Figure Legends
Figure 1: iTreg induced by unloaded or Bet-loaded IL-10 DC displayed an anergic phenotype after primary culture. PBMC for DC generation and T cells were obtained from birch-pollen allergic patients with associated hazelnut allergy. CD4+ T cells were primed with autologous unloaded IL-10 DC (IL-10 DC0) and Bet-loaded IL-10 DC (IL-10 DCBet) or mDC (mDC0/mDCBet), respectively, to induce non-specific and Bet-specific iTreg (iTreg0/iTregBet) and Teff (Teff0/TeffBet, as controls).(A) After 3 days of iTreg induction, the coculture was pulsed with [3H]TdR for 16-18 h to assess T cell proliferation which is shown as stimulation index (SI, mean ± SD) normalized to T cells stimulated with mDC0 (SI = 1). The data are pooled from 34 independent experiments. (B-F) Cytokine concentrations (B IL-5, C IL-9, D IL-13,E IL-2 and F IL-10) in the supernatants of primary cultures are depicted as mean ± SD normalized to ctrl (=1). P values calculated with paired student’s t-test are depicted as asterisks: **** p < 0.0001, ** p < 0.01, * p < 0.05, ns = not significant (p > 0.05). PC: primary culture (induction)
Figure 2: Bet-specific iTreg lost their anergic phenotype after Bet- and Cor-induced restimulation. In primary culture (PC), Teff0, TeffBet, iTreg0and iTregBet were primed for five days by coculture of CD4+ T cells with autologous unloaded and Bet-loaded mDC or IL-10 DC, respectively. After a subsequent resting phase of 3 days, T cells were restimulated (RS) with unloaded, Bet-loaded or Cor-loaded mDC and were pulsed with [3H]TdR for 16-18 h on day 3. T cell proliferation is presented as SI (mean ± SD) normalized to Teff0 stimulated with mDC0(SI = 1). (A) Bet- and Cor-stimulated T cell proliferation of Teff0, TeffBet and iTregBet was pooled from 8 independent experiments.(B) Data of 9 independent experiments were pooled demonstrating T cell proliferation of iTregBet and iTreg0 restimulated with mDCBet or mDCCor, respectively. P values calculated with paired student’s t-test are depicted as asterisks: ** p < 0.01, * p < 0.05, ns = not significant (p > 0.05). PC: primary culture (induction), RS: restimulation
Figure 3: Bet-specific iTreg inhibited allergen-specific and cross-reactive responder T cell responses in vitro. Bet- and Cor-specific proliferation of autologous CD4+CD25low responder T cells (Tresp) obtained from birch-pollen allergic donors with associated hazelnut allergy was induced by stimulation with mDCBet(green) or mDCCor (brown), respectively. For analysis of the suppressive activity iTreg (iTregBet or iTreg0) were added in a Tresp:iTreg ratio of 1:1 or 1:2, respectively. DC, Tresp and iTreg were stained with different proliferation dyes for cell identification and assessment of proliferation by flow cytometry (see Supplementary Figure 2 for gating strategy). (A, B) Percentage of proliferating Tresp are shown from one representative experiment (top) and as pooled data (mean ± SD) relative to control ( = 100%) from independent experiments (bottom, number of independent experiments indicated below). (A)Percentage (mean ± SD) of proliferating Bet- (upper panel, green) and Cor-stimulated (lower panel, brown) Tresp, cocultured with iTregBet in the ratios 1: 1 and 1:2 , respectively, are demonstrated (Pooled data: Bet-stimulation: n= 15; Cor-stimulation: n=14). (B) Function of iTregBet was compared to iTreg0 in suppressor assays with Tresp:iTreg = 1:2 and percentages (mean ± SD) of proliferating Bet- and Cor-stimulated Tresp are depicted as pooled data (Bet-stimulation: n= 10; Cor-stimulation: n=8). (C) IL-13 (n=5-9) and (D) IL-10 concentrations (n=5-10) in the supernatants of suppressor assay samples are depicted as mean ± SD normalized to ctrl (=1). P values calculated with paired student’s t-test are depicted as asterisks: **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05, ns = not significant (p > 0.05).
Figure 4: Bet- and Cor-specifically stimulated iTreg showed an activated and suppressive phenotype. Prior to flow cytometry analysis of suppressor assays, T cell populations were stained for expression of extra- and intracellular markers and with different cell proliferation dyes (see supplementary Figure 5 for gating strategy) to distinguish between proliferating and non-proliferating T cells. (A)Activation and (B) immunosuppressive parameters are pooled from independent experiments as indicated (CD45RO n=5, CD25 n=9, HLA-DR n=10, CTLA-4 n=10, TNFR2 n=9, PD-1 n=10, IL-10 n=5, ICOS n=5) and expression is shown as percentage (mean ± SD), except for ICOS which is presented as the mean fluorescence intensity (mean ± SD). (C) Five independent experiments were pooled to show the percentage (mean ± SD) of CD49b+LAG3+ T cells. P values calculated with paired student’s t-test are depicted as asterisks: **** p < 0.0001, *** p < 0.001, ** p < 0.01, * p < 0.05, ns = not significant (p > 0.05).
Figure 5: Bet-specific iTreg abrogated birch-specific allergic symptoms in humanized mouse models of allergen-induced intestinal and airway inflammation. (A) Immunodeficient mice were engrafted with human PBMC from allergic donors suffering from birch pollen and associated hazelnut allergy and were boosted twice (d0, d8) with birch pollen extract to induce allergic immune responses. iTreg were co-injected to analyze their impact on allergy development. After three weeks, blood samples for IgE analysis were collected and subsequently the mice were challenged with birch pollen extract either rectally for induction of allergic intestinal inflammation or intranasally for allergic asthma induction. (B) The allergic intestinal inflammation was monitored by mini-endoscopy and scoring of colitis activity. One representative set of pictures is shown. (C)Quantitative endoscopic assessment of colitis activity in all groups is shown as mean ± SD from 8 independent experiments. (D) Results of airway resistance as parameter of allergic asthma were pooled from four independent experiments and are depicted as relative changes to baseline in % (mean ± SD). (E) The concentration of birch-specific IgE was obtained from 7 independent experiments as indicated and are presented in kU/L. P values calculated with paired student’s t-test are depicted as asterisks: **** p < 0.0001, ** p < 0.01, * p < 0.05. Figure 5A was created using Servier Medical Art (http://smart.servier.com/).
Figure 6: Bet-specific iTreg facilitated a cross-tolerance to hazelnut in a humanized mouse model of allergic gut inflammation.Immunodeficient mice were treated as described in Figure 5A. After three weeks, the mice were challenged rectally with either birch pollen or hazelnut extract, respectively, 2 h prior to scoring of the intestinal inflammation by mini-endoscopy. (A) The endoscopic score of the allergic intestinal inflammation of four independent experiments is depicted as pooled data (mean ± SD). (B) Blood samples were collected and human hazelnut-specific IgE was analyzed and is shown as mean value ± SD. Lines connect values from individual experiments. P values calculated with paired student’s t-test are shown as asterisks: *** p < 0.001, ** p < 0.01, ns = not significant (p > 0.05)