Innate immune sensing of cell death in disease and therapeutics

Abstract

Innate immunity, cell death and inflammation underpin many aspects of health and disease. Upon sensing pathogens, pathogen-associated molecular patterns (PAMPs), damage-associated molecular patterns (DAMPs), or alarmins, the innate immune system activates inflammatory cell death, such as pyroptosis and PANoptosis. These genetically-defined, regulated cell death pathways not only contribute to the host defense against infectious disease, but also promote pathological manifestations leading to cancer and inflammatory diseases. Our understanding of the underlying mechanisms has grown rapidly in recent years. However, how dying cells, cell corpses, and their liberated cytokines, chemokines, and inflammatory signalling molecules are further sensed by innate immune cells, and their contribution to further amplify inflammation, trigger antigen presentation, and activate adaptive immunity, is less clear. Here, we discuss how pattern-recognition and PANoptosome sensors in innate immune cells recognize and respond to cell-death signatures. We also highlight molecular targets of the innate immune response for potential therapeutic development.

Healthy living cells do not cause inflammation, whereas cells, which have been infected, stressed, or on the verge of lytic cell death have the capacity to trigger inflammation. Cell death modalities that trigger inflammation are defined as inflammatory cell death. Inflammatory cell death can either be accidental, that is caused by physical or traumatic disruption of the plasma membrane, or regulated, that is executed by the activation of genetically-defined germline-encoded receptors and/or signalling proteins. A well-characterized regulated cell death pathway is inflammasome-dependent pyroptosis, defined by the activation of inflammatory caspases, human caspase-1, -4, -5, or mouse caspase-1 and -11 ^{1, 2}, leading to membrane-pore formation and lytic cell death driven by gasdermin proteins ^{3, 4}. PANoptosis is another innate immune cell death pathway initiated by innate immune sensors and driven by caspases and RIPKs through PANoptosomes ⁵. PANoptosomes are multi-protein complexes assembled by innate immune sensor(s) in response to pathogens, pathogen- or damage-associated molecular patterns (PAMPs or DAMPs), cytokines or homeostatic changes that drive PANoptosis. This pathway is predominantly activated under physiologically-relevant infectious and inflammatory disease conditions. Necroptosis is an alternate lytic cell death pathway that occurs in the absence of caspase-8, driven by receptor-interacting serine-threonine kinase 3 (RIPK3) ^6–9 and the effector mixed lineage kinase domain-like psuedokinase (MLKL) ^{10, 11}. The kinetics of cell death pathways differ among cell types and between host mammalian species. For example, human and mouse innate immune cells, such as macrophages, express an array of innate immune sensors and cell death proteins and are more susceptible to innate immune lytic cell death compared to non-immune cells such as fibroblasts that express a more limited subset of cell death proteins, suggesting a more prominent role for necroptosis in these cells¹².

The pro-inflammatory effects of innate immune inflammatory cell death are principally driven by cytokines and intracellular components that are actively or passively secreted as a cell dies. During lytic cell death, mammalian cells undergo balloon-like swelling, followed by physical permeabilization and rupture of the plasma membrane. Plasma membrane rupture during innate immune lytic cell death promotes the non-selective release of endogenous components ¹⁴. Liberated cytoplasmic components are no longer concealed from immunological detection by pattern-recognition receptors (PRRs), and are therefore, potentially immunostimulatory and capable of triggering signalling pathways that initiate and amplify the inflammatory response ¹⁵.

PRRs are important innate immune sensor proteins and include the Toll-like receptors (TLRs), NOD-like receptors (NLRs), AIM2-like receptors (ALRs), C-type lectin receptors (CLRs), and retinoic acid-inducible gene-I (RIG-I)-like receptors (RLRs). There also exist expanding families of cytoplasmic nucleic-acid sensors, including cGAS-like receptors (cGLRs) and Zα-domain-like receptors (ZLRs), containing the founding members cGAS and Z-DNA binding protein 1 (ZBP1), respectively ^16–18. These PRRs collectively sense a variety of pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs). PAMPs include any pathogen-derived components, toxins, and virulence factors that drive activation of the immune response and can be released from infected cells during cell death. DAMPs, or alarmins, are used to describe endogenous or newly-synthesized molecules released by damaged or dying cells. Examples of DAMPs include ATP, histones, metabolites, nuclear and mitochondrial DNA, RNA, S100 proteins, and uric acid, among others ¹⁹. The actual or relative contribution of many DAMPs and their cognate PRRs in driving inflammation under physiological conditions is still unclear. For example, ATP induces inflammation in cell culture experiments ^20–22; however, the pro-inflammatory contribution of ATP in mice is only physiologically observed under specific inflammatory conditions, such as thermal injury to the liver ²³, but not in response to exogenously-injected necrotic cells ²⁴. It is possible that not all DAMPs are physiologically pro-inflammatory and that the actual contribution might be obscured by functional redundancy amongst DAMPs released by dying cells and amongst PRRs from the host.

In addition to classical DAMPs, cytokines and chemokines are also secreted by mammalian cells during cell death and can activate immune responses ^{25, 26}, such as the pro-inflammatory cytokines interleukin (IL)-1α, a classical alarmin, and IL-1β and IL-18, secreted by pyroptotic cells to recruit and activate leukocytes ^{27, 28}. Immune sensing of PAMPs and DAMPs and immune responses to cytokines released during cell death therefore have important pathophysiological consequences that shape the course of infection and immunity, and the development of cancer, sepsis, and autoinflammatory diseases ^{29, 30}, providing a rationale to further understand how PAMPs and DAMPs are recognized by the immune system, how cytokine signalling drives pathogenesis, and how therapeutic intervention in these processes might be used in the treatment of diseases (Fig. 1).

Fig. 1. Dying or dead cells release PAMPs, DAMPs, and cytokines that contribute to innate immunity and disease pathogenesis.

Following infection or sterile injury such as trauma, dying or dead cells release a plethora of pathogen-associated molecular patterns (PAMPs), damage-associated molecular patterns (DAMPs), and cytokines that activate pattern-recognition receptors (PRRs) and other receptors such as cytokine receptors in responding immune cells. These cells respond by liberating more PAMPs (in the case of infection), DAMPs, or cytokines that contribute to the manifestation, or in some cases attenuation, of localized and systemic diseases. A subset of cytokines called chemokines induce the recruitment of additional immune cells to the site of injury, amplifying the inflammatory response. Collectively, PAMPs, DAMPs, and cytokines fuel the inflammation circuitry, resulting in the progression and perpetuation of inflammatory diseases.

