Interleukin-33 amplifies IgE synthesis and triggers mast cell degranulation via interleukin-4 in naïve mice1

Background The regulation and function of IgE in healthy individuals and in antigen-naïve animals is not well understood. IL-33 administration increases serum IgE in mice with unknown mechanism. We tested the hypothesis that IL-33 provides an antigen-independent stimulus for IgE production and mast cell degranulation. Methods IL-33 was administered to naïve wild-type (WT), nude and ST2−/−, IL-4−/−, IL4Rα−/− and T-or B-cell-specific IL-4Rα−/− mice. IgEand cytokines were quantified by ELISA. T- and B-lymphocyte numbers and CD40L expression were determined by flow cytometry. Anaphylaxis was measured by temperature, mast cell degranulation and histamine release. Results IL-33 enhanced IgE production in naïve WT, T-IL-4Rα−/− but not in ST2−/−, IL-4−/−, IL-4Rα−/− or B-cell-specific IL-4Rα−/− mice, demonstrating IL-33 specificity and IL-4 dependency. Moreover, IL-4 was required for IL-33-induced B-cell proliferation and T-cell CD40L expression, which promotes IgE production. IL-33-induced IL-4 production was mainly from innate cells including mast cells and eosinophils. IL-33 increased mast cell surface IgE and triggered degranulation and systemic anaphylaxis in allergen-naïve WT but not in IL-4Rα−/− mice. Conclusion IL-33 amplifies IgE synthesis and triggers anaphylaxis in naïve mice via IL-4, independent of allergen. IL-33 may play an important role in nonatopic allergy and idiopathic anaphylaxis.

IgE antibody is induced by allergen and requires immunoglobulin heavy-chain class switching and affinity maturation in B cells in a T-cell-dependent manner (1,2). IgE is also produced naturally (1,3) but less is known about the selective amplification and function of natural IgE in healthy subjects and antigen-naı¨ve animals (3). Natural IgE is secreted by B2 cells and contributes to the baseline concentrations of nonallergen-specific IgE in the serum and tissues of healthy individuals (3,4). In contrast to allergen-induced IgE, cognate T-helper cell via MHC class II and somatic hypermutation are not required for the generation of natural IgE (3). Instead, bystander stimulation from T-cell CD40L and low levels of IL-4 are necessary (3,4). The functions of natural IgE are poorly understood; they are polyreactive and have a critical role in immune priming (3)(4)(5)(6)(7). IgE can bind the high-affinity IgE receptor (FceRI) and prolong mast cell survival, release a spectrum of cytokines and have important roles in defence against infection and initiation of adaptive immunity (5)(6)(7)(8).
We aim to identify the mechanism and role of IL-33 in the induction and regulation of IgE production in allergen-naı¨ve mice. We report that IL-33 amplifies IgE synthesis via IL-4 from innate immune cells that is sufficient to trigger mast cell degranulation and anaphylaxis in the absence of allergen.

IL-33 treatment in vivo
Recombinant IL-33 was prepared as described (23). Mice were injected i.p. with IL-33 (2 lg/mouse) or PBS daily for 1-7 days unless stated otherwise. Cells were collected by peritoneal washout with 2 ml of cold PBS.
Anaphylaxis (AS): Mice were injected i.p. with IL-33 (2 lg) or PBS daily for 3-4 days. Changes in clinical signs and temperature were monitored every 10 min after each injection, using a rectal probe (Natsume Seisakusyo Co., Tokyo, Japan). Plasma histamine was measured by ELISA (IBL, Hamburg, Germany).

Cytokine and IgE measurements
Serum concentrations of cytokines were determined by Luminex (Luminex, Biosource; Invitrogen, Life Technologies Ltd, Paisley, UK) and IgE by ELISA (Bethyl Laboratories, Inc, Montgomery, TX, USA).

Histological analysis of mast cells
Skin samples were stained with toluidine blue to identify and count mast cells by microscopy. Degranulated mast cells were identified as those with >10% extrusion of granules (20).

Statistical analysis
Student's t-test was applied to in vitro studies. Analysis between individuals in groups in vivo was carried out by ANOVA followed by Student's t-test.

