IL-33 and Superantigenic Activation of Human Lung Mast Cells Induce the Release of Angiogenic and Lymphangiogenic Factors

Human lung mast cells (HLMCs) express the high-affinity receptor FcεRI for IgE and are strategically located in different compartments of human lung, where they play a role in several inflammatory disorders and cancer. Immunoglobulin superantigens (e.g., protein A of Staphylococcus aureus and protein L of Peptostreptococcus magnus) bind to the variable regions of either the heavy (VH3) or light chain (κ) of IgE. IL-33 is a cytokine expressed by epithelial cells that exerts pleiotropic functions in the lung. The present study investigated whether immunoglobulin superantigens protein A and protein L and IL-33 caused the release of inflammatory (histamine), angiogenic (VEGF-A) and lymphangiogenic (VEGF-C) factors from HLMCs. The results show that protein A and protein L induced the rapid (30 min) release of preformed histamine from HLMCs. By contrast, IL-33 did not induce the release of histamine from lung mast cells. Prolonged incubation (12 h) of HLMCs with superantigens and IL-33 induced the release of VEGF-A and VEGF-C. Preincubation with IL-33 potentiated the superantigenic release of histamine, angiogenic and lymphangiogenic factors from HLMCs. Our results suggest that IL-33 might enhance the inflammatory, angiogenic and lymphangiogenic activities of lung mast cells in pulmonary disorders.


Effects of Human IgG Anti-IgE on the Release of Angiogenic and Lymphangiogenic Factors from HLMCs
We have previously reported that IgG anti-IgE purified from the serum of a small percentage of atopic dermatitis patients induced histamine release from human basophils [85] and lung mast cells [94]. The activating effects of human IgG anti-IgE (H-aIgE) were mediated by the interaction with membrane-bound IgE on human basophils and mast cells. In a series of experiments, we used this human autoantibody to activate HLMCs in vitro. H-aIgE (10 −2 to 3 µg/mL) caused a concentration-dependent release of both angiogenic (VEGF-A) and lymphangiogenic factors (VEGF-C) from four different preparations for HLMCs ( Figure 1A). As a control, we found that the same concentrations of H-aIgE induced a concentration-dependent release of histamine. Similar results were obtained when HLMCs were activated by increasing concentrations (10 −1 to 3 µg/mL) of monoclonal antibody (mAb) anti-FcεRI (Table 1). Three preparations of human polyclonal IgG (10 −2 to 3 µg/mL) did not cause the release of histamine, VEGF-A, and VEGF-C (Table 2). These results indicate that mast cells isolated from human lung parenchyma express IgE bound to FcεRI. Figure 1B shows that there was a significant correlation between the production of VEGF-A and histamine release caused by H-aIgE (r = 0.76; p < 0.001). Similarly, there was a significant correlation between the production of VEGF-C and histamine release (r = 0.57; p < 0.05) ( Figure 1C) and between the production of angiogenic (VEGF-A) and lymphangiogenic (VEGF-C) factors (r = 0.89; p < 0.001) ( Figure 1D). production of VEGF-A and histamine release caused by H-aIgE (r = 0.76; p < 0.001). Similarly, there was a significant correlation between the production of VEGF-C and histamine release (r = 0.57; p < 0.05) ( Figure 1C) and between the production of angiogenic (VEGF-A) and lymphangiogenic (VEGF-C) factors (r = 0.89; p < 0.001) ( Figure 1D).

Figure 1. (A) Effects of increasing concentrations of human
IgG anti-IgE purified from the serum of an atopic dermatitis patient [85,95] on histamine release and the production of VEGF-A and VEGF-C from four different preparations of human lung mast cells (HLMCs). HLMCs were incubated (45 min at 37 °C) with the indicated concentrations of IgG anti-IgE for histamine secretion or (12 h at 37 °C) for VEGF-A and VEGF-C release. Each bar is the mean ± SEM; (B) Correlation (r = 0.76; p < 0.001) between VEGF-A release and the percent histamine secretion caused by human IgG anti-IgE from HLMCs; (C) Correlation (r = 0.57; p < 0.05) between VEGF-C release and the percent histamine secretion caused by human IgG anti-IgE from HLMCs; (D) Correlation (r = 0.89; p < 0.001) between VEGF-A and VEGF-C release caused by human IgG anti-IgE from HLMCs. Table 1. Effects of increasing concentrations of monoclonal antibody anti-FcεRI on histamine release and the production of VEGF-A (angiogenic) and VEGF-C (lymphangiogenic) from human lung mast cells.

