Combined exposure to the alarmins TSLP, IL-33 and IL-25 enhances mast cell-dependent contractions of human bronchi.

Asthma is a chronic inflammatory disease characterised by recurrent airflow obstruction where contraction of the smooth muscle is one main cause of the airway narrowing. Recent observations implicate thymic stromal lymphopoietin (TSLP), interleukin (IL)-33 and IL-25, collectively called epithelial alarmins, as key signalling molecules in asthmatic airway inflammation.1 When airway epithelium is exposed to external stimuli such as allergens, bacteria, viruses and air pollutants, alarmins are released as part of the host defence reaction. The actions of the epithelial alarmins initiate and maintain disease-driving inflammatory processes. Furthermore, tezepelumab, a biologic blocker the TSLP pathway, has recently been introduced as a new therapy in asthma treatment.1 Despite this, the molecular consequences of blocking TSLP or the signalling of other epithelial alarmins in human airways remain incompletely understood. So far, most mechanistic studies of actions of epithelial alarmins in humans have assessed their role in cell models in the context of type 2 (T2) inflammatory processes1; however, their possible effects on physiological processes in intact human airways have not been studied before. The current study was stimulated by the observation that the anti-TSLP antibody tezepelumab can attenuate allergen-induced bronchoconstriction in subjects with asthma.2 We therefore hypothesised that epithelial alarmins may enhance bronchoconstriction by previously unrecognised actions on human bronchi. Consequently, we used isolated human small bronchi to characterise the influence of the exposure to a combination of the three alarmins TSLP, IL-33 and IL-25 on controlled contractile responses. With permission from the regional ethical review board in Stockholm (ref. no. 2018/1819-31/1), tumour-free lung specimens were obtained with patient consent from 17 patients (12 women and 5 men, median age 70 [range 42-76 years]) undergoing lung surgery and consenting to donate removed lung tissue not required for diagnostic purpose. Small airway segments (inner diameter 0.5–2 mm) were isolated and exposed to TSLP, IL-33 and IL-25 (100 ng/mL each; hereafter referred to as the alarmin mix), for 24 or 48 h.3 After the preincubation with the alarmin mix or vehicle, the segments were mounted in separate myographs to study contractile responses. First, cumulative challenge with histamine (10–100 μM) was performed. Following a washout period, the same segments were challenged with two consecutive concentrations of anti-human IgE antibody (anti-IgE), (0.5 and 5 μg/mL, 15 min apart).4 The sensitivity of the bronchial segments to histamine was expressed as the negative logarithm of the concentration of histamine required to reach its half-maximal response (pEC50), and the amplitude of the response to the highest concentration of histamine was presented as the maximal response (Emax) in per cent of the segments' maximal contraction. We have previously shown using the same methodology that challenge of human small bronchi with anti-IgE triggers mast cell-dependent contractions.4 The contractile response to cumulative challenge with two concentrations of anti-IgE was enhanced after the preincubation with the alarmin mix for 48 h (Figure 1A), but not after 24 h (data not shown). In the alarmin mix pretreated group, the total area under the curve (AUC) was increased by 67% compared with the untreated group (1906 ± 274 vs. 1139 ± 185; p = .027; Figure 1B) and the maximal contractile response to the first dose of anti-IgE (0.5 μg/mL) was more than doubled compared with the untreated group (36.9 ± 6.8% vs. 14.5 ± 4.9%; p = .016; Figure 1C). The time to reach the peak contraction was shortened in the alarmin mix pretreated group compared with untreated (24.4 ± 1.0 min vs. 28.2 ± 1.2 min, p = .028; Figure 1D). Hence, the administration of the same concentration of anti-IgE to both treatment groups resulted in an enhanced contractile response in the group pretreated with the alarmin mix. We next investigated whether the enhanced contractile response to the anti-IgE challenge was due to increased release of mast cell mediators or increased responsiveness at the level of bronchial smooth muscle (Figure 2). To monitor mediator release, samples from the organ bath fluid were collected at baseline, 20 and 60 min after the start of the change with anti-IgE. The samples were then analysed for prostaglandin (PG) D2 using ELISA (PGD2-MOX kit: Cayman) and histamine using fluorometric analysis (RefLab ApS). The detected concentrations were normalised to the weight of the respective segment. Increased concentrations of PGD2 were detected in the organ bath fluid of the alarmin mix pretreated group 20 min after the initial challenge with anti-IgE (51.3 ± 14.0 ng/mg tissue; p = .034), but not in the bath fluid of the untreated group (Figure 2A). After 60 min, the concentrations of PGD2 were increased in the bath fluid from both treatment groups (87.4 ± 9.0 ng/mg tissue for vehicle control, p < .001; 146.8 ± 24.3 ng/mg tissue for alarmin mix; p < .001). Notably, at this time, the release of PGD2 was significantly greater in the alarmin mix pretreated group than in the untreated group (p = .031; Figure 2A). These findings support that the alarmin mix enhanced the IgE-triggered release of PGD2, presumably by increasing mast cell reactivity. In the control group, release of histamine into organ bath fluid was not significantly increased from baseline, either at 20 (fold change 1.2 ± 0.3) or at 60 min (fold change 0.8 ± 0.2). In contrast, in the alarmin mix pretreated group, there was a significant increase of histamine release at 60 min after the start of the