Mechanisms and Consequences of Noncanonical Inflammasome-Mediated Pyroptosis

Abstract

The noncanonical inflammasome, comprising inflammatory caspases 4, 5, or 11, monitors the cytosol for bacterial lipopolysaccharide (LPS). Intracellular LPS-elicited autoproteolysis of these inflammatory caspases leads to the cleavage of the pore-forming protein gasdermin D (GSDMD). GSDMD pore formation induces a lytic form of cell death known as pyroptosis and the release of inflammatory cytokines and DAMPs, thereby promoting inflammation. The noncanonical inflammasome-dependent innate sensing of cytosolic LPS plays important roles in bacterial infections and sepsis pathogenesis. Exciting studies in the recent past have significantly furthered our understanding of the biochemical and structural basis of the caspase-4/11 activation of GSDMD, caspase-4/11’s substrate specificity, and the biological consequences of noncanonical inflammasome activation of GSDMD. This review will discuss these recent advances and highlight the remaining gaps in our understanding of the noncanonical inflammasome and pyroptosis.

Introduction

The innate immune system is composed of an evolutionarily conserved set of germ-line encoded pattern recognition receptors (PRR) and effectors that serve critical roles in host surveillance and defense against pathogens [1,2]. PRRs are strategically positioned to recognize pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) in the intracellular and extracellular compartments. Once engaged, PRRs activate key transcription factors, including nuclear factor kappa B (NF-κB) or interferon regulatory factors (IRF)-3/7, to trigger the transcription of proinflammatory cytokines such as interleukin (IL)-6 and tumor necrosis factor (TNF) or type I interferons, respectively [3].

Inflammasomes, a special class of PRRs, are oligomeric complexes that assemble in response to various PAMPs and DAMPs [4,5]. Canonical inflammasomes consist of, in most cases, a sensor, an adaptor molecule, apoptosis-associated speck-like protein containing a caspase activation and recruitment domain (CARD) (ASC), and the effector protein caspase-1 [6]. Cellular perturbations including ion fluxes, bacterial secretion system proteins, and DNA activate the canonical inflammasomes such as NLRP3, NAIP/NLRC4, and AIM2, respectively, to induce the autoproteolysis of caspase-1 [5,7]. Once activated, caspase-1 cleaves the latent forms of proinflammatory cytokines IL-1β and IL-18 into their active forms. At the same time, caspase-1 cleaves gasdermin D (GSDMD), releasing the inhibitory C-terminal domain from the pore-forming N-terminal domain [8–11]. The N-terminal domain of GSDMD migrates to the plasma membrane and forms pores, leading to lytic death known as pyroptosis. During pyroptosis, GSDMD pores and plasma membrane rupture allow the release of cytokines, DAMPs, and alarmins that amplify local and systemic inflammation.

A noncanonical form of inflammasome devoid of dedicated receptor and adapter proteins has been uncovered almost a decade ago [12–15]. In this abridged version of the inflammasome, the effector caspases—caspases 4 and 5 in humans and caspase-11 in rodents that are closely related to the prototypical inflammatory caspase-1—function as the receptor also [12]. These inflammatory caspases (4/5/11) directly sense lipopolysaccharide (LPS), an abundant gram-negative bacterial cell wall component. The CARD domain of caspase-11/4/5 binds to the lipid A of LPS leading to caspase oligomerization and activation [12]. While the hexa-acylated lipid A of LPS is the optimal ligand for caspases 4 and 11 [12,14], caspase-4, not caspase-11, can sense tetraacylated lipid A [16]. This is one of the few functional differences between human and murine noncanonical inflammasomes. As pathogenic human bacteria such as Yersinia and Francisella underacylate their lipid A to evade TLR4 recognition [17,18], caspase-4’s capacity to detect such underacylated lipid A can be considered a host strategy to counter pathogen evasion mechanisms. The cytosolic localization of LPS occurs during both cytosolic and noncytosolic bacterial infections. In the latter scenario, LPS-rich outer membrane vesicles (OMV) secreted by bacteria play a critical role in the cytosolic localization of LPS [19].

