Fusobacterium nucleatum Secretes Outer Membrane Vesicles and Promotes Intestinal Inflammation

Abstract

Multiple studies have implicated microbes in the development of inflammation, but the mechanisms remain unknown. Bacteria in the genus Fusobacterium have been identified in the intestinal mucosa of patients with digestive diseases; thus, we hypothesized that Fusobacterium nucleatum promotes intestinal inflammation. The addition of >50 kDa F. nucleatum conditioned media, which contain outer membrane vesicles (OMVs), to colonic epithelial cells stimulated secretion of the proinflammatory cytokines interleukin-8 (IL-8) and tumor necrosis factor (TNF). In addition, purified F. nucleatum OMVs, but not compounds <50 kDa, stimulated IL-8 and TNF production; which was decreased by pharmacological inhibition of Toll-like receptor 4 (TLR4). These effects were linked to downstream effectors p-ERK, p-CREB, and NF-κB. F. nucleatum >50-kDa compounds also stimulated TNF secretion, p-ERK, p-CREB, and NF-κB activation in human colonoid monolayers. In mice harboring a human microbiota, pretreatment with antibiotics and a single oral gavage of F. nucleatum resulted in inflammation. Compared to mice receiving vehicle control, mice treated with F. nucleatum showed disruption of the colonic architecture, with increased immune cell infiltration and depleted mucus layers. Analysis of mucosal gene expression revealed increased levels of proinflammatory cytokines (KC, TNF, IL-6, IFN-γ, and MCP-1) at day 3 and day 5 in F. nucleatum-treated mice compared to controls. These proinflammatory effects were absent in mice who received F. nucleatum without pretreatment with antibiotics, suggesting that an intact microbiome is protective against F. nucleatum-mediated immune responses. These data provide evidence that F. nucleatum promotes proinflammatory signaling cascades in the context of a depleted intestinal microbiome.IMPORTANCE Several studies have identified an increased abundance of Fusobacterium in the intestinal tracts of patients with colon cancer, liver cirrhosis, primary sclerosing cholangitis, gastroesophageal reflux disease, HIV infection, and alcoholism. However, the direct mechanism(s) of action of Fusobacterium on pathophysiological within the gastrointestinal tract is unclear. These studies have identified that F. nucleatum subsp. polymorphum releases outer membrane vesicles which activate TLR4 and NF-κB to stimulate proinflammatory signals in vitro Using mice harboring a human microbiome, we demonstrate that F. nucleatum can promote inflammation, an effect which required antibiotic-mediated alterations in the gut microbiome. Collectively, these results suggest a mechanism by which F. nucleatum may contribute to intestinal inflammation.

Keywords: Fusobacterium nucleatum; TLR4; enteroids; epithelium; inflammation; intestine; microbiome; organoids; outer membrane vesicles.

FIG 1

F. nucleatum subsp. polymorphum adheres to colonic MUC2 and secretes OMVs. (A) Representative images of T84 cells after incubation with fluorescently tagged F. nucleatum subsp. polymorphum counterstained with nuclear dye Hoechst (scale bar, 50 μm). (B) Representative image of fluorescently tagged F. nucleatum subsp. polymorphum adhered to purified MUC2 (scale bar, 50 μm). (C) TEM images of F. nucleatum (cross-section) with OMVs attached and surrounding the bacterium. Images on the right-hand side depict the various sizes of OMVs (scale bar, 200 nm). (D) Nanoparticle tracking analysis of F. nucleatum subsp. polymorphum OMVs.

FIG 2

F. nucleatum compounds and OMVs promote IL-8, TNF, NF-κB, and MAPK activation. (A) Measurement of IL-8 (pg/ml) by ELISA in HT29 cell supernatant after 16 h of incubation with 25% uninoculated BHIS (BHIS), 25% F. nucleatum BHIS conditioned medium <50-kDa fraction (<50 kDa), 25% F. nucleatum BHIS conditioned medium >50-kDa fraction (>50 kDa), or 5% purified F. nucleatum OMVs (OMVs) in DMEM in the absence or presence of TLR4 inhibitor CLI-095 (n = 6 replicates/experiment, repeated three independent times). (B) Measurement of TNF (pg/ml) by ELISA in HT29 supernatant after 16 h incubation with 25% uninoculated BHIS (BHIS), 25% F. nucleatum BHIS conditioned medium <50-kDa fraction, 25% F. nucleatum BHIS conditioned medium >50-kDa fraction, or 5% purified F. nucleatum OMVs (OMVs) in DMEM in the absence or presence of TLR4 inhibitor CLI-095 (n = 6 replicates/experiment, repeated three independent times). (C) Quantification of secreted luciferase in HT29 cells transfected with a pNFκB-MetLuc2-Reporter treated for 16 h (n = 9/experiment, repeated two independent times). (D) Western blot analysis of phosphorylated ERK, phosphorylated CREB, phosphorylated iκB, total iκB, and actin at 30 min incubation in HT29 cells (n = 3/experiment). Treatments are the same as in panels A, B, and C. Quantification of Western blots was performed using Fiji software. (E) Analysis of metabolic activity/viability in HT29 cells by resazurin assay (excitation, 560; emission, 600 nm). *, P < 0.05 (multi-way ANOVA).

