Epithelial Sensing in Allergic Disease

Abstract

Epithelial cells provide a first line of immune defense by maintaining barrier function, orchestrating mucociliary clearance, secreting anti-microbial molecules, and generating sentinel signals to both activate innate immune cells and shape adaptive immunity. Although epithelial alarmins play a particularly important role in the initiation of type 2 inflammation in response to allergens, the mechanisms by which epithelial cells sense the environment and regulate the generation and release of alarmins has been poorly understood. Recent studies have identified new sensors and signaling pathways used by barrier epithelial cells to elicit type 2 inflammation, including a novel pathway for the release of interleukin-33 (IL-33) from the nucleus that depends on apoptotic signaling. These recent findings have implications in the development of allergic diseases from atopic eczema to food allergy, rhinitis, and asthma.

Introduction

Allergic diseases share core immunopathologic features that include the expansion of tissue-resident mast cells and type 2 innate lymphoid cells (ILC2s), the generation of IgE antibody, and tissue infiltration with eosinophils, basophils, and CD4⁺ Th2 cells. Despite these canonical features, the factors that initiate type 2 immunity in response to diverse environmental allergens remain poorly understood. In contrast to the well-defined role of pattern recognition receptors (PRRs) in recognizing conserved pathogen-associated molecular patterns (PAMPS) for the induction of type 1 or type 17 immunity, relatively few PRRs have been identified that recognize allergens directly and pattern type 2 immunity. Rather, allergens (or helminths or ticks) that elicit type 2 immunity are commonly associated with tissue damage or cell stress and the release of a repertoire of host molecules – so called ‘damage-associated molecular patterns’ (DAMPs) – that potentiate type 2 inflammation.

As a first point of contact with the environment, epithelial cells are well situated to sense allergens, endure damage and cell stress, and release alarmins. Epithelial DAMPs include nucleosides and nucleotides such as adenosine and adenosine triphosphate (ATP); nuclear proteins such as high mobility group box 1; S100 family members, reactive oxygen species (ROS), and even the cytokine interleukin-33 (IL-33). Additionally, epithelial cells can generate and secrete effector cytokines such as thymic stromal lymphopoietin (TSLP), IL-25, and granulocyte-macrophage colony-stimulating factor (GM-CSF) that pattern type 2 immunity in response to external stimuli. In this article, we review recent developments in our understanding of epithelial sensing of allergens and allergen-elicited cell stress and damage, and how that sensing contributes to the development and persistence of allergic disease.

Sensing pathogen-associated molecular patterns through classical PRRs

While PRRs typically recognize pathogen-associated patterns, some classical PRRs also recognize allergens or allergen-associated molecules to promote allergic inflammation (Fig. 1). For example, toll-like receptor 4 (TLR4) signaling in professional antigen-presenting cells is associated with the LPS-elicited generation of IL-12 and type 1 immunity, but TLR4 also binds house dust mite (HDM) and its activation in epithelial cells leads to the generation of key epithelial cytokines including TSLP, GM-CSF, IL-25, and IL-33 [1]. HDM potentiates TLR signaling because of structural homology between MD-2 (a lipid-binding component of the TLR4 signaling complex) and Der p2 [2], but other allergens have been shown to bind to MD-2 [3], suggesting a potential to elicit TLR4-mediated sensing.

Figure 1. Classical and non-classical pattern recognition receptors on epithelial cells that promote type 2 immunity.

House dust mite (HDM) activation of toll-like receptor 4 (TLR4) on epithelial cells leads to the upregulation of IL-33, IL-25, GM-CSF, and TSLP which promote type 2 inflammation. Additionally, HDM-associated muropeptides activate Nucleotide-binding oligomerization domain-containing protein 1 (NOD1) for IL-6 and IL-8 generation. In specialized tuft epithelial cells, microbial products such as succinate and a Trichinella spiralis (T.spiralis) product activate GPCRs such as SUCNR1 and bitter taste receptors, respectively, to elicit cysteinyl leukotriene (CysLT) generation and IL-25. Finally, bile acids derived from intestinal microbiome metabolism of inulin fibers bind the farnesoid X receptor (FXR), inducing IL-33 production. Created with BioRender.com.

