Potential O-acetylation of the sialic acid residues of Escherichia coli K1, groups W-135, Y, and C meningococci, and group B Streptococcus capsular polysaccharides modifies their immunogenicity and susceptibility to glycosidases. Despite the biological importance of O-acetylation, no sialic or polysialic acid O-acetyltransferase has been identified in any system. Here we show that the E. coli K1 O-acetyltransferase encoded by neuO is genetically linked to the endo-neuraminidase tail protein gene of a chromosomal accretion element, designated CUS-3, with homology to lambdoid bacteriophage. Molecular epidemiological analysis established concordance between O-acetyltransferase and CUS-3 in a set of E. coli K1 strains. Deleting neuO eliminated enzymatic activity, which was restored by complementation in trans, and confirmed by C-13-NMR analysis of the acetylated product. Analysis of mutants that accumulate intracellular polysialic acid because of export defects (kpsM and kpsS) or an inability to synthesize the sialic acid precursor, N-acetylmannosamine (neuC), indicated that NeuO does not require constant association with its substrate for activity. DNA sequencing and PCR analysis of neuO from strains that had undergone random capsule form variation showed that slip strand DNA mispairing or unequal recombination resulted in gain or loss of (5'-AAGACTC-3')(n) heptanucleoticle repeats (where n approximate to 14-39) located in the neuO 5' region. These repeats code for a previously undescribed structure designated the poly(Psi) motif. The unexpected discovery of the neuO contingency locus (hypervariable gene controlling expression of a surface epitope) in E. coli, and of a potential phage for redistributing variant neuO alleles, provides a robust system for investigating the functions of localized hypermutability in pathogen evolution.
