Roseburia intestinalis-derived extracellular vesicles ameliorate colitis by modulating intestinal barrier, microbiome, and inflammatory responses

Abstract

Inflammatory bowel disease (IBD) is a chronic disorder characterized by recurrent gastrointestinal inflammation, lacking a precise aetiology and definitive cure. The gut microbiome is vital in preventing and treating IBD due to its various physiological functions. In the interplay between the gut microbiome and human health, extracellular vesicles secreted by gut bacteria (BEVs) are key mediators. Herein, we explore the role of Roseburia intestinalis (R)-derived EVs (R-EVs) as potent anti-inflammatory mediators in treating dextran sulfate sodium-induced colitis. R was selected as an optimal BEV producer for IBD treatment through ANCOM analysis. R-EVs with a 76 nm diameter were isolated from R using a tangential flow filtration system. Orally administered R-EVs effectively accumulated in inflamed colonic tissues and increased the abundance of Bifidobacterium on microbial changes, inhibiting colonic inflammation and prompting intestinal recovery. Due to the presence of Ile-Pro-Ile in the vesicular structure, R-EVs reduced the DPP4 activity in inflamed colonic tissue and increased the active GLP-1, thereby downregulating the NFκB and STAT3 via the PI3K pathway. Our results shed light on the impact of BEVs on intestinal recovery and gut microbiome alteration in treating IBD.

Keywords: Ile‐Pro‐Ile; R. intestinalis; bacterial extracellular vesicle; inflammatory bowel disease; microbiome.

FIGURE 1

R‐EVs exert amicable alterations in dysbiosis, reduce the inflammatory response, and restore the intestinal barrier in IBD. (a) Schematic illustration of Roseburia‐derived extracellular vesicles (R‐EVs) as a critical mediator for DSS‐induced IBD treatment. (b) Species differences between IBD and nonIBD women in the Integrative Human Microbiome Project (634 IBD and 194 nonIBD stool samples). (c) Percentage in female stool samples.

FIGURE 2

R‐EV isolation and its differential protein expression. (a) Schematic illustration of tangential flow filtration (TFF)‐based isolation of R‐EVs. (b) Ultrathin‐sectioned transmission electron microscopy (TEM) and cryo‐TEM images of R.intestinalis (R) and R‐EVs (scale bar = 100 nm). Insets are magnified images (scale bar = 50 nm). (c) Size and surface charge of R‐EV. (d) The R‐EVs were analyzed for lipoteichoic acid content (Gram‐positive bacterial component) using ELISA (n = 5). (e) Volcano plot showing differentially expressed proteins in R and R‐EV. Proteins with log₂FC(R‐EV/R) ≤−1 or ≥1 and p < 0.01 (Welch t‐test) are coloured blue (R‐enriched) or red (R‐EV‐enriched). (f) Categorical composition of Clusters of Orthologous Groups (COG) based on the relative abundance of proteins assigned to COGs. (g) Bar plot showing differential abundance of COG categories in R and R‐EV. Those enriched in R and R‐EV are coloured blue and red, respectively. (h) Distinct COGs (q < 0.05) of significantly enriched COG categories (G: ‘Carbohydrate metabolism and transport’ and I: ‘Lipid metabolism’, q < 0.01) in R‐EV. Relative abundance of each COG was standardized into z‐score (coloured blue to red based on intensity). *p < 0.05 and **p < 0.01 versus the R‐treated group.

FIGURE 3

Orally administered R‐EVs effectively alleviate DSS‐induced colitis in the preventive mice model. (a) Experimental scheme for DSS‐induced colitis and R‐EV administration. (b) Representative images of colon sections stained with hematoxylin and eosin (H&E) and Alcian blue (AB) for histology assessment. (c) Colonic damage score, (d) total histological score, and (e) colonic goblet cell numbers per crypt on day 10. (f) Effect of R‐EVs on colon length. Mice were sacrificed on day 10, and colon lengths were measured and imaged. (g) DAI scores on day 10 and (h) daily changes in body weight of the R‐EV‐treated group over 10 days. (i) Endotoxin levels, serum nitric oxide (NO), IL‐6, and TNF‐α were examined by ELISA. (j) Distal colon was analyzed for mRNA expression of IL‐1β, REG3γ, Muc2, and ZO‐1. Data expressed as mean ± standard error of the mean (SEM) (n = 5). *p < 0.05 and **p < 0.01 versus the DSS‐treated control group. ‡p < 0.05 versus the R‐treated group.