Nucleotide-binding oligomerization domain- (NOD-)like receptors (NLRs) provide another mechanism for allergen sensing. NLRs are a class of intracellular PRRs that recognize PAMPs typically associated with gram positive bacteria. While early reports demonstrated that the best studied member of this family, NLRP3, was not relevant to allergic airway disease in mice [4], recent reports have identified a role for NOD1 in sensing HDM [5, 6]. Inhibition or deletion of NOD1 reduced HDM-elicited allergic airway inflammation in a murine model, and reduced airway epithelial generation of cytokines such as IL-6 and IL-8 through the receptor interacting serine/threonine kinase 2 (RIPK2) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway. Interestingly, NOD1 was activated by muropeptides in HDM extracts and its activation was independent of host microbiota, suggesting that epithelial recognition of HDM-associated microbiota can drive allergic inflammation [5]. Taken together, these findings provide evidence that some sensors typically associated with type 1 immunity can induce type 2 immunity as well. Given these findings, what dictates whether type 1 or type 2 immunity is patterned after activation of TLR4 or NOD1? While these PRRs lead to distinct downstream signaling in epithelial cells that may be sufficient to preferentially induce type 2 immune responses, other considerations include the magnitude and chronicity of activation and coincident signaling from the extracellular milieu that likely play important contextual roles in patterning type 2 responses.

Sensing microbial products through non-classical PRRs

Some G protein-coupled receptors (GPCRs) also detect microbial products and lead to the induction of allergic inflammation. Tuft cells are rare chemosensory epithelial cells poised to generate both IL-25 and cysteinyl leukotrienes (CysLTs) that pattern type 2 immunity [7–9]. Tuft cells sense succinate, a metabolite produced by some bacteria and parasites, through their expression of succinate receptor 1 (SUCNR1). Activation of SUCNR1 leads to ILC2 activation and cytokine generation, eosinophilia, and downstream goblet cell hyperplasia [10]. Additionally, while sensing of succinate produced by gut microbiota elicits tuft cell and ILC2 activation, a murine model of the atopic march demonstrated that increased serum succinate is detected after mechanical disruption of the skin, suggesting that local release of this metabolite or generation of succinate by skin bacteria could have distal effects, potentially activating airway and intestinal tuft cells as well [11]. Notably, tuft cells are distinguished by their expression of another class of GPCRs, bitter taste receptors, that are coupled to the Gα protein α-gustducin. Activation of taste receptors by a secretory product from the helminth Trichinella spiralis promotes the upregulation of IL-25 expression [12]. Taste receptors can also be activated by bacterial lactones, quorum-sensing molecules produced by Gram negative bacteria, but whether these ligands (or allergens) can elicit IL-25 release and downstream type 2 immunity through taste receptors is unknown.

Beyond succinate, bacteria also metabolize ingested fiber to produce bile acids that influence type 2 immunity. Specifically, feeding mice a diet high in inulin fiber results in a metabolic shift towards production of cholic acid (CA) by gut microbiota [13]. CA signals through farnesoid X receptor (FXR) resulting in upregulation of IL-33 in colonic stromal and lung epithelial cells and eosinophilia and ILC2 activation in the intestine and lungs. Taken together, these studies delineate potential pathways that could mediate the observed association between microbial metabolites and human allergic disease and highlight the importance of further studies of microbial metabolism to inform therapeutic strategies.

Sensing DAMPS through diverse receptors

Epithelial cells also sense allergens through cellular damage and the extracellular release of DAMPS, such as adenosine triphosphate (ATP) (Fig. 2). ATP is recognized by several GPCRs including the purinergic P2Y2 receptor, expressed on airway epithelial cells [14, 15]. In the upper airway, P2Y2 is expressed by nasal tuft cells, and ATP or uridine triphosphate (UTP) stimulation of these cells induces production of CysLTs, proinflammatory lipid mediators that promote type 2 responses [15]. Allergens such as HDM and Alternaria trigger P2Y2-dependent CysLT generation and downstream type 2 inflammation, presumptively due to allergen-elicited tissue damage and ATP release [9]. In the lower airway, ATP signaling through P2Y2 mediates allergic airway inflammation in response to HDM exposure. Here, the specific deletion of P2Y2 on airway epithelial cells in the lung reduces IL-33 expression in response to HDM [16]. This deletion also reduces type 2 cytokine secretion from mediastinal lymph node cells (MLN cells), methacholine-driven bronchial hyperreactivity, and eosinophilia after HDM challenge [16].

Figure 2. DAMP receptors on epithelial cells that promote type 2 immunity.

Allergen-induced cellular damage can lead to the extracellular release of nucleotides including adenosine triphosphate (ATP), adenosine diphosphate (ADP), and uridine triphosphate (UTP), which bind purinergic G protein-coupled receptors on epithelial cells including P2Y2 and P2Y13. This binding induces production of cysteinyl leukotrienes (CysLTs), IL-33, IL-25, and high mobility group box 1 (HMGB1). ATP also binds P2X7, a ligand-gated ion channel that induces DNA release into the extracellular space upon activation. Additionally, HMGB1 as well as S100A molecules activate the receptor for advanced glycation endproducts (RAGE), leading to epithelial production of IL-33 and IL-25. Created with BioRender.com.

