Several studies suggest that the presence or accumulation of mast cells at certain compartments of the lung are pathological features of allergic asthma (Figure 1). An increased number of mast cells were found in the airway smooth muscle of asthma patients in comparison to controls or subjects with eosinophilic bronchitis (42). The number of mast cells was also higher in the smooth muscle of allergic asthmatics in relation to non-allergic asthmatics (43). In support of these studies, isolated human bronchi with the ability to contract in response to allergens had a higher number of smooth muscle-associated mast cells than unresponsive human bronchi isolates (46). Moreover, an increased number of mast cells was found in the distal airways of subjects with non-fatal and fatal asthma compared to non-asthmatic controls (44). In the distal airways, the greatest increase in mast cells was found within the smooth muscle and mucous glands. In another study, the same authors found an increased percentage of degranulated mast cells in the mucous glands from fatal asthma in comparison to non-fatal asthma and controls, suggesting that mast cells are highly activated in fatal asthma (47). In biopsies from patients with severe asthma, the number of MC_TC and the MC_TC/MC_T ratio in the small airways were higher compared to normal subjects (48). However, a positive correlation between MC_TC in the region of small airways/alveolar attachments and lung function was found (48), suggesting a protective role of this subtype of mast cells.

Increased mast cell numbers are also found in other lung compartments of asthma patients. For example, sputum samples from asthmatics more often showed expression signatures of mast cell-specific protease genes (TPSAB1 and CPA3) and higher expression of these genes than samples from healthy subjects (49). In accordance with this, genetic analyses of samples from epithelial brushings have revealed increased gene expression of mast cell tryptase (TPSAB1, TPSB2) and CPA3 (but not CMA1, coding for chymase) in asthma patients (39–41). In the study by Singhania et al., a twofold increase in the number of intraepithelial mast cells was found in patients with “Th2-high” asthma compared to those characterized as “Th2-low” (39). Moreover, the presence of intraepithelial mast cells in the Th2-high asthma patients predicted responsiveness to inhaled corticosteroids. In a study of severe asthma, the proportion of MC_TC among all mast cells was higher in patients with severe asthma compared to those with mild asthma (50). A higher proportion of MC_TC was also found in the bronchi of uncontrolled atopic asthmatic subjects, and overall a higher number of mast cells was found in the alveolar parenchyma, paralleled by an increase in their expression of FcεRI (51). In a follow-up study, patients with mild atopic asthma were found to have an elevated number of alveolar mast cells with increased expression of FcεRI in comparison to healthy controls and non-asthmatic allergic rhinitis patients (45). Although the above-mentioned studies were performed on adults, another study demonstrated that submucosal mast cells are more frequent in bronchial biopsies from symptomatic children with severe asthma than in those with few symptoms (52). To summarize, mast cells accumulate in the smooth muscle, bronchial epithelium and alveolar parenchyma of patients with allergic and severe asthma, thereby presumably increasing the detrimental consequences of mast cell activation in the allergic lung. However, in some compartments of the lung, there might be an advantage to having chymase-expressing mast cells.

To address whether the increase in lung mast cells in patients with allergic asthma is due to the recruitment of MCps to the lung or a result of local proliferation of resident mast cells, mouse models have been used. In early studies, lung MCps were quantified by a limiting dilution and clonal expansion assay in an ovalbumin (OVA)-model of allergic airway inflammation (36). In OVA-sensitized and challenged mice, the number of lung MCps was around 30 times higher 1 day post-challenge than in control mice. The OVA-induced increase in lung MCps was absent in mice genetically deficient of endothelial VCAM-1 and in wild-type mice treated with blocking antibodies targeting VCAM-1, α4-, β1-, or β7-integrins during the challenge phase (36). This suggests that allergic airway inflammation induces the recruitment of MCps to the lung. After the acute phase, the accumulation of lung MCps gradually led to the appearance of toluidine blue⁺ mast cells in the epithelium of the trachea and the lung parenchyma (53, 54). Moreover, the OVA-induced recruitment of MCps to the lung was dependent on the presence of CD11c⁺ cells (54) and CD4⁺ cells (55), demonstrating that an adaptive immune response is required to stimulate this process.

In an effort to study a possible chemokine component in the antigen-induced recruitment of MCps to the lung, mice lacking CCR3 or CCR5 were used to test whether any of these chemokine receptors were involved in this process. However, CCR3^−/− and CCR5^−/− mice had an intact OVA-induced recruitment of MCps to the lung (53). Nevertheless, the OVA-induced recruitment of MCps to the lung was partly dependent on the presence of CXCR2 in stromal cells in the lung. Mice lacking CXCR2 had a reduced inflammation-induced upregulation of VCAM-1 on the lung endothelium, which could explain the reduced recruitment of MCp to the lung (53). CCR2-deficient mice also had a reduction in OVA-induced recruitment of MCps to the lung (56). Nonetheless, this was likely due to unidentified stromal defects and not to MCps lacking CCR2 per se (56). Therefore, any chemokine component required for the recruitment of MCps to the lung remains unknown.

The role of cytokines in OVA-induced recruitment of MCps to the lung has also been a matter of investigation. Interestingly, the OVA-induced recruitment of MCps to the lung occurs independently of genetic ablation of IL-4, IL-4Rα chain, STAT-6, IFN-γ, and IL-12 and antibody-mediated neutralization/blocking of IFN-γ, IL-3, IL-4, IL-5, IL-6, IL-13, IL-17A, IL-12p40, or IL-12p40Rβ1 during the challenge phase (55). However, IL-9 deficiency or IL-9 antibody neutralization efficiently prevented the OVA-induced recruitment of MCps to the lung. In an effort to identify the source of IL-9, we also found that genetic ablation of CD1d or blocking CD1d during the challenge phase inhibited the OVA-induced recruitment of MCps to the lung, but genetic ablation of invariant NKT cells (Jα18 deficient mice) had an intact infiltration of MCps to the lung (55). As blocking CD1d in IL-9-deficient mice or neutralizing CD1d in IL-9-deficient mice did not further inhibit the OVA-induced recruitment of MCp to the lung, type 2 NKT cells may provide or elicit IL-9 production (55). The importance of IL-9 in the accumulation of lung mast cells during allergic airway inflammation was also highlighted in a study where adoptive transfer of Th9 cells followed by challenge with OVA and TSLP increased the mast cell numbers estimated by histological analyses (57). Treatment with an anti-IL-9 antibody blocked the mast cell accumulation in both the adoptive transfer model and in an OVA sensitization and challenge model (57). In the same paper, decreased mast cell numbers were found in mice with PU.1-deficient T cells, which have reduced IL-9 levels in house dust mite (HDM)-induced allergic airway inflammation.

