The Expanding Role of ILC2s in Allergic Airways Disease

Abstract

Group 2 innate lymphoid cells (ILC2s) play a pivotal role in the initiation and propagation of allergic airways disease. These cells, first discovered 15 years ago, respond to a range of stimuli in a non-antigen-dependent manner. ILC2s produce copious amounts of cytokines including IL-5 and IL-13, which are critical in the pathogenesis of allergic asthma. While allergic airways diseases have long been considered T-helper 2 cell-driven diseases, ILC2s are capable of inducing allergic-type pathologies in mice even in the absence of the adaptive immune system. The role of ILC2s in driving the pathology of allergic airways disease remains an expanding field of knowledge that may unlock novel therapeutic approaches in the management of asthma, a disease that affects up to 300 million people worldwide. In this review, we survey current knowledge of ILC2 immunobiology and present an overview of ILC2 phenotyping, concurrent with insights into ILC2 plasticity, and an exploration of the roles of costimulatory molecules, neuroendocrine signals, and diet-derived nutrients in modulating ILC2 activity.

1 |. Introduction

Innate lymphoid cells (ILCs) are typically characterized into three groups, all originating from common lymphoid progenitor cells (CLPs) [1]. The first group comprises natural killer (NK) cells and group 1 innate lymphoid cells (ILC1s) [2]. While there are many differences between NK cells and ILC1s, and even variation between ILC1 subtypes, the production of IFNγ seems to be a unifying quality defining this group of ILCs [1]. The second group includes only group 2 innate lymphoid cells (ILC2s), characterized by their notable expression of transcription factor GATA3 as well as their production of IL-4, IL-5, IL-9, IL-13, and amphiregulin [1, 3]. Group 3 ILCs include lymphoid tissue inducer (LTi) cells and ILC3s. As with group 1 ILCs, notable differences exist between LTis and ILC3s; however, both cell types produce IL-22 and commonly IL-17 [2].

First identified in adipose tissue, and shortly after in the lung and gut, as a unique cell population producing TH2-associated cytokines, IL-5, IL-6, and IL-13, but distinctly lacking expression of T- and B-cell lineage markers and recombined antigen receptor [4, 5], ILC2s have since been found to play a key role in various diseases, notably allergic asthma. Moreover, in the past decade much has been discovered about ILC2 biology and their implication in the pathogenesis of allergic asthma, their role in type-2 diabetes, and their contribution to immune defense against helminth infections [6].

Whilst T-helper (TH)-2 cells have been fundamentally implicated in the pathogenesis of allergic airways disease since the 1990's [7], the complementary role played by ILC2s is now becoming increasingly apparent. ILC2s differ from T-cells in that they do not rely on antigen-specific activation pathways, allowing them to act as critical first responders to localized tissue damage. Whilst the majority of studies on ILC2s have so far relied on extracting these cells from homogenized tissue, there is a rapidly expanding literature highlighting the importance of the microenvironmental niche, particularly in the lung, where ILC2s most commonly are located. Recent advances in immune-histological localization have demonstrated that ILC2s are predominantly embedded in the micro-niche surrounding airway branchpoints in mid- and large-sized airways, where they rely on a unique relationship with IL-33-producing stromal cells to hardwire their lung-specific phenotype [8, 9].

ILC2s are recruited to sites of local inflammation in the lung by airway epithelial cells, alveolar macrophages, and dendritic cells releasing the chemokines CCL17 and CCL22, both of which act via the CCR4 receptor on ILC2s. ILC2s are rapidly activated in response to alarmins (IL-25, IL-33, TSLP) released by damaged epithelial cells. Once activated, ILC2s release copious amounts of type 2 cytokines including IL-5 and IL-13, two critical inflammatory signals implicated in the development of allergic asthma. ILC2-derived IL-5 plays a dual role in asthma: Firstly, IL-5 in the blood provides a strong signal to the bone marrow which promotes eosinophilopoiesis [10, 11] and secondly, in the lung parenchyma and airways, IL-5 acts synergistically with other chemoattractant molecules such as leukotriene B4, promoting the attraction of eosinophils to the site of inflammation [10–12]. IL-13 is another key cytokine, centrally involved in the development of asthma [13, 14]. The cytokine drives club cells to differentiate into goblet cells, leading to goblet cell hyperplasia in the inflamed airways resulting in increased mucus secretion in the inflamed airways [15–17]. Chronic allergic airway diseases such as allergic asthma also feature a high degree of lung remodeling, characterized by excessive peribronchiolar collagen deposition, driven by the effects of IL-13 on airway fibroblasts [18]. Furthermore, IL-13 causes direct damage to epithelial cells, causing them to produce dysfunctional tight junctions, rendering airways leakier. Epithelial cells will also release eotaxin-1 in response to IL-13 [19–21], which plays a key role in attracting more eosinophils to the inflamed airways. IL-13 also interacts with airway smooth muscle, causing hyperreactivity resulting in exaggerated bronchoconstriction. Immunoglobulin E (IgE) production by B-cells is another hallmark characteristic of allergic asthma. IgE binds to mast cells in the lung, forming dimers with other IgE molecules, which triggers the rapid degranulation of the mast cell, resulting in the release of vast quantities of cytokines and the bronchoconstrictor histamine. Subsets of murine ILC2s which express Fas (CD95) have been found to greatly enhance IgE production by B-cells in vitro [22]. Furthermore, human ILC2s are also reported to enhance IgE production by B-cells due to their expression of CD40 ligand, a molecule which binds B-cell receptor CD40 to stimulate IgE production [23].

