Type I immunity is induced in response to PAMPs and damage-associated molecular patterns (DAMPs) through pattern-recognition receptors (PRRs), which activate antimicrobial and pro-inflammatory responses to effect innate and acquired immunity. The major families of PRRs involved in viral sensing are the Toll-like receptors (TLRs), which are membrane proteins expressed on cell surfaces and in endosomes, and cytoplasmic PRRs. The latter include, most prominently, the RIG-like helicases (RLHs), the cyclic GMP-AMP synthase—Stimulator of Interferon Genes (cGAS-STING) pathway, and among the large family of NOD-like receptors (NLRs), NLRC2 (Evans et al., 2010; Motta et al., 2015; Ablasser and Chen, 2019). PAMPs that signal virus infections include double-stranded RNA (dsRNA), recognized by TLR3; single-stranded RNA (ssRNA), recognized by TLR7, TLR8, and NLRC2; and DNA, particularly unmethylated CpG motifs, recognized by TLR9. These signatures are also recognized by retinoic-acid inducible gene-I (RIG-I), coded by DDX58, and melanoma differentiation-associated 5 (MDA5), coded by IFIH1. RIG-I and MDA5 interact with mitochondrial antiviral signaling protein (MAVS) on the outer surface of mitochondria, causing MAVS to form aggregates with TNF receptor-associated factor 3 (TRAF3) and I-kappa-B kinase-epsilon/TANK-binding kinase 1 (IKKε/TBK1), leading to the activation of the transcription factors interferon regulatory factor (IRF) 3 and IRF7. These induce the innate anti-viral type I and type III IFNs, IFN-α and IFN-β, and IFN-λ1-4, respectively. STING is an ER-localized protein that acts as a powerful cytoplasmic DNA sensor by binding cyclic GMP-AMP generated by cGAS that directly binds DNA, and as a signaling adaptor for DEAD-Box Helicase 41 (DDX41) and other DNA sensors (Ablasser and Chen, 2019). It activates signal transducer and activator (STAT) 1/2 and IRF3 through TBK1 to strongly induce type I IFN production.

The principle cytokines capable of inducing Type 1 epithelial responses are IFNs released from epithelial cells in an autocrine and paracrine manner, or from leukocytes such as ILC1s, plasmacytoid dendritic cells, and lymphocytes (Schoggins, 2019). DAMPS that promote Type 1 epithelial responses are addressed in the next section.

Virus infection of airway epithelial cells results principally in the induction of the type III IFN-λs and IFN-β, though some, but not all of the 13 subtypes of IFN-α are also induced (Khaitov et al., 2009). The type I IFNs all bind the same surface receptor, composed of two proteins, IFNAR1 and IFNAR2, found on the surface of all nucleated cells (Schreiber, 2017). The type III IFNs signal via their common receptor IFNλR1 (IL28Rα) and IL-10Rβ, which is expressed mostly on epithelial surfaces of the respiratory and gastrointestinal tracts (Galani et al., 2017). Both receptors signal via the Janus family kinases (JAKs) to activate STAT proteins to induce hundreds of ISGs with wide-ranging direct anti-viral activities as well as immune stimulatory activities, including stimulation of adaptive immunity (Schoggins, 2019).

Studies of influenza virus infection in mice show that the IFN-λs are the first IFNs produced, and that they act at the epithelial barrier to suppress initial viral spread without activating inflammation. However, if those initial responses are insufficient and infection progresses, type I IFNs come into play to enhance anti-viral resistance. However, they also induce pro-inflammatory responses, thereby causing immunopathology (Galani et al., 2017). Recent studies in COVID-19 have confirmed that maladaptive IFN responses are associated with severe outcomes. Temporal studies of IFN induction in patients hospitalized with COVID-19 pneumonia showed that IFN-λ and type I IFN production were both diminished and delayed, and were induced in only a small fraction of patients as they became critically ill. Higher IFN-λ concentrations in these patients with COVID-19 correlated with lower viral loads in bronchial aspirates and faster viral clearance, a higher IFN-λ to type I IFN ratio correlated with improved outcomes (Galani et al., 2021), and recovery was paralleled by increased type I IFNs (Contoli et al., 2021).

There are a wide range of intracellular and extracellular DAMPs that may be released during respiratory virus infections, including heat-shock proteins, high-mobility group box 1 (HMGB1), hyaluronan fragments, ATP, uric acid, heparin sulfate, antimicrobial peptides (AMPs), dsRNA, ssRNA and DNA (Roh and Sohn, 2018; Zindel and Kubes, 2020). Mostly they induce the pro-inflammatory cytokines including IL-1β, IL-6, TNF, and CXCL8, which amplify innate and adaptive immune responses. However, one alarmin, IL-33, is notable because it is frequently induced by virus infection of airway epithelial cells and it potently induces Type 2 cytokine production from Th2 cells and ILC2s, resulting in the induction of an asthma phenotype (Jackson et al., 2014; Goldblatt et al., 2020). Type 2 epithelial responses do not seem to induce resistance to respiratory virus infection, rather to the contrary, the Type 2 cytokines IL-4, IL-5, and IL-13 suppress IFN induction (Chen et al., 2014; Contoli et al., 2015; Gielen et al., 2015; Hatchwell et al., 2015), so it is unclear whether the induction of IL-33 by respiratory viruses is a viral immune evasion mechanism or there is some adaptive value to the host that is not currently recognized.

