Pyroptosis in Antiviral Immunity

Abstract

Pyroptosis is a form of lytic, programmed cell death that functions as an innate immune effector mechanism to facilitate host defense against pathogenic microorganisms, including viruses. This type of proinflammatory cell death is orchestrated by proteolytic activation of human or mouse caspase-1, mouse caspase-11 and human caspase-4 as well as caspase-5 in response to infectious and inflammatory stimuli. Induction of pyroptosis requires either a canonical inflammasome responsible for caspase-1 activation or a noncanonical complex composed of caspase-11 in mice or caspase-4 or caspase-5 in humans. Recent studies have identified the pore-forming protein gasdermin D, a substrate of these inflammatory caspases, as an executioner of pyroptosis. The membrane pores formed by gasdermin D facilitate release of proinflammatory cytokines IL-1β and IL-18 and consequent biologic effects of these cytokines together with other released components. Pyroptosis, like other forms of programmed cell death, helps eliminate infected cells and thereby restricts the replicative niche, undermining survival and proliferation of intracellular pathogens. This includes viruses as well as bacteria where ample evidence supports a critical role for inflammasome effector functions and cell death in host defense. Viruses have evolved their own mechanisms to modulate inflammasome signaling and pyroptosis. Here, we review the current literature regarding the role of pyroptosis in antiviral immune responses.

1. Introduction

Pyroptosis, a form of programmed cell death, has roots that reflect a proinflammatory nature of the pathway. The name pyroptosis, first used to describe a lytic form of caspase-1-dependent cell death, is based on the Greek term ‘pyro’ meaning fire or fever and ‘ptosis’ meaning falling or dropping, distinguishes this pathway from apoptosis and necrosis (Cookson and Brennan 2001). The word ‘inflammation’ itself originates from a Latin term meaning ‘a setting on fire’ and aptly describes cardinal features such as calor (Dondelinger et al. 2015), rubor (redness) and dolor (pain) (Basil and Levy 2016). Inflammatory caspases (human and mouse caspase-1, mouse caspase-11 and human caspase-4 and caspase-5) function as effector caspases to initiate pyroptotic cell death (Jorgensen and Miao 2015; Man et al. 2017). The contribution of these inflammatory caspases distinguishes pyroptosis from either apoptosis or necroptosis, all mechanistically distinct forms of programmed cell death (Man et al. 2017).

Early reports failed to distinguish pyroptosis from apoptosis (Zychlinsky et al. 1992) despite evidence that cell death affecting macrophages in response to infection with Shigella flexineri and Salmonella typhimurium resulted in leakage (Zychlinsky et al. 1992; Chen et al. 1996; Monack et al. 1996). Cell death during infection with these pathogens was later found to be dependent on caspase-1 and/or caspase-11 and to involve morphological changes different from apoptosis (Hilbi et al. 1998; Hersh et al. 1999; Brennan and Cookson 2000; Watson et al. 2000). The initial focus on caspase-1 as a pro-pyroptotic caspase was revised when caspase-1-deficient cells were shown to also lack caspase-11, a closely linked gene that is responsible for pyroptosis induced by the Gram negative bacterial cell wall component, lipopolysacharride (LPS) (Kayagaki et al. 2011). Caspase-11 is now recognized as an interferon-inducable cytosolic sensor required for pyroptosis in response to LPS from Gram negative bacteria (Kayagaki et al. 2013). In contrast, viral infections may trigger caspase-1-dependent pyroptosis. Pyroptosis differs from the inflammatory and lytic necrosis, which occurs due to physical damage, independently of apoptotic caspases (Boise and Collins 2001). Pyroptotic cells are characterized by swelling followed by rupture of the plasma membrane resulting in release of cytoplasmic contents including inflammatory cytokines and alarmins. Dying cells are positive for Annexin V and TUNEL staining, but lack apoptotic markers such as nucleosome laddering and chromatin condensation (Fink and Cookson 2006), features that have been summarized more recently (Jorgensen and Miao 2015). The proinflammatory response initiated by pyroptosis can occur independently of the release of bioactive IL-1β and IL-18 (Man et al. 2017; Monack et al. 2001) due to the release of other immunostimulatory cellular contents (Davis et al. 2019). Unlike the apoptotic signaling cascade, the activation of caspase-1 is mediated by a multiprotein cytosolic complex called an inflammasome (Martinon et al. 2002) and the activation of mouse caspase-11 (human caspase-4 or caspase-5) is mediated directly by cytosolic LPS (Kayagaki et al. 2011; Kayagaki et al. 2013; Hagar et al. 2013; Aachoui et al. 2013). The inflammasome revealed the principle way IL-1 family cytokines become activated and established pyroptosis as an innate immune host defense mechanism acting against a range of pathogens (Man et al. 2017). Here, we provide an overview of the cellular and molecular events leading to activation and execution of pyroptosis and specifically discuss the beneficial and detrimental effects of pyroptosis in antiviral host defense.