In this review, we provide an overview on immune-sensing of cell death and the PAMPs, DAMPs, as well as cytokines associated with cell death through the perspective of PRRs. We also consider the cell biology and physiological and pathological consequences of the connections between innate immunity and cell death.

Toll-like receptors

TLRs are a family of transmembrane PRRs located on the cell surface or on the endosomal membrane that recognize a range of PAMPs and DAMPs. Ten human TLRs and twelve mouse TLRs have been identified ³¹. These TLRs activate the transcription factors NF-κB and STAT3 that drive the production of inflammatory cytokines and type I IFNs and stimulate many biological processes (Fig. 2a) ³¹. While the typical response of TLR engagement is pro-inflammatory and pro-survival signalling, TLR signalling can also induce cell death.

Fig. 2. Necrotic cells release DAMPs that activate Toll-like receptors and inflammasomes.

a, TLR2 and TLR4 share the immune sensing of many DAMPs. TLR2 can additionally recognize nucleosomes and versican, whereas TLR4 can additionally recognize DAMPs such as oxidized low-density lipoprotein-immune complexes (OxLDL-IC). TLR3 can sense RNA from necrotic cells, whereas TLR7 might sense single-stranded RNA (ssRNA) and TLR9 might sense single-stranded DNA (ssDNA). TLR9 can also sense DNA complexed with the antimicrobial peptide LL37 or the nuclear protein HMGB1, or DNA-containing immune complexes, and DNA from dying or dead hepatocytes or chemotherapy-killed cancer cells, mitochondrial DNA (mtDNA), and neutrophil-derived DNA. b, Priming via TLRs (also known as Step 1) induces transcription of the genes encoding NLRP3 and pro-IL-1β. Necrotic cells release DAMPs which trigger several common events that activate NLRP3 (also known as Step 2). The most widely accepted mechanism is the efflux of potassium ions (K⁺) from within the cell. In many cases, NLRP3 interacts with the kinase NEK7 in order to assemble an inflammasome complex. AIM2 binds directly to and is activated by double-stranded DNA, which is readily released by dying or dead cells. Both inflammasome sensor proteins form a separate inflammasome complex of 0.5 to 1 micron in diameter comprised of the inflammasome adaptor protein ASC and the cysteine protease caspase-1. Caspase-1 is activated to induce the proteolytic cleavage of pro-IL-1β and pro-IL-18, and the pyroptosis-inducing protein gasdermin D (GSDMD). The bioactive N-terminal fragment of GSDMD assembles into pores on the plasma membrane, through which these pores mediate the secretion of bioactive IL-1β and IL-18 and hundreds of other smaller DAMPs, including IL-1α and galectin-1 that can escape through the diameter of the GSDMD pores. The GSDMD pores eventually drive ninjurin-1 (NINJ1) activation and oligomerization, leading to physical plasma membrane rupture, allowing the release of lactate dehydrogenase (LDH), HMGB1, and other larger DAMPs. Sufficient tears on the plasma membrane eventually lead to pyroptosis. NLRP3 inflammasome specks are released by pyroptotic cells to further perpetuate inflammation in the extracellular environment. These specks are also phagocytosed by neighbouring macrophages to induce a second round of NLRP3 activation.

TLR2 and TLR4

TLR2 and TLR4 are sensors of PAMPs that recognize lipoproteins and LPS, respectively; additionally both TLRs can detect DAMPs (Fig. 2a). In mouse models of atherosclerosis, for example, genetic deletion of TLR2 in low-density lipoprotein receptor-deficient (Ldlr^–/–) mice ³², and genetic deletion of TLR4 in apolipoprotein E-deficient (Apoe^–/–) mice ³³, reduces the development of disease, hinting that these TLRs might recognize DAMPs. Many candidate DAMPs that can activate TLR2 and TLR4 have emerged, but whether these DAMPs originate specifically from dying or dead cells is not known. The glycosaminoglycan hyaluronan (also known as hyaluronic acid), found in the extracellular matrix, pericellular region, and blood, activates both TLR2 and TLR4 in individuals with acute lung injury, triggering the production of inflammatory cytokines and maintaining epithelial cell integrity during acute lung injury in mice ^{34, 35}. Histones released by dying or dead cells can directly interact with TLR2 and TLR4, leading to the production of TNF and IL-6 in mouse dendritic cells and in mice ³⁶.

Although TLR2 and TLR4 recognize many of the same DAMPs, they can also exhibit specificity. In human and mouse CD4⁺ T cells, histones activate TLR2, but not TLR4, and promote the differentiation of CD4⁺ T cells into inflammatory T_H17 cells that can potentially convey pathological effects ³⁷. Nucleosomes which carry histones, DNA, and other proteins are also thought to be sensed by TLR2 but not TLR4 ³⁸. Additionally, TLR2 but not TLR4 can sense the hyaluronan-binding proteoglycan versican, which is secreted by the Lewis lung carcinoma cell line ³⁹. In myeloid cells, activation of TLR2 by versican drives TNF-dependent metastasis in mice injected with the Lewis lung carcinoma cell line ³⁹. TLR4 recognizes free fatty acids to induce pro-inflammatory cytokine production in macrophages, adipocytes, and the liver in mice, leading to insulin resistance, a precursor to diabetes ⁴⁰. TLR4 is also a sensor of calprotectin proteins S100A8 and S100A9 (also known as MRP8 and MRP14, respectively) released by phagocytes, and these proteins are elevated in patients with severe sepsis ⁴¹. TLR4 and its accessory protein MD2 form a signalling-competent receptor complex, which can bind directly to S100A8, with both S100A8 and S100A9 amplifying TLR4-dependent TNF production in response to LPS- and E. coli-induced endotoxemia in mice ⁴¹. The extracellular matrix molecules fibronectin extra domain A and tenascin-C, which are elevated in patients with the skin condition scleroderma, can activate TLR4 in fibroblasts, triggering cutaneous fibrosis and/or delayed fibrosis resolution and healing in mice and human skin organoids ^{42, 43}.

Heme is another DAMP released during cell death and red blood cell (RBC) lysis in many infectious diseases as well as inflammatory and hemolytic diseases. In hereditary spherocytosis and sickle cell anemia, heme drives a feed-forward loop of chronic inflammation and tissue damage to cause disease ^44–46. TLR2 and TLR4 are activated by heme ^{47, 48}, and signal through MyD88 to drive the innate inflammatory response to heme plus PAMP stimulation ^{49, 50}. Taken together, it is clear that activation of these TLRs by DAMPs is largely detrimental to the host.

