Genomics, biofilm formation and infection of bladder epithelial cells in potentially uropathogenic Escherichia coli (UPEC) from animal sources and human urinary tract infections (UTIs) further support food-borne transmission

Abstract

Escherichia coli is the main cause of urinary tract infections (UTI). While genomic comparison of specific clones recovered from animals, and human extraintestinal infections show high identity, studies demonstrating the uropathogenicity are lacking. In this study, comparative genomics combined with bladder-cell and biofilm formation assays, were performed for 31 E. coli of different origins: 7 from meat (poultry, beef, and pork); 2 from avian-farm environment; 12 from human uncomplicated UTI, uUTI; and 10 from human complicated UTI, cUTI. These isolates were selected based on their genetic uropathogenic (UPEC) status and phylogenetic background. In silico analysis revealed similar virulence-gene profiles, with flagella, type 1 and curli fimbriae, outer-membrane proteins (agn43, ompT, iha), and iron-uptake (iutA, entA, and fyuA) associated-traits as the most prevalent (>65%). In bladder-cell assays, moderate to strong values of association (83%, 60%, 77.8%) and invasion (0%, 70%, 55.5%) were exhibited by uUTI, cUTI, and animal-derived isolates, respectively. Of interest, uUTI isolates exhibited a significantly lower invasive capacity than cUTI isolates (p < 0.05). All isolates but one produced measurable biofilm. Notably, 1 turkey meat isolate O11:H6-F-ST457, and 2 cUTI isolates of the pandemic lineages O83:H42-F-ST1485-CC648 and O25b:H4-B2-ST131, showed strong association, invasion and biofilm formation. These isolates showed common carriage of type 1 fimbriae and csg operons, toxins (hlyF, tsh), iron uptake systems (iutA, entA, iroN), colicins, protectins (cvaC, iss, kpsM, traT), ompT, and malX. In summary, the similar in vitro behaviour found here for certain E. coli clones of animal origin would further reinforce the role of food-producing animals as a potential source of UPEC. Bladder-cell infection assays, combined with genomics, might be an alternative to in vivo virulence models to assess uropathogenicity.