Th17-to-Tfh plasticity during periodontitis limits disease pathology

Abstract

Th17 cell plasticity is crucial for development of autoinflammatory disease pathology. Periodontitis is a prevalent inflammatory disease where Th17 cells mediate key pathological roles, yet whether they exhibit any functional plasticity remains unexplored. We found that during periodontitis, gingival IL-17 fate-mapped T cells still predominantly produce IL-17A, with little diversification of cytokine production. However, plasticity of IL-17 fate-mapped cells did occur during periodontitis, but in the gingiva draining lymph node. Here, some Th17 cells acquired features of Tfh cells, a functional plasticity that was dependent on IL-6. Notably, Th17-to-Tfh diversification was important to limit periodontitis pathology. Preventing Th17-to-Tfh plasticity resulted in elevated periodontal bone loss that was not simply due to increased proportions of conventional Th17 cells. Instead, loss of Th17-to-Tfh cells resulted in reduced IgG levels within the oral cavity and a failure to restrict the biomass of the oral commensal community. Thus, our data identify a novel protective function for a subset of otherwise pathogenic Th17 cells during periodontitis.

Graphical abstract

Figure 1.

Th17 cells continue to produce high levels of IL-17A during LIP. (A and B) LIP was induced in IL-17^KO mice (Il17a^Cre-HOMR26R^eYFP; gray bars) or littermate controls (white bars). Bone loss was examined 10 days after ligature placement. Graphs show changes in bone heights at defined sites (A) or in total (B) determined by subtracting the CEJ-ABC distance of LIP molars from CEJ-ABC distances of un-ligated molars. Data are pooled from two separate experiments with two to four mice per group. (C–H) LIP was induced in Il17a^CreR26R^eYFP mice and eYFP⁺CD4⁺ T cells examined in the gingiva (C–E) and dLN (F–H) 10 days after ligature placement. (C) Graph shows the proportions of CD4⁺ T cells that are eYFP⁺ in the gingiva of Il17a^CreR26R^eYFP naive control (white bar) and LIP (black bar) mice. (D) Graph shows the proportion of cytokine⁺ eYFP⁺ CD4 T cells in the gingiva of LIP mice. (E) Representative histogram showing RORγt expression in CD4⁺eYFP⁺ T cells in the gingiva of naive (blue histogram) and LIP (gray histogram) mice; fluorescence minus one (FMO; light gray histogram). (F) Graph shows the proportion of CD4⁺ T cells that are eYFP⁺ in the dLN of Il17a^CreR26R^eYFP naive control (white bar) and LIP (black bar) mice. (G) Graph shows the proportion of cytokine⁺ eYFP⁺ CD4 T cells in the dLN of LIP mice. (H) Representative histogram showing RORγt expression in CD4⁺eYFP⁺ T cells in the dLN of naive (blue histogram) and LIP (gray histogram) mice; FMO (light gray histogram). Data are pooled from three to six experiments with two to four mice per group. *P < 0.05, **P < 0.01, ****P < 0.0001 as determined by unpaired Student’s t test or Mann–Whitney test (C only). Results are expressed as mean ± SD or median ± interquartile range (C only).

Figure S1.

Th17-to-Tfh plasticity occurs in the dLN during LIP. (A) Representative FACS plots gated on eYFP⁺CD4⁺ T cells showing exemplar cytokine staining following ex vivo restimulation of cells. (B) Proportion of eYFP⁺ CD4⁺ T cells in the Peyer’s patches exhibiting a Tfh (CXCR5⁺PD-1⁺) phenotype in Il17a^CreR26R^eYFP mice with LIP (black circles) or naive controls (open circles). (C) Graph shows proportion of eYFP⁺ CD4⁺ T cells exhibiting a Tfh (CXCR5⁺PD-1⁺) phenotype in the spleen (SPL), mesenteric lymph nodes (MLN), and inguinal lymph nodes (iLN) of Il17a^CreR26R^eYFP mice with LIP or naive controls. (D and E) Representative FACS plots gated on eYFP⁺Tfh⁺ and eYFP⁺Tfh⁻ cells showing CCR6 (D) and RORγt and Bcl6 (E) staining. Data are pooled from three to six separate experiments with two to four mice per group. Results are expressed as mean ± SD.