TLR3, TLR7, and TLR9

The endosomal TLRs, TLR3, TLR7, and TLR9, function in sensing nucleic acids. TLR3 is a sensor of double-stranded RNA (dsRNA) that can be activated by RNA from necrotic neutrophils in macrophages, inducing inflammation and lethality in mice with tissue injury from gut ischemia and polymicrobial sepsis ⁵¹. TLR9, a sensor of CpG DNA, can bind to DNA released by dead cells, specifically DNA complexed with the endogenous antimicrobial peptide LL37 ⁵² or the nuclear protein HMGB1 released by necrotic cells ⁵³. When HMGB1 is complexed with DNA or DNA-containing immune complexes, TLR9 functions in partnership with the HMGB1-binding receptor RAGE to induce the secretion of TNF and IFN-γ in plasmacytoid dendritic cells and the expansion of autoreactive B cells ⁵⁴. The combined effects of these two cell types are thought to contribute to the development of the autoimmune condition systemic lupus erythematosus in mice ⁵⁴. HMGB1 amplifies the ability of CpG DNA to activate TLR9 in mouse macrophages and dendritic cells, possibly mediated by an interaction between HMGB1 and TLR9 ⁵⁵. A further study revealed that DNA binds directly to RAGE, such that RAGE delivers DNA to TLR9 on the endosome, effectively lowering the amount of DNA required to activate TLR9 ⁵⁶.

In a mouse model of liver injury, treatment with the analgesic drug acetaminophen (also known as paracetamol) leads to apoptotic hepatocyte death, freeing DNA to activate TLR9 in sinusoidal endothelial cells, driving inflammasome activation and liver inflammation ⁵⁷. In another sterile inflammatory model using mice lacking the phospholipase D3 (PLD3) and PLD4, which spontaneously develop inflammatory liver damage due to the accumulation of ssDNA and ssRNA, genetic deletion or antibody blockade of TLR9 or TLR7 partially prevents pathological inflammation and spontaneous fatal hemophagocytic lymphohistiocytosis (HLH) ⁵⁸.

TLR9 can also sense circulating mitochondrial DNA suspected to be released by mechanical trauma of cells. Mitochondrial DNA or RBC-bound mitochondrial DNA is increased in patients with trauma and during sepsis and pneumonia in mice and humans ^59–61. Mitochondrial DNA enhances the ability of CpG-DNA to activate TLR9 in human neutrophils, leading to the production of IL-8 ⁵⁹. TLR9 expressed on RBCs can also capture circulating DNA, which mediates clearance of RBCs by macrophage phagocytosis, and induction of local and systemic cytokine production in mice ⁶¹. TLR9 sensing can also be beneficial, when surviving cancer cells recognize DNA from chemotherapy-killed cancer cells and promote inflammation and anti-tumour responses ⁶². Therefore, endogenous nucleic acids can induce an inflammatory response through endosomal TLRs, with evidence demonstrating that the source of the nucleic acids can be from necrotic cells and mitochondria (Fig. 2a).

Inflammasomes and PANoptosomes

Inflammasomes are a family of cytosolic multi-component signalling complexes activated by inflammasome sensors that, in most cases, recruit the adaptor protein called PYCARD (also known as ASC), followed by the inflammatory caspase, caspase-1. Activation of caspase-1 leads to the proteolytic cleavage of the pro-inflammatory cytokines pro-IL-1β and pro-IL-18, and the membrane pore-forming protein gasdermin D which mediates pyroptosis (Fig. 2b) ^{3, 4}.

Multiple members of the PRR-sensing family, including NLRs, ALRs, and others, play a role in regulating or acting as the sensor for the formation of inflammasomes to drive pyroptosis. In addition to inflammasomes, many of the same PRRs may also regulate or assemble unique cell-death complexes called PANoptosomes, which mediate PANoptosis ⁵. In PANoptosis, in addition to gasdermin D-mediated IL-1β and IL-18 release, other gasdermins, including gasdermin E, and other executioners such as MLKL and ninjurin-1 (NINJ1) ¹³, also drive cell death and the release of hundreds of other proteins to perpetuate inflammation ⁶³, including the DAMP galectin-1 ⁶⁴ and the cytokine IL-1α ⁶⁵ from mouse macrophages, and S100A8 or S100A9 from human and mouse neutrophils ^66–69. PANoptosis is implicated in several inflammatory diseases, including Aicardi-Goutieres syndrome ^{70, 71}, cerebral ischemia ⁷², cytokine storm ²⁵, HLH ^{25, 50} and sepsis ¹⁷, as well as in bacterial ^{69, 73}, viral ^{17, 69, 74}, and fungal ⁷⁵ infections. Thus, determining the relative contribution of PAMPs, DAMPs, and cytokines in licensing PANoptosis is an active and emerging area of research. A holistic understanding of how PANoptosis can be blocked in systemic inflammatory diseases such as COVID-19, or activated during tumour development, is expected to provide therapeutic advances in the treatment of many chronic health conditions.

NLRP3 inflammasome and PANoptosome

NLRP3 is the most versatile inflammasome sensor protein identified to date, and it is often called the global sensor of PAMPs and DAMPs (Fig. 2b) ⁷⁶. Oligomeric inflammasome structures called ‘specks’ of 0.5 to 1 micron in diameter form in the cytoplasm of an inflammasome-activated cell; these specks are readily visualised under confocal microscopy and are comprised of ASC arranged in filaments of 8 to 15 nm in diameter ^{77, 78}. NLRP3 specks are also released from the cell corpse and activate caspase-1 outside the cell ^{79, 80}. These specks have been used as serum biomarkers for a group of autoinflammatory diseases known as cryopyrin-associated periodic syndromes or autoimmune diseases in humans, and are seen engulfed by neighbouring macrophages where the specks trigger a second round of NLRP3 inflammasome activation in the recipient cell ^{79, 80}. Therefore, NLRP3 specks are DAMPs and powerful perpetuators of inflammasome-mediated inflammation.

