An Interleukin-33-Mast Cell-Interleukin-2 Axis Suppresses Papain-Induced Allergic Inflammation by Promoting Regulatory T Cell Numbers

Abstract

House dust mite-derived proteases contribute to allergic disorders in part by disrupting epithelial barrier function. Interleukin-33 (IL-33), produced by lung cells after exposure to protease allergens, can induce innate-type airway eosinophilia by activating natural helper (NH) cells, a member of group 2 innate lymphoid cells (ILC2), to secrete Th2 type-cytokines. Because IL-33 also can induce mast cells (MCs) to secrete Th2 type-cytokines, MCs are thought to cooperate with NH cells in enhancing protease or IL-33-mediated innate-type airway eosinophilia. However, we found that MC-deficient Kit(W-sh/W-sh) mice exhibited exacerbated protease-induced lung inflammation associated with reduced numbers of regulatory T (Treg) cells. Moreover, IL-2 produced by IL-33-stimulated MCs promoted expansion of numbers of Treg cells, thereby suppressing development of papain- or IL-33-induced airway eosinophilia. We have thus identified a unique anti-inflammatory pathway that can limit induction of innate-type allergic airway inflammation mediated by NH cells.

Figure 1. Lung Natural Helper Cells Contribute to Papain-Induced Innate-Type Airway Eosinophilia

(A) Cell numbers and concentrations of cytokines in BALF and lungs from C57BL/6N wild-type mice 24 hr after the final inhalation of various doses of papain (n = 5–6). (B–D) Numbers of eosinophils and neutrophils in BALF from mice 24 hr after papain inhalation (25 μg) (heat-inactivated papain [HI], n = 3–6; and papain, n = 10–21). Wild-type and Il33^−/− mice (B), Rag2^−/− mice injected with anti-mouse CD25 mAb and control IgG (C), and wild-type, Rag2^−/− Il2rg^−/− mice, and Rag2^−/− Il2rg^−/− mice which had been engrafted with NH cells from wild-type mice (D). (E) Numbers of eosinophils and neutrophils in BALF and expression of IL-33 mRNA in the lungs from wild-type and Par2^−/− mice 24 hr after papain inhalation (25 μg). Heat-inactivated papain [HI], n = 7–8 for BALF, and n = 5 for IL-33 mRNA; and papain, n = 20–21 for BALF, and n = 8 for IL-33 mRNA. (F) IL-33 expression in lungs from C57BL/6N wild-type mice after papain inhalation (25 μg). Red, IL-33 staining; blue, DAPI staining. Scale bars represent 100 μm. The data show the mean ± SEM (A) or mean + SEM (B–E). *p < 0.05, **p < 0.01, and ***p < 0.005 versus 0 μg papain (A) and HI (B–E); †p < 0.05 and †††p < 0.005 versus papain-treated wild-type mice (B, D, E) and papain-treated control IgG-injected Rag2^−/− mice (C); and ††p < 0.01 versus papain-treated Rag2^−/− Il2rg^−/− mice engrafted with NH cells (Rag2^−/− Il2rg^−/− + NH).

Figure 2. Exacerbated Papain-IL-33-NH-Cell-Mediated Airway Eosinophilia in Rag2-Deficient and Mast-Cell-Deficient Kit^W-sh/W-sh Mice

(A and B) Numbers of eosinophils and levels of IL-4, IL-5, and IL-13 in BALF from wild-type mice and Kit^W-sh/W-sh mice (heat-inactivated papain [HI], n = 5; and papain; n = 24–25) (A) and wild-type mice and Rag2^−/− mice (heat-inactivated papain [HI], n = 5; and papain; n = 10–25). (B) 24 hr after the final inhalation of low-dose papain (25 μg). (C) Numbers of NH cells and MCs in BALF and/or lungs from wild-type mice, Kit^W-sh/W-sh mice, and Rag2^−/− mice 24 hr after the final inhalation of low-dose papain (25 μg) (heat-inactivated papain [HI], n = 5–10; and papain; n = 5–10). (D) Numbers of eosinophils and levels of IL-4, IL-5, and IL-13 in BALF from wild-type mice, Kit^W-sh/W-sh mice, and Rag2^−/− mice 24 hr after the final inhalation of low-dose IL-33 (0.1 μg) (saline, n = 4–6; and IL-33, n = 11–15). The data show the mean + SEM. *p < 0.05 and ***p < 0.005 versus HI (A–C) and saline (D), and †p < 0.05 and ††††p < 0.005 versus papain-treated wild-type mice (A–C) and IL-33-treated wild-type mice (D).