Figure 2.

Th17 cells exhibit a Tfh phenotype in the dLN during LIP. LIP was induced in Il17a^CreR26R^eYFP mice and eYFP⁺CD4⁺ T cells in the dLN examined 9–10 days after ligature placement. (A) Representative FACS plots show the gating of eYFP⁺ CD4⁺ T cells and their Tfh (CXCR5⁺PD-1⁺) phenotype. (B) The proportion of eYFP⁺ CD4⁺ T cells in the dLN exhibiting a Tfh (CXCR5⁺PD-1⁺) or Tfr (CXCR5⁺PD-1⁺Foxp3⁺) phenotype in naive and LIP mice. (C) The proportion of total Tfh in the dLN in naive control and LIP mice. (D and E) Representative FACS plots and bar graph showing Tfh cells in the dLN expressing eYFP⁺ during LIP. (F–H) Graphs show the expression of Bcl6 (F), IL-17 (G), and CCR6 (H) in eYFP⁻Tfh⁺ (orange symbol), eYFP⁺Tfh⁺ (blue symbol), and eYFP⁺Tfh⁻ (green symbol) cells in the dLN of LIP mice. (I) Representative histogram of RORγt expression and bar graph showing mean fluorescence intensity (MFI) of RORγt staining in eYFP⁺Tfh⁻ (open symbol; green histogram) and eYFP⁺Tfh⁺ (black symbol; blue histogram) T cells in the dLN of LIP mice. (J) eYFP⁺Tfh⁺ (blue symbol) and eYFP⁺Tfh⁻ (green symbol) cells were FACS sorted from the dLN of LIP mice and TCRα chain sequencing undertaken. Repertoire analysis shows Horn’s index of similarity between eYFP⁺Tfh⁺ and eYFP⁺Tfh⁻ cells. Data are pooled from three to six experiments with two to four mice per group, apart from J where cells were sorted from three separate mice. *P < 0.05, ***P < 0.0001 as determined by unpaired Student’s t test (C and I), one-way ANOVA (F), Mann–Whitney (B and E), or Kruskal–Wallis (G and H) tests. Results are expressed as mean ± SD (C, F, and I) or median ± interquartile range (B, E, G, and H).

Figure S2.

Th17-to-Tfh plasticity is not a generic feature of all Th17-driven inflammatory responses. (A and B) C. rodentium (Citro: left column), IMQ-induced psoriasis (IMQ: central column), and AIA (right column) were induced in Il17a^CreR26R^eYFP mice and eYFP⁺ and eYFP⁻ CD4⁺ T cells examined at the peak of disease in the appropriate draining lymph nodes. (A) Graphs show proportion of eYFP⁺ CD4⁺ T cells. (B) Graphs show proportion of eYFP⁻ Tfh T cells. Dashed lines on graphs indicate percent of cells in the dLN of LIP mice. Data are pooled from two to four experiments with two to four mice per group. (C) Experimental outline of LIP experiments in which TGFβ, primary IFN, or IL-6 signals were inhibited. (D–F) Representative FACS plots showing gating strategy and exemplar staining for B cell populations (D), PC (E), and Ig staining (F). *P < 0.05 as determined by unpaired Student’s t test and **P < 0.05 as determined by Mann–Whitney test. Results are expressed as mean ± SD (*, Student's t test) or median ± interquartile range (**, Mann–Whitney test).

Figure 3.