Effect of Superantigenic Protein A on the Release of Angiogenic and Lymphangiogenic Factors from HLMCs
S. aureus colonization is associated with bronchial asthma [52,95]. S. aureus superantigens trigger airway inflammation and increased airway responsiveness, and facilitate allergic sensitization in asthma models [96]. It has been shown that S. aureus and protein A can activate human mast cells through different mechanisms [47,97]. More recently, we have demonstrated that protein A induced the release of lipid mediators from human cardiac mast cells through the engagement of IgE V H 3 + bound to FcεRI [98]. Figure 2A shows that protein A (30 to 600 nM) caused a concentration-dependent release of both VEGF-A and VEGF-C from different preparations of HLMCs. The same concentrations of protein A caused a dose-dependent release of histamine. Protein A contains five homologous repeated domains, each of which binds to human Igs, including IgE [42,43]. Preincubation (15 min, 37 • C) of protein A (300 nM) with IgM V H 3 + (10 µg/mL), but not IgM V H 6 + (10 µg/mL), blocked the histamine-releasing activity of protein A (Table 3). These results suggest that the immunoglobulin superantigen protein A activates HLMCs through the binding to IgE V H 3 + bound to FcεRI.  There was a significant correlation between the release of VEGF-A and histamine secretion (r = 0.59; p < 0.05) caused by protein A ( Figure 2B). In addition, there was a significant correlation between VEGF-C production and histamine secretion (r = 0.82; p < 0.001) ( Figure 2C) and between VEGF-A and VEGF-C production (r = 0.64; p < 0.01) ( Figure 2D).

Effects of IL-33 on the Release of Angiogenic and Lymphangiogenic Factors from HLMCs
Several investigators have found that IL-33 can induce the release of different cytokines from human cord blood (CBMCs) and peripheral blood-derived mast cells (PBMCs) [17,[68][69][70][71]78,99]. By contrast, IL-33 had no effect on the release of preformed mediators from mouse and human mast cells [100]. Interestingly, long-term incubation of human skin mast cells (HSMCs) with IL-33 reduced anti-IgE-induced histamine secretion Each bar is the mean ± SEM; (B) Correlation (r = 0.59; p < 0.05) between VEGF-A release and the percent histamine secretion caused by protein A from HLMCs; (C) Correlation (r = 0.82; p < 0.001) between VEGF-C release and the percent histamine secretion caused by protein A from HLMCs; (D) Correlation (r = 0.64; p < 0.01) between VEGF-A and VEGF-C release caused by protein A from HLMCs from HLMCs.
There was a significant correlation between the release of VEGF-A and histamine secretion (r = 0.59; p < 0.05) caused by protein A ( Figure 2B). In addition, there was a significant correlation between VEGF-C production and histamine secretion (r = 0.82; p < 0.001) ( Figure 2C) and between VEGF-A and VEGF-C production (r = 0.64; p < 0.01) ( Figure 2D).

Effect of Short-Term Priming by IL-33 on Superantigenic Release of Mediators from HLMCs
It has been reported that short-term priming by IL-33 can potentiate the release of cytokines induced by different stimuli from mouse [73] and human mast cells [17,20,69,71,74,101]. In a series of experiments, HLMCs were preincubated (30 min, 37 °C) with IL-33 (30 ng/mL) before exposure to protein A (100 nM). The results presented in Figure 4A confirm that IL-33 alone has no effect on histamine release whereas it induced both VEGF-A ( Figure 4B) and VEGF-C ( Figure 4C) from HLMCs. Interestingly, short-term priming by IL-33 potentiated the release of VEGF-A ( Figure 4B) and VEGF-C

