IL ‐33 drives airway hyper‐responsiveness through IL ‐13‐mediated mast cell: airway smooth muscle crosstalk

1Asthma is a chronic inflammatory disorder of the airways characterized by inflammation, variable airflow obstruction and AHR. It is estimated that asthma affects 5–10% of adults. Genomewide association studies have consistently implicated IL-33 as an asthma susceptibility gene. IL-33 and its receptor ST2 function as an alarmin to alert the immune system after endothelial or epithelial cell damage during trauma, physicochemical stress or infection. IL-33 increased AHR in vivo and in lung slices and plays a key role in rhinovirus-induced asthma exacerbations. Additionally, disruption of IL-33/ST2 signalling in vivo during the course of experimental asthma or anaphylaxis reduced the severity of disease.

2In asthma, mast cell–ASM interactions are important in the development of disordered airway physiology. ASM cells from asthmatics express elevated levels of IL-33 compared to healthy subjects, and mast cells respond to IL-33 activation. We hypothesize that the IL-33/ST2 axis plays a role in mast cell–ASM interactions in asthma. We show that IL-33 expression was increased in the bronchial epithelium and ASM in asthma. IL-33 promoted ASM wound repair directly and, in an autocrine manner, augmented mast cell mediator release and, indirectly, increased ASM contraction following coculture with mast cells via upregulation of mast cell-derived IL-13. Similarly, in an in vivo mouse model of intratracheal cytokine installation, IL-33 induced AHR which was IL-13 dependent. Therefore, IL-33 may present an important target to modulate mast cell–ASM crosstalk in asthma.

3A more detailed methods section is provided in the supplement.

4Asthmatic subjects had a consistent history and evidence of asthma. The study was approved by Leicestershire Ethics Committee. All patients gave their written informed consent.

5Primary human ASM cells, human lung mast cells (HLMC), human epithelial cells and the human mastocytoma cell-1 (HMC-1) cell line were isolated and cultured as previously described.

6Lungs were taken from BALBc (8- to 12-week-old) and C57BL6 (16- to 24-week-old) mice for precision cut lung slicing (PCLS).

7Bronchial biopsy sections were stained for IL-33 and assessed using a semi-quantitative intensity score (SQS) and quantitative thresholding.

8IL-33 and ST2 expression was assessed by flow cytometry and immunofluorescence. Cells were counterstained with 4′,6′-diamidino-2-phenylindole (DAPI).

9Quantitative RT-PCR of ST2L, ST2 and IL-13 was performed and compared against the internal reference gene 18S.

10IL-33 and IL-13 concentrations were quantified by ELISA.

11The ratio of fluo-3/fura red within cells vs time was measured by flow cytometry. Following baseline measurements (1 min), cell flow was halted, IL-33 or calcium ionophore added, and data acquired for a further 3 min.

12ASM cells were treated as indicated in Fig. S1. The CellTiter 96 Aqueous One Solution was added as per the manufacturer's instructions. Apoptosis was assessed by DAPI staining of nuclear morphology and annexin-V ± propidium iodide staining according to manufacturer's protocol.

13ASM cells ± HLMC (4:1 ratio) were impregnated into collagen gels. Gel surface area was measured using ImageJ (http://rsb.info.nih.gov/ij).

14Cytokines and chemokines were measured in cells ± IL-33 by electrochemiluminescence detection (Mesoscale Discovery, Gaithersburg, Maryland).

15ASM cells ± IL-33, isotype control or anti-IL-33-neutralizing antibody were wounded as described previously. Wounds were photographed at baseline and after 18 h. Wound repair was analysed using cellF software.

16Histamine was measured in supernatants from activated HLMC (anti-FcεR1 antibody or IL-33 for 24 h) or HLMC incubated with ASM (1:4 ratio) for 5–11 days ± IL-33, isotype control or anti-IL-33-neutralizing antibody, as previously described.

17PCLS were prepared as described previously. Images were captured at baseline, then every 5 min for cumulative carbachol dose responses and 2–10 min post mouse IL-33. Airway lumen size was measured using ImageJ software.

