Enterohemorrhagic Escherichia coli pathogenesis: role of Long polar fimbriae in Peyer’s patches interactions

Enterohemorrhagic Escherichia coli (EHEC) are major food-borne pathogens whose survival and virulence in the human digestive tract remain unclear owing to paucity of relevant models. EHEC interact with the follicle-associated epithelium of Peyer’s patches of the distal ileum and translocate across the intestinal epithelium via M-cells, but the underlying molecular mechanisms are still unknown. Here, we investigated the involvement of Long polar fimbriae (Lpf) in EHEC pathogenesis. Of the 236 strains tested, a significant association was observed between the presence of lpf operons and pathogenicity. In sophisticated in vitro models of the human gastro-intestinal tract, lpf expression was induced during transit through the simulated stomach and small intestine, but not in the colonic compartment. To investigate the involvement of Lpf in EHEC pathogenesis, lpf isogenic mutants and their relative trans-complemented strains were generated. Translocation across M-cells, interactions with murine ileal biopsies containing Peyer’s patches and the number of hemorrhagic lesions were significantly reduced with the lpf mutants compared to the wild-type strain. Complementation of lpf mutants fully restored the wild-type phenotypes. Our results indicate that (i) EHEC might colonize the terminal ileum at the early stages of infection, (ii) Lpf are an important player in the interactions with Peyer’s patches and M-cells, and could contribute to intestinal colonization.


Supplemental Methods
Detection of promoter deletion in fim operon. The 16 bp deletion in the fim promoter resulting in the absence of expression of type I pili was investigated by PCR as previously described by Shaikh et al. 1 . Primers A and B (Supplemental Table 1) produce a 952 bp amplicon if the deletion is absent, but no amplicon in the presence of the deletion. Primers C and B (Supplemental Table 1) produce a 936 bp amplicon if the deletion is present in the fim operon.
Preparation of fimbrial crude extracts. EHEC O157:H7 EDL933 and AIEC LF82 (as a positive control) were grown for 3 h in DMEM (PAA) with 2% of bile salts. After centrifugation (10,000 g, 10 min, 4°C), bacterial pellets were washed in 5 mM Tris 5% NaCl, then centrifuged using same conditions. Pellets were suspended in 5 mM Tris and bacterial surface proteins were sheared off with a homogenizer set at low speed during 2 min (on ice, 3 times). Bacterial cells and debris were sedimented (10,000 g, 10 min, 4°C), while the supernatants constituting the crude extract of bacterial surface proteins were centrifuged (27,000 g, 30 min, 4°C). Pili extracts were obtained as a pellet after ultracentrifugation (270,000 g, 2 h, 4°C). After solubilization in 5 mM Tris buffer (overnight, 4°C), fimbriae were precipitated with 0.1 M MgCl2 (3 h, 4°C). Suspensions were centrifuged (27,000 g, 40 min, 4°C) and supernatants were discarded. Fimbriae were solubilized in PBS then subjected to depolymerization by trifluoroacetic acid (TFA) hydrolysis. Samples were acidified to pH 2 with final 0.16% TFA, heated at 100°C for 5 min, cooled and neutralized with NaOH before Immunoblotting. Fimbrial crude extracts were incubated with 4X Laemmli buffer containing 10% mercaptoethanol (3 min, 100°C), then resolved on a 12% SDS-PAGE followed by electroblotting onto nitrocellulose membrane (GE Healthcare Life Sciences) using a Trans-Blot® Turbo™ Blotting System (Bio-Rad). One percent of bovine serum albumin (BSA) in TBS-T (10 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.05% Tween 20) was used for diluting antibodies and blocking non-specific binding to nitrocellulose membrane.
Type I pili were detected by Western blotting using the primary anti-pili polyclonal antibody αF1 from rabbit (1:5000) and the secondary monoclonal HRP-conjugated anti-rabbit IgG from goat (1:5000). The blot was developed using enhanced chemiluminescence reagent (Bio-Rad).
Yeast agglutination assay. EHEC O157:H7 strain EDL933 and AIEC strain LF82 (as a positive control) were cultured in static LB broth at 37°C overnight. Bacterial cells were collected by centrifugation (5,000 g, 10 min, room temperature) and pellets were resuspended in 100 μl of 1X PBS. Subsequently, 50 μl of a 2% (vol/vol) suspension of Saccharomyces cerevisiae (Sigma) and an equal amount of undiluted or serial dilutions (from ½ to 1/20) of bacterial cells were mixed in 96-well plate. After 1 h incubation, visible agglutination was determined macroscopically.

Construction of isogenic mutants and transcomplementation. For double mutant
lpfAOI-141/lpfAOI-154 generation, kanamycin cassette was excised using pcp20 plasmid after the first gene deletion 2 . Primers used are listed in Supplemental Table 1. For transcomplementation assays, PCR products containing the entire major fimbrial subunit lpfAOI-141 or lpfAOI-154 were cloned into the pBAD24 and pBAD33 vectors, respectively, and introduced in their respective mutant strain as previously described 3 .
Shiga toxin production. EDL933-∆stx2 was used as a negative control. The Stx titre was expressed as the reciprocal of the highest filtrate dilution that cause 50% cell detachment after 24 h incubation, as judged by the dye density and by microscopic observation. Purified Stx2 (Toxin Technology) was used to estimate the range of toxin production. Each experiment was performed three times. percentages of total strains (in bracket). Prevalence of each lpf operon in the seropathotype A was compared to that in other seropathotypes and statistical differences were indicated by α (p < 0.01) or β (p < 0.001). Prevalence of lpf genes was determined and expressed as percentages of the total number of strains in each seropathotype.    (22)  Results are expressed as total number of lpf-positive strains and as percentages of total strains (in bracket). Prevalence of each lpf operon in the seropathotype A was compared to that in other seropathotypes and statistical differences were indicated by α (p < 0.01) or β (p < 0.001). Prevalence of lpf genes was determined and expressed as percentages of the total number of strains in each seropathotype.