2. Inflammasomes: Molecular platforms for activation of inflammatory caspases

Inflammasome assembly is mediated by the nucleotide and oligomerization domain, leucine-rich repeat–containing (NOD) receptor (NLR) family of cytosolic pattern recognition receptors (PRRs) in response to infection or other immunological challenges (Sharma and Kanneganti 2016). Sensors that can initiate inflammasome assembly include the apoptosis-associated speck-like protein containing a CARD domain (ASC)-dependent NLR family member NLRP3 (NOD family, pyrin domain-containing 3), HIN domain-containing family member AIM2 (absent in melanoma 2) and Pyrin and the ASC-independent NLRP1 and NLRC4 (NLR family, CARD domain-containing 4) (Broz et al. 2010; Pelegrin et al. 2008; Sharma and Kanneganti 2016). Apart from these five well-characterized receptors, the NLR family proteins NLRP6, NLRP7 and NLRP12 and certain nucleic acid sensors like retinoic acid inducible gene-1 (RIG-I) and IFNγ-inducible protein 16 (IFI16) are reported to facilitate caspase-1 activation (Broz and Dixit 2016). Inflammasome assembly is initiated upon recognition of pathogen-associated or damage-associated molecular patterns by these PRRs. After sensing the activating ligands or other stimuli, the sensor subsequently recruits and activates caspase-1 in either an ASC-dependent or -independent manner. Once activated, caspase-1 processes pro-IL-1β and pro-IL-18 into their biologically active forms and simultaneously triggers pyroptosis (Man and Kanneganti 2015; Sharma and Kanneganti 2016).

Caspase-11 (orthologous to human caspase-4 and caspase-5) initiates pyroptosis directly by sensing cytosolic LPS, the cell wall component of Gram-negative bacteria (Kayagaki et al. 2011; Kayagaki et al. 2013; Hagar et al. 2013; Aachoui et al. 2013). These caspases do not process pro-IL-1β and pro-IL-18 into their active forms (Shi et al. 2017), although caspase-11-mediated pyroptosis induces changes in intracellular ionic balance that may facilitate assembly of the NLRP3 inflammasome, resulting in caspase-1 activation and processing of pro-IL-1β and pro-IL-18 (Ruhl and Broz 2015).