Two other purinergic receptors on epithelial cells recognize ATP with relevance to allergic inflammation. P2Y13 is expressed by epithelial cells in the lower airway and recognizes both ATP and adenosine diphosphate (ADP) but has a higher affinity for ADP. Recognition of either ADP or ATP through this receptor triggers IL-33 and High mobility group box 1 protein (HMGB1) export to the cytoplasm, as well as extracellular secretion of HMGB1 [17]. Furthermore, nuclear to cytoplasmic export and release of HMGB1 triggered by HDM, by cockroach, and by Alternaria is inhibited by deletion or antagonism of P2Y13. These approaches are also sufficient to reduce HDM-elicited lung inflammation. A very interesting subsequent paper identified a mechanism for release of epithelial nuclear contents that depends on P2X7, an ATP-gated calcium channel [18]. Treatment of airway epithelial cells with Alternaria induces fragmentation and extracellular release of DNA, which potentiates type 2 inflammation. DNA release depends upon ATP activation of P2X7 and the downstream activation of caspase-3. Ultimately Alternaria-elicited inflammation is reduced by DNA scavenging or by deletion of caspase-3, underscoring a role for apoptosis pathways in the induction of allergic inflammation.

Finally, the receptor for advanced glycation end products (RAGE) recognizes a variety of damage-associated molecules including HMGB1 and members of the S100 Calcium Binding Protein A1 (S100A) family to amplify innate type 2 signals at multiple levels. An early study demonstrated through bone marrow chimeras that RAGE expression on non-hematopoietic cells increases HDM- and Alternaria-elicited IL-33 transcript and protein levels and promotes ILC2 accumulation in the lung [19]. Subsequent studies showed that epithelial RAGE signaling augments the apoptotic release of IL-25 and IL-33 in response to S100A11 [20], and that RAGE cooperates with EGFR in a novel manner to form a complex with the oxidized form of IL-33 and elicit goblet cell metaplasia [21]. Lastly, the organic compound toluene diisocyanate (TDI) triggers upregulation of Histone Deacetylase 1 (HDAC1) in a human bronchial epithelial cell (HBEC) line in a RAGE-dependent manner and inhibition of RAGE or HDAC1 reduces IL-4, IL-5, and IL-13 generation from restimulated lymphocytes in a murine model of TDI-induced asthma [22]. Thus, RAGE promotes IL-33 generation, release, and signaling to promote type 2 inflammation in response to allergen exposure.

Sensing Enzymatic Activity Through the Ripoptosome, Protease-Activated Receptors, and Other Pathways

Beyond sensing DAMPs released by nearby damaged cells, epithelial cells can also respond directly to exogenous enzymatic activity. Recent studies on the pathways by which IL-33 is secreted have identified a role for cell-intrinsic stress signaling. Chen et al. demonstrated that IL-33 can translocate from the nucleus to the cytoplasm via stress granule formation that is induced with exposure to protease allergens (e.g. papain, HDM)[23]. These proteases also induce fragmentation of gasdermin D with release of its N-terminal domain (GSDMD-N) that forms transmembrane pores in the cell membrane, allowing for IL-33 to escape into the extracellular space. Gasdermin D cleavage is independent of caspase 1 and 11, but it depends on the cysteine protease activity of papain. Brusilovsky et al. found that many indoor allergens (e.g. HDM, Alternaria, Aspergillus, Cockroach, and Cat) activate the ripoptosome in a cysteine protease-independent, but caspase 8-dependent fashion. This apoptotic signaling leads to the downstream activation of caspase 3 and 7, which cleaves IL-33 into its functional form before release [24]. This mechanism again points towards a key role for apoptotic signaling in IL-33-dependent initiation of type 2 inflammation. Mechanisms for IL-33 generation, nuclear export, intracellular maturation, and release are presented in Fig. 3.

Figure 3. Pathways triggering the production, nuclear export, maturation, and secretion of IL-33.

