Pyroptosis-induced inflammation and tissue damage

Abstract

Programmed cell deaths are pathways involving cells playing an active role in their own destruction. Depending on the signaling system of the process, programmed cell death can be divided into two categories, pro-inflammatory and non-inflammatory. Pyroptosis is a pro-inflammatory form of programmed cell death. Upon cell death, a plethora of cytokines are released and trigger a cascade of responses from the neighboring cells. The pyroptosis process is a double-edged sword, could be both beneficial and detrimental in various inflammatory disorders and disease conditions. A physiological outcome of these responses is tissue damage, and sometimes death of the host. In this review, we focus on the inflammatory response triggered by pyroptosis, and resulting tissue damage in selected organs.

1. Introduction

Pyroptosis is a form of pro-inflammatory programmed cell death [1]. The term “pyroptosis” was first coined by Cookson and Brennan in 2001, to describe “the screaming, alarm-ringing pro-inflammatory death of a potentially dangerous cell in an organism” [2]. The term is constructed from the Greek roots “pyro” for fire or fever and “ptosis (to-sis)” for falling. Although discovered relatively recently, pyroptosis has been observed to occur in different types of cells ranging from the immune system, digestive system, central nervous system, reproductive system, to cardiovascular system. Pyroptosis is provoked by stimuli including damage-associated molecular patterns (DAMPs) such as stress signals, uric acid crystals, oxidized lipoproteins, or pathogen-associated molecular patterns (PAMPs) such as flagellin, type 3 secretion system (T3SS) structure proteins, lipopolysaccharide (LPS), and nucleic acids [3].

Pyroptosis has initially been defined as a process of “caspase 1-dependent programmed cell death” [3]. However, later other caspases have been reported to lead to pyroptosis as well. Intracellular LPS from Gram-negative bacteria activates caspase-11 (in mice) and caspase-4/5 (in human), leading to pyroptosis [4–9]. More recently, caspase-8 has been shown to be activated by a bacterial virulent effector YopJ and flagellin, and subsequently cleaves both gasdermin D (GSDMD) and gasdermin E (GSDME) in murine macrophages, resulting in pyroptosis [10, 11]. In cancer chemotherapy, caspase-3 can cleave GSDME to induce pyroptosis in certain GSDME-expressing cancer cells [12]. GSDME serves as a switch molecule in the transformation between apoptosis and pyroptosis in tumor cells [12–14]. GSDME is cleaved by caspase-3, an apoptotic caspase activated by intrinsic and extrinsic apoptotic pathways [14]. Following cleavage by caspase-3, the N-terminal domain of GSDME behaves very much like the GSDMD N-terminal domain, forming pores in plasma membrane and induce pyroptosis [14]. In tumor cells with low GSDME expression, the hypermethylation of its promoter suppressed GSDME production. During chemotherapy using the DNA methyltransferase inhibitor decitabine, the hypermethylation of the GSDME promotor was inhibited and GSDME expression level increased. This change of GSDME level promoted the occurrence of pyroptosis in tumor cells in a caspase-3-dependent manner [12]. Other chemotherapy drugs that induce caspase-3-mediated apoptosis in GSDME-negative cell lines were shown to induce pyroptosis in GSDME-positive cell lines [12]. Due to the involvement of multiple caspases, more recently pyroptosis is defined as “Gasdermin-Mediated Programmed Necrotic Cell Death”, as gasdermin activation and pore formation in the cell membrane are features shared by both canonical (Caspase-1 dependent pyroptosis) and non-canonical (depending on caspases other than caspase-1) pyroptosis [15]. In addition to the difference in triggering signals and caspases involved, pyroptotic cells undergo membrane blebbing, swelling, and flattening before cell lysis, different from the explosive rupture observed in necroptotic cells, or the shrinkage and withering of apoptotic cells [16].

2. Mechanism of pyroptosis

The mechanism of pyroptosis has been the topic of several excellent reviews [3, 15, 17, 18] (Figure 1). Mammalian cells have sensors and receptors on their surface to detect environmental cues. DAMPs and PMAPs activate pattern recognition receptors (PRRs) on the cell surface [19]. Upon stimulation by extracellular signals, a cascade of responses inside the cells are triggered, including the activation of caspases and transcriptional effectors such as nuclear factor-κB (NF-κB). Activated NF-κB promotes the expression of pro-inflammatory cytokines, which are activated by caspases and secreted to escalate immune response. Activated NF-κB also upregulates the expression of NLPR3, which is critical for the activation of the NLRP3 inflammasome. In parallel, certain bacteria such as Salmonella typhimurium, Shigella flexneri, Pseudomonas aeruginosa, Vibrio parahaemolyticus, Burkholderia pseudomallei, and Yersinia enterocolitica can directly deliver effector proteins into the cytoplasm of host cells through the T3SS [20, 21]. Another route of entry of pathogenic signal molecules to the cytoplasm is endocytosis [22]. Together with host-derived cytoplasmic signal molecules, they interact with intracellular receptor NAIPs (NLR family apoptosis inhibitory proteins) and NLRs (Nod-like receptors) to trigger the activation and assembly of multiple protein complexes called inflammasomes [23, 24].

Figure 1. Inflammasome activation and pyroptosis.

Various PAMPs and DAMPs are recognized by both membrane and cytoplasmic receptors, leading to a series of signaling cascades, expression of cytokines and inflammasome sensers and adaptors, and assembly of either canonical or non-canonical inflammasome. As a result, caspases are activated and subsequently cleave gasdermins (primarily GSDMD) and pro-cytokines IL-1β and IL-18. Pore-formation and release of cytokines and tissue factor lead to cell lysis, inflammation, coagulation, and tissue damage. Created with biorender.com.

Several types of inflammasomes have been found to be involved in pyroptosis, including NLRC4 (NOD-like receptor family, CARD domain containing 4), NLRP3 (NOD-like receptor family, pyrin domain containing 3), NLRP1, NLRP6, NLRP9, AIM2, and PYRIN [25]. The NLRP3 inflammasome is unique in that it can be activated by internal stress signals and a large number of metabolites, responding to both DAMPs and PAMPs [26, 27], and is involved in infections as well as several inflammatory disorders [28–30]. Activation of the NLRP3 in macrophages is a two-step process involving priming and receptors including TLRs, IL-1R, and tumor necrosis factor receptor that signals via NF-kB-activating pathways [31, 32]. In certain cell types (such as monocyte), priming alone can activate NLRP3 inflammasome [33, 34]. How NLRP3 detects intracellular stimuli remains elusive. Direct binding to the stimuli seemed unlikely, since a variety of stimuli with drastically different structure can activate the NLRP3 inflammasome [35]. It has been suggested that NLPR3 could sense a common “cellular event” induced by its stimuli. While certain stimuli share a common feature such as the stimulation of K⁺ efflux or Ca²⁺ signaling, an event that is shared by all stimuli is yet to be determined [28]. ROS has been suggested as a mediator of NLRP3 activation, as ROS inhibitors were found to potently inhibit NLRP3 activation [36, 37]. Later it was found that ROS inhibitors block NLRP3 inflammasome activation through blocking the upregulation of NLRP3 expression, or the priming step instead [38]. In 2018 Chen and Chen proposed a role of the trans-Golgi network (TGN) as a common feature in mediating NLRP3 activation [39]. They found that stimulation of various NLRP3 stimuli led to the dissemble of the TGN into small vesicles, which they termed dTGN. Formation of dTGN was upstream of NLRP3 binding and activation. NLRP3 was recruited to bind and aggregate on the surface of dTGN through ionic interactions between a region containing polybasic residues in NLRP3 and the negatively charged phospholipid PtdIns4P on dTGN, forming multiple small puncta. These small puncta were then incorporated to form a large speck with the adaptor protein ASC, followed by the subsequent downstream inflammasome activation cascade. How these stimuli trigger the dissembling of the TGN remains elusive.

The NLRC4 inflammasome is mainly triggered by PAMPs from various microbial pathogens [40]. Oligomerization and assembly of NLRC4 inflammasome are triggered by PAMP binding to NAIPs [41]. While a mouse carries seven NAIPs, a human only has one [42]. Mouse NAIPs recognize and bind to specific ligands. For example, NAIP5 and NAIP6 detect flagellin, and NAIP1 and NAIP2 recognize T3SS needle and rod proteins, respectively [41, 43, 44]. Human NAIP binds with both flagellin and T3SS proteins [41, 45]. Binding of the PAMP signal leads to conformational changes in the corresponding NAIP, exposing a previously hidden surface, which promotes the interaction with NLRC4 and activation of the NAIP/NLRC4 inflammasome. NLRC4 contains an N-terminal CARD domain, which can recruit pro-caspase-1 and assemble inflammasome.

