Fig 2: TH9 polarization in DCs stimulated with large targets is dependent on Dectin-1 and CARD9(A) Fluorescence imaging of YLCA or HLCA. ß-glucan was visualized with soluble murine Dectin-1 receptor fused to human IgG1 Fc(sDectin-1),and recombinant human IgG1 Fc was used as a control.(B–D) Wild-type or (B) Clec7a-/-, (C) Clec4n-/-, or (D) Card9-/- DCs were stimulated with YLCA or HLCA, pulsed with Ova peptide (323–339), and co-cultured with naïve Rag-/- OT-II CD4+ T cells. Production of IL-9 was assessed by ELISA. (B and D) n = 3 biological replicates. (C) n = 6 biological replicates.(E) Wild-type or Clec7a-/- DCs were stimulated with 3, 6, 15, 25, or 45 µm polystyrene beads that were coated with BSA, mannan, or ß-glucan overnight and pulsed with Ova peptide (323–339) for 2 h prior to being co-cultured with naïve Rag-/- OT-II CD4+ T cells. Production of IL-9 was assessed by ELISA. n = 6 biological replicates.(F) Confocal images showing how DCs process the beads according to their size. DCs were stimulated with ß-glucan-coated beads overnight and were stained with Dectin-1 and DAPI.(G) Wild-type DCs were stimulated with intact or fragmented wild-type hyphal C. albicans, pulsed with Ova peptide (323–339), and co-cultured with naïve Rag OT-II CD4+ T cells. Production of IL-9 was assessed by ELISA. n = 8 biological replicates.Results are (B–D) mean ± SD or(E) mean ± SEM analyzed using two-way ANOVA followed by Tukey’s post hoc test. In(E), statistical comparisons are only shown for ß-glucan-coated beads of various sizes and between genotypes. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001; not significant if it is not denoted.

Fig 3: Duration of Dectin-1 signaling between small- versus large-target-stimulated DCs(A) Confocal images showing surface Dectin-1 expression of DCs 5 min or 1 h after stimulating DCs with propidium-iodide-stained YLCA or HLCA (left). Relative quantification of the mean integrated density of Dectin-1 signal (right). n = 158 cells.(B) DCs were stimulated with YLCA or HLCA for 5 min or 1 or 6 h. After stimulation, DCs were stained with fluorescein isothiocyanate (FITC)-labeled anti-Dectin-1 antibody. Representative histograms (left) show surface expression of Dectin-1 in DCs. Pooled percentages of Dectin-1low DCs from 3 independent experiments are shown (right). n = 3 biological replicates.(C–D) Representative immunoblots showing phosphorylation of SYK and ERK in DCs that were stimulated for 5 or 30 min or 1,2,4, or 6 h with (C) YLCA or HLCA or (D) ß-glucan-coated polystyrene beads (6 or 25 µm) (left). Quantification of phosphorylation of SYK and ERK (right). Data are representative of more than three independent experiments.(E) Experimental plan for in vitro DC:OT-II T cell co-culture. Solid line represents persistent Dectin-1 signaling, and dotted line represents inhibited Dectin-1 signaling.(F) DCs were stimulated with a ß-glucan-coated plate for 1 h and treated with an SYK inhibitor (25 µM piceatannol) overnight. The next day, DCs were transferred to a fresh plate not coated with ß-glucan and co-cultured with naïve Rag-/- OT-II CD4+ T cells. Production of IL-9 was assessed by ELISA. n = 3 biological replicates.Results are mean ± SD analyzed using one-way ANOVA followed by Tukey’s post hoc test. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001; not significant if it is not denoted.

Fig 4: IRF8 is required for Th9 transcription program setting-up. a Heat map showing RNA sequencing data of WT Th1, Th2, Th9, Treg and Th17 cells differentiated for 24 h. b MA-plot showing RNA sequencing data of WT or IRF8-/- Th9 cells after 24 h of differentiation. Genes upregulated or downregulated by at least 2-fold in IRF8-/- Th9 cells compared with WT Th9 cells are labelled in black, genes with similar expression in both IRF8-/- and WT Th9 cells are labelled in grey. Th9 cells-specific cytokines (Il9 and Il21) and genes of interest (Il4, Sfpi1, Batf, Irf4, Stat6) are highlighted in red. c Heat map showing RNA sequencing data of WT Th1, Th2, Th9, Treg, Th17 and Th9IRF8-/- cells differentiated for 24 h. d Venn diagram of genes downregulated in IRF8-/- Th9 cells at 24 h (green) and Th9 cells-specific genes (red). e Distribution of IRF8 ChIP sequencing peaks in WT Th9 cells. f, g Top panel presents a screenshot from the ECR (evolutionary conserved regions) Browser web site of the mouse Il9 (f) and Il21 (g) genes. Exonic regions are in blue, intronic regions in pink, UTRs in yellow and CNS are in red. Bottom panel presents IRF8-binding peaks in WT Th9 cells at Il9 (f) and Il21 (g) loci. h ChIP analysis of the interaction between IRF8 and the CNS1 of Il9 or Il21 promoter in Th2 and Th9 cells. ns, not significant; *P < 0.05, **P < 0.01; ***P < 0.001 (two-way ANOVA (h)). Data are from four h independent experiments (mean and s.e.m.)

Fig 5: IRF8 deficiency impairs Th9 cell development in vitro. a Immunoblot analysis of IRF8 in WT naive CD4+ T cells or after 1 day of differentiation into Th0, Th2, Th9, Treg, Th1, Th17 and Tfh cells. b, c WT naive CD4+ T cells were transfected with control siRNA (siCT) or siRNA against Irf8 (siIRF8), and then polarised under Th9 conditions. Relative expression of Il9 and Il21 mRNA (b) ELISA analysis of IL-9 protein in supernatant (c). d IL-9-eGFP naive CD4+ T cells were transfected with siCT or siIRF8, and then polarised under Th9 conditions. After 3 days of differentiation, eGFP-positive cells were assessed by flow cytometry (left: representative dot plot, right: means of four independent experiments). e, f WT naive CD4+ T cells were retrovirally infected with an empty-GFP vector (EV) or IRF8-GFP overexpressing vector (IRF8). GFP+ cells were sorted 2 days after infection and differentiated for 3 days into Th9 cells. Il9 mRNA expression e and IL-9 protein release assessed with ELISA (f) were evaluated. g-i, Irf8 f/f Cd4 cre and Irf8 +/+ Cd4 cre CD4+ T cells were differentiated into Th9 cells for 3 days. Il9 and Il21 mRNA expression (g) ELISA analysis of IL-9 protein in supernatant (h) and IL-9 intracellular staining in Th9 cells (i) (left: representative dot plot, right: means of four independent experiments). j IRF8-deficient naive CD4+ T cells were retroviral infected with an empty-GFP vector (EV) or IRF8-GFP overexpressing vector (IRF8). GFP+ cells were sorted 2 days after infection and differentiated for 3 days into Th9 cells. IL-9 protein release was evaluated by ELISA. ns, not significant; *P < 0.05, **P < 0.01; ***P < 0.001 (Mann–Whitney test). Data are from three b, c, e–h or two j independent experiments (mean and s.e.m.), d, i and one experiment representative of four independent experiments