18BALBc mice were dosed intranasally with three repeated doses of murine IL-33 (5 μg). Post 3 days, cell number in the lung tissue was assessed by lung digest. AHR was measured using a flexiVent system ± neutralization of IL-13 activity using fusion protein (IL-13Rα2) administered 2 h prior to each IL-33 administration. IL-13, Gob-5 and Muc5AC mRNA expression was determined by RT-PCR, and mouse serum mMCP-1 in serum by ELISA.

19Statistical analysis was performed using GraphPad PRISM using parametric and nonparametric tests as appropriate. A P < 0.05 was considered significant.

20IL-33 expression was identified within the ASM-bundles in most subjects with variable intensity of expression, and within the epithelium (Fig.1A). Mast cells within the ASM-bundle infrequently coexpressed IL-33 (data not shown). The semi-quantitative intensity score (SQS) for IL-33 expression was significantly increased in mild–moderate asthma compared to healthy controls (Kruskal–Wallis P = 0.033; post hoc Dunn's pairwise comparison P = 0.046, Fig.1B). The correlation between SQS IL-33 ASM expression and AHR was good (r = −0.63, P < 0.001, Fig.1C). There was no significant correlation between IL-33 expression and FEV1% predicted, bronchodilator reversibility, atopic status or sputum cell counts (data not shown). Epithelial IL-33 expression was also significantly increased in mild-to-moderate asthma compared to healthy controls (Kruskal–Wallis P = 0.047; post hoc Dunn's pairwise comparison P = 0.041, Fig.1D). The SQS and quantitative expression using thresholding were correlated for both ASM (r = 0.63, P = 0.004) and epithelium (r = 0.43, P = 0.013). Quantitative IL-33 expression was increased in the ASM and epithelium in asthmatics compared to healthy subjects, but did not reach statistical significance. Quantitative IL-33 expression in ASM correlated with AHR in those with asthma (r = −0.52, P = 0.007). The clinical characteristics of the subjects are shown in Table 1.

21IL-33 expression was identified in human ASM, HLMC, HMC-1 and epithelial cells by immunofluorescence (Fig.2A) and flow cytometry (Fig.2B,C). IL-33 expression was not different between ASM cells derived from asthmatic subjects compared to healthy controls (data not shown). IL-33 was spontaneously released from ASM, HLMC, HMC-1 and epithelial cells as measured by ELISA after 24 h (Fig.2D).

22mRNA expression of IL-33 receptors ST2L (long transducing isoform) and ST2 (short decoy soluble form) was evident in mast cells, but not in ASM cells (Fig.2E). A two- and threefold increase in ST2L and ST2 mRNA expression were observed respectively in HLMC following IL-33 stimulation (50 ng/ml, 24 h, Fig.2F), but not in ASM cells (n = 3, data not shown). Although ST2 cell surface expression was not identified in unstimulated ASM, HLMC and HMC-1 cells by flow cytometry, IL-33 stimulation (50 ng/ml, 48 h) significantly upregulated ST2 surface expression (Fig.2G). Total cell ST2 expression was apparent in all unstimulated cell types and increased poststimulation with IL-33 (Fig.2G).

23IL-33 (50–200 ng/ml) triggered Ca2+ flux with increased intracellular calcium in ASM and HMC-1 cells as indicated by an increase in the fluo 3/fura red ratio, with a maximum response at 100 ng/ml (Fig.3A). IL-13 mRNA expression (Fig.3B) and protein release (Fig.3C) were significantly upregulated in HLMC stimulated with IL-33 (50 ng/ml, 24 h) with increased histamine release independent of FcεR1 cross-linking also observed (Fig.3D); HMC-1 (n = 4), but not ASM (n = 6), cells released CCL2, 4, 5, CXCL8 and 10 significantly following IL-33 incubation (10 ng/ml, 24 h) compared to control (data not shown).

24ASM wound repair was promoted by both exogenous and ASM-derived IL-33 (Fig.3E) as demonstrated by an IL-33-neutralizing antibody reducing wound repair in both control and IL-33-treated cells (Fig.3E). Neither ASM proliferation nor survival was modulated by exogenous IL-33 or neutralization of ASM-derived IL-33 (see Fig. S1).