The importance of IgE for the survival of lung mast cells was demonstrated in a model of Aspergillus Fumigatus-induced allergic airway inflammation (58). However, no defect in the Aspergillus-induced recruitment of MCps to the lung could be detected in mice lacking IgE. Nevertheless, we found that the number of lung MCps increased significantly when sensitized wild-type mice were challenged with IgE-antigen immune complexes compared to control mice given the same dose antigen alone (59). The stimulating effect of IgE-immune complexes on the recruitment of MCps to the lung was lost in FcRγ-deficient mice, but not in CD23-deficient mice, indicating that MCp recruitment can be potentiated by Fc receptor-mediated activation (59). Thus, IgE-immune complex formation and IgE alone impacts the recruitment of MCps to the lung and the survival of lung mast cells, respectively.

The technical advances in multi-color flow cytometry have made it possible to distinguish different mast cell populations simultaneously. Using a prolonged protocol of OVA-induced allergic airway inflammation, three lung mast cell populations could be identified, MCps expressing high levels of integrin β7, resident mature mast cells and emerging/induced mast cells (37). In parallel, we demonstrated that influenza infection in mice induced the recruitment of MCps to the lung, which later gave rise to mast cells with intermediate expression of integrin β7 (immature/induced) and even later increased the number of mature mast cells with low expression of integrin β7 (38). While the recruitment of MCps induced by allergic lung inflammation was dependent on an adaptive immune response (54, 55), the influenza-induced MCp recruitment to the lung was due to the induction of innate immune responses (60). As respiratory infections commonly cause exacerbations of asthma symptoms, we think it is intriguing that influenza infections induces mast cell accumulation in the mouse lung.

Mice have frequently been used to model allergic asthma with the aim of investigating the mechanism behind the disease. When interpreting the results from mouse models of allergic airway inflammation it is necessary to understand that the lungs of mice and humans have anatomical and physiological similarities and differences. The right lung of humans and mice consists of five lobes. The left lung in humans consists of two lobes, whereas mice only have a single lobe on the left side (61). Another difference is that mice have monopodial branching whereas humans have dichotomous branching of the airways. Further, the intrapulmonary bronchi in mice lack cartilage, which suggests that the difference between bronchi and bronchioles is less obvious in mice than in humans. The number of goblet cells and submucosal glands is lower in mice, at least in laboratory mice that have not been infected by pathogens or subjected to a disease model. Moreover, whereas smooth muscle cell bundles populate the connective tissue surrounding the respiratory bronchioles in humans, mice have very few respiratory bronchioles that lack a smooth muscle layer.

When comparing human and naïve mouse lungs there is also an apparent difference in the quantity of mast cells and their distribution. In human lungs, mast cells are located throughout the airways and in the parenchyma, while in naïve mice mast cells are mainly found in the trachea and in the central airways (62). However, in mouse models of allergic asthma and influenza infection, the induced lung mast cells accumulate at places where they are not usually found such as in the epithelium, surrounding bronchioles, in the perivascular space and in the alveolar parenchyma (38, 53, 54, 63). Although allergic airway inflammation in mice induced by e.g. OVA and HDM stimulates airway hyperresponsiveness (AHR) to methacholine in vivo, these models rarely induce antigen-induced bronchoconstriction that can be measured in vivo. Importantly, neither lung mast cell expansion nor lung function have been analyzed in the majority of published studies of allergic airway inflammation in mice. Nevertheless, a high dose of HDM (125μg) given intranasally on five consecutive days/week over 3 weeks to induce allergic airway inflammation also induced mast cell expansion and an increase in the mast cell specific mediator mMCP-1, along with HDM-induced bronchoconstriction (63). Importantly, the HDM-induced bronchoconstriction was abrogated in Kit^{W−sh/W−sh} mice.

In contrast, when isolated mouse trachea from mice with allergic airway inflammation is analyzed ex vivo, antigen-induced contractions can be measured using OVA as the antigen (64). The antigen-induced contractions ex vivo are also abrogated in Kit^{W−sh/W−sh} mice (64). A possible reason for the discrepancy between the lack of OVA-induced bronchoconstriction in vivo and the presence of OVA-induced contraction in isolated airways may be that the majority of mast cells are found around the central airways and hence it is easier to measure their responsiveness to antigen in isolation (ex vivo). Nevertheless, it is currently unclear whether the reason behind why it is so hard to observe antigen-induced bronchoconstriction in vivo is only due to a less expanded lung mast cell population with lower doses of antigen and more acute protocols, or whether there are other more profound differences in lung physiology between the species that also play a role.

More often in vivo studies of allergic airway inflammation in mice measure AHR with increasing doses of methacholine. There are examples of protocols that bypass the dependence of mast cells for AHR, e.g. (65–67), while other protocols find that mast cells, or a mast cell mediator, are necessary for a full-blown AHR, e.g. (68–71). Moreover, when Fuch et al. compared Kit^{W−sh/W−sh} mice with Kit^{W−sh/W−sh} mice reconstituted by bone marrow-derived mast cells (BMMCs) to wild type sensitized and challenged with OVA, the reconstituted mice had a higher density of mast cells which were distributed differently compared to the wild-type mice (72). This resulted in an increased AHR in the Kit^{W−sh/W−sh} mice reconstituted with BMMCs as compared to wild-type mice treated in parallel. We speculate that this conundrum reflects the situation in human asthma, i.e. that not all phenotypes of asthma have a significant mast cell component, while some do.

Pseudo-allergic reactions typically occur in response to basic substances such as compound 48/80 or cationic peptide drugs. These reactions share characteristics with allergic responses induced by IgE-mediated responses, but trigger-specific IgE molecules are not detected (111). In 2015, MRGPRX2 (MAS-related G protein-coupled receptor X2) and the mouse ortholog MRGPRB2 were described as the main receptors involved in pseudo-allergic reactions due to their interaction with cationic drugs, which led to mast cell activation (112). In mice, MrgprB2 transcripts were only found in CTMCs (112). The transcript levels of MRGPRX2 are higher in human skin mast cells than in lung mast cells (113, 114). Nevertheless, the expression of MRGPRX2 on mast cells and the number of MRGPRX2⁺ mast cells are higher in lung biopsies from patients who died from asthma-related causes than in lung biopsies from patients who died from other causes (115).

Substance P (SP) was one of the identified ligands of MRGPRX2 (112). SP is elevated in BAL and sputum in asthma patients compared to healthy controls (116, 117) and is further increased in BAL immediately after allergen provocation (116). Therefore, one possibility is that MRGPRX2 is involved in a positive feed-back loop where mast cells are activated via e.g. allergen release histamine, which activates neurons to produce cationic peptides such as SP, which amplifies mast cell activation. The activation of MRGPRX2 may also be related with exacerbations of asthma symptoms that occur in connection with viral respiratory infections (118). Several respiratory viruses trigger the secretion of anti-microbial peptides such as β-defensins from epithelial cells (119, 120). β-defensins activate mast cells trough MRGPRX2 (121) and may thus contribute to virus-induced asthma exacerbations. To conclude, MRGPRX2-mediated activation of mast cells may contribute to the pathogenesis of asthma.

