Microbiota-Regulated IL-25 Increases Eosinophil Number to Provide Protection during Clostridium difficile Infection

Abstract

Clostridium difficile infection (CDI) is the most common cause of hospital-acquired infection in the United States. Host susceptibility and the severity of infection are influenced by disruption of the microbiota and the immune response. However, how the microbiota regulate immune responses to mediate CDI outcome remains unclear. Here, we have investigated the role of the microbiota-linked cytokine IL-25 during infection. Intestinal IL-25 was suppressed during CDI in humans and mice. Restoration of IL-25 reduced CDI-associated mortality and tissue pathology even though equivalent levels of C. difficile bacteria and toxin remained in the gut. IL-25 protection was mediated by gut eosinophils, as demonstrated by an increase in intestinal eosinophils and a loss of IL-25 protection upon eosinophil depletion. These findings support a mechanism whereby the induction of IL-25-mediated eosinophilia can reduce host mortality during active CDI. This work may provide targets for future development of microbial or immune-based therapies.

Figure 1. IL-25 was suppressed during human and murine Clostridium difficile infection (CDI). Related to Table 1

(A) Representative histology of human colonic biopsies from CDI negative (−) (n=9) and CDI positive (+) (n=5) patients stained for IL-25 protein expression (Scale = 50 μM). (B) Histology was scored for IL-25 expression; four independent blinded scorers; *p value<0.05 (C) Representative immunohistochemical staining for IL-25 in ceca of C57BL/6J mice that were untreated (UT), antibiotic treated (ABX), or infected with C. difficile (Scale = 100 μM). (D) IL-25 protein in mouse cecal tissue measured by ELISA. Data represents two combined experiments; n= 5–8 mice per group per experiment; mean +/− SEM; p value from antibiotic treated *< 0.05, p value from untreated #<0.05, ##<0.005, ###<0.0005. (E–F) Lamina propria (LP) and epithelial cells (EC) in the colon were separated and analyzed for IL-25 protein from untreated, antibiotic only, and day 3 post C. difficile infected mice. (E) Data represents IL-25 protein from combined time points (F) Data represents IL-25 protein in the epithelium alone on each time point. n=4–10 per group; mean +/− SEM; *p value < 0.05.

Figure 2. Recombinant IL-25 pretreatment protected against CDI-associated mortality and morbidity without changing C. difficile CFU or toxin. Related to Figure S1 and S2

C57BL/6J mice were treated with a daily dose of either 0.5 μg of recombinant IL-25 protein or PBS daily for five days prior to infection with C. difficile. (A) Survival and (B) clinical scores over the initial 6 days of infection. Data represents four combined experiments; n=6–10 mice per group per experiment; mean +/− SEM; * <0.05, ***<0.0005. (C) Representative H&E stained cecal sections of mice on day 3 after infection with C. difficile (Scale = 100 μM) and (D) pathology scores. (E) Toxin A/B levels and (F) C. difficile bacterial burden in cecal contents on day 3 post C. difficile. Data represents two combined experiments; n= 4–7 mice per group per experiment; mean +/− SEM; *<0.05, **<0.005.

Figure 3. IL-25 pretreatment increased IL-4 and mucin expression during CDI. Related to Figure S3 and S4

(A) ELISA analysis of protein expression of type 17 and type 2 cytokines cecal tissue of C57BL/6J mice on day 3 of C. difficile infection. Data represents two combined experiments; n=6–8 mice per group per experiment; mean +/− SEM; *<0.05, **<0.005. (B) Periodic acid-Schiff (PAS) staining of mucins in control and IL-25 treated cecal sections (Scale= 50 μM) and (C) scoring on day 3 of C. difficile infection. Data represents two combined experiments; n=4–6 mice per group per experiment; mean +/− SEM; **p<0.005. (D) Fold change of muc2 RNA in cecal tissue by qPCR relative to gapdh and actin. Data represents three combined experiments; n=4–6 mice per group per experiment; mean +/− SEM; *<0.05.

Figure 4. IL-25 increased lamina propria eosinophils during CDI. Related to Figure S3

(A) CD45⁺ CD11b⁺ CD11c^mid SiglecF⁺ Ly6g⁻ eosinophils and (B) CD45⁺ CD11b⁺ Ly6g⁺ Ly6c⁺ neutrophils were isolated from the colonic lamina propria and quantified by flow cytometry for absolute numbers and percentage of live cells on day 3 of CDI. Representative flow plot shows neutrophils and eosinophils as a percentage of live cells gated from CD11b+ cells. Data represents three combined experiments; n= 4–6 mice per group per experiment; mean +/− SEM; Students two tailed t-test **<0.005. #<0.05 from uninfected PBS treated group. (C) Absolute number and (D) percentage of live eosinophils in the lamina propria of PBS and IL-25 treated mice plotted against clinical scores on day 3 of CDI. Data is representative of three experiments; n= 4–6 mice per group per experiment.

Figure 5. IL-25 protected from CDI through an eosinophil-dependent mechanism. Related to Figure S5

(A–B) C57BL/6J mice +/− IL-25 treatment were given with 20 μg of anti-SiglecF or isotype control one day prior and one day after infection with C. difficile and assessed for (A) survival and (B) clinical scores during infection. Data is representative of two experiments; n=10 mice per group per experiment; mean +/− SEM; *<0.05. (C–D) C57BL/6J or PHIL mice +/− IL-25 treatment and infected with a sub-lethal dose of 10^{^3} CFU of C. difficile were assessed for (C) survival and (D) clinical scores. Data represents two combined experiments; n=5–10 mice per group; mean +/− SEM; p value *< 0.05, **<0.005, ***<0.0005 compared to IL-25+anti-SiglecF infected mice.

Figure 6. Eosinophils were necessary for IL-25-mediated maintenance of the intestinal epithelial barrier during CDI. Related to Figure S6

(A) H&E staining and (B) tissue pathology scores of cecal tissue from C57BL/6J mice treated with PBS, rIL-25, or rIL-25+anti-Siglecf on day 3 of CDI (Scale = 20μm). Data represents two combined experiments; n=4–7 mice per group per experiment; mean +/− SEM; *<0.05, **<0.005. (C) Albumin concentration in the cecal contents on day 3 of CDI. Data is from three combined experiments; n=2–5 mice per infected groups per experiment; mean +/− SEM; *<0.05, **<0.005. (D) Colon length and (E) toxin A/B level and (F) C. difficile bacterial burden in cecal contents on day 3 of CDI. Data is from two combined experiments; n=4–7 mice per group per experiment; mean +/− SEM; *<0.05, **<0.005.