Due to their significant contribution to many of the immune signaling pathways contributing to the pathogenesis of allergic asthma, ILC2s have begun to emerge as a potential diagnostic marker [24–26]. Understanding the biology of ILC2s could therefore have wide-ranging implications for the future diagnosis and management of allergic asthma, hopefully unlocking new and more effective therapeutic avenues. In this review, we explore notable advances in our understanding of ILC2s and how they operate in the context of allergic airways disease (AAD), a term used herein to refer to a collection of human diseases such as allergic asthma and murine models that seek to mimic the pathology of such diseases. In the past 15 years, the importance of costimulatory molecules, neuroendocrine signals, and intracellular micronutrients has started to be investigated, showing many novel stimulatory and regulatory pathways governing ILC2 activity that could be exploited therapeutically in the future.

2 |. Defining and Categorising ILC2s

Since their initial characterization as non-T, non-B cells implicated in the development of type-2 inflammatory diseases such as allergic asthma, much has been learned about ILC2s. These cells are commonly classified by the lack of expression of lineage markers (Mouse: CD3, CD34, CD45R, CD94, LY-6G, CD11b, CD11c, Ter119, NK1.1, TCR_γδ, TCR_αβ, FCεRI, Mac-1. Human: CD1a, CD3, CD14, CD16, CD19, CD20, CD34, CD94, CD123, CD235a, CD303, CD11b, FCεRI, TCR_γδ, TCR_αβ) and by positive expression of CD45, ST2/CRTH2, CD127 (Table 1). Developments over the past 15 years have led to the identification of a host of biomarkers associated with ILC2s, although not uniquely so.

TABLE 1 |.

Cell surface markers commonly used in the immunophenotyping of murine and human pulmonary ILC2s.

Marker	Mouse	Human
CD45	+	+
CD90	+	−
CD25	+	+
c-KIT (CD117)	+	+/−
ST2	++	+
IL-17Rb	+	+
CD161	−	+
CRTH2 (CD294)	+	++
KLRG1	+	+
Sca-1	+	−
CD194	−	+
CD44	+	+
CD69	+	+
ICOS (CD278)	+	+
ICOSL (CD275)	+	+
TNFR2 (CD120b)	−	+
CD200R1	−	+
CTLA4 (CD152)	+	+
TSLPR	+	+
CD30	−	+

Note: “+” to indicate positive identification, “++” to indicate where a marker is relatively more expressed in either mouse or human ILC2, “+/−” to indicate where large heterogeneity has been reported, and “−” to indicate lack of positive identification.

Many of the cell surface markers associated with ILC2s exhibit a degree of dynamism, varying in their expression between cells in a naïve versus an activated state (Table 2).

TABLE 2 |.

Relative expression of pulmonary ILC2 markers varies between activation states.

Marker	Naive ILC2s	Activated ILC2s	Notes
CD25	Low	High	Activation marker
CD69	Low	High	Early activation
TSLPR	Low	High	Activation
CD30	Low/absent	High	Activation
TNFR2	Low/absent	High	Activation
ICOS	Moderate	High	Activation
CRTH2	High	Downregulated	Human ILC2s
c-KIT	High	Downregulated	Activation
KLRG1	Moderate	High	Activation
GATA3	High	Variable	Down in instances of plasticity
ST2	High (mouse) low (human)	Downregulated	Reports suggest up to 40% of activated murine ILC2s downregulate ST2 expression [27]

3 |. ILC2 Plasticity

ILC2s have been reported to be able to take on the characteristics of both ILC1 and ILC3 cells under certain conditions (Figure 1). This functional plasticity and its implications for ILC2-driven diseases are still to be fully understood.

FIGURE 1 |.

ILC2s exhibit functional plasticity and can adopt both ILC1- and ILC3-like phenotypes. Under the influence of IL-12 and IL-1β, ILC2s can adopt ILC1-like phenotypes, characterized by expression of T-BET and production of IFN-γ. ILC1-like ILC2s can also be reverted to a traditional ILC2 phenotype under the influence of eosinophil-derived IL-4. Furthermore, in the presence of IL-23 and TGF-β, ILC2s possess the ability to adopt ILC3-like phenotypes, characterized by the expression of RORγt alongside production of IL-17. Vitamin D3 has been shown to inhibit the adoption of ILC3-like phenotypes by ILC2s.

3.1 |. ILC2↔ ILC1

Under the influence of IL-12 and IL-1β, ILC2 cells have been shown to exhibit an ILC1-like phenotype (Figure 1) [28, 29]. Notably, these cells produce IFNγ and express the transcription factor T-BET (Figure 1). Notch signaling is thought to play a role in potentiating this plasticity [28], as is the interleukin 12 receptor (IL-12Rβ1) since patients deficient in this particular receptor are found to have reduced ILC1 activity coupled with exaggerated ILC2 activity [28]. IL-1β influences the chromatin landscape in ILC2s, making the IFNγ gene locus accessible simultaneously with the IL-5 and IL-13 gene loci [29].