Mast cells are specialized in sensing external pathogens by recognition of pathogen-associated molecular patterns (PAMPs) and are equipped to respond to tissue damage by recognition of danger-associated molecular patterns (DAMPs) or alarmins (122). Interestingly, several allergens have the ability to stimulate toll-like receptors (TLRs), and these receptors may thus play a role in asthma development. For example, HDM extracts contain e.g. lipopolysaccharide (LPS) and proteins like Der p2 (Dermatophagoides pteronyssinus) mimicking MD-2, resulting in TLR4 activation (123). Mouse mast cells express several TLRs, and the expression pattern seems to differ between the mast cell subtypes, e.g. Tlr1-9 mRNA was detected in both MMC-like and CTMC-like BMMC but the level of Tlr3-5 expression differed (124). Human peripheral blood-derived mast cells express mRNA and protein for TLR1-4 and TLR6-8, and responded to double-stranded RNA stimulation via TLR3 by producing type I interferons (IFNs) (109). The expression of TLR2 was confirmed in isolated human lung mast cells by western blot, and in vitro stimulation with lipoteichoic acid (a TLR2 agonist) led to down regulation of FcεRI expression and decreased IgE-antigen-mediated mast cell degranulation (125). The expression of TLR4 has also been confirmed in human lung mast cells (126). In the same study, LPS stimulated release of TNF-α in peripheral blood-derived mast cells, and after pre-incubation with IFN-γ, LPS induced the expression of anti-viral genes (126). In vivo, TLR4-mediated mast cell activation was demonstrated to enhance eosinophilia and cytokine release in an OVA-model of allergic airway inflammation using mast cell-deficient (Kit^W/Wv) mice and reconstitution experiments (127). Furthermore, intranasal administrations of poly I:C expand the number of lung MCps in a TLR3-dependent fashion (60). To summarize, the activation of pattern recognition receptors may directly or indirectly induce or modulate mast cell responses in the allergic lung.

Inhaled allergens and respiratory viruses induce the release of alarmins such as IL-33 and thymic stromal lymphopoietin (TSLP) from airway epithelial cells (128). IL-33 is a member of the IL-1 family of cytokines, which primarily binds to the IL-1 receptor family member ST2. However, the IL-33/ST2 complex assembles with the IL-1 receptor accessory protein (IL-1RAcP), which is needed for signal transduction (129). The loci for IL1RL1 (ST2 gene) and IL-33 contains a single nucleotide polymorphism that was associated with asthma in a large-scale genome-wide association study (130). Moreover, IL-33 expression was higher in lung samples from patients with severe asthma than in those with mild asthma and IL-33 expression was mainly localized to epithelial and endothelial cells, neutrophils, fibroblast and mast cells (131). Another source of IL-33 in the airways of asthma patients are the smooth muscle cells (132). Interestingly, there was a strong inverse correlation between the concentration of IL-33 in BAL and lung function (pre-bronchodilator FEV₁) (133).

Human in vitro-derived mast cells from cord and peripheral blood express ST2 (134). In vitro, IL-33 accelerates mast cell maturation of CD34⁺ cells and induces the secretion of Th2 cytokines and chemokines (134). Moreover, pre-treatment with IL-33 increases the number and magnitude of degranulating in vitro-derived mast cells in response to a crosslinking anti-IgE antibody (135). IL-33 increases the survival of human skin mast cells by upregulation of the antiapoptotic protein, B-cell lymphoma-X large (BCLXL) (136). In cord blood-derived mast cells, IL-33 alone promoted adhesion to fibronectin and production of IL-8 and IL-13 (137). In mice, lung mast cells and their progenitors express ST2 protein on the cell surface and intranasal administrations of IL-33 induce an ST2-dependent increase in lung MCp (59, 60).

The role of IL-33 in the context of allergic airway inflammation has been studied in vivo. Generally, mice lacking ST2 and wild type mice administrated with ST2 blocking antibodies have reduced allergic airway inflammation (138). IL-33 induces an asthma-like phenotype in Rag2 ^−/− mice, which lack mature lymphocytes, demonstrating that adaptive immune responses are not required to induce an asthma-like phenotype (139). Further, in a new humanized mouse model (NOG IL-3/GM-CSF), administration of human IL-33 induced an asthma-like phenotype mediated by human IL-13 (140). The main producers of IL-13 in this setting were T cells and mast cells. Further, IL-33 given before the allergen challenge potentiated AHR in wild-type but not in mast cell-deficient Kit^{W−sh/W−sh} mice or mice treated with ketanserin (a nonselective 5-HT₂R antagonist with high affinity also for H1R) in a OVA model of allergic airway inflammation (141). In contrast, in papain-induced airway inflammation, which promotes IL-33 production and increased mast cell numbers in the lung, mast cell-deficient (Kit^{W−sh/W−sh}) mice had an exacerbated type 2 inflammatory response (142). The suppressive effect of mast cells in this experimental set-up was explained by the observation that IL-33-activated mast cells produced the IL-2 necessary for the expansion of CD4⁺CD25⁺Foxp3⁺ regulatory T cells that were inhibiting the type 2 inflammation by production of IL-10. Altogether, there is substantial evidence that the IL-33-ST2 pathway is involved in allergic asthma and that this pathway interacts with mast cells.

TSLP is another alarmin released from epithelial cells as a DAMP signal after allergen exposure. This IL-7-like cytokine binds to the TSLP receptor (TSLPR), which shares the alpha chain with the IL-7 receptor (143, 144). The expression of TSLP was first associated with acute and chronic dermatitis (145). In human bronchial biopsies from asthmatics, TSLP transcripts were increased in epithelial cells and in the submucosa (146). Interestingly, the number of epithelial cells or submucosal cells expressing TSLP was inversely correlated with FEV₁ (% of predicted) (146). TSLP was also increased in serum from asthmatics compared to healthy controls or patients with chronic obstructive lung disease (COPD) (147). In mice, the expression of TSLP is increased in allergic airway inflammation and TSLPR-deficient mice have an attenuated type 2 response (148). Moreover, overexpression of TSLP in epithelial cells in the mouse lung produced a spontaneous asthma-like phenotype (148). TSLP^−/− mice have reduced numbers of mast cells in the intestine, kidney, nasal mucosa, skin, liver and lung, suggesting that TSLP is involved in mast cell development (149). In support of that, IL-3-induced proliferation and differentiation of BMMCs is blocked by neutralization of TSLP (149).

