NAD auxotrophy of wild type uropathogenic Escherichia coli UTI89 by a single amino acid change in NadB

Background: Nicotinamide adenine dinucleotide (NAD) is indispensable for all living organisms. The nadA and nadB genes are required for the synthesis of quinolinate, an intermediate of NAD synthesis. Shigella flexneri strains typically carry mutations inactivating nadA and/or nadB. In S. flexneri, nadA and nadB were defined as anti-virulence loci, because expression of active nadA and nadB led to a severe reduction of virulence (Prunier et al., 2007a; Prunier et al., 2007b). Many E. coli O18:K1 isolates are also NAD auxotrophic (Achtman et al., 1983). In the present study, the molecular basis of the NAD auxotrophy of the uropathogenic E. coli O18:K1:H7 model strain UTI89 (Mulvey et al., 2001), isolated from an acute cystitis patient, was determined. In addition, the influence of this NAD auxotrophy on the pathogenicity of UTI89 was investigated in the murine ascending infection model.

Methods: The auxotrophy of UTI89 was identified by growing the bacteria on Minimal A agar plates supplement with vitamins, amino acids and nucleic acid bases. The complementation assay of UTI89 was performed by testing the growth of UTI89 on Minimal A medium, after introduction of the corresponding genes of the prototrophic E. coli K-12 strain MG1655. Point mutations of the gene were obtained by overlap PCR. Pathogenicity tests in the mouse ascending infection model were performed by mixed infection, followed by plating of bladder homogenates on the appropriate selective media.

Results: E. coli UTI89 is an NAD auxotroph. Addition of the metabolite quinolinate allows its growth on Minimal A medium, suggesting a mutation in the nadB gene. A prototrophic derivative of UTI89 was obtained by P1vir-mediated transduction of the wild type gene from E. coli K-12. Transposon mutagenesis and sequencing of the nadB and nadA genes of this transductant showed that the nadB gene of E. coli K-12 is present and is necessary for de novo synthesis of NAD. The auxotrophy of UTI89 was also complemented by the introduction of a plasmid carrying the wild type nadB gene of E. coli K-12. There are four amino acid differences between the NadB proteins of E. coli strains K-12 and UTI89. Each of these was introduced in the nadB gene of the complementing plasmid. Only the plasmid harboring the Ala28Val mutation did not allow growth of UTI89 on minimal medium. In the murine ascending infection model, UTI89 and a spontaneous prototrophic nadB+ revertant were equally pathogenic.

Conclusions: (i) The pathogenicity of E. coli UTI89 in the mouse model is not influenced by the NAD auxotrophy, which is different from Shigella, even though E. coli and Shigella are close relatives. (ii) Mutation of the 28th amino acid of NadB, from alanine to valine, causes the NAD auxotrophy of wild type E. coli UTI89.