromatography in solvent D, giving, two spots corresponding to 2,3,6-tri-O-methyl-D-glucose (R f 0.56) and 2,3-di-0-methyl-D-glucose (Rf 0.28). The same hydrolyzate was further treated with sodium borohydride (lOmg in 0.5ml water) and worked up as usual. Injection in chloroform, onto column a 60 at 200° and with a helium flow rate of 86 ml/min, gave two peaks i d e n t i c a l i n retention time to the a l d i t o l acetates of 2,3,6-tri-O-methyl-D-glucose (6.1 min) and 2,3-di-0-methyl-D-glucose (11.9 min). Samples' were c o l l e c t e d and the mass spectra were run and the fragmentation patterns were i d e n t i c a l , to that of a l d i t o l acetates of 2,3,6-tri^-methyl-D-glucose and 2,3-di-^-methyl-D-glucose, r e s p e c t i v e l y . A solution of methylated polysaccharide (0.17 g) in anhydrous tetrahydrofuran (20 ml) was added dropwise with s t i r r i n g to a suspension of lithium aluminum hydride (0.3 g) in tetrahydrofuran (30 ml) at room temperature. The reaction mixture was s t i r r e d for 18h more, then the excess lithium aluminum hydride was destroyed by cautious addition of a c e t i c a c i d . The whole reaction mixture was evaporated to dryness and the residue was extracted with b o i l i n g chloroform (3 X 30ml). The combined extracts were washed with water, dried over anhydrous sodium su l f a t e , then evaporated to dryness, giving 0.13g of reduced methylated polysaccharide. The p.m.r. spectrum (chloroformrd) showed the signals atX8.6 (pyruvate CH3,3H),6-7 (rin g protons and OCH^ protons, 40 H), 5.3-5.6 (anomeric protons, 3H). Two methyl esters at"c"6.2-6.4 were absent. The methylated carboxyl-reduced polysaccharide (56mg) was digested i n formic a c i d (907«) at 97° for l h , concentrated to dryness, and then hydrolyzed i n 2M t r i f l u o r o a c e t i c acid on a steam bath for 4h to give, by paper chromatography in solvent D, three spots of s i m i l a r i n t e n s i t y and m o b i l i t i e s i d e n t i c a l to 2-C|-methyl-D-mannose (Rf 0.10), 2,3-di-0-methy 1-D-glucose (R f 0.30), and 2,3,6-tri-0-methyl-D-glucose (Rf 0.58). A portion of the hydrolyzate (36mg) was chromatographicaHy 61 separated in solvent D. The guide strips were cut out from the chromatogram and developed by p-anisidine spray. The areas corresponding to the three components were cut out and eluted with water. Each fraction was concentrated to a syrup. The fraction from Rf 0.10 (6mg) was crystallized from ethanol, m.p. 137-139°, undepressed/ with authentic 2-0-methyl-D-mannose. The fraction from Rf 0.58 (9.7mg) was crystallized from ethyl ether, m.p. 113-115°, undepressed with authentic 2,3,6-tri-0-methyl-D-glucose. The fraction from Rf 0.30 (lOmg) had the same mobility as 2,3-di-O-methyl-D-glucose. The ald i t o l acetate was prepared and i t s retention time in g.l.c. analysis and the mass spectrum were both identical to the alditol acetate of authentic 2,3-di-O-methyl-D-glucose. Another portion of the hydrolyzate (lOmg) was reduced with sodium borohydride. After the usual work up the methylated alditols were acetylated and injected in chloroform onto column a programmed from 160 to 190° at 2°/min. Using a helium flow rate of 86 ml/min three peaks corresponding to the alditol acetates of 2,3,6-tri-0-methyl-D-glucose (16.0 min), 2,3-di-0-methyl-D-glucose (22.4 min), and 2-0-methyl-D-mannose (26.1 min) were obtained in a ratio of 1:1:1. Methylation of degraded polysaccharide A degraded polysaccharide with the pyruvic acid group removed was prepared by autohydrolysis of the original polysaccharide (0.5g) in 60ml aqueous solution-(pH 2.5) in a *'steam bath for 20h. The hydrolyzate was dialyzed against d i s t i l l e d water (2 X 300ml);T The:material in the dialyzing tubing was freeze-dried, giving degraded polysaccharide (0.25g) which did not show the pyruvate CH- in i t s p.miri^spectrum. 62 To this polysaccharide (200mg in 10 ml dimethyl sulfoxide) methylsulfinyl carbanion prepared from 0.5g sodium hydride was added at a temperature below 20°. The mixture was shaken overnight, and an excess of methyl iodide (3ml) was added while the reaction mixture was kept below 20°. The mixture was dialyzed against tap water for 24h, then the material in the dialyzing bag was freeze-dried. The freeze-dried methylated polysaccharide (170mg) was made anhydrous by evaporation to dryness with a small volume of benzene, and showed a small hydroxyl absorption band in the IR spectrum (in carbon tetrachloride). It was thus methylated by the method of Purdie to a permethylated polysaccharide without absorption at 3500 era"*. A portion of the methylated polysaccharide (30mg) was hydrolyzed with trifluoroacetic acid (2M, 8ml) on a steam bath for 8h. The solution was evaporated to dryness and the residue was dissolved in a small volume of water and examined chromatographically in solvent D. The chromatogram showed a major spot corresponding to trimethyl hexoses (Rf 0.55) and a faint spot corresponding to tetramethyl hexoses (Rf 0.78). The remaining hydrolyzate was treated with sodium borohydride (40mg) and the reaction mixture was allowed to stand at room temperature for 24h and worked up as usual. The partially methylated alditois were acetylated and injected in chloroform onto column a, programmed from 170 to 200°at 2°/min with a helium flow rate of 67 ml/min, giving two major peaks with retention times identical to those of the alditoi acetates of 2,4,6-tri-0-methyl-D-mannose (16.1 min) and 2,3,6-tri-O-methyl-D-glucose (18.8 min) along with a small peak corresponding to the alditoi acetate of 2,3,4,6-tetra-O-raethyl-D-glucose (9.3 min). 63 Samples were collected and the fraction at 16.1min was crystallized from petroleum ether (30-60°), giving crystals of m.p. 64-66°, undepressed with the authentic alditol acetate of 2,4,6-tri-0-methyl-D-mannose. Partial Hydrolysis of Capsular Polysaccharide Acidic polysaccharide (3.5g) was dissolved in water (300ml, i n i t i a l pH2.7), heated on a steam bath for 12h, and the solution was dialyzed against d i s t i l l e d water (2 X 600 ml). The non-dialyzable material was rehydrolyzed and the dialyzates were, combined and concentrated. Only a small amount of product was obtained and chromatography in solvent A revealed pyruvic acid (Rf 0.7), D-glucose, D-mannose, and two components having R^an 1.2 and 1.7 with traces of other compounds. The solution containing the non-dialyzable material was concentrated to 150ml, heated on a steam bath for 24h, and dialyzed against d i s t i l l e d water (2 X 500 ml). The process was repeated for a total of 148 h heating with periodic concentration to keep the volume below 300ml. The dialyzates were combined, concentrated, and passed through successive columns of Amberlite IR-120 (H +) and Duolite A-4 (OH") resins. The neutral components were fractionated by preparative paper chromatography in solvent B to give (30mg), ^ (20mg) and (15mg, R_, 0.36), and the components eluted from Duolite A-4 (OH") column —klc with 107. formic acid were separated on paper in solvent C to give Aj-A^. The neutral compounds and N 2 were chromatographically identical to D-mannose and D-glucose and were not further examined. The oligosaccharides were analyzed as follows. 64 Structural study of compound N3 Compound crystallized from water and had m.p. 164-166°, (°9D .7.3° (c 2.18, water) which agrees with the value (m.p. 165°) given for 3-0-^-D-glucopyranosyl-D-mannose (89). Compound N-j was hydrolyzed in t r i f luoroacetic acid (2M) at 100° for 6h. The hydrolyzate was evaporated to dryness and the residue was dissolved in a few drops of water and chromatographed in solvent B, giving two spots identical to the mobilities of D-mannose and D-glucose. Another portion of was reduced with sodium borohydride in water, worked up as usual then hydrolyzed in trifluoroacetic acid (2M) for 4h at 100°. The hydrolyzate was evaporated to dryness then dissolved in a few drops of water and examined on a duplicate f i l t e r paper using solvent E. One of the chromatograms was developed with silver nitrate-sodium hydroxide, revealing two spots with the mobilities identical to D-glucose and D-mannitol (R^^ 1.07). The other chromatogram was sprayed with p_-anisidine trichloroacetate, revealing only the spot equivalent to D-glucose. Compound N3 (5mg) was methylated according to the method of Hakomori , then hydrolyzed (2M trifluoroacetic acid, 100°, 4h) and concentrated to dryness. The residue was dissolved in a few drops of water and chromatogrammed in solvent D, revealing two spots of similar intensity and mobility identical to 2,3,4,6-tetra-O-methyi-D-glucose (R f 0.78) and 2,4,6-tri-O-methyl-D-mannose (Rf 0.56). The rest of the hydrolyzate was reduced with sodium borohydride (8mg) and acetylated with acetic anhydride-pyridine (1:1) in the usual way. Injection of the partially methylated a l d i t o i acetates in chloroform onto column a at 185° and using a helium flow rate of 60 ml/min, gave two peaks, in 1:1 ratio identical in retention times to the alditol acetates of 2,3,4,6-tetra-O-methyl-D-glucose (6.5 min) and 2,4,6-tri-0-methy1-D -mannose (13.1 min). The p.m.r. spectrum of the original N-j in Dr>0 was run after two deuterium-exchanges, showing signals at "Z'5.46 (J^ 2 6.0 Hz, IH), 5.l6(0.3H)and 4.81 ( J t 2 1Hz, 0.7H). The structure of N.j is therefore 3-0-^ -D-glucopyranosyl-D-mannose. Structural study of A^ Compound A^ (58mg, R^ c^ 0.23 solvent A) had an equivalent weight of 394 and£<x)D+9.1 (c 0.8, water) in agreement with the value given for cellobiouronic acid, (<x) D + 7. 0 2 - ° , water) (90). Compound A^ was hydrolyzed in sulfuric acid (IM) on a steam bath for 4h, The hydrolyzate was neutralized (barium carbonate), deionized with Amberlite IR-120 (H+) resin, concentrated, and examined by paper chromatography using solvent A. The chromatogram revealed two spots with mobilities identical to D-glucose and D-glucuronic acid. Compound A^ was dissolved in methanol (1ml), an excess of diazomethane (in ethyl ether) was added and after several minutes the solution was evaporated to dryness. The residue was dissolved in methanol (2ml) and sodium borohydride (lOmg) was added. After I6h the reaction mixture was worked up as usual. The reduced sugar was then hydrolyzed in sulfuric acid (0.5M) on a steam bath for 6h. The hydrolyzate was examined on duplicate f i l t e r papers using solvent E. One of the chromatograms was developed with s i l v e r nitrate-sodium hydroxide, revealing the presence of two spots with mobilities identical to D-glucose and D-glucitol. The other chromatogram was sprayed with p-66 anisidinetrichloroacetate and heated, revealing only a spot corresponding to D-glucose. The rest of the reduced and hydrolyzed sugars were concentrated to dryness, dissolved in pyridine (1ml), and hexamethyldisilazane (0.5ml) and trimethylsilylchloride (0.25ml) were added in that order. The solution was injected onto a column of SE52, programmed from 160 to 220° at l°/min and using a helium flow rate of 75 ml/min, to give two peaks with retention times identical to the trimethylsilyl derivatives of D-glucose (23.3 min (m ); 27.6 min (s) ), and one peak identical to the retention time of the trimethylsilyl derivative of D-gludLtol (25.6 min). The ratio of D-glucose to D-glucitol was 1:1. To a portion of Aj (lOmg) N, N-dimethyl formamide (2ml), methyl iodide (2ml) and silver oxide (0.6g) were added. The reaction mixture was stirred for 24h at room temperature, then more silver oxide (0.2g) and methyl iodide (0.2ml) were added. The reaction mixture was further stirred for 24h, filtered, and concentrated to dryness. The methylated sugar was refluxed with methanolic hydrochloride (4%, 6ml) for lOh. The cooled solution was neutralized(silver carbonate), centrifuged, and the supernatant was evaporated to dryness. The methyl glycosides were dissolved in a small volume of chloroform and injected onto column b at 150° and with a helium flow rate of 50 ml/min, giving two peaks characteristic of methyl 2,3,6-tri-O-methyi-D-glucoside (10.3 min (w), 13.8 min (s) ) and another two peaks characteristic of methyl 2,3,4-tri-0-methyl-D-glucuronic acid methyl ester (15.2 min (w); 21.2 min (s) ). To the rest of the methyl glycosides in methanol (2ml), sodium borohydride (15mg) was added. The reduction was allowed to proceed overnight and worked up as usual. The reduction mixture was hydrolyzed 67 by sulfuric acid (0.5M, 1ml) in a water bath of 100° for 6h. The hydrolyzate was examined by paper chromatography using solvent D. The chromatogram revealed two spots with the same mobilities as 2,3,6-tri-O-methyl-D-glucose (R f 0.58) and 2,3,4-tri-O-methyl-D-glucose (R f 0.51) along with faint slow moving spots. Compound (22ma) was dissolved in methanol (5ml) and an excess of diazomethane in ethyl ether was added. After a few minutes the solution was concentrated to a syrup which, in turn, was dissolved in acetic anhydride (3.5ml) containing zinc chloride (0.2g). The mixture was kept at 60° for 90 min. The dark solution was concentrated and chromatogrammed on f i l t e r paper using solvent D. The proper section was cut out and eluted with chloroform and the solution was evaporated to dryness. The residue was crystallized from ethanol, m.p. 248°, (cx) D +31.5°(c 0.1, chloroform), ( l i t . (86) 252° and|VjD +44° for hepta-0-acetyl-(^ -D-glucuronopyanosyl methyl ester)-o( -D-glucopyranose ). The p.m.r. spectrum of A^ was taken in D20 after two deuterium exchanges. The signals of the anomeric protons were as follows: T 5.44 (J , 0 7.5 Hz, IH), 5.31 (J. _ 7.9 Hz) and 4.78 (J, 3.3 Hz, IH combined). 1»* 1,2 Enzymatic hydrolysis of A^ (lmg in 0.2ml of c i t r i c buffer, pH 5) with ,<?