IL-33-induced IgE but not type-II cytokines is IL-4 dependent
Mice had seven daily injections of IL-33 or PBS. Low basal concentrations of serum IgE, IL-4 and IL-13 were maintained in antigen-naı¨ve or PBS-treated mice, but were significantly elevated in wild-type (WT), but not ST2 À/À mice, within 3 days of IL-33 administration (Fig. 1A,B) IL-4 production preceded the increase in IgE. As reported (9), IL-33 did not induce IFN-c in naı¨ve WT mice (data not shown). These results suggest that IL-33 is a powerful and rapid inducer of IgE in naı¨ve mice that might be mediated by type-II cytokines.
IL-4 in mice and IL-4 and IL-13 in humans are key cytokines for the initiation of type-II responses and IgE isotype class switching (4,24). We therefore determined the role of IL-4 and IL-13 in IL-33-induced IgE synthesis. WT and IL-4Ra À/À mice were injected with IL-33 resulting in higher serum concentrations of IL-4 and IL-13 in IL-4Ra À/À mice compared with PBS-treated IL-4Ra À/À mice or with IL-33-treated WT mice (Fig. 1C), probably due to reduced sequestration of these cytokines in the absence of IL-4Ra (25). More importantly, IL-33 failed to enhance IgE in IL-4Ra À/À mice (Fig. 1C). Together, these data indicate that IL-33-mediated induction of IgE is dependent on IL-4Ra responsiveness and that IL-4/IL-13 induction is not.

IL-33 induces type-II cytokines mainly via innate cells in naïve mice
We next investigated whether the increased IL-4 and IL-13 levels in IL-33-treated mice were produced by T cells or innate cells in vivo by comparing WT and athymic nude mice (lacking T cells) injected with IL-33 or PBS as described above. Nude mice produced significantly higher levels of serum IL-4 and IL-13 than IL-33-treated WT mice, but markedly lower IgE ( Fig. 2A). This suggests two points. First, T cells have minimal contribution to IL-33-induced type-II cytokine production in vivo. Second, while IL-33 could induce significant amounts of IgE in nude mice compared with PBS control, the levels of IgE were markedly lower than those of IL-33 in WT mice, suggesting that T cells have a modest but significant role in IL-33-mediated IgE induction, independent of enhanced type-II cytokine synthesis.
We next determined which were the IL-4-producing cells in vivo from IL-33-or PBS-treated WT mice by flow cytometry. IL-33 enhanced the IL-4 expression in CD117 (c-Kit) + mast cells (11% compared with 1% in PBS-treated controls) and in Siglec-F + eosinophils (31% compared with 2% in PBS controls) (Fig. 2B). However, IL-33 failed to increase the frequency of IL-4 + CD4 + T cells (Fig. 2B). IL-33 also increased the numbers of total cells, mast cells, eosinophils and to a lesser extent CD4 + T cells in the peritoneal cavity (Fig. 2C). Thus, IL-33 mainly induces type-II cytokines via innate immune cells in naı¨ve mice.

B cells expressing the IL-4Ra are required for IL-33-induced B-cell expansion and IgE production in vivo
We have previously shown that conventional CD19 + B cells do not express ST2 (26), suggesting that IL-33 may only induce IgE synthesis indirectly, likely via IL-4 as we have demonstrated earlier. We therefore assessed the role of IL-4Ra signalling on B cells in IL-33-induced IgE synthesis by comparing naı¨ve B-cell-specific IL-4Ra À/À (selectively deleted of IL-4Ra on B cells), IL-4Ra À/À (total IL-4Ra deletion) and WT mice treated with IL-33 or PBS. IL-33 induced IgE synthesis in WT but not in IL-4Ra À/À or B-IL-4Ra À/À mice (Fig. 3A). Serum IL-4 and IL-13 concentrations in B-IL-4Ra À/À and WT mice were similar but were significantly lower than those of the IL-4Ra À/À mice (Fig. 3B), suggesting that cells other than B cells may consume some of these cytokines. IL-33 increased the total cell number and the number of CD19 + B cells (but not CD4 + T cells) in the spleen of WT mice but not in IL-4Ra À/À or B-IL-4Ra À/À mice (Fig. 3C). Because IL-13 does not directly induce IgE in mouse B cells (24), together these results suggest that IL-33 indirectly elicits natural IgE production in naı¨ve mice by inducing IL-4, which then directly stimulates B-cell expansion and IgE isotype class switching.

