“Chemical modulation of gasdermin-mediated pyroptosis and therapeutic potential”

Abstract

Pyroptosis, a lytic form of programmed cell death, both stimulates effective immune responses and causes tissue damage. Gasdermin (GSDM) proteins are a family of pore-forming executors of pyroptosis. While the most-studied member, GSDMD, exerts critical functions in inflammasome biology, emerging evidence demonstrates potential broad relevance for GSDM-mediated pyroptosis across diverse pathologies. In this review, we describe GSDM biology, outline conditions where inflammasomes and GSDM-mediated pyroptosis represent rational therapeutic targets, and delineate strategies to manipulate these central immunologic processes for the treatment of human disease.

Introduction

Inflammation acts as a double-edged sword: on one edge it is essential to launch protective responses against foreign pathogens and to promote tissue repair while on the other it drives tissue damage and malignant transformation [1]. Modern medicine has developed a robust armamentarium to inhibit inflammation, initially in the form of non-specific anti-inflammatory agents and more recently via the development of novel small molecules and biologics that target pathogenic inflammatory pathways and mediators [2]. In parallel, our understanding of inflammatory mechanisms has grown immensely. The past two decades of research in the field have uncovered a set of cellular responders to a multitude of infectious agents and markers of cellular dysfunction that initiate inflammatory cascades via so-called inflammasomes. Activation of inflammasomes often, but not always, results in a programmed form of cell death with necrotic morphology known as pyroptosis. This immunostimulatory form of cell death can alert the immune system to an invading pathogen or damaged tissue in order to coordinate an appropriate response to restore organismal homeostasis. However, when aberrantly activated inflammasomes instigate pathology of a diverse complement of diseases, ranging from sepsis to atherosclerosis to cancer. The gasdermin (GSDM) family of pore-forming proteins has emerged as a critical facilitator of inflammasome biology and pro-inflammatory cell death. In fact, many now define pyroptosis as GSDM-mediated lytic cell death [3]. The seminal contribution of GSDMD to the inflammatory pathways that produce such immense morbidity and mortality has made it a natural focus for pharmaceutical targeting. Moreover, study of related family members has identified additional diseases where GSDM targeting may harbor therapeutic potential. In this review, we will outline pathways to GSDM activation, their known roles in human pathology and strategies– pharmaceutical and otherwise– under development to modulate GSDM biology for the treatment of disease.

Gasdermin family members

The GSDM family consists of six members in humans: GSDMA, GSDMB, GSDMC, GSDMD, GSDME (DFNA5), and pejvakin (PJVK, DFNB59). Mice have three paralogs of GSDMA and four paralogs of GSDMC, but lack a detectable ortholog of GSDMB. Structurally, GSDMA-E consist of a two domain architecture, with an N-terminal (NT) pore forming domain connected to a C-terminal (CT) auto-inhibitory domain through a flexible linker region [4,5]. PJVK contains a truncated CT relative to the other GSDMs (Figure 1). Within the family, the GSDM^NT display high sequence similarity with more variability in GSDM^CT. All GSDM^NT except for PJVK have been demonstrated to harbor cytolytic pore-forming activity. Tissue expression profiles differ among GSDM family members (summarized in several recent reviews [6–8]). While disease relevance for individual GSDMs likely coincides with the organs/cell types in which they are expressed, specific reported disease associations will be presented below.

Figure 1: Schematic overview of critical residues and targetable sites of the gasdermins.

GSDMA-E contain an NT pore forming domain (light blue) and a CT inhibitory domain (dark blue). PJVK contains a putative pore forming domain (light blue) and a truncated CT domain (light purple). Activating (green) and inhibitory (black) cleavage sites are indicated by scissors. Residues annotated in black indicate those with known function. Residues annotated in purple italics indicate those with putative function due to conservation with other GSDM family members. Residues in red indicate sites modified by current GSDMD targeting drugs.

Human GSDMD is activated by caspases 1,4,5,8 at D275 and by neutrophil elastase (ELANE) at C268 [16,41,42,49,52–54]. GSDMD is inactivated by caspase-3 at D87 and by enterovirus 71 protease 3C at Q193 [113,121]. Among the other gasdermin family members, the D87 residue is conserved in GSDMB (D91), GSDMC (D85), and PJVK (D85), although proteolytic cleavage has not been shown at this site for GSDMC or PJVK. Important residues for GSDMD pore formation include C191 (oligomerization) and F240 (dimerization) [43–45]. The F240 residue is conserved in all GSDM family members, whereas C191 is not. Mutation of I104N results in a hypomorphic gasdermin D allele, and this site is conserved in GSDMB [49]. Four residues in GSDMD^CT (L304, V308, V364, L367) are responsible for GSDMD-inflammatory caspase exosite interactions [50,51]. GSDMA and GSDME harbor an inhibitory phosphorylation site (T8, T6, respectively) [17]. A conserved residue is found in GSDMB, GSDMC and PJVK. There are no described proteases for GSDMA or PJVK. GSDMB is cleaved at K229 and K244 by granzyme A; some splice variants may undergo caspase-1 cleavage at D236 [28,32]. GSDMC is cleaved at D365 by caspase-8 [35]. GSDME is cleaved at D270 by caspase-3 or granzyme B [62,63,119,120]. Direct GSDMD inhibitors all modify C191 [44,45,57]. DMF additionally succinates C56, C268, C309, and C467. DMF additionally modifies GSDME on C45, C156, C168, C180, C235, C371, C408, and C489. See text for further details. Created with BioRender.com.

GSDMA

Discovery of the first identified GSDM homolog, GSDMA, occurred through positional cloning of a gene causing abnormal hair and skin phenotype in mice. The name comes from its localization to the gastrointestinal tract and skin [9,10]. In mice, mutations in GSDMA3^CT cause excoriation, alopecia, and keratosis that are believed to arise due to disrupted auto-inhibition of GSDMA3^NT [11–14]. In support of this view, no phenotype has been observed in a knockout mouse model [5,15]. At this time, it is not known whether GSDMA requires cleavage for activation or, if so, the identity of its activating protease. However, expression of GSDMA^NT is sufficient for plasma membrane permeabilization and lytic cell death [16] and this may be inhibited by phosphorylation at Thr8 by an unknown kinase [17].

In humans, polymorphisms in GSDMA have been linked to inflammatory bowel disease (IBD), systemic sclerosis and asthma susceptibility [18–21]. Furthermore, its expression is increased in asthmatic patients [22]. Conversely, GSDMA is silenced in gastric cancer. Early work demonstrated that restoration of GSDMA in these tumor lines increases susceptibility to TGF-β, suggesting a tumor suppressive role for the protein [9,23]. Thus, therapeutic modulation of GSDMA activity is likely to require disease specific strategies – it may be advantageous to inhibit GSDMA pore formation in some inflammatory diseases whereas reactivation may hold therapeutic potential in some cancers.

