Comparing Campylobacter jejuni to three other enteric pathogens in OligoMM12 mice reveals pathogen-specific host and microbiota responses

Abstract

Campylobacter jejuni, non-typhoidal Salmonella spp., Listeria monocytogenes and enteropathogenic/enterohemorrhagic Escherichia coli (EPEC/EHEC) are leading causes of food-borne illness worldwide. Citrobacter rodentium has been used to model EPEC and EHEC infection in mice. The gut microbiome is well-known to affect gut colonization and host responses to many food-borne pathogens. Recent progress has established gnotobiotic mice as valuable models to study how microbiota affect the enteric infections by S. Typhimurium, C. rodentium and L. monocytogenes. However, for C. jejuni, we are still lacking a suitable gnotobiotic mouse model. Moreover, the limited comparability of data across laboratories is often negatively affected by variations between different research facilities or murine microbiotas. In this study, we applied the standardized gnotobiotic OligoMM¹² microbiota mouse model and compared the infections in the same facility. We provide evidence of robust colonization and significant pathological changes in OligoMM¹² mice following infection with these pathogens. Moreover, we offer insights into pathogen-specific host responses and metabolite signatures, highlighting the advantages of a standardized mouse model for direct comparisons of factors influencing the pathogenesis of major food-borne pathogens. Notably, we reveal for the first time that C. jejuni stably colonizes OligoMM¹² mice, triggering inflammation. Additionally, our comparative approach successfully identifies pathogen-specific responses, including the detection of genes uniquely associated with C. jejuni infection in humans. These findings underscore the potential of the OligoMM¹² model as a versatile tool for advancing our understanding of food-borne pathogen interactions.

Keywords: Campylobacter jejuni; Citrobacter rodentium; Infectious diseases; Listeria monocytogenes; Salmonella Typhimurium; microbiota; mouse model.

Figure 1.

Experimental design. (a) Mice were infected with 10^9 CFU (C.j. ; C.r. ; L.m.), 5x10^7 CFU (S. Tm) or mock infected with phosphate buffer saline (PBS; vehicle control). The duration of infection was based on growth kinetics and disease severity for each pathogen (see methods). (b) The experiments were designed to allow comparison in two dimensions, between pathogens and also between OligoMM¹² and SPF mice.) (c) Number of mice in the different arms of the experiment. Infectious dose of three SPF mice in the C. rodentium group had a low infection dose and were thus excluded from the analyses in Figure 2. Mouse illustrations in this figure were purchased from VectorStock.

Figure 2.

Colonization of OligoMM¹² and SPF mice with the four pathogens. (a) Median (with range) of CFU per gram in feces of OligoMM¹² mice for the days post infection (gavage on day 0). (b) Median (with range) of CFU per gram in feces of SPF mice for the days post infection. (c) CFU per gram in the cecum content of OligoMM¹²on the last day of the experiment. (d) CFU per gram in the cecum content of SPF mice on the last day of the experiment. (e) CFU in the liver of OligoMM¹² mice on the last day of the experiment. (f) CFU in the liver of SPF mice on the last day of the experiment. Solid lines indicate median and dashed lines indicate the detection limit. The data was collected from at least two independent experiments per pathogen group and mouse model. (n of OligoMM¹² mice: PBS (8), S. Tm (13), C. rodentium (10), C. jejuni (10), L. monocytogenes (10); n of SPF mice: PBS (10), S. Tm (7), C. rodentium (5), C. jejuni (8), L. monocytogenes (6)).

Figure 3.

All four pathogens trigger inflammation in OligoMM¹² mice. (a) Light microscopy images of cecum tissue (H&E staining) of OligoMM¹² mice. (black bar = 100 µm; corresponding scores indicated with black filled data point in panel B) (b) total pathology score for cecum tissue (H&E staining) of OligoMM¹² and SPF mice. (mann-whitney test: p > .05 = ns, p < .05 = *, p < .01 = **, p < .001 = ***, p < .0001) (c) lipocalin-2 levels (as a marker for gut inflammation) in the feces of OligoMM¹² mice. (no data for L. monocytogenes on day 1 and 2) (d) lipocalin-2 levels (as a marker for gut inflammation) in the feces of SPF mice. (no data for L. monocytogenes on day 1 and 2) (e) correlation analysis of lipocalin-2 levels vs. pathogen densities in the feces of OligoMM¹²mice. Purple, green and red lines, highlighted with a shaded area, represent the linear model with the corresponding confidence interval. R² value and Pearson correlation were calculated and are shown on the corresponding graph.

Figure 4.

Host and microbiota response to pathogen infection in OligoMM¹² mice. (a) Number of significantly upregulated genes unique to and shared between pathogen groups. Every colored circle contains the total number of significantly upregulated genes in that pathogen group. The number of shared genes between the respective groups is shown wherever circles overlap. Triangles mark the number of genes that are unique to each pathogen group. The black star marks the number of genes that are significantly upregulated in all pathogen groups. (b) Top10 upregulated genes that are unique to the pathogen group. (c) Relative abundance of bacterial taxa in the feces of infected/mock treated (PBS) OligoMM¹² mice on day 3 and the last day of the experiment for the individual pathogen groups.

Figure 5.

Heatmap of metabolite abundance in the feces of OligoMM¹² and SPF mice relative to mock (PBS) treated mice. Color gradient indicates log2 fold change of average peak area of metabolites relative to PBS, and the stars indicate significance. (Wilcoxon test: p < .05 = *, p < .01 = **, p < .001 = ***, absence denotes non-significant values).