Furthermore, host factors such as HMGB1, secretoglobin-3A2 syndecan-1, and heparin-sensitive glycocalyx degradation have been shown to play a role in the intracellular access of LPS [20–22]. Since the lipid A moiety of LPS is buried in the outer membrane of bacteria and OMVs, caspase-4’s access to its ligand, lipid A, is limited. An ordered assembly of interferon-inducible guanylate binding proteins (GBPs) on bacterial membranes serves as a platform to recruit caspase-4 and facilitate its access to lipid A [23–31]. Once activated, caspase-4/5/11 directly cleave GSDMD to induce pyroptosis. Concomitantly, active GSDMD is considered to indirectly trigger NLRP3 inflammasome assembly leading to the processing of IL-1β and IL-18 by caspase-1 [32–35].

Unique pyroptotic substrate recognition by inflammatory caspases

Upon their activation by intracellular LPS, inflammatory caspases 4 and 11 cleave the pore-forming protein gasdermin D (GSDMD). In addition, caspase-4 has been shown to cleave IL-18 and the executioner caspase, caspase-7 [23,26,36]. Interestingly, IL-18 is a substrate for caspase-4 but not caspase-11 implying potential differences in the outcomes of noncanonical inflammasome activation in humans versus rodents especially in cells lacking a functional caspase-1. Moreover, TAR DNA binding protein of 43 kDa (TDP-43) has also been shown to be a substrate of caspase-4 but not caspase-11 [37]. As TDP-43 plays a role in amyotrophic lateral sclerosis (ALS) [38], caspase-4 cleavage of and subsequent cytosolic accumulation of TDP-43 may be involved in neurodegeneration in ALS. Furthermore, a recent proteomic study employing a degenerate monoclonal antibody library determined that the caspase-4 cleavage motif is found in over 300 proteins including several spliceosome components, indicating that there are additional physiological protein substrates for the noncanonical inflammasome [36].

Nonetheless, GSDMD remains the most predominant and well-characterized noncanonical inflammasome substrate identified thus far. GSDMD belongs to a family of gasdermins, which includes GSDMA, GSDMB, GSDME, and GSDMF [39]. Recent studies showed that IRF1 and IRF2 control the constitutive expression of GSDMD [40,41]. GSDMD is composed of two domains connected by a linker region containing a protease cleavage site [8,9]. Cleavage by the inflammatory caspases 1/4/5/11 at Asp276 (mouse) or Asp275 (human) releases the auto-inhibitory C terminal domain from the cytotoxic N terminal domain. It has been recently appreciated that there is a complex interaction between GSDMD and caspase-11 that is unique among caspases [42]. In general, the prevailing dogma is that caspases recognize a tetrapeptide motif (XXXD) within their substrate. For example, the preferred recognition motifs of apoptotic caspase 3, 8, and 9 are DEVD, LETD, and (W/L)EHD, respectively [43]. However, it has been demonstrated that upon LPS sensing, caspase-11/4 autocleaves at Asp289/285 in the p20-p10 interdomain linker Structural and biochemical studies of caspase-1/4/11-GSDMD complexes revealed that the autoproteolysis of caspase-11/4 at the p20-p10 interdomain linker [42,44] generates two anti-parallel β-strands in the p10 domain that bind tightly to the C-terminal domain of GSDMD at a hydrophobic interface [42]. This p10 exosite-driven interaction between caspase-1/4/11 and the C-terminal domain of GSDMD stabilizes dimerization-mediated caspase activation leading to the cleavage of GSDMD in the linker region in a tetrapeptide-independent manner [42]. The two distinguishing features of substrate binding by caspase-11/4, namely the exosite involvement and the tetrapeptide-independency, may explain their relatively narrow substrate range.

