Inflammasomes in respiratory diseases: from disease pathogenesis to drug development and use of multi-omics

Introduction

The respiratory system is a complex and interconnected network responsible for gas exchange. Due to the constant exposure to the external environment, host defences are continuously engaged to protect against noxious stimuli. Inflammation is a protective immune response initially triggered by innate immunity in response to endogenous and/or exogenous stimuli such as invading pathogens, dead or dying cells, toxins, smoke and air pollutants [1–3]. It involves the influx of inflammatory cells that perform phagocytosis and efferocytosis and release enzymes, antibacterial factors, cytokines, and chemokines. These responses aim to destroy the stimuli, a critical step for restoring homeostasis and overcoming cellular stress, but must be tightly regulated to prevent damage to host tissue [4]. Ineffective inflammation result in unrestricted pathogen invasion, while excessive and prolonged inflammation causes tissue damage and the development of chronic inflammatory diseases [5, 6]. In the respiratory system, this can be catastrophic, resulting in an influx of excessive inflammatory cell recruitment which can cause acute lung injury and acute respiratory distress syndrome, mucus hypersecretion that obstructs airflow, collagen deposition, fibrosis and narrowing of the airways, obstruction of airflow, alveolar damage and emphysema, and impaired lung function and gas exchange [5, 7].

Inflammasomes are multiprotein signalling complexes that control host inflammatory responses [8]. They are initially activated via a loss of cellular homeostasis [9]. This can occur when damage-associated molecular patterns (DAMPs) emerging from dead, dying or damaged cells, or from pathogen-associated molecular patterns (PAMPs) found on the surface of bacteria, fungi, protozoa or viruses damage important organelles in the cell or otherwise compromise essential functions [10, 11]. Inflammasome sensors are frequently kept in an inactive state, functioning as a conserved “guard mechanism” similar to plant innate immunity and are thus activated when homeostasis altering molecular processes relieve this constraint [9]. For example, pyrin is an inflammasome sensor that is kept inactive by 14-3-3 protein dimers; however, pathogen effectors acting on RhoGTPases prevent the 14-3-3 binding and trigger pyrin inflammasome formation [12]. Similarly, nucleotide-binding oligomerisation domain, leucine-rich repeat and pyrin domain containing family (NLRP) 3 is an inflammasome sensor that is kept inactive in a cage-like structure on the trans-Golgi network, but disruption of the structure allows the protein to oligomerise into an active inflammasome [13]. Although DAMPs/PAMPs do not typically interact with inflammasome sensors directly, they engage other pattern recognition receptors (PRRs) to provide the first (priming) signal for inflammasome activation by inducing expression of inflammasome sensors such as NLRP3 and pro-IL-1β [14]. A second activation signal, triggered by cellular stress or homeostasis-altering molecular processes, then promotes assembly of the inflammasome complex. PRRs are expressed by immune cells such as macrophages, neutrophils, dendritic cells (DCs) and B- and T-lymphocytes, as well as some nonclassical immune cells such as epithelial cells [15]. There are four major families of PRRs, inluding transmembrane proteins such as 1) Toll-like receptors (TLRs) and 2) C-type lectin receptors (CLRs), cytoplasmic proteins such as 3) nucleotide-binding oligomerisation domain (NOD) leucine rich repeat-containing receptors (NOD-like receptors or NLRs) and 4) retinoic acid-inducible gene (RIG)-I-like receptors (RLRs) [15, 16]. Each PRR induces several different signalling pathways which are largely dependent on the specificity of the DAMP/PAMP stimuli [16]. The most well-described is the NLRP3 inflammasome, but others such as the NLRP1, NLRP6, NLR family CARD domain-containing protein 4 (NLRC4) and absent in melanoma 2 (AIM2) inflammasomes have been previously described [16].