25Direct addition of IL-33 (100 ng/ml) to PCLS from BALBc and C57BL6 mice (n = 5) did not affect bronchoconstriction over 2–10 min (data not shown). Pretreatment of PCLS from BALBc (Fig.3F) and C57BL6 (n = 4, data not shown) mice with IL-33 (100 ng/ml, 20 h) had no effect on Cch-induced luminal diameter. Contraction of human ASM cells primed with IL-33 (50 ng/ml, 48 h) and embedded within collagen gels was not different to unprimed ASM cells over 3 days (Fig.3G).

26Previously, we have shown that HLMC/ASM cell coculture promotes HLMC survival/proliferation and results in increased α-SMA expression and histamine release. The contribution of endogenous IL-33 to these changes was assessed.

27HLMC cocultured with ASM for 7 days demonstrated increased proliferation compared to HLMC monocultures as determined by CFSE fluorescence (Fig.4B) and cell counts (Fig.4C and D). This was unaffected by IL-33-neutralizing antibody. ASM cells counts were significantly increased following coculture with HLMC lysate for 7 days compared to monoculture; however, this was unaffected by IL-33-neutralizing antibody (Fig.4E).

28ASM cells cocultured with HLMC lysate showed increased α-SMA GMFI compared to ASM monocultures, but this was unaffected by IL-33-neutralizing antibody (Fig.4F,G). Histamine release was increased from HLMC following coculture with ASM compared to HLMC monocultures reaching significance after 11 days; however, this was unaffected by IL-33 or IL-33-neutralizing antibody (Fig.4H).

29Critically, when both ASM and HLMCs are impregnated into collagen gels following coculture and then stimulated directly with exogenous IL-33, increased gel contraction is seen compared to untreated cells. This can be inhibited by an IL-13-neutralizing antibody but is unaffected by the corresponding isotype control antibody (Fig.4I). HLMCs alone did not elicit gel contraction, ASM/HLMC cocultures did not increase gel contraction compared to ASM alone in the absence of exogenous IL-33, with no effect of IL-13 neutralization over 3 days on this contraction (data not shown, P = 0.38, n = 3). HLMC IL-13 release was unaffected by coculture with ASM or incubation with ASM-conditioned media (data not shown). These data suggest IL-33 can augment ASM contractility indirectly via upregulation of HLMC IL-13 release. However, endogenous release of IL-33 by ASM is insufficient to activate HLMC IL-13 release in this system.

30IL-33 induced a profound AHR in naïve BALBc mice after intranasal challenge with increase in total lung cells (Fig.5B), mast cell activation with increased serum concentrations of mouse mast cell protease-1 (mMCP-1, Fig.5C) and increased expression in the airway of MUC5ac, Gob-5 (Fig.5D) and IL-13 (Fig.5E). Interestingly, similar to results in the human coculture system, neutralization of IL-13 activity (using IL-13Rα2 fusion protein administered 2 h prior to each IL-33 administration) abrogated AHR significantly (Fig.5F).

31We demonstrated IL-33 expression in vivo and in vitro in the bronchial epithelium and ASM and in primary mast cells. The ST2 receptor was expressed by mast cells and ASM by total cell staining and at the surface following IL-33 treatment. IL-33 promoted mast cell activation and ASM wound repair and indirectly promoted contraction via upregulation of mast cell-derived IL-13. This suggests that IL-33 may present an important target to modulate mast cell–ASM crosstalk in asthma.

32We report here that IL-33 expression was evident in the bronchial epithelium and ASM-bundle, with expression increased in mild–moderate asthmatics compared to healthy controls. This is consistent with earlier reports in adult asthma, but contrasts with paediatric severe asthma in which neither epithelial nor ASM IL-33 expression was increased compared to controls. We found expression in the ASM, but not epithelium was correlated to the degree of AHR. Primary mast cells and ASM expressed IL-33 and ST2 constitutively as assessed by total cell staining and at the surface following IL-33 treatment. Localization of IL-33 was detected primarily within the nuclei as seen in human nasal fibroblasts.

33We demonstrated for the first time that IL-33 had no effect on ASM proliferation, apoptosis and synthetic capacity, but both exogenous and ASM-derived IL-33 played an important role in ASM wound repair. Thus, endogenously expressed IL-33 in ASM observed in vivo in humans may contribute to ASM repair via migration following damage secondary to physical, mechanical or inflammatory insults. Further work is required to determine the mechanism via which IL-33 stimulates ASM migration; however, in support of our observations, recombinant IL-33 has been shown to have direct effects on chemotaxis of myofibroblasts, fibrocytes, neutrophils, nuocytes and microglia cells.