TSLP expression has also been demonstrated in human bronchial and submucosal mast cells as well as in the epithelium and airway smooth muscle (146, 150). Airway smooth muscle cells and human mast cells in this location expressed TSLPR (150). Further, the percentage of TSLP⁺ mast cells in bronchial samples was increased in asthmatic airways compared to healthy controls (151). In the same study, activation of human peripheral blood-derived mast cells by a crosslinking anti-IgE antibody induced TSLP expression, which was further enhanced by pre-incubation with IL-4. Moreover, TSLP potentiates the IL-33-stimulated secretion of type 2 cytokines and chemokines in human mast cells derived from peripheral blood or cord blood (134). The importance of TSLP in asthma is highlighted by clinical studies using a monoclonal anti-TSLP antibody. In 2014, an anti-TSLP antibody showed efficacy in several measures of allergen-induced early and late asthmatic responses including improved FEV₁ (% of predicted) (152). Furthermore, anti-TSLP treatment has been shown to reduce the frequency of asthma exacerbations and improved FEV₁ (% of predicted) in a randomized, double-blind, placebo-controlled trial of uncontrolled asthmatics with moderate to severe asthma (153). Altogether, the present data suggest that TSLP influences the development of mast cells and that mast cell-derived TSLP may contribute to allergic asthma.

Adenosine tri-phosphate (ATP) is a danger signal to the immune system when released to the extracellular milieu by many different cells types. Extracellular ATP is sensed by the class 2 purinergic P2Y and P2X receptors (154). In the extracellular space, ATP is quickly hydrolyzed by nucleoside triphosphate diphosphohydrolase (NTPDase or CD39) to adenosine mono-phosphate (AMP) via adenosine di-phosphate (155). Allergen challenge triggers an increase in ATP levels in BAL from asthma patients and mice with experimentally induced allergic airway inflammation (156). In the same study, neutralizing ATP by the ATP-hydrolyzing enzyme apyrase or inhibition with a broad-range inhibitor of P2 receptors before OVA challenge blocked type 2 inflammation and AHR.

That rodent mast cells are activated and degranulate in response to ATP has been known for several decades (157, 158). In an acute mouse model of colitis, ATP-mediated mast cell activation was demonstrated to occur through P2X₇ receptors (159). A role for the P2X₇ receptor in allergic airway inflammation and AHR was demonstrated using P2X₇ deficient mice and treatment of wild type mice with a specific P2X₇ antagonists given before each antigen challenge (160). Recently, CD203c or ecto-nucleotide pyrophosphatase-phosphodiesterase 3 (E-NNP3), a widely used activation marker of mouse mast cells and basophils, was demonstrated to negatively regulate IgE-antigen-mediated activation through hydrolyzation of extracellular ATP (161). Further, while E-NNP3^−/− mice had exacerbated allergic airway inflammation, mice lacking both E-NNP3 and P2X₇ had decreased responses to IgE-antigen-mediated activation of FcεRI. Therefore, ATP released by IgE-antigen-mediated activation of FcεRI stimulates mast cell activity through P2X₇, whereas E-NNP3 decreases the ATP concentration and suppresses mast cell (and basophil) activity in vivo.

In human lung mast cells, the expression of the purinergic receptors (P2X₁, P2X₄, P2X₇, P2Y₁, P2Y₂) has been confirmed by q-RT PCR and a gene array (162–164). An early study indicated that ATP stimulation of human lung mast cells did not directly induce degranulation but enhanced histamine release after anti-IgE mediated activation (162). The demonstration of P2X₇ receptor on mast cells in the colon from patients with Crohn's disease (159) suggests that this receptor may regulate mast cell function also in human disease. However, a reduced risk of asthma and asthma severity was detected in children with loss-of-function mutations in the P2X₇ receptor (165).

Adenosine is directly released from many cell types or generated from extracellular AMP by CD73 (ecto-5′-nucleotidase), and the extracellular levels increase under inflammatory conditions (166). In asthmatics and cigarette smokers, adenosine levels were increased in BAL (167, 168). Further, adenosine provoked bronchoconstriction in asthmatics (169, 170). Adenosine exerts its biological functions binding to four distinct G-protein coupled receptors: A1, A2a, A2b and A3 with different affinity (167). In human lung mast cells, adenosine potentiated IgE-antigen-mediated mast cell activation, thereby increasing the release of mediators such as LTC₄ and histamine (171). Human and mouse lung mast cells express mRNA for A2a, A2b and A3, and activation via A3 induces histamine release (172–175). In mice, nebulization of an A3 agonist for only 5 min caused mast cell degranulation (175). Moreover, adenosine induced airway contraction, which is lost in mast cell deficient (Kit^W/Wv) and in A3^−/− mice along with the loss of adenosine-induced mast cell degranulation (176). This suggests that mast cell activation via A3 is the major mechanism behind adenosine-induced bronchoconstriction. In a follow-up study, pre-exposure to aerosolized adenosine was demonstrated to increase methacholine-induced AHR in wild-type, but not in mast cell-deficient (Kit^W/Wv or Kit^{W−sh/W−sh}) or A3^−/− mice (177). Further, the increased AHR after adenosine pre-treatment was regained by reconstitution of Kit^{W−sh/W−sh} mice with wild-type BMMCs but not with A3^−/− BMMCs. These in vivo studies suggest that adenosine can promote bronchoconstriction or enhancing AHR by activating mast cells through the A3 receptor. In the context of human mast cells, adenosine or an A3-specific agonist potentiated FcεRI-induced activation of human lung mast cells (174). Altogether, ATP and adenosine may play an important role in promoting mast cell activation in the allergic lung.

Histamine is a pivotal molecule involved in allergic reactions, and mast cells were recognized early as the main source of histamine (208). In patients with allergic asthma, allergen challenge via bronchoscope stimulates increased histamine levels in BAL (179). Histamine mediates its biological effects by binding to four histamine receptors (H1-4R), which are expressed in many different cell types such as immune cells, nerves and smooth muscle cells (209). In the human airways, the H1R mediates bronchoconstriction and increases vascular permeability (182). Despite this, H1R antagonists have demonstrated variable efficacy in clinical asthma trials (210). In OVA models of allergic airway inflammation, mice lacking H1R had reduced Th2 responses and AHR (211, 212).

Besides mediating the effect on acid production in the stomach, the histamine H2R is associated with the regulation of immune responses (213). A recent study demonstrated that genetic ablation or blocking of H2R in a model of OVA-induced allergic airway inflammation increased the eosinophilia in BAL, type 2 cytokines and mucous production, while an H2R agonist suppressed these responses (214). The exacerbated type 2 inflammation in H2R-deficient lungs was associated with increased numbers of CD1d⁺ dendritic cells and iNKT cells with an increased capacity to secrete cytokines in response to lipid antigens, suggesting that iNKT cell activity is controlled by histamine via H2Rs in the allergic lung.

