Social Networking of Group Two Innate Lymphoid Cells in Allergy and Asthma

Allergic diseases including asthma, chronic rhinosinusitis, and atopic dermatitis are common conditions worldwide. While type 2 immune responses induced by T-cells significantly cause allergic inflammation, the recently identified group two innate lymphoid cells (ILC2s) are emerging as critical players in the development of allergy. Upon allergen exposure, ILC2s are rapidly activated by cytokines released by epithelial cells. Activated ILC2s release various effector cytokines altogether contributing to the pathogenesis of allergy and can even cause inflammation in the absence of T-cells, as observed in asthma. Although the factors inducing ILC2 activation have been identified, evidence suggests that multiple factors can enhance or repress ILC2 proliferation, trafficking, or secretion of effector cytokines upon allergic inflammation. In this review, we discuss the recent findings that influence ILC2 activation and the resulting effects on the pathogenesis of allergy. A better understanding of how ILC2s are modulated will open the door to the development of new therapeutic strategies against allergic diseases.


INTRODUCTION
Allergic diseases are highly diverse and common conditions caused by the inappropriate sensitization of the immune system to environmental antigens. Upon re-exposure to these allergens, sensitized individuals develop allergic symptoms including rash, wheezing, and shortness of breath. Most inflammatory responses seen in allergic diseases are caused by the release of type 2 cytokines from activated T helper 2 (Th2) cells. Main features of allergic reactions include smooth muscle cell contraction, mucus production, release of IgE, increased vascular permeability and recruitment of effector cells including eosinophils, basophils, and mast cells (1). However, type 2 cytokines are not only produced by Th2 cells, with Th9 (2), follicular T helper cells (3,4) and inflammatory cells further contributing to type 2 cytokine secretion. Invariant natural killer T-cells also produce large amounts of cytokines and induce airway inflammation independent of T-cells (5). Recently, group 2 innate lymphoid cells (ILC2s) were described as a source of cytokines during allergic inflammation.
ILC2s are a subset of the innate lymphoid cells family described in three independent studies (6)(7)(8), following pioneer work in the early 2000s (9). Mouse and human ILC2s are phenotypically comparable, lineage negative, non T-, non B-lymphocytes ( Table 1). Unlike T-cells, ILC2s lack antigen specific markers and instead are rapidly activated by alarmins released following tissue damage, pathogen recognition or allergen challenge. Activated ILC2s release high amounts of type 2 cytokines and contribute to a growing number of human diseases (10,11) including chronic rhinosinusitis with nasal polyps (CRSwNP) (12)(13)(14)(15)(16)(17) and atopic dermatitis (AD) (18,19), two common allergic diseases of the paranasal sinuses (upper airways) and skin, respectively. Several studies have described the involvement of ILC2s in human allergic asthma (20)(21)(22)(23)(24). Asthma is a common, heterogeneous chronic inflammatory disease of the lower airways characterized by airway hyperreactivity (AHR) and reversible bronchoconstriction. Patients with asthma have a greater number of total and activated blood ILC2s compared to healthy controls (20), with increased numbers of ILC2s further detected in bronchioalveolar lavage (BAL) fluids of asthmatics (21). In line with this, ILC2s in blood and sputum are increased in severe compared to mild asthma patients (24), and in the sputum of children with severe asthma (23). Interestingly, increased ILC2 numbers are correlated with increased eosinophilia (22,24). Altogether, these findings suggest that ILC2s are critical in human asthma. According to a recent genetic cluster analysis, there are at least five different clinical phenotypes of asthma (25). One cluster showed higher blood and sputum eosinophils, driven by a Th2dominant inflammatory response (26). Multiple studies have shown that ILC2s are directly involved in eosinophilic asthma (11). They are located near the basement membrane subjacent to the airway epithelium, residing within 70 µm of airway branchpoints (27). This strategic location allows them to act as sentinels and rapidly respond to allergen exposure. The main activators of ILC2s are alarmins released by activated epithelial cells such as IL-33, IL-25, or TSLP (18,(28)(29)(30)(31)(32). Activated ILC2s release various effector cytokines including IL-4, IL-5, IL-9, and IL-13 (6)(7)(8). IL-5 and IL-13 cause eosinophilia and smooth muscle cell contraction respectively, altogether contributing to the pathogenesis of asthma (33,34). Interestingly, activated ILC2s also release considerable amounts of pleiotropic cytokines IL-6 and GM-CSF (35)(36)(37). While IL-6 is known to induce the development of Th17 cells from naïve T-cells (38), ILC3derived GM-CSF is a key regulator of oral tolerance to dietary antigens by modulating macrophage effector functions (39). It remains to be elucidated whether ILC2-derived IL-6 and GM-CSF can have similar immunoregulatory effects in the context of allergic diseases. Furthermore, ILC2s contribute to tissue homeostasis through the secretion of amphiregulin (40). Several studies show that activated ILC2s enhance Th2-cell activation in response to allergens (41)(42)(43). However, in the absence of T-cells, effector cytokines released by ILC2s in response to intranasal challenge with alarmins are strikingly sufficient to induce airway inflammation and AHR (44,45). ILC2s are therefore emerging as important players in the pathogenesis of allergic diseases such as asthma, and a better understanding of their function will open the door to the development of new therapeutic strategies. Factors modulating ILC2 functions in allergic diseases will be discussed in this review, summarized in Figure 1.