Effect of Short-Term Priming by IL-33 on Superantigenic Release of Mediators from HLMCs
It has been reported that short-term priming by IL-33 can potentiate the release of cytokines induced by different stimuli from mouse [73] and human mast cells [17,20,69,71,74,101]. In a series of experiments, HLMCs were preincubated (30 min, 37 • C) with IL-33 (30 ng/mL) before exposure to protein A (100 nM). The results presented in Figure 4A confirm that IL-33 alone has no effect on histamine release whereas it induced both VEGF-A ( Figure 4B) and VEGF-C ( Figure 4C) from HLMCs. Interestingly, short-term priming by IL-33 potentiated the release of VEGF-A ( Figure 4B) and VEGF-C ( Figure 4C) induced by protein A from HLMCs.  Protein L is another immunoglobulin SAg, which binds to human Igs through a mechanism different from protein A [45,102]. This protein binds with high affinity (~10 10 M −1 ) only to subtypes (V K I, V K III, and V K IV) of κ light chains of human Igs, including IgE [46,48,49]. We have found that protein L is a superantigenic stimulus inducing the release of preformed and the de novo synthesized inflammatory mediators from human cardiac mast cells through the interaction with κ light chains of IgE [98]. In a series of preliminary experiments, we found that increasing concentrations (1-300 nM) of protein L caused the release of inflammatory mediators from HLMCs (data not shown). Table 4 shows that protein L (100 nM) induced histamine release from HLMCs. Preincubation of protein L with increasing concentrations (0-3-1 µg/mL) of human monoclonal IgE λ light chain did not modify the activating property of protein L. By contrast, preincubation with the same concentrations of human monoclonal IgE κ light chain concentration-dependently inhibits the release of histamine induced by protein L from HLMCs. These results are compatible with the hypothesis that protein L activates HLMCs through the interaction with the κ light chain of IgE on lung mast cells. We then examined the interactions between protein L and IL-33 on the activation of HLMCs. Figure 5 shows the results of a typical experiment showing that protein L induces the release of histamine from HLMCs ( Figure 5A). By contrast, IL-33 did not induce histamine secretion from HLMCs ( Figure 5A) but caused a small production of VEGF-A ( Figure 5B) and VEGF-C ( Figure 5C). Short-term incubation (30 min, 37 • C) of HLMCs with IL-33 (30 mg/mL) potentiated the secretion of histamine ( Figure 5A) and the release of VEGF-A ( Figure 5B) and VEGF-C ( Figure 5C) induced by protein L from HLMCs. Similar results were obtained in two additional experiments with different preparations of HLMCs (data not shown).

Discussion
Primary mast cells isolated from human lung parenchyma can be activated by a human IgG anti-IgE isolated from a patient with atopic dermatitis to release histamine, VEGF-A, and VEGF-C. Similar results were obtained by activating HLMCs with a monoclonal antibody anti-FcεRI. These findings indicate that HLMCs expressing FcεRI bind IgE, which is a central immunoglobulin in the pathogenesis of several allergic disorders [103,104] and pulmonary diseases [6,13]. Two bacterial superantigens, protein A and protein L, which bind to distinct regions of human IgE [43][44][45]49], activate HLMCs to release inflammatory (histamine), angiogenic (VEGF-A) and lymphangiogenic (VEGF-C) factors. IL-33, expressed and released by lung epithelial cells and endothelial cells [63,77], does not induce histamine release from HLMCs, but potentiates the release of VEGF-A and VEGF-C induced by superantigens from HLMCs. , 10, x FOR PEER REVIEW 11 of 21