-glucuronidase (lmg) at 37° was complete in 12h to give, by paper chromatography in solvent B then in A, D-glucuronic acid and D-glucose. Aj is therefore cellobiouronic acid. Structural study of Compound A„ (41mg, R . , , 0.65, solvent C) had r 2 °' -cellobiouronic acid an equivalent weight of 559 and (o<)D -24.2° (c 3.1, water). A 2 was 68 hydrolyzed in sulfuric acid (IN) on a steam bath for 8h. The hydrolyzate was neutralized (barium carbonate^, deionized with Amberlite IR-120 (H +) resin, and concentrated to a syrup which was examined by paper chromato-graphy using solvent B (8h) then A(28h). The chromatogram revealed a large spot corresponding to D-mannose, two spots corresponding to D-glucose and D-glucuronic acid, and a spot corresponding to Aj. A portion of A 2 (4mg) was dissolved in water (0.5ml) and reduced with sodium borohydride (8mg). The solution was allowed to stand for 24h. The excess sodium borohydride was destroyed with dilute acetic acid and the solution was passed through a column of Amberlite IR-120 (H +) resin, and worked up as usual. The reduced sugar was hydrolyzed in sulfuric acid (0.5M) on a steam bath for 16h. The cooled hydrolyzate was neutralized (barium carbonate), deionized with Amberlite IR-120 (IT*") resin and concentrated to a small volume which was examined by duplicate paper chromatography using solvent E. Half of the chromatogram, after drying and washing with ethyl ether to remove the phenol, was developed with silver nitrate-sodium hydroxide and revealed the presence of spots corresponding to cellobiouronic acid (RQ^ c 0.34), D-glucuronic acid (RJGIc 0.45), D-glucose, and D-mannitol (RQ^ c 1.2).the The other half of the chromatogram was developed with p-anisidine trichloroacetate and revealed a l l the spots above except D-mannitol. A portion of A^ (8mg) was methylated according to the method of Conrad and Sandford (15). The methylated sugar was hydrolyzed with hydrochloric acid (2M, 2ml) on a steam bath for 5h then evaporated to dryness at 20°. The hydrolyzate was chromatogrammed in solvent D, revealing an acidic component at the origin and a spot with mobility 69 similar to 2,3,6-tri-O-methyl-D-glucose or 2,4,6-tri-0-methyl-D-mannose (Rf 0.57). The remaining hydrolyzate was reduced with sodium borohydride for 24h. The reduction mixture, after the excess sodium borohydride was destroyed with dilute acetic acid, was passed through successive columns of Amberlite IR-120 (H+) and Duolite A-4 (0H&) resins. After the usual work-up, the partially methylated alditols were acetylated and analyzed on g . l . c , giving two peaks identical to the retention times of the a l d i t o l acetates of 2,4,6-tri-0-methyl-D-mannose and 2,3,6-tri-O-methyl-D-glucose in a ratio of 1:1. Samples were collected and the mass spectra were identical to those of the ald i t o l acetates of 2,4,6-tri-0-methyl-D-mannose and 2,3,6-tri-O-methyl-D-glucose, respectively. P.m.r. of A 2 in D20 showed "7/5.47 (J^ 2 7.3 Hz, IH), 5.43 ( J L 2 7.5 Hz, IH), 5.23 (Jj 2 1.2 Hz, 0.4H), and 4.88 (Jj 2 1.7 Hz, 0.6H). Carboxyl Reduction of K5 (55) To a solution of the acidic polysaccharide (0.8g) in formamide (60ml) pyridine (35ml) was added, and propionic anhydride (10ml) was then added dropwise into the stirred solution. After 48h the reaction mixture was poured into ice water containing 2% hydrochloric acid (500ml). The precipitate was collected by centrifugation and washed with ice water then dissolved in a small volume of acetone which was in turn poured with sti r r i n g into petroleum ether (30-60°, 500ml). The fine powder was collected by centrifugation and dissolved in pyridine (30ml), then propionic anhydride (8ml) was added dropwise into the solution. After 48h the reaction mixture was worked up as described above to yi e l d the propionylated polysaccharide (0.83g). 70 The propionylated polysaccharide (0.83g) in tetrahydrofuran (100ml) was treated with ethereal diazomethane (100ml) at -75°. The solution was stirred at -75° for one more hour then poured into petroleum ether (30-60°, 500ml), yield 0.89g. To a refluxing solution of the propionylated polysaccharide methyl ester (0.89g) in tetrahydrofuran (100ml), lithium borohydride (2g) in tetrahydrofuran (50ml) was added dropwise during a period of one hour. The refluxing was allowed to continue for 18h. The excess lithium borohydride was destroyed by careful addition of water to the ice-cooled, stirred reaction mixture. The mixture was dialyzed against tap water for 24h, deionized with Amberlite 1R-120 (H"1") resin and freeze-dried (0.4g). The p.m.r. spectrum showed a peak at ~C 8.60 but the integration suggested that some of the ketal grouping had been lost during the reduction procedure. The carboxyl reduced polysaccharide (8mg) was hydrolyzed in sulfuric acid (0.5M, 1ml) on a steam bath for lOh. After neutralization (barium carbonate)and deionization, the hydrolyzate was examined by paper chromatography in solvent B then in A. Spots corresponding to D-mannose and D-glucose along with traces of slower moving components were revealed on the chromatogram. The hydrolyzate was dissolved in a small volume of water and treated with sodium borohydride (15mg) and allowed to stand for lh. The excess sodium borohydride was destroyed by adding dilute acetic acid and the solution was worked up as usual. The alditols were acetylated with acetic anhydride-pyridine (1:1 v/v). After the removal of the excess acetylating agent the alditol acetates were dissolved in a small volume of chloroform and injected 71 onto column a, programmed from 180 to 220 at 1 /min and using a helium flow rate of 60 ml/min, giving two peakswith the retention times identical to those of the alditoi acetates of D-mannose (31.4 min) and D-glucose (37.5 min) in 1:2 ratio. Samples were collected and the f i r s t peak gave crystals of m.p. 119-121°, undepressed with the authentic D-mannitol hexaacetate; the second peak gave crystals of m.p. 93-95°, undepressed with the authentic D-glucitol hexaacetate. The carboxyl reduced polysaccharide (0.25g) was methylated according to the method of Conrad and Sandford (15). The methylation mixture was dialyzed against tap water for 24h and the material in the dialysis bag was extracted with chloroform (2 X 60ml). The aqueous layer was freeze-dried while the chloroform layer was evaporated to dryn and methylated twice according to the method of Purdie to give the methylated polysaccharide free of a hydroxyl band in the IR spectrum. G.l.c. analysis of the component sugars of the methylated polysaccharide on column a programmed from 160 to 190° at 2°/min and a helium flow rate of 86 ml/min, gave two peaks corresponding to alditoi acetates of 2,3,6-tri-O-methyl-D-glucose (16 min) and 2-0-raethy1-D-mannose (26 min) in a ratio of 2:1. Periodate Oxidation of K5 and Modified K5 Carboxyl reduced polysaccharide (ll.2mg) was dissolved in sodium periodate (0.0125M, 10ml) and kept at 1° in the dark. One ml of the solution was withdrawn and diluted to 250ml with d i s t i l l e d water for the absorbance measurement at 222.5 rap for the unreacted periodate ion. A limiting uptake of periodate (0.6 mole per sugar 0 72 unit) was reached in 40 h. The reaction was stopped by adding a few drops of ethylene glycol. After 30 min, the reaction mixture was dialyzed against tap water for 24h. The material in the dialysis tubing was treated with sodium borohydride (20mg). After 24 h of reaction the reduction mixture was dialyzed against tap water for 24 hf deionized with Amberlite IR-120 (H +) resin, freeze-dried, and evaporated to dryness three times with methanol. A portion of the carboxyl reduced polyol was hydrolyzed in sulfuric acid (0.5M) on a steam bath for 8 h. Paper chromatography of the hydrolyzate in solvent A revealed the presence of two majorspots with mobilities identical to those of erythritol and D-mannose along with a faint spot corresponding to D-glucose. p-Anisidine trichloro-acetate spray on a duplicate chromatogram revealed only the spot corresponding to D-mannose. Partial hydrolysis of a portion of the carboxyl reduced polyol was carried out in R^SO^ (0.5M) at room temperature. The hydrolyzate was examined on paper chromatogram in solvent A after 4 h and 26 h of hydrolysis. Both chromatograms showed the same hydrolysis pattern. Namely, major spots corresponding to erythritol and a disaccharide ( ^ j a n 0.48, solvent A), along with very faint spots at D-glucose and D-glucuronic region as well as traces of slower moving spots. p-Anisidine trichloroacetate spray failed to reveal any spots on the chromatogram. Original acidic polysaccharide (0.12g) was dissolved in o sodium periodate (0.05M, 40ml). The oxidizing mixture was kept at 1 in the dark. One ml of the solution was withdrawn and diluted to 1000ml with d i s t i l l e d water for the measurement of periodate absorbance 73 at 222.5 mjJ. The periodate uptake leveled off after 288 h to reach a limiting value of 0.35 mole of periodate consumed per sugar unit. Ethylene glycol (0.5ml) was added to the oxidizing mixture which was dialyzed against tap water after 30 min. The material remaining in the dialysis bag was treated with sodium borohydride (O.lg). The mixture was allowed to stand for 24h, then dialyzed against tap water for 24h, deionized with Amberlite IR-120 (H+) resin, freeze-dried, and evaporated to dryness with methanol three times at 20°, giving 92 mg of polyol. The p.m.r. spectrum of the polyol in D20 after three deuterium exchanges showed the presence of pyruvate CH3. The ratio of pyruvate CH3 to the ring protons (excluding anomeric protons) found from the „ integration of the peak areas was 3:18 (calculated 3:20). The polyol (lOmg) was hydrolyzed in sulfuric acid (0.5M) at 100° for 8h and the hydrolyzate was examined on paper chromatogram using solvent A. The chromatogram revealed the spots corresponding to D-mannose ( ^ j a n D, glucose (EQ\C erythritol ( J ^ j a n 2.04), and erythronolactone ( J ^ j a n 3.30). The hydrolyzate was further treated with sodium borohydride (16mg) and the reaction mixture was worked up in the usual way. The alditols were acetylated and the alditol acetates were injected in chloroform onto column a, programmed from 170 to 220° at l°/min. With a helium flow rate at 66 ml/min peaks identical in retention times to the alditol acetates of erythritol (7.6 min), D-mannose (36.1 min) and D-glucose (40.9 min) in a ratio of 2:2.7:1 were obtained. Samples were collected and the fraction corresponding to erythritol tetraacetate crystallized readily, giving m.p. 81-83°, undepressed with authentic erythritol tetraacetate. The fraction at 74 JL, 36.1 min gave m.p. 119-121°, undepressed with the authentic D-mannitol hexaacetate. The fraction Ry 40.9 min gave m.p. 93-95° undepressed with the authentic D-glucitol hexaacetate. Deacetylated polysaccharide (0.665g) was dissolved in sodium periodate (0.05M, 200 ml) and l e f t at 1° in the dark. Periodate uptake reached a limiting value of 0.35 mole per sugar unit after 288h of oxidation. To the oxidizing mixture ethylene glycol (lml) was added and after 30 min the solution was dialyzed against tap water for 24h. The material in the dialysis tubing was treated with sodium borohydride (0.6g) and the reduction was allowed to proceed for 24h. The solution was dialyzed against tap water for 24h then deionized with Amberlite IR-120 (H4") resin and freeze-dried. The f l u f f y white powder was evaporated to dryness with methanol three times at 20°. The resulting polyol was further oxidized with sodium periodate (0.05M, 100ml) for 120h to reach the limiting periodate uptake per sugar unit of 0.3 mole. After treatment with ethylene glycol (0.5ml) the oxidizing mixture was dialyzed against tap water for 24h. The material remaining in the dialysis tubing was treated with sodium borohydride (0.4g) for 24h, dialyzed against tap water for 24h, decationized and freeze-dried. The white powder was evaporated to dryness at 20° with methanol three times, yielding 0.45g of twice oxidized and reduced polyol. A portion of the polyol was hydrolyzed in hydrochloric acid (2M, 0.2ml) in a sealed tube in a steam bath for 4h. The hydrolyzate was directly applied to a f i l t e r paper which was subsequently developed in solvent A. The chromatogram revealed three main spots with mobilities identical to those of D-mannose, erythritol (R^^ 2.0), 75 erythronolactone ( J ^ ^ 3.3), and a faint spot corresponding to pyruvic acid (^Man.^*0). ' l^ ie s P o t ^ a n 3 , 3 0 w a s s n o w n t o b e a lactone by the hydroxylamine-ferric chloride spray (52). The rest of the hydrolyzate was evaporated to dryness and treated with sodium borohydride for lh. The excess of sodium borohydride was destroyed with dilute acetic acid and the solution was worked up as usual. The polyols were acetylated and the aldi t o i acetates were dissolved in a small volume of chloroform and injected onto column a, programmed from 175 to 220° at l°/min, and using a helium flow rate of 75 ml/min. Two peaks corresponding to peracetates of erythritol (5.2 min) and D-mannitol (27.9 min) were obtained. The rest of the polyol (ca. 0.45g) was chromatogrammed in solvent C. The fractions corresponding to £ e ry ti M rj_ t o]_ 1»2, 1«0» and the origin were cut out by the aid of guide strips. Each fraction was eluted with d i s t i l l e d water and freeze-dried. Fraction 3 (72mg) was examined by paper chromatography in solvent B (8h) then solvent A (8h). The chromatogram revealed a component streaking behind erythritol and giving a characteristic color reaction with a lactone spray (52). A portion of fraction 3 was hydrolyzed in hydrochloric acid (2M, 0.2ml) in a sealed tube in a steam bath for 4h. The hydrolyzate was directly applied on f i l t e r paper and separated in solvent A, giving spots identical in mobilities to the authentic compounds D-mannose, erythritol (R$j a n 2.0), erythronolactone (R^^ 3.3) and pyruvic acid (1^^4.6). The spot at 5 j j a n 3.3 gave a characteristic response to a lactone spray consisting of hydroxylamine-ferric chloride. The p.m.r. 76 spectrum of fraction 3 showed the proton ratio of pyruvate CRg to the ring protons to be 3:15 which indicated that fraction 3 was a 1:1 mixture of pyruvic acid-D-mannose-L-erythroholactone complex and erythritol. The chemical shift of the anomeric proton (T5.12) indicated the D-mannopyranosyl linkage to be in the j8 -configuration. The Location of 0-acetyl Groups in K5 , To the polysaccharide (30mg) in dimethyl sulfoxide (5ml) p-toluenesulfonic acid (30mg) and methyl vinyl ether (5g, condensed with dry ice-acetone) were added. The mixture was stirred below 20° for 40 min then poured into dry acetone (200ml) and stored overnight. The white precipitate was separated by centrifugation and the supernatant was concentrated and fractionated on a Sephadex LH-20 column (22 X 3 cm) using dry acetone as eluent. The acetalated polysaccharide, eluted in the void volume, was collected and subjected to Hakomori methylation according to the method of Conrad and Sandford (15). Paper chromatography (solvent B) of the hydrolyzate of the acetalated and methylated polysaccharide gave a spot corresponding to monomethylated sugar which in turn was identified as i t s a l d i t o l acetate on g.l.c. to have the same retention time (19.2 min) using column a (210° and helium flow rate 86 ml/min) as the alditol acetate of 2-0-methyl-D-glucose. A sample was collected and the m.p. mixed m.p. with the authentic alditol acetate of 2-0-methyl-D-glucose was 54-55°. 77 IV. THE STRUCTURE OF THE CAPSULAR POLYSACCHARIDE FROM KLEBSIELLA K- TYPE 62 78 SUMMARY Methylation, partial hydrolysis and periodate oxidation of the capsular polysaccharide of Klebsiella K62 show the structure to consist of a repeating unit. — - D_-Glcp ~ D-GlcAp ~ - D-Manp — ^ D-Galp ~ ~ 3~ cx 1 D-Manp The anomeric linkages were determined by p.m.r. spectroscopy of the polysaccharide and of oligosaccharides isolated following partial hydrolysis. 