Upon PRR recognition of DAMPs/PAMPs, transcription factors are induced, most notably NF-κB and interferon regulatory factors (IRFs), which drive the expression of inflammasome-related and pro-inflammatory genes, as well as type-I interferon (IFN) production to mount effective host defences [17, 18]. This priming step is followed by a second activation signal, resulting in inflammasome assembly and oligomerisation. Inflammasomes are differentiated by the inflammasome sensor molecule used in their assembly, such as NLRP3 or AIM2, which oligomerises with the inflammasome adaptor protein, apoptosis-associated speck-like protein (ASC) and caspase-1 to form the inflammasome complex [17, 19]. NIMA-related kinase 7 (NEK7) is a novel, essential component contributing to the oligomerisation of the NLRP3 inflammasome; however, it is not involved in NLRC4 or AIM2 inflammasomes [20]. Once the inflammasome proteins have assembled, caspase-1 is activated and proteolytically converts pro-IL-1β and pro-IL-18 into mature forms [17]. Caspase-1 also activates gasdermin D (GSDMD) by cleavage between the N-terminal and C-terminal domains allowing N-terminal GSDMD to bind to the inner leaflet of the cell membrane and form nanopores. This causes cell swelling and rupture, inducing a highly pro-inflammatory form of cell death known as pyroptosis [21–23]. This causes the release of mature IL-1β, IL-18 and cellular contents into extracellular spaces, leading to local inflammation and immune cell recruitment, ultimately triggering inflammatory cascades (figure 1). While pyroptosis can play a protective role during infections, it can also lead to pathological inflammation and chronic inflammatory diseases [24]. Thus, therapeutic targeting of inflammasome activation is an attractive target for mitigating inflammatory diseases. In fact, therapeutic targeting of inflammasome activation has been particularly successful for the treatment of autoinflammatory diseases. Clinical trials have demonstrated that treatment with IL-1 receptor agonist, anakinra, is highly effective at blocking the activity of IL-1 and subsequently is an effective treatment for cryopyrin-associated periodic syndromes, tumour necrosis factor (TNF) receptor-associated periodic fever syndrome, mevalonate kinase deficiency and colchicine-resistant familial Mediterranean fever [25]. For the treatment of other chronic inflammatory disease, however, treatment would be dependent on the stimuli and the inflammasomes involved.

In this review, we explore the role of inflammasomes in the context of respiratory diseases. We also discuss currently available therapeutics used to mitigate inflammasome activation within these diseases and the use of multi-omics to develop future therapies that can inhibit inflammasomes and prevent pathological inflammation in the lung.

Methods/search strategy

Literature searches were performed using the PubMed database with search terms including “inflammasomes” with any combination of the following search key words/criteria: “respiratory disease”, “chronic respiratory disease”, “therapeutics”, “drug discovery”, “chronic obstructive pulmonary disease”, “asthma”, “idiopathic pulmonary fibrosis”, “respiratory infections”, “bacterial infections”, “tuberculosis”, “Mycobacterium tuberculosis”, “viral infections”, “influenza”, “COVID-19”, “SARS-CoV-2”, “proteomics”, “lipidomics”, “metabolomics”, “transcriptomics” and “multi-omics”. Where possible, primary literature sources were used for citation in this review. Publications from the last 10 years were prioritised for inclusion to ensure that this review is timely and includes major advancements in the field. Where required due to the need to include primary literature sources and foundational knowledge in the field, publications that were from >10 years since publication were included in this review.

Asthma

Asthma is heterogeneous and characterised by chronic airway inflammation resulting in wheezing, cough, chest tightness and breathing difficulties caused by mucus production, airway remodelling and hyperresponsiveness (AHR) [26, 27]. Classically, asthma was thought to be driven by type-2 immune responses and eosinophilic inflammation [28]. However, studies in adult patients with moderate-to-severe asthma have identified disease with type-1/17 immune responses with noneosinophilic neutrophilic inflammation [26, 27]. Exposure to bacterial and viral infections, house dust mite (HDM), animal dander and smoking stimulate the production of IgE and type-2 cytokines including IL-4, IL-5 and IL-13, all of which have been implicated in asthma [29]. Repeated allergen exposure causes chronic disease with airway eosinophilic inflammation, mucus hypersecretion, bronchoconstriction, remodelling and AHR [26, 27]. Infection and a high fat diet and obesity are linked to the development and progression of severe asthma with a type-1/17 phenotype [26, 30, 31]. Critically, 5–10% of people with asthma have severe steroid-resistant (SSR) disease that do not respond to current therapies and accounts for 50–80% of asthma-related healthcare cost, highlighting the urgent need to understand steroid-resistance development in asthma patients [26, 27, 32, 33].

Inflammasomes and IL-1β responses have crucial roles in asthma development and progression [34–36]. Several studies have revealed the involvement of TLR2/4 and NLRP3 inflammasomes in asthma [37]; however, few have investigated other inflammasomes. In vivo studies demonstrate the significance of KN-62, a potent noncompetitive antagonist of the P2X7 receptor and selective inhibitor of calcium/calmodulin-dependent protein kinase II (CaMKII). KN-62 suppressed hallmark features of asthma by targeting pathways that promote inflammation and tissue injury. CaMKII-mediated phosphorylation of dynamin-related protein 1 (Drp1) and c-Jun N-terminal protein kinase 1 (JNK1) drives Drp1-dependent mitochondrial fission, reactive oxygen species (ROS) accumulation, and JNK1-dependent inflammasome activation [38–40]. This indicates the involvement of ATP-mediated NLRP3 inflammasome activation in asthma pathogenesis. Others have demonstrated the role of mitochondrial reactive oxygen species (mtROS) in allergic airway inflammation, which occurs via NLRP3 inflammasome activation using lipopolysaccharide (LPS) with ovalbumin and HDM mouse models [41]. NecroX-5, a novel mtROS inhibitor, supressed NLRP3 and caspase-1 levels in bronchoalveolar lavage fluid (BALF) and restored mitochondrial DNA content and integrity [41]. This revealed that inhibiting mtROS production inhibits asthma via modulation of NF-κB and NLRP3 inflammasome activation. Additionally, the NLRP3 inhibitor RRx-001 suppressed airway inflammation and mucus secretion in an NLRP3-dependent manner [42].

