Conditional expression of flagellar motility, curli fimbriae, and biofilms in Shiga toxin- producing Escherichia albertii

<i>Escherichia albertii</i> is an emerging foodborne pathogen. We previously reported that some avian Shiga toxin-producing <i>E. albertii</i> strains exhibited higher or comparable cytotoxicity in Vero-d2EGFP cells with several enterohemorrhagic <i>E. coli</i> (EHEC) outbreak strains. To better understand the environmental persistence of this pathogen, comparative genomics and phenotypic assays were applied to assess adhesion capability, motility, and biofilm formation in <i>E. albertii</i>. Among the 108 adherence-related genes, those involved in biogenesis of curli fimbriae, hemorrhagic <i>E. coli</i> pilus, type 1 fimbriae, and Sfm fimbriae were conserved in <i>E. albertii</i>. All 20 <i>E. albertii</i> strains carried a complete set of primary flagellar genes that were organized into four gene clusters, while five strains possessed genes related to the secondary flagella, also known as lateral flagella. Compared to EHEC strain EDL933, the eight chemotaxis genes located within the primary flagellar gene clusters were deleted in <i>E. albertii</i>. Additional deletion of motility genes <i>flhABCD</i> and <i>motBC</i> was identified in several <i>E. albertii</i> strains. Swimming motility was detected in three strains when grown in LB medium, however, when grown in 5% TSB or in the pond water-supplemented with 10% pigeon droppings, an additional four strains became motile. Although all <i>E. albertii</i> strains carried curli genes, curli fimbriae were detected only in four, eight, and nine strains following 24, 48, and 120 h incubation, respectively. Type 1 fimbriae were undetectable in any of the strains grown at 37°C or 28°C. Strong biofilms were detected in strains that produced curli fimbriae and in a chicken isolate that was curli fimbriae negative but carried genes encoding adhesive fimbriae K88, a signature of enterotoxigenic <i>E. coli</i> strains causing neonatal diarrhea in piglets. In all phenotypic traits examined, no correlation was revealed between the strains isolated from different sources, or between the strains with and without Shiga toxin genes. The phenotypic variations could not be explained solely by the genetic diversity or the difference in adherence genes repertoire, implying complex regulation in expression of various adhesins. Strains that exhibited a high level of cytotoxicity and were also proficient in biofilm production, may have potential to emerge into high-risk pathogens.