79 DISCUSSION The structure of K62 belongs to the prevalent structural pattern of the Klebsiella capsular polysaccharides so far published which is composed of a repeating unit involving a single sugar in the side chain with three to four monosaccharides in the backbone. Capsular polysaccharide from K-type 62 consists of one D-mannose lateral unit per five sugar units. The polysaccharide purified by Cetavlon- pr e c i p i t a t i o n , ' d e i b n i z a t i freeze-drying, had a neutralization equivalent of 806. One hexuronic acid to four hexose units requires 826. The optical rotation was determined on the methylated polysaccharide and found to be £c*)n~t~59.70 (c 2.3, carbon tetrachloride). The p.m.r. spectrum of a 2% solution of the polysaccharide in D20 at 95° showed signals of anomeric protons indicating two £ -linkages (T5.49 and 5.35) and three c* -linkages (T4.85, 4.81 and 4.58) (50,51). This p.m.r. pattern was confirmed by those of the oligosaccharides where better resolution was obtained due to the decreased viscosity. Hydrolysis of the polysaccharide, showed the rapid liberation of D-mannose and after 8 h D-mannose, D-galactose and D-glucose were found to be in the ratio of 2:1:1. When the polysaccharide was methanolyzed and the uronic acid was converted to the neutral sugar through the reduction of i t s methyl ester, the molar ratio of D-mannose, D-galactose and D_-glucose was found to be 2:1:2. The additional D-glucose must be from D-glucuronic acid. The results indicate that a five-sugar unit is the repeat block of the polysaccharide. 80 The quantitative analysis was carried out on the derived alditoi acetates. Samples of these were collected by g.l.c. and measurement of their c.d. spectra confirmed the assignment of the D-configuration to mannose and glucose while the D-configuration of galactose was demonstrated by using the partially methylated alditoi acetate (82). Glucuronic acid was collected from several batches of hydrolyzate and the positive c.d. spectrum of the lactone at 219 my was identical with that of authentic D-glucuronolactone thus indicating the D-configuration (37). Partial hydrolysis of the polysaccharide was conducted under two sets of conditions: (a) heating the polysaccharide with IM t r i -fluoroacetic acid for 5 h to obtain mainly an aldobiouronic acid and (b) for 1 h to obtain higher oligomers. The acidic oligosaccharides were separated from the monosaccharides by charcoal column chromatography(9 and individual components were obtained by Sephadex gel chromatography (G 15). Water was used as the eluent at a flow rate of ca. 5ml/min. The higher oligosaccharides were eluted f i r s t and followed by oligo-saccharides of lower molecular weight. The elution volume which covers this range was ca. 80ml to 150ml, and collection of ca. 2 ml fractions gave satisfactory separations. The content of each fraction was examined and a progressive appearance of higher to lower oligomers as well as the region of overlapping of two adjacent components were observed. Only the fractions containing pure components were pooled, freeze-dried, and analyzed. These were shown to be an aldobiouronic acid together with related aldotri-, pseudoaldotetra- and penta-uronic acids. The structures of these oligosaccharides are given in Table IV together with the chemical shifts of the anomeric protons. These data are in good agreement with the optical rotations, and enable the nature TABLE IV. OPTICAL ROTATIONS AND P.M.R. DATA ON KLEBSIELLA K 62 CAPSULAR POLYSACCHARIDE AND DERIVED OLIGOSACCHARIDES ^ Proton assignment y N23° T-value (coupling Ratio of ( a l l sugars have Repeating unit of compound ^C<J constant, H ) integrals D-configuration) GlcA Man-OH -32.0° 4.67 (2) 0.7 Ot - Man-OH (A 2) 5.00 0.3 J8- Man-OH 5.40 (7) 1 J3- GlcA i - 2 - Man GlcA ^f* Man ~ - Gal -OH +32.1° 4.67 0.75 CX - Gal-OH (A 3) 5.36 (7) 0.25 /$ - Gal-OH 4.80 1 CX - Man — Gal-OH 5.35 (7) 1 j3~ GlcA i - 2 - Man-OH Man — Gal ^ ~ erythritol +52.2° 4.91 1 CX - Man — Gal-OH 5.40 1 j8 - Gal — erythritol 1 2 1 2 1 3 Glc — GlcA Man Gal-OH +56.2° 4.71 0.4 CX . Gal-OH (A4> 5.36 (7) 0.6 J$ - Gal-OH 5.20 (7) 1 £ - GlcA U L Man-OH 4.71 I 1 3 Cx - Man Gal-OH 4.58 (3) 1 1 2 CX - Glc GlcA-OH 4 1 2 1 2 1 3 I G l c - ^ - GlcA Man Gal ~g~ 4.58 1 cx 3 1 4.81 1 Three CX-linkages Man 4.85 1 5.35 (5) 1 e i-5.49 (7) 1 Two jQ-linkages, * Spectra r u n in DoO with external tetramethvlsilane ( T = 10) at 100 M H_. 82 of the glycosidic linkages to be assigned. The structures given were determined by 1) hydrolysis with acid and enzyme, 2) progressive partial hydrolysis and 3) methylation as described in the experimental. Capsular polysaccharide was methylated (15,17), hydrolyzed and the partially methylated sugars were examined. 2,3,4,6-Tetra-O-methyl-D-mannose, 2,3,6-tri-0-methyl-D-glucose, 2,4,6-tri-0-methyl-D-galactose, 4,6-di-0-methyl-D-mannose were found in a molar ratio of 1:1:1:1 which was determined by g.l.c. of the al d i t o i acetates (92). The acidic component was found to be 3,4-di-0-methyl-D-glucuronic acid as identified by the derived 3,4-di-0-methyl-D-glucose. The analysis of methylated, carboxyl reduced polysaccharide showed that 3,4-di-O-methyl-D-glucose was present in equimoiar ratio to the other sugar moieties, A simple chromatographic differentiation on paper of 2,3,4,6-tetra-Onmethyl-D-mannose and 2,3,4,6-tetra-0-methyl-D-glucose was devised by making the chromatographic tank unsaturated with solvent at the beginning of chromatography. The presence of tetramethylated D-mannose indicates that D-mannose is in the side chain and the branch point i s another D-mannose since a dimethyl-D-mannose was the only dimethylated sugar obtained. The exact sequential arrangement is established by the study of the oligosaccharides. The identities of the individual partially methylated sugars were determined by comparison with standards using a) paper chromatography, 2) g.l.c. of the methyl glycosides, c) g.l.c. - m.s. of the alditoi acetates, d) m.p. of crystalline compounds, and e) demethylation to the parent sugar aldi t o i acetate in the case of tetramethyl-D-mannose. The methylation data in conjunction with the results of analysis of the acidic oligosaccharides enable the repeat unit of Klebsiella K 62 capsular polysaccharide to be written as: 4 1 2 1 2 1 3 1 D-Glcp D-GlcAp D-Manp D-Galp * 1 .D-Manp Confirmation of this structure was sought by the Smith degradation procedure. The polysaccharide was oxidized with periodate and 0.82 mole per hexose unit was consumed. The derived polyalcohol was totally hydrolyzed to give D-mannose, D-galactose, erythritol and glycerol in a molar ratio of 1:1:1:1. Erythritol i s derived from D-glucose and glycerol from D-mannose in the side chain (Figure 3). Mild acid hydrolysis by the acidification of the polyol with Amberlite IR-120 (H +) followed by evaporation to dryness at room temperature yielded, by gel f i l t r a t i o n , glycerol and a disaccharide glycoside composed of D-mannose, D-galactose and erythritol. The evidence presented shows clearly that the structure of the capsular polysaccharide of Klebsiella K-type 62 is as given above. 84 EXPERIMENTAL Isolation and Properties of K62 Capsular Polysaccharide Klebsiella K62 (5711-52) was grown in the medium as for K5 and harvested to give 400ml of slime and c e l l s from 41 of agar medium. Phenol (5g) in water (100ml) was added to the slime and the mixture was diluted to 1.5 1 with water. Centrifugation of the diluted slime at 27,000 r.p.m. (68,000g) for 1 h yielded a clear light yellow supernatant which was treated with Cetavlon (107., 25ml) to precipitate the acidic polysaccharide. The neutral polysaccharide remaining in the supernatant was dialyzed and freeze-dried, yielding 0.7g. The precipitated acidic polysaccharide was dissolved in sodium chloride (2M), then precipitated in five volumes of ethanol. The precipitate, collected by centrifugation, was dissolved in water, decationized, dialyzed against d i s t i l l e d water, then freeze-dried to give 3.7g of pure acidic polysaccharide. The electrophoretic mobility on cellulose acetate was 2.4cm for 30 min at pH 8.8 and 300 V. Ash content n i l : nitrogen n i l . This product, after methylation, had (<*)D +59.7° (c 2.3, carbon tetrachloride). The equivalent weight of this polysaccharide was found to be 806 by titration with 0.01M sodium hydroxide (phenolphthalein). The p.m.r. spectrum of a 27. solution of the original polysaccharide in D20 at 95° showed signals of anomeric protons at .T5.49 (IH), 5.35 (IH), 4.85 (IH), 4.81 (IH) and 4.58 (IH). A mildly degraded K62 polysaccharide was obtained by heating the original polysaccharide with trifluoroacetic acid (0.2M) at 97° for 30 min, followed by dialyzing against tap water for 24 h. The material in the dialysis tubing was freeze-dried, and i t s p.m.r. spectrum in D^ O gave signals of anomeric protons similar to that 85 of original polysaccharide with an improved resolution: f 5.49 (J 7Hz, IH), 5.35 (J, 5Hz, IH), 4.85 (IH), 4.81 (IH) and 4.58 (IH) (see Table IV). Analysis of Constituent Sugars The capsular polysaccharide (4mg) was hydrolyzed with trifluoroacetic acid (2M) at 100° for 9h. After evaporation, the hydrolyzate was examined by paper chromatography in solvent B revealing three major spots for D-mannose, D-glucose, D-galactose and an acidic component at the origin. Paper chromatography in solvent A showed the presence of D-glucuronic acid, and aldobiouronic acid (RQ^ 0.45) and a trace of aldotriouronic acid ( R ^ j ^ 0.16), in addition to the neutral hexoses. The polysaccharide (lOmg) was heated in hydrochloric acid (2M) on a steam bath for 8h. The solution was passed through a Duolite A-4 (OH") column to neutralize hydrochloric acid and to remove the acidic sugars. After evaporation, the neutral fraction was analyzed and found to contain D-mannose, D-galactose and D-glucose in an approximate 2:1:1 ratio as determined by the g.l.c. of the alditoi acetates using column a. The individual components were collected and identified as D-mannitol hexaacetate, m.p. 119-121°, galactitol hexaacetate, m.p. 161-163° and D-glucitol hexaacetate, m.p. 93-95°. The configuration of mannose and glucose were found by dissolving the mannitol hexaacetate and glucitol hexaacetate collected from g.l.c. in acetonitrile. The positive circular dichroism curves at 213 my , identical to those given by authentic samples, confirmed the D-configurations of mannose and glucose. A sample of polysaccharide (lOmg) was refluxed in methanolic 86 hydrochloride (3%) for lOh on a steam bath. After neutralization (s i l v e r carbonate), reduction (sodium borohydride) in methanol, the reaction mixture was neutralized with Amberlite 1R-120 (H+) cation exchange resin and d i s t i l l e d three times with methanol. The residue was further hydrolyzed (2M trifluoroacetic acid), reduced and acetylated. The acetylated mixture was found by g.l.c. analysis on column a to contain D-mannitol hexaacetate (m.p. and mixed m.p. 121-123°), galactitol hexaacetate (m.p. and mixed m.p. 162-164°), and p_-glucitol hexaacetate (m.p. and mixed m.p. 95-97°) in an approximate ratio of 2:1:2. Methylation of the Capsular Polysaccharide Dry polysaccharide (Ig) was dissolved in anhydrous dimethyl sulfoxide (40ml). To this solution methyisulfinyl anion (25ml, 3M) was added and the mixture was stirred for 12h. Methyl iodide (8ml) was added slowly while s t i r r i n g and keeping the temperature below 20°. After 0.5h the solution was dialyzed against running water for 24h then extracted with chloroform (3 X 100ml). The chloroform extract was dried over anhydrous sodium sulfate followed by evaporating the chloroform solution to dryness to give 0.9g of methylated polysaccharide which showed absorption of hydroxyl groups at 3600 cm"^. The incompletely methylated polysaccharide was dissolved in 8ral of N, N-dimethyl formamide and 20ml of methyl iodide. While stirr i n g , silver oxide (5g) was added. Another portion of silver oxide (5g) was added lOh later and the sti r r i n g was continued for 12h more. The reaction mixture was centrifuged and the solid was washed with chloroform. The combined centrifugate and washings were washed with water (2 X 10ml) 87 and the organic layer was dried and evaporated to dryness. The residue was methylated with Purdie's reagents to give a product showing no absorption at 3600 cm"l,(c*)D +59.7° (c 2.3, carbon tetrachloride). Methylated polysaccharide (O.lg) was heated at 100° in formic acid (90%) for lh, and the solution was evaporated to dryness. The residue was subsequently dissolved in hydrochloric acid (2M) and heated for 6h.AAf.ter evaporation to dryness,, analysis of the hydrolyzate by-paper chromatography (solvent D and F), showed the presence of 2,3,4,6-tetra-0-methyl-D-mannose (R f 0.80, solvent D; 0.82, solvent F), 2,3,6-tri-O-methyl-D-glucose (R f 0.55, D; 0.72, F), 2,4,6-tri-O-methyl-D-galactose (R^ 0.40, D; 0.63, F), 4,6-di-0-methyl-D-mannose (R f 0.28, D; 0.56, F), and an acidic component (origin, D; R^  0.29, F). Separation on anion-exchange resin (Duolite A-4) gave a neutral fraction which showed on paper chromatography components corresponding to 2,3,4,6-tetra-0-methyl-D-mannose, 2,3,6-tri-O-methyl-D-glucose, 2,4,6-tri-0-methyl-D-galactose and 4,6-di-0-methyl-D-mannose. The neutral fraction was reduced with sodium borohydride, and after removal of the borate, the material was acetylated in a sealed tube by using acetic anhydride and pyridine (1:1, 1 ml, 97°, 20min). After removal of the reagents, the residue was dissolved in a small volume of methanol and analyzed by g.l.c. on column a (195°, helium flow rate 71 ml/min) and column d (212°, helium flow rate 97 ml/min). Four peaks in the ratio of 1:1:1:1 were observed. They were identified by their retention times and mass spectra as acetates of 2,3,4,6-tetra-0-methyl-D-mannitol (8.6 min column 25.5 min, column d), 2,3,6-tri-0-methyl-D-glucitol (20.5 min, a; 32.7 min 2,4,6-tri-O-methyl-D-galactitol (18.4 min, a; 36.6 min, d), and 4,6-di-0-methyl-D-mannitol (26.1 min, a; 39.7 min, d). 88 A separate portion of the methylated polysaccharide (0.7g) was heated in formic acid (90%) on a steam bath for 40 min. After evaporation to dryness, the mixture was further hydrolyzed in trifluoro-acetic acid (2M) on a steam bath for 3h. The hydrolyzate, after removal of trifluoroacetic acid by evaporation, was separated into a neutral fraction (0.42g) and an acidic fraction (0.21g) using Duolite A-4 (OH-) ion exchange resin. The neutral fraction was separated into four components by preparative paper chromatography using solvent D. Fraction I (51mg) had chromatographic mobility in solvent D identical to 2,3,4,6-tetra-O-methyt-D-mannose, and could be differentiated from 2,3,4,6-tetra-O-methyl-D-glucose in solvent D i f the chromatographic tank were i n i t i a l l y not saturated with respect to the solvent system and with several drops of ammonium hydroxide on the bottom of the tank instead of using a tank saturated with the solvent. Under these special conditions i t took longer (ca. lOh) to develop the chromatogram, however, i t provided a ready separation of 2,3,4,6-tetra-OHnethyl-D-mannose and 2,3,4,6-tetra-0-methyl-D-glucose into two distinct spots with a ratio of mobility Me4 Man/Me^ Glc = ca. 0.94. A portion of this fraction (5mg) was methanolyzed with 3% methanolic hydrochloride, and after neutralization (silver carbonate) and concentration, the product was examined on column a, at 150° and a helium flow rate of 60 ml/min. Two peaks were obtained in approximately 10:1 ratio, identical in ratio and retention times to methyl 2,3,4,6-tetra-O-methyl-tf-D-mannoside (20.9 min) and methyl 2,3,4,6-tetra-O-methyl-^-D-mannoside (27.5 min) obtained by methanolysis of an authentic sample of the sugar. 