Search

Background Avian pathogenic Escherichia coli (APEC) and uropathogenic E. coli (UPEC) are the two main subsets of extraintestinal pathogenic E. coli (ExPEC). Both types have multiple iron acquisition systems, including heme and siderophores. Although iron transport systems involved in the pathogenesis of APEC or UPEC have been documented individually in corresponding animal models, the contribution of these systems during simultaneous APEC and UPEC infection is not well described. To determine the contribution of each individual iron acquisition system to the virulence of APEC and UPEC, isogenic mutants affecting iron uptake in APEC E058 and UPEC U17 were constructed and compared in a chicken challenge model. Results Salmochelin-defective mutants E058ΔiroD and U17ΔiroD showed significantly decreased pathogenicity compared to the wild-type strains. Aerobactin defective mutants E058ΔiucD and U17ΔiucD demonstrated reduced colonization in several internal organs, whereas the heme defective mutants E058ΔchuT and U17ΔchuT colonized internal organs to the same extent as their wild-type strains. The triple mutant ΔchuTΔiroDΔiucD in both E058 and U17 showed decreased pathogenicity compared to each of the single mutants. The histopathological lesions in visceral organs of birds challenged with the wild-type strains were more severe than those from birds challenged with ΔiroD, ΔiucD or the triple mutants. Conversely, chickens inoculated with the ΔchuT mutants had lesions comparable to those in chickens inoculated with the wild-type strains. However, no significant differences were observed between the mutants and the wild-type strains in resistance to serum, cellular invasion and intracellular survival in HD-11, and growth in iron-rich or iron-restricted medium. Conclusions Results indicated that APEC and UPEC utilize similar iron acquisition mechanisms in chickens. Both salmochelin and aerobactin systems appeared to be important in APEC and UPEC virulence, while salmochelin contributed more to the virulence. Heme bounded by ChuT in the periplasm appeared to be redundant in this model, indicating that other periplasmic binding proteins likely contributed to the observed no phenotype for the heme uptake mutant. No differences were observed between the mutants and their wild-type parents in other phenotypic traits, suggesting that other virulence mechanisms compensate for the effect of the mutations.