The H3R is preferentially expressed in the nervous system and regulates its own release from neurons (215). The H4R is the most recently discovered histamine receptor, and like the H1R, this receptor mediates mostly pro-inflammatory effects. The H4R is mainly expressed by immune cells (216). In human eosinophils, histamine potentiated the chemotactic response to eotaxin via the H4R and directly mediated chemotaxis of human basophils in vitro (183, 184). In basophils, a specific H4R agonist reduced the mediator release after IgE-antigen-mediated activation (184). The H4R is also expressed in human cord blood-derived and purified lung mast cells (185, 217). In human lung mast cells an H4R-specific agonist induced chemotaxis (185), whereas in cord blood-derived mast cells another selective H4R agonist stimulated degranulation, and generation of lipid mediators and cytokines (217). The H4R is also expressed by mast cells, basophils and eosinophils in mice, and mediates histamine-induced chemotaxis in BMMCs (218). In OVA-induced models of allergic airway inflammation, H4R-deficient and wild-type mice treated with an H4R antagonist during either the sensitization or challenge phase had diminished type 2 inflammation (219, 220). The reduced type 2 lung inflammation in the absence of functional H4R was first attributed to a reduced ability of Th2 cells to produce cytokines (219), and more recently to the fact that dendritic cells cannot be fully activated in the absence of signals through the H4R and therefore the CD4⁺ T cell activation is reduced (220). For detailed information about the history of histamine, its receptors and the use of antagonists, see (221).

Serotonin (5-hydroxytryptamine, 5-HT) is stored in mast cell granules and released upon mast cell degranulation. However, while mouse mast cells contain large quantities of 5-HT, human mast cells contain none or very small amounts (222, 223). Instead, platelets are the main source of 5-HT in human airways (187). 5-HT has wide biological effects and acts by binding to seven different 5-HT receptor families (5-HT₁₋₇R), the serotonin transporter (SERT) and by binding covalently to different effector proteins (224). Although many animal studies have demonstrated that 5-HT induces bronchoconstriction, 5-HT does not consistently induce bronchoconstriction in humans (187). Nevertheless, plasma levels of 5-HT are elevated in asthma patients and are associated with disease severity (186). Further, ketanserin decrease adenosine- and methacholine-induced bronchoconstriction in asthma patients (188, 189).

In mice, 5-HT plays a major role in airway contraction mainly via the activation of 5-HT₂ receptors (225). 5-HT induces a ketanserin-sensitive depolarization-response in cholinergic neurons in the trachea, suggesting that 5-HT works by inducing secretion of acetylcholine from nerves that innervate the airways (64). Also, selective 5-HT_2BR antagonists can block 5-HT-induced bronchoconstriction in mice ex vivo (226).

In human airways, 5-HT acts on airway smooth muscle via 5-HT_2AR, which induces contraction, and via 5-HT_1AR, which induces relaxation (227). Moreover, 5-HT may also work by inducing or augmenting the release of acetylcholine from cholinergic nerves via 5-HT₃R and 5-HT₇R (227). Nevertheless, 5-HT induced eosinophil migration via the 5-HT_2AR in both mice and humans (228, 229). Mast cells express mRNA for several 5-HT receptors. In a study by Kushnir-Sukhov et al., transcripts for the 5-HT_1A, 5-HT_1B, 5-HT_2A, 5-HT_2B and 5-HT₇ receptors were detected in BMMCs and peripheral blood-derived human mast cells (230). In addition, transcripts for the 5-HT_1E, 5-HT_2C, 5-HT₃ and 5-HT₄ receptors were also found in peripheral blood-derived human mast cells, while the BMMCs also expressed transcripts for 5-HT_1DR, 5-HT₆R and SERT. The same study revealed that 5-HT does not induce degranulation or cytokine release by in vitro-derived mast cells from mice or humans. Nevertheless, 5-HT stimulated adhesion to fibronectin and chemotaxis (230). In line with those findings, another study showed that intradermal injection of 5-HT induced 5-HT_1AR-dependent accumulation of mast cells at the site of injection 48 h post-injection (230).

Mast cell-specific proteases such as tryptase and chymase are stored inside the secretory granules in their active form. In addition, the mast cell granules contain other proteases such as CPA3, also expressed in basophils, as well as cathepsin C and G, found in e.g. neutrophils (231). Upon mast cell degranulation, the proteases are quickly released to the extracellular matrix. Tryptase, which is electrostatically bound to serglycin proteoglycans inside the acidic mast cell granules due to positively charged histidines, will gradually dissociate from the serglycin proteoglycans, possibly forming active monomers before losing its activity through the action of protease inhibitors (232–234). Chymase is also released as a complex with serglycin proteoglycans, but the association is not pH-dependent. Therefore, chymase may stay in complex with serglycin proteoglycan for an extended time and tends to remain close to the mast cell surface after degranulation (235, 236). As described earlier, the expression of tryptase and chymase has been used to classify human mast cells into subtypes. In the mouse, the major form of granule-stored tryptase is mMCP-6. In an acute OVA model of allergic airway inflammation, we found that mMCP-6-deficient mice had reduced methacholine-induced AHR but developed allergic airway inflammation (71). Moreover, the tryptase inhibitors nafamostat mesylate and gabexate mesylate reduced mast cell activation, AHR and eosinophil influx in an HDM model of allergic airway inflammation (237). A possible mechanism for the pro-inflammatory and broncho-constrictive effects of tryptase is cleavage and activation of protease activated receptor 2(PAR2). When trypsin-like proteases cleave PAR2 in the N-terminal, a tethered ligand sequence is exposed and the receptor auto-activates (238). For example, sensory neurons, which are frequently located in proximity to mast cells, express PAR2 and may thus be activated by mast cell tryptase (239). Moreover, immunostaining shows PAR2 expression in smooth muscle, epithelium, endothelium and glands of human bronchi (240). In the same study, tryptase and an artificial ligand peptide to PAR2 (SLIGKV-NH₂) were demonstrated to stimulate contraction of isolated human bronchi (240). In mouse models, PAR2 deficiency or blocking PAR2 with a monoclonal antibody reduces allergic lung inflammation and AHR (241–243).

In asthma patients who died due to an asthma attack, tryptase levels in serum were increased compared to those who died from other causes (190). Further, a study of 60 asthmatics classified according to pathological findings and standard clinical parameters to define asthma severity by Bayesian network analysis combined with topological analysis revealed six disease clusters (191). Two of the clusters demonstrated higher serum tryptase levels than the others. One cluster consisted of asthma patients with severe asthma who were generally older, had high BMI, poor lung function and many symptoms, while the other group consisted predominately of patients who were female, obese, non-atopic, had later onset and poor lung function. In children, basal serum tryptase levels were higher in those with mild, moderate or severe persistent asthma than in those with mild intermittent asthma and healthy controls (192). Nevertheless, more studies are needed to clarify the importance of mast cell tryptase and the mechanisms behind its role in allergic and non-allergic asthma.