Mice carry three NLRP1 paralogs (NLRP1a-c), while human only carry one. Among them NLRP1b is the most well characterized and is activated by bacterial and parasitic toxins and virulent factors [46–48]. Activation of the NLRP1b inflammasome is initiated by the autoproteolysis within the Function to Find Domain (FIIND) domain, which splits NLRP1b into a N-terminal and a C-terminal domain that remain noncovalently bound [48–50]. Several currently identified activators, including the anthrax lethal toxin protease and IpaH7.8, a Shigella flexneri ubiquitin ligase secreted effector, have been shown to initiate the ubiquitination and proteasomal degradation of the N-terminal domain, releasing the C-terminal domain to initiate inflammasome assembly [48]. This is consistent with the observation that NLRP1b activation can be blocked by proteasome inhibition in murine macrophages [51]. Thus, N-terminal degradation is likely a common mechanism of NLRP1b inflammasome activation.

Other well characterized inflammasomes involved in pyroptosis include the AIM2 and Pyrin inflammasomes. AIM2 directly binds to cytosolic double stranded DNA at least 70 bp in length in a sequence-independent manner [52–54]. DNA binding triggers inflammasome assembly. The Pyrin inflammasome indirectly senses inactivating modifications of host Rho GTPases by bacterial toxins. Rho-modifying proteins from pathogens induce pyrin-dependent activation of pyroptosis in macrophages [55–57]. In the resting state pyrin is phosphorylated at two sites and remains bound with its endogenous inhibitor. RhoA plays an important role in maintaining pyrin inhibited by facilitating phosphorylation through the recruitment of specific kinases [58]. Thus, bacterial toxins that inactivate RhoA lead to pyrin activation through disrupting the phosphorylation of pyrin and subsequently binding with its inhibitor [59].

As a direct consequence of inflammasome activation, pro-caspases are recruited and activated. Caspases that have been identified to be involved in pyroptosis include 1, 3, 4, 5, 8, and 11. Pro-inflammatory cytokines and gasdermins, primarily GSDMD, are digested by specific proteases. Gasdermin-mediated pore-formation is a key feature of pyroptosis. Activated caspases cleave GSDMD in the loop connecting the N-terminal and C-terminal domains, which lifts the autoinhibition of the C-terminal domain on the N-terminal domain. The N-terminal domain then migrates and inserts into the plasma membrane, oligomerizes and forms pores. While early studies reported a range of pore sizes, a recent cryo-EM study revealed a large pore of 21.5 nm inner diameter, with 33-fold symmetry [60]. Pore formation and cytokine release initiate a cascade of downstream consequences.

3. Consequence of pyroptosis at the cellular and subcellular level

Since its discovery in early 2000, pyroptosis has been observed to occur in many different cell types. The wide occurrence of this programed cell death process indicates its involvement in various disorders and disease conditions. Pyroptosis first occurs at the cellular level upon detection of pathogen invasion or cell stress/danger signals. Pyroptotic cells then release a large amount of pro-inflammatory cytokines and other mediators, triggering inflammatory responses from neighboring cells, similar as sparks that eventually lead to the burning of the entire forest. Distinct features of pyroptosis include the dependence on protease activation and cleavage of gasdermins, pore-formation in the plasma membrane, rapid cell swelling and lysis, and release of proinflammatory cytokines and microvesicles [3, 61]. Among them, pore formation is the first step of the physical changes of the pyroptotic cells. Pore formation leads to the release of pro-inflammatory cytokines, disruption of the ionic gradients and osmostasis, and eventually cell lysis and release of cellular contents [62]. Release of the inflammatory cellular contents and the cascade of reactions lead to tissue damage, and sometimes organ failure and host death.

Two pro-inflammatory cytokines generated directly in pyroptosis are IL-1β and IL-18. IL-1β and IL-18 lack secretion signals and their mechanism of release has been attributed to the pore-formation in the plasma membrane [62]. In a recent study, Xia et al. reported that the GSDMD pore conduit is predominantly negatively charged, which serves as an electrostatic gate to favor the passage of mature IL-1s over that of their precursors [60]. IL-1 precursors have a negatively charged acidic domain, which is proteolytically removed upon activation [63]. Mutation of acidic residues lining the conduit of the GSDMD pore was found to abolish this preference, suggesting that both charge and size are important factors determining the passage through the GSDMD pore. Similarly, a role in protein secretion has been previously proposed for caspase-1, which is likely the result of pore-formation [64]. Both cytokines play crucial roles in the pathogenesis of a range of inflammatory and autoimmune diseases [65, 66]. IL-1β binds to IL-1 receptors on the surface of immune cells and triggers various processes including fever, vasodilation, hematopoiesis, leukocyte tissue migration, antibody synthesis, and expression of cytokines and chemokines [66, 67]. IL-18 plays an important role in immune responses through inducing IFNγ production, activating T cells, macrophages, and natural killer cells, and plays a role in angiogenesis [65, 68, 69]. Both cytokines play crucial roles in promoting the beneficiary inflammatory process as well as the pathogenesis of a range of inflammatory and autoimmune diseases.

TNFα and IFNγ together had been found to induce pyroptosis and amplify pro-inflammatory responses [70]. Co-administration of TNFα and IFNγ caused a lethal cytokine shock in mice, leading to tissue damage and lethal inflammation, while treatment with neutralizing antibodies against TNFα and IFNγ protected mice from mortality during SARS-CoV-2 infection and sepsis. Additional mechanistic studies in vitro indicated co-treatment with TNFα and IFNγ induced the cleavage of GSDME and subsequent pyroptosis in BMDMs through the RIPK1/FADD/CASP8 Axis.

Activation of caspases is an important step in pyroptosis. While some substrates that are processed by activated caspases have been linked to pyroptosis, including IL-18, IL-1β, and GSDMD, other less documented substrates may also play important roles in inflammation and tissue damage. Through a proteomic study, 41 proteins were discovered that were directly cleaved by caspase-1, including chaperones, cytoskeletal and translation machinery proteins, cytokines, and interestingly, several proteins along the glycolysis pathway [71]. It was then confirmed that digestion by caspase-1 reduced the catalytic activity of these metabolic enzymes in vitro, and caspase-1 activation led to a significant digestion of these proteins in cells including macrophages. A more comprehensive set of caspase-1 substrates were identified through mass spectrometry-based proteomic studies coupled with a N-terminal enrichment method [72]. Analysis of THP-1 monocytic cell lysates treated with caspase-1 revealed 82 putative substrates, while activation of caspase-1 in THP-1 cells with proinflammatory stimuli led to the identification of 46 putative substrates, with 23 of them overlapping with the previous 82 substrates. Other than the well-established caspase-1 substrates GSDMD and IL-1β, a few substrates were identified in both studies, including the cytoskeleton actin and a few proteins involved in protein translation. The overall overlap between substrates identified through the two studies is small. This could be due to differences in sample handling and incomplete sampling.

4. Pyroptosis induced tissue damage

While pyroptosis is a host response that is evolved to protect against infection and cell stress, overactivation of pyroptosis can be detrimental, leading to tissue damage and even host death. Inflammation is initiated by harmful stimuli, as a means of self-protection by the cells. The production, activation, and release of cytokines, transcription factors, and proteases collectively promote the elimination of the harmful stimuli and protection of the host. Thus when under careful control, inflammation is a critical defense mechanism beneficial for the survival of the organism [73]. However, many molecules released as a result of inflammation could be toxic to the host cells as well, acting as a double-edged sword. When overactivated, this beneficial mechanism could cause damage to tissues and evolve into chronic inflammatory disease. IL-1β has been shown to trigger the generation of several cytokines in vivo [74]. Therefore, although IL-1β and IL-18 are the only known proinflammatory cytokines generated directly from inflammasome activation, inflammasome activation in vivo may lead to the generation of multiple proinflammatory cytokines indirectly, including TNF-α and IL-6, resulting in cytokine storm and tissue damage.

Recently, we and another research group reported that pyroptosis can trigger disseminated intravascular coagulation (DIC) [61, 75, 76], an acquired syndrome characterized by widespread intravascular activation of coagulation resulting in the formation of thrombi throughout the vasculature. We found that tissue factor, a protein playing an important role in blood coagulation, released from pyroptotic monocytes and macrophages was the initiator of coagulation cascade during sepsis. DIC can result from both infectious insults (such as COVID-19 and sepsis) and non-infectious insults (such as trauma and deep vein thrombosis) [77, 78]. DIC often leads to bleeding due to the consumption of coagulation factors and platelets, and formation of microvascular thrombi that inevitably results in multiple organ dysfunction.