GSDMB

Human GSDMB clusters in the same genetic locus as GSDMA and is thought to have arisen due to a gene duplication event [6,18,20,24]. Similar to GSDMA, genome-wide association studies have revealed two human polymorphisms that segregate together with an increased risk of IBD and asthma. These polymorphisms, p.G299S and p.P306S, are suggested to alter the structure of GSDMB but the biochemical and physiological effects of these alterations have not been defined [18,21,25–27]. A separate variant may alter caspase processing of GSDMB [28].

In contrast to GDSMA, the data on GSDMB in cancer are more variable. Increased expression of GSDMB has been linked to cancer proliferation and metastasis, with ectopic expression enhancing in vitro growth of cervical cancer cells and increasing cell motility and invasion of breast cancer cell lines [29–31]. Additional data showed GSDMB to promote metastases in xenograft models and identified an association of high GSDMB expression with poor prognosis in HER2-positive breast cancer. However, GSDMB shows variable expression changes in tumor relative to normal tissues across different cancer types [32].

A recent study discovered a novel role by which GSDMB converts cytotoxic lymphocyte (CTL)-mediated cell killing from apoptotic to pyroptotic in nature. Granzyme A from CTLs cleaves GSDMB, predominantly at K244, to induce target cell pyroptosis and promote efficacy of anti-PD-1 immunotherapy [32]. This work also showed that interferons and TNF-α variably induce GSDMB, highlighting a potential mechanism to augment anti-tumor immunity in some cancers.

More studies on the role of GSDMB in inflammatory disease and cancer immunology are necessary to inform future therapeutic targeting of this GSDM. Similarly to GSDMA, its inhibition may be beneficial in certain inflammatory disorders. However, the therapeutic efficacy of GSDMB activation or inhibition may be stage- and/or tissue-specific in cancer.

GSDMC

GSDMC was first identified as a marker for melanoma progression and given the name melanoma-derived leucine zipper-containing extranuclear factor (MLZE) [33]. As one of the least well-studied GSDM family members, the physiological role of GSDMC remains uncertain. Early work characterized GSDMC as an oncogene that accelerates proliferation of colorectal cancer cells [34]. High expression additionally correlates with poor prognosis in breast and lung cancers [35,36]. Inducers of GSDMC expression include ultraviolet light, NFATc1 and STAT3-PD-L1 signaling [35,37,38]. Under hypoxic conditions, the STAT3-PD-L1 pathway switches TNF-α-induced apoptosis to pyroptosis via caspase-8 cleavage of GSDMC to drive tumor necrosis [35]. Although this study showed that GSDMC-driven necrosis promoted tumor progression, the authors also showed enhanced tumor cell pyroptosis in response to certain chemotherapeutics known to cause immunogenic cell death (ICD) [39]. Together these data suggest that although GSDMC may contribute to carcinogenesis and cancer aggressiveness, its expression may be exploited as a therapeutic target to drive immunostimulatory cancer cell pyroptosis.

GSDMD

GSDMD was originally identified by homology to other family members and recognized as the most efficient substrate of inflammatory caspases [40]. It was not until 2015, however, that GSDMD was identified as the conduit for IL-1β and IL-18 release and the primary executioner of pyroptosis downstream of both multiple inflammasomes [16,41,42]. Since its initial characterization, it has rapidly become the best understood GSDM family member. Unlike its homologs, no polymorphisms have been described linking GSDMD expression to human disease, though some demonstrate functional consequences [43]. Deficiency or inhibition of Gsdmd has been demonstrated to be protective in numerous mouse models of inflammatory disease, including sepsis, experimental autoimmune encephalitis (EAE), familial Mediterranean fever (FMF), and neonatal-onset multisystem inflammatory disorder (NOMID; see below), highlighting the allure of GSDMD as a drug target for treatment of diverse inflammatory conditions [42,44–48].

The first proteases identified for GSDMD were the inflammatory caspases-1/4/5, which cleave GSDMD after D275 (caspase-1/11 after D276 in mice)[16,40–42,49]. Emerging data have identified a critical inflammatory caspase exosite interaction with GSDMD^CT [50,51]. This novel caspase-GSDM interface reveals a function for GSDMD^CT outside of its traditional view as solely an autoinhibitory domain. Other proteases, including caspase-8, neutrophil elastase (ELANE) and cathepsin G, cleave GSDMD in the same interdomain linker region as caspase-1 [52–55]. The diversity of upstream signals that activate GSDMD further supports its therapeutic potential as a common bottleneck for multiple pyroptotic stimuli.

In addition to its activation (and inhibition; see below) by protease cleavage, several post-translational modifications (PTMs) alter GSDMD function. Itaconate causes inhibitory covalent modification at C77 (present in mice but not humans) while fumarate and its derivatives block GSDMD through modification of C191 and other residues [56,57]. The former change prevents GSDMD cleavage whereas the latter inhibits its oligomerization and is the target of several GSDMD inhibitors (see below). In contrast, oxidation of GSDMD at multiple cysteine residues reportedly augments its cleavage by caspase-1 [58]. These PTMs highlight the potential influence of cellular metabolic state on GSDMD activation. Surprisingly little is known about other PTMs, such as phosphorylation or ubiquitination, though prediction tools suggest that both may occur and a single study found that some sites for PTM have little effect on GSDMD function [43]. Future studies may further clarify the roles played by PTMs in tuning the threshold for GSDMD pyroptosis.

Specific downstream consequences of GSDMD activation are cell type-dependent. Among the possible outcomes are inflammatory cytokine release, neutrophil extracellular trap (NET) extrusion, and/or pyroptotic cell death. As the primary GSDM downstream of inflammasomes, GSDMD has become a major focus of pharmaceutical development across divers inflammatory diseases, a topic addressed more thoroughly below [44,45,57,59].

GSDME

GSDME is the most ancient family member [60] and gain-of-function mutations were originally described as a cause of non-syndromic hearing loss in humans [61]. Two groups recently characterized GSDME as a pore-forming protein that triggers pyroptosis upon cleavage by apoptotic caspase-3 after D270. The generated GSDME^NT fragment in essence converts cells committed to apoptosis instead to a pyroptotic phenotype [62,63]. This cell fate switch appears to require a critical GSDME expression threshold, a phenomenon that may hold true for other GSDMs [64].

Recent studies of GSDME activation have fallen into three broad camps: caspase-9-driven intrinsic apoptotic activation of caspase-3 following chemotherapy regimens, caspase-8-driven activation of caspase-3 occurring via either inflammasomes or other signaling complexes, and cell extrinsic activation by granzyme B, all described below.

Targetable aspects of GSDME remain scant. One study provided evidence of inhibitory phosphorylation at the conserved T6 [17], and dozens of high-throughput studies have identified S252 and Y254 (www.phosphosite.org), both in the interdomain linker region, as potential phosphorylated sites conserved in mice but not zebrafish [65]. Further, the key modifiable C191 in human GSDMD is not present in GSDME. Whether a caspase exosite like in GSDMD exists for GSDME remains to be determined [50,51].

