Important and specific role for basophils in acute allergic reactions

Summary IgE‐mediated allergic reactions involve the activation of effector cells, predominantly through the high‐affinity IgE receptor (FcεRI) on mast cells and basophils. Although the mast cell is considered the major effector cell during acute allergic reactions, more recent studies indicate a potentially important and specific role for basophils and their migration which occurs rapidly upon allergen challenge in humans undergoing anaphylaxis. We review the evidence for a role of basophils in contributing to clinical symptoms of anaphylaxis and discuss the possibility that basophil trafficking during anaphylaxis might be a pathogenic (to target organs) or protective (preventing degranulation in circulation) response. Finally, we examine the potential role of basophils in asthma exacerbations. Understanding the factors that regulate basophil trafficking and activation might lead to new diagnostic and therapeutic strategies in anaphylaxis and asthma.

Despite an increasing number of studies using mouse models demonstrating an important role for basophils in orchestrating proallergic Th2-type immune responses and mediating chronic allergic inflammation, extrapolation to humans is highly problematical (Table 1). This is because of substantial differences in basophil morphology and relative expressions of various cell surface receptors, as well as different outcomes of their subsequent stimulation. 1,2 While recent studies suggest that murine basophils produce similar inflammatory mediators to human basophils, 3 sensitivities to the biological effects of these mediators differ from one species to another. For example, Berman & Munoz showed that the LD 50 of histamine (thought to be an important mediator of anaphylaxis) in mice was studies assessing the role of basophils in anaphylaxis, given that basophils are relatively uncommon in comparison with their tissue-fixed mast cell counterparts in both mice and humans. However, despite their relative rarity, human basophils are at least one order of magnitude more sensitive to IgE-mediated provocation than mast cells. 5

| IgE-versus IgG-mediated anaphylaxis
Multiple pathways of anaphylaxis are described in mice. It has been shown that, upon capture of IgG-allergen complexes, mouse basophils release platelet activating factor (PAF) that increases vascular permeability, leading to anaphylactic shock. In vivo depletion of basophils protects mice from fatal IgG-mediated anaphylactic shock, but has no effect on IgE-mediated anaphylaxis. Thus, Tsujimura and Karasuyama 6,7 postulated that there are two major distinct pathways of anaphylaxis in mice: one is mediated by basophils, allergen-IgG-FccRII-III receptor interactions and PAF release, whereas the other is mediated by mast cells, allergen-IgE-FceRI receptor interactions and histamine release. Previous murine studies similarly showed that only mast cells contributed to IgE-mediated anaphylaxis. 8,9 There are also alternative IgG pathways of murine anaphylaxis, mediated by IgG-FccRIII-macrophages or IgG-FccRIV-neutrophil interactions. 10,11 In addition, the role of neutrophils was also demonstrated in peanut-induced anaphylaxis in mice. 12 More recently, Finkelman et al reviewed the evidence for IgG versus IgE-mediated anaphylaxis in mice, arguing that dose of allergen is an important factor in determining the precise mechanism of induction. 13 In sharp contrast, human basophils cannot be activated through IgG receptors, as their function is inhibited by IgG-mediated triggering via FccRIIb receptors which are the predominant IgG receptor subtype on these cells. 14,15 Moreover, allergen-specific IgG antibodies are of questionable pathogenic relevance 16 and are more associated with blocking the effects of allergen-specific IgE. 17,18 Furthermore, there is little evidence that human anaphylaxis is in any way mediated by IgG antibodies in relation to either macrophages or neutrophils. Evidence for PAF production by human (as opposed to murine) basophils is also limited and inconsistent. 19,20 1.1.2 | Antigen presentation Murine basophils appear to be able to present antigens through MHC class II-dependent interactions. [21][22][23] However, the role of murine basophils as IL-4-releasing antigen-presenting cells (APC) is limited by the observations that basophils and dendritic cells (DCs) could efficiently co-operate, where basophils produce IL-4, whereas DCs present antigens. 24,25 Eckl-Dorna et al 26 and Kitzmuller et al 27 compared the antigen-presenting properties of different human cell types including basophils. Human basophils were not able to present allergens to T lymphocytes, whereas a mixture of APCs depleted of basophils did. Furthermore, human basophils lacked the machinery to uptake, process and present allergens, although a small increase of MHC-II was seen after incubating the basophils with both IFN-c and IL-3. There are some reports that basophils in patients with systemic lupus erythematosus express MHC-II, 28 but these data are not confirmed in other studies. 29 In addition, human basophils lack protease-activated receptor expression (PAR), and PAR ligands fail to induce activation of these cells. 30 In contrast, PAR activators, such as papain, which have been used in many of the mouse models, are able to elicit murine basophil-mediated Th2 response. 21 T A B L E 1 Differences in the pathophysiology of anaphylaxis in murine models compared to humans (adapted from Turner and Campbell 113 )

