Chitin promotes antigen-specific Th2 cell-mediated murine asthma through induction of IL-33-mediated IL-1β production by DCs

Chitin, which is a major component of house dust mites (HDM), fungi, crustaceans, etc., can activate immune cells, suggesting that it contributes to development of allergic disorders such as asthma. Although the pathophysiological sensitization route of asthmatic patients to allergens is considered via the respiratory tract, the roles of intranasally-administered chitin in development of asthma remain unclear. After ovalbumin (OVA) challenge, development of airway inflammation was profoundly exacerbated in mice sensitized with OVA in the presence of chitin. The exacerbation was dependent on IL-33, but not IL-25, thymic stromal lymphopoietin or IL-17A. Chitin enhanced IL-33-dependent IL-1β production by dendritic cells (DCs). Furthermore, chitin- and IL-33-stimulated DC-derived IL-1β promoted OVA-specific Th2 cell activation, resulting in aggravation of OVA-induced airway inflammation. These findings indicate the adjuvant activity of chitin via a new mechanism and provide important clues for development of therapeutics for allergic disorders caused by HDM, fungi and crustaceans.

even in the absence of acquired immune cells such as T cells and B cells, mast cells and STAT6-mediated signal transduction 11-13 . In addition, chitin particles can enhance Th2 cell-mediated airway eosinophilia as an adjuvant, like alum does. That is, intraperitoneal administration of ovalbumin (OVA) to mice in the presence of chitin particles enhanced activation of ovalbumin (OVA)-specific Th2 cells as well as Th1 cells and Th17 cells, contributing to development of OVA-induced Th2 cell-mediated airway inflammation 14 . TLR2 and IL-17 are required for the effect of chitin as an adjuvant in the setting. However, it remains unclear how IL-17, which is produced by macrophages through interaction of chitin with TLR2, contributes to Th2 cell-mediated airway eosinophilia.
In general, the route of sensitization of patients with asthma to allergens such as HDM is considered to be via the respiratory tract and/or skin, but not the peritoneum. Therefore, the effect of intranasally-administered chitin on Th2 cell-mediated airway inflammation needs to be elucidated.
In the present study, we show that even via the intranasal route chitin acts as an adjuvant on OVA-induced Th2 cell-mediated airway inflammation, independently of IL-17. We also found a novel pathway for enhancement of Th2 cell-mediated airway inflammation by chitin: chitin induced IL-33 in the lungs, followed by augmentation of IL-33-mediated IL-1β production by dendritic cells (DCs). Subsequently, that IL-1β promoted OVA-specific Th2 cell expansion, resulting in aggravation of OVA-induced Th2 cell-mediated airway inflammation.

Results
Chitin shows potent adjuvant effect on allergic airway inflammation. To investigate the effect of chitin on pulmonary immune responses, mice were treated intranasally with OVA or saline in the presence and absence of chitin (Fig. 1a). The mice were then intranasally challenged with OVA (Fig. 1a). One day after the last OVA challenge, the number of eosinophils, but not other types of cells, was profoundly increased in BALFs from mice sensitized with OVA in the presence of chitin compared with the mice sensitized with OVA alone, chitin alone or saline (Fig. 1b). Consistent with this, after the last OVA challenge, histological analysis showed airway inflammation in the mice sensitized with OVA in the presence of chitin, but not in mice sensitized with OVA alone, chitin alone or saline (Fig. 1c). In association with this, after the last OVA challenge, the levels of OVAspecific IgE, IgG 1 and IgG 2a were significantly increased in sera from mice sensitized with OVA in comparison with mice treated with chitin alone or saline (Fig. 1d). The levels of OVA-specific IgE and IgG 1 , but not IgG 2a , were approximately 2-to 3-fold increased in sera from mice sensitized with OVA in the presence of chitin compared with mice sensitized with OVA alone (Fig. 1d).