NLRP3 is highly responsive to lytic dead cell corpses ⁸¹ and DAMPs associated with cell death, notably ATP ^20–23, β-amyloids ^{82, 83}, biglycan ⁸⁴, endogenous crystals ^{85, 86}, free fatty acids ⁸⁷, histones ⁸⁸, mitochondrial DNA ^89–91, RNA ^{92, 93}, and S100A8 or S100A9 (Fig. 2b) ⁹⁴. NLRP3 does not directly bind to DAMPs, in general, and the responsiveness of NLRP3 to these DAMPs has been demonstrated largely under in vitro conditions. In a sterile inflammatory model in mice, NLRP3 has been reported to sense ATP released from necrotic neutrophils caused by thermal injury to the liver, promoting localized inflammation and adherence of circulating neutrophils to the liver sinusoids ²³. ATP released by chemotherapy-induced dying cells also activates the NLRP3 inflammasome–IL-1β pathway in dendritic cells and drives anti-tumour cytotoxic CD8⁺ T cell responses in mice ⁹⁵. However, there is no evidence that ATP drives inflammation in mice undergoing drug-induced liver injury or injection of necrotic cells ²⁴, suggesting that either the concentration of ATP necessary to trigger inflammation is not achieved or that NLRP3 has no role in these experimental models. Additionally, in response to NLRP3 triggers such as ATP and bacterial PAMPs, it is likely that NLRP3-dependent but caspase-1-, gasdermin-D-independent lytic cell death can occur through PANoptosis. In this context, the activation of caspases and executioners associated with PANoptosis would occur, and a PANoptosome complex containing ASC, caspase-8 and RIPK3 would form, positioning NLRP3 as a PANoptosome sensor activating PANoptosis. Furthermore, beyond its crucial roles in inflammasome-dependent pyroptosis and as a PANoptosome sensor, NLRP3 is also integral to several other PANoptosomes (Fig. 3), as discussed below.

Fig. 3. PAMPs, DAMPs, and cytokines activate PANoptosomes.

Heme in combination with PAMPs or TNF (not shown) triggers the activation of the NLRC5-PANoptosome, which also contains NLRP12. PAMPs induce the activation of the transcription factor IRF1 to upregulate the NLRC5 and NLRP12 proteins. Heme then induces PANoptosome formation by recruiting NLRP3, ASC, caspase-1 (CASP1), caspase-8 (CASP8), and RIPK3, leading to PANoptosis. The cytokine combination of TNF and interferon (IFN)-γ triggers a second PANoptosome. TNF and IFN-γ upregulate the production of nitric oxide (NO) via JAK, STAT1, IRF1, and inducible nitric oxide synthase (iNOS), triggering a cytokine-induced PANoptosome requiring CASP8, RIPK1, and RIPK3 to drive PANoptosis. The ZBP1-PANoptosome functions in sensing cytokines and DAMPs, as well as infections. In response to the combination of IFN and a nuclear export inhibitor (NEI), IFN upregulates the expression of ZBP1 and ADAR1, and NEI restricts ADAR1 to the nucleus, allowing ZBP1 to freely sense accumulated dsRNA and assemble a PANoptosome. The ZBP1-PANoptosome contains NLRP3, ASC, CASP1, CASP8, caspase-6 (CASP6), and RIPK3. The ZBP1-PANoptosome can also be activated by Influenza A virus infection (not shown). The RIPK1-PANoptosome senses the combination of the PAMP LPS and DAMPs in the form of TAK1 inhibition (TAK1i). TAK1i unleashes RIPK1, allowing LPS or other PAMPs or DAMPs to induce the assembly of the RIPK1-PANoptosome comprised of RIPK1, NLRP3, ASC, CASP1, CASP8, and RIPK3. The RIPK1-PANoptosome can also be activated by Yersinia infection (not shown). In response to PAMPs and DAMPs, such as those released during Francisella or herpes simplex virus 1 (HSV1) infection (not shown), dsDNA derived from these pathogens serves as a PAMP to activate the AIM2-PANoptosome. Other PAMPs carried by these pathogens and possibly DAMPs resulting from these infection events are likely required to drive the assembly of the AIM2-PANoptosome because transfection of dsDNA alone cannot recapitulate this response. The AIM2-PANoptosome comprises AIM2, Pyrin, ZBP1, ASC, CASP1, CASP8, RIPK1, and RIPK3. PANoptosomes activate or induce the proteolytic cleavage of many substrates, including gasdermin proteins, cytokines, and caspases, resulting in inflammatory lytic cell death and the secretion of cytokines and DAMPs. The symbols in the legend denote the colors and shapes used to depict each molecule; the number of shapes in a molecule reflects the general domain structure and Is not intended to suggest the number of molecules present.

NLRP3 also senses needle-shaped monosodium urate crystals ⁸⁵, formed by uric acid in joint tissues leading to gout. The ability of the NLRP3 inflammasome to drive the secretion of IL-1β contributes to this pathological manifestation. In mice, genetic deletion of the gene encoding IL-1R reduces neutrophil influx in monosodium-urate-induced peritonitis ⁸⁵. In humans, anti-IL-1 biologic therapies are highly effective in the treatment of gout ⁹⁶. Monosodium-urate-induced lytic cell death in macrophages is independent of NLRP3 and gasdermin D ^{97, 98}, implying that other DAMPs would still be released even if NLRP3 is pharmacologically inhibited. The effectiveness of NLRP3 blockade in the treatment of gout in humans is not yet clear, but a phase 2b trial has shown efficacy of an oral NLRP3 inhibitor Dapansutrile in reducing joint pain in patients with gout flares ⁹⁹.

NLRP3 also has an emerging role in sensing and responding to DAMPs in platelets. S100A8 and S100A9 are neutrophilic DAMPs comprising 45% of the total cytosolic protein content in human neutrophils ¹⁰⁰, and their levels are elevated in patients with severe sepsis ¹⁰¹. S100A8 or S100A9, released by human and mouse neutrophils via gasdermin D pores, can trigger the assembly of the NLRP3 inflammasome in mouse platelets ⁹⁴. These platelets then undergo pyroptosis and amplify the inflammatory feedback loop by liberating oxidized mitochondrial DNA and triggering more S100A8 or S100A9 release from neutrophils ⁹⁴.

AIM2 inflammasome and PANoptosome

AIM2 is a cytoplasmic inflammasome sensor that binds to double-stranded DNA and has been well-characterized in the context of infection ^{69, 102–104}, inflammatory skin and intestinal diseases ^105–107 and cancer ^{108, 109}. In infection, the exact source of DNA leading to AIM2 activation is presumed to be derived from microbes or the microbiome; however, identification of a role for AIM2 in the host defense against the RNA virus, Influenza A Virus (IAV), which does not carry DNA, suggests host DNA might be involved ¹¹⁰. Indeed, IAV infection causes the extracellular release of DNA into the lung fluid of mice ^{110, 111}, and the ion channel activity of the IAV M2 protein and the mitochondrial localization of the IAV PB1-F2 protein induce the release of mitochondrial DNA and oxidized DNA into the cytoplasm of macrophages and extracellularly from macrophages ¹¹².