Eosinophil-derived IL-4 has been shown to play an important role in regulating the plasticity of ILC2s by pushing IFNγ+ ILC2s back towards a classical ILC2 phenotype (Figure 1) [30]. This suggests an important relationship between classical ILC2s, which produce IL-5, a cytokine critical to the process of eosinophilopoiesis, and eosinophils, which produce IL-4, a cytokine crucial for maintaining a balance in ILC2 phenotypes [30].

3.2 |. ILC2↔ ILC3

ILC2s can adopt an ILC3-like phenotype and produce IL-17 under the influence of IL-23 and TGFβ (Figure 1) [31]. TGFβ furthermore suppresses GATA3 expression in these cells, reducing their classical ILC2 effector functions. Interestingly, vitamin D3 has been shown to prevent the adoption of ILC3-like phenotypes by ILC2s and may offer a novel way to alleviate IL-17-mediated neutrophilic airway inflammation (Figure 1) [31].

Another significant way through which ILC2 cells may adopt ILC3-like phenotypes is through migration-associated reprogramming whereby gut-native ILC2s enter peripheral circulation and migrate to take up residency in the lung. Whilst lung-native ILC2s respond to airway alarmins in the classical fashion, gut-native ILC2s express RORγt and have the dual ability to take on either classical ILC2 or ILC3-like phenotypes [32]. In humans, this phenomenon is aided by the expression of c-kit [33]. Human c-Kit expression in ILC2s has further been demonstrated to be promoted under the combined influence of IL-1β and IL-18, which also promote the IL-17-producing ILC3-like phenotype in patients with severe asthma [34].

4 |. Pulmonary ILC2s Are Embedded in a Complex Niche Relying on a Range of Costimulatory Molecules

ILC2s act as first responders to airway damage, interacting with leukocytes and other lung cells. We, and others, have shown how key ligand–ligand interactions modulate ILC2 function and metabolism, influencing airway hyperresponsiveness (AHR) and pulmonary eosinophilia in allergic airways disease (AAD). These costimulatory molecules can be subdivided into three subcategories: the tumor necrosis factor receptor superfamily (TNFRSF), immunomodulatory costimulatory receptors, and molecules with immunoreceptor tyrosine-based inhibitory motif (ITIM) signaling.

4.1 |. The Tumour Necrosis Factor Receptor Superfamily Is Widely Represented in ILC2 Surface Receptors

Receptors belonging to the tumour necrosis factor receptor superfamily (TNFRSF) are widely expressed on mammalian leukocytes, including ILC2s (Figure 2). These receptors commonly form trimeric complexes on the plasma membrane of ILC2s and signal via adaptor proteins such as tumour necrosis factor receptor-associated factor (TRAF), ribosome-inactivating protein (RIP), and Fas-associated protein with death domain (FADD) [35].

FIGURE 2 |.

Pulmonary ILC2s are embedded in a complex niche relying on a range of costimulatory molecules. Engagement of cell-surface receptors on ILC2s can influence their activity and crucially modulate their production of effector cytokines including IL-4, IL-5, IL-6, IL-9, IL-13, and GM-CSF. Such modulatory cell-surface receptors can be divided into three families: the TNF Receptor Superfamily (which includes RANK, TNFR2, GITR, DR3, and OX40L), receptors with ITIM motifs (such as PD-1, SIRPa, and LAIR-1), and costimulatory molecules (ICOS, ICOSL, CD200R, KLRG1, LFA-1, and ICAM-1) which rely on social interactions between ILC2s and other cell types such as epithelial cells, endothelial cells, and T-Helper cells.

We have shown that ILC2s isolated from visceral adipose tissue express the TNF receptor GITR, and that activating GITR improves ILC2 glucose homeostasis, suggesting therapeutic potential for type 2 diabetes [36]. GITR engagement in other contexts may however be deleterious, as we and others have shown that stimulating this pathway potentiates effector ILC2 function, increasing the production and release of type-2 inflammatory cytokines, particularly implicated in the pathogenesis of allergic airways disease, namely IL-5 and IL-13 (Figure 2) [36, 37].

Death receptor 3 (DR3) is another member of the TNF receptor superfamily expressed on the plasma membrane of ILC2s. TNF-like ligand 1A (TL1A), the DR3 ligand, is released by epithelial cells, dendritic cells, and macrophages after exposure to inflammatory stimuli [38, 39]. In cooperation with IL-33, TL1A stimulates pulmonary IL-2 expression and promotes IL-9 production by ILC2s, whilst also prolonging the maintenance of their activation state in both the murine and human lung [40].

DR3 engagement has been shown to be a critical factor regulating the expansion of both human and murine ILC2 populations [41, 42]. Moreover, ILC2s upregulate the expression of DR3 in response to epithelium-derived cytokines such as IL-33 and TSLP [39], and the expression of this receptor is not significantly affected by glucocorticoid treatments, suggesting that in steroid-resistant asthma, this key activation pathway may continue to drive asthma pathology [39]. Indeed, blocking DR3 signaling in mouse models of AAD and inflammatory bowel disease has been shown to lead to suppression of the inflammatory state, particularly through reducing cytokines such as IL-5 and IL-13 (Figure 2) [43].