To investigate the effect of chitin on antigen-specific Th2 cell function during OVA-induced airway inflammation, spleen cells from mice sensitized with OVA in the presence and absence of chitin and from mice treated with chitin alone or saline were cultured in the presence of OVA. The levels of IL-4 and IL-13 in the culture supernatants of spleen cells from the OVA-sensitized mice were comparable to those from mice treated with chitin alone or saline (Fig. 1e). On the other hand, the supernatant IL-4 and IL-13 levels of mice sensitized with OVA in the presence of chitin were markedly increased compared with mice sensitized with OVA alone, chitin alone or saline (Fig. 1e). In contrast to the responses of the above Type-2-cytokines, the level of a Type-1-cytokine (IFN-γ) was significantly reduced in the culture supernatants of spleen cells from mice sensitized with OVA compared with mice treated with chitin alone or saline (Fig. 1e). However, this reduction was not influenced by the presence of chitin (Fig. 1e). The supernatant level of a Type-3-cytokine (IL-17) from mice sensitized with OVA in the presence and absence of chitin was not altered when compared with that from mice sensitized with chitin alone or saline (Fig. 1e). In addition, the levels of IL-5 and IL-13 in the BALFs were significantly increased in mice sensitized with OVA in the presence of chitin, but not in the other groups, after the last OVA challenge (Fig. 2a). On the other hand, the levels of IFN-γ and IL-17 in the BALFs were not increased in any of the groups (Fig. 2a). It is thought that Th2 cells and group 2 innate lymphoid cells (ILC2) may be potential sources of type 2 cytokines in the lungs. We next examined the detection of IL-13 producing cells in the lungs and BALFs in each group. We found that the proportion and number of IL-13 + Th2 cells were significantly increased in both the lungs and BALFs from mice sensitized with OVA in the presence of chitin compared with the other groups (Fig. 2b,c). Although IL-13 + ILC2 were significantly detected in the lungs, but not the BALFs, from mice sensitized with OVA in the presence of chitin compared with the other groups (Fig. 2b,c), the number of IL-13 + ILC2 was relatively smaller than that of IL-13 + Th2 cells (Fig. 2b,c). In addition, airway inflammation was not observed in T/B cell-deficient Rag2 −/− mice (Supplemental Fig. 1). These observations suggest that the major source of IL-13 in our model was Th2 cells rather than ILC2, although we could not rule out a possible contribution of ILC2 in the setting. Taken together, these observations strongly suggest that chitin, which was administered by the intranasal route, has the potential to exert an adjuvant effect on Type 2-cytokine-associated immune responses during OVA-induced airway inflammation.
As in the case of alum, "intraperitoneal" administration of chitin particles facilitated OVA-induced airway inflammation 14 , but inhibited ragweed antigen-induced airway inflammation in mice 15 . The effects of chitin on T cell-mediated allergic inflammation may differ as a function of the size of its particles 9,10,14,15 . We thus investigated the effect of chitin particle size in our model, as shown in Fig. 1a. The number of eosinophils in the BALFs was similarly increased in wild-type mice sensitized with OVA in the presence of all sizes of chitin particles fractionated into <40 µm, 40-70 µm or 70-100 µm in diameter (Fig. 3a). These data show that any size of chitin particles less than 100 µm in diameter in our preparation has similar potency as an adjuvant of induction of allergic airway inflammation by sensitization with OVA via the intranasal route.