Pejvakin

PJVK, as previously mentioned, is the only member of the GSDM family without a complete CT domain. Further distinguishing it from other family members, loss of Pjvk in mice results in a distinct phenotype characterized primarily by hearing loss whereas other GSDM knock-out models lack a baseline phenotype [52,63,66–68]. While PJVK plays a role in protecting auditory hair cells from free radical-induced damage [69,70], no demonstration of pore-forming ability has been reported. Such data will be important to better understand its physiologic function.

Key residues involved in the regulation of individual human GSDMs are presented in Figure 1.

Pathways to GSDM Activation: Inflammasomes

The first recognized and most extensively studied pathways to GSDM cleavage and pyroptosis involve assembly of inflammasome complexes. Pathogen-associated molecular patterns (PAMPs) resulting from microbial and viral infections or damage-associated molecular patterns (DAMPs) released from host cells are detected by sensor proteins, including nucleotide-binding domain (NBD), leucine-rich repeat (LRR)-containing (NLR) proteins, pyrin and absent in melanoma 2 (AIM-2). Resultant activation of these proteins generally triggers recruitment of ASC [apoptosis-associated speck-like protein containing a caspase activation and recruitment domain (CARD)] and subsequent association and proximity-induced auto-cleavage of pro-caspase-1. Once activated, caspase-1 cleaves the inflammatory cytokines pro-IL-1β and pro-IL-18, as well as GSDMD.

The discovery that NLRP1 forms a caspase-1 activating platform in complex with ASC served as the first description of an inflammasome [71]. This general structural framework whereby a sensor protein (here NLRP1) enables supramolecular complex formation to facilitate caspase-1 activation holds true as a model for canonical inflammasomes. The NLRP1 protein contains an NT pyrin domain (PYD), an NBD, an LRR domain, a ‘function to find’ domain (FIIND) and a CT CARD [71]. NLRP1 auto-proteolysis within the FIIND is requisite for its activation [72,73]. Upon stimulus-triggered degradation of NLRP1^NT, the liberated CT CARD interacts with the ASC to promote its oligomerization into a single ASC focus, also called a speck, where pro-caspase-1 undergoes proximity-induced auto-activation [74,75].

While humans have one NLRP1 protein, mice express three paralogs, NLRP1a, NLRP1b, and NLRP1c, which all lack a PYD [76]. Mouse NLRP1b can be activated by the lethal toxin of Bacillus anthracis, whereas human NLRP1 is unaffected [76]. The NLRP1 inflammasome also responds to Toxoplasma gondii infection in mice and possibly humans [77–79]. Several recent studies report that multiple viral proteases cleave the auto-inhibitory NT fragment of NLRP1, releasing the active CT fragment to initiate inflammasome activation [80,81]. New evidence additionally describes NLRP1 as a double-stranded (ds)RNA sensor [82]. These studies expand the potential disease relevance of this earliest recognized inflammasome sensor.

Among the few proteins harboring both a FIIND and CARD, CARD8 shares a parallel activation mechanism with NLRP1 [72]. However, unlike NLRP1, human CARD8 reportedly forms an inflammasome independently of ASC [83]. Instead, direct CARD homotypic interactions between CARD8 and pro-caspase-1 enable proteolytic activity of the latter. Recent work has identified distinct features of the CARD of NLRP1 compared to that of CARD8 that explain the different composition of these two inflammasomes [84,85]. Of note, the ASC-independent inflammasomes appear to be inefficient at IL-1β and potentially GSDMD processing [86,87].

Perhaps the most studied inflammasome sensor, NLRP3 can form inflammasome complexes in response to a wide range of PAMPs and DAMPs, including bacteria, viruses, fungi, and protozoans, as well as monosodium urate and silica crystals, asbestos, extracellular ATP, DNA-RNA hybrids, and heme [88–93]. Given the diversity of activating stimuli, it has been proposed that NLRP3 responds to a common cellular perturbation. Proximal activating stimuli generated by NLRP3 agonists include reactive oxygen species, mitochondrial dysfunction, lysosomal damage, and ion flux; however, a common activation mechanism remains under investigation [94–100]. Multiple post-translational mechanisms also contribute to NLRP3 regulation (reviewed in [101]). Like GSDMD, the role of NLRP3 as a convergence point for diverse inflammatory stimuli increases its attractiveness as a therapeutic target.

In addition to canonical inflammasome activation, non-canonical activation occurs upon intracellular LPS binding to inflammatory caspase-4/5 in humans or caspase-11 in mice. This process activates GSDMD-dependent pyroptosis but these caspases lack activity toward pro-inflammatory cytokines pro-IL-1β and pro-IL-18 [102,103]. However, secondary NLRP3 activation can facilitate subsequent cytokine release, likely in response to potassium efflux through newly formed GSDMD pores [104,105].

Among other inflammasome sensors, the NLRC4 inflammasome is assembled in response to bacterial type 3 secretion systems in humans and mice, as well as flagellin in mice [106,107]. Recognition of cytosolic dsDNA activates the AIM2 inflammasome. Separately, the pyrin inflammasome responds to a number of bacterial toxins, including those of Clostridium difficile and Clostridium botulinum. However, rather than sensing them directly, these toxins inhibit RhoA GTPase activity [108], causing pyrin dephosphorylation and downstream inflammasome activation [109]. The reader is referred to recent reviews for more information about inflammasomes [110,111].

The central position of GSDMD as a bottleneck for inflammasome biology is depicted in Figure 2.

Figure 2: Understanding GSDMD as a bottleneck for inflammasome-driven effects.

Inflammasome-mediated GSDMD activation and the resulting downstream effects are divided into four general steps: (I) inflammasome activation, (II) GSDMD cleavage, (III) GSDMD oligomerization and membrane insertion and (IV) downstream processes/outcomes. (I) Numerous inflammasome activating stimuli are sensed by either a canonical or non-canonical inflammasome sensor protein. Depicted triggers for various canonical inflammasomes include bacterial components (type III secretion system, flagellin), dsRNA, viral proteases, and other stimuli that facilitate proteasomal activation of NLRP1 and CARD8. Depicted stimuli for NLRP3 canonical inflammasome activation include aspergillus, toxoplasma, monosodium urate (and other) crystals, amyloid β, ATP, heme released from lysed red blood cells, DNA:RNA hybrids, and K+ efflux. The non-canonical inflammasome is activated by LPS from gram-negative bacteria. (II) GSDMD^FL (full-length) can be cleaved by multiple caspases and other proteases to release the GSDMD^NT fragment. (III) The liberated GSDMD^NT associates with membranes, oligomerizes, and inserts into the membrane as a higher-order multimeric pore (~27mer) upon conformational change. (IV) Assembled GSDMD pores cause both cell intrinsic and extrinsic effects. Cell intrinsic effects include ion fluxes (both efflux and influx), NETosis, and pyroptosis. Some DAMPs additionally produce cell intrinsic effects. Other DAMPs and factors released via GSDMD pores or following GSDMD-mediated NETosis or pyroptosis cause cell extrinsic effects, including immune cell activation, clotting, and tissue repair or damage. To the right are agents that function at each step. Of note, GSDMD can be activated outside of the context of inflammasomes and a variety of stimuli trigger cleavage of other GSDMs to produce similar downstream effects (see text for details). PLC = phospholipase C. Created with BioRender.com.