89 A separate portion of fraction I (5mg) was dissolved in methylene chloride (lml) and chilled at -75°. Boron trichloride (2ml) withdrawn from a cylinder while chilled at -75° was added to the methylene chloride solution and the reaction mixture was kept at that temperature for 30 min, then allowed to stand at room temperature for I6h. After evaporation and d i s t i l l a t i o n s with methanol, the de-0-methylated product was reduced o o and acetylated and separated on column a programmed from 185-120 at 1 /min, with a helium flow rate of 100 ml/rain, to give a peak identical to the authentic D-mannitol hexaacetate (18 min). A sample was collected and the m.p. and mixed m.p. with D-mannitol hexaacetate was 120-122°. Fraction II (30mg) had the same chromatographic mobility as 2,3,6-tri-O-methyl-D-glucose (R f 0.55, solvent D). The m.p. was 115-117° (from ethyl ether) and undepressed with an authentic sample of 2,3,6-tri-0-methyl-D-glucose. Fraction III (32mg) had the same chromatographic mobilitiy as 2,4,6-tri-0-methyl-D-galactose (R^ 0.40, solvent D), and was crystallized from chloroform and recrystallized from ethyl ether to give crystals of m.p. 103-105°, undepressed with an authentic sample of 2,4,6-tri-0-methyl-D-galactose. The c.d. spectrum of 2,4,6-tri-0-methyl-galactitol acetate was identical to that of a standard sample of 2 , 4 , 6 - t r i -O-methyl-D-galactitol confirming the D-configuration of galactose. Fraction IV had the same chromatographic mobility as 4,6-di-0-methy1-D-mannose (R^ 0.28, solvent D). Attempted crystallization of fraction IV was not successful and i t was subsequently reduced, acetylated, and examined on column a (195°, helium flow rate 71 ml/min) to give a component identical in retention time (26.1 min) and mass spectrum to that of the alditol acetate of authentic 4,6-di-O-methyl-D-mannose. 90 Analysis of the acidic mixture A portion of the acidic fraction (20mg) was hydrolyzed in hydrochloric acid (2M) on a steam bath for 8h. After evaporation, the hydrolyzate was shown by paper chromatography to contain 4,6-di-O-methyl-D-mannose (Rf 0.28, solvent D) as the major component, together with small amounts of 2,3,6-tri-O-methyl-D-glucose (R f 0.55, solvent D) and 2,4,6-tri-0-methyl-D-galactose (Rf 0.40, solvent D) and having acidic components at the origin. A separate portion of the acidic fraction (lOOmg) was heated under reflux overnight with 3% methanolic hydrochloride (25ml) to give the ester glycosides which were reduced by refluxing with lithium aluminum hydride (0.5g) in tetrahydrofuran (15ml) for 4h. The excess lithium aluminum hydride was destroyed by adding ethyl acetate, evaporated to dryness, and extracted with chloroform. The chloroform extract was concentrated and the residue was hydrolyzed with triftuoro-acetic acid (2M). The hydrolyzate was examined by paper chromatography to show 4,6-di-O-methyl-D-mannose (Rf 0.28, solvent D) overlapped with another component, 2,3,6-tri-O-methyl-D-glucose (R^ 0.55, solvent D), and 2,4,6-tri-O-methyl-D-galactose (R f 0.40, solvent D). No acidic components were observed at the origin. The hydrolyzate was subsequently o reduced, acetylated, and examined by g.l.c. on column a (195 , helium flow rate 71 ml/min) to show peaks identical in retention times to 2,3, 4,6-tetra-0-methyl-D-mannose (8.6 min, trace), 2,3,6-tri-O-methyl-D-glucose (20.5 min), 2,4,6-tri-0-methyl-D-galactose (18.4 min), 4,6-di-O-methyl-D-mannose (26.1 min), as well as a peak at 42.1 min which was derived from the D-glucuronic acid moiety in the polysaccharide. The mass spectrum of the last peak showed the fragmentation pattern 91 characteristic of an ald i t o l acetate of a 3,4-di-0-methylhexose. Therefore the last component was 3,4-di-O-methyl-D-glucitoi acetate which reflected the fact that D-glucuronic acid in the polysaccharide was connected at position 2. Reduction of the Methylated Polysaccharide A portion of the methylated K62 (O.lg) was dissolved in tetrahydrofuran (30ml). To this solution lithium aluminum hydride (0.l6g) in tetrahydrofuran (8ml) was added,and the reaction was stirred overnight. The excess lithium aluminum hydride was destroyed by careful addition of ethyl acetate. After removal of the solvents by evaporation the residue was extracted with chloroform and the organic solution was evaporated to dryness, yielding 0.09g. The reduced product was heated with formic acid (90%) on a % steam bath for lh.- After concentration, the residue was further hydrolyzed with tfifluoroacetic acid (2M) for 4h. Quantitative analysis of the derived alditol acetates by g.l.c. on column a (195°, helium flow rate 71 ml/min) and column d (212°, helium flow rate 97 ml/mirt)showedthe p r e s e n c e o f i t h e a l d i t o l a c e t a t e s o f 2,3 , 4 , 6 - t e t r a - 0 - m a t h y1 - D - m a n n o s e ( 8 . 6 m i n , a ; 25.5 min, d), 2,3,6-tri-O-methyl-D-glucose (20.5 min, a; 32.7 min, d), 2,4,6-tri-0-methyl-D-galactose (18.4 min, a; 36.6 min, d), 4,6-di-0-methyI-D-mannose (26.1 min, a; 39.7 min, d), and 3,4-di-O-methyl-D-glucose (42.1 rain, a; 48.8 min, d) in a molar ratio of 1:1:1:1:1. The identity of the methylated sugars was confirmed by mass spectrometry of the derived alditol acetates. 9 2 Partial Hydrolysis of Capsular Polysaccharide Autohydroiysis of the capsular polysaccharide was carried out at pH 3.5 and 100° up to 37h. D-Mannose as the only neutral monosaccharide and traces of oligosaccharides were found by paper chromatography in solvents A and; C. Polysaccharide (0.4g) was hydrolyzed with trifluoroacetic acid (IM) on a steam bath for 5h. After evaporation, the hydrolyzate was found, by paper chromatography in solvent A, to contain D-mannose, D-glucose, D-galactose, D-glucuronic acid, and a component with the mobility of an aldobiouronic acid (15^  0.37). The hydrolyzate was separated into neutral and acidic fractions using ion-exchange resins (Amberlite IR-120 (H +) and Duolite A-4 (0H~) ). The acidic fraction (iOOmg) was applied to a Sephadex G 15 column (110 X 2 cm) which was irrigated with water at a flow rate of ca. 5 ml/h. The eluent was collected ca. 2ml each tube and the content in each tube was examined on paper in solvent C. The component corresponding to an aldobiouronic acid, A 2, (60mg) was collected from fractions about the elution volume of 140ml, followed by fractions containing the mixture of an aldobiouronic acid and D-glucuronic acid, and fi n a l l y followed by fractions containing only D-glucuronic acid (lOmg). The glucuronic acid was evaporated with the addition of a few drops of concentrated hydrochloric acid. The c.d. curves of the isolated D-glucuronic acid and that of a standard were both positive (^£ H2° +2.82 and +3.32, respectively). 219 In order to obtain oligosaccharides higher than the aldobiouronic acid (A 2), a milder hydrolysis condition was used by heating the polysaccharide (0.8g) in t r i f luoroacetic acid CW for lh. The solution was concentrated 93 and evaporated three times with water and fractionated using a charcoal-Celite column (a mixture of active carbon (30g) and Gelite (30g) was poured into a Buchner funnel (10cm i.d.) to make a column of 1.4 cm thickness). Monosaccharides were eluted by d i s t i l l e d water (2.8 1) and a series of oligosaccharides was eluted with 207. aqueous ethanol (2 1) and 107. aqueous isopropanol (500ml); yield of oligosaccharides about 0.34g. The mixture of oligosaccharides was separated (lOOmg per run) on a Sephadex G 15 column (110 X 2cm) into four fractions (J^j 0.46, 0.19, 0.12 and 0.06, solvent C), namely di-(A 2, 4mg), t r i - ( A 3 , 52mg), tetra- and pentasaccharides (50mg), and Ag (31mg). A successful separation of the third fraction into tetrasaccharide (A^ 28mg) and A 5 (13rag) was achieved by a prolonged paper chromatography (7 days) in solvent C. The oligosaccharides were analyzed as follows. Structural study of Aldobiouronic Acid (A 2) The aldobiouronic acid was found to have |e*)D-32.0° (c 6.37, H 20). (a) Constituent sugars ""' A portion (15mg) of A was hydrolyzed with hydrochloric acid (2M) at.97° for 8h, then evaporated to dryness. Paper chromatography in solvent B showed D-mannose as the only neutral sugar present, whereas the presence of D-mannose, D-glucuronic acid, and D-giucuronolactone was demonstrated by using solvent A. The rest of the hydrolyzate was dissolved in 37. methanolic hydrochloride (lOmg) and the solution was refluxed for 6h. The solution was neutralized (silver carbonate) centrifuged, and evaporated. The residue was dissolved in methanol and reduced with sodium borohydride. The solution was l e f t overnight at 94 room temperature and was then carefully neutralized and decationized by passage through an Amberlite IR-120 (H+) column. The residue, after removal of borate, was hydrolyzed with trifluoroacetic acid (IM) at 97° for 4h and shown in solvent A to give D-mannose and D-glucose. The mixture of neutral sugars was reduced, acetylated and found by g.l.c. on column a to contain D-raannitol hexaacetate (m.p. and mixed m.p. 120-122°;) and D-glucitol hexaacetate (m.p. and mixed m.p. 95-97°) in a ratio of 1:1. (b) Methylation. A^ (20mg) was methylated according to the method of Hakomori. One third of the methylated compound was hydrolyzed (hydro-chloric acid, 2M) to give a trimethyl sugar with the identical chromatographic mobility and characteristic color response of 3,4,6-tri-0-methyl-D-mannose (R^ 0.51, solvent D, a green color response to the p-anisidine spray), and an acidic component (R^ 0.2, solvent F). The remaining methylated aldobiouronic acid was treated with Purdie's reagent and the methyl ester methyl glycoside was then reduced in methanol with sodium borohydride and hydrolyzed. Paper chromatography showed two unresolved spots (R^ 0.56 and 0.58, solvent D). G.l.c. , analysis of the alditol acetates on column a at 180° and a helium flow rate of 43 ml/min gave two peaks identical in retention times to those of the alditol acetates of 3,4,6-tri-0-methyl-D-mannose (15min) and 2,3,4-tri-0-methyl-D-glucose (I8.4min) in a ratio of 1:0.8, The identity of these two compounds was confirmed by comparison of their mass spectra with those of authentic compounds. (c) Anomeric configuration A portion of A^ (20mg) was dissolved in Drfi (1ml) and freeze-dried. The deuterium exchange was repeated once more. P.m.r. of 95 this product (in Do0) showed peaks at T 5.40 (J 7.0 Hz, IH), 5.0 <s 1,2 (0.3H) and 4.67 (Jj 2 2 Hz, 0.7H), indicating the y8 -linkage of the D-glucuronic acid at the nonreducting end, and the predominant Ot-form (two thirds) to the J3 -form (one third) of the D-mannose in equilibrium at the reducing end. Enzymatic hydrolysis of the aldobiouronic acid (lmg, in 0,2ml of c i t r i c buffer, pH5) withy^ -glucuronidase (lmg) was carried out at room temperature for 4h. Paper chromatography (solvent A) of the reaction mixture showed the presence of D-mannose and D-glucuronic acid and the absence of the aldobiouronic acid, confirming the p.m.r. assignment of the ^ - l i n k a g e of the D-glucuronic acid. Structural analysis of aldotriouronic acid (A^) The aldotriouronic acid was found to have (°0D + 32.1 (c 10.9, water), (a) Constituent sugars A^ (20mg) was hydrolyzed with trifluoroacetic acid (0.2M) at 97° for 4h, Paper chromatography (solvent A) showed the presence of D-galactose, k^ and intact A 3 . The hydrolyzate was concentrated and further hydrolyzed in hydrochloric acid (2M) at 97° for 8h, and was examined by paper chromatography in solvent B then A to show the presence of D.?mannose, D-galactose and D-glucuronic acid. The hydrolyzate was evaporated to dryness then refluxed in 3% methanolic hydrochloride. The resulting methyl ester methyl glycoside and methyl glycosides were reduced with sodium borohydride in methanol, and hydrolyzed (trifluoroacetic acid, 2M) to give, by paper chromatography in solvent B, D-mannose, D-glucose and D-galactose. The ratio was found to be 1:1:1 by g.l.c. on 96 column a of the a l d i t o l acetates. (b) Methylation A3 (20mg) was methylated according to the procedure of Hakomori. One third of product was hydrolyzed (hydrochloric acid, 2M, 97°, 8h) to give (by paper chromatography) 3,4,6-tri-O-methyl-D-mannose (R f 0.60, solvent D), 2,4,6-tri-O-methyl-D-galactose (R f 0.41, solvent D), and an acidic component (R^ ca.0.2, solvent F). After removal of the acidic component (Amberlite IR-120 and Duolite A-4) the hydrolyzate was reduced and acetylated. G.l.c. of the alditol acetates show two peaks at 12 min and 13.6 min (column a, 185°, helium flow rate 60 ml/min) which were identical in retention times to the corresponding authentic alditol acetates. The remaining methylated aldotriouronlc acid was treated with Purdie's reagent, reduced in methanol with sodium borohydride, then hydrolyzed. Paper chromatography (solvent D) showed two unresolved spots corresponding to an authentic sample of 3,4,6-tri-0-methyl-D-mannose (R f 0.62) and 2,3,4-tri-0-methyl-D-glucose (R f 0.60), along with 2,4,6-tri-O-methyl-D-galactose (R^ 0.42). G.l.c. of the a l d i t o l acetates (on column a, 185°, helium flow rate 60 ml/min) showed three peaks identical in retention times to the al d i t o l acetates of the authentic 3,4,6-tri-0-methyl-D-mannose (12 min), 2,4,6-tri-O-methyl-D-galactose (13.6 min) and 2,3,4-tri-O-methyl-D-glucose (16 min) in an approximate ratio of 1:1:1. The structures of the partially methylated alditol acetates were confirmed by mass spectrometry. (c) Anomeric configuration The p.m.r. spectrum (in D„0) showed the anomeric protons at 97 T 5.36 (Jx 2 7.0 Hz, 0.25H), 5.35 2 7.0 Hz, ^ 3 1.5 Hz, IH), 4.80 (IH), and 4.67 (0.75H). Structural analysis of acidic tetrasaccharlde (A^) The acidic tetrasaccharlde was found to have (°0n +56.2° (c 4.45, water). (a) Constituent sugars A^ (15mg) was hydrolyzed in trifluoroacetic acid (IM) at 97° for 4h. Paper chromatography (solvent B then A) showed the presence of A 2, D-glucuronic acid, D-galactose, D-glucose and D-mannose. After concentration, the hydrolyzate was further hydrolyzed in hydrochloric acid (2M) at 97° for 7h to give (by paper chromatography in solvent B then A) components corresponding t o D-glucuronic acid, D-galactose, D-glucose and D-mannose. After removal of the D-glucuronic acid, g.l.c. (on column a) of the derived alditoi acetates of the neutral sugars showed three peaks corresponding to D-mannitol hexaacetate, galactitol hexaacetate and D-glucitol hexaacetate in a ratio of 1:1:1. (b) Methylation A^ (lOmg) was methylated according to the method of Hakomori and hydrolyzed. Paper chromatography in solvent D showed 2,3,4,6-tetra-0-methyl-D-glucose (R f 0.84), 3,4,6-tri-0-methyl-D-mannose (R f 0.61), 2,4,6-tri-0-methyl-D-galactose (R^ 0.41) and an acidic component near the origin of the chromatogram along with small amounts of tri-0-methyl-D-galactose (R f 0.68) possibly due to the methylation of the furanoside isomer. The chromatographic mobility of 2,3,4,6-tetra-O-methyl-D-glucose was faster than that of 2,3,4,6-tetra-O-roethyl-D-mannose and distinguishable from each other when the chromatographic tank was unsaturated with 98 solvent D In the beginning of development. The ratio of 2,3,4,6-tetra-O-methyl-D-glucose, 3,4,6-tri-O-methyl-D-mannose, and 