Akkermansia muciniphila and Parabacteroides distasonis synergistically protect from colitis by promoting ILC3 in the gut

Abstract

Inflammatory bowel disease (IBD) is a group of inflammatory conditions of the gastrointestinal tract. The etiology of IBD remains elusive, but the disease is suggested to arise from the interaction of environmental and genetic factors that trigger inadequate immune responses and inflammation in the intestine. The gut microbiome majorly contributes to disease as an environmental variable, and although some causative bacteria are identified, little is known about which specific members of the microbiome aid in the intestinal epithelial barrier function to protect from disease. While chemically inducing colitis in mice from two distinct animal facilities, we serendipitously found that mice in one facility showed remarkable resistance to disease development, which was associated with increased markers of epithelial barrier integrity. Importantly, we show that Akkermansia muciniphila and Parabacteroides distasonis were significantly increased in the microbiota of resistant mice. To causally connect these microbes to protection against disease, we colonized susceptible mice with the two bacterial species. Our results demonstrate that A. muciniphila and P. distasonis synergistically drive a protective effect in both acute and chronic models of colitis by boosting the frequency of type 3 innate lymphoid cells in the colon and by improving gut epithelial integrity. Altogether, our work reveals a combined effort of commensal microbes in offering protection against severe intestinal inflammation by shaping gut immunity and by enhancing intestinal epithelial barrier stability. Our study highlights the beneficial role of gut bacteria in dictating intestinal homeostasis, which is an important step toward employing microbiome-driven therapeutic approaches for IBD clinical management.

Importance: The contribution of the gut microbiome to the balance between homeostasis and inflammation is widely known. Nevertheless, the etiology of inflammatory bowel disease, which is known to be influenced by genetics, immune response, and environmental cues, remains unclear. Unlocking novel players involved in the dictation of a protective gut, namely, in the microbiota component, is therefore crucial to develop novel strategies to tackle IBD. Herein, we revealed a synergistic interaction between two commensal bacterial strains, Akkermansia muciniphila and Parabacteroides distasonis, which induce protection against both acute and chronic models of colitis induction, by enhancing epithelial barrier integrity and promoting group 3 innate lymphoid cells in the colonic mucosa. This study provides a novel insight on how commensal bacteria can beneficially act to promote intestinal homeostasis, which may open new avenues toward the use of microbiome-derived strategies to tackle IBD.

Keywords: Akkermansia muciniphila; ILC; Parabacteroides distasonis; colitis; gut immunity; microbiome.

Fig 1

Mice from different animal facilities display distinct susceptibility to colitis development. (A) C57BL/6 mice from two different animal houses were administered with 3% DSS in the drinking water and were monitored daily. (B) Disease progression was assessed by scoring the DAI throughout the experiment. (C) Representative colons were imaged, and colon length was measured at day 7, after excision. (D) Histological analysis of hematoxylin and eosin staining of mice prior and after colitis induction. (E) Colitis scores were obtained by the histological evaluation of colon samples at day 7. (F) Alcian blue/periodic acid-Schiff staining of the colonic tissues for goblet cells and mucus analysis. (G) Quantification of goblet cell numbers per crypt. For susceptible mice at day 7, no intact crypts were found. Data are presented as mean ± standard deviation. Statistically significant values are *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. DAI, disease activity index; DSS, dextran sulfate sodium; ND, not detected.

Fig 2

Resistant mice display alterations on epithelial barrier function and gut immunity. (A and B) Expression of Muc1, Muc2, Muc13 (A), and Cldn2, Cldn3, Cldn4, Cldn7, and Cdh1 (B) was analyzed by quantitative PCR in homeostatic conditions. (C) Intestinal permeability in homeostasis was measured after administration of FITC-dextran by oral gavage and quantified in the serum after 4 hours of administration. (D) The production of IL-10, IL-17, and IL-22 (pg cytokine/mg colon) was quantified in colonic extracts at homeostatic conditions. (E and F) Frequencies of Th17 cells (E) and ILC3 (F) in the gut of susceptible or resistant mice, under homeostatic conditions. For panels E and F , each dot corresponds to a pool of three mice. Data are presented as mean ± standard deviation. Statistically significant values are *P < 0.05, **P < 0.01; ****P < 0.0001. Th17, T helper 17.