Other GSDM Regulatory Mechanisms: Beyond Inflammatory Caspases

While the inflammasome pathways described above converge on inflammatory caspase activation to cleave GSDMD and additional relevant targets (pro-IL-1β, pro-IL-18, etc.), subsequent research has revealed greater diversity of GSDM proteolysis mechanisms. The implicated proteases target both GSDMD and other family members and include both activating and inhibitory cleavage events. Furthermore, GSDM proteolysis can be broadly divided into caspase-dependent and caspase-independent events.

In contrast to the roles of inflammatory caspases in pyroptosis, separate homologs exert important functions during apoptosis either as initiators (caspases-2,8,9,10) or executioners (caspases-3,6,7). Classically, apoptosis is thought to be a non-lytic, non-inflammatory form of programmed cell death. However, shortly after the recognition of GSDMD cleavage and pore formation by inflammatory caspases, GSDME was shown to be cleaved at D270 by caspase-3 to produce secondary pyroptosis in response to apoptotic stimuli; activation by caspases-6/7 is currently undetermined [62,63]. This and subsequent findings[112,113] have blurred the lines between pyroptosis and apoptosis as discrete forms of cell death.

Cytotoxic chemotherapeutics are among many stimuli that activate intrinsic apoptosis through the caspase-9/apoptosome pathway [63]. Another apoptotic caspase, caspase-8, is a component of many complexes and thus has multiple triggers for activation. Both caspase-8 and caspase-9 converge on caspase-3 activation during apoptosis. Of note, caspase-8 is also recruited by ASC CARD to inflammasomes where it mediates downstream caspase-3 activation [114,115]. Caspase-1 can similarly perform this function, especially in the absence of effective GSDMD-dependent pyroptosis [116]. Any of these pathways can therefore lead to GSDME-dependent pyroptosis. In addition to its role in activating caspase-3, caspase-8 also directly liberates pore-competent GSDMC^NT and GSDMD^NT through proteolysis, providing additional routes to secondary pyroptosis [35,54]. The necessary complexes and recognition sites for these various cleavage events are yet to be fully elucidated.

On the other hand, caspase-3 can also cleave at D87 within GSDMD to generate an inactive 43 kDa fragment [113]. This cleavage can be concurrent with inflammatory pathway cleavage at D275, where it retains its inhibitory function [117]. A conserved site, D91, in GSDMB is likewise targeted by caspase-3 (and caspases-6/7), resulting in inhibitory cleavage [27,118]. Thus, caspase-3 can both promote and inhibit GSDM-pyroptosis in a context-dependent manner.

Recent data demonstrate that granzyme B targets the same activation site in GSDME as caspase-3 [119,120] with similar consequences as granzyme A activity toward GSDMB. These compelling studies demonstrate critical functions for these GSDMs in the development of anti-tumor immunity, to be discussed more below. Thus, although CTL-mediated killing was long held to be immunologically “silent” like apoptosis, GSDMB or GSDME expression can convert this process to an immune-stimulatory pyroptotic cell death.

Adding to the long list of GSDM-cleaving enzymes, both ELANE and cathepsin G activate GSDMD via proteolysis at linker-region sites close to the canonical D275 (C268 and L273, respectively). The only reported caspase-independent GSDM inactivating cleavage is performed by enteroviral protease 3C on GSDMD at Q193 [121]. As this protease was recently reported to activate NLRP1 [80], the overall consequences of its proteolytic activity require further investigation. In time, additional GSDM proteases may be uncovered.

The variety of GSDM activating and inhibitory stimuli underscores the potentially broad set of conditions where targeting GSDM pyroptosis may harbor therapeutic potential. However, there is no standardized approach to these diseases since GSDMs have variably been demonstrated to both ameliorate and worsen infectious and inflammatory conditions. Moreover, in oncology settings GSDM activation can both promote immunogenic cancer cell death and cause normal tissue toxicity and uncontrolled inflammatory responses. Below we will briefly outline a number of diseases where GSDM pyroptosis has been implicated and where GSDM-targeted therapies are likely to be tested in the future.

Disease Relevance

Monogenic Inflammatory Diseases

Genetic mutations in inflammasome components produce rare, heritable auto-inflammatory diseases (Table 1). These pathogenic mutations drive spontaneous inflammasome activation, resulting in sterile inflammation driven by IL-1β, IL-18 and other factors. The most prevalent of these diseases is FMF, a disease attributed to likely gain-of-function mutations in the MEFV gene that encodes the pyrin protein [122,123]. FMF is a highly heterogeneous disease that is characterized by recurring attacks of inflammation with fevers, peritonitis, pleuritic, polyarthritis, and an erysipelas-like skin disease. The mainstay of treatment for FMF is the microtubule targeting agent colchicine, which is effective for many patients though treatment resistance does occur. Patients have elevated serum levels of inflammatory cytokines such as IL-1β, IL-6, and IL-18 [124]. Newer treatments with biologics targeting individual cytokines such as canakinumab (anti-IL-1β mAb) and tocilizumab (anti-IL-6R mAb) have shown success in patients resistant to colchicine [125].

Table 1.

Monogenic diseases associated with inflammasome genes

Gene	Disease	Current Treatments
MEFV	Familial Mediterranean Fever (FMF)	Colchicine, canakinumab
NLRC4	Autoinflammation with infantile enterocolitis (AIFEC)	Under investigation
NLRP3	Cryopyrin-associated Periodic Syndrome (CAPS)	Corticosteroids, canakinumab, anakinra, rilonacept
	Familial Cold Autoinflammatory Syndrome (FCAS)
	Neonatal-onset Multisystem Inflammatory Disease (NOMID)
	Muckle-Wells Syndrome (MWS)
NLRP1	Multiple Self-healing Palmoplantar Carcinoma (MSPC)	Retinoids, surgical removal of carcinomas
	Juvenile Recurrent Respiratory Papillomatosis (JRRP)	Surgical removal of papillomas
	Familial Keratosis Lichenoides Chronica (FKLC)	Retinoids, corticosteroids, phototherapy

Other inflammasome sensors have also been implicated in sterile auto-inflammatory diseases. Over 200 gain-of-function mutations in exon 3 of NLRP3 have been identified resulting in a spectrum of diseases that together are referred to as cryopyrin-associated periodic syndrome (CAPS). These syndromes range from the mildest presentation as familial cold auto-inflammatory syndrome (FCAS) to the most severe NOMID [126]. The most common treatments for CAPS are canakinumab or anakinra (IL-1R antagonist protein) [127]. Similar to NLRP3 mutations, NLRC4 inflammasomopathies also present with auto-inflammation, arthralgia, and rash. The observation that NLRC4 mutations can lead to FCAS and NOMID like diseases highlights the phenotypic convergence of aberrant inflammasome pathway activation [128–130].