FIGURE 4

R‐EV shapes gut microbiota in a DSS‐induced colitis mouse model through the growth of Bifidobacterium. (a) Bar plots of the phylum taxonomic levels in each group and (b) Bray‐Curtis‐based non‐metric multidimensional scaling (NMDS) plot of the cecal microbiota of each group from the prevention model. Relative abundance is plotted for each sample. (c) Alpha diversity (Chao1 index, Simpson diversity index, and Shannon diversity index) in each group (n = 8–12). (d) LDA score is computed from features differentially abundant between groups. The criteria for feature selection is an LDA score > 3.5. (e) Plots of differentially abundant microbiota at a genus level significantly enriched in each group (n = 8–12). Data expressed as mean ± SEM. *p < 0.05 and **p < 0.01 versus the DSS‐treated control group. ‡p < 0.05 versus the R‐treated group. (f) Schematic design of synthetic microbiome model using 16 core gut microbiota for R‐EV treatment. (g) Average relative abundance of the gut microbiota at the species level in each group from the synthetic microbiome model. (h) Heatmap representing the relative ratio of 16 microbiota in each group compared to the non‐treated group. (i) Biofilm formation of two Bifidobacterium strains (B. breve and B. longum subsp. infantis) using two different culture broths (Gifu Anaerobic Medium [GAM] or Bryant and Burkey [BB]) in response to R‐EV and several concentrations of Ile‐Pro‐Ile. Data represent relative biofilm formation (%) compared to the control group (n = 5) and are expressed as the mean ± SEM. *p < 0.05 and **p < 0.01 versus the R‐EV‐treated group.

FIGURE 5

Orally administered R‐EVs effectively alleviate DSS‐induced colitis in the therapeutic mice model. (a) Experimental scheme for administering R‐EV in the DSS‐induced therapy mouse model. (b) Representative images of colon sections stained with H&E and AB for histology assessment. (c) Colonic damage score, (d) total histological score, (e) colonic goblet cell numbers per crypt on day 10. (f) Effect of R‐EV_low on colon length. Mice were sacrificed on day 10, and colon lengths were measured and imaged. (g) DAI scores on day 10 and (h) daily changes in body weight of the R‐EV_low‐treated group over 10 days. (i) Survival rate and (j) NO, IL‐6, and TNF‐α serum levels in the R‐EV treated group. Data expressed as the mean ± SEM. (n = 5) *p < 0.05 and **p < 0.01 versus the DSS‐treated control group. ‡p < 0.05 and ‡‡p < 0.01 versus the R‐treated group. Effect of R‐EV on (k) dipeptidyl peptidase 4 (DPP4) activity and (l) plasma active glucagon‐like peptide‐1 (GLP‐1) in colon tissues of DSS‐induced colitis mouse. Data are expressed as mean ± SEM (n = 5). #p < 0.05 versus the control group. *p < 0.05 and **p < 0.01 versus the DSS group. (m) Western blot analysis and (n) quantification of the expression level of GLP‐1 and inflammation marker in colonic tissue of DSS‐induced colitis mouse. Data are expressed as mean ± SEM (n = 4). #p < 0.05 versus the control group. *p < 0.05 and **p < 0.01 versus the DSS group.

FIGURE 6

Biodistribution of Cy7‐labelled R‐EV in normal and DSS‐induced colitis mice. (a) In vivo fluorescence images of the Cy7‐labelled R‐EV (Cy7‐R‐EV) in a DSS‐induced mouse model. (b) Fluorescence intensity of the R‐EV in the entire body of the DSS‐induced colitis mouse as a function of time. (c) Ex vivo organ imaging of Cy7‐R‐EV at 3 and 12 h post‐injection. (d) Quantification of fluorescence intensity in major organs of normal and colitis models. (e) Serum fluorescence intensity of R‐EV in normal and colitis mouse models at 3 h post‐injection. Data are expressed as mean ± standard deviation (SD) (n = 3). **p < 0.01 versus the R‐EV (normal) group.