3. Gasdermins: The executioner of pyroptosis

Although inflammatory caspases mediate pyroptosis, the gasdermin family member gasdermin D is the substrate first identified to execute membrane lysis (Kayagaki et al. 2015; Shi et al. 2015; He et al. 2015). Activated caspases cleave gasdermin D to generate independent N-terminal and C-terminal fragments (Shi et al. 2015). Once free of the autoinhibitory C-terminal fraction, the N-terminus undergoes oligomerization to form membrane pores that drive leakage and lysis of cells (Shi et al. 2015; Kayagaki et al. 2015; Ding et al. 2016; Aglietti et al. 2016; Liu et al. 2016; Sborgi et al. 2016). Thus, the pore-forming function of cleaved gasdermin D carries out execution of pyroptosis and mediates release of IL-β as this cytokine is not secreted. Such proinflammatory cytokines are only released when cells lyse (Kayagaki et al. 2015; Shi et al. 2015). In addition to proteolysis by the inflammatory caspases, gasdermin D is cleaved by apoptotic caspases, caspase-3, caspase-7 and caspase-8, although this does not always lead to activation. Whereas cleavage by caspase-8 liberates the active N-terminal domain to drive pore-forming activity and pyroptosis similar to caspase-11 (Orning et al. 2018; Sarhan et al. 2018), caspase-3 and caspase-7 cleave and inactivate gasdermin D (Rogers et al. 2017; Taabazuing et al. 2017). Although gasdermin D is a key executioner, prolonged inflammasome activation induces pyroptosis in gasdermin D-deficient cells (Shi et al. 2015), consistent with additional pyroptosis executioners (Lamkanfi et al. 2008; Akhter et al. 2009), including additional gasdermin family members. Recently, pyroptosis has extended to include all gasdermin-mediated cell death because all the gasdermin family members except DFNB59 possess pore-forming activity (Shi et al. 2017). So far, only gasdermin E (DFNA5), activated by caspase-3, has been demonstrated to follow a parallel mechanism, in parallel with apoptosis (Rogers et al. 2017; Wang et al. 2017). In addition, gasdermin D drives secondary pyroptosis and IL-1β release even in the absence of caspase 1 (Schneider et al. 2017). Thus, there is still much to be discovered in this arena.

4. Pyroptotic cell death during viral infections: The good and the bad

Programmed cell death pathways remove the replicative niche and restrict the survival and proliferation of obligate intracellular pathogens (Jorgensen and Miao 2015; Dondelinger et al. 2016). Cell death also contributes to the pathogenesis of viral infections (Jorgensen et al. 2017). Moreover, inflammasomes and inflammasome-dependent cytokines contribute to antiviral immunity and resolution of infection (Lupfer et al. 2015; He et al. 2018). Although various DNA and RNA viruses induce pyroptosis via NLRP1, NLRP3 and AIM2 inflammasomes during infection (Lupfer et al. 2015), the overlap between pyroptosis and the release of IL-1β/IL-18 restricts the ability to distinguish the cytokine-independent contributions of pyroptosis in host defense (Jorgensen et al. 2017).

NLRP1 was the first NLR shown to form an inflammasome complex mediating caspase-1 activation (Martinon et al. 2002). NLRP1-mediated pyroptosis of hematopoietic progenitor cells causes prolonged cytopenia, bone marrow hypoplasia and immunosuppression in response to lymphocytic choriomeningitis virus (LCMV) infection (Masters et al. 2012). LCMV directly infects hematopoietic progenitor cells and induces caspase-1-dependent cell death that is enhanced in mice harboring an activating mutation (Nlrp1a^Q593P/Q593P). Consistent with this, NLRP1 deficiency in mice limits the induction of pancytopenia and increases virus-specific CD8⁺ T-cell response with improved recovery from LCMV infection (Masters et al. 2012). This study revealed detrimental effects of pyroptosis on hematopoiesis and generation of an effective immune response.

NLRP3, the most widely studied NLR, is activated during infection with diverse viruses (Lupfer and Kanneganti 2013; Lupfer et al. 2015). The NLRP3 inflammasome is thought to form in response to some change common to disruption of cellular homeostasis and ionic balance rather than any particular microbial or viral ligand (Sharma and Kanneganti 2016; Stewart and Cookson 2016). NLRP3 contributes to immunopathology as well as the quality of innate and adaptive immunity during viral infection above and beyond established microbial settings (Thomas et al. 2009; Tate et al. 2016; Lupfer and Kanneganti 2013). Although the importance of NLRP3 inflammasome-dependent antiviral responses is recognized, most effects have been attributed to the proinflammatory consequences of IL-1β and IL-18. As a result, the relative contribution of NLRP3-mediated pyroptotic cell death and inflammation during virus infection has not yet been defined.