Consequences of GSDMD processing by inflammatory caspases

Membrane perforation by GSDMD

The liberated N-terminal fragments of GSDMD oligomerize into a pre-pore complex that preferentially binds to acidic phospholipids. GSDMD most favorably binds phosphatidylinositol phosphates and phosphatidylserine on the inner leaflet of the plasma membrane, thereby forming plasma membrane pores [8–10,45]. The Ragulator-Rag complex-mediated production of reactive oxygen species (ROS) promotes the gasdermin D oligomerization and pore formation [46]. The inner diameter of GSDMD pores ranges from 10–20 nm, depending on the lipid composition at the insertion site. The pore is large enough to allow the passage of processed IL-1β and IL-18, ions, and smaller DAMPs such as galectins [47–50], but not the passage of high molecular weight intracellular proteins such as LDH and HMGB1, which require plasma membrane rupture for release [39]. It was recently appreciated that there is a degree of specificity in releasing molecules through GSDMD pores. Pore electrostatics specifically allow the robust release of mature IL-1β and IL-18 over the pro-forms of the cytokines, although the pore is physically large enough for the unprocessed cytokines to pass [47]. Furthermore, the N-terminal fragment of GSDMD (GSDMD-N) has an affinity for additional lipid species such as cardiolipin associated with mitochondrial and bacterial membranes [10]. Consistent with its affinity for a mitochondrial lipid, GSDMD-N targets and permeabilizes mitochondria leading to the release of cytochrome C and generation of mitochondrial ROS that activates the apoptosome and the canonical NLRP3 inflammasome, respectively [34,35]. Thus, plasma membrane and organelle membrane perturbations by GSDMD-N links the noncanonical inflammasome, canonical inflammasome, and apoptotic pathways. This crosstalk amplifies the cytotoxic effects of GSDMD and indirectly triggers additional cellular responses such as the activation of IL-1β and IL-18 downstream of the noncanonical inflammasome. GSDMD-N binding of cardiolipin has also been implicated in its targeting and direct killing of bacteria [10]. It has been shown that GSDMD pores are subject to a cellular repair mechanism; calcium ion influx through GSDMD pores activates the endosomal sorting complexes required for transport (ESCRT) proteins I and III. Upon assembly, ESCRT proteins encapsulate GSDMD pores at the membrane into vesicles and patch up the plasma membrane, halting the release of inflammatory proteins and progression to cell lysis [51]. These regulatory steps are necessary to allow for membrane repair in situations where the inflammasome activation is sub-optimal.

Cell death and plasma membrane rupture

Upon optimal inflammasome activation, the cell proceeds to die with several events occurring downstream of GSDMD pore formation [52]. First, extensive ionic flux through GSDMD pores disrupts cellular osmolality, leading to swelling of the cells [8]. During this process, the disruption of mitochondrial membrane potential and organelle structures coincides with cells irreversibly committing to death [52,53]. Eventually, the plasma membrane breaks apart, spilling the intracellular contents. The plasma membrane rupture (PMR) was assumed to be a passive event, however, a recent forward genetic screen identified the cell-surface protein ninjurin-1 (NINJ1) as the critical molecule mediating PMR [54]. NINJ1 is widely expressed in myeloid cells and has been associated with inflammation and tumor suppression [55,56]. NINJ1 is a small 15-kDa protein inserted into the plasma membrane such that both N- and C- terminal ends are in the extracellular space. BMDMs deficient in NINJ1 display attenuated release of larger intracellular proteins such as LDH and HMGB1 upon the activation of both canonical and noncanonical inflammasomes. However, NINJ1 deficiency did not affect the release of smaller molecules, including mature IL-1β and IL-18, which positions NINJ1 downstream of GSDMD pore formation. Interestingly, the secondary PMR that can occur following apoptosis is also dependent on NINJ1, but the PMR happening during MLKL-driven necroptosis is only partially dependent on NINJ1, which imply key differences in the plasma membrane perturbations by GSDMD and MLKL.

In vivo studies revealed that NINJ1^−/− mice are not protected against endotoxin shock, suggesting that pyroptotic death itself and the release of smaller DAMPs are sufficient to drive lethality. On the other hand, NINJ1 plays a role in host survival during Citrobacter rodentium infection [54]. Mechanistically, the N-terminal extracellular region is predicted to adopt an amphipathic helix, which is proposed to play a role in the PMR [54]. NINJ1 exists in unstimulated cells as a dimer or trimer and undergoes higher-order oligomerization upon pyroptotic signaling [54]. The exact signal for NINJ1’s transition to the oligomeric state remains to be identified.

Release of DAMPs and alarmins upon noncanonical inflammasome activation

Apoptosis is generally considered an immunologically silent form of cell death, whereas programmed necrosis is inflammatory partly due to DAMP/alarmin release [57,58]. In fact, DAMP/alarmin release is considered a hallmark of inflammatory cell death pathways, such as necroptosis and pyroptosis [58–60]. Unlike the conventional cytokines with an N-terminal signaling sequence that directs their release through the ER-Golgi secretory pathway, DAMPs and alarmins lack the signal sequence for the ER-Golgi secretion [61]. These intracellular proteins are released when plasma membrane integrity is compromised and function as DAMPs/alarmins in the extracellular space [58,60–62]. Emerging proteomic studies are revealing the global secretome of pyroptosis and the differences between the secretome of pyroptosis and necroptosis [50,63,64]. During pyroptosis, GSDMD pores, and microvesicles, and cell lysis contribute to the release of DAMPs/alarmins to varying degrees [48,65–67]. Secretory lysosomes also contribute to unconventional cytokine release [68]. Comparatively, most proteins are secreted in soluble form rather than in EVs during pyroptosis [50].

