IL-33 induces HDC activity and histamine release in various tissues, particularly in bone
A single i.p. injection of IL-33 (50 mg/kg) into C57BL/6 mice increased HDC activity in several tissues (Fig. 1a). IL-33 increased HDC activity in bone (tibia), spleen, liver, and lung, peaking 2 to 4 hours after the injection (Suppl. Fig. 1a). The increase in HDC activity was dependent on the dose of IL-33, and bone was the most sensitive, with 12.5 mg/kg IL-33 inducing a significant increase in bone HDC activity (Suppl. Fig. 1b). We confirmed HDC expression in the bone marrow regions of bone (tibia) by immunochemistry (Fig. 1b) and by using HDC-GFP mice (Fig. 1c). IL-33 increased HDC-expressing cells from 7.3% to 23.2% of live cells (Fig. 1c). These results indicate that IL-33 increases HDC activity in various tissues, particularly potently in the hematopoietic organs bone and spleen, bone being the most sensitive and greatest HDC-inducing organ.
Consistent with the increase in HDC activity, IL-33 increased histamine in plasma (Fig. 1d). However, changes in the level of histamine induced by 50 mg/kg IL-33 were variable among the organs tested (Suppl. Fig. 1c). Interestingly, histamine was reduced significantly in the bone, while it increased or tended to increase in the other tissues. It was also notable that the levels of HDC activity and histamine in non-stimulated mice (i.e., non-injected or saline-injected) were highest in bone (Figs. 1a and Suppl. Fig. 1c). These results indicate that (i) injection of IL-33 into mice induces HDC in various tissues, especially in the hematopoietic organs (bone and spleen), (ii) the histamine newly formed by such HDC-induction is largely released without storage, and (iii) in non-stimulated mice, bone marrow has the highest levels of both HDC and histamine.
Single i.p. injection of IL-33 increases IL-5, IL-13, and GM-CSF in serum, but not IgE
As described above, a single i.p. injection of IL-33 (12.5-100 mg/kg) induces HDC in various tissues. Repeated i.p. injection of IL-33 into mice (20-200 mg/kg, once a day for 7 days) reportedly induces Th2 cytokines (such as IL-4, IL-5, and IL-13) and IgE via stimulation of ST2 (the receptor for IL-33) [2]. In vitro, IL-33 reportedly stimulates bone-marrow-derived eosinophils to produce Th2 cytokines, including GM-CSF [43]. Interestingly, it has also been reported that (i) IL-33 induces degranulation of mast cells under the presence of IgE [32, 33] (i.e., IL-33 may induce histamine release), and (ii) aggregated IgE induces HDC in vitro in murine hematopoietic progenitors (but not mature cells, including mast cells) [44]. As described in Introduction, HDC is induced by various cytokines, and IL-1 (either a or b) is especially potent at inducing HDC. Hence, we measured these cytokines and IgE in the serum after a single injection of IL-33 (50 mg/kg) and examined whether these cytokines induce increases in HDC activity. As shown in Fig. 2a, a single i.p. injection of IL-33 into mice increased IL-5, IL-13, and GM-CSF (but not IL-1b or IL-4) in the serum at 4 hours after the injection (the peak time for induced HDC activity). However, IgE was not increased by such a single injection.
IL-33 exerts its effects via stimulation of its receptor ST2
In the following experiments, the spleen was used for measuring HDC activity. As shown in Fig. 2b, injection of an antibody to ST2 completely prevented the induction of HDC activity by IL-33. Anti-ST2 antibody also prevented any IL-33-induced increase in IL-5, IL-13, or GM-CSF in the serum (Suppl. Fig. 2a).
IL-33 may induce HDC activity in hematopoietic organs directly and/or via mediation by other cytokines, including IL-5
As shown in Fig. 2c, among the IL-33-induced cytokines (see Fig. 2a), injection of GM-CSF and IL-5 increased HDC activity, while IL-13 did not. IL-33 increased HDC activity in C3H/HeJ mice [TLR4-mutated/inactive mice (i.e., non-responsive to LPS)] and in mice deficient in both IL-1a and IL-1b (IL-1-KO mice) (Figs. 2d and 2e), indicating that neither IL-1 nor contaminating LPS is involved in the induction by IL-33 of HDC. Thus, it seemed possible that the HDC induction by IL-33 might be mediated by GM-CSF and/or IL-5. However, as shown in Fig. 2f, anti-GM-CSF antibody had no significant effect on the IL-33-induced HDC activity, although it did tend to reduce it. On the other hand, anti-IL-5 antibody significantly (but not completely) reduced the IL-33-induced HDC activity (Fig. 2g). The anti-GM-CSF and anti-IL-5 antibodies used here were effective at blocking the respective cytokines produced by IL-33 (Suppl. Fig. 2b and 2c). These results suggest that IL-33 may induce HDC in hematopoietic organs directly and/or via mediation by other cytokines, including IL-5.