Figure 3. Evidence that Treg Cells and Mast Cells Are Necessary for Suppression of Papain-IL-33-NH-Cell-Mediated Airway Eosinophilia

(A) Proportions of CD4⁺CD25⁺Foxp3⁺ Treg cells in BALF, thoracic LNs and spleen from naive, papain-inhaled and/or IL-33-inhaled C57BL/6J wild-type mice. Since the number of lymphocytes in BALF from naive mice was so small, the proportion of CD4⁺CD25⁺Foxp3⁺ Treg cells in these mice was hardly detectable. Data show representative results from three independent experiments. (B) Numbers of eosinophils in BALF from wild-type mice, Rag2^−/− mice, and Rag2^−/− mice which had been engrafted with CD4⁺CD25⁺ Treg cells from spleens of wild-type mice (Rag2^−/− + Treg) 24 hr after the final inhalation of 25 μg of papain or 0.1 μg of IL-33 (heat-inactivated papain [HI], n = 4–5; versus papain, n = 16; and saline, n = 5, versus IL-33, n = 13). (C–E) Numbers of eosinophils in BALF (C), and (D and E) proportions of CD4⁺CD25⁺Foxp3⁺ Treg cells in thoracic LNs from wild-type mice and Kit^W-sh/W-sh mice 24 hr after the final inhalation of 25 μg of papain or 0.1 μg of IL-33 (saline, n = 9; papain, n = 10; and IL-33, n = 10). FACS data show representative results in (D). (F) Numbers of eosinophils in BALF from wild-type mice, Kit^W-sh/W-sh mice, and Kit^W-sh/W-sh mice that had been engrafted with CD4⁺CD25⁺ Treg cells from spleens of C57BL/6J wild-type mice (Kit^W-sh/W-sh + Treg) 24 hr after the final inhalation of 25 μg of papain or 0.1 μg of IL-33. Data show the mean + SEM (HI, n = 4–5, versus papain, n = 10–15; saline, n = 5, versus IL-33, n = 13). *p < 0.05 and ***p < 0.005 versus HI or saline (B, C, and F), and †p < 0.05 and ††p < 0.001 versus papain-treated or IL-33-treated wild-type mice (B, C, and F) and versus papain-treated or IL-33-treated Rag2^−/− + Treg groups or Kit^W-sh/W-sh + Treg groups (B and F).

Figure 4. IL-33-Stimulated Mast Cells Can Enhance Treg Cell Expansion In Vitro

(A–F) Mouse (A–E) or human (F) CD4⁺ T cells were cultured with and/or without mouse bone-marrow-derived cultured mast cells (MCs) (A–E) or human peripheral blood stem-cell-derived cultured MCs (F) in the presence and absence of IL-33 for 3 days. The proportions and numbers of CD25⁺Foxp3⁺ and/or Helios⁺ Treg cells among c-Kit-negative CD4⁺ T cells were assessed by flow cytometry. Data show representative results from 3–5 batches of MCs per experiment and from 2–3 independent experiments, each of which gave similar results (A, B, D–F). (A) Co-culture of mouse CD4⁺CD25⁻ CD62L⁺ naive T cells and wild-type MCs. (B) Proportion of CD25⁺Foxp3⁺ Treg cells, and expression of Helios in CD4⁺ CD25⁺ T cells in co-culture of mouse CD4⁺ T cells (not naive T cells) and wild-type MCs in the presence of IL-33. (C) Co-culture of mouse CD4⁺CD25 ⁺ Treg cells and wild-type MCs. After flow cytometric analysis, the number of Treg cells was calculated. Data show the mean + SEM (n = 3). (D) Culture of mouse splenic CD4⁺ T cells with and without wild-type, Myd88^−/−, Il1rl1^−/− , or Il6^−/− MCs. (E) Co-culture of mouse CD4⁺ T cells with wild-type or H2-Ab1^−/− MCs. (F) Co-culture of human peripheral blood-derived CD4⁺ T cells with and without human peripheral blood stem-cell-derived cultured MCs.

Figure 5. Suppression of Airway Eosinophilia by Treg Cells Derived from IL-33-Stimulated MCs

