Microbiota of MR1 deficient mice confer resistance against Clostridium difficile infection

Abstract

Clostridium difficile (Cd) infection (CDI) typically occurs after antibiotic usage perturbs the gut microbiota. Mucosa-associated invariant T cells (MAIT) are found in the gut and their development is dependent on Major histocompatibility complex-related protein 1 (MR1) and the host microbiome. Here we were interested in determining whether the absence of MR1 impacts resistance to CDI. To this end, wild-type (WT) and MR1-/- mice were treated with antibiotics and then infected with Cd spores. Surprisingly, MR1-/- mice exhibited resistance to Cd colonization. 16S rRNA gene sequencing of feces revealed inherent differences in microbial composition. This colonization resistance was transferred from MR1-/- to WT mice via fecal microbiota transplantation, suggesting that MR1-dependent factors influence the microbiota, leading to CDI susceptibility.

Fig 1. Experimental design for experiments performed in this study.

(A) Standard model of CDI. Mice were given antibiotics in drinking water 5 days, normal water for two days rest, then inoculated with 1x10³ Cd630 spores. (B) Ten-day antibiotic treatment with inoculation of Cd630 or CdVPI strains. (C) Co-housing strategy. (D) FMT of conventional SPF WT and MR1^-/- mice. (E) FMT of SPF antibiotic treated or untreated donor stool into WT germ-free recipients.

Fig 2. MR1^-/- mice exhibit resistance to Cd colonization in multiple infection models.

(A) Cd colonization in WT and MR1^-/- mice in the standard cefoperazone model of CDI. WT n = 9, MR1^-/- n = 8. Data presented are mean ±SD. (B) Colonization levels were determined for WT and MR1^-/- mice pretreated with cefoperazone then inoculated with vegetative Cd cells and followed for three days PI. WT n = 5, MR1^-/- n = 6. Data presented are mean ±SD. (C) Colonization levels in WT and MR1^-/- mice for 14 days after streptomycin pretreatment and inoculation with Cd spores. n = 11 mice per group. Data presented are mean ± SD of colonization results from all mice across three independent experiments (n = 11 per group total). (D) Mice were pretreated with a cocktail of five antibiotics in the drinking water for five days followed by a switch to normal water and a single IP injection of clindamycin one day prior to inoculation with Cd spores. Cd colonization was followed for 17 days PI. n = 11 mice per group. Data presented are mean ±SD.

Fig 3. The gut microbiota composition of MR1^-/- mice contrasts that of WT mice before and after cefoperazone treatment.

(A) 16S rRNA gene sequencing was performed on fecal samples from mice before and after antibiotic treatment. Data shown are mean % relative abundance at the family level. “_unc” denotes sequences unclassified at the family level; the lowest level identified is listed. Only taxa with ≥ 1% are presented. Data presented are mean ±SD; before antibiotics: WT n = 6, MR1-/- n = 6; after antibiotics: WT n = 5, MR1-/-n = 7. (B) Chao1 richness estimations of alpha diversity among all samples analyzed in panel A before and after antibiotic treatment. (C) Principle coordinate analysis (PCoA) of ThetaYC distances. Before antibiotics: WT n = 6, MR1-/- n = 6. After antibiotics: WT n = 5, MR1-/-n = 7.

Fig 4. Cefoperazone absorption is the same in WT and MR1^-/- mice.

Levels of cefoperazone in intestinal contents were analyzed by UPLC/QTof-MS after one, three and five days of antibiotic treatment. Data shown are normalized based on the total intensity of each run and are mean ±SD. Day 1: p = 0.50, n = 2 mice per groups; Day 3: p = 0.72, n = 3 mice per group; Day 5: p = 0.70, n = 3 mice per group.

Fig 5. MR1^-/- mice are resistant to highly virulent strains of Cd.

(A) WT and MR1^-/- mice were inoculated with either 1x10³ Cd630 or CdVPI spores and assessed for disease daily. (B) H&E staining of colons harvested at time of euthanasia or at the study endpoint (4 days PI). n = 12 mice per group.

Fig 6. Altered Cd colonization resistance is not due to mouse strain breeding site differences.

(A) WT and MR1^-/- mice were co-housed for one month, separated for three weeks, and then put through the cefaperazone model of CDI. Colonization levels were determined for four days PI. Data presented are mean ±SD. WT n = 5 MR1^-/- n = 6. (B) β-diversity throughout co-housing. 2-D non-metric multidimensional scaling (NMDS) plot of ThetaYC distances for all samples collected during the experiment. WT samples taken before, on the last day of co-housing, and on the last day of separation are highlighted in red, black and blue, respectively. Remaining samples are depicted gray. Relative abundance of microbiota present in WT n = 5 (C) and MR1^-/- n = 6 (D) mice before co-housing (BCH), during (days 14, 24, and 31), three weeks after separation (day -8), upon initiation of cefoperazone treatment (day -7), after cefoperazone treatment (day 0), and after Cd inoculation (days 1–4).

Fig 7. Cd colonization resistance is transferrable to susceptible mice.

(A) WT and MR1^-/- mice were treated as pictured in Fig 1d and then taken off cefoperazone one day prior to FMT administration. Each strain was divided into two groups and inoculated with antibiotic treated stool from either autologous stool, or stool from the opposite strain. Colonization levels were tracked over time (B & C). Assessment of microbiota composition over time following FMTMR1^-/- → MR1^-/- n = 3, WT→ MR1^-/- n = 3, WT→WT n = 4 MR1^-/- →WT n = 3. Data presented are mean ±SD of one experiment.

Fig 8. Microbiota of MR1^-/- mice is responsible for colonization resistance.

(A) Germ-free C57BL/6NTac mice received FMT from SPF mice. Each donor (SPF WT and MR1^-/-) strain was separated into two groups, with and without antibiotic treatment. Colonization levels for each group were determined up to 14 days PI. (B-E) Microbial composition of previously germ-free mice following FMT. Data shown are mean ±SD of 2 independent experiments. n = 10 mice per group.

Fig 9. RLDA identifies families that contribute to colonization susceptibility.

(A) Regularized Fisher’s linear discriminant analysis (RLDA) was performed using the relative abundance data collected from germ-free mice administered FMT from WT and MR1^-/- SPF mice and colonization status as in Fig 5. The relative contribution of bacterial colonies differentiates the samples that undergo colonization and samples that do not. Bacteria are ordered in descending order of importance. (B) 3-D plot depicts RLDA of colonization status of previously germ-free mice. Colonized samples (red dots) vs non-colonized samples (blue dots) are arranged based on differentiator vectors. Samples are colored based on final colonization status, no stratification was performed based on timepoint.