Optimising the induction of inflammation within preterm infant-derived intestinal epithelial organoids

Abstract

Preterm infants born <32 weeks gestation have abnormal microbial colonisation and dysregulated inflammation within the gut. Preterm infant-derived intestinal organoids (PIOs) represent a valuable model for investigating gut microbiome-host interactions and inflammatory responses. We optimised an inflammation model in PIO monolayers incubated within an anaerobic co-culture system that recreates the physiological oxygen gradient of the intestinal epithelium. We trialled multiple stimuli, including live and heat-killed pathobiont consortia, lipopolysaccharide (LPS) and flagellin. We found that a combination of apical LPS and basolateral flagellin, incubated for 3 h, elicited the most robust response. This was characterised by enhanced pro-inflammatory cytokine secretion, the potential for chemokine-driven immune recruitment, TNFα and IL17C pathway signalling, shifts from NF-κB to AP-1-mediated responses, and signs of tissue remodelling. This provides a framework for appropriate study design to disentangle the impacts of microbiome-host interactions in health and disease using intestinal organoids.

Keywords: Gastrointestinal diseases; Gastrointestinal system; Inflammation; Intestinal stem cells.

Fig. 1. Comparisons of the effects of incubating preterm infant-derived intestinal epithelial organoids (PIOs) with purified bacterial stimuli (flagellin (100 μg/ml) and/or lipopolysaccharide (LPS) (100 μg/ml)), a cocktail of live bacterial pathobionts, or heat-killed bacterial pathobionts, for either 24 h or 3 h.

A Schematic of PIO inflammation assays. B Bar plots showing the changes in both apical and basolateral IL8 secretion from PIOs, for both the 24 h and 3 h incubations. Within each plot, conditions with the same letters are not significantly different. C Principal component analysis of PIO proteomics data for both the 24 h and 3 h incubations. Groups are coloured by condition. D Per condition comparison of the number of differentially abundant proteins identified within PIO proteomics data for both incubation durations. Thresholds for differential abundance were set at adjusted p ≤ 0.05 and log2(foldchange) ≥ +/− 0.58 (real term fold change of +/−1.5).

Fig. 2. Comparisons of the effects of incubating preterm infant-derived intestinal epithelial organoids (PIOs) with either flagellin (100 ng/ml) or lipopolysaccharide (LPS) (100 ng/ml), added either apically or basolaterally, for 6 h, and a 3 h incubation with basolateral flagellin and apical LPS (both 100 ng/ml).

A Schematic of PIO 6 h inflammation assay. B Bar plots showing the changes in both apical and basolateral IL8 secretion from PIOs, following 6 h incubation. Within each plot, conditions with the same letters are not significantly different. C Schematic of PIO 3 h inflammation assay. D Bar plots showing the changes in both apical and basolateral IL8 secretion from PIOs, following 3 h incubation. E Volcano plot showing changes in the proteomes of PIOs following 3 h incubation with apical LPS and basolateral flagellin versus untreated PIOs. The thresholds used for differential abundance are marked with dashed lines (adjusted p ≤ 0.05 and log2(foldchange) ≥ +/− 0.58). Differentially abundant proteins of interest are labelled on the plot, with blue signifying downregulation, and red upregulation. F Heatmap showing abundances of proteins from the 3 h incubation experiment, excluded from Limma fold change analysis due to being undetected in one or other condition. G KEGG pathways that were significantly enriched within PIO proteins upregulated during treatment with inflammatory stimuli for 3 h. H Reactome pathways that were significantly enriched within PIO proteins downregulated during treatment with inflammatory stimuli for 3 h.