Pyroptosis can occur in many cell types and affect almost all vital systems in the human body. In this special issue, the impact of pyroptosis and inflammation on liver, kidney, and the central nervous system have been reviewed, thus we will focus on system and organs that are not discussed, including the hematopoietic, cardiovascular, respiratory, digestive, and reproductive systems (Figure 2).

Figure 2.

Inflammasomes activation and pyroptosis induced organ damages and disease conditions. NLRP1, NOD-like receptor family, pyrin domain containing 1; NLRP3, NOD-like receptor family, pyrin domain containing 3; NLRC4, NOD-like receptor family, CARD containing 4; AIM2, absent in melanoma 2; NSAID, nonsteroidal anti-inflammatory drug.

4.1. Hematopoiesis

Inflammasome activation during the inflammatory response plays an important role in balancing multiple stages of hematopoietic homeostasis [79, 80]. Proper level of inflammasome activation is critical for the development of stem cell mobilization, homing and engraftment. On the other hand, pathological activation of inflammasome complex results in abnormal hematopoiesis. Early insights into the pathogenesis of abnormal hematopoiesis highlighted elevations of inflammatory cytokines as a mechanism to explain abnormal responses under infection, microenvironmental stress, or sterile inflammation [81–83]. It has been realized only recently that pyroptosis may also contribute to hematopoietic damage following inflammasome activation. Activation of inflammasome in hematopoietic cells (HSCs) may lead to irreversible pyroptotic cell death and contribute to the origin of hematopoietic malignancies. Here, we will focus on recent advances that demonstrate the impact of pyroptosis during infection and selected hematological diseases.

4.1.1. Bacterial and viral Infection

Systemic infection can result in profound alterations in bone marrow, leading to bone marrow failure. A hematological feature of many infections is peripheral blood cytopenia [84]. The development of cytopenia results in immunosuppression and can predict poor prognosis in septic patients. Direct killing of hematopoietic stem and progenitor cells (HSPCs) is effective in clearance of infected cells to limit dissemination of infection during hematopoiesis [85, 86]. Inappropriate inflammasome activation and pyroptotic cell death, however, may lead to cytopenia and immunosuppression during infection. Inflammatory cytokines such as IL-1β and IL-18 released from pyroptotic cells further impact emergency hematopoiesis [81, 82, 87, 88]. Masters et al. demonstrated that NLRP1a activation by lymphocytic choriomeningitis virus (LCMV) infection induced the death of hematopoietic progenitor cells, leading to neutrophilia and lethal inflammatory disease [86]. NLRP1a and caspase-1 deficiency in mice decreased the pyroptotic cell death of progenitor cells upon infection and led to faster recovery [86]. In consistence with this finding, deletion of caspase-1 is associated with elevated number of HSCs [89, 90]. Bone marrow cells from caspase-1 deficient mice had significantly increased longevity compared with control cells [89]. Caspase1/11 deficient neonatal mice had improved survival following bacterial challenge. This reduction in mortality is associated with increased number of HSCs in the bone marrow and spleen following bacterial challenge, suggesting a role of pyroptosis in immunosuppression during sepsis [90].

Several recent papers highlighted the role of pyroptosis during SARS-CoV-2 infection. SARS-CoV-2 infection may lead to a severe complication known as a cytokine storm, which results in uncontrolled hyperactivation of the innate immune response and organ damage [91]. It has been well established that SARS-CoV binds to the angiotensin-converting enzyme 2 (ACE2) receptor through its spike protein to invade human cells [92, 93]. A recent study showed that the SARS-CoV-2 spike protein activated NLRP3 inflammasome in HSCs and small embryonic-like stem cells, which may induce cell pyroptosis [94]. The expression of mRNA of NLPR3, IL-1β,IL-18 and ASC increased after exposure to the recombinant spike protein. Elevated levels of secreted IL-1β were detected using ELISA assay in the conditioned media from cells stimulated by spike protein. These data suggest that inflammasome was activated by the spike protein and pyroptosis likely occurred in these cells, which may contribute to the pathogenesis of SARS-CoV-2 infection.

4.1.2. Radiation induced hematopoietic damage

In addition to infection, pyroptosis also plays an important role in radiation induced hematopoietic damage. Several studies show that inflammasomes including AIM2 and NLRP3 are implicated in radiation induced bone marrow toxicity [95–98]. Stoecklein et al. reported the occurrence of pyroptosis in many types of immune cells in mice underwent whole body radiation [95]. Later, Hu et al. provided evidence of AIM2 inflammasome activation and subsequent pyroptotic cell death as a mechanism of radiation induced bone marrow toxicity [96]. AIM2 mediates pyroptosis of intestinal epithelial cells and bone marrow cells in response to double-strand DNA fragments resulted from ionizing radiation or chemotherapeutic agents. Consistent with the in vitro data, AIM2 and caspase-1 deficient mice showed decreased hematopoietic failure after total body irradiation due to lack of pyroptosis [96]. 5-Androstenediol, a natural steroid hormone, has been found to inhibit pyroptosis through disruption of the interaction between AIM2 and ASC and thus enhanced the recovery of the hematopoietic system after whole body irradiation in mice [98]. A recent study further confirmed that GSDMD, downstream of the NLRP3 and AIM2 inflammasomes, triggered signaling cascades in bone marrow in response to radiation [99].

4.1.3. Aging

The increased basal level of sterile inflammation developed in hematopoietic tissues with advanced aging is associated with increased risk of clonal hematopoiesis of indeterminate potential, myeloid neoplasia and anemia. Enhanced myelopoiesis with impaired lymphopoiesis is a hallmark of bone marrow aging [100]. Defective erythropoiesis that can progress to anemia is also common with aging [101, 102]. This alteration in hematopoiesis has been termed in the literature as “inflammaging”, which is a chronic and low-grade sterile inflammation that develops with advanced age [103]. Mitochondrial stress and DNA damage are two important mechanisms of HSC aging that result in cell death [104, 105]. Pyroptosis has been shown to play an important role in aging. Fali et al. reported an increase of pyroptosis in old HPCs, together with elevated levels of P2X7 transcripts and cleaved caspase-1 [106]. Similarly, Luo et al. showed significantly reduced HSC death in the presence of NAD-dependent deacetylase sirtuin 2 (SIRT2), an NAD+ (nicotinamide adenine dinucleotide)-dependent deacetylase [107]. In an earlier study SIRT2 was identified as among the most significantly repressed genes in old HSCs compared with young HSCs [108]. Deficiency of SIRT2 was found to increase pyroptosis and caspase-1 cleavage upon LPS and ATP induction in aged HSCs in vitro [107]. Specifically, the NLRP3 inflammasome activation was involved in this pyroptotic cell death, as inactivation of NLRP3 and caspase-1 in aged HSCs increased the reconstitution capacity and improved the differentiation into the lymphoid lineage. Furthermore, overexpression of SIRT3 and/or SIRT7, which are supposed to reduce mitochondrial oxidative stress, [104, 109] reduced caspase-1 activation. Thus in aged HSCs, reduction of SIRT2 likely enhances pyroptosis through increased mitochondrial stresses.

4.1.4. Myelodysplastic syndromes (MDS)

MDS is characterized by the clonal proliferation of HSCs, ineffective hematopoiesis, and peripheral blood cytopenia with high-risk of acute myeloid leukemia (AML) development [110]. Pyroptosis is an integrated part of this clonally heterogeneous driven process. Basiorka et al. described a common pathway that DAMP signals and/or somatic mutation were able to induce NLRP3 inflammasome activation and subsequent pyroptosis in HSPCs, resulting in the phenotypic features of MDS [111]. The expression levels of the inflammasome components, including NLRP3, caspase-1, IL-1β and IL-18, were elevated in hematopoietic stem and progenitor cells isolated from the MDS patients. Knockdown of NLRP3 or caspase-1 suppressed pyroptosis and restored effective hematopoiesis. Pyroptosis of the HSCs was triggered by an excess of S100 calcium binding protein A9 (S100A9) and subsequent generation of reactive oxygen species (ROS). During inflammation, S100A8/A9 is actively released from immunocytes, such as neutrophils and macrophages, and exerts a critical role in modulating the inflammatory response by stimulating leukocyte recruitment and inducing cytokine secretion. In a separate study, increased plasma circulation of A100A8/9 was also observed in MDS patients [112]. Overexpression of S100A9 in an S100A9 transgenic mouse model recapitulated progressive cytopenia mediated by NLRP3 inflammasome-induced pyroptosis [111]. Inhibition of inflammasome assembly suppressed pyroptosis, increased HSPCs numbers, and restored effective hematopoiesis in S100A9 transgenic mouse model. The S100A9/NOX/ROS/NLRP3 axis has been identified as the regulator of the Wnt/β-catenin signaling and pyroptotic cell death of HSPCs in MDS. Pyroptosis is clearly responsible for many of the hallmark features of MDS including macrocytosis and ineffective hematopoiesis. ASC specks in the peripheral blood and/or bone marrow plasma from patients with MDS is directly correlated with plasma concentrations of S100A8/A9, confirming the relationship to pyroptosis [113].

