Flagellin-elicited adaptive immunity suppresses flagellated microbiota and vaccinates against chronic inflammatory diseases

Abstract

Alterations in gut microbiota composition are associated with metabolic syndrome and chronic inflammatory diseases such as inflammatory bowel disease. One feature of inflammation-associated gut microbiotas is enrichment of motile bacteria, which can facilitate microbiota encroachment into the mucosa and activate pro-inflammatory gene expression. Here, we set out to investigate whether elicitation of mucosal anti-flagellin antibodies by direct administration of purified flagellin might serve as a general vaccine against subsequent development of chronic gut inflammation. We show, in mice, that repeated injection of flagellin elicits increases in fecal anti-flagellin IgA and alterations in microbiota composition, reduces fecal flagellin concentration, prevents microbiota encroachment, protects against IL-10 deficiency-induced colitis, and ameliorates diet-induced obesity. Flagellin's impact on the microbiota is B-lymphocyte dependent and, in humans, obese subjects exhibit increased levels of fecal flagellin and reduced levels of fecal flagellin-specific IgA, relative to normal weight subjects. Thus, administration of flagellin, and perhaps other pathobiont antigens, may confer some protection against chronic inflammatory diseases.

Fig. 1

Systemic flagellin administrations elicit systemic and mucosal antibodies to flagellin. a 4-week old C57BL/6 J mice, wild type and TLR5/NLRC4 DKO, were purchased from The Jackson Laboratory and housed for 2 weeks before procedure in order to favor microbiota stabilization. Subsequently, flagellin (10 μg per mouse) was administered by intraperitoneal injections weekly for 9 weeks, whereas control mice received vehicle (PBS). Serum collection occurred on days −14, 0, 28, 56, and 63. Body weight measurements and fecal collection occurred prior to every flagellin administration. b–c Serum anti-flagellin IgA and IgG throughout the experiment, d–e fecal anti-flagellin IgA and IgG, f–g serum anti-flagellin IgA and IgG at day 56, h serum interleukin-6, and i CXCL1 at day 56 were analyzed using ELISA kits. j–k Fecal anti-flagellin IgA j and IgG k were also quantified up to 11 weeks after the final flagellin administration in mice receiving 6 weekly intraperitoneal injections of flagellin (10 μg per mouse). Data are the means ± S.E.M. Significance was determined using t test (**p ≤ 0.01 ***p ≤ 0.001 ****p ≤ 0.0001, n.s. indicates non-significant) or using one-way ANOVA corrected for multiple comparisons with a Bonferroni test (^#p ≤ 0.05 ^##p ≤ 0.01 ^###p ≤ 0.001 ^####p ≤ 0.0001, n.s. indicates non-significant). (N=4–5 mice from one out of three representative experiment). Source data are provided as a Source Data file.

Fig. 2

Flagellin administration alters the intestinal microbiota towards a lower pro-inflammatory state. 4-week old C57BL/6 J Wild Type mice were purchased from The Jackson Laboratory and housed for two weeks before procedure in order to favor microbiota stabilization. Subsequently, flagellin (10 μg per mouse) was administered by intraperitoneal injections weekly for 9 weeks, whereas control mice received vehicle (PBS). Fecal microbiota composition was analyzed using Illumina sequencing of the V4 region of 16 S rRNA genes. a–b Principal coordinates analysis (PCoA) of the unweighted UniFrac distance matrix at a day −14 and b day 56 (post stabilization, post immunization). c LEfSe analysis was performed in order to investigate microbiota taxa that were significantly altered by immunization at day 56 (post stabilization, post immunization), with green and red colors highlighting taxa significantly more abundant in PBS- and flagellin-treated mice, respectively. d Percentage of IgA^±-coated bacteria in PBS- and FliC-treated mice, wherein the IgA⁻ and IgA⁺ gates were determined follows appropriate SSC-A/FSC-A gating of SytoBC⁺ cells in wild-type and μMT mice. e Principal coordinates analysis (PCoA) of the unweighted UniFrac distance matrix of IgA-coated bacteria. f Alpha diversity rarefaction using the Chao1 index of IgA-coated bacteria. g Taxa summarization of IgA-coated bacteria. In a and b, categories were compared and statistical significance of clustering were determined via Permanova. Data are the means ±S.E.M. Significance was determined using t test (*p ≤ 0.05). (N=4–5 mice from one out of three representative experiment). Source data are provided as a Source Data file.