2,4,6-tri-0-methyl-D-galactose was found to be approximately 1:1:0.8 as determined by g.l.c. of the alditoi acetates, (c) Anomeric configuration The p.m.r. spectrum in D20 showed the following signals: T5.36 ( J 1 2 7 Hz, 0.6H), 5.20 <Jj_ 2 7 Hz, IH), 4.71 (1.4H) and 4.58 (J. _ 3 Hz, IH). The foregoing evidence is consistent with the assignment of A^ as a pseudoaldotetraouronic acid of the following structure. D-Glcp ~ ~ D-GlcAp i ~ D-Manp ~ p_-Gal Structural analysis of A^ A 5 was found to have{©<)D +19.4 (c 1.08, water). A^ was hydrolyzed (hydrochloric acid, 2M, 97°, 8h) to show (by paper chromatography in solvent B then A) the presence of D-glucuronic acid, D-galactose, D-glucose and D-mannose. Methylation of A^ by the method of Hakomori followed by hydrolysis gave (paper chromatography in solvent D) 2.3.4.6-tetra-O-methyl-D-glucose (R f 0.80), 2,3,4,6-tetra-0-methyl-D-mannose (slightly slower than 2,3,4,6-tetra-O-methyl-D-glucose when the chromatographic tanks was i n i t i a l l y unsaturated with solvent D), 2,3,6-tri-O-methyl-D-glucose (Rj 0.55) unresolved from 3,4,6-tri-O-methyl-D-mannose, 2,4,6-tri-O-methyl-D-galactose (Rf 0.40) and 4,6-di-O-methyl-D-mannose. The identity of the sugars was confirmed by the retention time on g.l.c. of the ald i t o i acetates (column a). The presence of both 99 3,4,6-tri-O-methyl-D-mannose and 4,6rdi-0-methyl-D-mannose suggested that was a mixture comprising oligosaccharides with and without D-mannose as the branch unit. The ratio of sugars from g.l.c. analysis indicated that these oligosaccharides were pentasaccharides. Ag was found to have (c*) n + 40.7° (c 1.67, water). Similar analysis were done on Ag as for A 5 and the results indicated that Ag was also a mixture of oligosaccharides. Periodate Oxidation Capsular polysaccharide (0.86g) was dissolved in sodium periodate (200ml, 0.05M) in the dark at 1°. The periodate uptake per sugar unit leveled off at 0.82 mole in 360h. The excess periodate ions and iodate ions were precipitated with barium hydroxide, and the mixture was centrifuged. The supernatant, after reduction (sodium borohydride) and decationization (Amberlite IR-120), was evaporated to dryness, then d i s t i l l e d three times with methanol. The yield of polyol was 0.442g. The polyol (15mg) was hydrolyzed with hydrochloric acid (2M) at 97° for 6h. Paper chromatography in solvent A gave D-galactose, D-mannose (R_ t 1.24), erythritol (R_ , 1.97) and glycerol (R „ 2.34). — ~v»al. —vial ~Gal The hydrolyzate was reduced by sodium borohydride and, after removal of the borate, the product was acetylated in a sealed tube by using pyridine and acetic anhydride (1:1, 20 min, 97°). After the evaporation of excess reagents, the ald i t o l acetates were dissolved in a small volume of ethyl acetate and analyzed by g.l.c. on column a programmed from o o 150 at 3 /min and a helium flow rate of 50 ml/min, to give four peaks identical in retention times to the ald i t o l acetates of authentic glycerol 100 (8.8 min), erythritol (19.0 min, m.p. and mixed m.p. 80-82°), D-mannose (41.0 min, m.p. and mixed m.p. 120-122°) and D-galactose (43.9 min, m.p. and mixed m.p. 160-162°) in a ratio of 1:1:1:1. The polyol, after acidification with Amberlite IR-120 (H+) and evaporation at 30°, showed, by paper chromatography in solvent A, a component with a mobility of R 0.53. A solution of the polyol ""Gal in trifluoroacetic acid (2M) at room temperature showed the increased release of glycerol and erythritol, together with an oligosaccharide at R ^ 0.53 (solvent A). Polyol (O.llg), after acidification with Amberlite IR-120 (H +) and concentration, was passed through a Sephadex G 15 column (110X2.0lid. cm) at a water flow rate of about 4ml/h. Besides higher oligomers, a disaccharide glycoside (24mg,(o<)n +52.2°, c 4.48, water) and glycerol (14.3mg) were collected. The identity of glycerol was confirmed by the retention time of the acetate on g.l.c. (column a). Hydrolysis and paper chromatography of the disaccharide glycoside gave D-galactose, D-mannose, and erythritol (solvent A). G.l.c. analysis of the alditol acetates on column a (programmed from 200° at o 1 /min, with a helium flow rate of 50 ml/min gave peaks identical to erythritol tetraacetate (8.4 min, m.p. and mixed m.p. 80-82°), D-mannitol hexaacetate (39.9 min, m.p. and mixed m.p. 121-123°) and galactitol hexaacetate (43.5 min, m.p. and mixed m.p. 160-162°). The p.m.r. spectrum of the disaccharide glycoside (&2°» showed signals of anomeric protons at £"5.40 ( j ^ ^  7 Hz, IH) and 4.91 (IH). THE STRUCTURE OF THE CAPSULAR POLYSACCHARIDE FROM KLEBSIELLA K- TYPE 18 102 SUMMARY Methylation, partial hydrolysis, and periodate oxidation studies on the capsular polysaccharide of Klebsiella K-type 18 show the structure to be a repeating unit consisting of L 1 4 1 2 1 1 1 i 1 — - D-Glcp ~ D-GlcAp L-Rhap ~ L-Rhap ™ - D-Galp ~ D- Glcp The nature of the anomeric linkages was determined by p.m.r. spectroscopy of isolated oligosaccharides and the periodate degraded polysaccharide. 103 DISCUSSION L-Rhamnose has been found existing in one third of the 80 Klebsiella capsular polysaccharides so far examined (12,13). Capsular polysaccharide from K-type 18 is unusually complex in that i t consists of two L-rhamnoses in the backbone and one D-glucose as the side chain, with a total of six sugar units which constitute the repeating block of this polysaccharide. By comparison, the prevailing structural patterns of capsular polysaccharides which occur in this genus consist of four to five sugars in the repeating unit, including a side chain. The relatively high content of L-rhamnose in K 18 capsular polysaccharide renders i t less viscous and more acid l a b i l e . The capsular polysaccharide, after purification by precipitation with Cetavlon, had (c*) D +77.3° (c 0.96, water). The p.m.r. spectrum of o a 2% solution of the polysaccharide in D^ O run at 95 showed two doublets in equal intensity at "C8.74 and 8.67 characteristic of the CHj of a 6-deoxyhexose. The anomeric region indicated that the repeating unit i s composed of six sugar units of which two have J& -linkages (T5.52 and 5.33) and four; have cx-linkages (T4.91, 4.58 and 4.50). Integration of the methyl signals of the 6-deoxysugars and the anomeric protons suggested two 6-deoxysugars to six sugar units. Acid hydrolysis of the polysaccharide rapidly liberated D-glucose and complete hydrolysis of the polysaccharide gave L-rhamnose, D-glucose, D-galactose and D-glucuronic acid. The ratio of L-rhamnose, D-galactose and D-glucose was found by g.l.c. of the derived alditoi acetates to be approximately 2:1:2. After the conversion of D-glucuronic acid into D-glucitol through the reduction of the derived D-glucurone, 104 the r r a t i o of t h e . a l d i t o l acetates of L-rhamnose, DD-galactose and D-glucose was found to be approximately 2:1:3 , Samples of the alditol acetates were collected from g.l.c. and measurement of their c d . spectra confirmed the assignment of the D-configuration to glucose and L-configuration to rhamnose. The re-configuration of glucuronic acid was determined on the derived glucurone, while the D-configuration was assigned to galactose on the basis of the c d . spectrum of the derived 2,4,6-tri-O-methyl-galactose (82). In order to determine the linkages of the constituent sugars, a sample of the capsular polysaccharide was methylated and hydrolyzed, and the partially methylated monosaccharides were examined. The neutral portion was shown to contain 3,4-di-0-methyl-L-rhamnose, 2,4-di-0-methyl-L-rhamnose, 2,3,4,6-tetra-0-methyl-D-glucose, 2,4,6-tri-0-methy1-D-galactose and 2,6-di-0-methyl-D-glucose and the ratio was determined by g.l.c. of the derived alditol acetates to be 0.7:1:1:1:1. The a l d i t o l acetates of a l l components were resolvable oh column d except-2^4,6-tri-0-methyl-D-galactose'and 2,6-di-O-methyl-D-glucose which had to be separated on column a. Thus, a combination of two different columns, namely,Apiezon L and 3% ECNSS-M, was required for the complete separation of the constituent monosaccharides as pa r t i a l l y methylated al d i t o l acetates. The acidic fraction from the methylated and hydrolyzed polysaccharide was reduced (lithium aluminum hydride) and hydrolyzed to give 2,3-di-0-methyl-D-glucose and 3,4-di-0-methyl-L-rhamnose. The methylated polysaccharide, after carboxyl reduction, was hydrolyzed to give 3,4-di-0-methy1-L-rhamnose, 2,4-di-0-methyl-L-rhamnose, 105 2,3,4,6-tetra-O-methyl-D-glucose, 2,4,6-tri-O-methyl-D-galactose, 2,6-di-O-methyl-D-glucose and 2,3-di-O-methyl-D-glucose in a ratio of 1:1:1:1:1:1 by g.l.c. on the derived alditoi acetates using columns a and d. From the methylation data, i t is clear that D-glucose is a side chain and the branch point ln the backbone is another D-glucose linked at either 3 or 4 positions by the side chain; and that D-glucuronic acid is linked at position 4 as demonstrated by the derived 2,3-di-O-methyl-D-glucose. Partial hydrolysis of the polysaccharide was carried out under two sets of conditions; namely, heating the polysaccharide with trifluoro-acetic acid (0.5M) for 4h to obtain an aldobiouronic acid as the predominant component and heating for 30 min to produce a trisaccharide and higher oligomers. The structures of a neutral dlsaccharide (N^), the aldobiouronic acid (A 2) and the trisaccharide ( A 3 ) were found to be 1 4 1 2 1 4 1 2 D-Galp ~ D-Glc, D-GlcAp — L_-Rha and D-Glcp D-GlcAp ~z~ L-Rha by means of 1) total and partial hydrolysis by acid, 2) enzymatic hydrolysis, 3) methylation, and 4) p.m.r. spectroscopy as stated in detail in the experimental. Fractions A^, A^ a and Ag a were prepared carefully and were chromatographically homogeneous, but chemical analysis showed these oligosaccharides to be mixtures. However, each fraction contained one oligosaccharide as the major component thus information concerning the anomeric linkages of individual monosaccharides could s t i l l be obtained. A^ was shown to be related to A^ but with an extra D-glucose unit attached by an cx -linkage. A^ a was shown by the p.m.r. spectrum to be a mixture of pentasaccharides containing two different L-rhamnoses and the presence of a -linkage due to additional 106 D-galactose, as compared to the structures and p.m.r. spectra of N^ , Aj and A^, was evident. Since the only two j& -linkages in the polysaccharide are assigned to D-glucuronic acid and D-galactose, the remaining four sugar units must be o< -linked. The p.m.r. spectrum of Ag a indicated the presence of two different L-rhamnoses in six sugar units of which two are -linked, three are ex.-linked and one at the reducing end, confirming the p.m.r. spectra of Nj^  A 2, A^, A^, A^^ and the original capsular polysaccharide. From the results of methylation and partial hydrolysis, the structure of capsular polysaccharide K 18 may be proposed to consist of either one of the following repeating units. — - D-Glcp ~ D-GlcAp ~ ~ L-Rhap i - 3 - L-Rhap — D-Galp ~~-= - c < = - y3 - - o< = o<~ -J3 D-Glcp 1 D-Glcp ~ D-GlcAp ~ L-Rhap ~ " L-Rhap ~~~ D-Galp D-Glcp II Since the relative positions of the five monosaccharides in the hexaasaccharide repeating unit are fixed, the position of the sixth monosaccharide, i.e., one of the L-rhamnoses linked at C^, has to be preceding another L-rhamnose which is linked at C 2 as indicated in the structures I and II. The problem remaining to be solved is whether 107 D-galactose is in the backbone (I) or in the side chain (II). Periodate oxidation of K 18 polysaccharide resulted in the consumption of 0.64 mole of periodate per sugar residue and total hydrolysis of the derived polyalcohoi gave D-galactose, D-glucose, L-rhamnose, glycerol, L-erythronolactone and 1-deoxyglycerol, in good agreement with the proposed structures. However, when the polyalcohoi was treated with trifluoroacetic acid (2M) at room temperature for ,8h, followed by dialysis a preferential cleavage of the glycerol and glycolaldehyde complex derived from the D-glucose in the side chain and the deoxyglycerol from the degraded L-rhamnose was achieved. The degraded polyalcohoi was methylated to give 2,4-di-0-methyl-L-rhamnose, 2,3,6-tri-O-methyl-D-glucose and 2,4,6-tri-O-methyl -D-galactose in a ratio of 1:1:1, with the absence of tetra-0-methyl-D-galactose. This evidence can only allow the structure I to be the correct repeat unit for K18 slime polysaccharide. The p.m.r. spectrum of the degraded polyalcohoi indicated the presence of one L-rhamnose in three sugar units of which one is in ^ - l i n k a g e and two are in Oi -linkages, confirming the assignment of ^ - l i n k a g e to D-galactose, and cx"-linkages to L-rhamnose and D-glucose. 108 EXPERIMENTAL Isolation and Properties of K18 Capsular Polysaccharide A culture of Klebsiella K18 (1754-49) was obtained from Dr. I. 0rskov, Copenhagen, and grown in the medium as for K5. Cells and slime (450ml) were harvested from 41 of agar medium spread on four big trays, diluted with water (450ml) containing 2% phenol, and centrifuged at 68,000g for Ih. The clear supernatant was treated with Cetavlon-(57., 40ml) to precipitate an acidic polysaccharide which was then dissolved in sodium chloride (2M), precipitated in five volumes of ethanol, dissolved in water, decationized (Amberlite IR-120), dialyzed against d i s t i l l e d water, and freeze-dried, to give 4.2g of pure acidic polysaccharide. The acidic polysaccharide had (oij ^ +77.3° (c 0.96, water), and no absorption at 260 TIXJJ and 280 nyj . The elemental analysis showed the absence of ash and nitrogen. The supernatant from the Cetavlon precipitation was dialyzed against tap water for 4 days and freeze-dried, giving 1.8g of neutral polysaccharides. A sample of the neutral polysaccharide^) was hydrolyzed and shown (by paper chromato-graphy in solvent B) to contain D-galactose as the major component along with lesser amounts of D-glucose and D-mannose, and this neutral material was not further investigated. The p.m.r. spectrum of a 2% solution of the purified acidic polysaccharide in D20 at 95° showed signals of anomeric protons at 7T5.55 (iH), 5.33 (Jj 2 7 Hz, IH), 4.91 (2H), 4.58 (IH), and 4.50 (IH), along with methyl protons from L-rhamnoses at 8.74 ( J , 3 Hz, 3H) and = 3,6 8.68 ( J 5 6 3 Hz, 3H). 