4.1.5. Acute myeloid leukemia (AML)

In addition to MDS, the NLPR3 inflammasome and pyroptosis have also been implicated in AML [114]. A recent study shows that the oncogenic KrasG12D mutation, which occurs in various leukemias, has an inflammation-related effect contributing to the disease [115]. KrasG12D mutation caused NLRP3 inflammasome activation, leading to increased caspase-1 cleavage and IL-1β release, and increased cell death, which subsequently promoted myeloproliferation and cytopenia in vivo. Inhibitors of the IL-1 receptor or NLRP3 activation interfered with KRAS-driven myeloproliferation. The oncogenic KRAS triggered pyroptosis through ROS from the RAC1/NADPH axis [115]. In agreement with the observations in mice, ROS production increased in human Kras^mut cells derived from AML patients, and blocking RAC1 in the Kras mutant cells derived from AML patients inhibited ROS production and IL-1β expression. Thus, these findings highlight the important role of NLRP3 inflammasome and pyroptosis in hematological disorders.

Transforming growth factor β-activated kinase 1 (TAK1), a member of the mitogen-activated protein kinase kinase kinase (MAP3K) family, is required for NF-κB and MAPK signaling downstream of several cytokine receptors and pattern recognition receptors [116, 117]. TAK1 is also a crucial component in maintaining homeostasis, as TAK1 deletion or inhibition results in spontaneous cell death in a number of different cell types, including hematopoietic cells [118, 119]. Malireddi et al. observed spontaneous receptor interacting serine/threonine Kinase 1 (RIPK1)-dependent NLRP3 inflammasome activation and cell death in mouse bone marrow derived macrophages when TAK1 was absent [120], suggesting that TAK1 played a key role in maintaining NLRP3 inflammasome quiescence and preserving cellular homeostasis and survival. TAK1 appears to be a negative regulator of p38 and IKK activation, since TAK1-deficient neutrophils enhanced the phosphorylation of the kinases IKK, p38, and JNK and the production of IL-1β, IL-6, TNF-α, and ROS after LPS stimulation [121]. Interestingly, specific deletion of TAK1 in the myeloid compartment led to clonal myelomonocytic cell expansion, splenomegaly, and myelomonocytic leukemia [122]. It is not clear whether this phenomenon is associated with increased pyroptosis.

4.2. Cardiovascular system

Cardiovascular disease is a leading cause of death and disability worldwide [123]. Pyroptosis mediated by NLRP3 inflammasome can occur in various vascular cells, such as vascular endothelial cells (VECs), vascular smooth muscle cells (VSMCs), cardiomyocytes, and cardiac fibroblasts [124, 125]. Pyroptosis has been shown to contribute to the pathogenesis of cardiovascular diseases, including atherosclerosis, myocardial infarction, diabetic cardiomyopathy, and Kawasaki disease [126–129]. Here, we focus on the most recent discoveries about the molecular mechanisms and involvement of pyroptosis in cardiovascular diseases.

4.2.1. Atherosclerosis (AS)

AS is a chronic progressive disease characterized by the accumulation of lipids and infiltration of inflammatory cells in the arterial wall [130]. Endothelial cell death is a crucial and initial stage in the process of AS. VSMCs, normally located in the vascular media layer, play an important role in arterial wall remodeling. The prominent role of VSMCs in AS has been established since early 1970s [131, 132]. Recent studies indicate that pyroptosis of endothelial cells, VSMCs, and macrophages plays a critical role in the initiation and development of AS [133, 134]. In the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS), a human IL-1β neutralized antibody, Canakinumab, was shown to significantly reduce the rate of recurrent adverse cardiovascular events in patients with prior myocardial infarction, suggesting a link between inflammasome activation and atherosclerosis pathogenesis [135].

Many intracellular and extracellular activators of NLRP3 inflammasome in endothelial cells are associated with the pathogenesis of AS, including oxidized low-density lipoprotein (ox-LDL) [136], cholesterol crystal [137], nicotine [138], cadmium [139], trimethylamine N‐oxide (TMAO) [140], Lysophosphatidylcholine (LPC) [141], decabromodiphenyl ethane [142], and polychlorinated biphenyls (PCBs) [143]. Most of those activators mediate cell pyroptosis through the ROS/NLRP3/caspase-1 pathway [137–139, 142]. Briefly, the activators trigger induction of ROS production, which activates NLRP3 inflammasome, leading to the activation of caspase‐1. Activated caspase-1 triggers pore formation in cell membrane, DNA fragmentation, and release of mature IL-1β, IL-18, and high mobility group box 1 (HMGB1) from cells, causing a sterile inflammation response and further contributing to pyroptotic cell death and subsequently promoting AS. Several recent studies reported the signaling pathways that induced the production of ROS associated with AS. Zhaolin et al. reported that a MicroRNAs miR-125a-5p induced ROS generation, NLRP3 activation, and ultimately pyroptosis of VECs upon stimulation by ox-LDL, a crucial pathogenic factor for AS [136]. Wu et al. demonstrated that TMAO, a product from the phosphatidylcholine metabolism of gut flora [144], was able to promote the progression of atherosclerotic lesions in apolipoprotein E-deficient (ApoE^-/-) mice fed a high-fat diet by activating vascular endothelial cell pyroptosis [140]. They found that TMAO enhanced expression of succinate dehydrogenase complex subunit B (SDHB) in the vascular endothelial cells of ApoE^-/- mice and in cultured human umbilical vein endothelial cells (HUVECs).

Overexpression of SDHB in HUVECs impaired mitochondria and increased ROS levels. A ROS scavenger N-acetyl-L-cysteine (NAC) inhibited pyroptosis of the SDHB overexpressed endothelial cells.

Except for ROS production, several other signals, including potassium efflux, lysosomal rupture and mitochondrial dysfunction were also reported to activate NLRP3 inflammasome in VECs [141, 145, 146]. LPC is a major lipid component of the plasma membrane and plays a crucial role in the formation of atherosclerotic lesions [147]. Corrêa et al. reported that LPC induced pyroptosis of monocytes and endothelial cells by increasing potassium efflux and lysosomal damage [141]. In another study, potassium efflux was shown to play a role in the mixed lineage kinase domain-like (MLKL) protein triggered endothelial cell pyroptosis by ox-LDL stimulation [146]. MLKL is an executor of necrosis [148]. The expression of MLKL significantly increased in the pyroptotic endothelial cells induced by ox-LDL. Overexpression of MLKL induced an increase in NLRP3 inflammasome activation. Blocking of cell potassium efflux significantly attenuated the activation of caspase-1 and IL-1β.

Recently VSMC pyroptosis was reported to occur in mouse and human atherosclerotic plaques [149]. Distribution of active pyroptotic indicators, including cleaved caspase-1 and IL-1β, largely overlapped with α-smooth muscle actin, a marker of smooth muscle cells, especially near the necrosis core, at the plaque surface and in intra-plaque hemorrhage area. Importantly, VX-765, a specific inhibitor of caspase-1, reduced the pyroptosis of mouse VSMCs and IL-1β processing induced by ox-LDL, and inhibited progression of established atheroma and the development of AS [149].

AIM2 promoted pyroptosis of VSMCs has recently been reported to play a role in the progression of atherosclerotic plaques [150]. Oxidized low-density lipoprotein (ox-LDL) is a major risk factor of AS. Incubation of VSMCs with ox-LDL increased AIM2 expression and GSDMD-NT levels, and cell death. In consistent with the in vitro results, high-fat diet increased expression of AIM2 and the level of GSDMD-NT, and death of smooth muscle cells in ApoE^-/- mice. Importantly, overexpression of AIM2 in the ApoE^-/- mice by injection with lentivirus-AIM2 further increased high-fat diet-induced atherosclerotic lesions, while inhibition of AIM2 expression by injection with hRNA-AIM2 reduced atherosclerotic lesions.