Epithelial sensors promote IL-33 generation and release at distinct steps. 1) IL-33 Production. Signaling through the farnesoid X receptor (FXR), P2Y2, the receptor for advanced glycation endproducts (RAGE), and the aryl hydrocarbon receptor (AhR) upregulate transcription to increase nuclear IL-33. Epithelial ROS generation induces similar upregulation through an undefined signaling pathway. 2) IL-33 Export. Export of immature IL-33 is dependent on the production of stress granules, which carry immature IL-33 into the cytoplasm. Signaling through P2Y13 has been shown to induce this stress granule formation. 3) IL-33 Maturation. Immature IL-33 must be cleaved into its mature form by caspase-3 or caspase-7, which are in turn activated by caspase-8. This process is triggered in a cysteine protease-independent fashion by environmental allergens, such as house dust mite (HDM), Alternaria, Aspergillus, cat, and cockroach, through the assembly and activation of the ripoptosome. 4) IL-33 Secretion. Protease allergens (e.g. HDM and papain) cleave gasdermin D, allowing the N-terminal domain (GSDMD-N) to form pores, through which mature IL-33 can escape the cell. In the case of papain, this process was shown to be cysteine protease-dependent. Created with BioRender.com.

Epithelial cells also express Protease-Activated Receptors (PARs), a family of GPCRs that are activated by serine proteases associated with many common allergens. Notably, these receptors can also be activated by mast cell tryptase and by trypsin released from damaged epithelial cells after allergen exposure. The most important PAR in the context of allergic disease is PAR2, which has been previously implicated in sensing cockroach, HDM and the molds Penicillium and Alternaria (reviewed in [25]). PAR2 activation triggers release of IL-25 [26], TSLP [27], and granulocyte macrophage colony stimulating factor (GM-CSF) [28], among other mediators, in isolated human airway epithelial cells. While in vitro and in vivo studies have demonstrated conflicting reports regarding PAR2-dependent release of nuclear IL-33 [29–31], other endogenous proteolytic pathways have been identified for the cleavage and release of IL-33 from epithelial cells [23, 24], reviewed above. Thus, PAR2 signaling is a reproducible pathway through which diverse allergens elicit select epithelial alarmins.

Finally, many environmental, food, and venom allergens have additional enzymatic activities that are sensed by the epithelium. For example, the major honeybee venom allergens include phospholipase A2, hyaluronidase, venom acid phosphatase, and dipeptidylpeptidase IV with demonstrated or potential local cytotoxicity [32]. Here, phospholipases likely play a critical role in driving type 2 immunity. For example, the enzymatic activity of bee venom phospholipase A2 (bvPLA2) cleaves membrane lipids to produce lysophospholipids such as lysophosphatidylcholine that triggers IL-33 release and ST2-dependent inflammation [14]. This study also reported that phospholipase D shares the same activity for membrane degradation and initiation of type 2 inflammation. Finally, some allergens are not enzymes but may be sensed in an environment with robust enzymatic activity. Such is the case for galactose-alpha-1,3-galactose (alpha-gal) which is normally an innocuous antigen. However, at the time of a tick bite the antigen is introduced alongside high levels of phospholipase A2, among other components of tick saliva, and type 2 immunity to alpha-gal results [33]. Thus, epithelial cells have a variety of mechanisms for sensing enzymatic activity and this activity can be a critical switch to initiate type 2 immunity.

Sensing Oxidative Stress Through the Aryl Hydrocarbon Receptor and Undefined Mechanisms

Exposure to both allergens and environmental hazards also induces oxidative stress in epithelial cells which primes them towards a type 2 immune response. Diesel exhaust particles (DEPs) are another class of airborne irritants that have been implicated in worsening allergic airway inflammation through the induction of reactive oxygen species (ROS) (reviewed in [34]). DEPs have long been known to act as adjuvants, boosting antigen-specific IgE to innocuous protein antigens [35, 36], and eliciting oxidative stress in epithelial and immune cells [37, 38] DEPs are sensed by the aryl hydrocarbon receptor (AhR). In human airway epithelial cells this leads to ligand-bound AhR-AhR nuclear translocator (ARNT) complexes translocating to the nucleus and binding to the promoters for IL-33, IL-25, and TSLP [39]. This binding induces higher expression of each of these cytokines such that the expression of AhR is correlated with the expression of each alarmin in patients with severe asthma.