Further evidence that AIM2 can sense host DNA from dying cells came from a study showing that self-DNA released into the gastrointestinal tract of mice, due to treatment with the cytotoxic drug irinotecan, is captured by exosomes and delivered into intestinal cells to activate AIM2 ¹¹³. In another mouse model, DNA is taken up by leukocytes in the injured kidneys caused by ureteral obstruction ¹¹⁴. Sensing of DNA by AIM2 is supported by evidence using mice lacking AIM2, which have abrogated irinotecan-induced intestinal inflammation and diarrhoea ¹¹³, reduced influenza-induced lung inflammation, morbidity and mortality ¹¹¹, and attenuated renal injury, fibrosis, and inflammation ¹¹⁴.

In addition to forming an inflammasome, AIM2 also assembles a PANoptosome in response to physiologically relevant triggers such as HSV1 and Francisella tularensis ⁶⁹ (Fig. 3). In this complex, ZBP1 and Pyrin are recruited in conjunction with ASC and caspase-1, as well as caspase-8, RIPK1 and RIPK3 ⁶⁹. Activation of PANoptosis requires infection with the live pathogen, which releases PAMPs and DAMPs together; in contrast, treatment with the bacterial PAMP mimic poly(dA:dT) alone induces inflammasome activation rather than PANoptosome formation ⁶⁹. These findings highlight key distinctions between PANoptosis and pyroptosis in physiological settings in response to dying cells and liberated cytokines and DAMPs.

ZBP1-PANoptosome

The ZBP1-PANoptosome was the first PANoptosome identified, and it was discovered to form in response to IAV infection ¹⁷ (Fig. 3). It also plays a key role in the pathophysiology of IFN therapy during coronavirus infection ¹¹⁵, and the ZBP1-PANoptosome was also found to form upon combined stimulation with IFN and a nuclear export inhibitor (NEI) ¹¹⁶. In this case, NEIs, such as KPT-330, prevent the export of ADAR1 from the nucleus. Given that IFN upregulates the expression of ZBP1 and its negative regulator ADAR1, evidence suggests that the NEI restricts the activity of ADAR1 and indirectly promotes the assembly of the ZBP1-PANoptosome, which includes NLRP3, ASC, caspase-1, caspase-6, caspase-8, and RIPK3 ^{5, 74, 116}. This combination of IFN and KPT-330 promotes ZBP1-mediated PANoptosis and attenuates tumour growth in vivo in mice ¹¹⁶.

NLRC5- and NLRP12-PANoptosome

NLRC5 and NLRP12 were recently discovered to be key sensors in a PANoptosome complex (Fig. 3). NLRP12 was identified as a sensor of specific combinations of PAMPs and DAMPs, such as heme and LPS, Pam3, or TNF ⁴⁹. Subsequently, the NLRC5-PANoptosome was identified to activate PANoptosis in response to NAD⁺ depletion, a common occurrence during disease ⁵⁰. In this complex, there is coordinated activity among multiple NLRs, including NLRC5, NLRP12 and NLRP3, suggesting NLR networks similar to those seen in plants also form in mammals ⁵⁰. NLRC5- and NLRP12-dependent PANoptosis depends on the transcription factor IRF1, which upregulates NLRC5 and NLRP12 in response to heme and PAMPs or TNF ^{49, 50, 117}. Then, NLRP3, ASC, caspase-1, caspase-8 and RIPK3 are recruited to the PANoptosome, leading to PANoptosis. Nlrp12^–/– and Nlrc5^–/– mice are resistant to acute kidney injury and lethality during drug-induced hemolysis combined with LPS exposure ^{49, 50}. In addition, NLRC5 deficiency confers protection in colitis and HLH models ⁵⁰. Thus, inhibition of NLRC5 or NLRP12 could be therapeutically beneficial in the context of highly diverse stimuli, including pathogen-mediated hemolysis or other red blood cell disorders.

RIPK1-PANoptosome

The RIPK1-PANoptosome was initially shown to form downstream of LPS signalling in TAK1-deficient macrophages ¹¹⁸ (Fig. 3). This LPS-mediated cell death requires RIPK1 but not its kinase activity, implicating its scaffolding function in the formation of a multiprotein complex ¹¹⁸. Indeed, in the absence of TAK1, RIPK1 forms a complex with NLRP3, ASC, caspase-1, caspase-8 and RIPK3 to mediate PANoptosis ⁷³. While TAK1 promotes cell survival in response to stimuli such as LPS or TNF, TAK1 is also a cell-intrinsic negative regulator of PANoptosis via inhibition of RIPK1 ^{118, 119}. Inhibition of TAK1 in vivo induces a sepsis-like state ^{118, 119}, which indicates that constitutively active PANoptotic processes can generate unregulated and potentially lethal systemic inflammation.

Cytokine-induced PANoptosome

TNF and IFN-γ are a specific cytokine combination that is frequently elevated in inflammatory diseases ²⁵. TNF and IFN-γ upregulate the production of nitric oxide in macrophages via JAK, STAT1 and IRF1 signalling, inducing the formation of a PANoptosome which is responsible for cell death and lethality during SARS-CoV-2 infection, sepsis, HLH and cytokine shock ^{25, 26, 115} (Fig. 3). However, the specific sensor required to promote the formation of this PANoptosome remains unknown. Nevertheless, this cytokine-induced PANoptosome has potential therapeutic implications for the treatment of cancer ^{116, 120}, as TNF and IFN-γ trigger PANoptotic cell death in melanoma and leukaemia as well as colon and lung tumour cells ¹²¹.

C-type lectin receptors

C-type lectin receptors (CLRs) are a large class of cell-surface or secreted PRRs carrying one or more C-type lectin recognition domains ¹²². They primarily recognize carbohydrate structures and provide host defense against pathogens, but CLRs also bind to and phagocytose dead cells, sense DAMPs, and trigger or inhibit the production of inflammatory cytokines, antigen presentation and T cell responses (Fig. 4). Certain CLRs bind to endogenous ligands expressed on live mammalian cells ^{123, 124}, but other CLRs sense DAMPs that are liberated from dying or dead cells.