Regulatory Cytokines
We and others recently reported that regulatory T-cells (Treg)-derived cytokines suppress the development of ILC2dependent lung inflammation (46). Induced Tregs (iTregs), rather than natural Tregs (nTregs) efficiently suppress the production of ILC2-derived IL-5 and IL-13, ultimately inhibiting the development of airway inflammation and AHR. The suppressive effects of Tregs depend on ICOS:ICOSL interactions, but also on the release by Tregs of IL-10 and TGFβ. In line with our results, studies have confirmed the suppressive effects of IL-10 and TGF-β on mouse and human ILC2s in the context of asthma (47) and upper airway inflammation (48). Further studies are however required to better characterize the role of TGF-β, as epithelial-derived TGF-β was recently shown to drive pulmonary inflammation (49).
Interferons have various immunomodulatory functions and are classified in two families: Type I Interferons such as IFN-α and IFN-β, and Type II Interferons such as IFN-γ (50). We recently reported that plasmacytoid DC (pDC)-derived IFN-α directly suppresses ILC2 activation in models of asthma (51). In our study, we showed that IFN-α inhibits pulmonary ILC2-derived secretion of IL-5 and IL-13, ultimately preventing the development of airway inflammation and AHR. Depletion of pDCs strikingly reverses the suppressive effects on ILC2s. Interestingly IFN-β and IFN-γ also strongly suppress pulmonary ILC2 proliferation FIGURE 1 | Factors modulating ILC2 activation in the context of allergy. Main ILC2 activators are alarmins released by epithelial cells, as activated ILC2s release various effector cytokines including IL-4, IL-5, IL-6, IL-9, IL-13, and GM-CSF. Several groups of molecules can enhance or inhibit ILC2 activation. These include contact independent pathways such as regulatory cytokines, hormones, growth factors, complement peptides, neuropeptides, neurotransmitters, and lipid mediators. Contact-dependent pathways include co-stimulatory and adhesion molecules. and cytokine production, suggesting that both Type I and Type II Interferons can dampen ILC2-derived lung inflammation (52,53).
The classical view of the Th1/Th2 paradigm supports that Th1 cytokines inhibit Th2 cell differentiation and vice versa (54). This holds true for ILC2s, as they resemble Th2 cells. IL-12, a Th1 cytokine, reduces ILC2 activation and promotes their transition to T-bet high GATA-3 low ILCs (55), suggesting that ILC2s retain plasticity. Furthermore, IL-27, a member of the IL-12 cytokine family, suppresses ILC2 cytokine production in the lungs (53,56). On the other hand, Th2 cytokines were reported to enhance ILC2 cytokine production. A study shows in a model of lung inflammation that key type 2 cytokine IL-4 derived from basophils enhances ILC2 secretion of IL-5 and IL-13, ultimately favoring eosinophilia (57). Besides being known as a Th9 cell signature cytokine, IL-9 is also required for the survival and homeostasis of ILC2s (58,59). Furthermore, lung ILC2s themselves secrete IL-9 as autocrine IL-9 is crucial for ILC2 effector functions (60,61). Similar to T-cells, ILC2s require survival factors for efficient activation (62). IL-2 is crucial in the maintenance of ILC2 activation (43,63), as they further rely on IL-7 for their development (64,65) and efficient activation (20). A recent report however reveals that IL-7 is not strictly required for the development of ILC2s (66). Interestingly, human lung ILC2s are further activated by IL-1β and closely related cytokine IL-18 (37,67). In addition to affecting ILC2 activation, several lines of evidence suggest that cytokines from the local microenvironment affect ILC2 plasticity depending on the context. ILC2s may develop into specific subsets or even express an ILC1-phenotype, although further studies are warranted to better understand such processes (68)(69)(70).