In mice, the functional homolog of human chymase, mMCP-4, has a protective effect on AHR and lung inflammation. This was shown in two studies, first by analyzing mMCP-4^−/− mice in an OVA model of allergic airway inflammation and later in an HDM sensitization model (244, 245). The study by Waern et al. suggested that the protective effect of mMCP-4 is mediated by proteolytic cleavage of IL-33, leading to degradation of this important Type 2 cytokine (245). Moreover, mMCP-4 has the ability to suppress IL-13-induced enhancement of the contraction of the trachea in response to methacholine in vitro (246). The protective role of mast cell chymase is in accordance with a clinical study, mentioned earlier, where the number of chymase⁺ mast cells correlated positively with lung function in 20 severe asthmatics (48).

Mast cells are a well-known source of lipid mediators derived from arachidonic acid metabolized by different enzymes in response to various stimuli. Leukotrienes are metabolized from arachidonic acid through the 5-lipoxygenase (5-LO) pathway that generates LTA₄ as a main precursor of two different types of leukotrienes: the cysteinyl-leukotrienes (cysteinyl-LTs) LTC₄, LTD₄ and LTE₄, and LTB₄ (247).

The biological activities of the cysteinyl-LTs were discovered more than 30 years ago and included stimulation of smooth muscle contraction and vascular leakage (196–199). The levels of cysteinyl-LTs are increased in BAL (193) and sputum (194) of asthmatics, while LTE₄ levels are increased in the urine after allergen challenge (195). Inhalation of LTC₄ or LTD₄ causes bronchoconstriction, which is 1,000-fold more potent than histamine or methacholine (197–199). The cysteinyl-LTs mediate their biological effects through the receptors CySLTR₁, CySLTR₂ and GPR99 (248). Two other receptors, P2Y₁₂R and GPR17 are also involved in cysteinyl-LT and LTE₄-elicited responses, respectively. LTD₄ has a higher affinity for CySLTR₁ than LTC₄, whereas LTC₄ and LTD₄ have a similar affinity for CySLTR_2. However, LTE₄ has a high affinity for GPR99. CySLTR₁ is expressed by many different immune cells such as mast cells, neutrophils and monocytes/macrophages, whereas CySLTR₂ is expressed by eosinophils, macrophages and smooth muscle cells (248). The CysLTR₁ antagonist, montelukast, attenuates CysLT-induced bronchoconstriction and airway inflammation (200), confirming many earlier studies in animal models (249, 250). CysLTR₂ mediated vascular permeability in IgE-dependent passive cutaneous anaphylaxis and infiltration of macrophages and fibroblasts in a fibrosis model (251). Recently, a role for CysLTR₂ was demonstrated in IL-33-dependent type 2 immunity (252). Using various knockout strains, CysLTR₂ was shown to drive IL-33 expression induced by LTC₄ in two models of allergic airway inflammation. Human cord blood-derived mast cells express both CysLTR₁ and CysLTR₂ (253). In response to CysLTR₁ stimulation, human cord blood-derived mast cells proliferate in a process that is negatively regulated by CysLTR₂ (254). Moreover, stimulation of the human mast cell line LAD-2 with LTD₄ or LTE₄ leads to chemokine and prostaglandin D2 (PGD₂) release (255).

Inhaled LTE₄ induces the influx of inflammatory cells in the bronchial mucosa of asthma patients more potently than LTD₄, while they have similar bronchoconstrictive effects (256, 257). Nevertheless, LTE₄ has only weak affinity for the earliest discovered cysteinyl-LT receptors, CysLTR₁ and CysLTR₂. In mice, P2Y₁₂R and platelets were demonstrated to be required for LTE₄-induced pulmonary inflammation, although LTE₄ does not bind directly to P2Y₁₂R (258). Recently, a placebo-controlled randomized double-blind study of asthmatic patients demonstrated that the P2Y₁₂R antagonist prasugrel inhibited platelet reactivity but showed only a possible modest effect on mannitol-induced airway reactivity (259). After the discovery of the interaction of LTE₄-induced responses with P2Y₁₂R, another P2Y receptor family member called GPR17 was discovered to function as a negative regulator of CySLTR₁-mediated responses (260). However, while GPR17 has a role in regulating cysLTR₁-induced lung inflammation in a HDM model, it is unclear if this effect is directly mediated by Cys-LTs since different labs have conflicting data on whether Cys-LTs can activate GPR17 or not (261). Next, the P2Y receptor family member GPR99 was identified as the third high affinity cysteinyl-LT receptor, which preferentially binds LTE₄ (262). GPR99 was recently demonstrated to mediate the release of mucin from epithelial cells and swelling of nasal mucosa in mice subjected to a single intranasal dose of Alternaria extract or LTE₄ (263). However, in human asthmatics, LTE₄-induced bronchoconstriction was completely blocked by the CySLTR₁ inhibitor montelukast (264). The LTE₄-induced bronchoconstriction was associated with increases in urinary PGD₂ metabolites and other COX pathway products, which was also abrogated by montelukast, suggesting that LTE₄ activates mast cells (and possibly other cell types) via CySLTR₁ and that mast cell-derived products activate bronchial smooth muscle cells to constrict. These data suggest that CySLTR₁ mediates LTE₄-induced responses in the human lung or that montelukast also blocks GPR99.

The non-cysteinyl-LT, LTB₄, is increased in exhaled breath condensate and sputum of asthmatics (201, 265). In a recent study categorizing asthma patients according to GINA guidelines (step 1–3), patients treated with short-acting beta-agonist (step 1) or inhaled corticosteroids plus long-acting beta-agonist (step 3) had higher LTB₄ levels in sputum than those treated with inhaled corticosteroids (step 2) (266). This indicates that inhaled corticosteroids reduce LTB₄ levels in mild-moderate asthma as previously described in (267), but not in more severe asthma. LTB₄ is produced by various activated leukocytes including mast cells (268), and has a potent chemotactic effect on leukocytes such as neutrophils, T cells and immature BMMCs (202, 269, 270). LTB₄ mediates its effect through BLT₁R and BLT₂R, which are G-protein-coupled receptors. BLT₁R is a high affinity receptor for LTB₄ and is responsible for LTB₄-induced leukocyte migration (202, 248). In vivo, intradermal injection of LTB₄ induced the accumulation of intravenously injected immature BMMCs at the injection site (269).