Many studies have reported delay and deficiency in induction of type I and type III IFNs in response to virus infection of airway epithelial cells in asthma (Wark et al., 2005; Contoli et al., 2006), and these delays and deficiencies are implicated in increased severity of virus-induced exacerbations (Contoli et al., 2006). Similar studies report that IFN deficiency is implicated in increased chronic obstructive pulmonary disease (COPD) exacerbation frequency (Singanayagam et al., 2019). These studies led to a clinical development program to study the effect of inhaled IFN-β on worsening of asthma symptoms caused by viral infections (Djukanovic et al., 2014; McCrae et al., 2021). These studies reported encouraging improvements in symptom severity and lung function in people with moderate/severe asthma, but neither study met its primary endpoint as insufficient numbers of volunteers suffered virus-induced worsening of their asthma of sufficient severity for the treatment to have an impact.

IFN therapy has also been studied recently in COVID-19, with early subcutaneous IFN-β proving effective at reducing severe outcomes, and early subcutaneous IFN-β and IFN-λ both proving effective at speeding up virus clearance when administered in the first 7 days from symptom onset (Hung et al., 2020; Feld et al., 2021). Inhaled IFN-β was also reported effective at speeding up recovery of patients admitted to hospital with COVID-19 symptoms (Monk et al., 2021).

Many of the PAMPs that signal microbial infection by viruses and bacteria do not figure prominently in the recognition of infection by parasites or fungi. Thus, PRRs such as the TLRs and RLRs that are important in Type 1 and Type 3 responses are not usually involved in the initiation of Type 2 responses. An important exception is the recognition of allergic stimuli by TLR4 through the binding of fibrinogen cleavage products or by functional mimicry of the TLR4 binding protein MD-2 by allergens (Hammad et al., 2009; Trompette et al., 2009; Millien et al., 2013). Proteolytic activity is a common feature of allergens, and the release of extracellular proteases is also a prominent feature of fungal growth and parasite invasion (Li et al., 2021). Fibrinogen cleavage products directly upregulate MUC5AC production in isolated airway epithelial cells (Millien et al., 2013), and protease-activated receptors (PARs) also sense the presence of pathogens and allergens (Li et al., 2021).

An important molecular component of both helminths and fungi that signals infection is chitin, which is found in the cuticle and grinder of multiple helminths, and in the cell wall of multiple fungal species (Dickey, 2007; Przysucha et al., 2020). Introduction of chitin into the lungs of mice strongly induces airway epithelial mucous metaplasia and the accumulation of ILC2s and eosinophils (Reese et al., 2007). Mechanisms of the sensing of chitin in the airway have not been fully determined, but likely involve recognition by chitinases and chitinase-like proteins. In turn, upregulated expression of chitinases and chitinase-like proteins is a prominent feature of Type 2 responses. Other fungal cell wall components that activate Type 2 responses are β-glucans, galactomannans, and mannoproteins, which signal via C-type lectin-type receptors (CLR) on epithelial cells directly or indirectly through leukocytes (Li et al., 2021). Tuft cells are a rare cell type in the airway that express receptors for succinate and tastants, leading to the release of eicosanoids and IL-25 that strongly induce Type 2 responses (Schneider et al., 2019).

Multiple cytokines communicate between leukocytes and lung parenchymal cells in the development of Type 2 immune responses, but among these, IL-13 plays a central role in epithelial responses (Frey et al., 2020; Hammad and Lambrecht, 2021). It is capable of inducing many or all of the epithelial effector responses described in this section, and conversely, inhibition of IL-13 results in incomplete responses to most other stimuli.

While the pathologic role of mucus hyperproduction and rapid secretion (together termed “mucus hypersecretion”) in fatal asthma has been recognized for more than a century (Fahy and Dickey, 2010), the protective role of mucus hypersecretion in resistance to parasite migration through the lungs has only been demonstrated recently (Campbell et al., 2019). Specifically, lumenal mucus traps larvae in the airway lumen, preventing the completion of the parasite life cycle of migration to the gut. Similarly, chitinases and chitinase-like proteins promote resistance to both parasites and fungi, but can also contribute to the pathophysiology of asthma and other airway diseases (Bigot et al., 2020; Przysucha et al., 2020). The roles of multiple other molecules involved in Type 2 epithelial responses have a similar two-sided character, participating both in resistance to helminths and fungi and in the pathogenesis of non-infectious airway diseases (Hammad and Lambrecht, 2021).