It was also reported that TLR2-dependent IL-17 production is important for development of OVA-induced airway inflammation in mice sensitized "intraperitoneally" with OVA in the presence of chitin particles (40-to 70-µm in diameter) 14 . Therefore, we examined the contribution of IL-17 to induction of OVA-induced airway inflammation in Il17a −/− mice sensitized "intranasally" with OVA in the presence of various sizes of chitin particles. However, regardless of the chitin particle size, the number of eosinophils in BALFs from those mice was only slightly, without significance, increased compared with wild-type mice (Fig. 3a). These data suggest that IL-17 is not essential for induction of allergic airway inflammation in mice sensitized "intranasally" with OVA in the presence of various sizes of chitin particles. That finding suggests that different mechanisms underlie the adjuvant effects of chitin in the sensitization to OVA when administered intranasally and intraperitoneally. Therefore, we hereafter used <100-µm-diameter chitin particles to investigate the detailed mechanism of the adjuvant effect of chitin via the intranasal route. It is well established that the development of airway inflammation induced by "intraperitoneal" sensitization with OVA emulsified with alum is dependent on the IL-4/IL-13-IL-4Rα-STAT-6 pathway 16,17 . On the other hand, it was also reported that airway eosinophilia was induced independently of the IL-4/IL-13-IL-4Rα-STAT-6 pathway in certain settings, i.e., sensitization with OVA in the presence of a high dose of dsRNA 18 and inhalation of fungal antigens 19 . Therefore, it is important to clarify whether development of airway eosinophilia in mice sensitized with OVA in the presence of chitin requires IL-4/IL-13-IL-4Rα-STAT-6 signaling. Accordingly, we examined the role of that signaling in Stat6 −/− mice and Il4 −/− Il13 −/− mice. As expected, the eosinophil count and the EPO activity level in BALFs from wild-type mice sensitized with OVA in the presence of chitin were increased compared with wild-type mice sensitized with OVA in the absence of chitin or with saline alone (Fig. 3b). On the other hand, such increased responses were not observed in Stat6 −/− mice or Il4 −/− Il13 −/− mice (Fig. 3b). After the last challenge, the levels of total and OVA-specific serum IgE in wild-type mice, but not Stat6 −/− mice or Il4 −/− Il13 −/− mice, sensitized with OVA in the presence of chitin were markedly increased in comparison with in wild-type, Stat6 −/− mice and Il4 −/− Il13 −/− mice sensitized with OVA in the absence of chitin and/or with saline alone (Fig. 3c). In addition, the levels of IL-4 and IL-3 produced by OVA-stimulated spleen cells from Stat6 −/− and Il4 −/− Il13 −/− mice that had been sensitized with OVA in the presence of chitin were impaired and diminished, respectively, in comparison with wild-type mice (Fig. 3d). Taken together, it can be surmised that, in contrast to the special settings noted above 18,19 , IL-4/IL-13 and STAT-6 are required for induction of Type 2-cytokine-mediated immune responses in mice sensitized with OVA in the presence of chitin.
IL-33, but not IL-25 or TSLP, is required for allergic airway inflammation in the presence of chitin. IL-25, IL-33 and TSLP, which are considered to be produced by pulmonary epithelial cells, are well known to be involved in promoting Type-2 immune responses, including IL-4/IL-13-IL-4Rα-STAT-6-dependent allergic airway inflammation, through induction of Type-2 cytokine production by various types of cells and alteration of DC phenotypes 20,21 . In addition, these cytokines were increased in the lungs after inhalation of chitin alone 13 . In our model, after the last OVA challenge, the number of eosinophils in BALFs and the production of IL-13, but not IL-4 (data not shown), by OVA-stimulated spleen cells from Il33 −/− , but not Il25 −/− or Crlf2 −/− , mice sensitized with OVA in the presence of chitin were significantly reduced in comparison with wild-type mice (Fig. 4a,c). On the other hand, the levels of OVA-specific serum IgE were not altered in any of those strains of mice after the last OVA challenge (Fig. 4b). These observations indicate that IL-33, but not IL-25 and TSLP, is crucial for induction of OVA-induced airway eosinophilia in the presence of chitin.
Chitin abolishes tolerogenic immune responses. Inhalation of OVA prior to intraperitoneal sensitization with OVA emulsified with alum induces immune tolerance against OVA by generating tolerogenic pulmonary DCs, resulting in inhibition of development of OVA-induced airway inflammation by expansion of IL-10-producing Treg cells in mice 22 . On the other hand, the tolerance induction by inhalation of OVA is also known to be interfered with by persistent activation of pulmonary DCs by pathogens such as influenza virus 23,24 . Therefore, we investigated the effect of chitin on induction of tolerance by prior inhalation of OVA in mice (Fig. 5a). The number of eosinophils in BALFs, level of anti-OVA IgE in sera and production of IL-4 and IL-13 by OVA-stimulated spleen cells from wild-type mice that inhaled OVA prior to sensitization with OVA emulsified with alum was significantly reduced in comparison with wild-type mice that inhaled saline (Fig. 5b-d). On the other hand, interestingly, the tolerogenic responses induced by prior inhalation of OVA were partially but significantly restored in mice that inhaled OVA in the presence of chitin prior to sensitization with OVA emulsified with alum ( Fig. 5b-d). These data suggest that chitin can promote immune responses by blocking induction of tolerogenic immune responses to harmless aero-antigens, leading to subsequent development of allergic airway inflammation.