DAMPs can act on a wide range of cell types and regulate both innate and adaptive immune responses. Several biological functions ranging from inflammation to immunoregulation have been attributed to DAMPs [58–60]. As the circulating levels of DAMPs/alarmins are elevated during inflammatory conditions, they are also considered as biomarkers. HMGB1 is one such inflammatory DAMP released by cells undergoing pyroptosis [69]. Extracellular HMGB1 plays a detrimental role in several inflammatory conditions such as sepsis and ischemia-reperfusion injury. Increased levels of HMGB1 have been found in patients with severe sepsis and septic shock and were associated with morbidity and mortality [70–72]). HMGB1 exerts its inflammatory effects via TLR4 and receptor for advanced glycation end products (RAGE) [58].

While extracellular LPS signaling through TLR4 can trigger the conventional secretion of cytokines like TNF, cytosolic LPS sensing leads to the unconventional release of cytokines, like IL-1β and IL-18 as well as DAMPs/alarmins [13,73]. Whereas leaderless proteins released during canonical inflammasome-driven pyroptosis are known, those released upon noncanonical inflammasome activation remain poorly defined [73]. A liquid phase fractionation-based proteomic approach identified that the noncanonical inflammasome activation leads to the release of galectin-1 [48]. Galectin-1 belongs to a family of β-galactoside-binding proteins that lack signal peptides for the classical secretion [74]. As galectin-1 is a 14.5 kDa protein, its release is mediated via GSDMD pores without requiring terminal cell lysis. Moreover, galectin-1 release occurs upon canonical inflammasome activation and necroptosis, indicating that its release is a universal outcome of inflammatory cell death [48]. Human sepsis patients have increased levels of galectin-1 in the serum [48]. Importantly, galectin-1 binds the glycan ligands generated by Mgat5 and C2gnt1 and limits the anti-inflammatory activity of CD45, enhancing inflammatory responses during endotoxemia. In agreement with this, functional studies in vivo revealed a detrimental role for galectin-1 in sepsis. [69]

SQSTM1 is another DAMP released during caspase-11-GSDMD mediated pyroptosis [75]. In addition, SQSTM1’s release can also be triggered by TLR4 signaling; TBK1 activated by TLR4 can phosphorylate SQSTM1 at serine 403, leading to SQSTM1’s release. Furthermore, secreted SQSTM1 signals through the insulin receptor to activate NF-kB and mediates the pathophysiology of septic shock. Moreover, circulating SQSTM1 levels correlated positively with sepsis severity—as indicated by sequential organ failure assessment (SOFA) and disseminated intravascular coagulopathy (DIC) scores—and mortality in humans [75].

Functions of the noncanonical inflammasome in health and disease

Cytosolic LPS sensing by caspase-11 plays a protective role in several bacterial infections [73,76]. For example, caspase-11 deficient mice are highly susceptible to infection with cytosol-invasive Burkholderia thailandensis and B. pseudomallei [77,78]. Although T3SS of Burkholderia can activate GSDMD via the NLRC4 inflammasome-caspase-1 pathway to induce pyroptosis, the NLRC4 pathway is insufficient to protect against B. thailandensis infection, and GSDMD-activated by caspase-11 is crucial for bacterial clearance. Similarly, caspase-4 mediates intracellular restriction of B. pseudomallei in human alveolar epithelial cells [79]. Furthermore, caspase-11^−/− mice have defective neutrophilic responses and bacterial clearance during Klebsiella pneumoniae or Acinetobacter baumannii infections [80,81]. Caspase-11 activation has also been shown to trigger neutrophil extracellular traps and cell death and contribute to bacterial control during Salmonella infection [82].

While cytosolic LPS sensing has beneficial functions in certain bacterial infections, the uncontrolled activation has detrimental consequences for the host and can be lethal [13,69,78]. Several regulatory mechanisms have been shown to limit noncanonical inflammasome activation and its deleterious effects [83–86]. An interferon inducible immunity-related GTPase, Irgm2, has been found to negatively regulate caspase-11 activation and sepsis in cooperation with the noncanonical autophagy protein Gate-16 [85,87]. Furthermore, the oxidative stress sensor GPx8 binds covalently to caspase-4/11 and inhibits their oligomerization, and mice deficient for GPx8 have enhanced caspase-11 activation and mortality in endotoxemia [87]. SERPINB1 and the E3 ubiquitin ligase Nedd4 suppress noncanonical inflammasome responses by targeting caspase-11 oligomerization and protein level, respectively [88,89]. The second messenger, cAMP, has also been shown to inhibit caspase-11 activation and attenuate endotoxin shock [90].