Th17-to-Tfh plasticity is driven by IL-6. (A and B) C. rodentium (Citro: left column), IMQ-induced psoriasis (IMQ: central column), and AIA (right column) were induced in Il17a^CreR26R^eYFP mice and eYFP⁺CD4⁺ T cells examined at the peak of disease in the appropriate draining lymph nodes. Graphs and representative FACS plots showing the proportion of eYFP⁺CD4⁺ T cells exhibiting a Tfh (CXCR5⁺PD-1⁺) phenotype. Data are pooled from two to four experiments with two to four mice per group. Dashed lines on graphs indicate the percent of cells in the dLN of LIP mice. (C) Il17a^CreR26R^eYFP mice experienced gingival damage every other day for 11 days, then eYFP⁺CD4⁺ T cells were examined in the dLN. Graph shows the proportion of eYFP⁺CD4⁺ T cells that have a Tfh phenotype. Data are pooled from two experiments with three to four mice per group. (D) LIP was induced in IL-17^TGFbRII-KO (Tgfbr2^fl/flxIl17a^CreR26R^eYFP; gray symbols) and littermate control (Tgfbr2^fl/+xIl17a^CreR26R^eYFP or Tgfbr2^+/+xIl17a^CreR26R^eYFP; black symbols) mice that were examined 9–10 days after ligature placement. Graph shows the proportion of eYFP⁺CD4⁺ T cells that have a Tfh phenotype. Data are pooled from two experiments with two to four mice per group. (E) LIP was induced in Il17a^CreR26R^eYFP mice, and on days 2, 5, and 8 following induction mice received either isotype control (black symbols) or anti-IFNAR (gray symbols) i.p. before being examined at day 10. Graph shows proportion of eYFP⁺ CD4⁺ T cells that have a Tfh phenotype. Data are pooled from two experiments with two to four mice per group. (F–H) LIP was induced in Il17a^CreR26R^eYFP mice, and on days 2, 5, and 8 following induction mice received either isotype control (black bar/symbols) or anti-IL-6 (green bar/symbols) i.p. before being examined at day 10. (F) Graph shows total change in bone heights in LIP mice compared with un-ligated controls. (G and H) Graph and representative FACS plots show proportion of eYFP⁺CD4⁺ T cells that have a Tfh phenotype. Data are pooled from two to three experiments with one to three mice per group. *P < 0.05, ***P < 0.0001 as determined by unpaired Student’s t test (C), one-way ANOVA (D and G), or Kruskal–Wallis (E) test. Results are expressed as mean ± SD (A, C, D, F, and G) or median ± interquartile range (E).

Figure 4.

LIP induces B cell responses and increases in oral antibody. LIP was induced and B cell responses in the dLN and levels of antibody in oral rinses examined 9–10 days after ligature placement. (A–C) Bar graphs show proportions of B cells (A), switched B cells (B; IgD⁻IgM⁻), and GC B cells (C; GL7⁺Fas⁺) in the dLN of naive control (open symbols) and LIP (black symbols) mice. Data are pooled from eight experiments with two to four mice per group. (D) Bar graph shows proportion of PCs (CD45⁺B220⁻MHCII⁻CD138⁺) in the gingiva of naive control and LIP mice. Data are pooled from three experiments with two to four mice per group. (E) Graphs show expression of IgA and IgG on switched B cells (left) and GC B cells (right). Data are pooled from four to six experiments with two to four mice per group. (F and G) Graphs show concentration of total IgA (F) and total IgG (G) in the oral rinse of naive control or LIP mice. (H) Graphs show concentration of IgG1, IgG2b, IgG2c, and IgG3 in the oral rinse of naive control or LIP mice. Lines join dots from the same experiment. Data are pooled from five to eight experiments with two to four mice per group. *P < 0.05, **P < 0.005, ***P < 0.0001 as determined by unpaired Student’s t test (A, D, and E) or Mann–Whitney test (B, C, G, and H). Results are expressed as mean ± SD (A, D, and E) or median ± interquartile range (B, C, G, and H).

Figure 5.