Chitin enhances OVA-induced airway eosinophilia by promoting IL-33-induced IL-1β production by DCs. How can chitin break tolerance? It is known that IL-1 can break tolerance of T cells and B cells to certain antigens [25][26][27] . Therefore, we investigated the effect of chitin on IL-1β production by DCs after inhalation of chitin, OVA, or OVA + chitin. Pulmonary DCs migrated into draining LNs after inhalation of FITC-conjugated OVA 28,29 . We found that the proportion and number of IL-1β + DCs among 7-AAD − CD45 + MHCII hi CD11c + cells in draining LNs were comparable in each of naïve, OVA-treated and chitin-treated wild-type mice (Fig. 6ac). On the other hand, that proportion and number were significantly increased in OVA + chitin-treated wild-type mice compared with the other groups (Fig. 6a-c), suggesting that chitin is a potent activator of DCs in the presence of OVA. IL-1β + DCs could not be detected in Il1a −/− Il1b −/− mice in the settings (Supplemental Fig. 2).
As shown in Fig. 4, Il33 −/− mice showed attenuation of induction of OVA-induced airway eosinophilia in the presence of chitin, indicating that IL-33 is crucial in the setting. Interestingly, the proportion and number of IL-1β + DCs among 7-AAD − CD45 + MHCII hi CD11c + cells in draining LNs was significantly reduced in Il33 −/− mice compared with wild-type mice (Fig. 6d-f), suggesting that IL-33 induced by chitin may be involved in IL-1β production by DCs in the presence of OVA. To elucidate this, we cultured bone marrow-derived DCs (BMDCs) in the presence and absence of IL-33, with and without chitin. BMDCs produced IL-1β, IL-6 and TNF in response to IL-33, but not chitin (Fig. 7a). On the other hand, chitin enhanced IL-33-dependent IL-1β, but not IL-6 or TNF, production by BMDCs (Fig. 7a). These findings suggest that excessive IL-1β production by IL-33-stimulated DCs in the presence of chitin may be involved in the breaking of tolerance, resulting in excessive T-cell activation in response to allergens. Therefore, we next investigated the effect of IL-1β, produced by IL-33-stimulated BMDCs in the presence of chitin, on OVA-specific T-cell activation. BMDCs from wild-type mice were first incubated with or without IL-33 in the presence and absence of chitin, and then co-cultured with CD4 + T cells from OTII mice, which express OVA-specific TCR, in the presence of OVA peptides. As shown in Fig. 7b, the level of IL-13, but not IL-17 or IFN-γ, in the culture supernatants of the CD4 + OTII cells co-cultured with IL-33-and chitin-stimulated wild-type BMDCs were significantly increased compared with unstimulated, IL-33-stimulated and chitin-stimulated wild-type BMDCs. However, IL-13 production was not increased in the supernatants of the CD4 + OTII cells co-cultured with IL-33-or the chitin-stimulated Il1a −/− Il1b −/− BMDCs (Fig. 7b). These observations suggest that chitin can promote IL-33-induced IL-1β production by DCs, followed by enhancement of antigen-specific Th2-cell activation, in vitro.
In order to elucidate whether IL-1β production by DCs in response to IL-33 induced by chitin is important for induction of allergic airway eosinophilia in vivo, we established OVA-induced airway eosinophilia in the presence of chitin in Il1rl1 −/− mice transferred with wild-type or Il1rl1 −/− DCs (wild-type DCs → Il1rl1 −/− mice;  Fig. 1a. The number of eosinophils in BALFs was significantly increased in wild-type DCs → Il1rl1 −/− mice sensitized with OVA in the presence of chitin in comparison with wild-type DCs → Il1rl1 −/− mice sensitized with OVA alone at 24 h after the last OVA inhalation (Fig. 7c). On the other hand, the number of eosinophils in BALFs from Il1rl1 −/− DCs → Il1rl1 −/− mice was significantly reduced in comparison with wild-type DCs → Il1rl1 −/− mice in the setting (Fig. 7c). These observations indicate that activation of IL-1RL1/ST2-expressing DCs in response to IL-33 induced in the lungs after respiratory exposure to chitin is crucial for induction of OVA-induced airway eosinophilia. The number of eosinophils was also reduced in BALFs from Il1a −/− Il1b −/− DCs → Il1rl1 −/− mice compared with wild-type DCs → Il1rl1 −/− mice (Fig. 7c). These observations indicate that chitin-induced IL-33 stimulates DCs to produce IL-1β, which is involved in induction of OVA-induced airway eosinophilia in vivo.