The receptor activator of the nuclear factor-κB (RANK) is a further member of the superfamily of TNF receptors found to be expressed by ILC2 cells (Figure 2). Engagement of RANK by its corresponding ligand, RANKL, expressed on the cell membrane of CD4+ T cells as well as antigen presenting cells, has been shown to activate ILC2s via the NF-kB pathway [44]. Similarly, blocking of the RANK/RANKL signaling pathway with a competitive inhibitor, osteoprotegerin (OPG), has been shown to ameliorate disease states in the OVA model of AAD, particularly through a reduction in IL-4 producing CD4+ T cells (Figure 2) [45].

OX40L is a member of the TNFRSF that exhibits inducible expression on ILC2s in response to IL-33 (Figure 2) [46, 47]. OX40L on ILC2s binds to OX40 expressed by TH2 cells with the effect of promoting their survival and proliferation within the lung, leading to exacerbation of allergic inflammatory responses [46]. OX40L interactions can, however, promote regulatory immune responses through binding OX40 expressed on T-regs, although this seems to be of less importance compared to the interactions that occur between ILC2 OX40L and OX40 expressed by T-effector cells [46]. Notably, studies using ILC2-specific OX40L-deficient mice (Il7ra Cre/+Tnfsf4 fl/fl) demonstrated complete failure to mount adaptive type 2 immune responses, confirming the non-redundant role of ILC2-expressed OX40L in driving proinflammatory responses in the setting of allergic lung inflammation [46].

Tumour necrosis factor receptor 2 (TNFR2) is selectively expressed on ILC2s and, when bound by its ligand TNF-α in the lungs, promotes signaling pathways that significantly influence ILC2s by enhancing their survival, cytokine production, and functional roles in various disease contexts, including AAD (Figure 2) [48]. Our findings notably demonstrate that blocking the receptor or pharmacological inhibition of NF-κB-inducing kinase (NIK)-mediated receptor signaling both reduce the development of AHR in response to IL-33 or allergens in acute asthma models by reducing GATA-3 and p52 expression, which in turn results in lower production of inflammatory cytokines such as IL-5 and IL-13 [48].

4.2 |. ILC2s Are Collaborative and Crosstalk With a Range of Different Cell Types

In addition to regulation via the TNFRSF, ILC2s are highly collaborative and express many receptors that directly crosstalk with other cells within the ILC2 niche (Figure 2). Receptors like ICOS and ICOSL foster interaction between ILC2s and CD4+ T-cells. ILC2 expression of CD200R and KLRG1 emphasizes the importance of airway epithelial cells in influencing ILC2 activity. Trafficking ligands such as LFA-1 and ICAM-1 offer pathways through which ILC2s can interact with endothelial cells as well as other migratory leukocytes.

Inducible T-cell Co-stimulator (ICOS) is a receptor expressed on murine and human ILC2s, as well as T-cells, notably T-reg cells (Figure 2) [49, 50]. ICOS is not constitutively expressed on either ILC2s or T-cells but rather is rapidly induced in response to TCR (T-cells) and IL-33 (ILC2s) activation [51, 52]. The ligand for ICOS, ICOS-L, is also found on ILC2s, opening up the possibility of cis-activation via ICOS:ICOS-L interactions on the ILC2 cell membrane, as well as trans-activation interactions between ILC2s, APCs, and T-regs. We have described how ICOS:ICOS-L stimulation can promote pro-inflammatory ILC2 phenotypes by increasing cell survival and production of cytokines such as IL-5 and IL-13 [49, 53, 54].

CD200 is an immunomodulatory protein abundantly expressed by healthy airway epithelial cells [55]. It has been suggested that airway epithelial CD200 expression is of evolutionary advantage to prevent unnecessary activation of immune cells in the absence of inflammatory stimuli [56]. CD200 binds to CD200R, a receptor found on macrophages, dendritic cells, and indeed ILC2s (Figure 2) [57]. We have notably shown how CD200R engagement ameliorates AAD in mouse models by attenuating AHR and reducing the production of type-2 cytokines by ILC2s [57].

Killer cell lectin-like receptor G1 (KLRG1) is an immune checkpoint receptor expressed on many leukocytes including ILC2s (Figure 2). Cadherins (E-, N-, and R-) in the airway epithelium are key ligands for KLRG1 [58], triggering a signalling pathway that induces IL-10 production in ILC2s, shifting them to an anti-inflammatory phenotype [59–62]. Interestingly, soluble e-cadherin levels in human sputum samples have been linked to asthma severity, suggesting that patients with more severe asthma exhibit increased airway epithelial barrier disruption leading to “shedding” of e-cadherin, a process that would prevent the healthy interaction between these molecules and KLRG1 [63].

We also investigated the effects of endothelial trafficking markers on ILC2 accumulation and activity in the lung. The transendothelial migration cascade involves key cell–cell interactions, including crawling, mediated by binding to intercellular adhesion molecule 1 (ICAM-1) to lymphocyte function-associated antigen 1 (LFA-1) (Figure 2) [64]. We found that both LFA-1 and ICAM-1 knockout mice exhibit attenuated AHR in response to IL-33 challenge [65]. We demonstrated that LFA-1 deficient ILC2s do not effectively migrate to the inflamed lung; although the lack of LFA-1 does not interfere with ILC2 homeostasis, cytokine production, and metabolism. Conversely, ICAM-1 deficient ILC2s still home to the inflamed lung; however, this knockout was also associated with a downregulation of pro-inflammatory cytokines [66].

