A cell wall-associated polysaccharide antigen was isolated from Streptococcus mutans GS-5 and appeared to determine serotype c specificity. Ouchterlony double-diffusion analysis of crude formamide extracts derived from purified cell walls of two serotype c strains (GS-5 and JC-2) showed complete identity when reacted with anti-GS-5 sera. Immunoelectrophoresis of this extract demonstrated the typical mobility for this serotype as described by others. Column chromatography on BioGel P-100 of the crude formamide extracts derived from GS-5 walls resulted in a single antigenic peak being resolved. This material, when loaded onto a diethylaminoethylcellulose column and eluted with a linear gradient of ammonium carbonate (0.0 to 0.2 M), was resolved further into two serologically reactive peaks (I and II). Only two constituents, rhamnose and glucose, were detected in the purified column fractions. Peak I had a rhamnoseto-glucose molar ratio of 0.9:1.0, and peak II, the major resolvable fraction, had a molar ratio of 1.7:1.0. The peak II ratio was very similar to that found in the formamide extract residue pellet (1.6:1.0). Ouchterlony analysis of the crude formamide extract and the purified fractions revealed only partial identity between peaks I and II but complete identity between peak II and the crude extract. Likewise, immunoelectrophoresis showed no differences in mobility of peak II and the crude extract, whereas peak I moved towards the cathode. Possible structural relationships between the two antigenic fractions are discussed below. Hapten inhibition studies suggested that an a-glucosyl group is at the immunodeterminant site of the antigen.

Quantitative analyses of cell walls from Streptococcus mutans Ingbritt grown under carbohydrate limitation in the chemostat showed that growth conditions had no statistically significant effect on the composition of polysaccharide, peptidoglycan, or the proportion of polysaccharide in the cell wall. Lysis of cell wall preparations with a muramidase supported this conclusion and further indicated that there was little difference in their overall structure. In contrast, there was a consistent difference between the rates of lysis by this enzyme of organisms grown in 0.2% glucose and 0.5% glucose. Extremes of pH or dilution rate essentially did not influence the immunogenicity of type c antigen in whole organisms irrespective of whether the carbohydrate source was glucose or sucrose. However, differences were found in the immunogenicity of lipoteichoic acid under similar circumstances. The results indicated there was an inherent phenotypic stability in the cell walls of S. mutans Ingbritt despite changes in pH, generation time, and carbohydrate source, and that any changes that did occur were probably due to associated cell-surface components. The serological identification of strains of Streptococcus mutans has received considerable attention because of their relation to dental caries (2-4, 21, 25, 29). Most studies have assumed that the serological properties of organisms growing in plaque with a mean generation time of 12 to 24 h (8, 11) would be similar to the properties of organisms growing in batch culture under laboratory conditions. However, the surface properties of bacteria can be profoundly influenced by the growth conditions (8, 9), specific examples being the effect of phosphate limitation on teichoic acid formation (9), the effect of generation time on the toxicity of gram-negative bacteria (23, 24), and the immunogenicity of surface components of lactobacilli (17), as well as the effect of generation time, pH, and carbohydrate source on the production of lipoteichoic acid (LTA) by S. mutans strains BHT (14) and Ingbritt (13). In contrast, it was recently reported that the production of exopolysaccharide by Azotobacter vinelandii is unaffected by most environmental conditions (15). Serological type c strains of S. mutans generally predominate in plaque (2, 25, 29). The type c antigen, isolated from both strain Ingbritt (22)

Lipoteichoic acid (LTA) is a component of nearly all gram-positive membranes and recently has been found to be excreted into growth media by certain lactic acid bacteria. Cell-free extracts of LTA are usually contaminated with proteins, polysaccharides, and nucleic acids, thus causing problems to investigators studying the true biological function(s) of LTA. This report describes the preparation of purified extracellular LTA of Streptococcus mutans BHT and intracellular LTA of S. mutans AHT by three techniques: gel filtration, hydrophobic interaction chromatography, and adsorption to phospholipid vesicles. Gel filtration, the most commonly employed method for LTA purification, was found to remove nucleic acids, teichoic acids, and much polysaccharide while greatly concentrating LTA. But gross amounts of antigenic carbohydrate and protein remained associated with the LTA preparation. Hydrophobic interaction chromatography employing octyl Sepharose-4B allowed the separation of protein but not polysaccharide from partially purified BHT LTA preparations. By means of a new technique described in this paper, synthetic membranes (vesicles) were found to effectively separate all contaminants from the intracellular (AHT) and extracellular (BHT) LTA of S. mutans. This rapid method, on a comparative basis, proved to be the most effective approach for the purification of LTA from two widely differing sources.

The serotype-specific polysaccharide of Streptococcus mutans AHT (serotype a) was shown to be loosely associated with the cell surface of this organism. The antigen was extracted from whole cells by boiling in sodium acetate buffer, pH 4.0, for 10 min. The purified product was found to be a diheteroglycan of galactose and glucose (3.6:1, molar ratio). The antigen possessed serological characteristics similar to the a antigen previously extracted from purified cell walls with hot formamide. Its physicochemical structure was identical to the previously studied wall antigen. Electron micrographs, developed after immunocytological labeling of this antigen on whole cells, revealed it to compose a dense microcapsule surrounding the microbe. Analyses of spent culture fluids indicated that the antigen was released during exponential growth at a rate directly proportional to the increase in culture biomass. It is concluded that the serotype-specific antigen may be a prime immunogen due to its surface localization at both capsule and wall sites.