anti-IgE challenge (fold change 2.9 ± 0.7; p = .045) with a trend of almost reaching significance already at 20 min (fold change 3.3 ± 0.9; p = .059; Figure 2B). The release of histamine in response to IgE-mediated stimulation increased only in the alarmin mix pretreated group, indicating that, as shown for PGD2, the alarmin mix enhanced the IgE-triggered release of histamine, presumably by increasing mast cell reactivity. In contrast to the influence of the alarmins on the indirect contractile response triggered by exposure to anti-IgE, there was no effect of preincubation with the alarmin mix on the responsiveness to histamine. Preincubation with the alarmin mix for neither 24 (data not shown) nor 48 h altered neither the potency (pEC50 6.2 ± 0.1 vs. 6.1 ± 0.1 for vehicle control) nor the efficacy (Emax 99.9 ± 0.1% vs. 98.6 ± 1.1% for vehicle control) of histamine (Figure 2C). In summary, these findings confirmed our hypothesis that the epithelial alarmins have the potential to enhance bronchoconstriction. Specifically, 2 days of combined exposure of isolated small bronchi to TSLP, IL-33 and IL-25 enhanced the contractile response to anti-IgE and the associated release of the mast cell mediators histamine and PGD2, while having no effect on contractions triggered directly at the level of the smooth muscle by exposure to exogenous histamine. The lack of effect of the alarmin mix on responsiveness to histamine in the isolated bronchi is in line with the results of studies of human airway smooth muscle cells, where neither IL-335 nor TSLP has been shown to affect contraction.6 The effect of IL-25 on smooth muscle contractions has not been reported. Using isolated human small bronchi, we have recently shown that preincubation with the T2 cytokines IL-4 and IL-13 increases airway responsiveness to contractile agonists histamine, carbachol and leukotriene D4 that act directly on the smooth muscle.3 While 48-h exposure to IL-4 and IL-13 caused a marked increase in the potency of histamine, the alarmin mix had no direct effect on histamine responsiveness. Instead, the epithelial alarmins selectively enhanced the IgE-mediated indirect contractile response. Our experiments document that pretreatment with the alarmin mix resulted in a more pronounced release of PGD2 and histamine from isolated human bronchi in response to an anti-IgE challenge. This is in line with previous observations in other models, for instance, exposure to the combination of IL-33 and TSLP for 6 h increased release of PGD2 from human mast cells.7 Research on the influence of alarmins on human mast cells is at an early stage and has mostly been performed in combined stimulations. For example, Buchheit et al. showed that only the combined stimulation with IL-33 and TSLP, and not IL-33 alone increased the release of PGD2.7 Likewise, exposure to a combination of IL-33 and TSLP for 1 week before IgE crosslinking caused increase in cysteinyl leukotriene formation in mast cells from patients with exercise-induced bronchoconstriction, whereas exposure to either IL-33 or TSLP alone had no effect.8 The use of the combination of epithelial alarmins has also been found to trigger enhanced release of PGD2 and cytokines in type 2 innate lymphoid cells.9 Thus, for this first proof-of-concept study, it was considered appropriate to use an alarmin mix, such as in previously mentioned studies. This limitation will be addressed in follow-up studies where the actions of the individual epithelial alarmins and their combinations will be defined. There is also a need to establish the influence of longer periods of preincubation as well as the effects of different doses of the alarmins. In conclusion, to the best of our knowledge, this is the first study that explores the effects of epithelial alarmins in intact isolated human small bronchi. The findings implicate that the epithelial alarmins specifically increase the reactivity of airway mast cells. This mode of action may explain why the clinical effects of the first registered anti-alarmin tezepelumab include rapid attenuation of the allergen-induced bronchoconstriction and eosinophil-independent improvement of severe asthma.1, 2 MB performed experiments and data analysis and wrote the first version of the manuscript. MA-A and A-CO provided surgical specimens and contributed to study design and interpretation. MA and SED obtained ethics approval. MA, JS and SED had the general project responsibility, performed experimental design and finalised the manuscript together with MB. All authors read, revised and approved the manuscript. We thank Susanne Hylander for her dedicated management of the logistics for the collection of human lung specimens and all co-workers at the Heart and Vascular Theme as well as at Perioperative Medicine & Intensive Care for their contributions. We also gratefully acknowledge the patients at the Heart and Vascular Theme. This work was supported by Magnus Bergvall Foundation, Konsul Th C Berg Foundation, Swedish Heart-Lung Foundation, Swedish Research Council—Medicine and Health, Stockholm County Council Research Funds (ALF), Swedish Society of Medicine, Centre for Allergy Research Highlights Asthma Markers of Phenotype consortium which is funded by the Swedish Foundation for Strategic Research, Karolinska Institutet, AstraZeneca & Science for Life Laboratory Joint Research Collaboration and Vårdal Foundation. The funding sources were not involved in either of the following: preparation of the article, study design, collection, analysis and interpretation of data, writing of the report and decision to submit the article for publication. The authors have no conflicts of interest to declare. The data that support the findings of this study are available from the corresponding author upon reasonable request.