SSR asthma pathogenesis is linked to elevated NLRP3, caspase-1 and IL-1β responses, highlighting the importance of furthering our understanding of these pathways for future therapeutic development. TLR2, TLR4, IL-1β and IL-18 expression and elevated LPS levels were found in the sputum of patients with neutrophilic asthma compared to other asthma phenotypes and healthy controls [43]. Elevated expression of NLRP3, caspase-1/4/5, IL-1α/β and TNF/NF-κB signalling pathways have been found in sputum associated with neutrophilia in asthma [44, 45]. Specifically, increased IL-1β correlated with IL-8 levels indicating neutrophilic inflammation in SSR asthma [45]. In addition, alterations in the endoplasmic reticulum (ER)-unfolded protein response and dysregulation of genes associated with ER protein chaperones, degradation and apoptosis have recently been associated with impaired lung function and eosinophilic and neutrophilic inflammation. This correlated with T2, T17, inflammatory responses and inflammasome activation (NLRP3) in sputum from asthmatic participants [46, 47]. These findings highlight the complexity of asthma phenotypes and underscore the need for thorough characterisation in inflammasome responses.

Respiratory infections including Chlamydia, Haemophilus influenzae, influenza A virus (IAV) and respiratory syncytial virus (RSV) are associated with SSR asthma leading to release of active IL-1β, mediated by caspase-1 in an NLRP3-dependent manner [48–50]. NLRP3 and IL-1β sputum mRNA expression positively correlates with sputum neutrophil numbers and exacerbated AHR in people with stable asthma taking inhaled corticosteroids [33]. Importantly, administration of MCC950 (highly selective NLRP3 inflammasome inhibitor), ac-YVAD-cho (selective caspase-1 inhibitor) or anti-IL-1β monoclonal antibody in Chlamydia or Haemophilus infection-induced SSR asthma mouse models effectively suppressed increased IL-1β responses and steroid-resistant airway neutrophilia and AHR. Thus, targeting NLRP3 inflammasomes may be a novel therapeutic approach for SSR asthma.

Obesity is a risk factor for asthma, primarily SSR asthma but is associated with AHR particularly in females [51–53]. Body mass index correlates with increased NLRP3 expression and may result in neutrophilic inflammation in the airways as well as obesity-associated diseases [54]. Obesity also correlates with elevated type-2 cytokine expression with associated increases in eosinophils and C-C chemokine receptor type 3 in sputum in human and experimental asthma [52]. MCC950 treatment suppressed AHR but had no effect on lung type-2 cytokines; however, administration of anti-IL-5 and anti-IL-13 strongly suppressed NLRP3, AHR as well as lung eosinophils in experimental obese asthma [52, 55]. The relationship between T2 cytokine and NLRP3 inflammasome responses in obesity-associated asthma, highlights a potential for targeted therapies such as T2 cytokine biologics and inflammasome inhibitors.

Another inflammasome that is genetically linked to asthma is NLRP1 [56]. Perhaps surprisingly, NLRP1 activation and the subsequent production of IL-18 appear to have a beneficial role in mouse models of asthma by promoting host defence against respiratory pathogens. Typically, NLRP1 binds dipeptidyl peptidase 9 in vitro and suppresses eosinophilia in vivo [56]. In another study, NLRP1 was found to be a highly expressed inflammasome sensor in human bronchial epithelial cells of healthy and asthmatic patients. Human rhinoviruses are frequently the cause of asthma exacerbations and, among other pro-inflammatory responses, rhinovirus infection promotes activation of the inflammasome pathway. Viral proteases cleave NLRP1, leading to the formation of inflammasome and the secretion of cytokines IL-1β and IL-18, and small interfering RNA targeting NLRP1 was shown to inhibit rhinovirus-triggered inflammasome activation [57]. Thus, transient NLRP1-mediated IL-18 production may facilitate viral clearance, whereas prolonged or excessive inflammasome activation may contribute to airway inflammation and tissue damage. Given that NLRP1 is an innate immune sensor of multiple pathogens, there is likely a balance between the beneficial effects of inflammasome activation during infection and the detrimental consequences of sustained IL-1β, IL-18 and pyroptotic signalling. This has implications for targeted therapies in asthma and broader considerations for the NLRP1 inflammasome specifically (figure 2).