Pyroptosis also occurred in macrophages during AS. ox-LDL induced macrophage pyroptosis through the ROS/NLRP3/caspase-1 pathway [151]. Caspase-11/GSDMD was shown to be involved in ox-LDL-induced macrophage pyroptosis [152]. Depletion of caspase-11 or suppressing GSDMD attenuated the volume and macrophage infiltration of atherosclerotic lesions in the ApoE^-/- mice fed with a high-fat/high-cholesterol diet. Nicotine has been reported to play a role in accelerating AS by promoting macrophage pyroptosis [153]. Nicotine exacerbated atherosclerotic lesions in ApoE(-/-) mice, which was mediated by the thioredoxin-interacting protein (TXNIP), a protein binds to thioredoxin and modulates its antioxidant functions. Inhibition of TXNIP expression in mice reversed the effects of nicotine on macrophage invasion and vascular injury, and thus improved atherosclerotic plaque lesions [153]. Nicotine may induce macrophage pyroptosis via the histone deacetylase 6 (HDAC6)/NLRP3 signaling pathway [154]. HDAC6 is required for NLRP3 and pyrin inflammasome activation through microtubule transport and assembly of these inflammasomes [155]. HDAC6 modifies proteins by removing acetyls to regulate diverse biological functions [156, 157]. In vitro studies revealed that nicotine upregulated HDAC6, which deacetylated NF-κB p65. Enhanced nuclear translocation of NF-κB p65 mediated the transcription of NLRP3 and macrophage pyroptosis [154].

4.2.2. Myocardial infarction (MI)

MI is a severe coronary artery related disease, mainly resulted from disruption of coronary atherothrombosis or imbalance of myocardial oxygen supply-demand [158]. Both in vivo and in vitro studies have shown that pyroptosis in cardiomyocytes and cardiac fibroblasts play a significant role in MI [128, 159]. Recently, several studies reported that negative regulators of pyroptosis protected against heart injury after MI. Han et al. reported that CXADR‐like membrane protein (CLMP), a member of the CTX family, was highly expressed in the fibroblasts of the ischemic heart [160]. Since CLMP deficient mice died within 4 weeks after birth, they used Clmp^+/− mice to investigate the role of CLMP in MI. Pyroptosis of cardiac fibroblasts was more severe in the Clmp ^+/− mice. Accordingly, myocardial fibrosis and ventricular dysfunction post‐MI was more severe in the Clmp^+/− mice than wild‐type mice. Li and coworkers reported that the growth differentiation factor 11 (GDF11) contributed to cardio protection by preventing cardiomyocyte pyroptosis via the homeobox transcription factor A3 (HOXA3)/NLRP3 signaling pathway in MI mice [161]. The data showed that the expression of GDF11 was markedly lower in heart tissues of MI. Meanwhile, GDF11 overexpression significantly improved heart function in MI mice and also decreased the pyroptosis of hypoxic cardiomyocytes. In support of a role of cardiomyocyte pyroptosis in MI, Li et al. showed that miR-135b protected cardiomyocytes from infarction through the inhibition of the NLRP3/caspase-1/IL-1β pathway [162]. Han et al. showed that long noncoding RNA (lncRNA) H19 and CYP1B1 were associated with the MI progression and pyroptosis of cardiomyocytes [163]. The level of H19 was downregulated while CYP1B1 was upregulated in the peripheral blood in MI patients compared to those in healthy controls. Overexpression of LncRNA H19 suppressed pyroptosis of cardiomyocytes and thus attenuated myocardial infarction, suggesting that cardiomyocyte pyroptosis contributes to MI.

Coronary reperfusion is a main treatment method for acute MI. However, reperfusion itself may lead to accelerated and additional myocardial injury which is the so-called myocardial ischemia/reperfusion injury (MI/RI) [164]. MI/RI induces caspase-1-mediated pyroptosis by activating the NLRP3 inflammasome [165–167]. Importantly, several compounds alleviate MI/RI through targeting the NLRP3/GSDMD pathway, including sweroside [168], metformin [169], piperine [170], emodin [171], Kanglexin (a novel anthraquinone compound) [172], beta-asarone [173], and dexmedetomidine [174]. Nie and coworkers showed that hydrogen gas inhalation could also alleviate MI/RI through the inhibition of oxidative stress and NLRP3-mediated pyroptosis in rats [175].

4.2.3. Pulmonary hypertension

Pulmonary hypertension is a clinically common cardiopulmonary disease, which is characterized by pulmonary vascular remodeling and right ventricular hypertrophy [176]. The NLRP3 inflammasome has been shown to contribute to hypoxic pulmonary hypertension [177]. Consistently, deletion of the inflammasome adaptor ASC reduced hypoxia-induced pulmonary hypertension in mice [178]. ASC deficient mice displayed reduced muscularization and collagen deposition around arteries. He et al. reported that pulmonary artery smooth muscle cell (PASMC) pyroptosis played a role in chronic hypoxia-mediated pulmonary hypertension progression [179]. They also found that Glioma-associated oncogene family zinc finger 1 (GLI1), a transcriptional activator involved in many diseases, was involved in hypoxia-induced pyroptosis and pulmonary hypertension. GLI1 siRNA reversed pyroptotic cell death of pulmonary artery smooth muscle cells in vitro and specific GLI1 inhibition attenuates hypoxia-induced pulmonary hypertension and pulmonary vascular remodeling in vivo. Zhang et al. reported that pyroptosis occurred in the pulmonary arteries in both rat pulmonary hypertension model and hypoxic human pulmonary arterial smooth muscle cells. Importantly, caspase-1 inhibition attenuated pulmonary fibrosis and the pathogenesis of pulmonary hypertension, as assessed by vascular remodeling [180]. Circular RNAs (circRNAs), a unique class of noncoding RNAs, has emerged as important regulators of pulmonary hypertension. Zhang et al. reported the upregulation of a novel circRNA, circ-Calm4, in PASMCs in a mouse model of hypoxia-induced pulmonary hypertension [181]. Later, circ-Calm4 was found to contribute to hypoxia-induced PASMCs pyroptosis [182]. Inhibition of circ-Calm4 expression in mouse lungs abrogated pyroptosis of PASMCs and thus improved pulmonary hypertension [182].

4.2.4. Kawasaki disease (KD)

KD is a rare systemic inflammatory disease that predominately affects children younger than 5 years old, and endothelial cell damage and inflammation are two essential processes resulting in the coronary endothelial dysfunction in KD [183]. Jia et al. recently reported that endothelial cell pyroptosis played an important role in the pathogenesis of KD [184]. The coronary endothelial damage observed in KD was found to be associated with endothelial cell pyroptosis via the NLRP3 inflammasome activation. The high levels of high mobility group box 1 (HMGB1) lead to elevated expression of RAGE (receptor for advanced glycation end-products) and cathepsin B activity, which resulted in NLRP3 inflammasome-dependent caspase-1-mediated pyroptotic cell death in the ECs [184].

4.2.5. Non-ischemic dilated cardiomyopathy

Dilated cardiomyopathy is defined by the presence of left ventricular dilatation and contractile dysfunction [185]. In the myocardial tissues of patients suffering from non-ischemic dilated cardiomyopathy, hyperactivated NLRP3 inflammasome with pyroptotic cell death of cardiomyocytes has been observed [186]. Doxorubicin, an anthracycline antibiotic with anti-tumor activity, was reported to induce dilated cardiomyopathy in patients [187]. Deficiency of either NLRP3 or caspase-1 protected against dilated cardiomyopathy in a Doxorubicin-induced dilated cardiomyopathy mouse model, demonstrating an important role of inflammasome activation and pyroptosis in thiss disease condition[186]. Doxorubicin enhanced expressions of NOX1 and NOX4, which subsequently induced translocation of dynamin-related protein 1 (Drp1) into mitochondria and mitochondrial fission in myocardial tissues, leading to NLRP3 activation and pyroptosis via ROS generation.

4.2.6. Diabetic cardiomyopathy (DCM)

DCM is a common complication of diabetes, which is associated with inflammation in heart [188]. Several studies have demonstrated the participation of cardiomyocyte pyroptosis in the process of DCM. Chemerin and its receptor CMKLR1 (a G-protein-coupled receptor) have been shown to play an important role in DCM. Xie et al. showed that chemerin promoted DCM through NLRP3-dependent pyroptosis [189]. Yang et al. reported a role of a long non-coding RNA (lncRNA) Kcnq1ot1 in DCM [190]. The expression of Kcnq1ot1 increased in the cardiac tissues from patients with diabetes and diabetic mice, and silencing of Kcnq1ot1 alleviated pyroptosis and improved cardiac function and morphology in vivo. In contrast, overexpression of another lncRNA, GAS5, suppressed NLRP3 inflammasome-mediated pyroptosis and alleviated DCM [191]. All these data indicate that pyroptosis downstream from NLRP3 inflammasome plays an important role in diabetic cardiomyopathy.