Discussion
Primary mast cells isolated from human lung parenchyma can be activated by a human IgG anti-IgE isolated from a patient with atopic dermatitis to release histamine, VEGF-A, and VEGF-C. Similar results were obtained by activating HLMCs with a monoclonal antibody anti-FcεRI. These findings indicate that HLMCs expressing FcεRI bind IgE, which is a central immunoglobulin in the pathogenesis of several allergic disorders [103,104] and pulmonary diseases [6,13]. Two bacterial superantigens, protein A and protein L, which bind to distinct regions of human IgE [43][44][45]49], activate HLMCs to release inflammatory (histamine), angiogenic (VEGF-A) and lymphangiogenic (VEGF-C) factors. IL-33, expressed and released by lung epithelial cells and endothelial cells [63,77], does not induce histamine release from HLMCs, but potentiates the release of VEGF-A and VEGF-C induced by superantigens from HLMCs.
Asthma is a heterogeneous syndrome that has been subdivided into phenotypes and molecular endotypes [109]. Type 2 (T2)-high subtype asthma is characterized by IgE-mediated activation of lung mast cells and includes the majority of asthmatic patients [103,104]. T2-low asthma is less well-characterized and presumably includes different clinical and genetic variants [110,111]. T2-low asthma may be driven by abnormal neuronal activation, structural abnormalities involving airway smooth muscle as well as bacterial and viral superantigens [112,113]. We found that cross-linking the IgE-FcεRI network on HLMCs induces the release of histamine. More importantly, we found that IgE-mediated activation of primary HLMCs also induces the production of VEGF-A and VEGF-C. The latter findings extend previous results indicating that activation of different types of human mast cells causes the release of VEGF-A [14,16,17,94,98]. Interestingly, IgE-mediated activation of HLMCs induces the release of VEGF-C, which is the most potent lymphangiogenic factor [114,115]. Mast cells are strategically located in different compartments of human lung [3,4,78]. These cells and their mediators play a central role in the pathophysiology of bronchial asthma [1,2,9], lung remodeling [105], angiogenesis [14,16,17,98], lymphangiogenesis [14,18,94,98], chronic obstructive pulmonary disease (COPD) [106], and lung cancer [107,108].
Asthma is a heterogeneous syndrome that has been subdivided into phenotypes and molecular endotypes [109]. Type 2 (T2)-high subtype asthma is characterized by IgE-mediated activation of lung mast cells and includes the majority of asthmatic patients [103,104]. T2-low asthma is less well-characterized and presumably includes different clinical and genetic variants [110,111]. T2-low asthma may be driven by abnormal neuronal activation, structural abnormalities involving airway smooth muscle as well as bacterial and viral superantigens [112,113]. We found that cross-linking the IgE-FcεRI network on HLMCs induces the release of histamine. More importantly, we found that IgE-mediated activation of primary HLMCs also induces the production of VEGF-A and VEGF-C. The latter findings extend previous results indicating that activation of different types of human mast cells causes the release of VEGF-A [14,16,17,94,98]. Interestingly, IgE-mediated activation of HLMCs induces the release of VEGF-C, which is the most potent lymphangiogenic factor [114,115].
Several investigators have provided evidence that bacterial superantigens play a role in different allergic disorders [40,116]. We found that low concentrations of Staphylococcal protein A induces the release of histamine, VEGF-A, and VEGF-C from HLMCs. These findings might be relevant to explain the role of bacterial superantigens in the pathogenesis of various allergic disorders. The role of S. aureus in the pathogenesis of allergic diseases has been attributed to its capacity to activate T and B cells, resulting in cell proliferation and massive cytokine release [117]. On the other hand, it has been shown that Staphylococcal superantigens can induce the formation of IgE antibodies [95,118] and the presence of specific IgE has been associated with the severity of airway and skin allergic disorders [52,119,120]. There is also evidence that S. aureus can trigger the production of cytokines from human mast cells through the engagement of TLR2 and CD48 molecules [97]. Our results indicate that protein A might contribute to the role played by S. aureus in allergic diseases by induc-ing the release of histamine and angiogenic and lymphangiogenic factors from HLMCs through the interaction with the V H 3 region of IgE.
Protein L is another Ig superantigen which specifically interacts with high affinity with the κ light chains of human Igs, including IgE [42]. In this study, we found that protein L is a potent stimulus to induce the secretion of histamine and the release of VEGF-A and VEGF-C from HLMCs. These results extend previous findings indicating that protein L induced the release of preformed (i.e., histamine) and de novo synthesized mediators (i.e., prostaglandin D 2 : PGD 2 ) from human cardiac mast cells [98]. Collectively, these findings indicate that protein L is a complete mast cell secretagogue capable of releasing inflammatory, angiogenic and lymphangiogenic mediators implicated in cardio-pulmonary pathophysiology [121][122][123].
Angiogenesis plays a role in pulmonary pathophysiology [18,134,135]. VEGF-A is a major mediator of angiogenesis and can be produced by several immune cells [14,36,37,136,137]. To our knowledge, this is the first evidence that superantigens protein A and protein L induce the release of angiogenic factors from HLMCs raising the possibility that these cells can contribute to angiogenesis, a process of pivotal relevance in bronchial asthma [18,134] and lung cancer [135]. Further studies are needed to comprehensively define the contributive role of IL-33 and superantigens to angiogenesis in pulmonary disorders.