COPD

COPD is characterised by persistent and progressive airflow limitation and pro-inflammatory responses in the lung, peripheral airways and parenchyma. It is the fourth leading cause of death worldwide and current research aims to better understand COPD pathophysiology and develop novel therapies [58, 59]. Since chronic inflammation and pyroptosis are important clinical features in COPD, the role of inflammasomes in COPD pathogenesis has been explored [60, 61].

Multiple studies using cigarette smoke (CS)-induced cell and animal models and human samples show that NLRP3, IL-1β and IL-18 mRNA levels are elevated [62–64]. Canonical NLRP3 inflammasome activation requires a two-step process comprising an initial priming signal, induced by PAMPs/DAMPs, which upregulates NLRP3 and pro-IL-1β expression, followed by a second activation signal that triggers assembly of the NLRP3–ASC–caspase-1 complex and maturation of IL-1β and IL-18. Examination of sputum and serum from COPD and healthy patients showed that NLRP3 activation is limited to the airways with increased mRNA expression in sputum, but not serum, from patients with stable COPD compared to the healthy controls [65]. NLRP3, caspase-1 and IL-1β mRNA levels were further increased during acute COPD exacerbations; demonstrating the association between NLRP3 activation and COPD severity [65]. Another study assessed a CS-induced COPD mouse model and BALF from patients with COPD to demonstrate that post-pyroptosis ASC specks, a hallmark feature of inflammasome activation, accumulate in extracellular spaces where they promote IL-1β activation [19]. Phagocytosis of the ASC specks by macrophages induced lysosomal damage and activation of IL-1β in recipient cells, triggering inflammatory cascades. These findings suggest that inflammasome activation and subsequent ASC speck accumulation might play critical roles in prolonged inflammatory processes in COPD. Interestingly, in total lung tissue from a CS-induced COPD mouse model, CS-induced inflammation is attenuated upon IL-1α or IL-1β neutralisation but is not affected by NLRP3 or caspase-1 deficiency [66]. Moreover, in stable COPD the gene expression of NLRP3 and IL-1β are upregulated; however, caspase-1 and ASC remain in inactive forms [67]. It was only during infectious COPD exacerbations that the protein levels of extracellular caspase-1, IL-1β and IL-18 were increased. This highlights that under stable conditions, the NLRP3 inflammasome is primed but not active and only induces inflammation during challenge. Only limited efficacy of anti-IL-1β, IL-1R1 and IL-18 antibodies were shown in randomised clinical trials of inflammasome therapies in COPD [68]. The exact role of NLRP3 inflammasomes in COPD pathogenesis remain to be fully understood.

Since DNA damage from oxidative stress contributes to the pathogenesis of COPD, there has been considerable interest in the role of AIM2 [69, 70]. AIM2 expression is increased in COPD lungs which associated with increased IL-1β and CD14⁺ peripheral blood mononuclear cells and was responsible for IL-1α release in a caspase-1/4-dependent manner [71, 72]. A positive correlation between AIM2 ASC speck foci and COPD disease severity stages was identified unlike with NLRP3 [73]. More recently, AIM2 deletion was shown to increase airway neutrophilia, but decrease caspase-1 activation in the parenchyma and lung neutrophils in a CS-induced mouse model of COPD [74]. Anti-Ly6G treatment in Aim2^−/– mice had no effect on neutrophil levels or neutrophil caspase-1 levels in the lungs. CS-induced AIM2 activation induced higher recruitment of immunosuppressive cells, such as nonactive DCs and myeloid-derived suppressor cells. Thus, AIM2 regulates caspase-1 activation in neutrophils but also has caspase-1-independent effect. Due to the heterogeneity of COPD, further studies are required to elucidate these interactions and develop new therapies that regulate inflammasome activation and disease pathology (figure 2).

Idiopathic pulmonary fibrosis (IPF)

IPF is a relatively rare, fatal condition of older people with an estimated prevalence of 1.25–63 per 100 000 of the population, with a median survival of 3–5 years from diagnosis [75]. There are no treatments that have a major survival benefit, with current frontline therapies, nintedanib and pirfenidone, merely slowing disease progression by 50% [76, 77]. IPF is characterised by pulmonary fibrosis and remodelling, where normal lung tissue is interspersed with interstitial fibrosis, honeycomb cysts and fibroblast foci [78, 79]. The lung extracellular matrix (ECM) is a complex architectural framework consisting of >400 proteins that provide structural support, cell anchorage and regulate cell phenotype [80]. During normal wound healing, ECM-secreting cells, myofibroblasts, are rapidly recruited to the site of active repair. Once repair processes are underway, myofibroblasts are removed through apoptosis and lung injury tissue resolves [81]. Abnormal deposition of ECM proteins in the airways and parenchyma, and apoptosis-resistant myofibroblasts are hallmark features of IPF, resulting in irreversible lung tissue scarring. Subsequently, myofibroblasts cluster into myofibroblastic foci and ECM deposition extends into the interstitium, resulting in airway narrowing, parenchymal stiffening, alveoli damage and lung function decline leading to severe breathing difficulties [82].