In much of the world, parasite infestation remains a major problem. However, in industrialized countries, parasite infection is rare whereas allergic diseases are a greater problem. Thus, therapies targeting Type 2 signaling, such as monoclonal antibodies directed against IL-13, IL-5, IL-33, TSLP (thymic stromal lymphopoietin) or their receptors, or against IgE, have been effective in treating allergic asthma without inducing a substantial cost in parasite clearance (Hammad and Lambrecht, 2021). Neither have these therapies been associated with an increase in airway fungal infection, though whether this is due to the effectiveness of baseline mucociliary clearance, the effectiveness of Type 3 antifungal immune mechanisms that act in parallel to Type 2 mechanisms, or some other reason, is not clear. Targeting Type 2 effector mechanisms, such as chitinases, MUC5AC, and ClCa (chloride channel accessory) signaling in allergic diseases is also being explored (Keeler et al., 2018; Przysucha et al., 2020), and might be expected to have similar benefit without incurring substantial cost in parasite or fungal defense. By the same line of reasoning but in the opposite direction, augmentation of Type 2 epithelial immune responses would generally not be expected to have therapeutic value in industrialized countries.

PAMPs that signal bacterial infections include bacterial lipopolysaccharide (LPS), endotoxins found on the cell membranes of Gram-negative bacteria (recognized by TLR4), bacterial flagellin (recognized by TLR5), and lipoteichoic acid and peptidoglycans from Gram-positive bacteria (recognized by TLR1/2/6). NLRs similarly recognize many of the same PAMPs, and activation of both TLRs and NLRs by the same stimulus can promote particularly vigorous responses. Besides these well-recognized PRRs, numerous other epithelial receptors sense the presence of microbes directly or indirectly, such as CLRs and complement receptors (Evans et al., 2010; Motta et al., 2015; Ablasser and Chen, 2019).

Many DAMPs also activate Type 3 responses in the recognition of tissue injury and cell death (Roh and Sohn, 2018; Zindel and Kubes, 2020). These include macromolecules released from extracellular matrix such as hyaluronan and decorin, proteins released from the cytoplasm such as heat shock and S100 proteins, proteins released from the ER such as calreticulin, proteins released from mitochondria such as formyl peptides and mitochondrial transcription factor A (TFAM), and nucleic acids and proteins released from the nucleus such as HMGB1 and IL-1α. Many of these activate TLRs and NLRs, and they may also activate distinct receptors such as the receptor for advanced glycation endproducts (RAGE) and formyl peptide receptor 1 (FPR1). Besides these macromolecules, small molecules such as ATP and uric acid can signal tissue damage via purinergic receptors and NLRs.

Cytokines that mimic PAMPs and DAMPs in activating Type 3 epithelial immune responses are IL-22 and multiple IL-17 family members (Dudakov et al., 2015; Swedik et al., 2021). IL-22 and some IL-17 family members are secreted by leukocytes to activate epithelial cells, while other IL-17 family members are secreted by epithelial cells in response to PAMPs and DAMPs to signal to neighboring epithelial cells and to leukocytes.

Historically, most attention on Type 3 responses was focused on neutrophil recruitment, induction of inflammatory cytokines such as IL-1β, IL-6, and TNF, and the induction of AMPs. However, as noted above, ROS generation by epithelial cells was identified early as a critical antimicrobial defense in Drosophila (Ha et al., 2005). In the mammalian airway, antimicrobial resistance can be powerfully induced by exposure to an aerosolized bacterial lysate or to the combination of TLR agonists ODN M362 (agonist of TLR9 and multiple cytoplasmic nucleic acid sensors) and Pam2CSK4 (agonist of TLR2/6) (Evans et al., 2011), as well as other PRR ligands (Evans et al., 2010), or IL-22 or IL-17 cytokines (Dudakov et al., 2015; Swedik et al., 2021). Recently, the principal effector of the resistance response to ODN-Pam2CSK4 was found to be ROS generated both by DUOXs (dual oxidases) and mitochondria (Kirkpatrick et al., 2018; Ware et al., 2019). Thus, ROS generation appears to be a central defense mechanism from flies to mammals. A multipronged airway epithelial response is characteristic of Type 3 immunity, with cooperation among the limbs of the response as exemplified by secretion of both ROS and the antimicrobial protein lactoperoxidase, resulting in the generation of highly potent antimicrobial oxidized halides (Kirkpatrick et al., 2018; Ware et al., 2019; Lazzaro et al., 2020; Yamakaze and Lu, 2021). In a maladaptive context, all limbs of the response—neutrophils, inflammatory cytokines, AMPs, and ROS—are thought to contribute to the pathogenesis of airway diseases including asthma, COPD, and bronchiectasis.

Type 3 epithelial responses can be both adaptive and maladaptive, so the potential for therapeutic modulation exists in both directions. As noted above, aerosolized PRR ligands and Type 3 cytokines have the potential to augment resistance to microbial infections of the lungs, and some are in clinical trials for this purpose (e.g., NCT04312997). Conversely, in view of the multiple potential contributions of Type 3 responses to the pathogenesis of airway diseases, antioxidants, PRR antagonists, and cytokine antagonists are being studied in the mitigation of chronic airway diseases (Dudakov et al., 2015; Swedik et al., 2021). An important therapeutic principle for future studies will be disentangling beneficial short-term Type 3 responses from pathological chronic responses.