Co-stimulatory Molecules
ILC2s express multiple receptors on their surface that bind to ligands present on other immune cells. Among those are co-stimulatory molecules, known to modulate T-cell activation (71). Both mouse and human lung ILC2s express Inducible T-cell co-stimulator (ICOS) at steady state but also upon inflammation (72)(73)(74)(75), and we were the first to show that they also express ligand ICOS-L (76). We show that the ICOS:ICOS-L trans-interaction is crucial for ILC2 homeostasis and effector functions in models of lung inflammation. Compared to controls, mice genetically deficient in ICOS develop less AHR and lung inflammation as a result of a defect in pulmonary ILC2-derived IL-5 and IL-13 secretion and increased apoptotic rates. The observed effects were strictly ILC2-dependent, as alymphoid mice adoptively transferred with ICOS-deficient ILC2s develop less AHR compared to control mice. In a separate study, we further studied the interaction of pulmonary ILC2 surface expression of ICOS with its ligand on T-cells. Strikingly, we found that binding of iTregs via ICOS-L to ILC2s via ICOS inhibits ILC2 cytokine secretion and development of AHR (46). A recent study reveals that PD-1, another costimulatory molecule, is expressed on ILC2s and acts as a negative regulator of ILC2s by inhibiting proliferation and IL-13 cytokine production in a model of lung inflammation (77). Although PD-1 is expressed on ILC2s, further studies are required to better understand its function in the context of allergic asthma. For example, it is not clear whether a PD-1 agonist can be used as a therapeutic agent in the context of ILC2-dependent asthma.
Members of the tumor necrosis factor receptor superfamily (TNFRSF) and their ligands (TNFSF) provide key co-stimulatory signals to T-cells (78). They contribute to T-cell homeostasis and induce or restrict immune responses. Such findings have led to the design of treatments of autoimmune diseases and tumors (79,80). Similar to T-cells, several TNFRSF and their ligands are involved in ILC2 homeostasis and activation in the context of allergic diseases. TNFRSF25 (DR3) was first shown to be required for mouse and human ILC2 expansion and function (81,82). ILC2s express TNFRSF25, and engagement with TNFSF15 induces ILC2 expansion, survival and cytokine secretion in the lungs (82). Interestingly, TNFSF15 alone is sufficient to activate ILC2s. In a mouse model of dermatitis, a recent study further shows that TNFSF15 activates skin ILC2s in a TNFRSF25-dependent manner (83). A report describes the co-stimulatory role of TNFRSF18 (GITR) in ILC2-dependent lung inflammation (61). Interaction of the receptor with GITR-L (DTA-1) enhances autocrine IL-9-induced IL-5 and IL-13 secretion by ILC2s, ultimately driving lung inflammation. However, further studies are required to better characterize the function of TNFRSF18 on ILC2s during inflammation. Upregulation of the tumor-associated surface molecule B7-H7 is observed in human AD lesions (84). Binding to its receptor NKp30 on ILC2s induces type 2 cytokine secretion, suggesting this pathway may be involved in ILC2-derived skin inflammation (84). Although no effect on ILC2 activation was reported, a recent study describes ILC2s as a source of TNFSF4 (OX40L), promoting IL-33driven Th2 and Treg lung inflammation by binding to TNFRSF4 (OX40) (85,86). Altogether, these studies suggest that co-stimulatory molecules are potent modulators of ILC2 activation.