4.3. Respiratory system

The human respiratory tract is constantly exposed to harmful microorganisms and air pollutants. In patients with community acquired pneumonia, asthma, chronic obstructive pulmonary disease (COPD), pulmonary fibrosis, and acute lung injury (ALI), inflammasome activation regulates respiratory infections and pathological airway inflammation. The PAMPs and DAMPs resulted from these disease situations induce inflammasome activation, leading to secretion of IL-1β and IL-18 as well as pyroptosis [192]. It is known that several inflammasomes, including NLRP3, NLRC4, and AIM2, are involved in lung damage in different diseases. Inhaled environmental stimuli such as smoke, asbestos, silica, particulate matter, allergens, and air pollution can also activate the NLRP3 inflammasome, leading to pyroptosis [36, 193–196].

4.3.1. Pneumonia

Pneumonia is caused by infection and subsequent inflammation of the respiratory tract. S. pneumonia infection is the most common cause of community acquired pneumonia, which is the leading cause of death in many countries [197]. Using a S. pneumonia infection murine model, Witzenrath et al. identified streptococcal exotoxin “pneumolysin” as a key activator of the NLRP3 inflammasome [198]. Pneumolysin destroyed plasma membrane in cells via forming of pores on the cell membrane and caused cell lysis, and the release of cell components led to activation of the NLRP3 inflammasome and caused cell pyroptosis [199]. Influenza A virus and adenovirus infection stimulated ATP and ROS release from respiratory epithelial cells, suggesting a mechanism of NLRP3 inflammasome activation and pyroptosis [200–202]. Lee et al. reported that influenza A infection triggered respiratory epithelial cells pyroptosis and apoptosis in a mutually exclusive manner. While apoptosis played an important role in the early influenza A infection phase, pyroptosis was dominant at the late phase of infection. Type I interferon signaling regulated the epithelial cell apoptosis shift to pyroptosis to promote proinflammatory cytokine production and initiate proinflammatory response [203].

4.3.2. Asthma

Approximately 10% of people in western countries and 300 million people worldwide currently suffer from asthma, a chronic airway inflammatory disease [204]. Increasing clinical and experimental studies have shown the activation of the NLRP3 inflammasome and/or the production of excessive IL-1β are related to the pathogenesis of asthma [205–212]. Baines et al. studied the gene expression profile of inflammatory cells in induced sputum [206]. The expression of IL-1β signaling pathway-related genes, such as IL-1β, IRAK2, IRAK3, and IL-1R2, significantly increased in the sputum of patients with neutrophilic asthma [206]. Simpson et al. reported that the expressions of NLRP3, caspase-1/4/5 and IL-1β in the airways of patients with neutrophilic asthma increased, and macrophages and neutrophils were the main cell sources of NLRP3 and caspase-1 in this cohort [212]. The expression levels of TLR2, TLR4 and IL-8 in these asthmatic patients also elevated, suggesting abnormal inflammasome activation in asthma [212].

A large amount of clinical evidence suggests that bacterial respiratory infections are related to steroid-insensitive asthma. Chlamydia pneumoniae is an obligate intracellular bacterial pathogen and is associated with severe, steroid-insensitive asthma [213–215]. H. Influenza bacillus is a Gram-negative bacteria and is the most commonly isolated bacteria from the airways of steroid-insensitive asthma patients [216, 217]. Essilfie et al. reported that infections by Chlamydia and H. influenza caused an increase of neutrophils, as well as T helper lymphocyte type (Th) 1 and/or Th17 responses and leading to severe, steroid-insensitive neutrophilic allergic airways disease [218]. Both Chlamydia and Haemophilus respiratory tract infections induced the release of active IL-1β in an NLRP3 inflammasome-dependent and caspase-1-mediated manner [219, 220].

4.3.3. COPD

COPD is a progressive and obstructive lung disease, including chronic bronchitis and emphysema. It has become the fourth leading cause of death world-wide [221, 222]. Clinical studies have shown that cigarette smoking, a leading cause of COPD, induced the release of IL-1β in lung tissues [223–225]. The concentration of HMGB1 significantly increased in sputum and bronchoalveolar lavage fluid in patients with COPD [226, 227]. HMGB1 activated NLRP3 inflammasomes in a TLR4-dependent manner in hemorrhagic shock mice model, mediated the activation of NAD(P)H oxidase and ROS generation, which subsequently promoted NLRP3 inflammasome activation and subsequently induced active IL-1β secretion in endothelial cells [228]. HMGB1-mediated inflammasome activation plays a clear role in the pathogenesis of COPD.

4.3.4. Pulmonary fibrosis

Pulmonary fibrosis refers to a series of lung diseases, which is characterized by irreversible destruction and remodeling of the lung structure due to excessive deposition of collagen and extracellular matrix proteins. The role of inflammasomes in the pathogenesis of pulmonary fibrosis remains elusive. However, fibrosis-inducing irritants (such as silica, asbestos, cigarette smoke, and bleomycin) are known to damage the lung epithelium through directly activating NLRP3 inflammasome [229, 230].

4.3.5. ALI

ALI is a main cause of acute respiratory distress syndrome (ARDS) [231]. ALI/ARDS is identified by severe inflammation, leading to diffuse alveolar injury, varying degrees of ventilation/perfusion differentiation, poor lung compliance, and severe hypoxemia [232]. Many factors can lead to ALI, such as severe shock, infection, mechanical damage, and bacterial infection. LPS is a major component of the outer membrane of Gram-negative bacteria and LPS-induced acute lung injury has been extensively studied [233]. Pyroptosis has been revealed as a critical step in LPS-induced inflammation [234–236].

Several studies have shown that alveolar macrophages (AMs) had a profound impact on the occurrence and prognosis of ALI by releasing proinflammatory cytokines and regulating other immune cells [237–239]. Pyroptosis of AMs has been shown to be involved in the development of ALI and the NLRP3/ASC inflammasome was essential in this process [240–242]. Knocking down of ASC or inhibiting the function of NLRP3 reduced AM pyroptosis and release of HMGB1, subsequently reduced the inflammatory response in ALI [241].

IL-1R1 up-regulates the sensitivity of AMs to IL-1β, causing elaboration of pyroptosis of AMs, and subsequent aggravation of ALI [243]. In addition, myeloid differentiation protein 2 (MD-2), an essential factor required for TLR4 binding of LPS, was required for the expression of NLRP3 and IL-1β [244]. Interferon regulatory factor-1 (IRF-1) played an essential role in mediating the pyroptosis of AMs in LPS-induced ALI [245]. MyD88 is also known to recruit IRF-1 transcription factors to activate the TLR4 pathway [246]. Thus, IRF-1 may be the bridge between the LPS-TLR4 signaling pathway and the caspase-1-dependent pyroptosis pathway in AM.

Liu et al. suggested a new mechanism in macrophages to explain how mitochondrial DNA (mtDNA) activated the pyroptosis pathway, thereby aggravating the inflammatory response in ALI [247]. Specifically, the sensory receptor of cytosolic DNA (cyclic GMP-AMP synthase, cGAS) detected mtDNA leaked into the cytoplasm and activated stimulator of interferon gene (STING) [247, 248]. LPS also upregulated the expression level of the transcription factor c-Myc, which directly promoted the expression level of STING. The upregulation and activation of STING activated the NLRP3 inflammasome and caspase-1, which ultimately leads to cell pyroptosis and ALI [247, 248]. STING deficiency blocked LPS‐induced pyroptosis in lung tissues, evidenced by the reduced expression levels of cytokines (IL‐1β and IL‐18) and GSDMD, which was restored by NLRP3 overexpression. However, the exact mechanism by which STING regulate the activation of NLRP3 remains elusive [247].

4.3.6. Radiation lung injury (RILI)

RILI is a common and serious complication in the treatment of thoracic tumors. It not only limits additional fractional radiotherapy, but also seriously affects the prognosis of patients. The development of RILI is divided into an early phase of radiation pneumonitis and a late stage of radiation fibrosis. Acute inflammation occurs within a few weeks after radiation and is characterized by the release of pro-inflammatory cytokines and the accumulation of immune cells in the lung tissue, while chronic fibrosis occurs months to years later, and ultimately leads to permanent damage of lung function [249–252]. More and more evidences indicate that AIM2 inflammasome-mediated pyroptosis of macrophages and epithelial cells play a key role in radiation-induced tissue damage [96, 97]. Under radiation, AIM2 recognizes the double-strand DNA fragments resulted from ionizing radiation and chemotherapeutic agents, forming specks and recruits ASC, thereby activating caspase-1 and leading to pyroptosis [253–257]. Gao et al. demonstrated that inhibition of AIM2 activation attenuated the inflammatory cascade, acute pneumonitis, and subsequent chronic fibrosis, illustrating the role of AIM2 in RILI [258].