Discussion
Chitin, β-(1-4)-poly-N-acetyl D-glucosamine, is known to be a common component of the exoskeleton of arthropods, such as house dust mites, crabs, shrimp and insects, and the microfilarial sheath of parasitic nematodes [1][2][3][4][5] . It was reported that inhalation of chitin alone resulted in induction of airway eosinophilia in mice 11,13 , suggesting that chitin may be somehow involved in the pathogenesis of HDM-mediated asthma. Induction of type 2-cytokine-mediated airway inflammation by ragweed antigen was inhibited by intraoral administration of chitin particles (1-to 10-µm in diameter) in the antigen sensitization phase by enhancing IFN-γ production 15 . Development of type 2-cytokine-mediated eosinophilia and IL-17-mediated neutrophilia during OVA-induced airway inflammation was promoted by intraperitoneal administration of chitin particles (40-to 70-µm in diameter) in the antigen sensitization phase by promoting IL-17 production by macrophages via TLR2 and/or Dectin-1 9,10,14 . Those findings suggest that the effects of chitin on T cell-mediated allergic inflammation differ as a function of the size of chitin particles.
IL-33, IL-25 and TSLP, which are preferentially produced by lung epithelial cells, are considered to be involved in development of type 2 cytokine-associated allergic airway inflammation 20,[30][31][32] . They have been reported to be crucial for induction of OVA-induced airway inflammation in mice sensitized "intraperitoneally" with OVA in the presence of alum 29,33,34 . On the other hand, only IL-33, but not IL-25 or TSLP, was important for induction of HDM-induced airway inflammation in mice sensitized "intranasally" with HDM 35,36 . Similar to the latter observation, we demonstrated that IL-33, but not IL-25 or TSLP, is responsible for induction of OVA-induced airway inflammation in mice sensitized "intranasally" with OVA in the presence of chitin. In addition, in the present study, we found that the administration route of chitin affects the development of OVA-induced airway inflammation. We demonstrated that chitin administered intranasally has the potential to exert an adjuvant effect on Type 2-cytokine-associated immune responses during OVA-induced airway inflammation. Especially, the mechanisms for the effect of chitin via the intranasal route differed from those via the intraperitoneal route, as reported previously 14 . That is, the mechanisms for the effect of chitin via the intranasal route are mediated in an IL-4/IL-13-STAT6-dependent, but IL-17A-independent, manner. On the other hand, it was reported that the mechanisms via the intraperitoneal route are mediated in an IL-17A-dependent manner 14 .  In addition, inhalation of chitin in mice resulted in increased IL-33 expression in the lungs 13 . In association with this, we found that 1) IL-33 induced IL-1β production by DCs, 2) chitin promoted IL-33-dependent IL-1β production by DCs, and 3) DC-derived IL-1β enhanced antigen-specific Th2 cell activation (Fig. 8), contributing to induction of Th2 cell-mediated IL-4/IL-13-STAT6-dependent allergic airway inflammation.
IL-33 was reported to induce cytokine production (i.e., IL-1β, IL-6 and TNF) and enhance co-stimulatory molecule expression (i.e., CD40, CD80 and OX40L) in DCs, followed by enhancement of Th2 cell differentiation via that activation of DCs 37 . In addition, adoptive transfer of OVA-pulsed, IL-33-stimulated DCs resulted in aggravated OVA-induced airway inflammation compared with OVA-pulsed, unstimulated DCs in mice 37 . We found that chitin alone did not induce cytokine production (i.e., IL-1β, IL-6 and TNF), but it enhanced IL-33-induced IL-1β production, by DCs (Fig. 6a). On the other hand, chitin did not enhance CD40 or OX40L expression on IL-33-stimulated DCs (data not shown). It is known that IL-1 can break the tolerance of T cells and B cells to certain antigens [25][26][27] , suggesting that excessive IL-1β production by DCs in response to IL-33 in the presence of chitin may be involved in breaking the tolerance of T cells to allergens, such as OVA used in this study. It is well known that respiratory exposure to allergens leads to tolerance of T cells to those allergens, contributing to protection against induction of allergic asthma through induction of tolerogenic pulmonary DCs in non-atopic individuals 22 . On the other hand, it was reported that such tolerance to allergens was canceled by concurrent viral infection, such as influenza A, resulting in provocation of allergic airway inflammation 23,24 . Likewise, we found that inhalation tolerance to OVA in a mouse model 22 was abolished by concurrent inhalation of chitin. We also demonstrated that IL-1β derived from DCs in response to IL-33 and chitin is important for Th2 cell activation in the sensitization phase, which is involved in aggravation of allergic airway inflammation.