Innate immunity plays a role in the pathogenesis of IPF [83, 84], whereby macrophages and monocytes are key drivers of aberrant tissue repair. Recently, a unique subset of alveolar macrophages (AMs) was identified that contribute to fibrotic lung pathogenesis [85], supporting extensive research implicating inflammatory pathways in IPF aetiology [84, 86–88]. Furthermore, IL-18 is elevated in serum, BALF and fibroblastic foci in IPF [89], and IL-1β is expressed in AMs during lung fibrosis [90, 91]. Thus, therapies targeting inflammasome activation hold promise for early IPF, though understanding the inflammasomes implicated, and their mode of priming and activation, are paramount.

NLRP3 inflammasome activation is increased in many forms of pulmonary fibrosis including silicosis [92], asbestos [92], sarcoidosis [93] and IPF [94]. It is activated by distinct stimuli, including PAMPs/DAMPs such as extracellular ATP [17], ROS accumulation [95], toxins and particulate matter [96]. In IPF, NLRP3 and caspase-1 are elevated in AMs and lung fibroblasts, as were IL-1β and IL-18 [94, 97]. However, surprisingly there was impaired inflammasome activation following LPS and LPS/ATP stimulation of BALF AMs from patients with IPF. This dysregulation of NLRP3 activation supports the concept that impaired innate immunity and the inability to clear infections [98, 99], could be a major factor in the progression of early IPF. Recent findings suggest NLRP3 inflammasome components, NLRP3 and pro-IL-1β, were upregulated in lung fibroblasts from mice with bleomycin-induced pulmonary fibrosis and that NLRP3 inflammasomes are required for the proliferation of fibrosis-associated lung fibroblasts and myofibroblast differentiation [97].

AIM2 inflammasomes are pathogenically linked to DNA damage associated with lung fibrosis, with an increased expression in IPF AMs [100]. Upon activation of AIM2 with poly(dA:dT), peripheral mononuclear cells from patients with IPF secreted high levels of the pro-fibrotic mediator transforming growth factor (TGF)-β, and pro-inflammatory cytokines IL-1α and IL-18. This was found to be caspase-4 dependent and antifibrotic drugs nintedanib and pirfenidone did not reduce TGF-β levels following AIM2 activation [101]. AIM2 inflammasomes were also activated during Streptococcus pneumoniae infection and was critical for glucose transporter 1-dependent glycolysis and subsequent induction of lung parenchymal fibrogenesis [102]. AIM2 protein levels are increased in lungs of patients with IPF, along with heightened Drosha ribonuclease-III (DROSHA) crucial for microRNA biogenesis [100]. DROSHA deficiency in LPS-primed mouse AMs challenged with poly(dA:dT) suppressed AIM2 inflammasome-dependent caspase-1 activation and cytokine release, identifying a novel molecular target for inflammasome activation and lung fibrosis. Therapeutic studies to target AIM2 activation in lung fibrosis models are limited but promising. Andrographolide, an active component from the plant Andrographis paniculate, reduced pro-inflammatory cytokine release, immune cell infiltration and collagen deposition following radiation-induced lung injury through the suppression of AIM2 inflammasome-mediated pyroptosis in AMs [103].