4.3 |. Receptors With ITIM Motifs Can Modulate ILC2 Function; However, Concrete Evidence of ITAM Signalling Remains Lacking

Cell surface receptors with immunoreceptor tyrosine-based activation motifs (ITAMs) and immunoreceptor tyrosine-based inhibition motifs (ITIMs) form two distinct families of costimulatory molecules. Whilst we still have not found evidence of ITAM signalling in ILC2s, a number of ITIM-containing receptors have been identified in both murine and human ILC2s in the naïve and activated state. ITIMs have a consensus sequence of amino acids: Serine, Isoleucine, Valine, Leucine, x, Tyrosine, xx Isoleucine, Valine. These residues are hydrophobic and are located in the cytoplasmic tails of inhibitory receptors. Upon phosphorylation of their tyrosine residues, ITIMs recruit phosphatases like Small Heterodimer Partner (SHP)-1, SHP-2, or SH2-domain-containing inositol polyphosphate 5-phosphatase 1 (SHIP1). These phosphatases dephosphorylate signalling molecules, thereby dampening ILC2 activity [67–69].

Programmed cell death protein 1 (PD-1), an immune checkpoint receptor which we have previously shown is present on both naïve and activated murine and human ILC2s, is upregulated once these ILC2s are activated, particularly by IL-33 (Figure 2) [70]. Binding of PD-1 by its known ligands PD-L1 and PD-L2 has been shown to lead to T-cell inhibition and exhaustion [71, 72]. Likewise, we have shown the PD-1 interactions are linked to restraining ILC2 activity and expansion. We demonstrate that on a mechanistic level, this phenotype is likely due to metabolic reprogramming, whereby ILC2 metabolism is shifted from primarily being reliant on fatty acid oxidation and glutamine catabolism to preferentially scaling up reliance on glycolysis [70]. Moreover, we find that PD-1 deficient mice exhibit increased levels of AHR, greater levels of pulmonary eosinophilia, and increased pulmonary populations of ILC2 cells, indicating that engagement of the PD-1 receptor could present itself as a novel therapeutic opportunity in the treatment of AAD.

Recently we outlined the importance of signal regulatory protein-α (SIRPα) in regulating the function of ILC2s (Figure 2). SIRPα is a transmembrane receptor previously identified on macrophages and dendritic cells, which binds to CD47, initiating a signaling cascade culminating in the abortion of pro-phagocytic pathways [73, 74]. We identified the expression of SIRPα on both murine and human ILC2s in a naïve state and demonstrated its upregulation following IL-33 exposure in both cases [75]. We further showed the potential benefits of targeting SIRPα in AAD [75]. The SIRPα-CD47 axis modulates mitochondrial metabolism through the JAK/STAT and ERK/MAPK signaling pathways, regulating NF-κB activity and the production of type 2 cytokines. We further show how targeting this pathway reduces AHR and pulmonary inflammation in mouse models of AAD [75].

We have previously explored the role of the immune checkpoint receptor LAIR-1 in regulating ILC2 induced AHR. LAIR-1 is a transmembrane glycoprotein in the immunoglobulin family which, when stimulated, has an inhibitory function on a range of leukocytes (Figure 2) [76]. LAIR-1 is known to have two important endogenous ligands: C1q (a complement protein) and collagens, particularly collagen-I and collagen-III [77, 78]. IL-33 stimulation has been shown to increase LAIR-1 expression on murine ILC2 cells in vivo and both murine and human ILC2 cells in vitro, contributing to the suppression of their pro-inflammatory phenotypes [67]. We notably show that LAIR-1 signalling in ILC2s is mediated via inhibitory pathways, including SHP1/PI3K/AKT [67].

Despite the identification of ITIM-containing receptor signalling pathways that are able to modulate ILC2 activity, the identification of ITAM-containing receptors currently lacks concrete evidence. The receptor NKp30, found on ILC2s, has, for example, been observed to play a role in inducing type-2 cytokine secretion in human atopic dermatitis lesions. However, the mechanism of action in ILC2s remains unclear since NKp30 traditionally utilizes ITAM-containing adapter proteins such as CD3ζ, a protein not expressed in ILC2s per se.

5 |. The ILC2 Niche Is Highly Innervated and Receives Multiple Forms of Neuroendocrine Stimuli

The lung receives three forms of innervation: sympathetic, parasympathetic, and sensory (Figure 3) [79, 80]. Sympathetic innervating fibres use the neurotransmitter noradrenaline to signal to airway smooth muscle, pulmonary blood vessels, and the airway epithelium, typically resulting in bronchodilation, pulmonary blood vessel constriction, and decreased airway epithelial mucus secretion [81]. Parasympathetic innervation arises from the vagal nerve and relies on the neurotransmitter acetylcholine. Acetylcholine signaling results in bronchoconstriction, blood vessel dilation, and increased mucus secretion [81]. Sensory neurones innervate the lung from either the thoracic dorsal root ganglia or the vagal ganglia [82–84]. Sensory neurones have an incredibly diverse role in the lung and release many neuropeptides such as CGRP, SP, NMU, and VIP, which are important signals for modulating immune cell function. In addition to the input of sympathetic, parasympathetic, and sensory nervous systems, the ILC2 niche is further influenced by club cells and pulmonary neuroendocrine cells (PNECs) in the epithelial barrier, which are important sources of neuroendocrine stimuli (Figure 3).