4.4. Digestive system

Inflammasome activation and pyroptosis play a significant role in immunopathogenesis of several different conditions related to gastrointestinal tract. The effector molecules of pyroptosis, the Gasdermin proteins, were actually first discovered in cells lining the gastrointestinal tract in early 2000s [259]. However, only recently the role of Gasdermin proteins in pyroptosis was identified [6, 253]. Gastrointestinal tract has been attributed as the biggest immunological organ in the body, hence, the role of innate immune system is pivotal in this system [260]. Si Ming Man, in a review in 2018, listed the expression profile of inflammasome components in cells lining the gastrointestinal tract [261]. The majority of inflammasome components are expressed in these cells with a slight difference between human and mice. For example, human enterocytes express caspase-1, -4,- 5, IL-18, NLRP6, and NLRP9 whereas murine enterocytes express AIM2, ASC, Caspase-1, -11, GSDMD, Il-1β, IL-18, NAIPs, NLRC3, NLRC4, NLRP6 and NLRP9b [261]. Similarly, human gastric cells express AIM2, ASC, Caspase-1, NLRC4, NLRP1,3, and 6 whereas murine gastric cells express ASC, IL-1β and IL-18 [261]. Gastrointestinal system is a site where there is a need of perfect balance between gut microbiome and immune system. Imbalance, either through overactivation or through decreased activation of inflammasome, leads to severe pathogenic conditions. Below we discuss some of the conditions where inflammation and pyroptosis are involved in the immunopathogenesis in gastrointestinal tract and pancreas.

4.4.1. Bacterial and viral infection

Gastrointestinal lining acts as a barrier and maintains a perfect homeostasis between the gut microbiome and the host body [260, 262]. Intestinal epithelial cells (IECs) lie at the interface of gut microbiota and host systemic sites. Unchecked inflammation and pyroptosis of IECs causes disruption of this intestinal homeostasis between the host immune system and gut microbiome [263]. Unregulated disruption of intestinal barrier promote invasion by gut microbiome, leading to systemic infection and sepsis, and excessive inflammation and cell death in the intestinal lining cause tissue damage [263, 264]. However, suboptimal inflammation leads to improper colonization of the intestinal environment by microbes, rendering the host susceptible for systemic infection [265].

NLRP3 and ASC have been found to be important in restricting bacterial colonization in a mice model infected orally with Citrobacter rodentium. During the early stage of infection, NLRP3 activation in intestinal epithelial cells acted as a defense barrier to limit bacterial colonization and systemic infection [266]. Similarly, Crowley et al. demonstrated that caspase-1 and caspase-11 were essential to restrict the intracellular bacteria Salmonella typhimurium within the gut [267]. These caspases were activated after the detection of bacterial virulence factors and triggered pyroptotic cell death, leading to shedding of infected cells and expulsion of bacteria in the gut. Similarly, NAIP and NLRC4 inflammasome has also been shown to be important to restrict the intracellular replication of Salmonella during gastrointestinal infection. Lack of these inflammasome components led to increased bacterial colonization of the intestinal epithelial cells in oral Salmonella infection model mice [268–270]. However, protection from bacterial dissemination into deeper tissues by IEC pyroptosis and expulsion is a very controlled process coordinated by IL-18 and eicosanoids. In addition, extracellular pathogens cause pyroptosis of IECs causing intestinal inflammation and diarrhea like symptoms. For example, enteropathogenic E. coli causes the pyroptosis of IECs in Ca²⁺ influx dependent LPS internalization [271]. Excessive activation of IEC pyroptosis and expulsion led to severe damage in the epithelial layer and compromise the intestinal barrier to expose the systemic sites [270].

In addition to bacterial infection, pyroptosis is also important during gastrointestinal viral infection. It has been shown that pyroptosis has both positive and negative consequences in terms of viral mediated lethality depending on the mouse genotype and extent of activation [272, 273]. Pyroptosis contributes towards the detrimental effects during gastrointestinal murine norovirus infection in STAT1^-/- mice. STAT1 has been found to be protective and block norovirus replication and infection, making the deficient mice more susceptible to infection [272, 274]. Pyroptosis played a detrimental role in STAT1^-/- mice but not in WT mice during norovirus infection. Norovirus infection was observed to be causing the NLRP3 inflammasome activation and pyroptosis. In WT mice, STAT1 was shown to control the expression of pro-IL-1β, so that during STAT1 deficiency, murine norovirus induced inflammasome activation and pyroptosis caused excess release of IL-1β [272]. Transmissible gastroenteritis virus (TGEV), which causes fatal severe diarrhea, has recently been found to induce NLRP3 inflammasome and pyroptosis in porcine intestinal epithelial cells [275]. Pyroptosis could also be protective during gastrointestinal viral infection. For instance, Zhu et. al. showed that NLRP9b/ASC/Caspase-1 inflammasome mediated pyroptosis played a protective role during intestinal rotavirus infection [273]. They demonstrated NLRP9b deficient mice were more susceptible to rotavirus infection. Similarly, GSDMD deficiency but not IL-18 deficiency also showed increased susceptibility independent of intestinal microbiota, suggesting a protective role of pyroptosis against rotavirus infection. Thus, similar to bacterial infection, inflammasome activation and pyroptosis may play roles in controlling microbe intracellular replication, however, excessive inflammation could be detrimental.

4.4.2. Inflammatory bowel disease (IBD)

One of the major diseases involving the imbalanced inflammation in intestine is the IBD, which generally comprises of two conditions, Crohn’s disease (CD) and Ulcerative colitis (UC) [276]. UC is more limited in the colon, whereas CD can be present in any part of gastrointestinal tract. IBD can be triggered by various genetic causes, environmental influences, imbalance in gut microbiome and immune system dysregulation [265, 277, 278]. Complete loss of inflammasome can exacerbate the IBD, making the mice more susceptible to experimental colitis [279]. However, overactivation of these inflammasome also creates problem [278, 279]. The role played by inflammasome and pyroptosis remains elusive as both protective and detrimental effects have been observed in various studies [278, 280–283]. Although the reason for the discrepancy between different studies is not fully understood, plenty of evidence suggests that pyroptosis contributes to the pathogenesis of IBD and its proper regulation is one of the keys for the treatment of IBD.

Several studies using dextran sodium sulfate (DSS)-induced colitis models suggested a protection role of pyroptosis. Elinav et al. reported that NLRP6 and ASC deficiency were correlated with the severity of DSS-induced colitis [284]. Allen et. al. reported cancer progression associated with DSS-induced colitis increased in mice lacking NLRP3, ASC or caspase-1 [281]. Deficiency in NLRP3, ASC, caspase-1, and caspase-11 has been shown to promote epithelial cell damage and morbidity, and administration of IL-1β and/or IL-18 attenuated the disease progression [285, 286].

However, a large number of studies reported an opposite effect of pyroptosis in IBD. Some early studies reported a correlation of increased IL-1β, IL-18 among other pro-inflammatory cytokines with the severity of the inflammatory bowel disease [287–289]. Similarly, inhibition of caspase-1 was found to be protective against the DSS-induced acute experimental colitis suggesting a role of pyroptosis during colitis [290, 291]. Besides, neutralization of IL-18 also helped to ameliorate the DSS-induced acute colitis [292]. Data from these studies suggest that inflammatory cytokines from inflammasome activation contribute to the pathogenesis of IBD. Recently, Xu et. al. reported that CD147, a transmembrane protein which induces matrix metalloproteinases and angiogenic factors, exacerbated the IBD in patients by increasing pyroptosis in intestinal epithelial cells [293]. CD147 promotes phosphorylation of NF-kB, leading to upregulation of NLRP3 inflammasome components and increased pyroptosis [293]. Xiong et. al. demonstrated that the protective effect of cholecalciferol cholesterol emulsion for IBD was through downregulation of pyroptosis, specifically by inhibiting the NF-kB mediated expression of pro-forms of inflammasome components and cytokines [278].

Similarly, Chen et al. found that DSS-induced chronic colitis was significantly reduced upon knockdown of NEK7 (NIMA related kinase 7), an important component of macrophage’s NLRP3 inflammasome [277]. Reduction of expression of genes involved in pyroptosis including NLRP3, ASC and GSDMD has been shown to be protective in IBD mice models [294, 295].

4.4.3. Intestinal ischemia-reperfusion (I/R) injury

Another condition where pyroptosis plays a role in tissue damage in gastrointestinal tract is the intestinal I/R injury. I/R injury is a condition where there is obstruction in proper blood flow in certain organs leading to tissue/organ damage and failure [296]. Yang et al. reported the occurrence of pyroptosis during renal I/R injury [297]. Recently, Jia et al. reported the activation of inflammasome and pyroptosis as a downstream process during intestinal I/R injury [298]. Pyroptosis contributed toward the prognosis of injury by cell death and disruption of intestinal barrier, which may further lead to sepsis. Metformin was found to be protective against intestinal I/R injury by inhibiting pyroptosis and inflammatory response by blocking the interaction between TXNIP and NLRP3, which otherwise would activate the NLRP3 inflammasome [298].

