Pyroptosis was first observed in 1986, when Friedlander reported that Anthrax lethal toxin induced rapid and massive cell death in mouse peritoneal macrophages (1). Similar observations were made by Zychlinsky and colleagues in macrophages challenged with Shigella flexneri and Salmonella typhimurium (2, 3), and their studies revealed that those pathogen-induced macrophage deaths featured caspase-1 activation and were accompanied by IL-1β and IL-18 release (2, 4), but these cell deaths were classified into apoptosis (2, 3). In 2000, Cookson and Brennan reported Salmonella-induced cell death as caspase-1-dependent necrosis (5), which exhibits distinct features compared to apoptosis. Pyroptosis displays giant membrane ballooning, while apoptosis exhibits cell shrinkage and small blebbing. In 2001, they named this proinflammatory cell death “pyroptosis”, in which the Greek root “pyro” means fire or fever, and “ptosis” denotes falling (6). Later studies confirmed that Anthrax lethal toxin treatment, Shigella challenge and Salmonella infection all induce pyroptosis in macrophages (7, 8). Thus, pyroptosis could be originally defined as a caspase-1-mediated inflammatory cell death triggered by sensing of invasive pathogens and danger signals in macrophages. Later on, this pyroptosis was also reported in other immune and non-immune cells, such as epithelial cells and cardiomyocytes.

This classical caspase-1-dependent pyroptosis is mediated by inflammasome, manifesting IL-1β/IL-18 secretion and cell membrane rupture. The inflammasome is a multi-protein platform responsible for caspase-1 activation, proposed by Tschop and coworkers in 2002 (9). Recognition of pathogen-associated molecular patterns (PAMPs) or danger-associated molecular patterns (DAMPs) by cytosolic pattern recognition receptors (PRRs) initiate the assembly of inflammasome, during which the PRR recruits an adaptor protein called ASC (Apoptosis-associated speck-like protein containing a CARD) as well as pro-caspase-1 to form a supramolecular complex, the inflammasome. ASC contains a pyrin domain (PYD) and a caspase recruitment domain (CARD), bridging the upstream PYD-containing PRR receptor and downstream CARD-containing caspase-1. The proximity-induced auto-processing of pro-caspase-1 leads to its activation, and active caspase-1 further maturates IL-1β and IL-18, and triggers pyroptosis (10). The most studied inflammasomes include but are not limited to NLRP1b inflammasome that is activated by lethal toxin (11, 12), AIM2 inflammasome for cytoplasmic double-strand DNA recognition (13–15), NLRP3 inflammasome that detects multiple danger signal molecules such as crystalline material, extracellular ATP and pore-forming toxins (16, 17), NLRC4 inflammasome that recognizes bacterial flagellin or type III secretion system components (18, 19), and Pyrin inflammasome that monitors the modification or inactivation of Rho GTPases by pathogens (20).

In addition to caspase-1-dependent canonical inflammasomes, a non-canonical inflammasome pathway has recently been identified (21–23), in which human caspase-4/5 and mouse caspase-11 act as cytosolic receptors for lipopolysaccharides (LPS) (24), a cell wall component from Gram-negative bacteria. Upon LPS binding, pro-caspase-4/5/11 oligomerize and are auto-activated to trigger pyroptosis. Although caspase-4/5/11 do not process IL-1β or IL-18, secretion of those cytokines are also detected during non-canonical inflammasome activation, which could be attributed to the secondary activation of NLRP3 inflammasome (25–27). The extensive studies on inflammasome pathways have been reviewed in detail (10, 28) and will not be further discussed here.

Inflammasome-activated pyroptosis (IAP) is a primary defense mechanism against intracellular bacterial infections (29). During lytic pyroptosis, intracellular pathogens, immune alarmins and other cellular components are released from the dead cells. Pyroptosis promotes the clearance of bacteria by disrupting the bacterial intracellular replication niche (30), directly damaging or eliminating the bacteria (31, 32), and enhancing host immune responses such as neutrophil-mediated bacterial killing (30, 31). Nonetheless, the molecular mechanism underlying pyroptosis remained unknown for decades until the seminal discovery of gasdermin D (GSDMD) in 2015.

Through unbiased genome-wide CRISPR-Cas9 screens and a forward genetic screen with ethyl-N-nitrosourea-mutagenized mice, respectively, Shao and Dixit group independently identified gasdermin D (GSDMD) as the executor of pyroptosis in 2015 (33, 34). GSDMD is cleaved after D276 by inflammatory caspase-1/4/5/11, and the resulting GSDMD N-terminal (NT) fragment is required for pyroptosis. Later on, Shao, Lieberman and other groups discovered that GSDMD-NT induces pyroptosis by forming membrane pores (32, 35–38). GSDMD consists of a pore-forming N-terminal domain (GSDM-NT) and an autoinhibitory C-terminal domain (GSDM-CT), connected by a flexible linker. The cleavage of GSDMD in the linker by inflammatory caspases unleashes GSDM-NTs, which oligomerize and perforate on the plasma membrane, leading to the disruption of the plasma membrane and pyroptosis. The structure of the GSDMD-NT membrane pore further validated these discoveries (39).

