In this study, we used mouse ileal loops to investigate the interaction of enterohemorrhagic Escherichia coli (EHEC) O157:H7 with the mouse intestinal mucosa. With a dose of 109 and 3 h incubation, EHEC O157 was detected in the lumen and to a lesser extent associated with the epithelium. Typical attaching and effacing (A/E) lesions were seen, albeit infrequently. While the effector protein Tir was essential for A/E lesion formation, the bacterial type III secretion system adaptor protein TccP was dispensable. These results suggest that A/E lesions on mouse intestinal mucosa can be formed independently of robust actin polymerization.
Enterohemorrhagic Escherichia coli (EHEC) colonize the intestinal mucosa and cause attaching and effacing (A/E) lesions on epithelial cells. These lesions are characterized by localized effacement of microvilli and intimate adherence of bacteria to the apical plasma membrane. Actin-rich pedestal-like structures are frequently associated with A/E lesions; however, they may not be essential for colonization (Deng et al., 2003; Ritchie et al., 2008; Bai et al., 2008; Frankel & Phillips, 2008). Ruminants are an important reservoir of EHEC, particularly serogroup O157:H7 (Gansheroff & O'Brien, 2000), although other serogroups including O26, O103, and O111 have also been isolated (Bettelheim, 2000). In humans, EHEC infections can cause disease of variable severity, including mild diarrhoea, haemorrhagic colitis and in a subset of patients a sequela called the haemolytic uraemic syndrome (HUS) that has a mortality rate of 5% and is characterized by thrombocytopaenia, haemolytic anaemia, and kidney lesions (Paton & Paton, 1998). Shiga toxins, produced by EHEC, are considered to be responsible for both haemorrhagic colitis and HUS (Griffin & Tauxe, 1991).
The genes required for the formation of A/E lesions by EHEC O157:H7 in vitro are carried on the locus of enterocyte effacement (LEE) (McDaniel et al., 1995) and the prophage CP-933U/Sp14 (Campellone et al., 2004; Garmendia et al., 2004). The LEE encodes transcriptional regulators, the adhesin intimin, a type III secretion system (responsible for effector protein translocation), chaperones, translocators and six effector proteins including translocated intimin receptor (Tir) (Kenny et al., 1997). Prophage CP-933U/Sp14 encodes the effector protein TccP (Garmendia et al., 2004) (also known as EspFU) (Campellone et al., 2004).
Intestinal colonization of mice by EHEC following oral gavage has been reported in the literature, although it is only made possible by reducing or eliminating the normal intestinal microbiota by streptomycin treatment or using gnotobiotic mice, respectively (Wadolkowski et al., 1990; Fujii et al., 1994; Isogai et al., 1998). We recently reported that, although being able to persist in C57BL/6J mice, EHEC O157:H7 bacteria are rapidly eliminated from C3H/Hej mice and show the same colonization dynamics as a commensal E. coli strain (Mundy et al., 2006).
The aim of this study was to characterize the interaction of EHEC O157:H7 with the mouse intestinal mucosa using an ileal loop model. In particular, we investigated the role of the effector proteins Tir and TccP in colonization of the mouse mucosa by EHEC O157:H7.
Material and methods
The strains used in this study included the nalidixic acid-resistant (Nalr) variant of the wild-type O157:H7 strain 85–170 (Stevens et al., 2004), and its isogenic tir (85–170Δtir) (Stevens et al., 2004) and tccP (85–170ΔtccP) (Vlisidou et al., 2006) deleted mutants. All strains were grown overnight in Luria–Bertani broth (LB, Oxoid), centrifuged, and the pellet washed and resuspended in fresh, sterile Hank's buffered salt solution without phenol red (HBSS, Sigma). Nalidixic acid (100 μg mL−1) or kanamycin (50 μg mL−1, used to grow 85–170ΔtccP) was used when appropriate.
Mouse loop model
A total of five 8-week-old female SLC:ICR mice were used in three different studies. Feed was withdrawn for 8 h before the studies. Mice were placed in an induction chamber (small anaesthetic induction chamber; Vet TECH Solutions Ltd, Cheshire, UK) for anaesthetic induction with isoflurane in oxygen. Anaesthesia was maintained with isoflurane in oxygen delivered via coaxial face mask (rodent mask delivery system, size 1; Vet TECH solutions Ltd, Cheshire, UK). Analgesia was provided with 5 mg kg butorphanol (Torbugesic; Fort Dodge Animal Health, Southampton, UK) administered subcutaneously. Intraoperative fluid therapy consisted of 1 mL per mouse of warmed Hartmann's solution (Aqupharm number 11; Animal Care Ltd, York, UK) injected subcutaneously. Rectal temperature was monitored continually throughout the procedure (TES 1319 K-Type thermometer with thermocouple probe TP-K02; TES, Ontario, Canada), and warming (using a hot water bottle), or cooling (with a small hand-held rotary fan) instigated to maintain near normothermia (37.4 °C). Eyes were protected with a bland ophthalmic ointment. A midline laparotomy was performed to exteriorize the distal ileum. Once the ileum was isolated four 2 cm loops were made incorporating a Peyer's patch, each with spacer loops to avoid any cross contamination from one loop to another. The loops were inoculated with 100 μL (1 × 109 CFU) of test bacteria and reinstated into the abdominal cavity. The vital signs of the mouse were continuously monitored (temperature, colour and respiration). Following a 3 h incubation period the loops were harvested under terminal anaesthesia and the mouse euthanized. The tissues were either immediately fixed for histopathology or electron microscopy, or were embedded in optimal cutting temperature (OCT) medium (Raymond A Lamb Limited, UK) and snap frozen using ice-cold isopentane (VWR, UK). All procedures were conducted under the jurisdiction of HO license 70/6103.