The diverse array of DNA and RNA viruses that activate the NLRP3 inflammasome (Lupfer et al. 2015) include RNA viruses of the families Orthomyxoviridae (influenza virus), Paramyxoviridae (respiratory syncytial virus and measles virus), Rhabdoviridae (vesicular stomatitis virus and rabies virus), Picornaviridae (polio virus, enterovirus and encephalomyocarditis virus) and Flaviviridae (Hepatitis C virus, dengue virus and West Nile virus) as well as DNA viruses in the families Poxviridae (vaccinia virus and myxoma virus), Herpesviridae (herpes simplex 1, cytomegalovirus and varicella zoster) and Adenoviridae (adenovirus 5). DNA viruses such as herpes simplex virus (HSV) 1, murine cytomegalovirus (MCMV) and adenoviruses activate the NLRP3 as well as the AIM2 inflammasome (Lupfer et al. 2015).

Of particular interest is influenza virus, a long-recognized potent activator of the NLRP3 inflammasome (Kuriakose and Kanneganti 2017). The critical role of inflammasome-dependent cytokines and cell death responses have been observed during influenza virus infection in mice. Recent studies have demonstrated parallel and complementary activation of various programmed cell death pathways in response to influenza virus infection (Kuriakose et al. 2016; Kesavardhana et al. 2017; Thapa et al. 2016). Z-nucleic acid binding protein-1 (ZBP1), also known as DAI and DLM1, an IFN-inducible receptor interacting protein (RIP) homotypic interaction motif (RHIM)-containing protein, initiates the NLRP3 inflammasome as well as RIP kinase 3 (RIPK3), inducing combined pyroptosis, apoptosis and necroptosis in influenza virus-infected cells (Kuriakose et al. 2016; Thapa et al. 2016). ZBP1 senses the viral ribonucleoprotein complex of influenza A virus recruiting the RHIM-containing RIPK3 protein to transduce these diverse cell death signaling outcomes (Kesavardhana et al. 2017; Kuriakose et al. 2016). The ZBP1-RIPK3 interaction mediates influenza-induced inflammasome activation and programmed death of infected cells. Both processes are abrogated in ZBP1- and RIPK3- deficient cells (Kuriakose et al. 2016). The distinct contributions of each of these complementary pathways during influenza infection was supported by studies in cells lacking NLRP3, caspase-1, gasdermin D, RIPK3 and combined caspase-8 and RIPK3 (Kuriakose et al. 2016; Thapa et al. 2016; Nogusa et al. 2016). Based on these studies, influenza virus-infected cell death is inhibited in the absence of ZBP1 or when the three programmed cell death pathways and inflammasome activation are blocked or absent. Studies from our group added to prior understanding of innate immune signaling during influenza infection both in vitro and in vivo (Kuriakose and Kanneganti 2017), revealing a specific role for ZBP1 as a PRR that activates assembly of the NLRP3 inflammasome and the execution of pyroptosis in response to influenza infection, but not to other RNA virus infections that have been examined. We propose this unique phenomenon as ‘virus-induced unconventional NLRP3 inflammasome activation’ via ZBP1, which we refer to as the ZBP1-NLRP3 inflammasome. ZBP1 in association with RIPK3 and caspase-8 facilitate activation of the NLRP3 inflammasome that underlies IL-1β and IL-18 release (Kesavardhana et al. 2017; Kuriakose et al. 2016). Whether ZBP1-NLRP3-activated caspase-1 or caspase-8 facilitates gasdermin D cleavage or pyroptosis induction remains to be established.