As NLRP1 is predominantly expressed in the skin, mutations in this receptor lead to a clinical diseases such as multiple self-healing palmoplantar carcinoma (MSPC), juvenile recurrent respiratory papillomatosis (JRRP) and familial keratosis lichenoides chronica (FKLC) [131–133]. Some mutations lead to severe disease that closely mimics systemic inflammasome activation and presents as auto-inflammation with arthritis and dyskeratosis (AIADK) [132,133]. Since dipeptidyl peptidase 9 (DPP9) is an endogenous inhibitor of NLRP1, it not surprising that a recent preprint details a family with pathogenic DPP9 variants that allow NLRP1 activation and produce similar disease phenotypes [134]. Because DPP9 also restrains CARD8 activity, some phenotypes associated with DPP9 dysfunction may be due to pathologic CARD8 activation.

Despite the genetic diversity of the above inflammasomopathies, the affected pathways all converge on inflammasome-mediated GSDMD cytokine release and pyroptosis. Thus, the inflammasome-GSDMD axis represents a rational set of targets to modulate these diseases. While biologic drugs targeting disease-associated cytokines have proven to be beneficial, upstream targeting of inflammasomes or GSDMD may be more efficacious by simultaneously blocking the release of multiple inflammatory mediators. In mice, many of these diseases are rescued by knockout of ASC, inflammatory caspases, or GSDMD, supporting the viability of this approach [46–48].

Monogenic disorders, although rare, often serve as testing platforms for novel pathway modulating agents, benefitted by their orphan disease status and uniform pathophysiology. As such, these inflammasomopathies may become a primary point-of-entry into human use for novel inflammasome and GSDMD targeting therapies. The established efficacy of therapeutically blocking downstream mediators further enforces the potential for this approach. Identification of suitable orally bioavailable small molecules would also benefit patients who currently rely on regular injection of biologics. Demonstration of safety and efficacy of inflammasome-directed therapies in these rare genetic diseases will additionally pave the way for expanded use in diverse inflammatory diseases where inflammasomes and GSDM-mediated pyroptosis contribute to morbidity and mortality.

Diseases with Aberrant Inflammasome Activation

While the initial entry of inflammasome- and GSDM-targeted therapeutics into the clinical realm may be for the treatment of the above rare genetic disorders, the long-term impact of such therapeutics is expected to be greatest for complex diseases with an inflammasome-driven component to their pathophysiology.

Acute pathologies that fit this mold include sepsis, disseminated intravascular coagulation (DIC) and cytokine release syndrome (CRS). Multiple studies have shown protection from sepsis in Gsdmd-deficient mice [42,44,45,135]. The importance of GSDMD for releasing NETs and activating tissue factor support its importance for DIC [136–139]. Finally, GSDME has been shown to cause CRS in models of chimeric antigen receptor (CAR) T cell therapy [119]. Whether other GSDMs play roles in acute inflammatory diseases remains to be seen. Still, the preponderance of data show that GSDM activation can contribute to an overwhelming inflammatory response. Given the profound disease burden attributed to these and similar life-threatening conditions, GSDM inhibitors may deliver groundbreaking therapeutic advances in these contexts.

In terms of chronic diseases with demonstrated inflammasome- and GSDM-mediated pathology, the list includes atherosclerosis, liver disease, diabetes, cardiovascular disease, ischemic stroke, graft-versus-host-disease, sickle cell anemia, neurodegenerative diseases and multiple sclerosis (MS), among others (reviewed in [7]). This laundry list of conditions (Table 2) includes some of the biggest killers and most costly diseases in medicine. The implication of inflammasomes and GSDMD in the pathogenesis of complications across such a prevalent and burdensome set of disease raises the potential impact of successfully targeting GSDM-mediated pyroptosis to revolutionary in its scope. Interested readers are referred to other reviews in this issue for more specifics regarding some of these disease topics.

Table 2.

Acute and chronic diseases associated with inflammasome activation

	Disease	References
Acute	Sepsis	[44,45,135,210]
	Disseminated Intravascular Coagulation (DIC)	[136–139]
	Cytokine Release Syndrome (CRS)	[119]
	Tumor Lysis Syndrome	[211]
Chronic	Multiple Sclerosis	[212,213]
	Atherosclerosis	[214,215]
	Diabetes	[216–218]
	Gout (and other crystal/particulate disorders)	[90,91,96,187,215,219,220]
	Alzheimer’s Disease	[221,222]
	Asthma	[20,25,26,80,223–226]
	Inflammatory Bowel Disease	[193,227,228]
	Non-alcoholic Fatty Liver Disease	[229]
	Ischemic Stroke	[230]
	Sickle Cell Anemia	[231–233]
	Rheumatoid Arthritis	[234,235]

The data on GSDMs, especially GSDMD, in infectious diseases reveals a mixed picture. While GSDMD contributes to defense against some microorganisms [140–142], Gsdmd-deficient mice show better outcomes in other infection models [52,143]. Likely, the balance between GSDMD-mediated eradication of replicative environments and constructive immune stimulation versus inflammatory tissue pathology will determine whether therapeutic benefit can be derived from targeting GSDMD in distinct infectious diseases. Such information will need to be generated on a case-by-case basis.

Anti-Tumor Immunity

The extent to which GSDMs contribute to the immunogenicity of cancer cell death is incompletely understood, yet emerging evidence supports the sufficiency of multiple GSDMs to produce ICD [32,120]. The strongest data supporting the function of GSDMs in anti-tumor immunity come for GSDMB and GSDME, which are cleaved by granzyme A and B, respectively, from CTLs. Expression of GSDMB by cancer cells promoted efficacy of anti-PD-1 therapy in immune competent tumor-bearing mice [32]. For GSDME, two separate papers have now shown its role in driving ICD, as GSDME expressing tumors in immunocompetent mice fail to regrow tumors after either tumor re-challenge [120] or therapy cessation [144]. Furthermore, while low-level GSDMC cleavage may be responsible for tolerizing tumor necrosis in some tumors, activation of caspase-8-mediated cleavage by certain chemotherapeutics promotes cancer cell pyroptosis [35]. Likewise, secondary GSDMD activation can occur downstream of apoptotic stimuli [112]. These data highlight the potential for GSDMs to activate anti-tumor immunity in response to CTLs or chemotherapy, though much remains to be learned about the full extent to which individual GSDMs contribute to ICD in different malignancies. Approaches to therapeutically control GSDM activity to direct anti-tumor immunity will be discussed below.

Hematopoietic Disorders

Cytopenias (decreased blood cell counts) arise as part of many inflammatory and autoimmune disorders. Mechanistically, chronic inflammation drives hematopoietic impairment that involves inflammasome activation. For example, anemia of chronic disease shows contribution from caspase-1 activation and prolonged IL-1β exposure drives hematopoietic stem cell exhaustion in model systems [145,146]. Yet, the contributions of GSDMs to these effects remain largely unknown. This area deserves further investigation both to prevent bone marrow failure caused by chronic inflammation and to understand potential hematological consequences of therapeutically modulating the inflammasome-GSDMD axis.