Mitochondrial function and metabolic dysregulation has emerged as a key determinant of IPF progression [104] and a source of DAMPs such as cardiolipin, succinate and ATP. Stimulation of NLRP3 and AIM2 with ATP and poly(dA:dT) respectively in IPF AMs, increased IL-1β release from all lung fibrosis groups (IPF and non-IPF-interstitial lung disease patients) compared to healthy controls [95]. Furthermore, increased mtROS was crucial for hyperactivation of NLRP3 and AIM2, which was reduced with MitoTEMPO treatment. Aged mitochondria studies show that NLRP3 activation and bleomycin-induced fibrosis were augmented in aged versus young mice and reduced in Nlrp3^−/– mice [105]. Inhibiting NLRP3 inflammasomes shows promise as a therapeutic strategy for IPF. MCC950 [97, 106], lycorine [107], scutellarin [108] and anakinra (IL-1RA) [97] all inhibit the NLRP3 inflammasome, reduce inflammation and prevent lung fibrosis development. Pre-treatment of bleomycin-challenged mice with NecroX-5, an antioxidant and necrosis inhibitor, suppressed oxidative stress, NLRP3 inflammasome activation and pulmonary fibrosis by inhibiting the TGF-β1/SMAD2/3 pathway [109]. Recently, a dual NLRP1/3 inflammasome inhibitor, ADS032, reduced pulmonary inflammation in a silicosis mouse model in an NLRP3-dependent manner [110]. cAMP signalling is increasingly recognised as a therapeutic target in IPF. Although cAMP-mediated protein kinase A activation is a negative regulator of the NLRP3 inflammasome, the role of its downstream effector, the exchange protein directly activated by cAMP 1 (Epac1), appears more complex. Increased Epac1 expression has been observed in IPF lungs and pharmacological inhibition with the small-molecule inhibitor AM-001 reduced fibroblast proliferation, profibrotic marker expression and bleomycin-induced fibrosis, potentially through suppression of IL-6/signal transducer and activator of transcription 33, TGF-β and neural precursor cell expressed, developmentally downregulated 8 signalling [111]. Conversely, BI 1015550, an oral preferential phosphodiesterase 4B inhibitor that increases intracellular cAMP, slowed lung function decline in patients with IPF [112], further supporting modulation of cyclic nucleotide signalling as a promising therapeutic strategy (figure 2).

Thus, therapeutically targeting the inflammasome pathway may provide a novel strategy for IPF treatment; however, understanding the mechanisms of inflammasome activation remains paramount in producing effective inflammasome therapies for IPF.

Respiratory infections

Pathogenic infections trigger inflammatory responses by binding of PAMPs such as TLRs to host-sensing receptors that prime inflammasomes [113, 114]. Inflammasomes have complex roles in driving infection-induced inflammation and can be protective or detrimental to infection outcomes depending on the specific pathogen involved and the timing of inflammasome activation [115].

Bacterial infections can broadly be split into either Gram-positive or Gram-negative bacilli. Gram-positive bacteria such as S. pneumoniae contain a thick outer layer known as the peptidoglycan which protects bacilli and contains many antigenic proteins such as lipoteichoic acid (LTA) [116], while Gram-negative bacteria such as Pseudomonas aeruginosa possess a thin peptidoglycan layer which uncovers the highly antigenic molecule LPS [117]. Importantly, there are key differences in the inflammasome activation between these two groups, with Gram-positive LTA predominantly activating the NLRP6 inflammasome [116], while Gram-negative LPS binds to TLR4 and primes the NLRP3 and AIM2 inflammasomes [118]. Importantly, other significant respiratory bacterial pathogens such as Mycobacterium tuberculosis (Mtb) are not typically classified as either Gram-positive or negative, rather they contain a high concentration of long-chain fatty acids called mycolic acids that coat the outer bacterial surface and directly interact with CLRs to activate the NLRP3 inflammasome [119]. Mtb, the causative agent of tuberculosis (TB), is responsible for ∼1.6 million deaths annually and represents the leading cause of death by infectious disease globally [120, 121]. Mtb is effectively contained within granulomas, which are organised structures formed by a network of recruited inflammatory cells. Granulomas successfully limit the spread of the infection but typically cannot eliminate the pathogen entirely from the host's body [122, 123]. The role of inflammasomes in TB is complex and whether they are protective or detrimental remains unclear. Some studies that show Mtb directly inhibits NLRP3 [124, 125], and others show that Mtb dampens NLRP3 activation by upregulating NOS and IFN-β [126–130], and limiting potassium efflux [131]. Mtb-induced NLRP3 activation also causes macrophage plasma membrane rupture and pyroptosis dependent on a functional type-7 secretion system (ESX-1), while MCC950 was able to reduce NLRP3 activation and ACS speck formation (figure 2) [129, 132–134]. Others have reported that genetically diverse Mtb isolates manipulate NLRP3 inflammasome and IL-1β independently of host macrophages and requires the presence of live, metabolically active bacteria within the infected macrophages [135]. The role of AIM2 in TB has not been as thoroughly investigated; however, it is critical for IL-1β production and Aim2^−/– mice are hypersusceptible to Mtb infection [136].

