Microbial Ecology of Intestinal Microbiome Reassembly and Mucosal Immune Response Post Amoxicillin Treatment in the Presence and Absence of Candida albicans in a Murine Model

Abstract

The yeast Candida albicans is an indigenous member of the gastrointestinal (GI) tract of most healthy individuals; however, it can overgrow during antibiotic treatment. Broad spectrum antibiotics such as amoxicillin are prescribed frequently, but recovery from the dysbiosis caused by this therapy remains an active area of research. The first objective of my dissertation was to study the effect of C. albicans colonization on microbiome reassembly in C57BL/6 mice after amoxicillin treatment. Several studies have reported that the prototype C. albicans strain SC5314 does not colonize the intestinal tract long-term. The Huffnagle lab has isolated strain CHN1, capable of persistent colonization. Thus, we utilized these two strains of C. albicans from genomically distinct clades as a tool to model the effect of differences in C. albicans colonization dynamics. These experiments reveal that mice inoculated with strain SC5314 progressively cleared it from the GI tract over time and the bacterial community structure of these mice increasingly became more similar to the pre-antibiotic structure over time. This included GI tract levels of the lactic acid bacteria Lactobacillus johnsonii and Enterococcus faecalis, which quickly recovered to pre-antibiotic levels. Importantly, this recovery was associated with a loss of C. albicans colonization. In contrast, all mice inoculated with strain CHN1 maintained significantly elevated levels of colonization throughout the study. In CHN1-inoculated mice, the bacterial community structure never returned to its pre-antibiotic state and this dysbiosis accompanying stable C. albicans colonization was evident at phyla-level taxonomic changes. Colonization by strain CHN1 was associated with markedly lower L. johnsonii levels and elevated E. faecalis levels. Notably, I developed a molecular approach for differentiating between SC5314 and CHN1 by exploiting allelic differences in Ece1, a gene linked to hyphal formation. My second objective was to study the effect of transient oral amoxicillin therapy (1 week) in the presence and absence of C. albicans colonization on (a) intestinal microbiome reassembly dynamics and (b) host intestinal tissue gene expression in BALB/cJ mice. These studies revealed that amoxicillin treatment resulted in a significant initial change in bacterial community structure that began to return to baseline structure beginning about one week post-antibiotic therapy. 3 days post-amoxicillin treatment, the microbiota was dominated by E. faecalis and was associated with an upregulation of antimicrobial peptide (AMP) gene expression: Reg3b, Reg3g, Defa1, Defa20, and Defa28, as well as neutrophil-associated genes Cxcl1 and Il36g. In C. albicans¬-colonized mice, the recolonization dynamics of the bacterial microbiota was similar to that observed in amoxicillin-only treated mice, with a couple distinct exceptions. The presence of C. albicans triggered a response pathway characterized by the upregulation of the calprotectin genes, S100a8 and S100a9 that remained elevated through the end of the experiment. Our data also signaled the involvement of mast cells in controlling C. albicans, depicted by an upregulation of histamine receptors Hrh2 and Hrh4 and the mast cell protease Mcpt1 in the colon. Altogether, these data illustrate that recovery of an amoxicillin-disrupted community, in the absence or presence of C. albicans, takes at least two weeks and results in upregulation of host response pathways during this period of time.