Additionally, inflammasomes have emerging functions in age-related clonal hematopoiesis, also known as clonal hematopoiesis of indeterminate potential (CHIP). These terms describe the age-dependent acquisition of neoplasia-associated mutations without overt hematopoietic neoplasm. In addition to being a precursor to malignancies such as myelodysplastic syndromes (MDS) or acute myeloid leukemia (AML), CHIP is a risk factor for cardiovascular morbidity and mortality [147–149]. Atherosclerosis and heart failure are exacerbated in a CHIP mouse model through an NLRP3-dependent mechanism [150,151]. Inflammasome activation may, in fact, be a general feature of CHIP that provides a selective advantage to these clones [152–154]. These data suggest that CHIP-associated inflammasome activation may contribute to a variety of diseases of aging.

The exact functions exerted by inflammasomes and GSDMs in myeloid leukemogenesis should be the focus of future research. Additionally, whether they can be inhibited to limit myeloid leukemogenesis or selectively activated to treat myeloid leukemia remain open questions. Such efforts may turn CHIP from a risk factor to an actionable biomarker to prevent inflammatory pathology and forestall the development of MDS and AML.

Strategies to Target GSDM-Mediated Pyroptosis

Having laid the foundation to understand GSDM biology and disease associations, we now turn to approaches to modulate GSDM function therapeutically. Conceptually, chemical modulation of GSDM-mediated pyroptosis can be divided into three general categories of targets: the upstream pathways, the GSDMs themselves and the downstream effectors (Table 3).

Table 3.

Current pharmacological strategies for modulation of pyroptosis

Target	Drug	Phase	References/ Clinical Trials
Inhibitors of Pyroptosis
NLRP3	MCC950	Failed Phase 2 trial in rheumatoid arthritis patients due to hepatotoxicity	[164]
	Inzomelid	Completed Phase 1	NCT04015076
	IZD334	Completed Phase 1	NCT04086602
	Tranilast	Approved outside the US for asthma, keloids, and conjunctivitis; Phase 2 trial ongoing in CAPS patients	NCT03923140
	DFV890 (IFM-2427)	Phase 2 trials for knee osteoarthritis, FACS, and COVID-19	NCT04868968; NCT04886258; NCT04382053
	Dapansutrile	Phase 2 trial completed for gout and in progress for CRS	[168]; NCT04540120
	RRx-001	Phase 3 for oncology applications	[167]
	CY-09	Pre-clinical	[169]
	Oridonin	Pre-clinical	[166]
	INF39	Pre-clinical	[170]
Caspase-1	Pralnacasan (VX-740)	Failed Phase 2 trial in rheumatoid arthritis patients due to hepatotoxicity	[175]
	Belnacasan (VX-765)	Failed to meet primary endpoints in Phase 2 trial in epilepsy patients	[175,178]
GSDMD	Necrosulfonamide	Pre-clinical	[44]
	Disulfiram	Approved for alcohol dependence; Phase 1–3 for oncology applications	[45,171]
	Dimethyl fumarate	Approved in the US for multiple sclerosis; Approved in the EU for plaque psoriasis	[172]
Unknown	Punicalagin	Pre-clinical	[200]
	LDC7559	Pre-clinical	[138]
Activators of Pyroptosis
NLRP1/ CARD8	Val-boroPro (talabostat, BXCL701)	Phase 1–3 for oncology applications	NCT03910660; NCT03910660; NCT04171219
NLRP3	BMS-986299	Phase 1	NCT03444753
GSDMA	Nanoparticle-conjugated delivery of GSDMA3	Pre-clinical	[199]
GSDMD	AAV delivery of GSDMDNT	Pre-clinical	[198]

AAV = adeno-associated virus; EU = European Union

Targeting upstream pathways

The most advanced therapeutic strategies directed at upstream drivers of GSDM-mediated pyroptosis target the NLRP3 inflammasome, for reasons related to its broad relevance to inflammatory diseases as mentioned above. Selective pharmacologic inhibition of the NLRP3 inflammasome was first demonstrated with the sulfonylurea glyburide in 2009 [155]. A more potent diarylsulfonylurea analog, MCC950, also selectively inhibits the NLRP3 inflammasome by binding to the NBD and blocking ATP hydrolysis [156–158]. While MCC950 has shown therapeutic benefit in mouse models of myocardial infarction, atherosclerosis, Parkinson’s disease, Alzheimer’s disease, EAE and colitis, it is unable to inhibit some CAPS-associated NLRP3 mutants [157,159–163]. Unfortunately, development of MCC950 stalled due to hepatotoxicity in a phase 2 clinical trial for rheumatoid arthritis [164]. Two analogs of MCC950, inzomelid and IZD334, have demonstrated safety for healthy adults in phase 1 clinical trials and offer the latest hope for this class of inflammasome inhibitors (https://clinicaltrials.gov/ct2/show/NCT04015076, https://clinicaltrials.gov/ct2/show/NCT04086602).

Tranilast, a tryptophan derivative, also binds within the NBD of NLRP3 but inhibits inflammasome activation by blocking oligomerization [165]. Tranilast is approved for use outside of the US in asthma, allergic conjunctivitis, and keloids and has shown efficacy in mouse models of gout and CAPS. A phase 2 clinical trial for its use in CAPS patients is ongoing (https://clinicaltrials.gov/ct2/show/NCT03923140). Additional NLRP3 inhibitors are in various stages of development ([166–172] and see recent reviews on the topic [110,164,173]).

Currently, specific inhibitors of other inflammasomes are not available. Proteasome inhibitors can block NLRP1 and CARD8 activation but have much broader cellular activity [174]. Specific inhibition of other inflammasomes may prove efficacious in diseases where that inflammasome drives detrimental pathology rather than playing a prominent role in disease resolution or pathogen eradication. However, such strategies may simply compromise immunity to specific infectious agents.

Another upstream approach to targeting GSDM-mediated pyroptosis is inhibition of caspase-1. Tetrapeptide inhibitors targeting the catalytic cysteine within the caspase-1 active site have failed clinically due to unfavorable pharmacology and toxicity profiles [175]. Synthetic peptidomimetic prodrugs were developed to overcome these issues; these drugs include pralnacasan (VX-740) and belnacasan (VX-765). The metabolic products of both of these drugs are reversible inhibitors of caspase-1/4/5 [176,177]. A phase 2 clinical trial of VX-740 in rheumatoid arthritis patients was stopped due to hepatotoxicity, while a phase 2 clinical trial of VX-765 in partial epilepsy patients was stopped after it failed to reach its primary endpoints [175,178]. At this time, the future of this therapeutic strategy remains uncertain.

