Extracellular vesicles produced by the human commensal gut bacterium Bacteroides thetaiotaomicron affect host immune pathways in a cell-type specific manner that are altered in inflammatory bowel disease

Abstract

The gastrointestinal (GI) tract harbours a complex microbial community, which contributes to its homeostasis. A disrupted microbiome can cause GI-related diseases, including inflammatory bowel disease (IBD), therefore identifying host-microbe interactions is crucial for better understanding gut health. Bacterial extracellular vesicles (BEVs), released into the gut lumen, can cross the mucus layer and access underlying immune cells. To study BEV-host interactions, we examined the influence of BEVs generated by the gut commensal bacterium, Bacteroides thetaiotaomicron, on host immune cells. Single-cell RNA sequencing data and host-microbe protein-protein interaction networks were used to predict the effect of BEVs on dendritic cells, macrophages and monocytes focusing on the Toll-like receptor (TLR) pathway. We identified biological processes affected in each immune cell type and cell-type specific processes including myeloid cell differentiation. TLR pathway analysis highlighted that BEV targets differ among cells and between the same cells in healthy versus disease (ulcerative colitis) conditions. The in silico findings were validated in BEV-monocyte co-cultures demonstrating the requirement for TLR4 and Toll-interleukin-1 receptor domain-containing adaptor protein (TIRAP) in BEV-elicited NF-kB activation. This study demonstrates that both cell-type and health status influence BEV-host communication. The results and the pipeline could facilitate BEV-based therapies for the treatment of IBD.

Keywords: extracellular vesicles; host-microbe interactions; single-cell data analysis; toll-like receptor pathway; ulcerative colitis.

FIGURE 1

Computational workflow to analyse cell‐type specific effects of BEVs. Numbers indicate the sequence of the main steps: 1, Extraction of BEV proteins from the proteomic dataset 2, Processing the raw single‐cell transcriptomics from human colon 3/a, Creating cell‐type specific network using protein‐protein interactions from OmniPath (Türei et al., 2016) 3/b, Predicting protein‐protein interactions (PPIs) between BEV and host proteins in each cell‐type separately 4, Reconstruction of Toll‐like receptor pathway using Reactome database (Jassal et al., 2020) 5, Combining cell‐specific signalling with BEV targeted human proteins

FIGURE 2

Interactions of 48 BEV proteins with monocytes, macrophages and dendritic cells in healthy (a) and UC (b) conditions. Number of expressed genes/number of interacting proteins are highlighted for each cell‐type

FIGURE 3

Overlap of biological processes over‐represented in the BEV‐host interactomes corresponding to cell‐types in healthy (a) and uninflamed UC (b) conditions

FIGURE 4

TLR pathway in DCs. Edges between nodes represent protein‐protein interactions. Figures have been created with Cytoscape (Shannon et al., 2003)

FIGURE 5

TLR pathway in monocytes. Edges between nodes represent protein‐protein interactions. Figures have been created with Cytoscape (Shannon et al., 2003)

FIGURE 6

TLR pathway in THP‐1 monocytes (based on bulk transcriptomic datasets). Edges between nodes represent protein‐protein interactions. Figures have been created with Cytoscape (Shannon et al., 2003)

FIGURE 7

TLR pathway in macrophages. Edges between nodes represent protein‐protein interactions. Figures have been created with Cytoscape (Shannon et al., 2003)

FIGURE 8

Inhibition of TLR4 and TIRAP signalling pathway abrogates THP1‐Blue cells activation by Bt BEVs. NF‐κB activation was assessed using different doses of BEVs in 5 × 10⁵ THP1‐Blue cells/ml in the presence or absence of the TLR4 inhibitor CLI‐095 (a) or TIRAP inhibitor (b) and by measuring absorbance at 620 nm after incubation with the colorimetric assay reagent Quanti‐Blue.LPS from E. coli was used as a positive control and PBS as a negative control. Data are presented as mean ± SD (n = 9). Significant differences were determined by using two‐way ANOVA followed by Bonferroni's multiple comparison post hoc test. ** (P  <  0.01), **** (P  <  0.0001)