Lack of Th17-to-Tfh plasticity during LIP enhances periodontal pathology. LIP was induced in IL-17^Bcl6-KO and littermate control mice and immune parameters examined 9–10 days after ligature placement. (A and B) Graph (A) and representative FACS plots (B) showing the proportion of eYFP⁺CD4⁺ T cells that exhibit a Tfh (CXCR5⁺PD-1⁺) phenotype in the dLN of IL-17^Bcl6-KO (black symbols) and control (open symbols) mice. (C) CEJ-ABC distance was measured at defined points across the molars and change in bone heights determined in IL-17^Bcl6-KO and control mice by comparing with un-ligated molars. Bar graph shows total change in bone heights in LIP mice compared with un-ligated controls. (D) Graph shows total cellularity of the dLN of IL-17^Bcl6-KO and control mice with LIP. (E) Graphs show (left) the proportion of CD44^hiCD4⁺ T cells staining positive for IL-17A and (right) MFI of that IL-17A staining normalized to MFI in control mice from the same experiment. (F) Graphs show the percent of neutrophils in the gingiva of IL-17^Bcl6-KO and control mice with LIP. (G) Graphs show the concentration of IgG1, IgG2c, and IgG2b in the oral rinse of IL-17^Bcl6-KO and control mice with LIP; box and whisker graphs show minimum to maximum values. (H and I) Graphs showing oral microbiome composition at the phyla (H) and genus (I) levels in control and IL-17^Bcl6-KO LIP animals elucidated by 16S rRNA sequencing of oral swab–derived bacteria; asterisk indicates significantly different genera. Data from n = 8 control samples and n = 9 IL-17^Bcl6-KO samples. (J and K) Graph shows total bacterial load in the oral cavity of IL-17^Bcl6-KO and control mice with LIP as determined by (J) 16S rRNA-based real-time PCR assay and (K) CFU enumeration from aerobic and anaerobic culture. Apart from 16S sequencing data, data are pooled from n = 10–26 mice per group (LIP mice) or n = 5–8 mice per group (naive mice) from three to seven different experiments. *P < 0.05, **P < 0.01, as determined by unpaired Student’s t test (C, F, and K), Mann–Whitney test (D, G, and J), or Kruskal–Wallis test (A). Results are expressed as mean ± SD (C, F, H, and I) or median ± interquartile range (A, D, and J).

Figure S3.

Elevated periodontitis pathology is seen in IL-17^Bcl6-KO mice. (A) LIP was induced in IL-17^Bcl6-KO and littermate control mice and immune parameters examined 9–10 days after ligature placement. Graph shows proportion of eYFP⁺ CD4⁺ T cells which exhibit a Tfh (CXCR5⁺PD-1⁺) phenotype in the Peyer’s patches of IL-17^Bcl6-KO (black symbols) and control (open symbols) mice. (B and C) Graphs show mouse weight and fecal IgA concentrations in adult naive IL-17^Bcl6-KO and control mice. Data are pooled from two to three experiments with two to four mice per group. (D) Relative expression of indicated genes in the colon and small intestine of IL-17^Bcl6-KO (black symbols) and control (open symbols) mice. Expression in IL-17^Bcl6-KO mice is presented relative to that in controls; data are pooled from four to six mice per group. (E) Representative FACS plots showing gating strategy and exemplar staining for gingival neutrophils (L/D; live/dead stain). (F and G) Graphs show proportions of gingival neutrophils staining positive for cytokine (F) or ROS (G) from IL-17^Bcl6-KO and control mice with LIP. Data are pooled from two experiments with one to three mice per group. (H) Graphs show proportions of total B cells, switched B cells, and GC B cells in the dLN of IL-17^Bcl6-KO and control mice with LIP. (I) Graphs show proportions of total B cells, PCs, and IL-10⁺ CD5⁺ B cells in the gingiva of IL-17^Bcl6-KO and control mice with LIP. Data are pooled from five to six (H) or two to four (I) experiments with two to three mice per group. (J) LIP was induced in IL-17^Bcl6-KO and control mice and IgA levels in the oral rinse examined; graph shows OD values from ELISA. Data are pooled from 15 to 17 mice from four separate experiments. (K) Left: Bar graph shows alpha-diversity of oral microbial communities as determined by Shannon diversity index. Right: Bray-Curtis–based non-metric multidimensional scaling (NMDS) plot of samples from control and IL-17^Bcl6-KO LIP animals (n = 8 control samples [open squares] and n = 9 IL-17^Bcl6-KO samples [black circles]). Results are expressed as mean ± SD (except for J, which is mean ± interquartile range).