1. Walls of Staphylococcus epidermidis I2 contain 30% (w/w) of a glycerol teichoic acid containing phosphate, d-alanine and d-glucose in the molecular proportions 1:0·25:0·50. 2. The teichoic acid was isolated by extraction with trichloroacetic acid and with dilute aqueous NN-dimethylhydrazine at pH7, and was shown to be a (1→3)-linked poly(glycerol phosphate) containing β-d-glucopyranosyl and d-alanyl ester substituents. 3. 2-O-β-d-Glucopyranosylglycerol was isolated and characterized as its crystalline hexa-O-acetate. 4. Unlike that of certain other bacteria, the peptidoglycan component of the wall is not solubilized by NN-dimethylhydrazine. 5. The membrane teichoic acid is also a (1→3)-linked poly(glycerol phosphate) but contains a smaller proportion of glucosyl substituents.
1. The teichoic acid from walls of Staphylococcus lactis I3 was isolated by extraction with trichloroacetic acid and shown to contain glycerol, N-acetylglucosamine, phosphate and d-alanine in the molecular proportions 1:1:2:1. The alanine is attached to the polymer through ester linkages. 2. Hydrolysis with acid gave alanine, glucosamine and glycerol diphosphates. Under mild acid conditions a repeating unit was produced; this consists of glycerol diphosphate joined through a phosphodiester group to N-acetylglucosamine. 3. Hydrolysis with alkali gave glycerol diphosphates, saccharinic acid and two phosphodiesters containing glucosamine whose structures were elucidated; these both contain glucosamine 1-phosphate, and N-acetylglucosamine 1-phosphate was isolated by a degradative procedure. 4. The unusual properties of the teichoic acid are explained by a polymeric structure in which N-acetylglucosamine 1-phosphate is attached through its phosphate to glycerol phosphate. 5. The biosynthetic implications of this structure are discussed.
It has been known for some time that Concanavalin A will agglutinate bacteria whose wall teichoic acids contain a-linked glucosyl substituents [1]. The ability of the lectin to precipitate appropriately substituted teichoic acids has been used in structural [2] and preparative [3] studies and its interaction with whole bacteria or isolated walls has been studied in attempts to gain information on the location of teichoic acids in the cell wall and at the bacterial surface [4,5]. Interpretation of the latter studies has been complicated by the lack of any information on how the amount of Concanavalin A bound is related to the amount of teichoic acid present in the bacteria, or whether variations in the amount of teichoic acid present affect both the amount bound and the affinity of binding. We have recently [6] shown that the amount of teichoic acid present in walls of B. subtilis 168 can be controlled by growing the bacteria under balanced conditions with varied phosphate supply. We now report a study of the interaction of these bacteria with Concanavalin A. Our results show that the association constant of the binding reaction is little affected by variation in teichoic acid content but the amount of Concanavalin A bound increases in proportion to the amount of teichoic acid present although it ap-
The N-acetylgalactosamine in the walls of Bacillus subtilis 168 occurs in two polymers. One of these contains N-acetylgalactosamine, glucose and phosphorus and is attached to the peptidoglycan through an alkali-labile bond; preliminary studies indicate that a repeating unit of this polymer is glucosyl-N-acetylgalactosamine 1-phosphate. N-Acetylgalactosamine is also associated with the peptidoglycan in a component that is not converted into the free sugar or other soluble compounds on treatment of the walls with alkali. The two polymers containing N-acetylgalactosamine are released on autolysis of the walls and can be separated by ion-exchange chromatography. As glucose 6-phosphate is produced by gentle hydrolysis of the wall with acid a third phosphate polymer, poly(glucose 1-phosphate), may occur in this wall. However, as no polymer with this structure could be separated from that containing galactosamine, its existence has not been established unequivocally. The methods described permit the study of variations in N-acetylgalactosamine content with respect to growth conditions.
The pattern of cross-linking in the peptidoglycan of Bacillus megaterium has been studied by the pulsed addition of radiolabeled diaminopimelic acid. The distribution of label in muropeptides, generated by digestion with Chalaropsis muramidase and separated by high-performance liquid chromatography, stabilized after 0.15 of a generation time. The proportion of label in the acceptor and donor positions of isolated muropeptide dimers stabilized over the same period of time. The results have led to the formulation a new model for the assembly of peptidoglycan into the cylindrical wall of B. megaterium by a monomer addition process. Single nascent glycan peptide strands form cross-linkages only with material at the inner surface of the wall. Maturation is a direct consequence of subsequent incorporation of further new glycan peptide strands, and there is no secondary cross-linking process. The initial distribution of muropeptides is constant. It follows that the final pattern of cross-linking in the wall is determined solely by, and can be forecast from, this repetitive pattern of incorporation. In a modified form, this model can also be applied to assembly of cell walls in rod-shaped gram-negative bacteria.