T cells expressing IL-4Ra contribute to IL-33-induced IgE synthesis
Our aforementioned results ( Fig. 2A,B) suggest that T cells also contribute to IL-33-enhanced IgE production in naı¨ve mice other than by providing type-II cytokines. To assess the potential role of IL-4R responses by T cells, we injected IL-33 or PBS into WT or T-IL4Ra À/À (Lck cre Il4ra À/lox mice, selectively deleted of IL-4Ra on T cells) mice. IL-33 modestly but significantly increased serum IgE levels in T-IL-4Ra À/À mice (Fig. 4A). Consistent with the result from nude mice ( Fig. 2A), IL-33 elicited abundantly increased level of IL-4 and IL-13 in T-IL4Ra À/À compared with WT mice (Fig. 4B), suggesting that these cytokines are unlikely to play an important role in T-cell function in IL-33-induced IgE production. An alternative contribution therefore should be considered.
CD40L expression on T cells is required for B cells to produce IgE, and the expression of CD40L on T cells can be enhanced by IL-2 receptor signals (3,27). We therefore assessed whether IL-33 could up-regulate the expression of CD40L and CD25 (IL-2 receptor a chain) on T cells. Because murine CD4 + T cells only respond to IL-4 but not IL-13 (because of the lack of IL-13 receptor) (24), we examined IL-33-mediated T-cell function in IL-4 À/À mice ex vivo. IL-33 increased the surface expression of CD40L and CD25 on CD4 + T cells in WT mice but not in IL-4 À/À mice (Fig. 4C). Consistent with our finding (Fig. 3C), IL-33 also increased the total number of B cells (but not T cells) in the WT but not in IL-4 À/À mice (Fig. 4D). Thus, IL-33 stimulates CD40L and CD25 expression on T cells via IL-4, which may contribute to the IL-33-induced IgE production in naı¨ve mice.
IL-33 triggers mast cell degranulation in WT but not ST2 À/À or IL4Ra À/À mice Because natural IgE can bind to mast cell FceRI and impact mast cell functions (5-7), we sought to investigate what effect IL-33 and IL-33-derived IgE may have on mast cell function in naı¨ve mice.
Naı¨ve WT, ST2 À/À or IL4Ra À/À mice were injected with IL-33 or PBS daily for up to 3 days. IL-33 treatment markedly increased the mast cell surface IgE in naı¨ve WT but not in IL4Ra À/À mice as evidenced by enhanced levels of membrane-bound IgE and reduced levels of unoccupied FceRI on mast cells (Fig. 5A). IL-33 treatment enhanced total and degranulated skin mast cells compared with PBS controls in WT but not ST2 À/À or IL4Ra À/À mice (Fig. 5B). More importantly, 30 min after the third IL-33 challenge, WT but not ST2 À/À or IL4Ra À/À mice showed clear signs of shock (piloerection, prostration, reduced response to stimuli), with a rapid reduction in body temperature lasting for about 15 min before returning to normal (Fig. 5C), and a substantial increase in serum histamine levels (Fig. 5D). Together, these results demonstrate that IL-33 can initiate mast cell degranulation and systemic anaphylaxis in naı¨ve mice via IL4Ra.