| Skin
The skin might be an important route of allergen exposure, 35 especially in the case of skin barrier disruption, 36 and significant increases in the numbers of basophils were previously observed 6 hours after intradermal injection of allergen, 37 or, in patch-test skin sites, for house dust mite allergen. 38 Furthermore, basophil infiltration into skin lesions seems to be more common than previously thought, indicating that they may play important roles in a variety of inflammatory skin diseases. 39 Higher number of basophils were detected in inflammatory skin diseases where eosinophils are present, 39 and those observations are consistent with a recent study which demonstrated a significant correlation between airway basophils and eosinophils in asthma patients. 40

| Controlled allergen challenge studies
Currently, there are two models for studying anaphylaxis in humans: emergency department (ED) studies and controlled challenge models (mostly to food, but also to Hymenoptera venom). 42 Smith et al 43  Allergen-induced reactions often manifest themselves as an early asthmatic response, and bronchial allergen challenge may be another model for study of basophils during the acute allergic reaction. 51 In addition, nasal allergen challenges could also be employed as an experimental set-up to study the role of basophils in local allergic reactions. 31,52

| Emergency department-based studies
The ED-based anaphylaxis study was first described by Lin et al, 53  there are several commercial kits (basophil activation tests). 67 Additional options to assess basophil activation include measurement of CD203c and CD11b, which are located in a rapidly expressed vesicular compartment that is distinct from the histamine-containing granules, and CD69, which is not related to secretion but is expressed when basophils are exposed to cytokines, such as IL-3. 65 Under in vitro basophil stimulation experiments with different types of stimuli (allergens, anti-IgE, anti-FceRI mAbs or fMLP), upregulation of CD63 generally parallels degranulation and histamine release. 66,67 However, the situation in vivo is not so clear. Turner et al 68 reported increased expressions of CD63, CD107a and CD203c on basophils following double-blind, placebo-controlled peanut challenge in 13 peanut-allergic subjects (P < .01). This is consistent with data from another food challenge study which included 12 subjects with IgE to galactose-alpha-1,3-galactose who experienced a delayed clinical response to mammalian meat. 49 Two subjects experienced anaphylaxis, and 8 experienced mild reactions. Nine of those subjects (including 2 asymptomatic) showed increased expression of CD63 (median 30% basophils, range 17%-67%). However, in the same study, 5 of 13 healthy controls, without IgE to galactose-alpha-1,3-galactose, showed comparable increase in CD63 expression (median 34% basophils, range 17%-46%) after meat challenge, but without any clinical symptoms.
Gober et al 48  CCL5 binds both to chemokine receptors CCR1 and CCR3, but with higher affinity to CCR1 than to CCR3. CCR1 is also present both on basophils and on eosinophils. CCL2 binds to the chemokine receptor CCR2, which is present on basophils but is undetectable on human eosinophils 78 and it fails to induce eosinophil transendothelial migration. Therefore, in contrast to CCL11 and CCL5, which also induce eosinophil migration, CCL2 preferentially induces basophil migration and may represent a unique mechanism for the selective migration of human basophils.

| Migration during anaphylaxis
The results of the experimental allergen challenge in the nose, airways and skin have demonstrated the influx of basophils to inflammatory sites several hours after allergen exposure. 31,37,51,79 A recent study has suggested that basophils migrate from the circulation during anaphylaxis, both in ED and controlled allergen challenge models. 50 In the ED study, which included predominantly venom-allergic adult patients, there was a substantial reduction (80%) in circulating basophils during anaphylactic reactions, and these findings were CRTH2 + Basophils F I G U R E 1 Basophil activation without evidence of degranulation following oral challenge in peanut-allergic subjects (n = 4). Blood samples were collected prior to, during and 24 h after objective allergic reaction at oral food challenge, as previously described. 50 Surface expression of CD63 (A) on basophils was evaluated (without further ex vivo stimulation) by flow cytometry. 50 Basophils were isolated by Ficoll-density centrifugation and purified to over 90% purity by immunomagnetic cell sorting, using a negative selection technique which we have previously described. 111 Cells were incubated for 15 min at 37°C before stimulation with crude peanut extract (CPE) or anti-IgE for 8 min after which histamine release was assessed by spectrofluorometric autoanalysis according to Shore et al 112 (B). Data are shown as mean percentage histamine releases AE SEM. Despite increased surface CD63 expression on ex vivo, unstimulated basophils (A), there was no difference in IgEmediated histamine release in the same basophils compared with baseline (B). This implies that circulating basophils have become "activated"or rather, have increased surface expression of CD63, an activation marker-but with no evidence of degranulation, at least in terms of histamine release. These data presented at the 45th annual meeting of the European Histamine Research Society (EHRS) in Florence, 2016 expressed in populations of T cell progenitors and thymic T cells. 83 HDC catalyses the formation of histamine from L-histidine, and in hematopoietic cell lineages, the gene is expressed only in mast cells and basophils. 84 Importantly, the expression of all three genes significantly decreased during anaphylaxis, and correlated with the absolute number of circulating basophils, indicating that the decrease in whole blood gene expression of FCER1A, CPA3 and HDC was due to reduced number of basophils in blood.