4.4.4. Non-steroidal anti-inflammatory drugs (NSAID) induced intestinal injury

Use of NSAID may cause severe injury and bleeding in gastrointestinal tract. This was ameliorated by using the acid suppressants and histamine H2 receptor antagonists in the stomach but not in the intestinal injury, where they can even worsen the case [299, 300]. Recently, Otani et al. demonstrated that NSAID-induced intestinal injury could be averted by colchicine through the inhibition of NLRP3 inflammasome [299]. In NSAID-induced small intestinal injury, NF-kB upregulation through TLR4 signaling and extracellular ATP caused the activation of NLRP3 and pro-IL-1β expression [301, 302]. The release of IL-1β has been shown to aggravate the pathogenesis of this intestinal injury and use of IL-1β neutralizing antibodies ameliorated the injury. Otani et. al. observed that the levels of NLRP3, caspase-1 and IL-1β significantly increase in the small intestine upon indomethacin induced small intestine injury, and colchicine could inhibit the activation of NLRP3 inflammasome [299].

4.4.5. Pancreatic diseases

Prolonged exposure to inorganic arsenic has been shown to cause diabetes and recently pyroptosis of pancreatic beta cells has been found to contribute to the pathogenesis of arsenic induced diabetes [303, 304]. By using insulin secreting beta cell derived cell line (INS-1 cells), Pei et. al. demonstrated that prolonged exposure to As₂O₃ caused elevated expression of ER stress sensor IRE1α and TNF-α, activating the NLRP3 inflammasome and caspase-1 cleavage [305]. Inhibition of IRE1α, TNF-α and NLRP3 by using their inhibitors ameliorated the As₂O₃ induced pyroptosis and β-cell dysfunction.

Pyroptosis was also observed to contribute to the pathogenesis of acute pancreatitis. Acute pancreatitis is associated with large-scale damage and death of pancreatic acinar cells. Although several different mechanisms can cause the death of acinar cells, pyroptotic death has been shown to be related with the worst prognosis of the disease [306]. Wang et al. showed that a circular RNA, circHIPK3, stimulated the pyroptotic cell death of acinar cells by regulating the miR-193a-5p/GSDMD axis [306]. In the plasma of acute pancreatitis patients, level of circHIPK3 was elevated. CircHIPK3 regulates the expression of multiple miRNAs, including miR-193a-5p, the level of which reduced in the samples from acute pancreatitis patients. MiR-193a-5p negatively regulates the GSDMD expression, thus enhanced expression of circHIPK3 led to an increased expression of GSDMD, and subsequently pyroptosis.

Pyroptosis is also involved in the intestinal injury during severe AP. GSDMD mediated pyroptosis causes cell deaths in the intestinal mucosal lining, and the downregulation of intestinal epithelial cell tight junction proteins, further weakening the intestinal barrier [307]. Recently, Lin et. al. demonstrated that downregulating the GSDMD expression by using siRNA protected the intestinal injury and helped to alleviate the symptoms during acute pancreatitis in mice model [308].

4.5. Reproductive system

Activation of inflammasome has been associated with many diseases in the reproductive system, including infectious diseases such as vaginitis and sextually transmitted infections (STIs), and non-infectious disease including female and male infertility, recurrent miscarriages, endometriosis, preeclampsia, placental inflammation, and preterm birth [309, 310]. Inhibitors targeting the inflammasome activation have been shown to benefit the treatment of certain reproductive diseases. For example, neutralization of ASC by adding anti-ASC polyclonal antibody to semen significantly increased mean sperm motility by 50% in men with spinal cord injury [311]. Using a mouse model for sterile intra-amniotic inflammation, Gomez-Lopez et al. showed that inhibition of the NLRP3 inflammasome via the injection of a specific inhibitor MCC950 prevented preterm labor/birth by 35.7% and reduced neonatal mortality by 26.7% [312]. Omega-3 fatty acid has been shown to reduce the rate of preterm birth induced by infection and trophoblast inflammation through inhibiting inflammasome-related molecule synthesis and cathepsin S (CTSS) and caspase-1 activation [313].

While a clear role of inflammation has been established early on in many pathological conditions related to the reproductive system, the contribution of pyroptosis was not as clear since mechanistic studies have been limited [309, 314]. Many studies focused on the examination of elevated level or activity of biomarkers related to inflammasome activation, including NLRP3, caspase-1, ASC, IL-1β, and IL-18 in patient, while the underlining mechanism remains elusive. Here we focus our discussion of several recent papers aiming at probing the connection between the molecular trigger and the pathogenesis of disease conditions.

Maternal obesity has been identified as a major risk factor for adverse pregnancy complications and associated with pro-inflammatory cytokine release in the placenta [315]. In the plasma of obese patients, the levels of free fatty acid, including palmitic acids (PAs), are elevated. Using a human trophoblast cell line, Shirasuna et al. discovered that PA could activate the NLRP3 inflammasomes, leading to significant caspase-1 activation and IL-1β secretion, and disruption of caspase-1 activity diminished PA-induced IL-1β release, clearly indicating the involvement of pyroptosis in this process [316]. Furthermore, when production of ROS was disrupted by the addition of antioxidant, the PA-induced cytokine release was abolished. While the role of ROS in the process remains elusive, these studies suggest that free fatty acid such as PA plays an important role in the pathogenesis of obesity-related placental inflammation and subsequent pregnancy complications.

Yang et al. investigated the mechanisms linking inflammatory dysregulation and endoplasmic reticulum (ER) stress to preeclampsia [317]. In preeclampsia, trophoblast cells express inflammatory cytokines including IL-1β, a biomarker of inflammasome activation [318, 319]. Since TXNIP has been reported to activate the NLRP3 inflammasome [320, 321], and NLRP3 inflammasome activation has been shown to activate IL-1β in the placenta of patients with preeclampsia, the authors hypothesize that TXNIP was the linkage between ER stress and NLRP3 inflammasomes to activate IL-1β production in preeclampsia [322]. Consistent with their hypothesis, ER stress was found to enhance TXNIP activation and induce NLRP3 inflammasome formation; suppression of the TXNIP expression using siRNA and lentivirus or treatment with ER stress inhibitors both reduced NLRP3 inflammasome activation. Furthermore, TXNIP has been shown to be involved in compromised trophoblast invasion, a feature involved in the pathogenesis of PE [314, 323, 324]. Through a series of studies, they first established that it is pyroptosis, not necroptosis, that lead to the inflammatory response in the placenta from women with early onset preeclampsia. Then using an ER stress cellular models, they further examined the signaling pathways leading to pyroptosis in human trophoblasts, and observed increased expression of TXNIP and NLRP3 in both human primary trophoblast cells and in placental tissues from early onset preeclampsia deliveries. Based on these results, a mechanism was derived in which excessive ER stress and unfolded protein response lead to activation of the NLRP3-pyroptotic inflammatory pathway through TXNIP, which is pivotal to the inflammation response in early onset PE pathology.

5. Future direction and perspectives

While the importance of pyroptosis has been reported in the pathogenesis of various disease conditions, many studies are focused on the examination of activation of inflammasome and generation of pro-inflammatory cytokines associated with pyroptosis. Studies focusing on mechanistic insights have been limited. Additional insight from mechanistic studies are necessary to lay the foundation for the development of better therapies and to identify new biomarkers for diagnosis and monitoring.

As discussed extensively above, inflammasome activation and pyroptosis could be both protective and detrimental in various inflammatory disorders and disease conditions, depending on many factors. Thus, inhibition of inflammasome as a therapeutic strategy can be both promising and challenging. For example, while the direct outcome of inhibiting the inflammasome activation and pyroptosis had been proven beneficial in many disease models including sepsis [325–327], preterm birth [312, 328], PE [329], AS [148], AML [114], MI/RI [159–165], pulmonary hypertension [170, 173], the overall clinical outcome demands thorough evaluation and careful consideration of the entire body as inhibiting of normal house-keeping inflammasome activation could lead to unintentional immunosuppressive effects.

Highlights.

Pyroptosis occurs in many cell types and affect almost all vital systems in the human body.

Pyroptosis is involved in pathogenesis in the hematopoietic, cardiovascular, respiratory, digestive, and reproductive systems

Inflammasome activation is a double-edged sword with both protective and detrimental effect to host wellbeing