ZBP1-dependent inflammasome activation and cell death restricts influenza virus replication, but contributes to lung inflammation as well as disease pathogenesis (Kuriakose et al. 2016; Thapa et al. 2016; Nogusa et al. 2016). Dengue virus infection induces activation of combined pyroptotic and apoptotic pathways that control virus replication (Suwanmanee and Luplertlop 2017). Additionally, NLRP3-mediated induction of pyroptosis is observed along with apoptosis during hepatitis C virus (HCV, strain FH1_T) infection of cultured HuH-7.5 hepatoma cells, suggesting a role for combined death pathways in HCV-induced liver pathology (Kofahi et al. 2016). Any contribution of ZBP1 as a PRR in dengue or HCV infection remains to be explored.

Inflammatory caspases contribute to central nervous system pathology following infection with an attenuated Evelyn-Rotnycki-Abelseth (ERA) strain of rabies virus (Kip et al. 2017). Morbidity and clinical disease scores are elevated in Casp1^−/−Casp11^−/− mice compared to either WT or Il1b^−/−Il18^−/− mice following infection with this strain (Kip et al. 2017). Pyroptosis mediated by inflammatory caspases may therefore contribute to host defense against rabies virus independently of IL-1β and IL-18. However, inflammasome-dependent responses did not modulate disease severity during infection with Challenge Virus Standard strain-11 (CVS-11), a highly virulent neurotropic strain of rabies virus (Kip et al. 2017). NLRP3-driven inflammasome activation was also recently revealed to be a critical component of inflammation during Zika virus (ZIKAV) infection of patients (He et al. 2018).

AIM2-dependent pyroptosis in human monocyte-derived dendritic cells was reported in response to immune-complexed human adenovirus (Eichholz et al. 2016). This demonstrates how humoral immunity engages professional antigen-presenting cells to initiate an innate immune response during persistent or recurring infections. The impact of immune complex-dependent pyroptosis in host defense is unknown although the inflammatory consequences may facilitate the T cell response to adenovirus infection (Eichholz et al. 2016). Induction of AIM2 inflammasome-mediated pyroptosis also occurs during MCMV infection (Rathinam et al. 2010), however this infection induces simultaneous activation of multiple cell death pathways. Apoptosis, necroptosis and pyroptosis were observed in an experimental model of retinitis suggesting involvement of cell death in the development of AIDS-related necrotizing retinitis caused by human CMV infection (Chien and Dix 2012). In addition to these viruses, the role of caspase-1 and caspase-11 was also investigated during infection with encephalomyocarditis virus (EMCV), murine gammaherpesvirus-68, vesicular stomatitis virus (VSV) and West Nile virus (WNV) (Rajan et al. 2011; Cieniewicz et al. 2015; Ramos et al. 2012). However, the individual contributions of pyroptosis and inflammatory cytokines have not been investigated.

Apart from the well-characterized inflammasomes, the DNA sensor IFI16 and RNA sensor RIG-I are also able to complex with ASC and caspase-1 to mediate caspase-1 activation and cytokine secretion (Kerur et al. 2011; Poeck et al. 2010; Pothlichet et al. 2013). RIG-I-dependent, NLRP3-independent caspase-1 activation and IL-1β secretion occur in normal human bronchial epithelial cells infected with influenza virus and in LPS-primed BMDCs infected with VSV (Poeck et al. 2010; Pothlichet et al. 2013). Despite the fact that Kaposi’s sarcoma-associated herpes virus (KSHV) encodes a protein that can inhibit NLRs and block NLRP1-mediated inflammasome activation (Gregory et al. 2011), IFI16 associates with the replicating KSHV DNA in the nucleus of endothelial cells and assembles a complex with ASC to mediate caspase-1 activation (Kerur et al. 2011). IFI16-dependent caspase-1 activation and inflammatory cytokine production also occur in B-cell lines latently infected with KSHV or Epstein Barr virus (EBV) and, possibly, in fibroblasts infected with HSV1 (Ansari et al. 2013; Singh et al. 2013; Johnson et al. 2013), noting that later studies failed to observe similar results with HSV1 (Diner et al. 2015). Although these studies suggest that inflammasome activation occurs downstream of IFI16 and RIG-I, pyroptosis has not been observed.