Lipid Mediators
Lipids are generally known as a source of energy for the human body and crucial components of cellular membranes (87). However, eicosanoids such as prostaglandins (PG), cysteinyl leukotrienes (cystLT) and lipoxins (LX) are bioactive lipids also involved in cell signaling (88,89). In the context of ILC2-driven allergic inflammation, several prostaglandins were described to modulate ILC2 functions (90)(91)(92). The most studied prostaglandin is PGD2, which binds to CRTH2 and induces human and mouse ILC2 chemotaxis and type 2 cytokine production in the inflamed lungs (90,93,94). Unlike PGD2, other prostaglandins were shown to inhibit ILC2 functions (91,92). In a mouse model of airway inflammation, PGI2 binds to PGI2 receptor IP on ILC2s and reduces the number of lung-expressing IL-5 and IL-13 ILC2s (91). Another recent study shows that PGE2 inhibits human tonsillar ILC2 proliferation and cytokine secretion by binding to EP2 and EP4 on ILC2s (92). Cysteinyl leukotrienes are important inflammatory mediators in the context of allergy. Mouse and human ILC2s express cystLT-R1, as LTD4 was first shown to induce ILC2 proliferation and production of cytokines during lung inflammation (95). Furthermore, it was recently reported that LTC4 given intranasally with low dose IL-33 increases lung ILC2 proliferation and type 2 cytokine secretion in mice (96). In another recent study using an AD model, LTE4 induces migration, reduces apoptosis and enhances cytokine secretion in human ILC2s (97). Finally, lipoxins are are generally associated with resolution of inflammation (98), with LXA4 shown to inhibit IL-13 production on activated human ILC2s (99). Altogether, these studies suggest that lipid mediators are potent modulators of ILC2 activation and chemotaxis.

Adhesion Molecules
Besides their role in cell migration, adhesion molecules are required for efficient, tight cell-to-cell interactions and can function as co-stimulatory molecules (100). ILC2s express integrins such as leukocyte function-associated molecule (LFA-1, α L β 2 ), interacting with members of the intercellular cell adhesion molecule (ICAM) family of ligands (101). Interestingly, disruption of LFA-1 and ICAM-1 binding impaired the development of airway inflammation (102). A study recently showed that mouse and human ILC2s express both LFA-1 and ICAM-1 (103). This study elegantly shows that LFA-1 is required for ILC2 migration from the circulation to the lungs during airway inflammation, although it did not affect ILC2 functions. This study strongly supports that ILC2s are not only resident cells and can also be recruited to inflamed lungs, an observation also made by others (104). Besides these observed effects, ICAM-1 is furthermore required for ILC2 homeostasis and efficient activation in the lungs, as absence of ICAM-1 specifically on ILC2s significantly inhibits IL-5 and IL-13 secretion and development of airway inflammation (105). Killer cell lectin-like receptor G1 (KLRG1) -E-cadherin interactions inhibit ILC2 activation in the context of human AD lesions (19). Skin ILC2s of AD lesions express high levels of KLRG1 compared to healthy controls. Interaction with E-cadherin, an adhesion molecule found on epithelial cells, reduces ILC2 proliferation and cytokine secretion in response to both IL-25 and IL-33. The KLRG1-E-cadherin interaction is of particular interest, as loss of E-cadherin on lung epithelium in humans is linked to asthma severity (106). Altogether, adhesion molecules are emerging as efficient modulators of ILC2 homeostasis, function, and trafficking.

Neuropeptides and Neurotransmitters
Neuropeptides and neurotransmitters are proteins used by neurons to communicate with each other. Interestingly, the immune and nervous systems are closely linked (107). Neurotransmitter acetylcholine binds to receptors including nicotinic acetylcholine receptors (nAChRs). As opposed to muscarinic AChRs, nAChRs also respond to nicotine, ultimately affecting immune responses (108). We were the first to show that human and mouse ILC2s express a7nAChR, regulating ILC2-mediated lower airway inflammation and AHR (109). Engagement of the a7nAChR with GTS-21, a specific agonist, inhibits ILC2-derived IL-5 and IL-13 secretion, ultimately inhibiting the development of airway inflammation and AHR. ILC2s also express β2-adrenergic receptor (β2-AR), known to interact with neurotransmitter epinephrine, and was recently described as a negative modulator of ILC2 activation. This study shows that mice treated with a β2-AR agonist exhibit less lung ILC2 proliferation and cytokine production in response to IL-33 (110).
Neuropeptides can amplify allergic lung inflammation (27,111). First, a study shows that ILC2s express Neuromedin U receptor 1 (NMUR1) at steady state and during inflammation (111). NMUR1 ligand Neuromedin U (NMU) amplifies IL-25induced allergic lung inflammation by increasing ILC2-derived IL-5 and IL-13 secretion. As a result, mice co-challenged with NMU and IL-25 develop increased airway inflammation and AHR compared to controls. Another study recently shows that neuropeptide calcitonin gene-related peptide (CGRP) induces ILC2-driven allergic lung inflammation by enhancing ILC2-derived IL-5 secretion (27). Interestingly, the source of CGRP are pulmonary neuroendocrine cells (PNECs) that reside in close proximity to ILC2s at airway branchpoints. Additionally, lung ILC2s release IL-5 after stimulation with the neuropeptide vasoactive intestinal peptide (VIP). VIP signals through VIP receptor type 2 (VPCA2) expressed on ILC2s, as they release IL-5 after stimulation with a specific agonist (34). Altogether, these studies suggest that neuronal products are emerging as potent modulators of ILC2 activation.