Human GSDMD belongs to a gasdermin family of six proteins, namely GSDM A-D, GSDME (a.k.a. DFNA5) and DFNB59 (40). Gasdermin A was initially named in 2000 based on its selective expression in the gastrointestinal tract and dermis (41). In mice, there are no GSDMB counterpart but 3 gsdmas (gsdma 1-3) and 4 gsdmcs (gsdmc 1-4). GSDM A-E exhibit pore-forming activity because overexpressing the NT domains of those GSDMs trigger pyroptosis in HEK293T cells (36). GSDM-mediated pyroptotic cell death can be triggered by proteases other than inflammatory caspases, including apoptotic caspases, serine protease granzymes secreted by cytotoxic lymphocytes, proteases from neutrophil granules and pathogens (40). For instance, cysteine protease SpeB from Streptococcus pyogenes cuts GSDMA to induce pyroptosis in keratinocytes (42, 43). GSDMB can be cut and activated by GzmA (44). Both GSDMC and GSDMD could be cleaved and activated by caspase-8 (45–47), and GSDMD is also cleaved by neutrophil elastase (48, 49) and cathepsin G (50), which are implicated in neutrophil pyroptosis (netosis). GSDME is processed and activated by both Caspase-3 (51, 52) and GzmB (53, 54). More details can be found in other reviews (40). All those proteolytic cleavages similarly release the pore-forming GSDM-NTs to activate pyroptotic cell death. Therefore, pyroptosis has been redefined as gasdermin-mediated programmed necrotic cell death (55, 56), which means any cells expressing GSDMs could undergo pyroptosis, no matter through what activating mechanisms. From another perspective, GSDMs act as intracellular sensors of mislocalized cytosolic proteases. If considering SpeB as a PAMP from bacteria and Gzms as DAMPs from within, the GSDMs function as unconventional PRRs that sense danger-associated proteases and alarm the immune system by triggering pyroptosis. It will also be interesting to investigate whether GSDMs could be activated through mechanisms other than proteolytic cleavage.

The identification of gasdermin family proteins and their pore-forming activities decouples the inflammasome pathways and pyroptosis, which sets the stage for subsequent discoveries of various forms of pyroptosis that are implicated in distinct biological and pathological processes, like cancer. Although GSDMD-mediated IAP has attracted considerable attention, an era of cancer-associated pyroptosis (CAP) mediated by other GSDM proteins is in sight ( Figure 1 ). In contrast to IAP, this CAP is independent of inflammasome or caspases. It occurs in tumor cells but not immune cells and is essential in igniting anti-tumor immunity.

Cancer-associated pyroptosis, rather than immunosilent apoptosis, not only directly kills some susceptible tumor cells, but also elicits robust tumor-specific immune responses, leading to the control or rejection of the entire tumor (54, 72). The role of pyroptosis in anti-tumor immunity makes GSDM proteins attractive therapeutic targets for cancer treatment. Because GSDME could be cleaved and activated by caspase-3, multiple caspase-3-activating chemotherapy drugs have been used to activate GSDME pyroptosis in tumors (92). For instance, BRAF inhibitors plus MEK inhibitors treatment can activate GSDME pyroptosis to enhance anti-tumor immunity in GSDME-expressing melanoma (62). Similar strategies have also been reported for pyroptosis mediated by other GSDMs. As mentioned, α-KG has shown its ability to trigger GSDMC-mediated pyroptosis and suppress tumor growth (47). Serine dipeptidase DPP8/9 inhibitors, which activate GSDMD-mediated pyroptosis in human myeloid cells, induce pyroptosis in most human acute myeloid leukemia cell lines and primary cells but not cells from other lineages, presenting a novel therapeutic strategy for acute myeloid leukemia (93). Multiple nano-technology-based strategies have also been developed to activate pyroptosis in tumor cells (94, 95).