IL-33 increases HDC in hematopoietic cells in vitro and releases the newly-formed histamine without it being stored
As described above, IL-33 induces HDC especially in the hematopoietic organs bone and spleen (Fig. 1a). So, we examined whether IL-33 increases HDC activity and histamine in vitro in cells isolated from bone marrow and spleen. As shown in Fig. 3a, bone marrow cells exhibited greater increases in HDC activity and histamine than spleen cells. Thus, the following experiments were performed using bone marrow cells. As shown in Fig. 3b, HDC activity and histamine each increased in a manner dependent on the concentration of IL-33. In the following experiments, we used 100 ng/ml IL-33. As shown in Fig. 3c, the increase in HDC activity peaked within 24 hours, and histamine continued to accumulate in the culture supernatant throughout the 96 hours of incubation. Although histamine increased both in the cells and in the supernatant (Fig. 3d), its increase was greater in the supernatant. These results indicate that IL-33 induces HDC in hematopoietic cells and that most of the newly formed histamine is released from the cells without first being stored.
HDC-inducing cells in bone marrow in response to IL-33 are c-kit-positive cells
SCF is a cytokine essential for the proliferation of hematopoietic stem cells (HSCs), and c-kit is the receptor for SCF. Since bone marrow cells exhibited a greater ability to induce HDC than spleen cells (Fig. 3a), we compared the profiles of the cells expressing c-kit and ST2 between bone marrow and spleen. In the route underlying hematopoiesis, c-kit is expressed in HSCs and myeloid progenitor cells, while sca-1 (stem cell antigen-1) is expressed in HSCs but not in myeloid progenitor cells [45, 46]. In mice, lineage marker-negative (lin-) (see Methods) sca-1+ and c-kit+ cells [i.e., lin−sca-1+c-kit+ cells (called LSK cells)] are used as HSCs. LSK cells differentiate to lin−sca-1−c-kit+ cells (called LK cells). Both LK cells and neutrophils (CD45+CD11b+CD11c−Gr-1+) were more abundant in the bone marrow than in the spleen (Fig. 4a and 4b), and ST2 was observed only in LK cells (Fig. 4c). Hence, we examined whether LK cells include HDC-inducing cells. IL-33 increases HDC-expressing cells in LK cells from 21.2% to 38.3% (Fig. 5a). Next, c-kit+ cells were depleted from bone marrow cells (see Methods) and the c-kit+ cell-depleted bone marrow cells were stimulated with IL-33 for 72 hours. As shown in Fig. 5b, IL-33-induced HDC activity was markedly reduced by such depletion of c-kit+ cells. W/Wv mice are mast-cell-deficient mice due to c-kit mutation, and these mice lack LK cells, too (Fig. 5c). As shown in Fig. 5d and Suppl. Fig. 3a, IL-33-induced HDC activity in the bone and spleen was much smaller in W/Wv mice than in control +/+ mice. These results indicate that the HDC-inducing cells in bone marrow in response to IL-33 are c-kit+ cells (i.e., cells expressing SCF receptors).
Other HDC-inducing cells in response to IL-33
Mast cells: In mice, mast cells are the main source of histamine in skin, lung, thymus, spleen [47], skeletal muscle [48], and in bone marrow-removed bone [40]. So, we examined whether IL-33 induces HDC in mast cells. As shown in Fig. 1a, HDC activity is detected at a significant level in the ear-pinnae of saline-injected (i.e., non-stimulated) mice and is augmented by IL-33. Mast cells in WT ear-pinnae express the ST2 receptor (Fig. 6a). Experiments using HDC-GFP mice demonstrated that IL-33 induced HDC in ear-pinnae without affecting the frequency of HDC-expressing cells in mast cells (Fig. 6b), indicating that IL-33 induces HDC mostly in the same mast cells. In the skin (ear-pinna) of W/Wv mice, mast cells were few (Fig. 6c), and IL-33 did not increase HDC activity at all in the ear-pinna of W/Wv mice (Fig. 6d). Mast cells purified from bone marrow (BMMCs) (see Methods) also express ST2 (Fig. 6f), and HDC activity and histamine were markedly increased by in vitro stimulation of BMMCs with IL-33 (Fig. 6g). These results clearly indicate that IL-33 can induce HDC in mature mast cells in various tissues, as well as skin.