FIGURE 3 |.

The ILC2 niche is highly innervated and receives multiple forms of neuroendocrine stimuli. ILC2 cells respond to neuroendocrine stimuli from the sympathetic, parasympathetic, and sensory nervous systems in addition to airway epithelium-derived neuroendocrine signals. Parasympathetic inputs influence ILC2s via the a7nAChR receptor. Sensory nervous signals such as VIP, NMU, and SP interact with their respective receptors (VPAC2, NMUR1, NK1) to further modulate ILC2 activity. Additionally, PNEC and club cell-derived molecules such as ADM, CGRP, and serotonin play a further important role in modulating the activity of ILC2s within their niche by interacting with the receptors RAMP2, RAMP1, and HTR2A, respectively.

5.1 |. Input From the Parasympathetic Nervous System Ameliorates ILC2 Driven Allergic Airways Disease

The parasympathetic nervous system relies on signaling via the binding of acetylcholine to muscarinic and nicotinic acetylcholine receptors. ILC2s express the α7 nicotinic acetylcholine receptor (α7nAchR) which can be stimulated by both acetylcholine as well as nicotine (Figure 3). We found that exposure to alarmins such as IL-25 and IL-33 upregulates expression of the α7nAchR in ILC2s [85]. Moreover, engagement of this receptor with an agonist has been shown to suppress ILC2 cytokine production and ameliorate disease states in mouse models of AAD, suggesting it may be a potential novel therapeutic target in the treatment of allergic asthma [85].

5.2 |. The Sensory Nervous System Influences ILC2s Through Various Mechanisms

Two main sub-types of sensory neurone have been identified in the lung: Na_v1.8+ neurones and TRPV1+ neurones [86]. TRPV1+ neurones have been shown to play a crucial role in regulating ILC2-mediated inflammation in the lung of allergen-exposed mice, as studies show that by blocking TRPV1+ neurones with resiniferatoxin, ILC2 and eosinophil cell populations expand in the inflamed lung alongside increased mucus production in the small airways [87].

Substance P (SP) is produced primarily by sensory neurones around the airway epithelium and airway smooth muscle, although leukocytes are also known to produce small quantities [88]. Interestingly, SP production is not uniform throughout the respiratory tree but rather seems to concentrate on specific branchpoint regions [89], the same regions where ILC2 cells are commonly found. The effect of SP on ILC2 cells directly is as of yet unknown; however, SP signalling is observed to precipitate aggravated pulmonary inflammation with increased BALF levels of ILC2-related cytokines such as IL-5 and IL-13. Indeed, ILC2 cells do express NK1, the highest affinity SP receptor, and so it would be reasonable to hypothesise that SP has a direct stimulatory effect on these cells (Figure 3).

Neuromedin U (NMU) is a 25 amino acid neuropeptide found in multiple organ systems including the lung. NMU is produced in the lung from sensory neurones originating from the thoracic dorsal root ganglia, but not from those originating from vagal ganglia [90]. ILC2 cells are known to express NMUR1, the receptor for NMU, and activation of this receptor activates ILC2s, increasing their secretion of cytokines such as IL-5 [90–92]. In human subjects, NMUR1 expression in ILC2s has been found to be greater in mild asthma compared to severe asthma, and has also been shown to be inversely correlated with inhaled corticosteroid dose [93]. IL-13 has further been shown to be a powerful signal for TPRV1+ neurones, increasing their expression of the Nmu gene and therefore producing a positive feedback loop between these neurones and activated ILC2 cells driving the pathology of AAD (Figure 3) [94].

Both Na_v1.8+ and TRPV1+ neurones, part of the nodose ganglion of the vagus nerve, can release vasoactive intestinal peptide (VIP) [86]. VIP is a 28-amino acid neuropeptide belonging to the glucagon/secretin family [95, 96]. VIP can directly modulate ILC2 function via the VPAC2 receptor (Figure 3). VIP-VPAC2 signalling has been linked with increasing effector cytokine production by ILC2 cells [97–99]. Nodose ganglion neurones, specifically Nav1.8+ neurones, have been shown to express IL-5 receptor and demonstrate the ability to be activated by IL-5, suggesting a potential positive feedback loop between these VIP-secreting neurones and IL-5-secreting ILC2s [94]. Interestingly, it has also been shown that immune cells such as TH2 cells produce VIP endogenously and that this production is upregulated following allergen exposure [100].

5.3 |. Epithelium-Derived Neuroendocrine Immune Modulatory Signals Provide Further Regulatory Cues to ILC2s in Their Microenvironmental Niche

The airway epithelium is a key source of neuropeptides that modulate ILC2 function. Adrenomedullin (ADM), a 52 amino acid neuropeptide initially isolated from adrenal pheochromocytoma cells, is also found in the lung and binds two receptors: Receptor Activity-Modifying Protein 2 (RAMP2) and Calcitonin Receptor-Like Receptor (CRLR) (Figure 3). Club cells in the airway epithelium are an important source of ADM, particularly in the context of AAD [101]. Emerging evidence highlights club cells as key coordinators of airway epithelial responses to pathogens, allergens, and inflammatory stimuli [28, 29]. Club cells produce ADM in a hypoxia-dependent manner, reliant on the function of transcription factor HIF-2α [101]. In vivo ADM administration has been shown to potentiate ILC2 activity, in particular by increasing IL-5 production by ILC2 cells, even in the absence of alarmins such as IL33 [101].