Nonetheless, it is well documented that tumors can evade CAP by suppressing the expression of GSDMs, which hampers the usage of pyroptosis-activating drugs. GSDME is suppressed by promoter hypermethylation in gastric, colorectal and breast cancers (41, 59, 60). GSDMA and GSDMB are similarly silenced by methylation in some cancer cell lines (96, 97). DNA methyltransferase inhibitor 5-aza-2’-deoxycytosine (decitabine) upregulates GSDME expression and restores pyroptosis in culture cells (52, 60). Decitabine combined with chemotherapy nano-drugs triggers GSDME-mediated pyroptosis and enhances the efficacy of chemotherapy in mice (98). Thus similar strategies might also be used to increase the expressions of other GSDMs in tumors. In addition, some cytokines, chemokines and other signaling molecules could upregulate the expression of GSDMs. For instance, GSDMA is upregulated by transforming growth factor-β (TGF-β) (99), GSDMB by interferons and TNF-α (44), and GSDME by corticosteroid dexamethasone and forskolin (100). It is worth testing whether these GSDM transcriptional modulators could be combined with pyroptosis-inducing drugs to de-repress CAP and potentiate tumor-specific immune responses.

Regardless of endogenous GSDM expressions and activating mechanisms, strategy to directly introduce active GSDM proteins into tumor cells has also been proposed. A recent study described nanoparticle delivery combined with the desilylation-based release of GSDMA3-NT in tumors (72). Phenylalanine trifluoroborate (Phe-BF3) is a cancer-imaging probe that can enter cells to desilylate and ‘cleave’ a designed linker that contains a silyl ether. GSDMA3-NT proteins were conjugated to nanoparticles by a triethylsilyl linker, which can be cleaved by Phe-BF3. Those nanoparticles and Phe-BF3 were both concentrated in tumors, leading to the selective release of GSDMA3-NT in tumor cells to trigger CAP. Interestingly, pyroptosis of around 15% of tumor cells was sufficient to reject the entire transplanted tumor in a T cell-dependent manner, confirming the potential of CAP in activating anti-tumor immunity and eliminating tumors.

While there are great therapeutic benefits of targeting GSDMs and pyroptosis for tumor suppression, the safety of those strategies needs to be carefully evaluated. Uncontrolled and excess activation of pyroptosis may cause severe tissue damage and other adverse effects. DPP8/9 inhibitors, which induce pyroptosis in acute myeloid leukemia, also activate pyroptosis in other cells, such as B cells and CD34+ cells (93) and augment treatment-associated toxicity. Nonspecific activation of GSDME-mediated pyroptosis in normal tissues resulted in extensive tissue damage upon chemotherapy treatments in mice (52). Secondary activation of GSDMD pyroptosis by CAP alarmins in tumor-associated macrophages lies at the root of CAR-T cell-triggered cytokine release syndrome, a severe adverse effect of CAR-T immunotherapy (53). Those observations represent the possible limitations of pyroptosis-based immunotherapeutic strategies. Thus, CAP induction in vivo needs to be finely controlled to achieve the specific targeting of tumor cells, which is a major challenge for pyroptosis-based immunotherapeutics.

Identifying GSDMs and their pore-forming activity has revolutionized our understanding of pyroptosis. Pyroptosis has been redefined as a gasdermin-mediated programmed necrotic cell death. In addition to classical inflammasome-activated pyroptosis, recent studies have identified many novel forms of pyroptosis mediated by different GSDMs and activated by various proteases, most of which occur in tumor cells. Pyroptosis in tumors is a highly immunogenic cell death that activates anti-tumor immune protection. Therefore, inducing pyroptosis in tumor cells could have therapeutic utility. Nonetheless, many questions about cancer-associated pyroptosis remain elusive. 1) How does pyroptosis activate anti-tumor immunity? It has been shown that CAP is associated with more active tumor-infiltrating immune cells such as dendritic cells, CD4 and CD8 T cells and NK cells (54, 62), which lead to the control or rejection of the implanted tumors. However, the exact mechanism underlying how pyroptosis recruits and activates those immune cells remains largely unknown. 2) How do cancer cells evade anti-tumor pyroptosis? It is well documented that GSDME is silenced or mutated in multiple tumors (41, 59, 60, 71). However, GSDMB is overexpressed in multiple tumors, and it is controversial if it acts as a pore-forming protein (44, 73, 101). Further investigation is required to clarify if GSDMB-mediated pyroptosis activates anti-tumor immunity and how tumors evade this CAP. 3) How is pyroptosis regulated in cells? The regulation of IAP has been extensively studied (10, 102), but how cancer-associated pyroptosis, mediated by GSDMB, C, and E, is regulated remains elusive and awaits further investigation. 4) How can we harness pyroptosis to ignite anti-tumor immunity to treat tumors, and how can we block pyroptosis-associated toxicity? As we discussed, many strategies have been proposed to employ pyroptosis to kill tumor cells and activate the tumor-specific immune response. How to efficiently and specifically induce pyroptosis in tumor cells is the key to suppressing tumor growth but reducing pyroptosis-associated toxicity. Tumor-targeting GSDM agonists could be a possible way to address the problem.

ZZ and QK wrote and revised the manuscript. Both authors contributed to the article and approved the submitted version.