In experimental allergic airway inflammation, BLT₁R mediates the infiltration of T cells to the lung (271) as well as AHR, eosinophilic inflammation and goblet cell hyperplasia (272). An indication of a role for mast cells in LTB₄-mediated allergic airway inflammation was demonstrated in a study using mice lacking LTA₄ hydrolase (LTA₄H), which thereby were unable to produce LTB₄ (273). The LTA₄H^−/− mice had reduced AHR and BAL eosinophilia compared to controls in an active systemic sensitization and challenge model, as well as when mice were passively sensitized and challenged. However, transfer of LTA₄H^+/+ BMMCs into LTA₄H^−/− mice normalized airway reactivity to methacholine assessed by electrical field stimulation of tracheal smooth muscle preparations, and partly restored eosinophilia in passively sensitized and challenged mice but not in mice subjected to active systemic sensitization and challenge (273). Thus, LTB₄ plays a role in allergic airway inflammation, and LTB₄ from mast cells may mediate some of these effects.

The low affinity LTB₄ receptor BLT₂R is expressed in the mouse lung. Intriguingly, BLT₂R-deficient mice displayed enhanced eosinophilia in an OVA model of allergic airway inflammation (274). This might be explained by the fact that a cyclooxygenase metabolite (12(S)-hydroxyheptadeca-5Z,8E,10E-trienoic acid) was shown to have higher affinity than LTB₄ for BLT₂R (275). To summarize, leukotrienes and many of their receptors represent important targets in allergic asthma; for a recent detailed review, see (248).

Prostaglandins are metabolites initially derived from the conversion of arachidonic acid by the cyclooxygenases (COX-1 and COX-2) to PGH₂, which subsequently is a substrate of PGD synthase forming PGD₂ or PGE synthase forming PGE₂. There are also other primary prostaglandins (PGF_2α, PGI₂, and thromboxane A₂). However, here we limit the discussion to PGD₂, which is the major mast cell-derived prostaglandin produced in large amounts through the action of hematopoietic prostaglandin D synthase (HPGDS) in response to IgE-antigen-mediated activation (276). Nevertheless, mouse and human eosinophils also produce PGD₂ (277). Extracellularly, PGD₂ is metabolized to different compounds such as 9α,11β-PGF₂, which can be measured in plasma as a possible indication of mast cell activation (204).

Allergen challenge in allergic asthmatics leads to an increased level of PGD₂ in BAL, and PGD₂ metabolites in plasma and urine (178, 203, 204). Inhalation of PGD₂ leads to bronchoconstriction in normal subjects and to an even stronger bronchoconstrictive response in patients with asthma (205). Despite corticosteroid use, the PGD₂ synthesis pathway is upregulated in asthmatics with severe and poorly controlled asthma (278). In the same study, the levels of HPGDS transcripts in epithelial brushings correlated strongly with the levels of TPSAB1/TPSAB2 (tryptase) transcripts, suggesting that mast cells are the main source of PGD₂, at least at this site.

PGD₂ acts via activation of the D-prostanoid receptors DP1 and DP2. Another name for DP2 is chemoattractant receptor-like protein expressed on Th2 cells, abbreviated CRTH2 (279, 280). The DP1 receptor is expressed on Th2 cells, dendritic cells, basophils, eosinophils, goblet cells and vascular endothelium (276). Recently, mast cell maturation was shown to be mediated by PGD₂ acting on DP1, a process which was driven by phospholipase A2 group III (PLA2G3) secreted from mast cells, which activated fibroblasts to produce PGD₂ by the action of lipocalin-type PGD₂ synthase (281). An amplifying role of PGD₂ in experimental asthma was discovered by comparing wild-type and mice lacking DP1 in an OVA-induced mouse model of allergic airway inflammation (282). In this study, DP1^−/− mice had attenuated AHR and the production of type 2 cytokines was diminished. Further, pre-treatment of sensitized mice with aerosolized PGD₂ before challenge amplified the type 2 response (283). In contrast, PGD₂ acts in an anti-inflammatory fashion via DP1 expressed on dendritic cells during the sensitization phase. PGD₂ or a DP1 agonist, but not a DP2 agonist, instilled intratracheally with FITC-OVA temporarily inhibited the migration of dendritic cells to the lung draining lymph nodes, which limited the expansion of adoptively transferred T cells (284). In a later study by Hammad et al. PGD₂ or an agonist for DP1 or DP2 was intratracheally delivered 30 min before each OVA challenge in an OVA model of allergic airway inflammation (285). Interestingly, the DP1 agonist (but not PGD₂ or the DP2 agonist) suppressed inflammation and AHR by a mechanism involving T regulatory cells producing IL-10. As the suppressive effect of the DP1 agonist could be mimicked by adoptively transferred dendritic cells pre-treated with the DP1 agonist. Hence, the DP1 agonist likely acted on lung dendritic cells which acquired a regulatory function by stimulating the expansion of T regulatory cells producing IL-10 (285). DP2/CRTH2 is expressed by various immune cells such as eosinophils, basophils, Th2 cells and ILC2s (286, 287). Moreover, Ptgdr2 transcripts have been found in BMMCs (288), and CRTH2 immunopositive human mast cells were found in nasal mucosa (289) and nasal polyps (290), although the expression seemed to be intracellular. Early studies demonstrated that CRTH2 mediates chemotaxis of human eosinophils, basophils and Th2 cells (206, 207). Further, patients with severe asthma have an increased level of CRTH2 transcripts in BAL cells compared to patients with mild to moderate asthma and heathy controls (278). In the same study, immunohistochemical analyses of CRTH2⁺ cells in BAL revealed a higher percentage of CRTH2⁺ cells in asthma patients compared to healthy controls, with the highest percentages found among those with severe asthma and as well as those with mild asthma but who did not have corticosteroid treatment (278).

In mice, sensitization with OVA followed by nebulization of a CRTH2 agonist before each challenge enhanced AHR and the eosinophilia in the BAL and lung tissue (291). In contrast, mice lacking Ptgdr2 have an enhanced BAL eosinophilia in OVA-induced allergic airway inflammation (292). This was likely due to the fact that T cells lacking CRTH2 produced higher levels of cytokines (292). However, a CRTH2 antagonist given in connection with OVA challenge in a model of allergic airway inflammation reduced eosinophilia and mucus hyperplasia (293). Similarly, mice sensitized and challenged with cockroach allergen given a CRTH2 antagonist before the final challenge had significantly reduced AHR, levels of type 2 cytokines in the lung and allergen-specific IgE and IgG2a in serum (294). For these reasons, CRTH2 antagonists have been tested as a treatment for asthma. Several different CRTH2 antagonists have been taken into clinical trials, but some of them have been discontinued for various reasons like poor effect or pharmacokinetics (295). However, other CRTH2 antagonists have been demonstrated to improve lung function, symptom scores and reduced sputum eosinophilia (296–298). Still more studies are needed to define what asthma phenotypes will benefit (most) from treatment with a CRTH2 antagonist.

