Allergic inflammation is initiated by IL-33-dependent crosstalk between mast cells and basophils

Abstract

IgE-primed mast cells in peripheral tissues, including the skin, lung, and intestine, are key initiators of allergen-triggered edema and inflammation. Particularly in severe forms of allergy, this inflammation becomes strongly neutrophil dominated, and yet how mast cells coordinate this type of response is unknown. We and others have reported that activated mast cells--a hematopoietic cell type--can produce IL-33, a cytokine known to participate in allergic responses but generally considered as being of epithelial origin and driving Type 2 immune responses (e.g., ILC2 and eosinophil activation). Using models of skin anaphylaxis, our data reveal that mast cell-derived IL-33 also initiates neutrophilic inflammation. We demonstrate a cellular crosstalk mechanism whereby activated mast cells crosstalk to IL-33 receptor-bearing basophils, driving these basophils to adopt a unique response signature rich in neutrophil-associated molecules. We further establish that basophil expression of CXCL1 is necessary for IgE-driven neutrophilic inflammation. Our findings thus unearth a new mechanism by which mast cells initiate local inflammation after antigen triggering and might explain the complex inflammatory phenotypes observed in severe allergic diseases. Moreover, our findings (i) establish a functional link from IL-33 to neutrophilic inflammation that extends IL-33-mediated biology well beyond that of Type 2 immunity, and (ii) demonstrate the functional importance of hematopoietic cell-derived IL-33 in allergic pathogenesis.

Fig 1. IL-33 and ST2 are necessary for ADTI.

C57BL/6, St2^–/–, or Il33^–/–mice were used in the PCA model. Mice were intradermally injected with DNP-IgE followed by retroorbital challenge with DNP-HSA. (A) Ear thickness was measured at 0, 1, and 24 hours after antigen challenge. (B) Inflammatory cell infiltration was analyzed 24 hours after challenge by histological staining by H&E. (C, D) Flow cytometry flow plots and analysis of CD45⁺Gr-1^highSiglec-F^low neutrophils and (E) cytospun BAL cells stained with Diff-quick. Arrows illustrate neutrophils. *P ≤ 0.05, **P ≤ 0.01 and ***P ≤ 0.001 (one-way ANOVA). (F) Ear thickness was measured 24 hours after application of croton oil. Data are from at least 3 independent experiments, and the mean ± SEM from n = 10–18 mice per group in (A) and n = 3–7 mice per group in (D) and (F) are displayed. Data in B, C, and E are representative of at least 3 independent experiments with similar results.

Fig 2. Mast cell-derived IL-33 is necessary and sufficient to elicit ADTI.

(A) Ear skin of Kit^W-sh/W-sh mice was reconstituted after 8 weeks by intradermal injection of WT or Il33^–/–BMMCs, and mice were rested for 1 week. Mice then underwent the PCA model as described in the methods. Ear thickness was measured at 0, 1, and 24 hours after challenge. (B) Il33 mRNA was analyzed by real-time RT-PCR from peritoneal-cell derived mast cells harvested from Cpa3-cre and Cpa3-Cre×Il33^{fl/fl-IRES-eGFP} mice and stimulated with DNP-IgE/DNP-HSA for 4 hours. (C) Cpa3-cre and Cpa3-Cre×Il33^fl/fl mice underwent the PCA model as described in the methods and ear thickness was measured at 0, 1, and 24 hours after challenge. (D) Ear skin from WT or Il33^–/–mice were intradermally injected with WT or Il33^–/–BMMCs and rested for 1 week. Mice then underwent the PCA model as described in the methods. Ear thickness was measured at 0, 1 and 24 hours after challenge. **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001 (one-way ANOVA). Data are from at least 2 independent experiments, and the mean ± SEM of n = 4–12 mice per group in (A), n = 5–6 mice per group in (B), n = 22–27 mice per group in (C) and n = 6–7 mice per group in (D) are displayed.

Fig 3. ST2-expressing basophils are required for ADTI responses.

(A) Representative gating for basophils (MCPT8-YFP+) and (B) quantification of total cells and basophils in PBS and DNP-IgE/DNP-HSA treated ears 2 hours after challenge. (C) WT x iDTR (control) and Basoph8 x iDTR mice received 3 consecutive doses of DT (1 μg) for basophil depletion. Mice then underwent the PCA model as described in the methods. Ear thickness was measured at 0, 1 and 24 hours. (D) C57BL/6 WT and St2^–/–mice received WT or St2^–/–BMBs, and the PCA model was performed 1 hour later. Ear thickness (D) was measured at 0, 1, and 24 hours after challenge, and ear neutrophil population (CD45⁺Gr-1^highSiglec-F^low) was analyzed by flow cytometry (E). *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001 (one-way ANOVA). Data from (A, B) are from at least 2 independent experiments. Data from (C, D) are from at least 4 independent experiments, and the mean ± SEM from n = 11 (B), n = 9–17 (C), n = 12–17 (D) and n = 7–15 mice per group (E) are displayed.

Fig 4. ST2-driven stimulation of basophils promotes a unique inflammatory profile in which basophil-derived CXCL1 is functionally important in their involvement in ADTI.

(A–C) BMBs from C57BL/6J mice were activated with IgE/DNP or different doses of mIL-33. Gene expression 4 hours after activation was analyzed by microarray (A) and real time RT-PCR (B, C). Protein production 24 hours after activation was determined by ELISA (D). (E) Network diagram generated from the microarray comparison of activated BMBs in (A) showing select genes altered by IgE-stimulation, IL-33-stimulation and shared by both stimuli. (F) Ear tissues from C57BL/6J mice undergoing the PCA model were collected at 20 and 24 hours after challenge and homogenized. IL-6, CXCL1, CXCL2 and CCL24 expression was determined by ELISA and normalized to total protein concentration. Data are from 3 individual mice at each time point (mean ± SEM). (G) WT or St2^–/–underwent the PCA model. Some mice underwent repletion with WT or Cxcl1^–/–BMBs 1 hour before challenge. Ear thickness was measured at 0, 1 and 24 hours after challenge. *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001 (one-way ANOVA). Data are from at least 2 independent experiments, and mean ± SEM for n = 3–11 per group (B, C) n = 4–9 per group (D), n = 3 per group (F), and n = 6–10 per group (G) are displayed.