ILC2 cells are most frequently located within 70μm of airway branchpoints, bringing them into close association with pulmonary neuroendocrine cells (PNECs) [102]. PNECs comprise roughly 0.5% of airway epithelial cells, but despite this relative rarity, they are emerging as crucial cell types involved in immune signaling with the potential to modulate immune responses to a range of inflammatory stimuli. PNECs produce multiple signals that interact with ILC2s, including serotonin and calcitonin gene-related peptide 1 (CGRP-1) [103]. Serotonin, produced from the metabolism of tryptophan, suppresses ILC2 activation and proliferation through binding 5-hydroxytryptamine receptor 2A (HTR2A) (Figure 3) [104, 105]. CGRP-1 interacts with ILC2s via the RAMP1 receptor [88], initiating a pro-inflammatory signaling cascade, resulting in increased IL-5 secretion (Figure 3) [102]. The role of CGRP-1 in driving the pathology of asthma is, however, not yet fully understood, as despite it being found to increase ILC2 activation, some studies associate it with attenuated inflammatory responses, reducing AHR and eosinophilia [106, 107].

6 |. ILC2s Rely on Diet-Derived Nutrient and Micro-Nutrient Cues to Regulate Their Metabolism and Phenotype

Nutrients serve as fundamental substrates for immune cell metabolism, directly influencing cellular energy production, biosynthetic pathways, and effector functions. Glucose, amino acids (e.g., glutamine), fatty acids, and micronutrients such as iron or zinc govern immune cell activation by modulating metabolic checkpoints like glycolysis, oxidative phosphorylation, and lipid synthesis. Diet-derived nutrients, including vitamins and microbiota-generated short-chain fatty acids (SCFAs), critically shape ILC2 function, while the regulatory roles of essential elements such as calcium, zinc, and iron in ILC2 biology are increasingly recognized.

Short chain fatty acids such as butyrate have been identified as important regulatory molecules for ILC2s (Figure 4). SCFAs are primarily produced by commensal bacteria in the gut, largely of the Firmicute phylum, which digest dietary fiber. We have previously shown that dietary fiber supplementation, particularly with pectin, reduces susceptibility of mice to AHR by suppressing ILC2 activity [108]. This is largely due to the expansion of Firmicutes in the gut microbiome which break down pectin to produce butyrate. Others have also found that butyrate supplementation reduces AHR and lowers the expression of pro-inflammatory genes such as IL-5, IL-13, and Gata3, likely through epigenetic modulation via HDAC (Figure 4) [109].

FIGURE 4 |.

ILC2s rely on diet-derived nutrient and micro-nutrient cues to regulate their metabolism and phenotype. Various diet-derived nutrients and micro-nutrients have been identified as important regulators of ILC2 immunometabolic pathways. Short chain fatty acids, such as butyrate, suppress ILC2 effector activity by reducing the expression of Gata3, il-5, and il-13. Calcium also suppresses ILC2 activity by repressing the NF-kB pathway via increasing KLF2 expression. Contrastingly, Zinc has been identified as an essential cofactor for the transcription of pro-inflammatory genes in ILC2s, where zinc supplementation precipitates an increase in MAPK and NF-kB activity. Changes to intracellular iron concentrations have also been shown to alter the metabolism of ILC2s, particularly via the HIF1a pathway, whereby increased iron availability has been associated with reduced HIF1a signaling, and by extension, a reduction in pro-inflammatory phenotypes. Retinoic acid, a derivative of retinol (Vitamin A), has been shown to increase the production of the anti-inflammatory cytokine IL-10 by ILC2s.

Retinoic acid (RA), an essential form of vitamin A, has likewise been shown to contribute to regulating ILC2-mediated inflammation (Figure 4). RA supplementation has been associated with inducing phenotypic shifts in ILC2s whereby they take on a regulatory cell phenotype, producing greater quantities of IL-10 [110].

Calcium signaling exerts dual regulatory control over ILC2 activity, balancing proinflammatory and suppressive pathways (Figure 4). We first showed that Orai1/2-mediated Ca²⁺ entry drives ILC2 effector functions by sustaining mitochondrial fatty acid oxidation and oxidative phosphorylation [111]. Conversely, mechanosensitive channels Piezo1-dependent Ca²⁺ influx restrains ILC2 activation by enhancing KLF2 transcription factor activity, which suppresses NF-κB signaling, limiting IL-5/IL-13 secretion [112]. This dichotomy highlights the context-dependent role of calcium: while Orai1/2 sustain ILC2-driven inflammation, Piezo1 activation counterbalances these effects. Zinc (Zn²⁺) is an essential trace element found to have myriad functions in regulating the activity of intracellular enzymes and bolstering the stability of many structural proteins [113]. Epidemiologically, zinc deficiency has been associated with an increased prevalence of childhood asthma and atopic dermatitis, two allergic inflammatory diseases in which ILC2s are implicated [114, 115]. Curiously, in ILC2s, it has recently been shown that Zn²⁺ plays a stimulatory role, driving pro-inflammatory phenotypes via MAPK and NF-kB pathways (Figure 4) [116]. While zinc chelation notably suppressed IL-5 expression in human ILC2s, Zn²⁺ supplementation was remarkably able to enhance IL-5 expression ex vivo. The conflicting evidence between the association of zinc deficiency with childhood asthma and the stimulatory role of Zn²⁺ warrants further investigation. Indeed, zinc is a critical co-factor for a myriad of immune system processes, particularly in the context of mucosal immunity and the maintenance of epithelial barriers via tight junction proteins such as ZO.1 and Occludin [117, 118].