Other cells: It was notable that in W/Wv mice, IL-33 augmented the low, but significant, HDC activity present in bone (Fig. 5d), lung, and spleen (Suppl. Fig. 3a), indicating that HDC is induced by IL-33 in cells other than mast cells and c-kit-positive cells. It has been reported that IL-33 directly activates basophils and induces degranulation [35]. So, we examined basophils. In mice depleted of their basophils by an antibody specific for mouse basophils (Ba103), the HDC activity induced by IL-33 in the spleen was essentially the same as that induced in mice not depleted of their basophils (Suppl. Figs. 3b and 3c), suggesting that basophils may be not involved in HDC induction by IL-33. Thus, it remains to be clarified which cell-types, in addition to mast cells and c-kit+ cells, can exhibit HDC-induction in response to IL-33.
Comparison of IL-33 with other HDC-inducing cytokines
In the present study, we found that IL-33 induced HDC activity in various tissues, but particularly potently in hematopoietic organs (bone and spleen). As described in Introduction, IL-1 and LPS are potent HDC inducers, and induce HDC in a number of tissues. So, here, we compared induced HDC activities between hematopoietic (bone and spleen) and non-hematopoietic (lung and liver) tissues in mice injected i.p. with IL-33, IL-1, and LPS. In this study, we also compared the effects of other cytokines (IL-3, IL-5, and IL-13). As shown in Suppl. Fig. 4, their abilities to induce HDC were as follows. In the bone: IL-33 ³ LPS ³ IL-3 ³ IL-1b > IL-5; and in the spleen: LPS ³ IL-1b ³ IL-3 ³ IL-33 ³ IL-5. In the lung, the ability of IL-33 was much lower than those of IL-1b and LPS. In the liver, the abilities of IL-33, IL-1b, and IL-3 were similar to each other, while LPS was much more potent. IL-5 did not induce HDC in the lung or liver, and IL-13 induced HDC in none of the tissues examined. Thus, LPS is the most potent HDC-inducer in tissues other than bone. In this study, we found that like IL-33, IL-3 can induce HDC in non-hematopoietic tissues, too.
IL-33-induced newly formed histamine negatively regulates eosinophilia partly via H4Rs
Repeated i.p. injection of IL-33 into mice (once a day for 7 days) has been reported to induce eosinophilia [2]. Such a treatment of mice with IL-33 is also reported to induce a systemic increase in IL-5, and an anti-IL-5 antibody ablates the IL-33-induced eosinophilia [49]. As described above, IL-33 increases HDC activity in various tissues (including bone marrow and spleen) and IL-5 was also shown to increase HDC activity in the spleen. Finally, therefore, we examined whether HDC/histamine is involved in IL-33-induced eosinophilia. Mice were i.p. injected with IL-33 and/or histamine every other day for a total of 3 times, and eosinophils in the bone marrow and IL-5 in the serum 24 hours after the last injection were analyzed (Fig. 7a). IL-33-induced eosinophilia was much greater in HDC-KO mice than in WT mice, and this effect was markedly reduced by the addition of histamine (Figs. 7a and c). This IL-33-induced eosinophilia in the bone marrow of WT mice was slightly but significantly augmented by the H4R antagonist JNJ7777120 (Figs. 7b and c). In addition, IL-33 increased the serum level of IL-5 (Fig. 7d). Although histamine and histamine-receptor antagonists did not influence this level in WT mice, the IL-33-induced serum IL-5 level was greater in HDC-KO mice than in WT mice, and this elevated level was markedly reduced by the addition of histamine (Fig. 7d). These results indicate that (i) IL-33-induced newly formed histamine suppresses or negatively regulates IL-33-induced IL-5-mediated eosinophilia, and (ii) H4Rs are partly involved in this effect of histamine.