While intense pharmaceutical efforts are directed toward inhibiting pyroptosis, it is also possible to induce inflammasome activation using the anticancer drug Val-boroPro (Talabostat, BXCL701) and similar agents. Val-boroPro is a nonselective inhibitor of post-proline cleaving serine proteases, including DPP8 and DPP9. While it was initially proposed that Val-boroPro treatment promoted caspase-1 activation directly, it has more recently been shown that inhibition of DPP8/9 activates the NLRP1 and CARD8 inflammasomes, leading to downstream caspase-1 activation and pyroptosis [86,174,179,180]. The mechanism for DPP8/9 regulation of inflammasome formation is still incompletely understood, with data supporting roles for both binding and enzymatic activity to repress CARD8 and NLRP1 [180,181]. In line with CARD8 being ASC-independent in humans, Val-boroPro appears to cause low-level caspase-1 and GSDMD cleavage. The therapeutic implications of this inefficient cleavage remain unclear. Yet, Val-boroPro shows efficacy in mouse models of AML. Earlier studies demonstrated immune activation against cancers in multiple mouse models [182,183]. However, this agent also kills normal B and T lymphocytes and CD34-positive hematopoietic progenitor cells in vitro [86,184]. Likewise, NLRP1 activation kills hematopoietic stem cells [185], raising questions about the therapeutic window for this strategy. Val-boroPro and a distinct inflammasome activator are in early clinical trials in solid tumors (https://clinicaltrials.gov/ct2/show/NCT03910660, https://clinicaltrials.gov/ct2/show/NCT03444753) [186]. Continued research may identify therapeutically tractable approaches to inflammasome activation that will break immune tolerance in solid tumors or directly treat cancers via induction of pyroptosis.

Targeting GSDMs: GSDMD Inhibitors

As a convergence point for upstream inflammasome pathways and as a conduit for Il-1 family cytokine and DAMP release, GSDMD represents a logical drug target for disease with inflammasome-mediated pathology. Based on the protective effects of GSDMD knockout across numerous murine inflammatory disease models, therapeutic inhibition of GSDMD may become an efficacious strategy to prevent inflammasome-mediated pathology.

Several inhibitors of GSDMD-mediated pyroptosis have been identified over the past few years. The first direct inhibitor of GSDMD, the cysteine-reactive alkylating agent necrosulfonamide(NSA), was found to bind directly to GSDMD at C191 and consequently to inhibit oligomerization and pore formation by GSDMD^NT [44]. Other groups have since shown that NSA may also inhibit upstream steps on the way to GSDMD pyroptosis, such as caspase-1 activation and GSDMD cleavage [45,187]. While NSA shows efficacy in mouse models of sepsis[44], it lacks specificity for GSDMD, as it was originally identified as an inhibitor of human (but not murine) mixed-lineage kinase domain-like (MLKL) protein [188].

Identification of other small molecules that target C191 underscores the importance of this residue for GSDMD function and its potential as a site for rational drug development. Among them, disulfiram (DSF), a drug used to treat alcohol dependence, was identified through a liposome-leakage based screen as an inhibitor of GSDMD. Similar to NSA, it was found to bind directly to C191 on GSDMD and prevent IL-1β release as well as lytic cell death. Both DSF and NSA decreased serum cytokine levels and extended survival of wild-type mice after a lethal dose of LPS [44,45]. Likewise for both compounds, mutation of C191 in GSDMD reduces their ability to inhibit pyroptosis. However, both compounds additionally have upstream inhibitory effects. While GSDME lacks a residue corresponding to C191, DSF apparently inhibits GSDME processing by an uncertain mechanism [189].

A third inhibitor of GSDMD driven pyroptosis is the MS drug dimethyl fumarate (DMF). Endogenous fumarate, which accumulates in activated macrophages, causes GSDMD succination at C191 and other sites, which impairs inflammatory caspase binding [57]. The fumarate analog DMF functions similarly. Murine treatment with DMF protected mice against a lethal dose of LPS and FMF activating mutations in MEVF, as well as showing efficacy in an EAE model. Mass spectrometric analysis identified 5 cysteines in humans and 7 in mice that are modified by DMF. Furthermore, it was found to similarly modify GSDME at C45 and other sites, resulting in GSDME blockade. While this result again highlights the lack of specificity of GSDMD inhibitors, this dual inhibition could be beneficial given the compensatory role for GSDME in IL-1β release [57,64].

The non-cysteine-reactive molecule LDC7559 was identified through screening to block both PMA and cholesterol crystal-induced NET formation in neutrophils and silica-induced IL-1β release in human monocytes [138]. While it associates with GSDMD via affinity chromatography, direct binding was not verified and the compound did not appear to inhibit GSDMD^NT activity in a liposome leakage assay [45]. Lanthanides, trivalent cations with broad channel blocking activity, also inhibit inflammasome-driven propidium dye influx and pyroptosis [190]. The recent recognition of important acidic residues within the GSDMD channel interior may help to explain these effects [191]. However, lanthanides did not significantly impair IL-1β release. Further studies are needed to validate the precise mechanism of action for these agents.

It is important to remember that the efficacy of GSDMD inhibition may be disease specific. For example, GSDMD exerts anti-inflammatory properties in some contexts and is critical for clearance of certain pathogens. Contradictory effects of GSDMD deletion have been seen in models of IBD [192,193]. Moreover, our recent work demonstrates that GSDME can compensate for GSDMD to release IL-1β [64]. Furthermore, other studies have shown that GSDMD loss does not prevent IL-1β release or cell death in some inflammasome-driven pathologies, suggesting that not all conditions resulting in inflammasome activation will benefit from therapeutic targeting of GSDMD alone [187]. The extensive ongoing efforts toward understanding GSDMD biology are likely to determine diseases where GSDMD inhibition will be of benefit.

Currently, only molecules targeting GSDMD (some with additional activities as above) have been reported. However, advances in the biology of other GSDMs are likely to reveal novel agents that inhibit other GSDM family members. Efforts will likely include pharmacological attempts to specifically target the C191 site in GDSMD and homologs. The recent identification of a caspase-1-GSDMD^CT exosite binding interface may represent an additional site for rational drug design [51]. Other GSDMs may have similar or unique protease interaction sites that could serves as scaffolds for structure-based inhibitor development. Emerging data show that electrostatic forces additionally control mature IL-1β release via GSDMD pores, demonstrating an additional potential targetable feature [191]. Finally, pathway targeting agents that alter GSDM PTMs may exhibit therapeutic effects in GSDM-mediated pathologies.

Targeting GSDMs: Therapeutic Activation

Given the transformational nature of recent therapeutic advances harnessing the immune system to treat cancer [194], the excitement surrounding GSDM functions in anti-tumor immunity is justified. This burgeoning area of GSDM biology includes several approaches to activate GSDMs for cancer therapy. For example, GSDME is hypermethylated in many cancers and its expression can be induced by hypomethylating agents such as decitabine [63,195,196]. As mentioned earlier, research into other GSDMs has identified positive regulatory factors for other GSDMs that may present additional means to enhance GSDM expression by cancer cells. These inducing factors include interferons and TNF-α for GSDMB [32], STAT3 among other stimuli for GSDMC [35], and IRF2 for GSDMD [197]. Agents to enhance GSDM expression in cancers may work alone or in combination with therapeutics to activate appropriate GSDM-cleaving proteases.