Hormones, Growth Factors and Complement Peptides
Studies have shown that asthma incidence differs depending on the sex (112). In human asthmatics, the number of blood ILC2s is increased in women compared to men (113). This highly suggests a role for sex hormones as regulators of the development of asthma, as they are already known to affect T-cell differentiation and cytokine secretion in a different context (114,115). Three studies recently established a role for androgens in ILC2-driven airway inflammation (113,116,117). A study shows that male mice develop less severe IL-33-induced allergic asthma compared to females (116). Interestingly, this difference is due to an increase in androgen receptor (AR) signaling, which ultimately decreases ILC2-dependent airway inflammation. In line with this, a group recently found that a derivative of testosterone, 5α-dihydrotestosterone (5α-DHT) directly inhibits lung ILC2 proliferation and secretion of IL-5 and IL-13 in response to IL-33 (113). As a result, testosterone decreases Alternaria extract-induced airway inflammation. Interestingly, lung ILC2s from gonadectomized females secrete less IL-5 compared to controls, suggesting that ovarian hormones may also affect ILC2 homeostasis and function (118). In line with this, ILC2s were shown to be regulated by female sex hormones in the uterus (119).
Members of the vascular endothelial growth factors (VEGF) including VEGF-A, VEGF-C, and VEGF-D are secreted by multiple immune cells (120). Recently, a study shows that both human and mouse ILC2s strikingly promote AHR via the production of VEGF-A (121). ILC2s stimulated with IL-33 release autocrine VEGF-A, which binds to surface VEGFR2, altogether promoting cytokine secretion and lung inflammation. Treatment of mice with a specific VEGFR2 inhibitor significantly inhibits the development of AHR in response to allergen. Interestingly, ILC2s from patients with asthma express increased VEGF-A transcripts. Finally, complement system activation and generation of anaphylatoxins, or complement peptides, induces and regulates the development of type 2 responses at mucosal surfaces (122). A study recently shows that complement peptide C3a increases ILC2 numbers in the lungs, as well as their secretion of IL-13 and GM-CSF in response to IL-33 (36). This novel mechanism by which C3a drives type 2 immunity in the lungs is of particular interest as elevated levels of C3a were found in asthmatics airways (123).

Concluding Remarks and Future Directions
ILC2s are potent producers of type 2 cytokines, and it is therefore not surprising that they are involved in the development of various allergic diseases including asthma. Treating such diseases by targeting upstream ILC2 activators such as alarmins will likely have unwanted adverse effects on other immunological pathways. In recent years however, multiple pathways were described to modulate ILC2 effector functions, ultimately affecting the pathogenesis of allergic diseases. Such findings provide valuable information for the design of novel therapeutic strategies, largely dependent on corticosteroids in the context of asthma and AD. Open questions however remain to be tackled. First, although research over the past years has revealed the expression of multiple key receptors on ILC2s, no specific marker has yet to be identified (Table 1). Second, although ILC2s are considered as resident cells (124), emerging data suggests that they express chemotactic/trafficking molecules upon inflammation and are therefore also recruited to inflamed tissues, as discussed (103,104). A better understanding of ILC2 trafficking and tissue tropism will provide valuable information for the treatment of allergic diseases. Third, several lines of evidence suggest that ILC2s retain plasticity and adapt to signals from the local microenvironment, such as composition of the local cytokine pool. ILC2s can develop into specific sub phenotypes, such as the newly described IL-10 producing ILC2 10 (69) or IL-17 producing ILC2 17 (70). Furthermore, several lines of evidence describe conversion from ILC2 to an ILC1-like phenotype such as following smoke exposure (68). It will be crucial to delineate the factors inducing plasticity or commitment among ILC2s particularly in the context of allergic diseases. Finally, since asthma is a heterogeneous disease, it will be essential to find a biomarker characterizing the cohort of asthma patients with increased airway ILC2 activity.

FUNDING
This article was financially supported by National Institutes of Health Public Health Service grants R01 ES025786, R01 ES021801, R21 ES024707, R21 AI109059 (OA). BH is supported by the Swiss National Foundation for Early Postdoc.Mobility #181286.