Fig. 4. Necrotic cells release DAMPs that activate C-type lectin receptors.

MINCLE binds to spliceosome-associated protein 130 (SAP130) and associates with the signalling receptor FcRγ and signals via SYK and CARD9 to drive the production of TNF and chemokines, inducing neutrophil recruitment. MINCLE activation by SAP130 also promote pro-tumorigenic immunosuppression in pancreatic cancer. MINCLE also recognizes β-glucosylceramide and activates CARD9, leading to cytokine and chemokine production, increased co-stimulation, and activation of adaptive immunity. Clec1a binds to the histidine-rich glycoprotein, mediating phagocytosis of bacteria. Clec1a also binds to TRIM21 and inhibits cross-presentation, promoting tumour growth. Clec2d is a sensor of DNA-free histones (not shown) and DNA-bound histones. Clec2d delivers DNA to TLR9 within the endosome to trigger cytokine production, enhancing liver injury and lethality. Clec9a binds to actin filaments and F-actin, activating SYK and promoting the delivery of the necrotic cell debris to the recycling endosome and cross presentation to CD8⁺ T cells. Further, Clec9a activates the tyrosine phosphatase SHP-1 and inhibits the production of the chemokine CXCL2, thereby preventing neutrophil recruitment, necrotizing pancreatitis, and pathology during infection with Candida albicans. Clec12a senses monosodium urate crystals from dead cells and inhibits SYK-dependent cytokine and chemokine production, reducing neutrophil influx. Clec12a also mediates the production of type I interferons (IFNs) by enhancing RIG-I signalling. Dectin-1 binds to N-glycan structures found on the cell-surface of live tumour cells and presumably also dead tumour cells, and via the membrane-associated protein MS4A4A, activates IRF5 and induces the cell-surface expression of IRF3-dependent NK-activating molecule (INAM) on dendritic cells. INAM mediates enhanced contact between dendritic cells and NK cells, promoting NK-cell anti-tumour immunity. Dectin-1 also senses galectin-9 found on the cell-surface of tumour cells and promotes immunosuppression and the growth of pancreatic ductal adenocarcinoma.

MINCLE

Human and mouse MINCLE (also known as Clec4e) are expressed on macrophages and dendritic cells, and recognize pathogenic fungal Malassezia and bacterial Mycobacterium species, as well as dead cells and DAMPs ¹²⁵. MINCLE binds to the small nuclear ribonucleoprotein SAP130, and potentially to related proteins SAP49, SAP145, and SAP155 ¹²⁵. In response to SAP130, MINCLE associates with the immunoreceptor tyrosine-based activation motifs (ITAM)-containing adaptor FcRγ and signals via the adaptor proteins SYK and CARD9 to drive the production of TNF and chemokine CXCL2 (also known as MIP-2) in macrophages, leading to the recruitment of neutrophils to the site of injury in mice ¹²⁵.

Additionally, MINCLE can recognize the metabolite glycolipid β-glucosylceramide released from damaged mammalian cells ¹²⁶. β-glucosylceramide is normally found in the endoplasmic reticulum or Golgi and is unlikely to be exposed to cell-surface-associated MINCLE during homeostasis. In response to LPS stimulation or bacterial infection, increased production of β-glucosylceramide can occur ^{127, 128}, and subsequent cell death offers a physiological scenario by which β-glucosylceramide can be liberated and interact with MINCLE on neighbouring cells. In response to β-glucosylceramide, MINCLE activates CARD9, driving the secretion of TNF and MIP-2, and inducing the expression of costimulatory molecules CD40, CD80 and CD86 and MHC class II on mouse dendritic cells ¹²⁶.

Inhibition of MINCLE using a neutralizing DNA aptamer or an antibody reduces the severity of chemically-induced inflammatory bowel disease in mice ¹²⁹, suggesting that targeting MINCLE in inflammatory diseases might be beneficial. However, comparison between wild-type and Mincle^–/– mice identified no role for MINCLE in two other experimental models of cell-death-induced inflammation: injection with necrotic EL4 tumour cells resulting in a neutrophilic inflammatory response and treatment with acetaminophen causing liver cell necrosis ²⁴. In the context of cancer development, pancreatic ductal adenocarcinoma cells expressing recombinant SAP130 injected into wild-type mice grew faster than in mice lacking MINCLE ¹³⁰, indicating that MINCLE activation by SAP130 is detrimental and promotes oncogenesis in pancreatic cancer.

Clec1a

Clec1a is expressed on human and mouse dendritic cells, macrophages, monocytes, neutrophils and endothelial cells ^131–133. Clec1a binds directly to the plasma protein histidine-rich glycoprotein ¹³⁴, and the E3 ubiquitin ligase TRIM21 expressed on the surface of necrotic cells ¹³⁵. Clec1a has different roles depending on the disease context and cell types. In response to infection in the presence of histidine-rich glycoprotein, Clec1a mediates phagocytosis of bacteria in human neutrophils ¹³⁶. In response to human or mouse cells killed by UV or X-ray radiation, Clec1a binds to TRIM21 expressed on dead cells and inhibits cross-presentation by conventional type 1 dendritic cells to CD8⁺ T cells, promoting tumour growth in mice ¹³⁵. Blockade of Clec1a using a Clec1a fusion protein prolongs the survival of mice injected with MC38 adenocarcinoma ¹³⁵, indicating a disease-promoting role of Clec1a.

Clec2d

Clec2d is found on the plasma membrane and endosomes of mouse macrophages and can sense histones ¹³⁷. Histones are released by activated neutrophils as part of neutrophil extracellular traps, or they can be found in either a DNA-bound form in a nucleosome or in an octameric protein form free of DNA ¹³⁸. Clec2d senses both DNA-free histones and DNA-bound histones, including histone variants H1, H2A, H2B, H3, and H4, by recognizing the N-terminal tail regions of all histones and the C-terminal tail region of H1 ¹³⁷. Unlike most CLRs, Cled2d cannot induce signalling on its own, because it lacks an ITAM ¹³⁷. On activation by DNA-bound histones, Clec2d delivers DNA to TLR9 within the endosome to trigger cytokine production, which enhances the liver injury and lethality induced by acetaminophen overdose in mice ¹³⁷.