Fig 3

FMT from resistant mice is able to prevent the development of acute colitis and avoid relapse in chronic colitis in susceptible mice. (A) Susceptible mice were treated with antibiotic for 5 weeks and then received fecal contents from resistant mice by oral gavage for 3 days. After 3 weeks, to allow colonization, mice were challenged with 3% DSS for 7 days. (B) Disease progression was assessed by scoring the DAI throughout the experiment. Images are representative of at least three independent experiments; n = 5 per group. (C) Susceptible mice were treated with 2% DSS for 5 days. After remission, mice received FMT from resistant mice by oral gavage for 5 days. The control group was treated with the vehicle (PBS). Two weeks later, both groups were given 2% DSS as previously mentioned. (D) Disease progression was assessed by scoring the DAI throughout the experiment. (E) Histological analysis of hematoxylin and eosin and alcian blue/periodic acid-Schiff stainings of the colonic tissues from mice that received FMT or PBS at 7 weeks of treatment. (F) Colitis scores were obtained by the histological evaluation of colon samples at week 7. (G) Quantification of goblet cell numbers per crypt. Images are representative of at least three independent experiments; n = 5 per group. Data are presented as mean ± standard deviation. Statistically significant values are *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. FMT, fecal microbiota transplant.

Fig 4

16S rRNA sequencing of fecal samples from resistant and susceptible mice revealed a distinct microbiota composition. (A) Principal coordinate analysis of susceptible, resistant, and susceptible + FMT mice regarding intestinal microbiota composition. Resistant vs susceptible: P = 0.001; resistant vs susceptible + FMT: P = 0.005; susceptible vs susceptible + FMT: P = 0.006. (B) Relative abundance of bacterial genera identified from DNA from stool samples of susceptible, resistant mice and susceptible + FMT mice; A. muciniphila and P. distasonis are marked by purple and beige arrows, respectively. (C) Number of operational taxonomic units and (D) diversity of species found in susceptible, resistant, and susceptible + FMT mice. (E) Absolute abundance quantification of A. muciniphila and (F) P. distasonis in susceptible, resistant, and susceptible + FMT mice. Images are representative of at least three independent experiments; n = 5/6 per group. Data are presented as mean ± standard deviation. Statistically significant values are *P < 0.05, **P < 0.01, ****P < 0.0001.

Fig 5

The combination of Akkermansia muciniphila and Parabacteroides distasonis attenuates colitis development. (A) Mice were supplemented with Am, Pd, or a combination of both (Am + Pd) for 12 days by daily oral gavage, followed by administration of 3% DSS for 7 days. (B) Disease progression was assessed by scoring the DAI throughout the experiment. (C–E) Quantification of goblet cell numbers per crypt after colitis induction, as well as colitis scores obtained by the histological evaluation of colon samples after DSS treatment. N = 5 per group. (F) A relapse-remission experiment was performed in which susceptible mice were subjected to colitis induction with 2% DSS for 5 days. After recovery, mice were supplemented with Am, Pd or the combination of both (Am + Pd) during 12 days by daily oral gavage, followed by a second cycle of colitis induction. Control mice (unsupplemented) received PBS as vehicle. (G) Disease progression was assessed by scoring the DAI. * in black corresponds to comparison between control and Am + Pd. * in blue corresponds to comparison between Am and Am + Pd. * in yellow corresponds to comparison between Pd and Am + Pd. $ in blue corresponds to comparison between control and Am. x corresponds to comparison between Am and Pd. (H) AUC was calculated based on the disease course upon colitis induction. (I–K) Quantification of goblet cell numbers per crypt after colitis induction, as well as colitis scores obtained by the histological evaluation of colon samples after DSS treatment. n = 5 per group. Data are presented as mean ± standard deviation. Statistically significant values are *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Am, Akkermansia muciniphila; AUC, area under the curve; Ctr, control; Pd, Parabacteroides distasonis.

Fig 6

Supplementation with Akkermansia muciniphila leads to increased frequencies of ILC3 in the colon and tissue integrity. (A–F) Frequencies of ILC3 (A), IL-17-producing ILC3 (B), IL-22-producing ILC3 (C), Th17 cells (D), IL-17-producing Th17 (E), and IL-22-producing Th17 (F) in the colon of control mice and mice supplemented with Am, Pd, or the combination of both (Am + Pd). (G) Correlation between the number of copies of Am with the frequencies of ILC3 in the colonic tissue. (H and I) Mice were supplemented with Am for 12 days by oral gavage and treated with anti-CD90 monoclonal antibody (Am + aCD90) at days −3, 0, and 3 of 3% DSS treatment or no antibody treatment (Am + vehicle). As experimental controls, an isotype control (Am + isotype) was given instead of anti-CD90, or no antibody treatment was performed (Am + vehicle). Mice were given 3% DSS, and colitis development was assessed by DAI. The control group is related to mice that were not supplemented or treated. (J) Absolute abundance of Akkermansia muciniphila, alone (Am) or in combination with Parabacteroides distasonis (Am + Pd) in the colon of supplemented mice. (K) Correlation between the number of copies of Am with the DAI. Data are presented as mean ± standard deviation. Statistically significant values are *P < 0.05, **P < 0.01, ***P < 0.001.