Pathogenesis of Escherichia coli O157:H7 strain 86-24 following oral infection of BALB/c mice with an intact commensal flora

Abstract

Escherichia coli O157:H7 is a food-borne pathogen that can cause hemorrhagic colitis and, occasionally, hemolytic uremic syndrome, a sequela of infection that can result in renal failure and death. Here we sought to model the pathogenesis of orally-administered E. coli O157:H7 in BALB/c mice with an intact intestinal flora. First, we defined the optimal dose that permitted sustained fecal shedding of E. coli O157:H7 over 7 days ( approximately 10(9) colony forming units). Next, we monitored the load of E. coli O157:H7 in intestinal sections over time and observed that the cecum was consistently the tissue with the highest E. coli O157:H7 recovery. We then followed the expression of two key E. coli O157:H7 virulence factors, the adhesin intimin and Shiga toxin type 2, and detected both proteins early in infection when bacterial burdens were highest. Additionally, we noted that during infection, animals lost weight and approximately 30% died. Moribund animals also exhibited elevated levels of blood urea nitrogen, and, on necropsy, showed evidence of renal tubular damage. We conclude that conventional mice inoculated orally with high doses of E. coli O157:H7 can be used to model both intestinal colonization and subsequent development of certain extraintestinal manifestations of E. coli O157:H7 disease.

Fig. 1

Colonization of BALB/c mice by E. coli O157:H7strain 86-24Nal^R. (A) Percent of infected mice colonized over time. Animals were intragastrically infected with 86-24Nal^R at inocula of 10⁵ (●), 10⁶ (), 10⁷ (▲), 10⁸ (), and 10⁹ CFU (♦). Mice were considered colonized if there were recoverable 86-24Nal^R in the feces. (B) Colonization level of 86-24Nal^R overtime at various doses administered intragastrically (symbols as above). For panels C and D, animals were infected with 86-24Nal^R at doses of 10⁸ CFU or 10⁹ CFU either by intragastric administration (dashed lines with for 10⁸ CFU and ♦ for 10⁹ CFU) or pipette feeding (solid lines with for 10⁸ CFU and ■ for 10⁹ CFU). (C) Effect of administration route on percent of mice colonized. (D) Impact of administration route on colonization level. For panels A and C, animals that died during the study were excluded from the study after the time of death; these included 2 mice on day 4 from the 10⁹ dose (gavage) and one animal each on days 4 and 5 from the 10⁸ dose (gavage). For panels B and D, each point represents the GM of the number of 86-24Nal^R shed into the feces from surviving, infected animals, and the limit of detection (indicted by the black dashed line) was 10² CFU/g feces.

Fig. 1

Colonization of BALB/c mice by E. coli O157:H7strain 86-24Nal^R. (A) Percent of infected mice colonized over time. Animals were intragastrically infected with 86-24Nal^R at inocula of 10⁵ (●), 10⁶ (), 10⁷ (▲), 10⁸ (), and 10⁹ CFU (♦). Mice were considered colonized if there were recoverable 86-24Nal^R in the feces. (B) Colonization level of 86-24Nal^R overtime at various doses administered intragastrically (symbols as above). For panels C and D, animals were infected with 86-24Nal^R at doses of 10⁸ CFU or 10⁹ CFU either by intragastric administration (dashed lines with for 10⁸ CFU and ♦ for 10⁹ CFU) or pipette feeding (solid lines with for 10⁸ CFU and ■ for 10⁹ CFU). (C) Effect of administration route on percent of mice colonized. (D) Impact of administration route on colonization level. For panels A and C, animals that died during the study were excluded from the study after the time of death; these included 2 mice on day 4 from the 10⁹ dose (gavage) and one animal each on days 4 and 5 from the 10⁸ dose (gavage). For panels B and D, each point represents the GM of the number of 86-24Nal^R shed into the feces from surviving, infected animals, and the limit of detection (indicted by the black dashed line) was 10² CFU/g feces.

Fig. 2

Levels of E. coli O157:H7 over time at different sites in the gastrointestinal tract of BALB/c mice orally-infected with 6–8 × 10⁸ CFU of strain 86-24Nal^R. The GM CFU/g tissue (A) or luminal contents (B) from groups of 8–17 mice each are shown for the small intestine (●), cecum (▲), and large intestine () as well as the luminal contents of the small intestine (○), cecum (Δ), and large intestine (). Bars indicate the 95% confidence interval.