5. AIDS: When pyroptosis goes off limits

Human immunodeficiency virus (HIV) infection is one of the viral infections where the detrimental effects of pyroptosis are quite apparent (Doitsh and Greene 2016). Infection with HIV leads to a spectrum of diseases resulting in an immunocompromised state referred as acquired immunodeficiency syndrome (AIDS). The progressive depletion of CD4⁺ T cells is the major pathophysiological feature of AIDS, however, the mechanisms responsible for massive CD4⁺ T cell death has remained unresolved. Recent studies have identified pyroptosis as the mechanism mediating CD4⁺ T cell depletion and subsequent progression of AIDS (Doitsh et al. 2014). Interestingly, cell death during HIV infection does not involve virus-infected cells (Finkel et al. 1995). Rather, most of the dying cells in infected patients are bystander CD4⁺ T cells that acquire HIV by cell-to-cell spread within lymphoid tissues (Finkel et al. 1995; Doitsh et al. 2014; Galloway et al. 2015). The virus undergoes abortive replication in these bystander T cells leading to accumulation of viral DNA, which is sensed by the DNA sensor IFI16 (Monroe et al. 2014) and drives ASC-dependent activation of caspase-1 and induction of pyroptosis independently of NLRP3. The pyroptotic cell death of CD4⁺ T cells is amplified by the release of proinflammatory mediators including ATP, which subsequently amplifies NLRP3 inflammasome signaling and thereby perpetuates a chronic inflammatory state in infected individuals (Doitsh et al. 2014). Treatment with caspase-1 inhibitors or sh-RNA-mediated knockdown of caspase-1 or ASC prevented HIV-associated CD4⁺ T cell death in an ex vivo human lymphoid aggregate culture system (Doitsh et al. 2014), suggesting that the host immune response to HIV infection exacerbates pathology. Because pyroptosis contributes to CD4⁺ T cell depletion as well as chronic inflammation, the two major consequences of HIV infection, combined treatment with inhibitors of pyroptosis and antiretroviral agents is a suggested therapeutic option that may be explored (Doitsh and Greene 2016).

6. Pyroptosis in inflammation: Effector functions by IL-1β and IL-18

The identification of gasdermin D as the main executioner of pyroptosis helps to further confirm the key role of pyroptosis in release of bioactive IL-1β and IL-18. Because these cytokines lack a signal peptide directing them to the secretory pathway, caspase-1-processed forms pass through membrane pores dependent on gasdermin-D (Kayagaki et al. 2015; Shi et al. 2015). In addition to IL-1β and IL-18, many other proinflammatory mediators are released from cells as a consequence of pyroptosis (Man et al. 2017).