A more unknown lipid mediator that is generated and released from mast cells after IgE-antigen-mediated activation is sphingosine-1-phosphate (S1P) (299). S1P is produced as a result of sphingosine (Sph) phosphorylation by two Sph kinases (SphK1 and SphK2). In mast cells, IgE-antigen-mediated crosslinking of FcεRI induces the activation and translocation of SphK1 to the plasma membrane and leads to increased levels of S1P, which then becomes secreted extracellularly (299). S1P is produced by most cell types but is usually degraded or dephosphorylated intracellularly, and the level in the body is tightly regulated (300). S1P binds to five different receptors (S1PR1-S1PR5) expressed by both innate and adaptive immune cells. The main outcome of S1P binding to these receptors seems to be directed migration. In patients with asthma, the S1P concentration is increased in BAL after allergen challenge (301). In the same study, stimulation of human airway smooth muscle cells by the S1P-activated signaling pathways was involved in contraction, proliferation and stimulated IL-6 release. In addition, S1P induced contraction of human airway smooth muscle cells embedded in collagen matrices (302).

In mice, an inhibitor of Sph kinases given before each OVA challenge in a model of allergic airway inflammation reduced inflammatory BAL cells and type 2 cytokines, AHR and mucus production (303). In isolated bronchi from mice, S1P enhanced acetylcholine-induced contraction (304). Further, S1P induced contraction of OVA-sensitized lungs but not in naïve lungs as demonstrated both in isolated bronchi and in vivo measured by whole-body plethysmography (304). In a follow-up study, Roviezzo et al. demonstrated that subcutaneous administration of S1P day 0 and day 7 dose-dependently increased bronchial responsiveness in vivo and ex vivo (305). The effect was time-dependent, with the greatest effect found 21 days after S1P administration. At this time point the number of mast cells quantified in BAL was twice that found in the vehicle controls. Using the same protocol of S1P-induced asthma-like inflammation, the same authors recently demonstrated that while LPS potentiated S1P-induced AHR, TLR4-defective mice (C3H/HeJ) or BALB/c mice pre-treated with an TLR4 blocking antibody were protected from S1P-induced AHR (306). Moreover, S1P induced higher expression of TLR4 in the proximity of the bronchi, and immunoprecipitation revealed an increased association between S1PR1 and TLR4. Together, this study suggests a functional interaction between S1PR1 and TLR4 that amplifies allergic airway inflammation and airway reactivity.

Mast cells were implicated in S1P-mediated responses in vivo in a model of IgE-mediated systemic anaphylaxis. In this study, pre-treatment with a S1P neutralizing antibody delayed and reduced peri-vascular lung inflammation along with BAL tryptase and histamine serum levels in mice injected intraperitoneally with dinitrophenyl (DNP)-specific IgE 12 h before injection of DNP coupled to human serum albumin (307). In the same model, mast cell deficient mice (Kit^{W−sh/W−sh}) or mice pre-treated with an S1PR2 antagonist or an anti-S1P antibody had reduced early peri-vascular lung inflammation. This suggests that mast cells providing and/or responding to S1P via S1PR2 are responsible for early peri-vascular lung inflammation observed in the model of IgE-mediated systemic anaphylaxis. In humans, an inhibitor of all S1PRs except S1PR2 (FTY720, fingolimod) is approved as a drug to treat multiple sclerosis (308). This drug was tested on moderate asthmatics in a double-blind, placebo-controlled 10-day study (309). Patients that received high doses of fingolimod showed only mild reductions in some lung function parameters and needed more short acting beta agonists, perhaps suggesting that although S1PR2 may be a target for asthma, the other S1P-receptors should not be targeted.

Cytokines play a crucial role in coordinating and maintaining allergic inflammation in the asthmatic lung. For a recent update on the general importance of cytokines in asthma, please see (310). Mast cells have the ability to secrete a broad spectrum of cytokines e.g. IL-1β, IL-2,-3,-4,-5,-6,-9,-10,-11,-12,-13, TNF-α, IFN-γ, GM-CSF, SCF, and TGF-β, as recently reviewed (311). In asthma patients, mast cells have been demonstrated to produce and release IL-4 (312). Interesting to note is that mouse MCp also have the ability to produce IL-4 (313). Possibly, this Th2-skewing cytokine derived from mast cells may promote or support early Th2-differentiation, although the contribution of mast cell-derived IL-4 in asthmatic lungs is unknown. Tryptase⁺ mast cells producing IL-4, as well as IL-5, IL-6 and TNF, were found in the bronchial mucosa from both healthy and asthmatic subjects (314). Moreover, isolated human mast cells produce IL-13 upon IgE-crosslinking (315), and IL-13 and IL-4 expressing mast cells are found within the airway smooth muscle of asthma patients (316). In mice, a role for IL-13, FcεRI and mast cells in the development of AHR was demonstrated in a model of OVA-induced allergic airway inflammation (77). Human lung mast cells also produce TNF-α (317). As already described, mast cell-derived TNF-α has been suggested to contribute to the development of allergic airway inflammation, Th2/Th17 cytokine production and AHR in mice (70, 73).

IL-9 is a pleiotropic cytokine, produced by many types of immunes cells including mast cells (318). Genetic studies have revealed that the IL-9 receptor is associated with asthma susceptibility (319, 320), and elevated IL-9 mRNA and IL-9 immunoreactive cells are found in lungs from asthmatics compared to patients with chronic bronchitis, sarcoidosis and healthy controls (321). In mouse models, lung selective IL-9 overexpression leads to mast cell accumulation, along with massive airway inflammation and AHR in the absence of an antigen (322). Nevertheless, as reviewed in (323), there is some conflicting evidence for the significance of IL-9 in mouse models of asthma, indicating that an alternative pathway exists. In the context of allergic asthma, mast cells responding to IL-9 signals are likely more important than mast cells as a source of this cytokine, as discussed earlier and demonstrated in different experimental set-ups of allergic lung inflammation in mice. In principle, IL-9 seems to be required for antigen-induced accumulation of lung mast cells in mice. Interestingly, a humanized anti-IL-9 monoclonal antibody (MEDI-528) has been tested on 36 mild asthmatics (injected subcutaneously) (324). The results demonstrated a trend toward a protective effect of MEDI-528 on lung function (324). However, in a randomized, controlled trial of 329 patients in which MEDI-528 (or placebo) was added to their usual asthma treatment over 24 weeks, no improvements in quality of life, asthma exacerbation rates or FEV₁ values (pre-bronchodilator) were found (325). Thus, the future of targeting IL-9 in asthma is unclear although it is still possible that this pathway may be a valid target for a subgroup of asthma patients. To summarize, mast cells can produce many different cytokines that have been demonstrated to play significant roles in allergic asthma. However, whether or not the mast cell-produced cytokines play a major critical role in the human disease is currently unclear.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.