Fig. 3

Flagellin administration alters the intestinal microbiota toward a lower pro-inflammatory state. a Fecal pro-inflammatory potential was analyzed using HEK 293 cells expressing mTLR5 or mTLR4 measuring bioactive flagellin and lipopolysaccharide, respectively. b Colonic myeloperoxidase quantification of 4-week old, wild-type C57BL/6 J mice after receiving either vehicle or 10 μg of flagellin by intraperitoneal injections weekly for 9 weeks. c–f Colonic microbiota localization analysis of wild type and μMT mice treated with PBS, Salmonella-derived flagellin, or Bacillus-derived flagellin. c, e Confocal microscopy analysis of colonic microbiota localization; Muc2 (green), actin (purple), bacteria (red), and DNA (blue). d, f Distances of closest bacteria to colonic intestinal epithelial cells (IEC) per condition over 2–3 high-powered fields per mouse. g Fecal bacterial load determined by qPCR analysis of 16 S bacterial DNA in the fecal contents of mice treated with PBS or flagellin. Data are the means ±S.E.M. Significance was determined using t test (*p ≤ 0.05 **p ≤ 0.01 ***p ≤ 0.001 ****p ≤ 0.0001, n.s. indicates non-significant). (N=4–5 mice from one out of three representative experiment). Source data are provided as a Source Data file.

Fig. 4

Flagellin administrations protect against immune dysregulation-induced colitis. 4–8-week old C57BL/6 J wild-type and μMT mice received either vehicle or 10 μg of flagellin by intraperitoneal injections weekly for 9 weeks. Subsequently, animals were treated weekly for 4 weeks by 1 mg of anti-IL-10R antibody intraperitoneally to induce intestinal inflammation. Biometric data of Wild Type animals represented by a body weight, b adipose weight, c spleen weight, d colon weight, e colon length, f and colon weight/length ratio. g Colonic myeloperoxidase levels. h Colon pathohistological scoring. i–j Serum interleukin-6 and CXCL1 following anti-IL-10R antibody regimen. Severity of colitis in μMT animal represented by k colon pathohistological scoring and l colon length. m–o Principal coordinate analysis of the Bray–Curtis distance using a matrix containing all the morphometric and molecular parameters presented in a–l. Data are the means ± S.E.M. Significance was determined using t test (*p ≤ 0.05 **p ≤ 0.01 ****p ≤ 0.0001) or using one-way ANOVA corrected for multiple comparisons with a Bonferroni test (#p ≤ 0.05 ^##p ≤ 0.01 ^###p ≤ 0.001 ^####p ≤ 0.0001, n.s. indicates non-significant). (N=4–5 mice from one out of two representative experiment). Source data are provided as a Source Data file.

Fig. 5

Beneficial effects of flagellin immunization are abolished in TCRβ KO mice. 4-week old C57BL/6 J TCRβ KO mice were purchased from The Jackson Laboratory and housed for 2 weeks before procedure in order to favor microbiota stabilization. Next, flagellin (10 μg per mouse) was administered by intraperitoneal injections weekly for 9 weeks, whereas control mice received vehicle (PBS). Subsequently, animals were treated weekly for 4 weeks by 1 mg of anti-IL-10R antibody intraperitoneally to induce intestinal inflammation. a–b Fecal anti-flagellin IgA and IgG quantified using ELISA. c Confocal microscopy analysis of colonic microbiota localization; Muc2 (green), actin (purple), bacteria (red), and DNA (blue). d Distances of closest bacteria to colonic intestinal epithelial cells (IEC) per condition over 2–3 high-powered fields per mouse. e–f Fecal flagellin and LPS quantified using HEK 293 cells expressing mTLR5 or mTLR4. g Body weight, h spleen weight, i colon length, j colon weight, and k colon pathohistological scoring. Data are the means ±S.E.M. Significance was determined using t test (*p ≤ 0.05, n.s. indicates non-significant). (N=4–5). Source data are provided as a Source Data file.

