Altering host resistance to infections through microbial transplantation

Abstract

Host resistance to bacterial infections is thought to be dictated by host genetic factors. Infections by the natural murine enteric pathogen Citrobacter rodentium (used as a model of human enteropathogenic and enterohaemorrhagic E. coli infections) vary between mice strains, from mild self-resolving colonization in NIH Swiss mice to lethality in C3H/HeJ mice. However, no clear genetic component had been shown to be responsible for the differences observed with C. rodentium infections. Because the intestinal microbiota is important in regulating resistance to infection, and microbial composition is dependent on host genotype, it was tested whether variations in microbial composition between mouse strains contributed to differences in "host" susceptibility by transferring the microbiota of resistant mice to lethally susceptible mice prior to infection. Successful transfer of the microbiota from resistant to susceptible mice resulted in delayed pathogen colonization and mortality. Delayed mortality was associated with increased IL-22 mediated innate defense including antimicrobial peptides Reg3γ and Reg3β, and immunono-neutralization of IL-22 abrogated the beneficial effect of microbiota transfer. Conversely, depletion of the native microbiota in resistant mice by antibiotics and transfer of the susceptible mouse microbiota resulted in reduced innate defenses and greater pathology upon infection. This work demonstrates the importance of the microbiota and how it regulates mucosal immunity, providing an important factor in susceptibility to enteric infection. Transfer of resistance through microbial transplantation (bacteriotherapy) provides additional mechanisms to alter "host" resistance, and a novel means to alter enteric infection and to study host-pathogen interactions.

Figure 1. Analysis of microbial 16S rRNA gene composition indicates successful transplantation of microbiotas between mouse strains.

(a) Similarity tree using Bray Curtis metrics of bacterial 16S rRNA gene terminal restriction fragment profiles from untreated C3H/HeJ (HeJ) and NIH Swiss (NIH) mice and HeJ or NIH mice having received oral microbial transplantation from HeJ mice (HeJ-HeJ or NIH-HeJ) or NIH mice (HeJ-NIH or NIH-NIH). Mice receiving transplantation were from 4 separate experiments and represent multiple cages in each experiment. The number denotes which experiment the samples were from. (b) Classification of 16S rRNA gene clone libraries at the family level generated from fecal samples from HeJ, NIH and HeJ-NIH mice.

Figure 2. Mice with NIH microbiota showed reduced susceptibility to C. rodentium infection.

After microbiota transplantation, mice were challenged with 10⁸ or 10⁴ colony forming units (CFU) of C. rodentium. Shedding of C. rodentium, body weight loss and survival were assessed for host susceptibility. Data is representative of three independent experiments with high infection dose (a,b,c), or two independent experiments with low infection dose (d,e,f). HeJ-NIH mice carrying NIH microbiota had significantly lower C. rodentium shedding in feces at the early time points post infection, delayed weight loss and mortality. Weights are represented as mean±SEM. (n = 8 for each experiment).

Figure 3. Increased C. rodentium colonization (a), weight loss (b) and intestinal pathology (c) in NIH-HeJ as compared NIH-NIH mice after infection.

Pathology was assessed in four regions including the lumen, surface epithelium, mucosa and submucosa. Data is representative of 2 independent experiments. * P<0.05 ** P<0.01 (n = 8 for each experiment).

Figure 4. NIH microbiota induced expression of IL-22 and reg3β.

Relative expression in the ileum (a) for IL22-, Reg3γ-, and Reg3β-specific mRNA in NIH Swiss (NIH), C3H/HeJ transplanted with NIH microbiota (HeJ-NIH), C3H/HeJ transplanted with HeJ-Naïve (HeJ-HeJ) and C3H/HeJ (HeJ-Naïve) mice relative to HeJ-HeJ mice. (b) Immunostaining for Reg3β in ileal sections shows abundant peptide in ilea of NIH and HeJ-NIH mice, but not in HeJ-HeJ and HeJ-Naïve mice. The scale bar is equal to 50 µm. Data represents mean±SEM. (c) IL-22 and Reg3β expression in NIH mice transplanted with HeJ microbiota (NIH-HeJ) or NIH microbiota (NIH-NIH). All figures representative of two independents experiments with n = 4.

Figure 5. Immunoneutralization of IL-22 impacts survival in HeJ-NIH but not HeJ-HeJ mice.

Transcript abundance of Reg3β before infection (a) and survival curves of HeJ mice receiving (b) NIH microbiota or (c) HeJ microbiota after infection with C. rodentium and intraperitoneal injection of anti-IL-22 or isotype control. (n = 4).

Figure 6. NIH microbiota and improved survival is maintained in the following generation.

(a) Similarity tree using Bray Curtis metrics of bacterial 16S rRNA gene terminal restriction fragment profiles from the offspring of untreated C3H/HeJ (HeJ) and NIH Swiss (NIH) mice and HeJ mice having received oral microbial transplantation from HeJ mice (HeJ-HeJ) or NIH mice (HeJ-NIH). Parent mice are identified with a the letter p. (b) Relative expression in the ileum for IL22- and Reg3β-specific mRNA in the offspring of HeJ-Naïve, HeJ-NIH and HeJ-HeJ mice (n = 5). (c) Delayed colonization and (d) mortality in the offspring of HeJ-NIH as compared to HeJ-HeJ mice (n = 8).