Dysregulated responses to cytosolic LPS is a significant contributor to lethality during sepsis as caspase-11- or GSDMD-deficient mice are highly protected in murine models of sepsis [13,48,91]. The mechanisms underlying caspase-11 mediated lethality include pyroptosis and the simultaneous release of DAMPs such as HMGB1, galectin-1, and SQSTM1 [13,21,48,69,71,75]. DAMPs secreted through GSDMD pores and cell lysis fuel inflammation, and their pharmacological or genetic inhibition reduces caspase-11-driven lethality [21,48,69,71,75]. Disseminated intravascular coagulation (DIC) also contributes to noncanonical inflammasome-induced death [92,93]. Calcium influx induced by GSDMD pore formation triggers the exposure of phosphatidylserine on the plasma membrane via the activation of transmembrane protein 16F. Phosphatidylserine enhances the coagulant activity of tissue factor-3, which leads to DIC [92,93]. The noncanonical inflammasome has also been shown to be involved in multiple non-infectious diseases such as colitis, colitis-associated cancer, colonic dysmotility, allergic airway inflammation, multiple sclerosis, atherosclerosis, and nonalcoholic steatohepatitis [76,94] [95,96] [97] [98][93] [99] [100][101,102]. In summary, cytosolic LPS sensing is important for the defense against certain bacterial infections; however, its hyperactivation has detrimental effects.

Cell type-specific physiological and pathological roles of the noncanonical inflammasome

A broad range of hematopoietic and nonhematopoietic cells—such as macrophages, dendritic cells (DC), neutrophils, intestinal epithelial cells (IEC), endothelial cells, airway epithelial cells, fibroblasts, and neurons [8,9,12,82,91,103–105]—express caspase-4/11 and GSDMD and are responsive to intracellular LPS. Growing evidence suggests cell-type-specific contributions of the noncanonical inflammasome to antibacterial host responses and pathophysiological manifestations of sepsis. Caspase-11 activation in endothelial cells (EC) has been shown to contribute to acute lung injury [91]. Conditional deletion of endothelial caspase-11 conferred protection against endotoxin shock and polymicrobial sepsis. Hepatocyte-specific expression of caspase-11 also contributes to the release of DAMPs such as HMGB1 and IL-1α and lethality during endotoxin shock [21].

Furthermore, we recently identified differential contributions of monocyte/macrophage-, DC, neutrophil-, and IEC-specific caspase-11 to a spectrum of host responses to cytosolic LPS [106]. Mice lacking caspase-11 in monocytes/macrophages were more resistant than wild-type mice to LPS challenge. Surprisingly, mice lacking caspase-11 in DCs and IECs were also protected, albeit to a small extent, against lethal LPS shock. On the other hand, neutrophil-specific deletion of caspase-11 did not protect mice against LPS-induced lethality. Moreover, splenic and hepatic GSDMD cleavage mainly occurs in monocytes, macrophages, and DCs during endotoxemia [106]. Additionally, cytosolic LPS sensing in monocytes/macrophages is a primary contributor to DAMP release and organ damage.

On the other hand, caspase-11 in neutrophils and monocytes/macrophages was crucial for bacterial clearance and host survival during infection with B. thailandensis [106,107]. DC- and IEC-intrinsic caspase-11 expression is dispensable for the host protection against B. thailandensis infection [106]. Overall, cytosolic LPS sensing in monocyte/macrophages and neutrophils provides antibacterial protection. In contrast, cytosolic LPS sensing in monocytes/macrophages, DCs, endothelial cells, hepatocytes, and IECs contribute to the pathophysiology of sepsis to varying degrees. Furthermore, enteric-associated neuron-intrinsic caspase-11 mediates rapid and persistent neuronal loss and long-term gastrointestinal abnormalities during enteric infections [105]. All these studies show that the noncanonical inflammasome functions in several cellular compartments to control inflammatory and antimicrobial responses.