109 Analysis of Constituent Sugars The polysaccharide (15mg) was hydrolyzed with trifluoroacetic acid (2M) at 97° for 18h, giving (by paper chromatography in solvent B then A) D-glucuronic acid, D-galactose, D-glucose and L-rhamnose. The hydrolyzate was evaporated to dryness, reduced with sodium borohydride in methanol and acetylated. G.l.c. analysis on column a programmed from 150° at 3°/min and a helium flow rate of 60 ml/min gave peaks corresponding to the alditoi acetates of L-rhamnose (11.I min), D-galactose (30.1 min, m.p. and mixed m.p. 160-162°) and D-glucose (35.5 min, m.p. and mixed m.p. 98-99°) in an approximate ratio of 2:1:3. Another portion of the polysaccharide (lOmg) was heated in hydrochloric acid (2M) on a steam bath for 8h. The solution was passed through a Duolite A-4 (OH") column to remove a i l the acidic components. After evaporation, the neutral fraction was analyzed and found to contain L-rhamnose, D-galactose and D-glucose in an approximate 2:1:2 ratio as determined by the g.l.c. of the alditoi acetates using column a. The individual components were collected and circular dichroism curves were taken in acetonitrile at 213 mjj. The negative c.d. (•^ E 213** -1'50) and the positive c.d. (AE +0.33) confirmed the L- and D-configuration of rhamnose and glucose, respectively. The c.d. curves of authentic samples were used for comparison. Methylation of the Capsular Polysaccharide To vacuum dried polysaccharide (lg) in anhydrous dimethyl sulfoxide (40ml) an excess of methylsulfinyl anion (25ml, 3M) was added and the mixture was stirred for 12h. Methyl iodide (7ml) was added slowly while sti r r i n g and keeping the temperature below 20°. The 110 solution was dialyzed against tap water for 30h then freeze-dried (yield 1.05g). The product showed a slight absorption at 3600 cm"*, and was dissolved in methyl iodide (20ml) and N, N-dimethyl formamide (50ml). To the solution of methylated polysaccharide silve r oxide (5g) was added while st i r r i n g . Another portion of silve r oxide (3g) was added 12h later and the sti r r i n g was allowed to continue for 8h. The reaction mixture was centrifuged and the solid was washed with chloroform. The combined centrifugate and washings were washed with water then dried over anhydrous sodium sulfate and evaporated to dryness. The residue was methylated with Purdie's reagents to give a product showing no hydroxyl absorption at 3600 era"*. The methylated polysaccharide (O.lg) was digested in formic acid (90%) at 97° for l h . The solution was concentrated, dissolved in hydrochloric acid (2M), and heated for 6h. Analysis by paper chromato-graphy (solvent D and F), of the hydrolyzate, after evaporation to dryness, showed the presence of components corresponding to 2,3,4,6-tetra-0-methyl-D-glucose (R f 0.82, solvent D; 0.84, solvent F), 2,4-dir 0-methyl-L_-rhamnose (green color response to p-anisidine spray; 0.65, D; 0.74, F) slightly ahead yet unresolved from 3,4-di-0-methyl-L-rhamnose (brown color response to p-anisidine spray), 2,4,6-tri-O-methyl-D-galactose (R f 0.40, D; 0.62, F), 2,6-di-0-methyl-D-glucose (R f 0.19, D; 0.54, F), and an acidic component (origin, D, Rf ca. 0.25, F). Passage through a column of anion-exchange resin (Duolite A-4) gave a neutral fraction which showed (on paper) the presence of 2,3,4,6-tetra-O-methyl-D-glucose, the mixture of 2,4- and 3,4-di-0-methyl-L-rhamnose, 2,4,6-tri-O-methyl -D-galactose and 2,6-di-0-methyl-D-glucose. When the I l l chromatographic tank was i n i t i a l l y unsaturated with solvent 0 and a few drops of aqueous ammonia were added to the tank, the developed paper chromatogram showed the separation of a mixture of 2,3,4,6-tetra-O-methyl-D-glucose and 2,3,4,6-tetra-O-methyl-D-mannose, The former travelled ahead of the latter and had the same mobility to the assigned 2,3,4,6-tetra-0-methyl-D-glucose in the hydrolyzate. The neutral fraction was reduced with sodium borohydride, acetylated with acetic anhydride and pyridine (1:1) in a sealed^tube at 97° for 20 min, and the solution was evaporated to dryness. The residue was dissolved in a small volume of methanol and analyzed by g.l.c. on column d (212°, a helium flow rate of 91 ml/min) to give four peaks corresponding to the alditoi acetates of 3,4-di-0-methyl-L-rharanose (18.6 min), 2,4-di-0-methyl-L-rhamnose (20.7 rain), 2,3,4,6-tetra-O-methyl-D-glucose (23.6 min), and the mixture of 2,4,6-tri-O-methyl-D-galactose and 2,6-di-0-methyl-D-glucose (35.3 min) in a ratio of 0.7:1:1:2. Samples were collected, and the substitution patterns of the f i r s t three sugars were confirmed by mass spectrometry. A sample of 2,3,4,6-tetra-O-methyi-D-glucitol diacetate was dissolved in methylene chloride (0.5ml) and chilled at -75°. Liquified boron trichloride (2ml, -75°) was added to the methylene chloride solution, and the mixture was kept at that temperature for 30 min, then allowed to stand at room temperature for I6h. After evaporation, and d i s t i l l a t i o n s with methanol, the de-0-methylated product was acetylated and examined on column a to give a peak identical in retention time to D-glucitol hexaacetate. A sample was collected and crystallized, m.p. 93-95°, undepressed by the authentic D-glucitol hexaacetate. 112 The fourth fraction (R- 35.3 min) was further examined on column a, at 220° and a helium flow rate of 50 ml/min, giving two peaks corresponding to the alditol acetates of 2,4,6-tri-O-methyl-D-galactose (7.5 min) and 2,6-di-0-methyl-D-glucose (11.5 min) in a ratio of 1:1. Samples were collected and the substitution patterns of the sugars were confirmed by mass spectrometry. The circular dichroism curve of 2,4,6-tri-O-methyl-D-galactitol triacetate was identical to that of the authentic compound ^^213^ + ^*^°^» thus, the D-configuration of galactose i s confirmed. 2,6-Di-O-methyl-D-glucitol tetraacetate was de-0-methylated (boron trichloride) and acetylated to give (by g.l.c. on column a) D-glucitol hexaacetate, m.p. 92-94°, undepressed by an authentic compound. The acidic fraction eluted from Duolite A-4 column with 10%, formic acid was refluxed in 3% methanolic hydrochloride, neutralized (silver carbonate), reduced with sodium borohydride in methanol, then hydrolyzed. Paper chromatography in solvent D gave a faint spot corresponding to 3,4-di-0-methyl-L-rhamnose (R^ 0.65), and a major spot corresponding to 2,3-di-O-methy1-D-glucose (R f 0.30). The hydrolyzate was concentrated to dryness, reduced, acetylated, and analyzed by g.l.c. on column d, at 211° and a helium flow rate of 60 ml/min, to give peaks with the same retention time as the ald i t o l acetates of 3,4-di-0-methyl-L-rhamnose (18.0 min, (ra)) and 2,3-di-O-methyl-D-glucose (42.9 min, (s)V). Samples were collected and the identity of the sugars was confirmed by mass spectrometry. Reduction of Methylated Polysaccharide To a suspension of lithium aluminum hydride (0.2g) in anhydrous tetrahydrofuran (10ml) a solution of the methylated polysaccharide (O.lg) in anhydrous tetrahydrofuran (10ml) was added dropwise at room temperature with s t i r r i n g . The reaction mixture was stirred overnight, then the excess lithium aluminum hydride was destroyed by cautious addition of ethyl acetate. The whole reaction mixture was evaporated to dryness and the residue was extracted with boiling chloroform (3 X 30ml). The chloroform extract was washed with water and evaporated to dryness, giving O.lg of carboxyl reduced methylated polysaccharide. The carboxyl reduced methylated polysaccharide (O.lg) was o digested in formic acid (90%) at 97 for Ih. After removal of formic acid, the residue was further hydrolyzed with trifluoroacetic acid (2M) on a steam bath for 6h. The solution was evaporated to dryness, dissolved in a small amount of water and examined on paper in solvents 0 and F to show spots corresponding to 2,3,4,6-tetra-0-methyl-D-glucose (R^ 0.82, solvent D; 0.84, solvent F), 2,4-di-0-methyl-L-rhamnose and 3,4-di-O-methyl-L-rhamnose (R_£ 0.65, D; 0.74, F), 2,4,6-tri-O-methyl-D-galactose (R f 0.30, D; 0.59, F) and 2,3- and 2,6-di-0-methyl-D-glucose (j^O.19, D; 0.54, F). The hydrolyzate was reduced with sodium borohydride, acetylated with acetic anhydride and pyridine (1:1) in a sealed tube at 97° for 20 min, and the acetylating reagents were removed by evaporation. The residue was dissolved in a small volume of methanol and analyzed by g.l.c. on column d (221°, a helium flow of 60 ml/min) to give peaks corresponding to the alditoi acetates of 3,4-di-0-methyl-L-rharanose (17.8 min), 2,4-di-0-methyl-L-rhamnose (19.8 min), 2,3,4,6-tetra-0-methyl-D-glucose (22.7 min), the mixture of 2,4,6-tri-O-methyl-D-galactose and 2,6-di-O-methyl-D-glucose (33.6 min), and 2,3-di-O-methyl-D-glucose 114 (42.7 rain) in a ratio of 1:1:1:2:1. Samples were collected. The fourth peak was resolved on column a (220°, a helium flow rate of 50 ml/min) into two components corresponding to the alditoi acetates of 2,4,6-tri-0-methyl-D-galactose (7.8 min) and 2.6-di-O-methyl-D-glucose (12.1 min) in a ratio of 1:1. A sample from the f i f t h peak (R^ 42.7 min) was analyzed by mass spectrometry and was found to be a 2,3-di-O-methyl hexitol tetraacetate. Another sample of this compound was dissolved in methylene chloride (1ml) and chilled at -75°. Liquified boron trichloride (1.5ml, -75°) was added to the methylene chloride solution, and the reaction mixture was kept at that temperature in an acetone-dry ice bath for 30 min, then l e f t at room temperature for I6h. After concentration and d i s t i l l a t i o n s with methanol, the product was acetylated and examined on column a to give a peak corresponding to D-glucitol hexaacetate (m.p. and mixed m.p. 94-96°). Partial Hydrolysis of Capsular Polysaccharide Autohydrolysis of the acidic polysaccharide (pH 2.5) at 97° for 13h gave, by paper chromatography in solvent B, D-glucose as the only neutral sugar; while in solvent A, the presence of D-glucose and a faint spot at 0.39 was observed. Progressive hydrolysis of the polysaccharide in trifluoro-acetic acid (0.5M) at 97° was carried out to determine the optimum conditions for obtaining the oligosaccharides. Ten minutes of hydrolysis released predominantly D-glucose and a minor component corresponding to D-galactose. Half an hour of heating liberated mainly higher oligomers, while heating for 4h gave rise to an aldobiouronic acid as the major component. The presence of L-rhamnose in the 115 oligosaccharides increased the mobilities of these oligosaccharides in paper chromatography using solvent A,B and C. The polysaccharide of K 18 (lg) was hydrolyzed in trifluoro-acetic acid (0.5M) for 4h on a steam bath. The solvent was evaporated at room temperature on a rotary evaporator and the residue was twice d i s t i l l e d with water, then passed through a column of Duolite A-4 (0H~) anion exchange resin. The column was eluted with d i s t i l l e d water until the effluent gave a negative Molisch test (93). The aqueous solution was concentrated to a syrup (0.5g) which was shown by paper chromato-graphy in solvent B to contain L-rhamnose, D-glucose, D-galactose and small amounts of neutral oligosaccharides. Chromatographic separation of the neutral mixture in solvent B on Whatman No. 1 f i l t e r paper gave a disaccharide (Nj_, R^ - 0.20, lOmg). The acidic components in the Duolite A-4 column were displaced with 107. formic acid and the eluent was evaporated to dryness, followed by two d i s t i l l a t i o n s with d i s t i l l e d water to give an acidic mixture (0.19g). Separation of the acidic mixture on Whatman No. 3 f i l t e r paper in solvent C gave the following fractions: A 2 (75mg, 5QJ_c A 1 . 1 0 ) , A| (21mg, D-glucuronic acid), and small amounts of slow moving components. The glucuronic acid was evaporated with the addition of a few drops of concentrated hydrochloric acid. The c.d. curves of the derived glucurono-lactone and that of a standard D-glucuronolactone were both positive ( 4 c + 2.90 and + 3.32, respectively) thus confirming the assignment of D-configuration to the glucuronic acid. In order to obtain higher oligomers, milder conditions for hydrolysis were employed by heating the polysaccharide (2g) in 116 trifluoroacetic acid (0.5M) for 30 min at 97°. The hydrolyzate was evaporated to dryness then d i s t i l l e d twice with d i s t i l l e d water. The syrup was applied on a Celite-charcoal column (10 X 1.4 cm, made from mixing 30g active carbon and 30g Celite) and eluted with water (2 1) followed by 207. aqueous ethanol (1 1) and warm 257. aqueous isopropanol (1.5 I ) . The aqueous solution was concentrated to give a syrup (0.58g) containing (by paper chromatography in solvent B) L-rhamnose, D-glucose, D-galactose and small amounts of oligosaccharides. The ethanol and Isopropanol fractions were evaporated to dryness to give a residue (1.03g) which was subsequently separated on Whatman No. 3 f i l t e r paper in solvent C and gave the following components: A 2 (trace, EQIC A 1*08), Aj^  (trace, glucuronic acid), A 3 (30mg, R ^ A 0.53), A^ (22mg, A 0.44) A. (58rag, R 0.23), A f t (63mg, R 0.07), A. (36mg, R 0.0 4) "~ C l c A ~Glc A 7 ~Glc A and the substance at the origin. Fractions A^  and Ag were each further separated into two components, namely, A ^ , A , and Ag & , A ^ by prolonged paper chromatography (200h) in solvent C. The. oligosaccharides were analyzed as follows. Structural study of neutral disaccharide-(Nj) (a) Gonstituent sugars Component was hydrolyzed with trifluoroacetic acid (2M) o for 6h at 97 , and the hydrolyzate was examined on paper in solvent B to give spots with approximately the~ same intensity equivalent to D-galactose and D-glucose. (b) Methylation Component Nj^  (4mg) was methylated according to the method of Hakomori. The methylated material was hydrolyzed with trifluoroacetic 117 acid (2M) at 97° for 6h. Paper chromatography (solvent D) gave spots corresponding to 2,3,4,6-tetra-0-methyl-D-galactose (R^ 0.68) and 2,3,6-tri-O-methyl-D-glucose (Rj 0.56). The mixture was reduced (sodium boro-hydride), acetylated and analyzed by g.l.c. on column a (200°, a helium flow rate of 30 ml/min) and column d (220°, a helium flow rate of 38 ml/min) to give peaks corresponding to the alditol acetates of 2,3,4,6-tetra-O-methyl-D-galactose (9.5 min, aj 26.2 min, d) and 2,3,6-tri-O-methyl-D-glucose (20.5 min, a; 30.6, d) in a ratio of 1:1. Samples were collected and the identity of the sugars was confirmed by mass spectrometry. (c) Anomeric linkage The p.m.r. spectrum of (in DgO at 95°) gave the following signals of anomeric protons: ?"5.46 (J. 8Hz, 0.5H), 5.40 (J 8 Hz, IH) and 4.78 (J^ 2 3 , 5 H z » 0.5H), indicating *j8-disaccharide. Structural study of aldoblouronic acid (A2) The aldoblouronic acid was found to have £°0n " 12.8° (c 2.9, water). (a) Constituent sugars (2mg) was hydrolyzed with hydrochloric acid (2M) at 97° for 6h and the hydrolyzate was examined on paper in solvent B then A to give spots equivalent to D-glucuronic acid and L-rhamnose. The same sugars were released when A2 (lrag) was digested in c i t r i c buffer (0.2ml), pH5 with y& -glucuronidase (lmg) at 37° for lOh. (b) Lithium aluminum hydride reduction and reducing end analysis A portion of A2 (30mg) was reduced with sodium borohydride in water and the reduction mixture was l e f t at room temperature over-night. One tenth of the disaccharide ald i t o l was hydrolyzed in hydro-chloric acid (2M) at 97° for 7h. The hydrolyzate was examined on duplicate f i l t e r 118 papers in solvent A. One of the paper chromatograras was developed with silver nitrate-sodium hydroxide to reveal the components corresponding to D-glucuronic acid and L-rhamnitol while p-anisidine spray of the other chromatogram only revealed the presence of D-glucuronic acid. The rest of the disaccharide alditoi crystallized from the syrup, m.p. 118-121°. (c) Methylation &2 (ISmg) was methylated according to the procedure of Hakomori and Purdie. One third of the methylated material was hydrolyzed with hydrochloric acid (2M) at 97° for 7h. Examination of the hydrolyzate on paper in solvent