Fig. 6

Specificity of the beneficial effects induced by flagellin immunization. 4-week old C57BL/6 J, wild-type mice, were purchased from The Jackson Laboratory and housed for 2 weeks before procedure in order to favor microbiota stabilization. Subsequently, mice were treated with either flagellin (10 μg per mouse), TNF-α (50 μg/kg body weight), or Poly (I:C) (10 μg/kg body weight) via intraperitoneal injections weekly for 9 weeks, whereas control mice received vehicle (PBS). Subsequently, animals were treated weekly for 4 weeks by 1 mg of anti-IL-10R antibody intraperitoneally to induce intestinal inflammation. a–b Fecal anti-flagellin IgA and IgG quantified using ELISA. c Confocal microscopy analysis of colonic microbiota localization; Muc2 (green), actin (purple), bacteria (red), and DNA (blue). d Distances of closest bacteria to colonic intestinal epithelial cells (IEC) per condition over 2–3 high-powered fields per mouse. e–f Fecal flagellin and LPS quantified using HEK 293 cells expressing mTLR5 or mTLR4. g Body weight, h spleen weight, i colon length, and j colon weight. Colitis severity was assessed by k fecal lipocalin-2 concentration and l colon pathohistological scoring. Data are the means ±S.E.M. Significance was determined using t test (*p ≤ 0.05 **p ≤ 0.01 ***p ≤ 0.001 ****p ≤ 0.0001, n.s. indicates non-significant). (N=4–5 mice). Source data are provided as a Source Data file.

Fig. 7

Flagellin immunization protected against high-fat diet-induced obesity. a Flagellin load (y axis) inversely correlate with anti-flagellin IgA concentration (x axis) in humans. R2 represents the coefficient of determination. b Mean concentration ± S.E.M. of flagellin load and anti-flagellin IgA concentration for human subjects segregated by their BMI to normal (18.5–24.9, N = 17), overweight (25–29.9, N = 11) or obese (> 30, N = 15). Flagellin concentration is denoted on the left y axis and anti-flagellin IgA concentration is denoted on the right y axis. ^#P < 0.05; one-way analysis of variance (ANOVA). c–j 8-week old, C57BL/6 J mice were purchased from The Jackson Laboratory and housed for two weeks before procedure in order to favor microbiota stabilization. Subsequently, flagellin (10 μg per mouse) was administered by intraperitoneal injections weekly for 9 weeks, whereas control mice received vehicle (PBS). Subsequently, animals were treated with high-fat diet (60% kcal from fat) for 4 weeks. c Fecal flagellin quantified using HEK 293 cells expressing mTLR5. d Body weights were measured weekly and expressed as relative values, day 63 (post immunization, pre high-fat diet treatment) being define as 100%. e Adipose weight. f Colon length. g Spleen weight. h–j Day 63 fecal anti-flagellin IgA correlated with adipose and spleen weights, as well as, colon length. Data are the means ± S.E.M. Significance was determined using linear regression analysis (for h, p values shown), t test (for c), or one-way ANOVA corrected for multiple comparisons with a Bonferroni test (^#p ≤ 0.05 ^##p ≤ 0.01 ^####p ≤ 0.0001, n.s. indicates non-significant) for e–g. (N=4–5 mice from one out of two representative experiment). Source data are provided as a Source Data file.