The gut microbiota-independent virulence of noninvasive bacterial pathogen Citrobacter rodentium

Abstract

Attaching and effacing (A/E) bacterial pathogens consist of human pathogens enteropathogenic Escherichia coli, enterohemorrhagic E. coli and their murine equivalent Citrobacter rodentium (CR). Emerging evidence suggests that the complex pathogen-microbiota-host interactions are critical in conferring A/E pathogen infection-induced severe symptoms and lethality in immunocompromised hosts; however, the precise underlying mechanisms remain enigmatic. Here we report that CR infection causes severe colitis and mortality in interleukin 22 knockout (Il22-/-) and Rag1 knockout (Rag1-/-) mice under germ-free (GF) conditions. In a gut microbiota-independent manner, CR colonizes in GF Il22-/- and Rag1-/- animals, triggers colonic epithelial tissue damage and systemic dissemination of CR, and results in lethal infections. Pretreatment with cefoxitin, a broad-spectrum antibiotic, exacerbates CR-induced colitis and lethality in specific-pathogen-free (SPF) Il22-/- and Rag1-/- mice. Together our results reveal that CR possesses a gut microbiota-independent virulence, which is better illustrated during infections in immunocompromised hosts associated with severe outcomes.

Fig 1. CR causes lethal infection in SPF and GF Il22^-/- mice.

(A-D) Fecal live Citrobacter rodentium (CR) burden (A), body weight loss (B), clinical scores (C), and survival (D) of specific-pathogen-free (SPF) Il22^-/- mice in C57Bl/6J background at indicated days post inoculation (dpi) with 2 × 10⁹ CFU of CR or vehicle control. (E) CR burdens in the liver (top) and the spleen (bottom) derived from SPF Il22^-/- mice at 10 dpi infected with vehicle control or CR. No CR burden was detected in vehicle controls. (F-I) Fecal live CR burden (F), body weight loss (G), clinical scores (H) and survival (I) of germ-free (GF) Il22^-/- mice at indicated dpi with 2 × 10⁹ CFU of CR or vehicle control. Data are representative results of at least two independent experiments. The error bars (A and F) are too small to be visible. ** p < 0.01, *** p < 0.001, **** p < 0.0001 with Student’s t tests (A-C, and G-H) and Long-rank test (D and I).

Fig 2. CR colonization causes lethality in GF Il22^-/- mice.

(A) Representative images of the cecum and the colon derived from GF Il22^-/- mice at 5 days post inoculation (dpi) infected with vehicle control or 2 × 10⁹ CFU of CR. Scale bars, 1 cm. (B) Colon lengths of GF Il22^-/- mice infected with vehicle control or CR. (C-D) Hematoxylin and eosin staining (C) and histopathology scores (D) of colon sections derived from GF Il22^-/- mice at 5 dpi infected with vehicle control or CR. Scale bars, 100 μm. (E) Crypt lengths measured in colonic tissue sections from at least 4 GF Il22^-/- mice infected with CR or vehicle control at 5 dpi. (F) Immunofluorescence micrographs of Claudin-3 and CR in the colons derived from GF Il22^-/- mice at 5 dpi infected with vehicle control or CR, with nuclei counterstained by DAPI. Lum indicates the colon luminal space. Scale bars, 100 μm. (G) Live CR burdens in the liver (left) and the spleen (right) derived from GF Il22^-/- mice at 5 dpi infected with vehicle control or CR. No CR burden was detected in vehicle controls. (H) FITC-dextran concentrations in the sera of GF Il22^-/- mice at 5 dpi infected with vehicle control or CR, at 4 h post oral administration of FITC-dextran. (I) Representative ex vivo tissue images of GF Il22^-/- mice infected with vehicle control or CR at 5 dpi, with an IVIS50 camera and displayed as pseudo colors. The color scale bar indicates bioluminescence signal intensity (photons sec^-1 cm^-2 sr^-1). Data are representative results of at least two independent experiments. **** p < 0.0001 with Student’s t tests (B, D and E).

Fig 3. CR colonization leads to lethal colitis in GF Rag1^-/- mice.

(A) Fecal live CR burden of germ-free (GF) Rag1^-/- mice at indicated dpi with 2 × 10⁹ CFU of CR or vehicle control. (B) Representative ex vivo tissue images of GF Rag1^-/- mice infected with vehicle control or CR at 5 days post inoculation (dpi), with an IVIS50 camera and displayed as pseudo colors. The color scale bar indicates bioluminescence signal intensity (photons sec^-1 cm^-2 sr^-1). (C) FITC-dextran concentrations in the sera of GF Rag1^-/- mice at 14 dpi infected with vehicle control or CR, at 4 h post oral administration of FITC-dextran. (D) Live CR burdens in the liver (left) and the spleen (right) derived from GF Rag1^-/- mice at 14 dpi infected with vehicle control or CR. No CR burden was detected in vehicle controls. (E-F) Body weight loss (E) and survival (F) of GF Rag1^-/- mice at indicated dpi with vehicle control or CR. Data are representative results of at least two independent experiments. ** p < 0.01, *** p < 0.001, **** p < 0.0001, with Student’s t tests (E) and Long-rank test (F).

Fig 4. The gut microbiota offers protection against CR-caused lethality in SPF Il22^-/- and Rag1^-/- mice.

(A) Schematic of Citrobacter rodentium (CR) infection experiment in specific-pathogen-free (SPF) Il22^-/- mice. The indicated mice were pretreated with normal drinking water (Vehicle) or drinking water containing Cefoxitin (500 mg/L) for 48 h. 24 h after removal of antibiotic-containing drinking water, the mice were orally inoculated with 2 × 10⁹ CFU of CR. (B-D) Fecal live CR burden (B), body weight loss (C) and survival (D) of vehicle control- or Cefoxitin-pretreated SPF Il22^-/- mice at indicated days post inoculation (dpi) with CR. (E) CR burdens in the liver (left) and the spleen (right) derived from vehicle control- or Cefoxitin-pretreated SPF Il22^-/- mice at 5 dpi infected with CR. (F-G) SPF Rag1^-/- mice were pretreated with normal (Vehicle) or Cefoxitin-containing drinking water, followed with CR infection, as described in (A). Body weight loss (F) and survival (G) of vehicle control- or Cefoxitin-pretreated SPF Rag1^-/- mice at indicated dpi with 2 × 10⁹ CFU of CR. Data are representative results of at least two independent experiments. ns, not significant; * p < 0.05, ** p < 0.01, with Student’s t tests (C and F) and Long-rank test (D and G).