D gave spots corresponding to 3,4-di-0-methyl-L-rhamnose (R^ 0.65, brown in response to p-anisidine spray) and an acidic component at the origin. The hydrolyzate was de-ionized by passage through a column of Duolite A-4 (OH") anion exchange resin and the neutral eluent was concentrated, reduced with sodium borohydride and acetylated. The derived aldi t o i acetate was dissolved in a small volume of methanol and examined by g.l.c. on column d (215°, helium flow rate 100 ml/min) to give a peak identical in retention time to 3,4-di-0-methyl-L-rhamnose (10.8 min). A sample was collected and i t s identity was confirmed by mass spectrometry. The rest of the methylated was reduced with sodium borohydride in methanol for 24h, then hydrolyzed in trifluoroacetic acid (2M) at 97° for 8h. The hydrolyzate was found on paper in solvent D to contain 3,4-di-0-methyl-L-rhamnose (R^ 0.65) and 2,3,4-tri-O-methyl-D-glucose (R^ 0.59). Reduction and acetylation of the mixture gave rise to two components on g.l.c. (column d, 215°, helium flow rate 100 ml/min) corresponding to the alditoi acetates of 3,4-di-O^nethyl-L-rhamnose (10.8 min) and 119 2,3,4-tri-0-methyl-D-glucose (17.7 min). Samples were collected and the identity of the second component was confirmed by mass spectrometry, (d) Anomeric linkage A sample of A (20mg) was dissolved in DO and the p.m.r. measurement was run at 95° to give signals at 7J" 8.70 (J^ g 6 Hz, 3H, CH3 of L_-rhamnose), 5.25 ( J x % 7 Hz, IH), 5.14 (0.3H) and 4.63 Ul 2 Hz, 0.7H). ' 1 2 Therefore the structure of A 2 is D-GlcAp L-Rha. Structural study of A^ Aj was found to have (c*) n +50.0 (c 2.9, water). (a) Constituent sugars A 3 (6mg) was hydrolyzed in trifluoroacetic acid (0.1M) at 97° for 6h to give (by paper chromatography in solvent B then A) spots corresponding to D-glucose (major), A^ (major) and traces of intact A^ and L-rhamnose. The solvent was evaporated and the residue was further hydrolyzed in hydrochloric acid (2M) at 97° for 6h to give L-rhamnose, D-glucose, and D-glucuronic acid. The hydrolyzate was passed through a column of Duolite A-4 (OH**) anion exchange resin to remove acidic components, then reduced and acetylated. The a l d i t o l acetates were analyzed by g.l.c. on column a at 200° and a helium flow rate of 85 ml/min to give peaks corresponding to the ald i t o l acetates of L-rhamnose (8.7 min) and D-glucose (36.3 min) in a ratio of 1:1. (b) Methylation A^ (20rag) was methylated according to the method of Hakomori and Purdie. Half of the methylated A^ was hydrolyzed (2M hydrochloric acid, 97°, 7h) and chromatographed in solvent D to give 2,3,4,6-tetra-120 O-methyl-D-glucose (R f 0.80), 3,4-di-0-methyl-L-rhamnose (R f 0.65) and an acidic component at the origin. The hydrolyzate was passed through a column of Duolite A-4 (0H~) anion exchange resin and the neutral effluent was concentrated, reduced and acetylated. The product was analyzed by g.l.c. on column d at 210° and a helium flow rate of 86 ml/min to give peaks corresponding to the aldit o i acetates of 3,4-di-0-methyl-L-rhamnose (14.8 min) and 2,3,4,6-tetra-O-methyl-D-glucose (16.9 min) in a ratio of 1:1. Samples were collected and the identity of the sugars was confirmed by mass spectrometry. The rest of the methylated A3 was reduced in methanol with sodium borohydride then hydrolyzed, giving (by paper chromatography in solvent D) 2,3,4,6-tetra-0-methyl-D-glucose (R f 0.81), 3,4-di-0-raethyl-L-rhamnose (R f 0.65) and 2,3-di-0-methyl-D-glucose (Rj 0.29). The hydrolyzate was reduced, acetylated and analyzed by g.l.c. on column d at 211° and a helium flow rate of 100 ml/min to give peaks corresponding to the al d i t o i acetates of 3,4-dl-0-methyl-L-rhamnose (18.4 min), 2,3,4,6-tetra-0-methyl-D-glucose (23.5 min) and 2,3-di-0-methyl-D-glucose (43.9 min) in a ratio of 1:1:1. (c) Anomeric linkage A sample of A3 was dissolved in D20 and the p.m.r. spectrum was run at 95° to show sharp peaks at T 8.73 (J' - 6 Hz, 3H, CH. of 5,6 J L-rhamnose), 5.30 (^ 2 8 Hz, IH), 5.15 (0.3H), 4.69 (^ • 1 Hz, 0.7H) and 4.55 (^ 2 3 Hz, 1H). A 3 is therefore composed of A 2 and D-glucose and can be 1 4 1 2 written as the following pseudotriouronic acid, D-Glcp-^- D-GlcAp-^- L-Rha. 121 Structural study of higher oligomers A. was found to have (ex) + 30.0 (c 3.27. water). A^ was hydrolyzed in hydrochloric acid (2M) at 97° for 7h to give (by paper chromatography in solvent B then A) L-rhamnose, D-glucose, and D-glucuronic acid. Slow development of dark color by silver nitrate-sodium hydroxide is an indication that L-rhamnose i s at the reducing end. A sample of A^ (lOmg) was methylated according to the method of Hakomori, then hydrolyzed and examined on paper in solvent D. The chromatogram showed the presence of 2,3,4,6-tetra-0-methyl-D-glucose (R f 8.0), the mixture of 2,4- and 3,4-di-O-methyl-L-rhamnose (R^ 0.65), 2,3,6-tri-O-methyl-D-glucose (R^ 0.57, trace) and 2,4,6-tri^O-methyl-D-glucose (Rf 0.49). G.l.c. analysis on column d at 200° and a helium flow rate of 76 ml/min of the aldi t o i acetates gave peaks corresponding to the aldi t o i acetates of 3,4-di-0-methyl-L-rhamnose (17.4 min), 2,4-di-O-methyl-L-rhamnose (19.9 min), 2,3,4,6-tetra-O-methyl-D-glucose (21.3 min) and 2,4,6- and 2,3,6-tri-O-methyl-D-glucose (25.5 min) in a 1 3 ratio of 2:1.3:1:2. A^ i s therefore a mixture containing D-Glcp -^p D-Glcp ~ ~ D-GlcAp L-Rha as the major component. The p.m.r. spectrum of A^ was run in D20 at 95°. Two different L-rhamnoses were obvious from the signals of CH 3»s at 8.69 ( J 5 g 6 Hz, (s) ) and 8.67 ( J $ 6 6 Hz (w) ). Four anomeric protons were found by the presence of the following peaks: ^5.30 (^ 2 7 Hz, IH), 5.13 (0.4H), 4.89 (IH), 4,77 (0.6H) and 4.55 (IH). The presence of only one J} -linkage i s a contribution from D-glucuronic acid, as also shown in A 2 and A^. The rest of sugars (except the one at the reducing end) have c*. -linkages as 122 represented by the proposed structure of the major component. A^ a was found to have {W^  + 38.0 (c 1.37, water). Hydrolysis of and paper chromatography in the solvent B then A showed the presence of L-rhamnose, D-glucose, D-galactose, and D-glucuronic acid. The p.m.r. spectrum of A 5 a was run at 95° in D20 and showed two different CHo's of L-rhamnose with signals at 8.69 ( J . , 6 Hz, = P t o (s) ) and 8.67 (J 6 Hz, (w) ). Comparison of the integration of 5,6 the area of CH^'s of L-rhamnose to that of the anomeric protons indicates that there are two L-rhamnoses to five anomeric protons with signals at T5.72 ( J t % 7 Hz, 0.3H), 5.50 2 7 Hz, IH), 5.31 2 7 Hz, IH), 4.91 (IH) 4.80 ( j 2 Hz, 0.7H) and 4.45 (IH). The presence of an extraj3 -linkage (T'5.50) accompanying the additional sugar moiety of D-galactose as compared to N^, A^ and A^ suggests that D-galactose i s ^ - l i n k e d . A^ a (lOmg) was methylated according to the procedure of Hakomori, and hydrolyzed to show (on paper in solvent D) the presence of 2,3,4,6-tetra-O-methyl-D-glucose (R f 0.80, minor), 2,3,4,6-tetra-0-methyl-D-galactose (R^ 0.68), 3,4- and 2,4-di-0-methyl-L-rhamnose (R f 0.65), 2,3,6-tri-O-methyl-D-glucose (R f 0.57, minor), 2,4,6-tri-O-methyl -D-galactose (R^ 0.39 and 2,6-di-O-methyl-D-glucose (R f 0.19). G.l.c. analysis of the aldi t o l acetates on column d did not give a proper sugar ratio for the assignment of structure. Therefore A^ a is a mixture. Ag a was found to have (otJQ + 62.0 (c 1.27, water). The p.m.r. spectrum of A, in D_0 was run at 95°. Two oa 2 different L-rhamnoses were evident by the signals at T ' S ^ (J 6 Hz, (s) ) = 5,6 123 and 8.67 (J_ 6 Hz, (m) ). Comparison of the area of CH,»s of 5,6 J L-rhamnoses and that of the anomeric protons suggests the presence of two L-rhamnoses in six sugar units with signals of anomeric protons at T 5.72 Ux 2 7 Hz, 0.3H), 5.50 ( J ^ 2 7 Hz, IH), 5.31 ( J ^ 2 7 Hz, IH), 4.91 (J. „ 3 Hz, IH), 4.72 (J, _ 2 Hz, 0.7H), 4.53 (J, . 2 Hz, IH) 1,2 and 4.45 (IH). Methylation study of Ag a gave a complex mixture of partially methylated monosaccharides as found in A 5 a with a slightly different sugar ratio, and the assignment of a single structure i s impossible. Therefore A^ a is also a mixture. A5b' A6b a n 0" A7 w e r e similarly analyzed and found to be mixtures. Periodate Oxidation Capsular polysaccharide (0.977g) was dissolved in an aqueous solution of sodium periodate (0.05M, 200ml) and the solution was kept at 1°C in the dark. One ml of the oxidizing solution was withdrawn and diluted to 1000 ml with d i s t i l l e d water for the measurement of the unreacted periodate ion at 222.5 ay/. The periodate uptake per sugar unit leveled off at 0.64 mole In 360h. The reaction was stopped by adding ethylene glycol (2ml) and the mixture was l e f t standing at room temperature for 30 min, dialyzed for 24h, reduced (sodium borohydride), again dialyzed, deionized, and freeze-dried. The yield of polyol was 0.77g. The polyol (20mg) was hydrolyzed with hydrochloric acid (2M) o at 97 for 7h. Paper chromatography in solvent B gave D-galactose, D-glucose, L-rhamnose (R(;i c 3.13), glycerol (IL 3.35), D-erythronolactone 124 (JR^ 5.26; R^ 0.60, standard obtained from the periodate degradation of K 5), and 1-deoxyglycerol (R 8.95; R^  0.80, 1,2-propanediol as Glc —%, standard). The lactone was distinguished by the hydroxylamine.ferric chloride spray. The hydrolyzate was reduced with sodium borohydride and, after removal of the borate, the product was acetylated in a sealed tube by using pyridine and acetic anhydride (1:1, 20 min, 97°). After the evaporation of the solvents, the a l d i t o l acetates were dissolved in a small volume of methanol and examined by g.l.c. on column a programmed from 160°at l°/min and a helium flow rate of 43 ml/min to give four peaks identical in retention times to the a l d i t o l acetates of glycerol (5.2 min), erythritol (16.4 min, m.p. and mixed m.p. 80-82°), L-rhamnose (29.2 min), D-galactose (60.8 min, m.p. and mixed m.p. 160-162°), and D-glucose (64.1 min, m.p. and mixed m.p. 98-99°) in a ratio of 0.4:1:1:1:1. A loss of glycerol triacetate during the operations is obvious. A sample of the polyol (200mg) was dissolved in t r i f l u o r o -acetic acid (2M) at room temperature for 8h to cleave preferentially the side chain acetal group derived from D-glucose at the branch chain and the deoxyglycerol derived from>the degraded L-rhamnose. T A f t e r dialysis against running water and deionization, the product was freeze-dried to give a degraded polyol (121mg). The degraded polyol was hydrolyzed (2M hydrochloric acid, 97°, 7h) and examined on paper in solvent B to contain D-galactose, D-glucoee, L-rhamnose and D-erythronolactone (Kglc 5.26). A separate portion of the degraded polyol (40mg) was methylated according to the method of Hakomori, dialyzed against running water for 24h, then hydrolyzed and examined on paper in solvent D and F to give 125 2,4-di-0-methyl-L-rhamnose (R^ 0.62 solvent D; 0.73, solvent F), 2,3,6-tri-O-methyl-D-glucose (R^ 0.54, D; 0.70, F) and 2,4,6-tri-0-methyl-D-galactose (R f 0.35, D; 0.53, F). Authentic standards were used for comparison. The remaining hydrolyzate was reduced (sodium borohydride), acetylated and analyzed by g.l.c. on column a (180°, helium flow rate 55 ml/min) and column d (180°, helium flow rate 60 ml/min) to give the alditoi acetates of 2,4-di-O-methyl-L-rhamnose (6.1 min, column a; 15.6 min, column d), 2,3,6-tri-O-methyl-D-glucose (16.3 min, aj 26.5 min, d) and 2,4,6-tri-O-methyl-D-galactose (14.7 min, a; 30,9 min, d) in a ratio of 1:1:1. Samples were collected and the structures of the sugars were confirmed by mass spectrometry. The acid degraded polyol (30mg) was twice dissolved in DO and freeze-dried. The p.m.r. spectrum of the solution (in D 20) was run at 95° and showed a sharp signal at 1^8.65 ( J 5 g 6 Hz, 3H, CH3 of L-rhamnose) and signals of anomeric protons a t T 5.43 (lH,^*-linkage) and 4.83 (2H, -1inkages). Same mild Smith degradation of the polyol was carried out in trifluoroacetic acid (2M) at room temperature for 4h and 12h in addition to 8h. Similar degree of hydrolysis of the polyol was observed by the analysis of the products. 126 APPENDIX A: Immunochemical Studies on Klebsiella K-Type 5 Polysaccharide Many serological studies and cross-reactions have been reported in the Klebsiella group (2,3,8,12,13,94,95). However, quantitative inhibition studies of the homologous antisera, using oligosaccharides derived from the parent capsular polysaccharides have not been reported. Data were desired on the relative antigenic potency of the subunits of the polysaccharides as well as the size of the antibody reacting site in terms of the size of the oligosaccharide inducing 100% inhibition of the antiserum for the subsequent reaction with the homologous antigen. In the course of the structural study of K5, i t was found that this capsular polysaccharide was composed of such potentially antigenic deterrainents as D-glucuronic acid, pyruvic acid and acetyl group, and that a series of oligosaccharides from the capsular poly-saccharide were available (37). These properties make K5 an excellent material for the immunochemical study i f a potent homologous antiserum could be e l i c i t e d from the experimental animals. It has been found that purified capsular polysaccharides from Klebsiella are immunogenic only to mice and men but not immunogenic to rabbits (96). 