The mammalian lung is rich in lymphatic vessels [138] which are increased in human lung following infections [139][140][141]. We provide the first evidence that a superantigenic activation of HLMCs leads to the production of VEGF-C, a major mediator of lymphangiogenesis [142]. Lymphangiogenesis is canonically considered pivotal for the diffusion of metastasis to draining lymph nodes [143,144]. However, recent evidences indicate that VEGF-C can potentially exert protective effects, since inflammation-associated lymphangiogenesis can improve the resolution of inflammation [115,145]. Therefore, the contribution of bacterial superantigens to lung mast cell-mediated lymphangiogenesis requires additional investigations.
To the best of our knowledge, we provide the first evidence that IL-33 can induce the release of the lymphangiogenic factor VEGF-C from HLMCs. This finding extends a previous observation indicating that immunologically-activated human cardiac mast cells release VEGF-C [98]. The production of VEGF-C by activated primary human mast cells is intriguing because these cells are at the interface of the lymphatic and immune systems [146]. In several clinical and experimental studies, mast cells play a pro-tumorigenic role, whereas in others, they play an anti-tumorigenic role [25,26,143,147]. VEGF-C is mostly viewed as the most potent lymphangiogenic factor [115] controlling the formation of metastasis. However, increasing evidences indicate that, under certain circumstances, lymphangiogenesis and VEGF-C have protective effects in cancer [148]. Moreover, VEGF-C can exert a protective role in several inflammatory disorders [149,150] by favoring the resolution of inflammation [142,151,152]. The pathophysiological role of VEGF-C released by human lung and cardiac mast cells [98] deserves further investigations.
Several studies have provided evidence that IL-33 can exert a priming effect on the activation of rodent [73] and human mast cells [17,20,69,71,74]. A previous study reported that preincubation of LAD2 cells and CBMCs with IL-33 augmented SP-induced VEGF-A mRNA and VEGF-A protein secretion [17]. In our study, we found that shortterm incubation of HLMCs with IL-33 potentiates the release of histamine, VEGF-A, and VEGF-C induced by superantigens protein A and protein L. The interactions between IL-33 and IgE-mediated stimuli (i.e., protein A and protein L) is unlikely mediated by an overexpression of FcεRI because it has been demonstrated that IL-33 does not increase FcεRI expression on human mast cells [74].
IL-33 is an alarmin overexpressed in lung epithelial cells [99,129] in asthmatic patients. IL-33 expression has been also identified in human airway smooth muscle (ASM) and HLMCs in mild-to-moderate asthma [99]. Using quantitative morphometry of the airway wall, it has been demonstrated that IL-33 causes a shift in mast cells from the submucosa to the airway epithelium associated with type 2 inflammation and airway hyperresponsiveness (AHR) [78]. A recent study highlighted a novel pathogenetic mechanism of interaction between IL-33 and S. aureus as inducers of airway inflammation in mice [118]. Intratracheal exposure to S. aureus derived serine protease-like protein (Spl) D upregulated IL-33 production in the lung leading to eosinophilia, bronchial hyperreactivity, and goblet cell hyperplasia in the airways. Interestingly, blocking IL-33 activity with a soluble ST2 receptor significantly reduced airway inflammation. Our findings highlight a novel immunologic mechanism by which IL-33 and superantigen protein A can amplify the immune response in inflammatory disorders involving lung mast cells. Collectively, these findings might have translational relevance emphasizing the relevance of IL-33/S. aureus-derived proteins in inducing inflammatory response in the airways [78,99,153]. The translational significance of the interactions between IL-33 and bacterial superantigens deserves further investigations in experimental models of asthma.
Our study has several limitations that should be pointed out. The in vitro experiments were performed using primary mast cells purified from lung parenchyma of patients undergoing thoracic surgery for lung cancer. The purity of HLMCs used in these experiments ranged from 49% to 81%. Although the activating properties of protein A, protein L and IL-33 were not affected by mast cell purity, we cannot exclude the possibility that contaminating cells may have influenced some of our results. In addition, IL-33 is released by tumor cells [84] and by a myriad of immune and non-immune cells localized in human lung [64]. There is the possibility that in vivo exposure of HLMCs to IL-33 could explain some of our results. In addition, HLMCs, although obtained from macroscopically normal lung tissue, are in close proximity to lung cancer cells. There is the possibility that the in vivo exposure to altered tumor microenvironment, such as low pH [154], hypoxia [155,156], lactate [157], or adenosine [37,158,159], may have affected the phenotypic expression and the functional activity of pulmonary mast cells.
In conclusion, our results indicate that two immunoglobulin superantigens, protein A and protein L, can interact with different domains of human IgE bound to FcεRI to induce the release of inflammatory, angiogenic and lymphangiogenic factors from human lung mast cells. IL-33 synergistically potentiates the superantigenic release of mediators from these cells. Future studies are needed to investigate whether these in vitro observations can help to understand the in vivo interactions between IL-33 and S. aureus in inflammatory airway disorders.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.