Clec9a

Clec9a (also known as DNGR-1) is expressed on mouse CD8α⁺ dendritic cells and human dendritic cells, and it recognizes dead mammalian cells undergoing secondary necrosis following UV treatment ¹³⁹. Human and mouse Clec9a bind directly to actin filaments secreted by necrotic cells ^{140, 141}, with modest affinity for F-actin ¹⁴². The binding between Clec9a and F-actin, however, is strengthened by multivalent interactions defined by the ligand-binding domain of Clec9a interacting with three actin subunits at the interface of two actin protofilaments ¹⁴².

In addition, the F-actin-associated motor protein, myosin II, also exposed in necrotic cells, promotes the binding between Clec9a and F-actin ¹⁴³. This pathway of dead-cell recognition activates SYK signalling ¹⁴⁰, but does not result in cytokine production in CD8α⁺ dendritic cells ¹³⁹. Instead, Clec9a promotes the delivery of necrotic cell debris to the recycling endosome, leading to phagosomal rupture where the antigens are cross-presented to CD8⁺ T cells ^{144, 145}. Further, Clec9a activates the tyrosine phosphatase SHP-1 and inhibits the production of the chemokine CXCL2 by conventional type I dendritic cells during systemic infection with the fungus Candida albicans ¹⁴⁶. Sensing of early damage by Clec9a, presumably via actin release caused by C. albicans infection, reduces CXCL2-mediated neutrophil recruitment, necrotizing pancreatitis, and pathology ¹⁴⁶.

Clec12a

Clec12a can directly bind dead mammalian cells leading to crosslinking between Clec12a proteins and inhibition of SYK ¹⁴⁷. SYK is activated by cell-surface receptors carrying ITAM or ITAM-like motif; however, Clec12a carries an immunoreceptor tyrosine-based inhibition motif (ITIM), which indicates that the functional activity of Clec12a is anti-inflammatory. Clec12a senses the monosodium urate crystals of dead cells in humans and mice ¹⁴⁷. Mice lacking Clec12a are hypersusceptible to neutrophil influx following injection of monosodium urate crystals or X-ray irradiation to induce thymocyte killing ¹⁴⁷. In response to X-ray irradiation in mice, Clec12a can also drive type I interferon (IFN) responses by enhancing RNA-sensing RIG-I signalling pathways ¹⁴⁸. The physiological RIG-I ligands in this case are not known, nor are the potentially broader effects of Clec12a in other nucleic-acid-sensing pathways.

cGAS-like receptors

cGAS is a cytoplasmic PRR of the cGLR family that recognizes DNA ¹⁴⁹. Following activation, cGAS generates the second messenger cyclic guanosine monophosphate-adenosine monophosphate (cGAMP), which binds to and activates the adaptor protein STING ¹⁵⁰. STING recruits and activates the kinase TBK1 and promotes the activation of the transcription factor IRF3 and type I IFN production. cGAS directly senses DNA from infectious agents and mislocalized self-DNA within the same cell (Fig. 5).

Fig. 5. Necrotic cells release DAMPs that activate the cytosolic DNA-sensing and RNA-sensing pathways.

In phagocytic cells, cGAS recognizes DNA from dying or dead cells. DNA that results from DNase II deficiency (not shown), apoptotic cells, erythroid precursor cells or mitochondria accumulates within the cytoplasm of phagocytes. cGAS catalyzes cGAMP production for STING activation to trigger the production of inflammatory cytokines via NF-κB and type I interferons (IFNs) via IRF3/7, resulting in embryonic lethality and inflammatory arthritis. DNA from dying tumour cells and tumour-derived mitochondrial DNA taken up by dendritic cells is also sensed by cGAS, leading to IFN-β production and cross-presentation to CD8⁺ T cells, inducing anti-tumour immunity. Due to a loss of the mitochondrial RNA helicase SUV3 and polynucleotide phosphorylase PNPase (not shown), mitochondrial double-stranded RNA (dsRNA) from inside the cell or from dying or dead cells accumulates and escapes into the cytoplasm to activate MDA5, and via MAVS, triggers the production of inflammatory cytokines and type I IFNs. RIG-I can sense cytoplasmic RNA (not shown), but there is currently no evidence suggesting that RIG-I can sense RNA directly from dead cells. Defects in adenosine-to-inosine RNA editing induced by deficiencies or mutations in ADAR1 also lead to accumulation of endogenous RNA that is sensed by MDA5.

The recent model proposed for DNase II deficiency-mediated disease provides the strongest evidence that DNA from dying neighbouring cells also activates cGAS. DNase II deficiency in mice leads to the lysosomal accumulation of undigested DNA, driving TNF and IFN-β responses, embryonic lethality and autoimmune diseases ^{151, 152}. The disease-initiating DNA has been reported to come from apoptotic cells ^{153, 154}, erythroid precursor cells ¹⁵⁵, or within the nucleus of DNase II-deficient cells ¹⁵⁶. Self-DNA can also accumulate within the cytoplasm of cells lacking the DNA nuclease TREX1, but the evidence that the self-DNA comes from another dying cell is limited. Earlier studies showed that genetic deletion of TLR3, TLR9 or MyD88 and TRIF in DNase II^−/− mice does not prevent embryonic lethality and disease ¹⁵⁴, but deletion of the type I IFN receptor 1, IFNAR1, prevents embryonic lethality but not polyarthritis ^{151, 152}, suggesting the involvement of other DNA-sensing PRRs and additional cytokines. Later studies demonstrated that deletion of either STING or cGAS in DNase II^−/− mice leads to reduced cytokine production and prevents both embryonic lethality and inflammatory arthritis ^{157, 158}, corroborating with in vitro experiments showing that necrotic cells trigger STING-dependent production of IFN-β in mouse macrophages ¹⁵⁷. Consistent with the role of cGAS in catalysing cGAMP production for STING activation, cGAMP does not accumulate in the fetal liver of DNase II^−/− mice lacking cGAS ¹⁵⁸.

