Innate immune detection of flagellin positively and negatively regulates salmonella infection

Abstract

Salmonella enterica serovar Typhimurium is a flagellated bacterium and one of the leading causes of gastroenteritis in humans. Bacterial flagellin is required for motility and also a prime target of the innate immune system. Innate immune recognition of flagellin is mediated by at least two independent pathways, TLR5 and Naip5-Naip6/NlrC4/Caspase-1. The functional significance of each of the two independent flagellin recognition systems for host defense against wild type Salmonella infection is complex, and innate immune detection of flagellin contributes to both protection and susceptibility. We hypothesized that efficient modulation of flagellin expression in vivo permits Salmonella to evade innate immune detection and limit the functional role of flagellin-specific host innate defenses. To test this hypothesis, we used Salmonella deficient in the anti-sigma factor flgM, which overproduce flagella and are attenuated in vivo. In this study we demonstrate that flagellin recognition by the innate immune system is responsible for the attenuation of flgM(-) S. Typhimurium, and dissect the contribution of each flagellin recognition pathway to bacterial clearance and inflammation. We demonstrate that caspase-1 controls mucosal and systemic infection of flgM(-) S. Typhimurium, and also limits intestinal inflammation and injury. In contrast, TLR5 paradoxically promotes bacterial colonization in the cecum and systemic infection, but attenuates intestinal inflammation. Our results indicate that Salmonella evasion of caspase-1 dependent flagellin recognition is critical for establishing infection and that evasion of TLR5 and caspase-1 dependent flagellin recognition helps Salmonella induce intestinal inflammation and establish a niche in the inflamed gut.

Figure 1. Wildtype Salmonella efficiently evades flagellin detection during acute mucosal infection in streptomycin-pretreated mice.

Bacterial burden of WT C57BL/6 (n = 11) and caspase-1−/− (n = 8) mice infected with 1000 cfu WT SL1344 Salmonella in the cecum (A), spleen (B), liver (C), MLN (D). Bacterial burden of WT C57BL/6 (n = 12) and TLR5−/− (n = 13) mice infected with 1000 cfu wildtype Salmonella in the cecum (E), spleen (F), liver (G), MLN (H). Figures A–D are the combined data from two independent experiments. Figures E–H are the combined data from three independent experiments. Mann-Whitney test. * = p<0.05. ** = p<0.01. *** = p<0.001.

Figure 2. Flagellin detection accounts for the caspase-1 increased susceptibility during acute mucosal infection in streptomycin-pretreated mice.

Bacterial burden of WT C57BL/6 (n = 10) and caspase-1−/− (n = 16) mice infected with 1000 cfu flagellin⁻ Salmonella in the cecum (A), spleen (B), liver (C), MLN (D). Bacterial burden WT C57BL/6 (n = 15) and TLR5−/− (n = 16) mice infected with 1000 cfu flagellin⁻ Salmonella in the cecum (E), spleen (F), liver (G), MLN (H). Figures A–D are the combined data from two independent experiments. Figures E–H are the combined data from three independent experiments. Mann-Whitney test * = p<0.05. *** = p<0.001.

Figure 3. flgM⁻ Salmonella has an attenuated in vivo phenotype.

C57BL/6 mice were infected with 1000 CFU of either WT SL1344 (n = 12) or flgM⁻ (n = 15) Salmonella. (A) Bacterial burden in the cecum. (B) Spleen. (C) Liver. (D) MLN. Representative histology of the cecum (E+F) and liver (G+H) infected with WT (E+G) or flgM⁻ Salmonella (F+H). (I) Histological scores for changes in the cecum. Submucosal expansion (S); mucosal neutrophilic infiltrate (M); lymphoplasmacytosis (L); goblet cells (G); epithelial integrity (E). Liver images (400X magnification), cecal images (200X magnification). Figures A–D are the combined data from three independent experiments. Mann-Whitney test. ** = p<0.01, *** = p<0.001.

Figure 4. flgM⁻ Salmonella attenuated phenotype is dependent on flagellin.

C57BL/6 WT mice were infected with 1000 CFU of either flgM^−/flagellin⁻ (n = 5) or flgM⁻ (n = 5) Salmonella. Bacterial burden in the cecum (A). MLN (B). Spleen (C). Liver (D). (E) Histological scores for changes in the cecum in cecal pathology as described in Fig. 4. Figures A–D represent data from one experiment (flgM⁻ n = 5; flgM^−/flagellin⁻ n = 5). Mann Whitney test. * = p<0.05, ** = p<0.01, *** = p<0.001.

Figure 5. Caspase-1 is required for attenuation of flgM-deficient Salmonella.

C57BL/6 WT (n = 9) and caspase 1−/− (n = 8) mice were infected with 1000 CFU flgM⁻ Salmonella. Bacterial burden in the cecum (A). Spleen (B). Liver (C). MLN (D). Representative histology of the cecum of WT mice (E) or caspase-1−/− (F), and liver of WT mice (G) or caspase-1−/− (H). (I) Histological scores for changes in the cecal pathology as described in Fig. 4. Figures A–D represent data from two independent experiments. Liver images (200X magnification), cecal images (100X magnification). Mann-Whitney test, * = p<0.05, ** = p<0.01, *** = p<0.001.

Figure 6. Salmonella reside within F4/80+ cells the cecum.

C57BL/6 WT (n = 5) (A+C) or Caspase-1−/− (n = 5) (B+D) mice were infected with 1000 cfu of flgM⁻ Salmonella containing a stable GFP expressing plasmid. Intestinal epithelial cells were stained using TROMA (A+B), and phagocytes were stained using F4/80 (C+D) antibody. Red = TROMA-1 (A, B) and F4/80 (C, D); green = GFP (Salmonella). Shown are representative images from 10 mice examined.

Figure 7. The innate immune receptor TLR5 is dispensable in the attenuation of flgM⁻.

C57BL/6 WT (n = 14) and TLR5−/− (n = 16) mice were infected with 1000 CFU flgM⁻ Salmonella. Bacterial burden in the cecum (A). Spleen (B). Liver (C). MLN (D). Representative histology of the cecum for WT mice (E) or TLR5−/− (F), and liver of WT mice (G) or TLR5−/− (H). Liver images (200X magnification), cecal images (100X magnification). Figures A–D represent data from three independent experiments. Mann-Whitney test * = p<0.05. ** = p<0.01.(I) Histological scores for changes in the cecal pathology as described in Fig. 4. Figures A–D represent data from three independent experiments. Mann-Whitney test *** = p<0.001.

Figure 8. Composite analysis of Salmonella WT, flagellin- and flgM⁻ infections in WT, caspase-1−/− and TLR5−/− mice.

We combined the CFU data for all experiments using WT, flagellin deficient and flgM⁻ S. Typhimurium SL1344, and the data was analyzed using two-way ANOVA and Bonferroni’s multiple comparisons post tests. * = P<0.05, ** = P<0.01, and *** = P<0.001. Cecum (A), Spleen (B), Liver (C), and MLN (D).