Viruses and their PAMPs are also recognised by TLRs, RLRs and DNA sensors such as cyclic-GMP-AMP synthase, as are DAMPs that are released due to cell injury/death following infection [137]. IAV is a seasonal respiratory virus responsible for mild to severe respiratory infections. Up to 10% of people globally are infected with ∼5 million cases of severe disease each year due to seasonal influenza [138]. IAV activates NLRP3 and AIM2 inflammasomes via the M2 protein, a proton-selective ion channel important for IAV replication [49], and triggers oxidised DNA release in macrophages [139]. Conversely, IAV nonstructural protein 1 inhibits both NLRP3 and NF-κB independently, reducing pro-inflammatory cytokine production, which represents an immune evasion strategy [140]. NLRP3 inflammasome inhibition with MCC950 from day 3 post-infection improved survival, reduced pro-inflammatory cytokine production and did not affect viral clearance in an IAV mouse model (figure 2) [141]. However, treatment from day 1 post-infection caused hyper-inflammation and lethality. This suggests that NLRP3 has a protective role during early infection but may be detrimental during later stages [142]. In a similar study, treatment with dual NLRP1/3 inflammasome inhibitor, ADS032, administered after infection resulted in attenuated weight loss and improved survival, indicating that NLRP3 inhibition mitigates harmful inflammation during the later stages of infection. However, in contrast to previous observations with MCC950, when ADS032 treatment was initiated from day 1 post-infection, mice also exhibited reduced weight loss and protection from IAV-associated mortality. This demonstrates that dual NLRP1/3 inflammasome inhibition confers protection when administered either early or late during infection [110]. AIM2 is also induced by IAV infection, with Aim2^−/– mice having attenuated lung injury and improved survival, however viral burden was unaffected [143].

COVID-19, with the aetiological agent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is the cause of the most recent global pandemic, resulting in >7 million deaths worldwide since 2020. The SARS-CoV-2 nucleocapsid protein directly binds to NLRP3, promotes binding of ASC and NLRP3, and drives NLRP3 inflammasome assembly in human macrophages, dendritic cells and mouse models [144]. Similarly, the spike protein activates the NLRP3 inflammasome in human and transgenic mouse microglial cells in an angiotensin-converting enzyme 2-dependent manner [145]. MCC950 reduced pro-inflammatory cytokine production and improved survival, highlighting the detrimental role of NLRP3 in COVID-19 severity [144, 145]. Interestingly, human monocytes can take up antibody-opsonised virus, which in turn activates both NLRP3 and AIM2 inflammasomes causing pyroptosis of infected cells [146]. Moreover, post-COVID-19 fibrosis was related to AIM2 activation implying roles in fibrosis [147], but it was not involved in lung pathology [70], and circulating cells were responsive to AIM2 activation and the caspase-1/4-dependent and independent inflammatory and fibrotic factors. Furthermore, inflammasome responses emerge from tissue-resident macrophages and not infected epithelial or endothelial cells in COVID-19 lung autopsy specimens [146].

Although this review focuses on respiratory disease, inflammasome activation can have important systemic consequences. Sustained pulmonary inflammation and excessive production of IL-1β and IL-18 may contribute to extrapulmonary complications through systemic inflammatory signalling, which can exacerbate cardiovascular pathology, including myocarditis, dilated cardiomyopathy, acute coronary syndromes and heart failure [148, 149]. Additionally, respiratory viral infections are recognised triggers of cardiac injury, highlighting the close interplay between pulmonary and cardiovascular inflammation. Accordingly, vaccination against influenza, SARS-CoV-2, RSV and pneumococcal disease to reduce respiratory infections, and vaccination against herpes zoster vaccines, have been associated with reductions in cardiovascular complications in at-risk populations [150, 151]. While a detailed discussion of cardio-pulmonary interactions is beyond the scope of this review, these systemic effects should be considered when evaluating inflammasome-targeted therapies.

In summary, inflammasomes play critical roles in mediating host responses to both bacterial and viral pathogens. Critically, the success of pharmacological interventions targeting inflammasome activation is largely dependent on the pathogen and the timing of inhibition. Further studies are needed to define the roles of other inflammasomes in respiratory infections.

Multi-omics to identify novel inflammasome biomarkers and therapeutics

Recent advances in omics technologies have enabled researchers to better characterise changes at the transcriptomic, epigenetic and proteomic level in different diseases, which has enabled the discovery of novel disease biomarkers and therapeutic targets using computational approaches [152–154]. These include gene expression based, network-based, mathematical models and machine learning-based approaches [155, 156]. Gene expression-based approaches can encompass various methods, including supervised and unsupervised approaches. Supervised methods such as gene expression differential analysis can help identify disease-specific genes by comparing diseased to healthy samples, while other methods such as network based and machine learning can help identify co-expressed genes. Various powerful tools have been developed using these approaches, including gene expression profiling interactive analysis (GEPIA) [157], the Tumor Immune Estimation Resource (TIMER) [158], Oncomine [159], KM plotter [160], LOGpc [161], OncoLnc [162], MethSurv [163], PrognoScan [164], SurvExpress [165] and UALCAN [166], and have been applied to inflammasomes. These tools have primarily been designed to identify differentially expressed and co-expressed genes, such as components of the inflammasome, and analyse the abundance of specific genes or gene sets in disease versus control samples. These can also conduct survival analyses to correlate gene expression with patient outcomes and investigate the relationship between gene expression and immune cell infiltration. Therefore, they can effectively identify components of the inflammasome that are differentially co-expressed and discriminate their abundance in disease versus controls [157].