IgE stimulation for basophil migration
The results of experimental allergen challenge in various organs 31,37,51,79 and recent anaphylaxis studies reveal that basophils migrate during acute allergic reactions. However, the specific mechanism(s) at play causing basophil migration during allergic reactions is unclear. In the previously mentioned study assessing human anaphylaxis, 50

| TH E ROLE OF BASOPHILS IN ASTHMA
Basophil infiltration in the airways and subsequent activation or immunomodulatory roles might be an important part of asthma pathogenesis and/or exacerbation.

| Early and late asthmatic response
Inhalation of allergen leads to an early asthmatic response, which is associated with a decrease in lung function that occurs within 2 hours, caused by the release histamine and cysteinyl leukotrienes from mast cells. 95 In some patients, the early response is However, this study was performed only on stable patients who had no exacerbations for at least the preceding two months. Salter F I G U R E 2 Hypothetical role of basophil migration in anaphylaxis. Upon allergen challenge, basophil-directed chemokine CCL2 (possibly secreted from mast cells) induces a rapid migration of basophils out of the circulation. This may reactogenic, with migration to target organs resulting in activation and degranulation. Alternatively, the migration may be a protective response, removing basophils from the circulation so that they are unable to degranulate in response to circulating allergen. CCL2, chemokine (C-C motif) ligand 2; CCR2, C-C chemokine receptor type 2; IgE, Immunoglobulin E; IgE receptor, high-affinity IgE receptor (FceRI) F I G U R E 3 Hypothetical role of basophil migration in asthma exacerbation. Following exposure to allergen or respiratory viral infection, basophil chemotactic factors are released in lungs leading to recruitment of basophils from the circulation to the airways where they may contribute to the early asthmatic response. In some patients, a Th2-type immune response orchestrated by basophils, mast cells and infiltration of eosinophils can cause late asthmatic response, resulting in prolonged swelling of the airway mucosa and aggravating the airway obstruction. CCL2, chemokine (C-C motif) ligand 2; CCR2, C-C chemokine receptor type 2; IgE, Immunoglobulin E; IgE receptor, high-affinity IgE receptor (FceRI); IL-4, interleukin 4 et al 98 also demonstrated increased expression of CD203c in blood, bone marrow and sputum basophils after allergen challenge. However, CD203c is located only in a rapidly expressed vesicular compartment that is distinct from the histamine-containing granules, 65 and thus, it can be not concluded whether basophil degranulation and/or secretion of immediate mediators occurs during asthma exacerbation. The source of basophils in the airways of asthmatic patients should be circulating basophils, but there is no current direct experimental evidence which can confirm basophil migration from the circulation to the airways during asthma exacerbations. Assessing the basophil absolute count and/or whole blood expression of genes specific for basophils during asthma exacerbation or after allergen challenge, and comparing them with baseline values, would be an obvious approach. In anaphylaxis models, the induction of migration seems to be related to IgE-and FceRI-cross-linking upon allergen contact. 87 However, this might not be the case in asthma, as basophils are also increased in the airways of non-allergic asthmatics, 40,41,99  shown that in vitro TSLP-primed basophil migrate to CCL11 chemokine by upregulation of CCR3 expression. 98 Finally, basophil chemotactic factors such as CCL2 may also be important for basophil migration in asthma patients, similar to anaphylaxis. 50 This is supported by recent observations that CCL2 is released by airway smooth muscles in asthma patients, and that levels of CCL2 are increased in the serum of asthma patients. 110 However, substantially broader studies are required to confirm or refute these speculations.