34IL-33 triggered calcium flux in both ASM and HMC-1 cells, in keeping with other studies showing that IL-33 can enhance calcium elevation autonomously or in synergy with other mediators and that IL-33 can activate calcium-dependent downstream signalling. Although the mechanism by which IL-33 causes calcium elevation has not been studied, other members of the IL-1 family have been shown to induce calcium signalling in a manner which is GPCR dependent involving both extracellular calcium and intracellular calcium stores. Due to the rapid response, the effect of IL-33 on [Ca2+]i elevation is likely to be a direct effect on ASM, but in the mast cells it could be a synergistic response in conjunction with preformed mediators released by mast cells.

35Importantly, IL-33 is a critical cytokine in the initiation and exacerbation of inflammatory responses and enhanced IgE production in naïve wild-type mice, histamine release and tryptase expression in mouse mast cells. Similarly, human mast cells respond to IL-33 activation. Here, we found that IL-33 upregulated ST2, IL-13 and histamine release acutely by mast cells independent of FcεR1 cross-linking. Mast cell localization to the ASM-bundle is a notable feature of asthma, and therefore, ASM-derived IL-33 might play an important role in IgE-independent mast cell activation in the asthmatic airway. In addition, mast cell proteases have recently been shown to increase the activity of IL-33.

36Indeed, mast cell number within the ASM-bundle is related to the degree of AHR. Coculture of primary ASM and mast cells promotes mast cell activation, differentiation, survival, proliferation and phenotypic changes in ASM. Neutralization of IL-33 in ASM/mast cell cocultures had no effect on mast cell proliferation or histamine release, or α-SMA expression by ASM cells. Interestingly, coculture of ASM and mast cells together with the addition of IL-33 increased collagen gel contraction. This was IL-13 dependent, and the enhanced contraction in response to IL-33 was normalized following IL-13 neutralization. IL-33 had no direct effect on human ASM contraction or ex vivo PCLS from BALBc or C57BL6 mice. We found that in a mouse model system, IL-33 induced AHR, mMCP-1, MUC5ac, Gob-5 and IL-13 expression that was abrogated with IL-13Rα2 fusion protein supporting the human findings of IL-13-dependent IL-33 induction of ASM contraction possibly via mast cell activation. These findings are supported by two independent recent studies. Barlow and colleagues examined the response to methacholine in the PCLS ex vivo model and demonstrated that IL-33 mediated increased AHR that was IL-13 dependent. Saglani and colleagues found that IL-33 induction of AHR was IL-13 dependent prior to prolonged exposure to house dust mite, but following this exposure was partly IL-13 independent. However, the exact mechanism via which IL-33 mediates IL-13-dependent ASM contraction remains to be elucidated.

37Critically, in contrast to human asthma, mast cell localization to the ASM is not a feature of murine models of asthma. Therefore, although the animal models support the concept that IL-33-induced AHR is IL-13 dependent, whether mast cells are critical in these models is uncertain. Indeed, IL-13 has been shown to be produced by Th2 cells, NKT cells, basophils and ILC2s, the latter two of which can be dependent on IL-33. Nevertheless, IL-13 release by mast cells in human disease secondary to IL-33 activation remains likely to be important in human disease as these mast cells express IL-13 in vivo and are the most abundant inflammatory cell in the ASM-bundle. Interestingly, we were unable to demonstrate that constitutive ASM-derived IL-33 was sufficient to induce IL-13 release from mast cells in coculture, and therefore, it is likely that either upregulation of IL-33 release by ASM in vivo or contributions from other cellular sources such as the epithelium might be important in activating IL-13 release from mast cells in asthma. The exact role of IL-33 in human disease will require future clinical studies targeting the IL-33 axis.

38In conclusion, our findings showed that IL-33 promoted mast cell activation and ASM wound repair and indirectly promoted both ASM contraction in vitro via upregulation of mast cell-derived IL-13 and also IL-13-dependent AHR in vivo. Therefore, IL-33 might be an important novel therapeutic target to modulate mast cell–ASM crosstalk in asthma.