Suppression of the noncanonical inflammasome by bacterial pathogens

It is becoming increasingly apparent from recent studies that bacterial pathogens have evolved to inhibit the noncanonical inflammasome. OspC3, an effector protein of Shigella flexneri, directly targets caspase-4 to prevent its dimerization and activation [108]. A S. flexneri mutant lacking OspC3 triggers enhanced cell death and is more sensitive to bacterial clearance in a guinea pig model. Coxiella burnetii, the causative agent of Q fever, has been shown to target caspase-11-LPS interaction and suppress noncanonical inflammasome responses via its T4SS effector IcaA [109]. A type III secretion system effector, NleF, produced by Enteropathogenic E. coli (EPEC) inhibits the catalytic activity of caspase-4 by binding to its catalytic domain. Similarly, NleF of C. rodentium targets caspase-11 to suppress IL-18 activation and early neutrophil influx [110,111]. Furthermore, bacteriophage-encoded Shiga toxin (Stx) of enterohemorrhagic E. coli (EHEC) has recently been shown to suppress caspase-11 activation [112]. An EHEC mutant strain lacking Stx induces stronger noncanonical inflammasome responses than wild-type EHEC, and purified Stx2 is sufficient to suppress noncanonical inflammasome responses in macrophages and mice. Importantly, a genetically engineered C. rodentium strain that expresses Stx, a recently established murine model for EHEC infection, inhibits noncanonical inflammasome responses in vivo [112]. The molecular basis by which Stx inhibits the caspase-11-GSDMD interaction is yet to be fully elucidated, and it appears that Stx targets GSDMD downstream of caspase-11 activation. Given the evolutionary advantage of subverting the innate immune sensing, it is likely that additional bacterial proteins also interfere with the noncanonical inflammasome pathway.

Conclusions & outstanding questions

Rapid progress has been made in mapping out the signaling and functions of the noncanonical inflammasome pathway within a decade of its discovery. The interaction between the noncanonical inflammasome and the effector GSDMD is critical to the innate inflammatory response and host defense against bacterial insults. While this interaction’s structural and mechanistic basis is emerging, there is still much to learn regarding certain aspects of the noncanonical inflammasome. The cell biological mechanisms involved in the cytosolic translocation of LPS are still unclear, and whether endosomal transporters actively mediate this process is an area of interest. Caspases 4 and 11 have been presumed to have a limited set of substrates. However, recent work suggests additional substrates for the noncanonical inflammasome. How the processing of these potential new substrates shapes the responses to intracellular LPS during bacterial infections warrants further investigation. Additional studies are also needed to further understand the roles DAMPs and alarmins play in mediating noncanonical inflammasome functions. The discovery of NINJ1 as a mediator of PMR—challenging the long-held notion that the terminal lysis of cells undergoing pyroptotic death is a passive event—prompts many exciting questions including whether additional proteins participate in this programmed event. Though both human caspase-4 and caspase-5 are similar to caspase-11, the role of caspase-5 in innate immunity remains less clear. Finally, additional work on the role of noncanonical inflammasome in human diseases including sepsis could facilitate the development of new therapeutics.

Figure 1. Noncanonical inflammasome sensing of cytosolic LPS and activation of GSDMD.

LPS associated with intracellular Gram-negative bacteria or outer membrane vesicles (OMVs) gains access to the cytosol. GBPs act on the bacterial membranes to liberate LPS for recognition by the noncanonical inflammasome. The lipid A moiety of cytosolic LPS interacts with the CARD of caspase-11/4, resulting in caspase-11/4 oligomerization and activation. Active caspase-11/4 interact with GSDMD in an exosite-dependent but tetrapeptide-independent fashion and cleave it. The freed N-terminal domain of GSDMD forms of pores on the plasma membrane leading to the release of DAMPs/alarmins and the processing of IL-1β and IL-18 by the NLRP3 inflammasome. The pyroptotic cascade activates ninjurin-1, which mediates the plasma membrane rupture and terminal cell lysis (Illustration created with BioRender.com.)

Highlights.

• Inflammatory caspases of the noncanonical inflammasome function as intracellular LPS sensors.

• Active caspase-11/4 interact with GSDMD in an exosite-dependent but tetrapeptide-independent fashion and cleave it.

• N-terminal domain of GSDMD forms of pores on the plasma membrane leading to the release of DAMPs/alarmins and activation of the NLRP3 inflammasome.

• The pyroptotic cascade activates ninjurin-1, which mediates the plasma membrane rupture and terminal cell lysis.

• Noncanonical inflammasome-mediated sensing of cytosolic LPS plays a critical role in numerous bacterial infections and sepsis pathogenesis.