Fig. 3

Immunofluorescent E. coli O157:H7 in the cecum. Mice were infected with E. coli O157:H7 (A–F) or mock-infected (G–I) for 6 hours. Images display infected or control ceca stained for O157 (A, D, and G), the phase contrast image of the cecal section (B, E, and H), and the merged images (C, F, I). Panels D, E, and F are a higher magnification of the boxed area of panel B. For panels A–C and panels G–I the original image was obtained at a magnification of 40×. For panels D–F the original image was obtained at a magnification of 100×.

Fig. 4

Presence of intimin and Stx2 over time at different sites in the gastrointestinal tracts of BALB/c mice orally infected with E. coli O157:H7 strain 86-24Nal^R. (A) Percent of different intestinal samples positive for intimin over time. Each bar represents the mean of 8–17 samples tested. For panels A–D, the tissues are represented by black bars (cecum), white bars (cecal contents), hatched bars (large intestines), and gray bars (large intestinal contents). (B) Levels of intimin in different intestinal sites over time. Each bar represents the mean value of a subset of those samples presented in panel A (n=5–10). The limit of quantifiable intimin was 360 ng/g tissue or tissue contents. For panels B and D, each individual sample is represented by a ○. (C) Percent of different intestinal samples positive for Stx2 over time. Each bar represents the mean of 7–16 samples tested. (D) Levels of Stx2 in different intestinal sites over time. Each bar represents the mean value of a subset of those samples presented in panel C (n=5–10). The limit of quantifiable Stx2 was 6.4 ng/g tissue or tissue contents.

Fig. 4

Presence of intimin and Stx2 over time at different sites in the gastrointestinal tracts of BALB/c mice orally infected with E. coli O157:H7 strain 86-24Nal^R. (A) Percent of different intestinal samples positive for intimin over time. Each bar represents the mean of 8–17 samples tested. For panels A–D, the tissues are represented by black bars (cecum), white bars (cecal contents), hatched bars (large intestines), and gray bars (large intestinal contents). (B) Levels of intimin in different intestinal sites over time. Each bar represents the mean value of a subset of those samples presented in panel A (n=5–10). The limit of quantifiable intimin was 360 ng/g tissue or tissue contents. For panels B and D, each individual sample is represented by a ○. (C) Percent of different intestinal samples positive for Stx2 over time. Each bar represents the mean of 7–16 samples tested. (D) Levels of Stx2 in different intestinal sites over time. Each bar represents the mean value of a subset of those samples presented in panel C (n=5–10). The limit of quantifiable Stx2 was 6.4 ng/g tissue or tissue contents.

Fig. 5

Effect of intragastric infection of BALB/c mice with ~ 10⁹ CFU of E. coli O157:H7 strain 86-24Nal^R on body weight and survival. Animals were starved overnight prior to intragastric inoculation with E. coli O157:H7 on day 0. (A) Weight was monitored for the control, mock-infected animals (●) and the 86-24Nal^R-infected, experimental group (■). Each point on the graph represents the mean weight of 10 (or less in the case of groups with mortality) animals per group. The bars depict ± one standard error of the mean of weight. Asterisks indicate days that demonstrated a significant difference between weights of infected and control mice (p≤0.001). (B) Survival of animals infected with 86-24Nal^R or mock-infected. The percent of surviving animals (10 mice per group) on each day post-infection inoculated with 86-24Nal^R (■) or mock-infected (●).

Figure 6

Renal lesions in mice infected by gavage with E. coli O157:H7 strain 86-24Nal^R. Panels A, B, and C display kidney sections stained with H&E and taken from infected (panels A and B) or control (panel C) mice. The original images were obtained at 40× magnification. (A) The mouse was necropsied on day 5 post-infection. Multiple dilated cortical tubules are lined by epithelial cells that are hypereosinophilic, shrunken, angular, and pyknotic [indicative of necrosis (arrowheads)] or hypertrophic with vacuolated cytoplasm [indicative of degeneration (arrow)]. (B) The mouse was necropsied on day 5 post-infection. Cortical tubule epithelial cells have basophilic cytoplasm, are karyomegalic (arrowheads), and exhibit anisocytosis (presence in the blood of erythrocytes with excessive variation in size) with mitotic figures (arrow), all of which is indicative of regeneration.