Chemical and Macromolecular Structure of O-Antigens from Salmonella anatum Strains Carrying Mutants

The conversion of O-antigens by bacteriophage was discovered by Iseki and Sakai (I) and has been investigated from the points of view of virology and serology in a number of laboratories (2-6). The phenomenon involves the change in structure of the cellular polysaccharide O-antigen associated with the presence of a temperature bacteriophage (7, 8). Since the role taken by the virus in modifying the metabolic or biosynthetic activities of the bacterium must give the cell some kind of selective advantage, the relationship may serve as a model for the types of interaction that are assumed to take place in other virus-cell relationships. Dulbecco (9), in fact, suggests that the growth and differentiation of cells following infection with Rous sarcoma virus may be influenced by changes in the cell surface caused by the presence of the virus. Whether or not the situations are parallel, however, the bacteriophage-O-antigen interrelationship probably provides the most accessible system for studying the control of a normal cellular process by a viral genome at the structural and enzymatic level. This being the case, we have undertaken a detailed study of the situation in Salmonella anatum and related strains in which the polysaccharide O-antigen is modified by the presence of the bacteriophages cl5 and e34. Previous studies have shown that the polysaccharide consists of n-galactosyl-n-mannosyl-nrhamnose repeating units linked end-to-end (8). In nonlysogenic cells the galactosyl-mannose linkage is a-l, 6-D, and the galactose unit carries an 0-acetyl residue. In cells carrying phage $5 the sequence and linkage position of the chains are the same but the galactosyl-mannose linkage is p-1 ,6-D, and 0-acetyl residues are absent. In cells carrying both e15 and c34 the ,&n-galactosyl groups are modified by the attachment of an ol-n-glucosyl residue to position 4 of the galactose. The present paper deals with the structural changes produced by three mutants of t15. The following paper (10) provides kinetic and enzymatic evidence that the proposed changes caused by cl5 and its mutants can be observed during the early stages of infection: thus ruling out artifacts based on cellular selection. The two studies taken together then provide a working model for the mechanism of action of @.