Bacillus subtilis-mediated protection from Citrobacter rodentium-associated enteric disease requires espH and functional flagella

Abstract

Commensals limit disease caused by invading pathogens; however, the mechanisms and genes utilized by beneficial microbes to inhibit pathogenesis are poorly understood. The attaching and effacing mouse pathogen Citrobacter rodentium associates intimately with the intestinal epithelium, and infections result in acute colitis. C. rodentium is used to model the human pathogens enterohemorrhagic Escherichia coli and enteropathogenic E. coli. To confirm that Bacillus subtilis, a spore-forming bacterium found in the gut of mammals, could reduce C. rodentium-associated disease, mice received wild-type B. subtilis spores and 24 h later were infected by oral gavage with pathogenic C. rodentium. Disease was assessed by determining the extent of colonic epithelial hyperplasia, goblet cell loss, diarrhea, and pathogen colonization. Mice that received wild-type B. subtilis prior to enteric infection were protected from disease even though C. rodentium colonization was not inhibited. In contrast, espH and hag mutants, defective in exopolysaccharides and flagellum production, respectively, did not protect mice from C. rodentium-associated disease. A motAB mutant also failed to protect mice from disease, suggesting that B. subtilis-mediated protection requires functional flagella. By expanding our current mechanistic knowledge of bacterial protection, we can better utilize beneficial microbes to prevent intestinal disease caused by pathogenic bacteria, ultimately reducing human disease. Our data demonstrate that wild-type B. subtilis reduced disease caused by C. rodentium infection through a mechanism that required espH and functional flagella.

Fig 1

Effect of wild-type B. subtilis on C. rodentium-associated disease 10 days postinfection as demonstrated by histological analysis of colonic tissues. (A to D) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received PBS only (A), wild-type (wt) B. subtilis only (B), C. rodentium (Cr) only (C), or wild-type B. subtilis spores prior to C. rodentium infection (D). Orange arrows indicate goblet cells. (E and F) Colonic crypt heights (E) and goblet cell measurements (F). Results are averages from three independent experiments, and a total of 6 to 12 mice were assessed for each group. P values were calculated by Student's t test.

Fig 2

Effect of wild-type B. subtilis on C. rodentium-associated disease 10 days postinfection as demonstrated by diarrhea scores. Results are averages from three independent experiments; a total of 8 to 12 mice were assessed for each group. P values were calculated by Student's t test.

Fig 3

C. rodentium colonization was assessed by culturing serial dilutions of fecal homogenates. N.D., none detected (the level was less than 10³ CFU/g feces). ●, values for mice receiving C. rodentium only; ■, values for mice receiving wild-type B. subtilis prior to C. rodentium infection. Results are averages from three independent experiments, and a total of 8 to 10 mice were assessed for each group.

Fig 4

Effects of espH and hag spores on C. rodentium-associated disease 10 days postinfection. (A to D) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received espH B. subtilis spores (A), espH B. subtilis spores prior to C. rodentium infection (B), hag B. subtilis spores (C), and hag B. subtilis spores prior to C. rodentium infection (D). H&E-stained sections from PBS-treated and C. rodentium-infected mice are shown in Fig. 1. (E to G) Colonic crypt height measurements (E), goblet cell measurements (F), and diarrhea scores (G). Results are averages from three independent experiments, and a total of 6 to 12 mice were assessed for each group. ∗, P < 0.05 (Student's t test).

Fig 5

Assessment of B. subtilis colonization, germination, motility, and biofilm development. (A) B. subtilis colonization was assessed by culturing serial dilutions of fecal homogenates. Results are averages from two independent experiments, and a total of 5 or 6 mice were assessed. (B) Germination of B. subtilis spores in LB medium. (C) B. subtilis motility in soft-agar plates. (D) Biofilm formation by B. subtilis strains. Germination, motility, and biofilm results are averages from three independent experiments.

Fig 6

Effect of coadministration of espH and hag spores on C. rodentium-associated disease 10 days postinfection. (A and B) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received espH and hag B. subtilis spores only (A) or espH and hag B. subtilis spores prior to C. rodentium infection (B). H&E-stained sections from PBS-treated and C. rodentium-infected mice are shown in Fig. 1. (C to E) Colonic crypt height measurements (C), goblet cell measurements (D), and diarrhea scores (E). Results are averages from three independent experiments; a total of 6 to 8 mice were assessed for each group. ∗, P < 0.05 (Student's t test).

Fig 7

Effect of motAB spores on C. rodentium-associated disease 10 days postinfection. (A and B) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received motAB B. subtilis spores only (A) or motAB B. subtilis spores prior to C. rodentium infection (B). H&E-stained sections from PBS-treated and C. rodentium-infected mice are shown in Fig. 1. (C to E) Colonic crypt height measurements (C), goblet cell measurements (D), and diarrhea scores (E). Results are averages from three independent experiments; a total of 6 to 8 mice were assessed for each group. ∗, P < 0.05 (Student's t test).