0rskov (8) had also reported immunological paralysis induced in rabbits by a heavilycapsulated Klebsiella K5 strain. In this study, attempts have been made to immunize rabbits and mice with Klebsiella K-type 5 bacteria using either whole c e l l s or the purified capsular polysaccharide. The schedule for immunization and bleeding was as follows. New Zealand albino rabbits (three month old) No. 1 and 2 were 127 immunized intravenously with 0.1ml of bacteria suspended in saline (1.2XK?ce&sjmO which was k i l l e d by heating the bacterial culture and washed by saline. The subsequent immunizations were carried out using bacteria k i l l e d by formalin (0.5%). The animals were immunized every 4 days unt i l 4 injections were given. The doses of the 2nd, 3rd and 4th injections were 0.2, 0.4 and 0.8ml of a c e l l suspension of the same concentration. Both rabbits were bled 3 days after the 3rd immunization. Eight days after the 4th injection, the animals were challenged with 1 ml of r l i v i n g bacteria which had been cultured for 48h. Both rabbits died of the challenge. Rabbits No. 3 and 4 were immunized with formalin-killed 8 bacterial culture (1.8 X 10 cells/ml) and challenged by li v i n g bacteria according to the schedule stated in the preceding paragraph. The rabbits were bled by cardiac puncture 4 days after the challenge. A group of three Swiss albino mice (8 weeks old, ca. 20g in weight) was given 4 subcutaneous injections with 0.2ml of K5 polysaccharide (17. in saline) at 4 day intervals. Fourteen days after the last injection the mice were challenged with 0.1ml of a liv e culture (6 X 10 cells/ml). Bleeding was performed 4 days after the challenge using the technique of cutting the t a i l vein. Another group of three Swiss albino mice were immunized with I Q 0.2ml of a formalinized, washed broth culture (3 X 10* cells/ml) by giving 4 injections at 4 day intervals. Fourteen days after the last immunization the mice were challenged with 0.1ml of a l i v i n g broth culture of the bacteria. One mouse died on the 9th day of challenge. The rest were further challenged 10 days after the f i r s t challenge with 128 o 0.2ml of•living bacteria (6 X 10 cells/ml). Mice were bled 4 days after the second challenge by cutting the t a i l vein. For the preliminary test of the antisera, the ring precipitin test was employed by layering 0.15ml of 0.5, 0.1, 0.02 and 0.004% of the capsular polysaccharide in saline (pH adjusted to 8 by IM sodium hydroxide) over 0.15ml of serum in narrow tubes at room temperature. No precipitation was observed over a period of 24h for a l l the sera from rabbits and mice. Ouchterlony gel diffusion tests were also performed on each serum by putting the serum in the central well of an Ionagar plate which was surrounded by five perlferal wells containing 10-fold serial dilutions of the antigen beginning from 0.5% of the polysaccharide in saline (pH adjusted to 8 by IM sodium hydroxide). No precipitin bands were observed up to 7 days. Plate agglutination tests were conducted on the heat-killed K 5 bacteria by gently mixing the bacteria with each of the serial dilutions of a serum which had been heated at 56° for 30 min to inactivate the complement. No flocculation was observed for any of the tested sera. APPENDIX B: P.M.R. SPECTRA 130 K5 CAPSULAR POLYSACCHARIDE 8.0 SWEEP OFISEl (Hi): SPECTRUM AMPLITUDE INTEGRAL AMPLITUDE SPINNING UAH HPS) 2«f CQROiaQ C M A i i t a J 7.0 6.0 MANUAL 5.0 4.0 I AUTO [3] 1 SAMPIE; I UM 3.0 1 • 2.0 REMARKS: 1.0 ^ SWEEP IIME (SEC): i M M u » i , . SWEEP WIDTH (Hi): jfV 1» j«y">»l,jy 1 IK»I FILTER: 1^1' I'J'J'.I' i.'ii.l ! I > l | JJi^ RF POWER LEVEL: . O.O\Q | I o> I , SOLVE II:')-0 K 5 PotroL DATE: . . . \ T \ \ I 1 •* OPERATOR: . P.O. TWICE PERIODATE OXIDIZED AND SODIUM BOROHYDRODE REDUCED K5 60 MHl NMR SPECTRUM NO. 5 ro 133 METHYLATED K5 134 METHYLATED AND REDUCED K5 135 ! I I I ! ! I ! 1 U U I I 1 ! I I ! 1 I I I ! I '~~T~ I 1 I ' ' I PPM : ' 1 I • • • i | ! ' ; i | ! f ! ! | ! I i i | i i ! i | i i i ! | i i i : | i i ! i | i i i ; | i ; i ! | : i ; ' | -CELLOBIOURONIC ACID (Al) FROM K5 136 I I J I i t i I L U — L _ J i i i l i i J—!— I i~~T- l i i i i I I I | ; i i i | i i i i i i I I i i 1 TT~1 i . i . I T ~ n — | i i i i | i I T i l l M i l -- . .-U... . ! . i :•<> .Gfi'-fGlcjinan. • 5.U- • 500 10° • ! i i • •• I t ' l l ' 1 1 1 I I I I I- I ' t I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I 1 I ALDOTRIOURONIC ACID (A2) FROM K5 137 NEUTRAL DISACCHARIDE (N3) FROM K5 PYRUVIC ACID-MANNOSE-ERYTHRONIC ACID COMPLEX FROM K5 139 K62 CAPSULAR POLYSACCHARIDE 140 i i i i ' i i i i i 11 i i i i i i i i i i i i i i i r~ i i i i i i P P M I I I I I I I I ! I I I I ! | I I I I | I I I I | ! I I I | I j I I | M I I | I I I I | I I I I | I I i i i i | i i i i 111 i i i | ; i i i | I I I I | I I I I | I I I I | I I I MILDLY DEGRADED (DILUTE ACID) K62 141 i i i i I i i II i I | | 1 I 1—!—I—I—L J ' 1 1 1 1 ! 1 1 1 1 1 1 1 ,1 \ri 1 1 1 • i._J ...L J _J—1—1—1—LJ—1 l_l t i i : i i i i i T - r l 111 i i i i i T T T T r n - i i i i i i i i " i ALDOBIOURONIC ACID ( A 2 ) FROM K 6 2 142 ALDOTRIOURONIC ACID (A3) FROM K6 2 143 ' ' ' » " ' ' H i "I ' i i " i i i ' i ' i - t i "'I . i—r J—I I I I I I I I _L "' M i i ; i | ; i i i | | i | | i l l ! ' I ' I I I I I j I I I | ] | | | | | "| 1 ,' ' ' ' I ' ', ' ' , ' 1 ' ', ,' ' I ' I ' I I I | I I I I | | I | , | | | | | | | ; - r - T 1500-I r I l I i i i i„ i.. i i i i | i i 7: ....)., 100'.; I' I I I / 3 ~i r r r T PSEUD0TETRA0UR0NIC ACID (A4) FROM K62 144 DISACCHARIDE GLYCOSIDE FROM SMITH DEGRADATION OF K62 1 4 5 J I I I LU i i i i I I I I L I I l ~ T I I ' 1 ' ' ' ' ' ' i • i ' i I M I I | I I I I | i I I ! I. | I I I I I I I I I I J . I I I : | I I I I | ~\—i— i—i— i—rh— i—i—i—i—i—i—r-„ i , I,.! I I I H I I I • I I II I f, I I I I I JL_J—l—L—L I I i ' l . l I I I I | I I I I I I I • ! I I I I I I I I 1 I I I I ! I ! I I I I I I I I I III I I I I I I I I I I I I I I I I ! I t T T 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 I I III K18 CAPSULAR POLYSACCHARIDE 146 n i i i i i i i i i i r ~ r i i L * t M I ' ' i 1 i i i i I i ; i i | i i i i I i i i i 1 i ! i i I , i i ! i i i i i i i u i i i ALDOBIOURONIC ACID (A2) FROM K18 147 "B—J 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I I I t I I I I I I I I I I II I NEUTRAL DISACCHARIDE (Nl) FROM K18 148 J 1 ! l I I I I I " I I I I | PSEUDOTRIOURONIC ACID (A3) FROM K18 149 i i 11 i I i l I i l i i i i i i i r r i i i i i i i i i , i i i i , , , • , I ' I 11 ' ' I I i I I* I I I' I I I" I'"I i T r w r i Tl I I T T 1 T"l t I | I • I I TETRASACCHARIDE (A4) FROM K18 150 A5a FROM K18 151 I i it I l I l l I l i i i i i i i i i i i i i i i i i i i i i i i 't i- n I • P P M . ' ! 1 1 1 ' 1 1 ' ' ' I ' I I I I "I I I 1 I i I I I I I I I I I V I I I I -J A6a FROM K18 152 i i II i i J L J I I L - l _ J ! I I L_J I i i I ' i j m_ MILD ACID TREATED POLYALCOHOL FROM K18 153 APPENDIX C: Expressions of Raw Sugar Ratio Data from G.L.C••Ana l y s i s Raw sugar r a t i o s g . l . c . a n alysis have been rounded o f f to the nearest i n t e g r a l numbersi to represent the simple i n t e g r a l molar r a t i o of the repeating unit i n the polysaccharides as well as the r a t i o s of the monosaccharides in the oligosaccharides derived from the polysaccharides. The relati'onshlpbbet'Weentt'heaaetiual experimental values represented by the p r i n t outs of peak areas by the d i g i t a l integrator to the corresponding rounded o f f sugar r a t i o s i s exemplified by the following seven examples. i :-l H 1 I I I i ! i i I i i ! I IliTil ANALYSIS OF K5 CAPSULAR POLYSACCHARIDE AFTER GLUCURONIC ACID HAS BEEN TRANSFORMED INTO GLUCOSE 01 -o i t ! h r r ' J IT! •"?" j n i l n-HJ! ri I I I ! I i ANALYSIS OF AUTO HYDROLYZED K5 1 2- !-I ! -l! I I i i i i i i .1 : j i ! i I-: -I J;! j J! | i : ! ! H I M ! .11 i i • i U i i I I ANALYSIS OF METHYLATED K62 CAPSULAR POLYSACCHARIDE ANALYSIS OF THE METHYLATED, REDUCED K18 OO -----| ••• T I - | I - . T : I ; i• i - : !-{'-!"! |.;. .1....).,.. ;.i ••hlTr'rf-TI Lj:!:; •I •!--1 H7 7 i 5 7 5 3 , 1 7 9 1 * 8 2 5 -i ]r':':-r .1! b j t . '^ " • I il. • r:].:::/ II: :!: j.FH'fi •it[-:.f:n. t-H mmkmm A *.<£..b:±L it-lr-fl [./ip^:J{:.:.|./.a :::; j [;; ii Liu;: H:; • i ; i RESOLUTION OF PEAK NO.4 IN PAGE 158 TREATED POLYALCOHOL FROM K18 o l'6-I BIBLIOGRAPHY 1. R.R. Porter, Essays Biochem., 3> 1-20 (1967). 2. S.D. Henriksen and Jorunn Eriksen, Acta path. et. microbiol. scandinav., 51^ 259 (1961). 3. M. Heidelberger and G.G.S. Dutton, J . Immunol., 111. 857 (1973). 4. W.F. Dudman and J.F. Wilkinson, Biochemical J., 62, 289 (1956). 5. P. J . Garegg, B. Lindberg, T. Onn, and T. Holme, Acta Chem. Scand., 25, 1185 (1971). 6. P. J. Garegg, B. Lindberg, T. Onn, and I.W, Sutherland, Acta Chem. Scand., 25, 2103 (1971). 7. K.J. Carson and R.G. Eagon, Can. J . Microbiol., 10, 467 (1964). 8. I. 0rskov, Acta path, et microbiol. scandinav., 38., 375 (1956) 9. E.A. Kabat, Blood Group Substances, Their Chemistry and Immuno-chemistry. New York, Academic Press, Inc., 1956. 10. T. Hubscher and A.H. Eisen, Internat. Arch. Allergy and Applied Immunology, 42, 466 (1972). 11. 0. Luderitz, K. Jann and R. Wheat, Comprehensive Biochem., 26A. 105 (1968). 12. W. Nimmich, Z. Med. Mikrobiol. Immunol., 154. 117 (1968). 13. W. Nimmich, Acta b i o l . med. germ., 26, 397 (1971). 14. R.W. Wheat, C. Dorsch and G. Godoy, J. Bacterid., 89, 539 (1965). 15. P.A. Sandford and H.E. Conrad, Biochemistry, 5, 1508 (1966). 16. H.E. Conrad, J.R. Bamburg, J.D. Epley, and T.J. Kindt, Biochemistry, 5, 2802 (1966). 17. S.I. Hakomori, J. Biochem. (Tokoyo), 55, 205 (1964). 18. S. Hirase, Bull, Chem. Soc. Japan, 30, 75 (1957). 19. ^  J.H. Sloneker and O.G. Orentas, Can. J . Chem., 40, 2188 (1962). 20. Y.M. Choy and G.G.S. Dutton, Can. J. Chem., 52, 684 (1974). 21. P.A.J. Gorin and J.F.T. Spencer, Can. J. Chem., 42, 1230 (1964). 22. Y.M. Choy and G.G.S. Dutton, Can. J. Chem., M, 198 (1973). 162 23. P.J. Garegg, B. Lindberg, T. Onn, and T. Holme, Acta Chem. Scand., 25, 1185 (1971). 24. P.J. Garegg, B. Lindberg, T. Onn and I.W. Sutherland, Acta Chem. Scand., 25, 2103 (1971). 25. B.J. Gormus, R.W. Wheat, and J.F. Porter, J . Bacteriol., 107. 150 (1971). 26. M. Heidelberger, W.F. Dudman, and W. Nimmich, J. Immunol., 104. 1321 (1970). 27. L.C. Gahen, P.A. Sandford, and H.E. Conrad, Biochemistry, 6, 2755 (1967). 28. I.W. Sutherland, Biochemistry, 9, 2180 (1970). 29. B. Lindberg, J. Lonngren, and J.L. Thompson, Carbohydr. Res., 25, 49 (1972). 30. H. Bjorndal, B, Lindberg, J . Lonngren, K. Rosell, and W. Nimmich, Carbohydr. Res., 27, 373 (1973). 31. Y.M. Choy and G.G.S. Dutton, Can. J. Chem., 51, 3015 (1973). 32. Y.M. Choy, G.G.S. Dutton, and A.M. Zanlungo, 51, 1819 (1973). 33. B. Lindberg, K. Samuelsson, Carbohyd. Res., 30_, 63 (1973). 34. H. Bjorndal, B. Lindberg, J . Lonngren, and M. Meszaros, Carbohyd. Res., 31, 93 (1973). 35. Y.M.Choy and G.G.S. Dutton, Can. J. Chem., 51, 3021 (1973). 36. G.G.S. Dutton and M.T. Yang, Can. J. Chem., 50, 2382 (1972). 37. G.G.S. Dutton and M.T. Yang, Can. J. Chem., 51, 1826 (1973). 38. CK. De-Bruyne and J . Wouters-Leysen, Carbohyd. Res., 17_, 45 (1971). 39. P. Albersheim, D.J. Nevis, P.D. English, and A. Karr, Carbohydr. Res., 5, 340 (1967). 40. B. Capon, Chemical Review, 69, 407 (1969). 41. G.A. Adams, Methods Carbohyd. Chem., 5, 269 (1965). 42. D.G. Medcalf and K.A. G i l l e s , Cereal Chem., 45, 550 (1968). 43. A.S. Perlin, Cereal Chem., 28, 382 (1951). 44. B. Lindberg and J; Lonngren, Carbohydr. Res., 23, 47 (1972). 45. C.T. Bishop and F.P. Cooper, Can. J. Chem., 41, 2742 (1963). 163 46. V. Smirnyagin and C.T. Bishop, Can. J . Chem., 46, 3085 (1968). 47. V. Smirnyagin, C.T. Bishop, and F.P. Cooper, Can. J. Chem., 43, 3109 (1965). 48. G. Entlicher and J.N. BeMiller, Carbohydr. Res., 16, 363 (1971). 49. P.A. Sandford, P.R. Watson and A, Jeans, VI International Symposium on Carbohydrate Chemistry. Madison, U.S.A. August 1972. Abstracts of paper presented. P. 35. 50. Y.M. Choy, G.G.S. Dutton, A.M. Stephen and M.T. Yang, Anal. Lett., 5, 675 (1972). 51. G.M. Bebault, Y.M. Choy, G.G.S. Dutton, N. Funneli, A.M. Stephen, and M.T. Yang, J. Bacteriol., 113, 1345 (1973). 52. M. Abdel-Aker and F. Smith, J . Amer. Chem. Soc, 73, 5859 (1951). 53. R. Kuhn, H. Trischmann, and I. Low, Angew. Chem., 67, 32 (1965). 54. T. Purdie and J.C. Irvine, J . Chem. Soc, 83, 1021 (1903). 55. G.G.S. Dutton and S. Kabir, Anal. Lett., 4, 95 (1971). 56. P.D. Bragg and L. Hough, J. Chem. Soc, 4347 (1957). 57. S.C. Churms, Advances Carbohydrate Chem., 25, 13 (1970). 58. I.J. Goldstein, G.W. Hay, B.A. Lewis and F. Smith, Methods Carbohydrate Chem., 5, 361 (1965). 59. B. Lindberg, Private Communication. 60. R.J. Yu and C.T. Bishop, Can. J. Chem., 45, 2195 (1967). 61. T. Painter and B. Larsen, Acta Chem. Scand., 24, 813 (1970). 62. M.F. Ishak and T. Painter, Ibid., 25, 3875 (1971). 63. A.M. Stephen, J . Chem. Soc, 2030 (1962). 64. CC. Sweeley, R. Bentley, M. Makita and W.W. Wells, J . Amer. Chem. Soc, 85, 2497 (1968). 65. S.W. Gunner, J.K.N. Jones, and M.B. Perry, Can. J . Chem., 39, 1892 (1965). 66. CO. Aspinall, J. Chem. Soc, 1676 (1963). 67. J.S. Sawardeker, J.H. Sioneker and A. Jeanes, Anal. Chem., 37, 1602 (1965). . 164 68. H. Bjorndal, C.G. Hellerqvist, B. Lindberg and S. Svenssen, Angew. Chem. internat. Edit., 9, 610 (1970). 69. G.O. Aspinall, J . Chem. Soc, 1676 (1963). 70. E. Percival, Carbohydr. Res., 4, M l (1967). 71. A.G. Cooper, J.F. Codington, and M.C. Brown, Proc, Nat. Acad. Sci. U.S.A., 71, 1224(1974). 72. D. Allan, J . Auger and M.J. Crumpton, Nature New Biology, 236. 23 (1972). 73. K. Biemann, "Mass Spectrometry, Application to Organic Chemistry" McGraw-Hill, New York, 1962. 74. M.S.B. Munson and F.H. Field, J . Amer. Chem. Soc, 88, 2621 (1966). 75. H.R. Schulten and H.D. Beckey, Org. Mass Spectrometry, J_, 861(1973). 76. W. Heyns and H. Scharmann, Liebigs Ann. Chem., 667. 183 (1963). 77. O.S. Chizkov, N.N. Moldostov and N.K. Kochetkov, Carbohydr. Res., 4, 273 (1967). 78. G. Pettersson, 0. Samuelson, K. Anjou and E. von Sydow, Acta Chem. Scand., 21, 1251 (1967). 79. O.S. Chizhov, L.S. Golovkina and N.S. Wulfson, Izv. Akad. Nauk. SSSR, Ser. Khim., 1915 (1966). 80. J . Hoffman, B. Lindberg and S. Svensson, Acta Chem. Scand., 26, 661 (1972). 81. S.L. Patt and B.D. Sykes, J . Chem. Phys. 56, 3182 (1972). 82. G.M. Bebault, J.M. Berry, Y.M. Choy, G.G.S. Dutton, N. Funnell, L. D. Hayward, and A.M. Stephen, Can. J. Chem., 51, 324 (1973). 83. M. Duckworth and W. Yaphe, Chem. Ind. (London), 747 (1970). 84. J.H. Sloneker and Jeanes, Can. J. Chem., 40, 2066 (1962). 85. J.O. Deferrari, E.G. Gros, and 1.0. Mastronardi, Carbohydr. Res., 4, 432 (1967). 86. B. Lindberg and L. Selleby, Acta Chem. Scand., 14, 1051 (1960). 87. J.G. Slonecker, D.G. Orentas, C.A. Knutson, P.R. Watson, and A. Jeanes, Can. J. Chem., 46, 3353 (1968). 88. A.N. de Belder and B. Norrman, Carbohydr. Res., 8, 1 (1968). 165 89. R.W. Bailey, Oligosaccharides. Pergamon, Oxford, 1965, p. 99. 90. J.K.N. Jones and M.B. Perry, J. Aner. Chem. Soc, 79, 2787 (1957). 91. R.L. Whistler and D.F. Durso, J . Amer. Chem. Soc, 72, 677 (1950). 92. H. BjSrndal, B. Lindberg and S. Svensson, Acta Chem. Scand., 2JL, 1801 (1967). 93. Z. Dische, ,: Methods Carbohyd. Chem., 1, 478 (1962). 94. S.D. Henriksen, Acta path, et microbiol. scandinav., 34, 249 (1954). 95. M. Heidelberger, W.F. Dudman and W. Nimmich, J. Immunol., 104. 1321 (1970). 96. J . Erikseri; ; ^, Bakteriologiske Institutt, Rikshosptalet, Oslo, Norway; Personal communication. 97. M. Stacey and S.A. Barker, Polysaccharides of Micro-organisms. Oxford; Clarendon Press, 1967. 98. M.J. How, J.S. Brimacombe, and M. Stacey, Adv. Carbohydr. Chem., 19, 313 (1964). 99. E.J. Bourne and S. Peat, Adv. Carbohydr. Chem., 5, 145 (1950). 100. M. Heidelberger, Bacteriol. Rev., 3, 49 (1939). PUBLICATIONS Lung Ching L i n and Mo-tai Yang, Anomalous N i t r a t i o n Reactions of Halotropones and Tribromotropolone, J. Chinese Chem. S o c , I I , L3, 178 (1966.) J.J.M.-Rowe, K.B. Gibney, M.T. Yang, G.G.S. Dutton, . Periodic Acid-Dimethyl Sulfoxide Mixtures, a P o t e n t i a l Hazard, J.A.C.S., 90, 1924 (1968) G.G.S. Dutton and M.T. Yang, 4,6-0-(1-Carbox-yethylidene)-D_-mannose as a Str u c t u r a l Unit • i n Capsular Polysaccharide of K l e b s i e l l a K-• type 5, Can. J . Chem., 50, 2382 (1972) G.G.S. Dutton 'and M.T. Yang, 'The Structure of the '< Capsular' Polysaccharide of K l e b s i e l l a • i K-type 5, Can. J.' Chem., 51_, 1826 (1973)' Y.M. Choy, G.G.S. Dutton, A.M. Stephen, and M.T. Yang, P.M.R. Spectroscopic Analysis of B a c t e r i a l Polysaccharides Containing Pyruvic Acid, A n a l y t i c a l L e t t e r s , 5(10), 675 (1972) G.M. Bebault, Y.M. Choy, G.G.S. Dutton, N. Funnell, A.M. Stephen, and M.T. Yang, Proton Magnetic Resonance Spectroscopy of K I e b s i e l l a Capsular Polysaccharides, J . B a c t e r i d . , 113, 1345 (1973) i 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061058/manifest

Comment

Related Items