The cGAS-STING pathway can also activate NF-κB, leading to the production of TNF and IL-6. This additional signalling capacity may in part explain why DNase II^−/−Ifnar1^−/− mice ^{151, 152} or DNase II^−/− mice carrying a mutant form of STING (called STING^S365A/S365A), which cannot induce type I IFN signalling but retains the ability to induce NF-κB activation ¹⁵⁹, still develop polyarthritis. Neutralizing antibodies against TNF or the IL-6 receptor alleviate polyarthritis in DNase II^−/−STING^S365A/S365A mice ¹⁵⁹. Additional genetic deletion of STING, AIM2 or the TLR-chaperon protein UNC93B1 in DNase II^−/−Ifnar1^−/− mice also reduces the severity of polyarthritis ^{160, 161}. Unravelling the complexity of PRRs has allowed the development and fine-tuning of therapeutic targeting for autoinflammatory disease instigated by self-DNA.

cGAS and STING are also implicated in sensing DNA from dying cells in tumorigenesis ^162–165. Labelled DNA from tumour cells injected into mice are found inside dendritic cells ¹⁶⁵, with cGAS, STING and IRF3 driving the induction of IFN-β in dendritic cells exposed to irradiated tumour cells or dying cells ^{162, 163}. STING-dependent IFN-β production and cross-presentation are both inhibited by physical separation between mouse dendritic cells and irradiated tumour cells or by an actin polymerization inhibitor ¹⁶², supporting a model where phagocytosis of dead cells by dendritic cells is essential. This process drives dendritic-cell cross-presentation of tumour-associated antigens to CD8⁺ T cells, inducing anti-tumour immunity in mice ^{162, 165, 166}. In response to antibodies blocking the immunoglobulin-like domain–containing molecule CD47, tumour-derived mitochondrial DNA is phagocytosed by dendritic cells to activate the cGAS-STING pathway that promotes tumour rejection in mice ¹⁶⁷. This occurs through antibody-mediated interference of binding between CD47 on tumour cells and signal-regulatory protein alpha (SIRPα) on dendritic cells, activating a NOX2-dependent delay in acidification of the phagosome and degradation of engulfed mitochondrial DNA. This delay maintains the necessary DNA concentration for cGAS sensing ¹⁶⁷. How genomic or mitochondrial DNA from tumour cells are released into the cytoplasm of recipient cells to activate the cGAS-STING pathway is still unknown. A different model suggests that constitutively-active cGAS in tumour cells catalyses cGAMP production, with cGAMP activating STING-dependent IFN-β production in neighbouring non-tumour cells, leading to anti-tumour responses by natural killer cells ¹⁶⁸. However, a direct transfer of cGAMP between tumour cells and non-tumour cells licensing STING activation and immune responses in vivo remains to be tested, noting that cGAMP transfer in vivo is largely limited by the degradation of extracellular cGAMP by Ectodomain phosphatase/phosphodiesterase-1 (ENPP1) ¹⁶⁹.

RIG-I-like receptors

RIG-I-like receptors (RLRs) are found in the cytoplasm, with the family member RIG-I sensing shorter dsRNA, and another family member MDA5 sensing longer dsRNA ¹⁷⁰. Both receptors trigger the production of type I IFNs and pro-inflammatory cytokines via the adaptor protein MAVS. While RLRs are most associated with immune responses to viral pathogens, they also sense DAMPs in the form of endogenous RNA (Fig. 5). Mitochondrial dsRNA, which is normally rapidly degraded, can accumulate due to a loss of the mitochondrial RNA helicase SUV3 and polynucleotide phosphorylase PNPase ¹⁷¹. Mitochondrial dsRNA escapes into the cytoplasm to activate MDA5 and type I IFN production, which may contribute to the clinical manifestations of humans with mutations in PNPase ¹⁷¹. Additionally, defects in adenosine-to-inosine RNA editing induced by deficiencies or mutations in ADAR1 can also lead to the accumulation of endogenous RNA that can be sensed by MDA5 in humans and mice ^172–174, and the loss of ADAR1 in some tumours sensitizes them to immunogenic cell death ^{116, 175, 176}. While these studies point to the presence of mislocalized or accumulating self-RNA in triggering RNA-sensing mechanisms within the same cell, whether extracellular RNA liberated from dead cells can activate RLRs in neighbouring cells has not been reported. RNA can be released passively from dying or dead cells or via extracellular vesicles, but this RNA is degraded rapidly by extracellular RNAses ¹⁷⁷. However, with functional defects in extracellular RNAses, it may be possible that extracellular RNA would accumulate to a quantity that can activate RLRs.

Conclusions and outlook

While inflammatory cell death is a key outcome of innate immune activation, it is a driving and perpetuating force of PRR sensing that results in chronic inflammation and pathology. Unravelling the connections between innate immune cell death and pathological inflammation is expected to be crucial for understanding fundamental cell biology that enables the identification of therapeutic strategies. The biological effects of DAMPs derived from dying or dead cells and their sensing by PRRs are largely characterized by the amplification of inflammation and immune cell recruitment, which can drive detrimental tissue damage, delayed healing, cytokine storm, metastasis and insulin resistance. These effects contribute to the development of cancer, diabetes, sepsis and other inflammatory conditions. However, some protective effects of PRR-sensing of cell death have been documented, including lowering the magnitude of inflammatory responses, induction of cross-presentation, and anti-tumour immunity. Therefore, PRRs can sense the same cell corpse, DAMPs and cytokines but generate converging or opposing outcomes. How PRRs precisely coordinate with one another holistically to produce a coherent cellular response to the same dying or dead cell has remained unclear. The answers may lie in spatially-distinct multi-protein complexes whose organization and execution of the various activities of PRRs, caspases, and RIPKs within a cell is orchestrated at a singular signalling hub, such as the PANoptosomes.

From a therapeutic perspective, strategies used to block PRRs that sense PAMPs, DAMPs and cytokines released during cell death can prevent inflammation and pathology. Inhibiting PANoptosome activation using anti-TNF and anti-IFN-γ antibodies prevents mortality during SARS-CoV-2 infection, sepsis, HLH, and cytokine shock ²⁵. Furthermore, blocking activation of the NLRP12-PANoptosome protects against hemolytic disease models ⁴⁹, and inhibiting the NLRC5-PANoptosome prevents PANoptosis activation in response to NAD⁺ depletion and reduces morbidity and mortality during hemolytic disease, colitis, and HLH models ⁵⁰. Blocking DAMPs such as HMGB1 has also shown promising effects in reducing pathology and mortality during pneumococcal meningitis ¹⁷⁸, stroke ¹⁷⁹, Alzheimer’s disease ¹⁸⁰, and several other CNS and inflammatory diseases in animal models. For therapeutic activation of PRRs to harness beneficial effects, activation of the ZBP1-PANoptosome using IFN and nuclear export inhibitors could be considered in anti-tumour therapies ¹¹⁶, and additional strategies to target PANoptosis through other PRRs should also be considered. Continued investigation to identify critical PRR-sensing pathways in responding to cell death is expected to pave the way for new diagnostic and prognostic biomarkers and therapeutic advancement for the treatment of diseases.