IL-1β and IL-18 are known to mediate a wide range of immune and inflammatory responses. During influenza virus infection, IL-1R supports cytotoxic T-cell and antibody responses that promote survival of mice (Pang et al. 2013; Ichinohe et al. 2009). Whereas IL-1β is protective early during infection, this cytokine contributes to immunopathology and disease severity later in infection (Schmitz et al. 2005). Similar to IL-1β, IL-18 enhances immune responses and promotes viral clearance, but also contributes to immunopathology and tissue damage (Denton et al. 2007; Lupfer et al. 2013). The current understanding is that the NLRP3 inflammasome and pyroptosis are both important for antiviral responses to influenza virus. IL-1β is also crucial for control of WNV infection within the CNS such that mice lacking components of the NLRP3 inflammasome or IL-1 signaling show increased susceptibility to WNV infection (Ramos et al. 2012). Elevated levels of IL-1 β in plasma of patients infected with WNV suggests a role for this cytokine in disease pathogenesis (Ramos et al. 2012). Recently, inflammasome-driven IL-1β has also been shown to promote an IRF3-dependent interferon response during WNV infection of myeloid cells, suggesting a tight link between the inflammasome and interferons for the antiviral innate response (Aarreberg et al. 2018). IL-1β and IL-18 similarly contribute to inflammatory responses during dengue virus infection (Tan and Chu 2013; Chirathaworn et al. 2010). Additionally, HCV infection induces production of IL-1β and IL-18 from both circulatory and liver resident macrophages, suggesting a role for these cytokines in hepatic inflammation (Shrivastava et al. 2013). The NLRP3 inflammasome is regarded as a key mediator of neuroinflammation during murine Japanese encephalitis virus infection (Kaushik et al. 2012). IL-18 supports IFNγ production from NK cells and contributes to early control of MCMV infection, further demonstrating the role of inflammasome-dependent cytokines in antiviral immune and inflammatory responses (Rathinam et al. 2010).

7. Interference with pyroptosis: Virus-encoded inhibitors and decoy proteins

Viruses have evolved mechanisms that interfere with inflammasome functions that control inflammatory cytokine and cell death responses (Lamkanfi and Dixit 2011). Inflammasome assembly and signaling as well as caspase-1 enzymatic activity are inhibited by specific viral modulators (Kanneganti 2010; Stewart and Cookson 2016; Lamkanfi and Dixit 2011). The vaccinia virus F1L protein and KSHV ORF63 protein are examples of decoys that inhibit NLRP1 inflammasome assembly, caspase-1 activation and IL-1β secretion (Gregory et al. 2011; Gerlic et al. 2013). Additionally, F1L can inhibit apoptosis by interfering with cytochrome c release (Stewart et al. 2005). ORF63 also interacts and disrupts NLRP3 inflammasome assembly (Gregory et al. 2011). Measles virus V protein interacts with NLRP3 to inhibit IL-1β secretion (Komune et al. 2011). Pyrin-only proteins (POPs) encoded by poxviruses interfere with inflammasome signaling (Kanneganti 2010; Lamkanfi and Dixit 2011). Orthopoxviruses myxomavirus M13L protein and Shope fibroma virus S13L protein also inhibit inflammasome assembly (Johnston et al. 2005; Dorfleutner et al. 2007), possibly by interacting with ASC. Viruses lacking these decoy proteins are attenuated, consistent with a role in viral pathogenesis.

Orthopoxvirus-encoded serpins (serine protease inhibitors or SPIs), including cytokine response modifier A (CrmA, also known as SPI-2) encoded by cowpox and conserved broadly in variola virus, vaccinia virus and ectromelia virus, acts as a pseudo-substrate inhibitor of the proteases caspase-1, caspase-8 and granzyme B, and also inhibits TBK1 (Ray et al. 1992; Quan et al. 1995; Tewari et al. 1995). Consistent with its role in inhibiting inflammasome- and TNF-death receptor-induced caspases, immune cell granzyme and interferon induction, CrmA deficiency dramatically attenuates cowpox virus.

