Abstract
6 were partiallypurified from a K99 - variant of E. coli B41. One was identical to the anionic antigen, adhered to calf brush borders and did not have a regular fimbrial structure. The second was a polymer of two polypeptide subunits with molecular weights of 49,500 and 48,000; it was antigenically distinct from K99 and the anionic antigen, did not adhere to brush borders or isolated villi and had a fimbrial structure. An immunoperoxidase test indicated that the second haemagglutinin was produced in the ileum of a gnotobiotic calf infected with E, coli B41. Agglutination of sheep erythrocytes by the second haemagglutinin was inhibited by any of ten sugars. The receptor was present in glycolipid extracts of sheep erythrocytes. K99 was the only detectable adhesin on E. coli B117. This organism produced three polypeptides of similar molecular weight to the anionic antigen which did not adhere to brush borders or isolated villi. Results of an immunoradiometric assay suggested that the K99 of different strains was not identical. Calves fed colostrum which contained natural IgM antibodies that were opsonic were not protected against infection by K99 E. coli. However, the presence of a normal gut flora together with colostrum provided some protection. A blocking immunoradiometric assay which detected antibody to K99 in calves previously infected with K99 E. coli failed to detect antibody in sera from 248 calves collected between 1971 and 1975 on a progeny testing station. During the winters of 1981-1983 a survey of calves with diarrhoea in South England revealed that K99 E. coli caused approximately 4% of the outbreaks examined. An atypical E. coli isolated from calves with dysentery caused dysentery in experimentally infected gnotobiotic calves by colonising the mucosa of the colon and by production of petechial haemorrhages.