Additionally, enforced expression or selective delivery of GSDMs can induce ICD. Established experimental techniques to this end include adenoviral expression of GSDMD^NT under a tumor specific promoter and delivery of nanoparticles conjugated to a modified active GSDMA3 molecule that co-accumulates in tumors with its releasing agent [198,199]. Of note, one of these studies showed that pyroptosis of fewer than 15% of tumor cells was sufficient to promote complete tumor regression. If appropriate specificity can be achieved to prevent off-target pyroptosis, this finding suggests that similar approaches may be able to activate an anti-tumor immune response without causing tumor lysis syndrome or CRS.

Beyond target specificity, one great hurdle for GSDM activation as a cancer therapeutic will be the avoidance of normal tissue damage and systemic inflammatory complications. The initial description of GSDME pyroptosis showed that it contributed to the tissue toxicity of chemotherapy [63]. Subsequent work showed that GSDME contributed to CRS upon CAR T cell therapy [119]. Likewise, while DPP8/9 inhibitors can induce pyroptosis in AML blasts, they also show on-target toxicity toward normal hematopoietic cells [86]. Thus, whether such strategies can be safely applied will be an important area of research moving forward. Of note, loss-of-function mutations in GSDMs are seen in cancer [120]. Such mutations would be predicted to impair anti-tumor immunity.

Targeting Downstream Processes and Effectors

Billed as a GSDMD inhibitor, punicalagin, a complex polyphenol found in pomegranate extract, was found to prevent ATP-induced IL-1β secretion and pyroptosis. The mechanism by which it inhibits cell death is unknown. However, it was also shown to protect against detergent lysis and may inhibit GSMD membrane insertion via lipid membrane stabilization [200]. Glycine, a cytoprotective agent, has long been known to prevent pyroptosis by an uncertain mechanism that likewise occurs downstream of GSDMD [190,201]. Further studies are needed to clarify the targets of these agents.

Cellular mechanisms in place to repair membrane integrity can be engaged upon insertion of GSDMD pores. The endosomal sorting complex required for transport (ESCRT) machinery is activated by GSDMD-dependent Ca²⁺ influx to excise GSDMD pore-containing membrane vesicles. However, it seems that pyroptosis proceeds when a critical threshold of GSDMD pores accumulate [202]. At this time the involvement of other membrane repair pathways and the contribution of the ESCRT pathway to forestalling pyroptosis downstream of other GSDMs remain to be determined [203]. Additionally, a recent report showed that the small transmembrane protein NINJ1 facilitates a common pathway for plasma membrane rupture downstream of both pyroptotic and apoptotic stimuli [204]. The therapeutic potential of these downstream pathways is uncertain but merits further investigation. Inhibiting ESCRT may prevent replication of bacteria through rapid pyroptosis or cause a dramatic systemic inflammatory response syndrome. Inducing ESCRT pathway activity will allow prolonged inflammatory cytokine egress through GSDMD pores prior to repair which may also result in a hyperactive state [205]. Therefore, the effects of ESCRT modulation on systemic immune responses need to be addressed in appropriate models.

Other Ca²⁺ mediated events downstream of GSDMD pore formation include activation of calpains (whose targets include vimentin and IL-1α), TMEM16F (which can accelerate coagulation via tissue factor-phosphatidylserine interaction) and phospholipases [135,206–208]. Oligomerization of GDMD appears to require Ca²⁺ and this step can be blocked by supraphysiological levels of Mg²⁺ [209]. The diversity of processes engaged downstream of GSDMD pores further highlights its critical placement as a limiting step in inflammasome-driven mediator release and pyroptosis. The utility of targeting these downstream events individually is not known.

As mentioned above, several monoclonal antibody therapies targeting inflammatory cytokines released upon inflammasome activation have become clinically available for treatment of inflammatory diseases. These agents come with their own drawbacks, including infection risk, route of administration by injection, lack of central nervous system penetration and specificity for a single downstream mediator. Still, the burden of proof will fall on the field to demonstrate acceptable safety and efficacy of inflammasome- and GSDM-targeted therapies before such agents can overtake non-specific anti-inflammatory drugs and these biologic agents as mainstays of treatment for inflammatory diseases.

Future directions

The discovery of pore-forming ability of GSDMD in 2015 answered an important question in the field of inflammasome biology. It simultaneously opened new avenues of inquiry and created opportunity for the development of novel therapeutics for a wide variety of diseases. A robust body of research has now determined regulatory mechanisms and highlighted potential translational relevance for many of the GSDMs. Critical next steps for inflammatory diseases will include the development of specific GSDM inhibitors, the determination of appropriate disease contexts and the evaluation of safety and efficacy of targeting at each step outlined in this review. For cancer, the contribution of GSDMs to ICD and anti-tumor immunity should be determined on a global scale. These data may lead to personalized approaches to anti-tumor immunity involving combinatorial therapies targeting GSDM-expression in tumors. Less than a decade since the initial discovery of GSDMD pore formation, the broad potential of therapeutic modulation of GSDM-mediated pyroptosis is just beginning to be realized. Much work remains to be done, yet the promise of these strategies is undeniable.

Highlights.

1. Gasdermins have emerged as critical effectors of inflammatory mediator release and pyroptosis

2. The central position of gasdermins as a bottleneck between diverse upstream activating pathways and release of myriad effectors makes them a prime target for therapeutic intervention

3. This review will discuss the known function and regulation of gasdermins in human disease and describe emerging strategies to target gasdermin-mediated pyroptosis therapeutically

Abbreviations

AML

Acute myeloid leukemia

ASC

Apoptosis-associated speck-like protein containing a CARD

CAPS

Cryopyrin-associated periodic syndrome

CAR

Chimeric antigen receptor

CARD

Caspase activation and recruitment domain

CHIP

Clonal hematopoiesis of indeterminate potential

CRS

Cytokine release syndrome

CT

C-terminal/C=terminus

CTL

Cytotoxic T lymphocyte

DAMP

Danger-associated molecular pattern

DIC

Disseminated intravascular coagulation

DMF

Dimethylfumarate

DPP9

Dipeptidyl peptidase 9

Ds

Double-stranded

DSF

Disulfiram

EAE

Experimental autoimmune encephalitis

ELANE

Neutrophil elastase

ESCRT

Endosomal sorting complex required for transport

FCAS

Familial cold auto-inflammatory syndrome

FIIND

Function to find’ domain

FKLC

Familial keratosis lichenoides chronica

FMF

Familial Mediterranean fever

GSDM

Gasdermin

IBD

Inflammatory bowel disease

ICD

Immunogenic cell death

JRRP

Juvenile recurrent respiratory papillomatosis

MDS

Myelodysplastic syndrome

MLKL

Mixed-lineage kinase domain-like

MS

Multiple sclerosis

MSPC

Multiple self-healing palmoplantar carcinoma

NET

Neutrophil extracellular trap

NLR

Nucleotide-binding domain (NBD), leucine-rich repeat (LRR)-containing

NOMID

Neonatal-onset multisystem inflammatory disorder

NSA

Necrosulfonamide

NT

N-terminal/N-terminus

PLC

Phospholipase C

PTM

Post-translational modification

PYD

Pyrin domain