Human Intestinal Enteroid Monolayers as a Physiologically Relevant Model to Study Clostridium difficile Toxin Activity

Abstract

BackgroundClostridium difficile is a spore‐forming microbe that has emerged as an important nosocomial pathogen globally. C. difficile infection (CDI) involves alterations of the host microbiota, germination of C. difficile spores, and production of toxins A and B that are primarily responsible for the clinical manifestations. Both toxins bind cell surface receptors, are endocytosed, and inactivate GTP‐binding proteins. This leads to an influx of calcium (Ca2+) and collapse of the actin cytoskeleton, which can be assayed by cell rounding. Current C. difficile model systems have used immortalized cell lines to assess toxin‐mediated cell rounding. However, the biological relevance of these model systems is limited by variability in response to C. difficile toxins, making standardization of activity challenging. To date, no study has examined the distribution of known C. difficile toxin receptors in the human intestine or determined the sensitivity of the human epithelium to C. difficile toxins. We hypothesized that human intestinal enteroids (HIEs), as the most physiologically relevant in vitro model system, expresses the important native toxin receptors and provide an ideal model to dissect C. difficile toxin activity.Methods & ResultsWe generated biopsy‐derived HIE cultures stably expressing a GFP‐based calcium sensor (GCaMP6s) to monitor cytosolic Ca2+ and LifeAct‐Ruby labeling F‐actin to monitor actin cytoskeleton rearrangement by live cell imaging. For comparison we also generated stably expressing LifeAct‐Ruby and GCaMP6s cell lines, Hela, Vero, HT29 and T84. By qPCR we found that jejunal and colonic HIEs expressed far higher concentrations of the known toxin receptors compared to Vero cells. The highest expressing receptor for HIEs was the toxin B receptor FZD7 (~60,000 fold increase in jejunum and colon), followed by toxin A receptor GP96 (~500 fold increase in jejunum, ~3,500 fold increase in colon). Live imaging revealed toxins from pathogenic C. difficile strains elicited robust Ca2+ signaling in cell lines and HIEs, while non‐toxigenic strains did not. Increased intracellular Ca2+ correlated with actin cytoskeleton rearrangement in cell lines and HIEs. Interestingly, the HIEs were less sensitive to toxins derived from several C. difficile strains (R20291, 196, 630, M68, M929972) or purified toxins when compared to the cell lines, particularly the Vero and HeLa cells. HIE's were also tenfold more sensitive to toxin A (0.001 μg) than toxin B (0.01 μg).ConclusionsWe found that HIEs have high expression of toxin receptors, and yet decreased sensitivity to C. difficile toxins when compared to traditionally used cell lines. We speculate that this may reflect components that are present in the HIEs and absent in immortalized models, such as mucins and secreted peptides. Our work demonstrates the need for more robust models of C. difficile infection and that the use of HIEs provides a new biologically relevant system for dissecting the intricate signaling between toxins and the host epithelium.This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.