Amongst RNA viruses, the influenza virus NS1 protein is also known to regulate caspase-1 activation as well as IL-1β and IL-18 secretion (Stasakova et al. 2005; Kuriakose and Kanneganti 2017; Chung et al. 2015). However, just like the role of NS1 in controlling type I interferon expression (Rajsbaum et al. 2012), the effect of NS1 on NLRP3 inflammasome activation is strain specific (Park et al. 2018). While the NS1 C-terminus of the A/swine/Saskatchewan/18789/2002/H1N1 strain does not suppress NLRP3 inflammasome activation in primary porcine alveolar macrophages, the NS1 of the pandemic H1N1 strain shows a strong inhibitory effect on NLRP3 inflammasome activation (Park et al. 2018). Meanwhile, pandemic H1N1 NS1 inhibits NLRP3 inflammasome activation by counteracting ASC ubiquitination, but the other strains such as A/PR/8/34 (H1N1), A/WSN/33 (H1N1) and A/Hong Kong/483/1997 (H5N1) disrupt caspase-1 cleavage by interacting with NLRP3 (Chung et al. 2015; Moriyama et al. 2016). It has been reported that the N-terminal RNA-binding domain of NS1 is important in blocking inflammasome activation (Moriyama et al. 2016). The N-terminus of NS1 can block viral RNA through its RNA binding domain and is critical for binding to both TRIM25 and RIG-I in order to suppress antiviral type I IFN responses (Jureka et al. 2015; Gack et al. 2009; Hale et al. 2008) such that NS1 inhibition of inflammasome responses may be a consequence of antagonizing type I IFN production (Kuriakose et al. 2016; Kochs et al. 2007). In addition to targeting the inflammasome components, viruses have also evolved to inhibit pyroptosis by directly processing the executioner gasdermin D. For example, the viral protease 3C of enterovirus 71 is able to cleave in the N-terminus to inactivate gasdermin D, which is critical for virus replication (Lei et al. 2017).

8. Conclusions and future perspectives

Programed cell death during viral infections is well established as an antiviral host defense mechanism to curtail survival and replication of these intracellular pathogens. Because of its proinflammatory nature, pyroptosis triggers activation of additional inflammatory cascades and immune surveillance systems to facilitate viral clearance and recovery from infection. Whereas the antiviral functions of inflammasomes and inflammasome-dependent cytokines are investigated in detail, only limited studies have explored the importance of pyroptotic cell death in antiviral host defense. Like any other inflammatory response, pyroptosis also can be a double-edged sword with both beneficial and detrimental effects. Recent studies have uncovered the molecular and biochemical mechanisms governing activation and execution of pyroptosis and discovered additional molecules in the cascade leading to pyroptosis. However, the functional relevance of many of these components including caspase-11 and gasdermin D has not been investigated in the context of viral infections. The activation of complementary cell death pathways and overlapping effects of cell death and inflammatory cytokine secretion complicates our understanding of pyroptosis in antiviral responses. Although gasdermin D is highly expressed in the epithelial cells of the gastrointestinal tract, the induction or effects of pyroptosis during viral infections in the gastrointestinal tract is not known. The newly generated gasdermin D-deficient mice in parallel with Casp1^−/−, Casp11^−/−, Il-1b^−/− and Il18^−/− mice provide an unprecedented opportunity to dissect the unique contributions of pyroptosis during viral infections. A better understanding of the mechanisms and functions of pyroptosis during viral infections may provide therapeutic options targeted either to harness the beneficial effects or to block the harmful effects of pyroptosis. Therefore, ‘Pathfinders’ are needed to further probe the role of pyroptosis in antiviral immunity.

Figure 1: Induction of pyroptosis during viral infections.

The cytosolic pattern recognition receptors NLRP1, NLRP3 and AIM2 sense cytosolic pathogen-associated or danger-associated molecular patterns in virus-infected cells and trigger the formation of inflammasome complexes with the adaptor ASC and the cysteine protease caspase-1. Assembly of the inflammasome complex facilitates proteolytic activation and cleavage of the caspase-1 zymogen. Active caspase-1 in turn cleaves the inactive pro-IL-1β and pro-IL-18 into bioactive cytokines. Caspase-1 also cleaves the membrane pore-forming protein gasdermin D into N-terminal and C-terminal fragments. The N-terminal fragment of gasdermin D binds to lipids in the plasma membrane and oligomerizes to form pores. The formation of pores results in loss of cell membrane integrity and lysis of the cell. The membrane pores formed by gasdermin D also facilitate release of bioactive IL-1β and IL-18.