Several studies have discovered single gene biomarkers in different human diseases but recent advances in genomics and transcriptomics now allow for high-throughput screening of biomarkers using microarrays, and next-generation sequencing (RNA-Seq). For instance, one study performed unsupervised hierarchical clustering on sputum cell transcriptomics from moderate-to-severe asthmatic and healthy subjects, and identified three distinct transcriptome-associated clusters (TACs). TAC1 is associated with IL-13/T-helper cell type 2 and group 2 innate lymphoid cell signatures, severe eosinophilia and significant airflow obstruction. TAC2 is related to IFN, TNF and inflammasome pathways, neutrophilic inflammation, and elevated c-reactive protein levels. TAC3, involving metabolic, ubiquitination and mitochondrial genes, is linked to mild eosinophilia and improved lung function. This analysis highlighted potential biomarkers and therapeutic targets, offering a new framework for asthma phenotyping [167]. Follow-up studies showed that the TAC2 inflammasome signature was not responsive to inhaled corticosteroids in COPD patients indicating the importance of implementing improved therapies for this patient cohort [168].

Network-based approaches can identify hub genes that potentially control entire networks of disease development or progression and serve as more effective therapeutic targets [169, 170]. A recent study used network-based approach to integrate transcriptomics and epigenomics data from COPD and control lungs, revealed that many genes with consistent positive correlations across RNA and methylation layers are co-regulated in response to stimuli such as CS, which has also been found in asthma [171]. Key immune-related genes, such as TNFSF8 and IL10, were identified, along with novel gene–gene interactions. Notably, AIM2 was highlighted as a significant node, suggesting an important role in COPD pathogenesis [172]. Overall, these techniques can identify differentially expressed and co-expressed genes involved in inflammasome pathways and evaluate potential therapeutics using advanced bioinformatics and machine learning methods [157, 173].

Proteomics investigations have also been performed. The methods used are relatively routine as they are predominantly centred around traditional “shotgun” liquid chromatography–tandem mass spectrometry. In inflammasome studies, the protein-centric approach has focused on specific protein–protein interactions with the functional caspase and pyrin domains contained within NLRP3 [174, 175]. Notable studies used comprehensive proteomics to fully characterise inflammasome-activated protein release in macrophages and provided a resource for further mass spectrometry-based investigations [176, 177]. Lipidomics analysis is crucial to properly understand the effects of inflammasome activation, since the triggering of systemic inflammation includes various lipid species such as poly unsaturated fatty acids and lipoxygenase [178]. In BALF from community-acquired pneumonia, lactosylceramides were the most elevated lipids, which have been shown to induce ROS production and activate the NLRP3 inflammasome [179]. Further, targeted mass spectrometry approaches to specifically isolate the metabolite itaconate showed that it can suppress NLRP3 inflammasome activation in its derivatised state [180]. Recently, a comprehensive study on asthma metabolomics made a clear distinction between targeted and untargeted approaches and the need for future untargeted studies [181].

Taken together, multi-omics datasets represent a powerful method to tie together the transcriptional and proteomic signatures in respiratory diseases (figure 3). Using these datasets, we can then identify biomarkers of disease and important pathways/networks for the discovery of novel therapeutics that can then be assessed through screening differentially expressed molecules and prioritising candidates through advanced bioinformatics or statistical methods including machine learning, random forest and Bayesian reasoning [173]. Furthermore, co-expression network-based approaches can identify hub genes that potentially control entire networks of disease development/progression and be more effective therapeutic targets [169, 170]. More recently, advances in multi-omics analyses have led to the development of online webservers and databases helping researchers identify more effective therapeutic targets which can be found in table S1.

Conclusions

Inflammasomes are at the core of host inflammatory responses; however, our understanding of inflammasomes is incomplete and discovery biology is crucial to further our knowledge of the mechanisms of pathogenesis underpinning acute and chronic inflammasome activation and associated disease. Interrogating representative animal models of acute and chronic lung respiratory diseases will elucidate the delicate balance between protective and detrimental inflammasome responses, identify effective therapeutic targets and develop drugs to treat the sequalae of inflammasome activation in the lung. A summary of the findings as well as leading drug candidates for each inflammasome in the respiratory diseases discussed in this review is presented in table S2 and figure 2.

Supplementary material