(A) Numbers of eosinophils in BALF from wild-type mice, Kit^W-sh/W-sh mice and Kit^W-sh/W-sh mice engrafted with IL-33- and MC-derived wild-type CD4⁺CD25⁺ Treg cells (Kit^W-sh/W-sh + WT Treg) or Il10^−/− CD4⁺CD25⁺ Treg cells (Kit^W-sh/W-sh + Il10^−/− Treg) at 24 hr after the final IL-33 or saline inhalation. Data show the mean + SEM (saline, n = 5; and IL-33, n = 10). ***p < 0.005 versus saline, and †p < 0.001 versus values from wild-type or Kit^W-sh/W-sh + WT Treg groups. (B) Expression of IL-10R1 on NH cells from BALF of C57BL/6J-wild-type mice after the final inhalation of 25 μg of papain or 0.1 μg of IL-33. BAL cells were incubated with anti-IL-10R1 mAb or isotype matched control IgG in the presence and/or absence of rmIL-10. The expression of IL-10R1 on NH cells (gated by NH cell markers) in BAL cells was determined by flow cytometry. Gray lines = isotype-matched control IgG staining, red lines = anti-mouse IL-10R1 mAb staining or a green line = anti-mouse IL-10R1 mAb staining in the presence of rmIL-10. Since the proportion of lymphocytes in BALF from naive mice was very small, the proportion of NH cells in these mice was hardly detectable. Data show representative results from one of three mice. (C) Numbers of NH cells and percent inhibition of numbers of NH cells, and IL-5 and IL-13 production by NH cells, in response to 10 ng/ml IL-33 in the presence and absence of 0.5 2 and 50 ng/ml IL-10. Data show the mean + SD (n = 3). *p < 0.05 versus 0 (no added IL-10). (D) Numbers of NH cells, and IL-5 and IL-13 production by NH cells, in response to 10 ng/ml IL-2, 10 ng/ml IL-25, 10 ng/ml IL-33, and IL-2 + IL-25 in the presence and absence of 10 ng/ml IL-10. Data show the mean + SD (n = 3) *p < 0.05 versus Medium. (E) Expression levels of IL-10R1 mRNA in human ILC2s, Jurkat cells and HeLa cells were determined by quantitative PCR. Data show mean + SEM (n = 3). (F) Human ILC2s were co-cultured with and without human Treg cells in the presence of 50 ng/ml IL-33 and 10 U/ml IL-2. Data show representative results from the two independent experiments performed, each of which gave similar results.

Figure 6. Importance of MC-Derived IL-2 and Costimulatory Molecules for IL-33- and MC-Mediated Treg Cell Expansion In Vitro

(A–E) Mouse CD4⁺ T cells (not naive T cells) were cultured with and/or without mouse bone-marrow-derived cultured MCs (MCs) in the presence and absence of IL-33 for 3 days. The proportion of CD25⁺Foxp3⁺ Treg cells among c-Kit-negative CD4⁺ T cells was assessed by flow cytometry. (A) Co-culture of CD4⁺ T cells with wild-type MCs separated by transmembrane filters (Separated) or cultured together (Not separated). (B) Co-culture of CD4⁺ T cells with wild-type or Il2^−/− MCs in the presence and absence of IL-33. (C) Co-culture of CD4⁺ T cells with Tgfb1^+/+ or Tgfb1^−/− MCs in the presence and absence of IL-33. (D) Co-culture of CD4⁺ T cells with wild-type or Il10^−/− MCs in the presence and absence of IL-33. (E) Co-culture of CD4⁺ T cells with wild-type MCs in the presence of neutralizing mAb for adhesion/co-signaling molecules. (A–D) Data show representative results from 3–5 batches of MCs per experiment and from the 2–3 independent experiments performed, each of which gave similar results. (E) Data show the mean + SEM (n = 5). **p < 0.01 versus control IgG groups.

Figure 7. Importance of IL-2 Produced by IL-33-Stimulated Mast Cells for Suppression of Papain-IL-33-NH-Cell-Mediated Airway Eosinophilia

(A–D) Mice were intranasally administered 0.1 μg IL-33 or saline, or 25 μg papain or heat-inactivated papain (HI) for 3 days (once per day). Numbers of eosinophils in BALF and proportions of Foxp3⁺ Treg cells among CD4⁺ cells in thoracic LNs from wild-type mice, Kit^W-sh/W-sh mice and Kit^W-sh/W-sh mice that had been engrafted with wild-type (WT), Il2^−/−, or Il1rl1^−/− MCs before the start of IL-33 or saline challenge) at 24 hr after the final IL-33 or saline inhalation (saline, n = 5; IL-33, n = 10) (A–C) or at 24 hr after the final papain or heat-inactivated papain (HI) inhalation (HI, n = 4–6; papain, n = 10–11) (D). Data show the mean + SEM (A, C, and D). *p < 0.005 versus saline or HI and †p < 0.05 versus wild-type mice or Kit^W-sh/W-sh mice that had been engrafted with wild-type MCs (Kit^W-sh/W-sh + WT MCs). FACS data show a representative result in each IL-33-treated group, as shown in (C). (E) IL-2 production by wild-type MCs co-cultured with or without lung epithelial cells (ECs) from wild-type (WT) or Il33^−/− mice in the presence of papain and monensin for 6 hr. Green, IL-2; Red, MC; and Blue, Hoechst. Scale bars represent 10 μm. Representative results in each group are shown. No staining was obtained with the isoptype control for the anti-IL-2 Ab (data not shown).