Other allergens and environmental irritants can also induce oxidative stress, although their direct receptors are unidentified. Pollen extracts including ragweed, plantain, oak tree, and timothy grass induce ROS in cultured airway epithelial cells [40]. In the case of ragweed, both eosinophilic airway inflammation and ragweed-specific IgE are reduced by oxidase inhibition. Epithelial ROS also induces epithelial cytokine production. For example, vanadium, a metal component of particulate matter, promotes lung eosinophilia and HDM-specific IgE in a mouse model of HDM-induced allergic airway inflammation through ROS-dependent epithelial production of TSLP and IL-25 [41]. Another study demonstrated that treatment of keratinocytes with HDM alone induces ROS production that elicits higher expression of IL-33, TSLP, and IL-25 and is reduced by anti-oxidant treatment [42]. Interestingly, several common household laundry detergents have also been reported to elicit ROS-dependent eosinophilic lung inflammation even in the absence of allergens [43]. Detergents induce the ROS-dependent upregulation of IL33 from HBECs in vitro and elicit IL-33 upregulation and eosinophilic lung inflammation in vivo that are reduced by N-acetyl-cysteine. As the concentrations of detergents needed to induce eosinophilic inflammation in this study were significantly higher than those identified in house dust, further studies are needed to understand whether exposure to these compounds in house dust can influence long-lived Th2 immunity to protein allergens.

Epithelial Sensing in the Context of Allergic Inflammation

IL-4 and IL-13 induce profound changes to barrier epithelial tissues. These include increasing the formation of antigen passages; increasing goblet cell, basal cell, and tuft cell frequency; and inducing epigenetic alterations, all of which are likely to alter epithelial sensing. For example, IL-13 activates a PI3K-CD83-cADPR pathway in epithelial cells in the small intestine. This promotes formation of secretory antigen passages (SAPs) through which antigens are chaperoned from the lumen to immune cells in the lamina propria, such as professional APCs and mast cells, leading to anaphylaxis [44]. This stands in contrast to homeostatic goblet cell antigen passages which do not drive anaphylaxis but instead feed intestinal antigens to tolerogenic CD103⁺ dendritic cells to facilitate intestinal homeostasis [45]. Moreover, while few studies have evaluated the effects of IL-13-elicited tuft cell hyperplasia on environmental sensing, it seems reasonable to infer that expansion of this rare cell population is likely to potentiate recognition of succinate and bitter taste receptor ligands which may further exacerbate type 2 inflammation. Finally, type 2 cytokines like IL-4 and IL-13 induce profound epigenetic changes in the epithelium. A genome-wide study found that exposure of cultured airway epithelial cells to IL-13 induces widespread epigenetic changes over time that are largely conserved when compared to DNA methylation patterns in primary airway epithelial cells from asthmatic patients [46]. Similarly, in the upper airway, serially passaged basal epithelial cells from patients with chronic rhinosinusitis with nasal polyps (CRSwNP) demonstrate enhanced expression of IL-4 and IL-13-stimulated β-catenin target genes, as compared to CRSsNP, recapitulating the increased IL-4/13 signatures, epigenetic changes, and expression of Wnt/β-catenin target genes that distinguished primary basal epithelial cells of CRSwNP from controls [47]. However, further studies in this area are needed to understand the sequelae of these functional, cellular, and epigenetic changes on mucosal defense and to understand ‘feed forward’ mechanisms through which type 2 inflammation may be perpetuated.

Conclusion

Epithelial cells play key roles in integrating countless environmental signals at the mucosal interface. Remarkably, the diversity of signals that elicit type 2 inflammation seem to converge on the release of a relatively small set of conserved epithelial mediators generated in response to epithelial cell distress. While allergen-elicited release of nuclear IL-33, HMGB1, and DNA require activation of apoptotic pathways, the evolutionary advantage conferred by this additional signal is not clear given that these mediators and other alarmins elicit seemingly similar downstream pathways for type 2 inflammation (ILC2 activation, IL-13 and Th2 generation, tissue eosinophilia, etc). It is possible that this redundancy confers an advantage simply through signal amplification, generating an innate inflammatory response proportionate to the degree of tissue injury. By contrast, it is possible that research in the next few years will identify additional effector pathways (or regulatory pathways) uniquely activated by these nuclear signals. Finally, while many studies have assessed epithelial sensing in the context of short-term type 2 inflammation, the importance of individual pathways in producing durable epigenetic changes that result in type 2 epithelial memory remain to be determined.

Highlights:

• Allergens elicit epithelial generation and release of cytokines and cysteinyl leukotrienes leading to type 2 inflammation

• Epithelial cells sense allergens through many receptor families including: 1) classical and non-classical pattern recognition receptors, 2) receptors for damage-associated molecular patterns, 3) protease-activated receptors, and 4) sensors for intracellular enzymatic activity and oxidative stress.

• In contrast to the initiation of type 1 and type 17 immunity, there are not structurally conserved allergen-associated molecular patterns that initiate type 2 immunity. Rather, damage, cell stress, and apoptotic signaling are conserved host pathways leading to type 2 immunity.

Data Availability:

No data were used for the research described in this review article.