In conclusion, we demonstrated that intranasal administration of chitin induced IL-33 in the lungs, and IL-1β production by DCs in response to IL-33 is important for Th2 cell polarization, resulting in aggravation of OVA-induced Th2 cell-mediated allergic airway inflammation that is dependent on an IL-4/ IL-13-STAT6 pathway. These observations suggest that intranasal exposure to chitin can potentially act as a type 2-cytokine-associated natural adjuvant, contributing to the pathogenesis of allergic disorders induced by a variety of allergens, including HDM.

Methods
Mice. BALB/c-and C57BL/6-wild-type mice were obtained from Japan SLC.  Preparation of chitin suspension. Chitin (Dextra Laboratories) was treated and reduced in size using an established procedure 47 . In brief, chitin was partially hydrolyzed with concentrated hydrochloric acid for 30 min at 30 °C, followed by neutralization with 1 M Tris-HCl buffer (pH 8.0) and centrifugation at 3,000 rpm for 5 min at room temperature. After washing with saline, the chitin was routinely passed through 100-µm nylon mesh filter, or fractionated by size-exclusion nylon mesh filters into 70-100 µm, 40-70 µm and <40 µm diameter particles. The chitin was resuspended in saline at the appropriate concentration, sterilized by autoclaving and stored at 4 °C.
Animal model of allergic asthma. As shown in Fig. 1a, mice were intranasally treated with 10 µg OVA (grade V; Sigma) in saline in the presence and absence of 10 µg chitin in 20 µl of saline or saline alone on each of seven alternate days (days 0, 2, 4, 6, 8, 10 and 12). On days 28, 31 and 34, the mice were intranasally challenged with 100 µg OVA in 20 µl of saline. One day after the last challenge, sera, lungs, bronchoalveolar lavage fluids (BALFs) and spleens were harvested. Bronchoalveolar lavage (BAL) cell analysis. BALFs were collected as described elsewhere 48 . The total cell count and leukocyte profile were determined with a hemocytometer (XT-1800i; Sysmex), as described previously 34 . Histology. Lungs were harvested, fixed in 10% neutral buffered formalin and embedded in paraffin.
Three-µm-thick lung sections were stained with hematoxylin and eosin (HE).
ELISA for immunoglobulins. OVA-specific IgE, IgG 1 and IgG 2a levels in sera were determined by ELISA, as described elsewhere 34 . The data were normalized to the value for arbitrary serum obtained from OVA-immunized mice as a standard calibrator. Mouse IgE ELISA Quantitation Kit (Bethyl Laboratories) was used to measure the total IgE in sera.
Spleen cell culture. After elimination of RBCs with an RBC lysing buffer (Sigma), spleen cells were suspended in RPMI1640 medium supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin. Spleen cells (4 × 10 6 cells/ml) were cultured in the presence of 100 µg/ml OVA in 24-well flat-bottom plates for 4 d. Inhalation tolerance model. An animal model of tolerogenic immune responses induced by respiratory exposure to antigen was previously reported 22 . In brief, mice were exposed intranasally to 100 μg of OVA in 20 μl of saline in the presence and absence of 100 μg of chitin (<100 µm-diameter), or saline alone, on 3 consecutive days as shown in Fig. 4a. Ten days after the last inhalation, the mice were immunized intraperitoneally with 10 μg of OVA emulsified in 2 mg of alum hydroxide (alum). Two weeks later, the mice were challenged with 50 μg of OVA in 20 μl of saline, or saline alone, intranasally on 3 consecutive days. One day after the last challenge, sera, BALFs and spleens were harvested.