Tissues were collected, rinsed gently in phosphate-buffered saline (PBS), and fixed in 10% buffered formalin for microscopic examination, as previously described (Girard et al., 2005).
Indirect immunofluorescence assay
An indirect immunofluorescence assay (IFA) derived from previously described methodology (Girard et al., 2005, 2007) was used for the detection of O157 bacteria on formalin-fixed, paraffin-embedded (FFPE) sections or on cryosections (intimin-γ, Tir, TccP) fixed in 3% paraformaldehyde in PBS. Appropriate sections were immunostained using the following antibodies: goat anti-O157:H7 (Fitzgerald Industries International Inc., Concorde, MA), rabbit anti-intimin-γ (Girard et al., 2007), rabbit anti-Tir EHEC (Batchelor et al., 2004), and rabbit anti-TccP (Ogura et al., 2006). Cyanine 2 (Cy2)-conjugated donkey anti-rabbit, or tetramethyl rhodamine iso-thiocyanate (TRITC)-conjugated donkey anti-goat (Jackson ImmunoResearch Europe Ltd, Soham, Cambridgeshire, UK) secondary antibodies were used in respect of the primary antibodies. Phalloidin-Alexa Fluor 633 (Invitrogen, UK) was used to stain F-actin, while DNA of both bacteria and epithelial cells was counterstained with Hoechst 33342 (Sigma-Aldrich Co., UK). Sections were examined with an Axio Imager M1 microscope (Carl Zeiss MicroImaging GmbH, Germany), images were acquired using an AxioCam MRm monochrome camera, and computer-processed using AxioVision (Carl Zeiss MicroImaging GmbH), Adobe Photoshop 5.0 and Adobe illustrator 8.0 software (Adobe Systems Incorporated, CA).
Additional tissues were processed for electron microscopy. Explants were fixed overnight at 4 °C in 2.5% glutaraldehyde, and then rinsed in cacodylate buffer (0.1 M sodium cacodylate, pH 7.3) for 1.5 h with regular changes. Explants were further postfixed in aqueous 1% osmium tetroxide (OsO4), rinsed in water, and then dehydrated in a graded ethanol series.
For scanning electron microscopy (SEM), samples were critical-point dried, mounted on aluminium stubs, sputter-coated with gold–palladium and examined without knowledge of the strain used, at an accelerating voltage of 25 kV using a JEOL JSM-5300 scanning electron microscope (JEOL (UK) Ltd, Herts, UK).
For transmission electron microscopy (TEM), methods were as described previously (La Ragione et al., 2006). Ultrathin sections at 70–90 nm thickness were prepared on copper grids and stained with uranyl acetate and lead citrate. Sections were observed using a Philips CM10 transmission electron microscope (Philips, UK).
Analysis of the interaction of EHEC O157:H7 with the mouse ileal epithelium by immunofluorescence
Ileal epithelium and villous structure were well preserved when loops were inoculated with sterile HBSS for 3 h (data not shown). When inoculated with either the wild-type EHEC (strain 85–170) or the tccP mutant, shorter villi and some desquamated epithelial cells were frequently observed (data not shown). Bacteria in the intestinal lumen or associated with the ileal epithelium were frequently seen following incubation with both the wild-type and the tccP mutant (data not shown). These bacteria were confirmed as O157 serotype by immunostaining (Fig. 1, IFA panels). O157-positive bacteria were mostly found in the intestinal lumen in loops inoculated with a tir mutant (Fig. 1, Δtir IFA panel). No bacteria were observed in HBSS-inoculated loops (Fig. 1, HBSS panel).
Analysis of the interaction of EHEC O157:H7 with the mouse ileal epithelium by electron microscopy
We investigated the interaction of EHEC 85–170 with the mouse intestinal mucosa using SEM and TEM. Only a few bacteria were found to be associated with the epithelium by TEM for both the wild type and the tccP mutant (Fig. 1, TEM panels). Typical A/E lesions, characterized by a localized effacement of the epithelial brush border microvilli and intimate bacterial attachment associated with pedestal-like structures were observed, albeit infrequently, for both the wild type and the tccP mutant (Fig. 1, SEM and TEM panels). No A/E lesions were observed in loops inoculated with the tir mutant (Fig. 1, Δtir panel). Neither changes in the epithelium morphology nor A/E lesions were observed in the control loops inoculated with HBSS (Fig. 1, HBSS panel).