We have recently shown that iron influences ILC2 activity and metabolism (Figure 4) [119]. Activated murine and human ILC2s have an increased demand for iron, corresponding with increased expression of the transporter TfR1. Using iron chelators, we were able to show that iron deprivation reduces cellular OXPHOS, inducing a HIF1a-driven anaerobic glycolysis. In murine models of acute AHR, iron deprivation reduced ILC2 function and ameliorated AHR. Notably, human ILC2s from patients with asthma exhibit elevated TfR1 expression, correlating with disease severity, suggesting iron-dependent activation pathways are conserved across species.

7 |. Concluding Remarks

Over the past 15 years, our understanding of ILC2s has expanded far beyond their initial characterization as IL-5 and IL-13-producing non-T, non-B cells. Recent advances in immunology, metabolism, and neuroendocrine regulation have shed light on the complex network of signals governing ILC2 function. These discoveries offer new insights into their role in AAD and broader immune regulation.

The identification of key costimulatory and inhibitory receptors, such as PD-1, GITR, ICOS, and DR3, has highlighted the intricate balance of signals required to modulate ILC2-driven inflammation. Furthermore, emerging evidence suggests that neuropeptides, including VIP, NMU, and CGRP, exert direct and indirect control over ILC2 activity, positioning the pulmonary nervous system as a crucial regulator of allergic inflammation. The influence of diet-derived metabolites and micronutrients on ILC2 metabolism, particularly the roles of short-chain fatty acids, retinoic acid, and essential metals like iron and zinc, also underscores the importance of systemic factors in shaping immune responses.

Despite these advances, many unanswered questions remain. The precise mechanisms by which neuroendocrine signals and dietary components shape ILC2 phenotypes need further elucidation, as do the interactions between ILC2s and other innate and adaptive immune cells. Future research should focus on leveraging these findings to develop targeted therapies that selectively modulate ILC2 function while minimizing off-target effects. Given the global burden of allergic asthma and other type-2 inflammatory diseases, understanding and therapeutically targeting ILC2s could offer novel strategies to improve patient outcomes.

As research continues to unravel the complexities of ILC2 biology, the challenge now lies in translating these mechanistic insights into clinically effective interventions that harness ILC2 regulation for therapeutic benefit.

Abbreviations:

a7nAchR

alpha 7 nicotinic acetylcholine receptor

AAD

allergic airways disease

ADM

adrenomedulin

AHR

airway hyperresponsiveness

CD

cluster of differentiation

CGRP

calcitonin gene related peptide

CRLR

calcitonin receptor-like receptor

CRTH2

chemoattractant receptor-homologous molecule expressed on Th2 cells

DR3

death receptor 3

ERK

extracellular signal-regulated kinase

FADD

Fas-associated protein with death domain

GATA3

GATA binding protein 3

GITR

glucocorticoid-induced TNFR-related protein

HIF-2a

hypoxia inducible factor 2 alpha

HTR2A

5-hydroxytryptamine receptor 2A

ICAM-1

intercellular adhesion molecule 1

ICOS

inducible T-cell co-stimulator

ICOSL

inducible T-cell co-stimulator ligand

IFNy

interferon gamma

IL-

interleukin-

ILC

group 2 innate lymphoid cell

ITIM

immunoreceptor tyrosine-based inhibitory motif

JAK

Janus kinases; KLRG1, killer cell lectin-like receptor G1

LAIR-1

leukocyte-associated immunoglobulin-like receptor 1

LFA-1

lymphocyte function-associated antigen 1

MAPK

mitogen-activated protein kinase

NaV 1.8

voltage-gated sodium channel 1.8

NMU

neuromedin-U

NMUR-1

neuromedin U receptor 1

OX40L

tumour necrosis factor ligand superfamily member 4

PNEC

pulmonary neuroendocrine cell

RAMP2

receptor activity-modifying protein 2

RANK

receptor activator of the nuclear factor-κB; RIP, ribosome-inactivating protein

SHIP

SH2-domain-containing inositol polyphosphate 5-phosphatase

SHP

small heterodimer partner

SP

substance P

ST2

suppressor of tumourigenicity 2/IL1RL1

T-BET

T-box transcription factor 21

TH

T-helper

TL1A

TNF-like ligand 1A

TNFR

tumour necrosis factor receptor

TRAF

tumour necrosis factor receptor-associated factors

TRPV

transient receptor potential vanilloid

TSLP

thymic stromal lymphopoietin

VIP

vasoactive intestinal peptide

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.