FIG 3

F. nucleatum >50-kDa compounds promote TNF, NF-κB, and MAPK signaling in human colonoid monolayers. (A) Representative images of human colonoid monolayers treated with 25% BHIS (BHIS) or 25% F. nucleatum conditioned medium >50-kDa fraction (>50 kDa) in DMEM, 1× HEPES, 1× GlutaMAX, and pyruvate for 16 h (scale bar,100 μm). (B) Measurement of TNF (pg/ml) by ELISA in colonoid monolayers treated with 25% BHIS (BHIS) or 25% F. nucleatum BHIS conditioned medium >50-kDa fraction after 16 h incubation (n = 4 monolayers/experiment, repeated two independent times). (C) Quantification of secreted luciferase in human colonoid monolayer cells transfected with a pNFκB-MetLuc2-Reporter treated for 16 h (n = 4 monolayers/experiment). (D and E) Luminex Magpix multiplex analysis of phosphorylated ERK (D) and CREB (E) in human colonoid monolayers treated with 25% BHIS (BHIS) or 25% F. nucleatum conditioned medium >50-kDa fraction for 1 h (n = 3 monolayers/experiment). *, P < 0.05 (Student t test).

FIG 4

F. nucleatum subsp. polymorphum is unable to promote inflammation in the presence of a complete gut microbiome. (A) Representative images of H&E stains of control animals and F. nucleatum subsp. polymorphum-treated animals at day 3 and day 5 postinfection (scale bar, 100 μm). (B) FISH staining of Fusobacterium (red) counterstained with MUC2 (yellow) and Hoechst (blue) at day 3 and day 5 postinfection (scale bar, 100 μm). (C) Analysis of mouse weights at days 1, 2, 3, and 5 postinfection (n = 6/group). *, P < 0.05 (repeated-measures ANOVA). (D) Colonic mRNA expression of proinflammatory related genes on day 3 postinfection (n = 6/group). *, P < 0.05 (two-way ANOVA). (E) Colonic mRNA expression of proinflammatory related genes on day 5 postinfection (n = 6/group). *, P < 0.05 (two-way ANOVA).

FIG 5

F. nucleatum subsp. polymorphum drives inflammation in vivo with antibiotic disruption of the gut microbiome. (A) Representative images of H&E stains of control animals and F. nucleatum subsp. polymorphum-treated animals who received antibiotics at day 3 and day 5 postinfection (scale bar, 100 μm). Blue arrows highlight immune infiltration. (B) FISH staining of Fusobacterium (red) counterstained with MUC2 (yellow) and Hoechst (blue) at day 3 and day 5 postinfection. Enlarged insets demonstrate Fusobacterium at day 3 and 5 (scale bar, 100 μm). (C) Analysis of mouse weights at days 1, 2, 3, and 5 postinfection (n = 6/group). *, P < 0.05 (repeated-measures ANOVA). (D) Colonic mRNA expression of proinflammatory related genes on day 3 postinfection (n = 6/group). *, P < 0.05 (two-way ANOVA). (E) Colonic mRNA expression of proinflammatory related genes on day 5 postinfection (n = 6/group). *, P < 0.05 (two-way ANOVA).

FIG 6

Proposed model for F. nucleatum subsp. polymorphum-driven inflammation. F. nucleatum adheres the intestinal mucus layer in the setting of an altered gut microbiome, in which it delivers secreted compounds as “cargo” in OMVs. The OMVs activate epithelial TLRs, including TLR4, which promotes phosphorylation and activation of ERK, CREB, and NF-κB, thereby driving the production of proinflammatory cytokines and initiating inflammation.