Localization of bacterial proteins at the site of intimately adherent bacteria
Foci of intimate bacterial adherence were further characterized by immunostaining for various virulence factors known to be involved in EHEC pathogenesis, for example the outer membrane protein intimin, the translocated intimin receptor Tir, and the adaptor protein TccP, using cryosections. Although A/E lesions were found only infrequently, intimin-γ and Tir were found on the bacterial surface and at the apical pole of the infected enterocytes, respectively, for both the wild type and the tccP mutant (Fig. 2a and b WT, Fig. 3a and bΔtccP). TccP was detected underneath adherent bacteria in loops inoculated with the wild-type strain (Fig. 2c); no staining for TccP was observed at the site of intimately adherent bacteria in loops inoculated with the tccP mutant (Fig. 3c). The rim of F-actin, which was found to be continuous and regularly distributed at the luminal surface of the enterocyte layer on HBSS-inoculated loops (Figs 2d and 3d), appeared to be disrupted, at least locally, in the presence of intimately adherent bacteria (Figs 2 and 3). Neither Tir nor TccP staining was observed on cryosections derived from loops inoculated with the tir mutant (data not shown).
In this study, we report the interaction of the EHEC O157:H7 strain 85–170 with mouse intestinal epithelium using the ileal loop model. Mice have been used as a model for EHEC in many previous studies (Lindgren et al., 1993, 1994; Isogai et al., 1998, 2000, 2001; Robinson et al., 2006; Sheng et al., 2006), although the main studied features were bacterial shedding and the systemic effects of the Stx toxins.
Bovine and ovine intestinal loops have been used in the past to study EHEC pathogenesis (Stevens et al., 2002; Wales et al., 2002; Vlisidou et al., 2004, 2006). The mouse loop model used in this study revealed that EHEC were mostly observed loosely associated with the ileal villi and triggered mild remodelling of the epithelium. Infrequent, typical, A/E lesions were detected in this mouse loop model, suggesting that EHEC has the potential to induce A/E lesions in mouse. Such lesions have not been identified when C57BL/6J mice have been inoculated by oral gavage with EPEC (Savkovic et al., 2005); at this stage we cannot rule out the possibility the difference is due to SLC:ICR mice being more susceptible. The fact that bacteria are trapped when inoculated into the mouse loops may explain why some A/E lesions are observed in this model. The inability to induce A/E lesions efficiently might explain why EHEC cannot expand and establish a long term infection in mice after oral gavage but instead exhibit infection dynamics similar to that of a commensal E. coli (Mundy et al., 2006). This suggests that intimin-Tir interaction is the dominant mechanism of EHEC gut colonization and that other potential adhesins, for example F9 fimbriae (Low et al., 2006) and long polar fimbriae (LPF) (Jordan et al., 2004) may have an accessory function following A/E lesion formation. Interestingly, previous ovine gut loop studies have also demonstrated that O157-specific lesions are often small and sparse in susceptible hosts and that A/E lesions may vary between specific loops and EHEC strains (Wales et al., 2002).
Immunofluorescent staining revealed that intimin-γ was found on the surface of intimately adherent O157-positive bacteria. Tir was translocated and concentrated at the apical surface of the infected enterocytes. We found that tccP is not required for A/E lesion formation on the mouse ileal mucosa, although the protein was translocated and found underneath adherent wild-type O157-positive bacteria. This result is in agreement with the reports showing that EHEC O157 can induce A/E lesions in a tccP-independent mechanism in bovine ileal loops (Vlisidou et al., 2006) and on human in vitro organ cultures (Garmendia et al., 2004). Although we cannot make any definitive conclusions in respect to actin polymerization at the site of EHEC O157 adhesion to the mouse gut, the integrity of the apical actin rim is clearly disrupted in loops with the wild type and EHEC O157 tccP mutant.
Taken together, our results demonstrate that after inoculation of mouse ileal loops with a high dose and 3 h incubation EHEC O157:H7 has the potential to induce A/E lesions, albeit inefficiently. Nonetheless, we show that while Tir plays a vital role, TccP appears to be dispensable. As such, our results emphasize the fact that the ability to trigger robust actin polymerization in vitro is distinct from formation of A/E lesions on the mouse mucosal surfaces, and raises the question of the role of actin polymerization at the site of bacterial adhesion in pathogenesis.
This work was supported by the Defra grant OZO713. F.G. is supported by a Postdoctoral Fellowship from the Natural Sciences and Engineering Research Council (NSERC) of Canada, and by a Canada–United Kingdom Millennium Research Award, awarded by the NSERC and the Royal Society of London, UK.