The group-specific antigen was isolated from a type Ia group B streptococcal strain and is a complex polysaccharide composed of alpha-L-rhamnopyranosyl, alpha-D-galactopyranosyl, 2-acetamido-2-deoxy-beta-D-glucopyranosyl, D-glucitol, and phosphate residues. The complexity of the group B polysaccharide antigen is evident from the fact that when depolymerized by basic hydrolysis it yielded three structurally related, but nevertheless significantly different, oligosaccharides. These oligosaccharides were obtained in different molar quantities as their monophosphate esters. This evidence strongly suggests that they are linked by phosphodiester bonds in the original group B antigen. If these oligosaccharides are in fact randomly situated throughout the linear polysaccharide, then this type of heterogeneous repeating unit is unusual for a polysaccharide of bacterial origin. However, this structural arrangement of the oligosaccharides has yet to be unambiguously established because the alternate explanation of there being three different polysaccharides in the group B antigen cannot be discounted in the evidence presented here. The oligosaccharides were enzymatically dephosphorylated, and the structures of two of the three oligosaccharides are (formula: see text) Despite their structural differences, the two oligosaccharides are related by the smaller being an integral part of the larger. In the structural analysis of the group B antigen, methylation analysis, periodate oxidation, nuclear magnetic resonance (NMR) spectroscopy, mass spectrometry, fast atom bombardment mass spectrometry, and various specific chemical and enzymatic degradations were the principal methods used. Of particular interest was the use of an alpha-rhamnosidase to selectively degrade the larger oligosaccharide. This facilitated the assignment of signals in its 1H and 13C NMR spectra.

The structures of the 0-antigens of all known serotypes and subserotypes of Shigella jlexneri have been reinvestigated. The results support the assumption that these antigens are composed of a basic tetrasaccharide repeating unit (l), to which a-D-glucopyranosyl and/or 0-acetyl groups are attached at different positions.The immunological determinants responsible for 0-factors I, 11, IV, V and 7,8 contain a-D-glucopyranosyl groups, the locations of which have been determined. 0-Factor 6 is due to 0-acetyl groups, linked to 0-2 of the 3-substituted a-L-rhamnopyranosyl residue in unit 1 and 0-factor 111 seems to be due to the same groups. The chemical natures of the determinants responsible for 0-factors 4 and 3,4 are still obscure. The structural studies indicate that the immunological classification of Sh. ,flexneri serotypes and subserotypes, as regards these 0-factors, may need revision.Sh.Jlexneri has been divided into a number of serotypes and subserotypes, which should differ in the structure of their 0-antigens [l]. We recently demonstrated that the structures proposed for such 0-antigens should be revised [2,3], and proposed that all Sh. JZexneri 0-antigens contain the same basic repeating unit [4,5]. Variation should be provided by adding a-D-glucopyranosyl groups and 0-acetyl groups to different positions of this repeating unit.The structure of this proposed basic repeating unit (1) was determined for the Sh..flexneri variant Y 0-antigen, which does not contain a-D-glucopyranosyl groups [4]. The three L-rhamnopyranosyl(1) residues in unit 1 will, in the following, be called Rha I, Rha I1 and Rha 111, as indicated in the formula.Studies on the 0-antigens from types 5a, 5b and variant X supported the assumption of a basic repeating unit [5]. The locations of the a-D-glucopyranosyl groups in these antigens were also determined. We now report structural studies of the 0-antigens from all the other known serotypes and subserotypes of N-acetylglucosamine; Rha, rhamnose.Abbreviations NMR, nuclear magnetic resonance; GlcNAc, Sh.JZexyleri, with the exception of type 6. The structure of the 0-antigen from this type differs considerably from those of the other 0-antigens [6] and it is questionable if it should be classified as a Sh. ,fZexneri. MATERIALS AND METHODSGeneral methods are described elsewhere [4]. The I3C NMR spectra were obtained on a JEOL FX-100 spectrometer operating at 25.0 MHz with complete proton decoupling, using 100000 accumulations of free induction decay. Pulse-width of 7 ps (45" flipangle), acquisition time of 0.3996 s and pulse repetition time of 0.5 s were the parameters used in the experiments. A 10-mm NMR tube was used containing 100 mg polysaccharide/ml 'Hz0. External tetramethylsilane was used as a reference. Optical rotations of the polysaccharides were recorded on a Perkin-Elmer 241 instrument with x 5 mg/ml solution. Preparation of Lipopolysaccharides and PolysaccharidesShigella flexneri standard strains of serotype la, 2a, 2b, 3a, 4a and 4b (Dysentery Reference Laboratory, London), s...

The rate of coupling of oligosaccharides having aldose end groups to protein by reductive amination was significantly increased by changing the temperature and pH of the reaction, and even more significantly by the addition of borate ions. Under optimized conditions half of the lysine residues of bovine serum albumin could be derivatized by lactose in 7 h and their complete derivatization was achieved in approximately 24 h. All attempts to carry out similar reductive amination procedures using oligosaccharides having ketose (D-fructose, 3-deoxy-D-manno-octulosonic acid (KDO), and sialic acid) failed owing to the slowness of the reaction. Model studies on the coupling of D-fructose and KDO to glycine indicate that any coupling procedure based on reductive amination of ketose residues would of necessity require the prior introduction of a small functionalized spacer molecule.

No abstract