Discussion
Data reported in this study reveal a hitherto unrecognized effect and mechanism by which IL-33 induces IgE production and anaphylaxis in the absence of specific allergen (Fig. 6). IL-33 triggers IL-4 secretion by innate cells, which then stimulates B-cell expansion and IgE synthesis. In parallel, IL-4 enhances T-cell expression of CD40L, which interacts with CD40 on B cells to complement IgE production. IL-33 and IgE then synergistically trigger mast cell degranulation and anaphylaxis. Thus, IL-4 is the key intermediate in IL-33-mediated anaphylaxis in vivo. Because the mice were allergen naı¨ve and we found that serum samples even from 7 daily injections of IL-33, which showed enhanced levels of total IgE, failed to show any detectable recombinant IL-33-specific IgE (data not shown) by ELISA (Data S1), it is likely, although not proven, that this IgE is natural IgE as described previously (3).
We demonstrated here that eosinophils and mast cells can produce IL-4 upon IL-33 treatment in naı¨ve mice. However, it is likely that other ST2 + innate cells such as basophils and NKT cells can similarly produce IL-4 (10,14,16). By contrast, T cells, which are the key cells expressing type-II cytokines in allergy, are not the major IL-4 producers in this innate immune context. This may be because naı¨ve T cells do not express ST2 The levels of CD40L and CD25 on CD4 + T cells and the number of CD4 + CD40L + , CD4 + CD25 + T cells in the spleen of WT and IL-4 À/À mice were determined by FACScan and differential counting. (D) Total CD4 + T and CD19 + B cells in the spleen of WT and IL-4 À/À mice were determined by differential cell count. Data are from two experiments, n = 6 mice per group, *P < 0.05, **P < 0.01 compared to PBS control. and hence are not responsive to IL-33 in the absence of T cell receptor activation (17,18). In addition, it is known that IL-33 induces mainly IL-5 but not IL-4 in activated T cells and therefore cannot polarize Th2 cells (17). Using nude mice and T-IL-4Ra-deficient mice, we demonstrated that T cells nevertheless play a significant role in IL-33-derived IgE production in naı¨ve mice. However, this process is unlikely to involve IL-4 production, because nude mice and T-IL-4Ra-deficient mice produced normal or elevated levels of IL-4 and yet had significantly less IgE compared to WT mice. Previous studies have shown that T cells expressing CD40L are required for the induction of natural IgE and that the CD40L expression can be enhanced by IL-2 (3,27). Furthermore, IL-33 can induce IL-2 production in T cells (10). We show here that IL-33 can increase CD40L and CD25 expression on CD4 + T cells via IL-4. It is therefore likely that IL-33-derived IL-4 and IL-2, alone or together, stimulate CD4 + T cells to express surface CD40L, which subsequently interacts with B cells to promote IgE production.
Anaphylaxis (AS) is an acute, life-threatening allergic reaction in which IgE-mediated mast cell degranulation plays a critical role (28,29). The aetiology of AS in particular idiopathic AS is not fully understood (28,30). Our result demonstrated for the first time that IL-33/ST2 signals can trigger anaphylaxis in naı¨ve mice in the absence of specific allergen (Fig. 5). While the detailed molecular mechanism is still unresolved, we found the IL-33-mediated allergen-independent mast cell degranulation and systemic AS described here was dependent on ST2, IL-4Ra and required repeated IL-33 stimulation (Fig. 5C). Thus, the mechanism may comprise a combination of the following two components: 1 It is known that increased mast cell surface IgE expression and mast cell activation are the hallmark of mast cell degranulation and AS. We found that IL-33 can markedly enhance the levels of serum and mast cell surface IgE that is dependent on ST2 and IL-4Ra in antigen-naı¨ve mice. In addition, mast cells also express high levels of ST2, IL-4Ra and IgE FceR, and it is known that signals via these receptors play critical roles in mast cell activation and function. IL-33 can directly stimulate mast cell proliferation, cytokine production and activation (Fig. 2B, C and ref. 19), and the signals from the interaction between IL-4/IL-13 and IgE with their receptors can activate and prolong the survival of mast cells as well as can exacerbate AS (6,8,31). 2 We also found that at least three daily IL-33 treatments were required for IL-33 to trigger mast cell degranulation and AS. Intriguingly, this was also the critical time required for IL-33 to elicit significant production of IgE and IL-4/IL-13 and mast cell proliferation (Figs 1A, 2B,C and 5). Thus, our results suggest that while IL-33, IL-4 or IgE alone is insufficient to trigger mast cell degranulation without allergen, the accumulated number and activation status of mast cells, strong FceR signals derived from continued and intensive interaction between IgE and its high-affinity receptor and repetitive IL-33/ST2 and IL-4Ra signals may cumulatively or synergistically reach a signal threshold that leads to mast cell degranulation and AS.
IL-33 is clearly detected in various clinical allergic diseases including asthma and dermatitis (15,32,33). Given that many allergic reactions occur without a known allergen, including nonatopic asthma and idiopathic anaphylaxis (30,34), IL-33 may play a critical role in these diseases. Supporting this hypothesis, we have measured serum IL-33 levels in atopic and nonatopic patients with severe asthma by ELISA. Our results showed that asthma subjects have a wide range of serum IL-33 levels and that there was no significant difference between atopic and nonatopic subjects (Fig. S1). This result is consistent with our finding in the animal model, as well as with a recent report that IL33  Figure 6 Schematic representation of the mechanism of IL-33induced IgE synthesis and anaphylaxis in naive mice. IL-33 binding to ST2 + innate cells, such as mast cells and eosinophils, leads to mast cell proliferation and increased IL-4 synthesis. IL-4 would then activate B cells to proliferate and to produce IgE. IL-4 produced by the innate cells could also activate T cells to express CD40L that interacts with CD40 on B cells to further enhance IgE production. IgE, together with IL-33 and IL-4, would stimulate mast cells to degranulate, resulting in anaphylaxis.
polymorphism is correlated with clinical asthma but not necessarily atopy (35). Because IL-4 and IL-13 also play a pivotal role in antigenspecific IgE production, IL-33 may therefore have a general effect on all IgE-mediated allergic conditions. This concept suggests that IL-33 may represent a novel therapeutic target for a range of allergic diseases.

Disclosures
The authors have no financial conflicts of interest.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Figure S1. Serum IL-33 levels are not associated with atopy.
Data S1. Methods. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.