GSDMD autoinhibition is maintained by an intramolecular interaction between its C-terminal (C-GSDMD) and N-terminal (N-GSDMD) domains. Crystal structure analysis of the human GSDMD C-terminal domain (resolved at 2.64 Å) reveals that a key loop inserts into the N-terminal domain to stabilize this inhibited state [34]. This interaction masks the positively charged surface of the N-terminal domain, preventing membrane binding. Upon proteolytic cleavage by inflammatory caspases, the N-terminal domain is released and undergoes higher-order oligomerization via charge-charge interactions [35, 36]. These oligomers form dynamic pores (10–14 nm in diameter) across the plasma membrane, disrupting ion homeostasis and facilitating the secretion of pro-inflammatory cytokines, including IL-1β and IL-18. Additionally, internal cysteine residues contribute to the plasticity and modulation of this pore-formation process [37].

The canonical pathway is initiated by the assembly of multiprotein inflammasome complexes (e.g., NLRP3) in response to pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs) [38, 39]. These cytosolic sensors recruit procaspase-1 via the adaptor protein apoptosis-associated speck-like protein containing a caspase activation and recruitment domain (CARD, an adaptor protein) (ASC), triggering its autocatalytic cleavage into active caspase-1 [40, 41]. Active caspase-1 then cleaves GSDMD at a specific aspartate residue (Asp276 in mice) within the linker region, while simultaneously processing pro-IL-1β and pro-IL-18 into their mature forms [42, 43]. Beyond inducing pyroptosis, caspase-1 plays a dual role in neutrophils by modulating both apoptotic and inflammatory signaling to balance host defense [44].

The non-canonical pathway involves the direct sensing of cytosolic Gram-negative bacterial LPS by caspase-4/5 (human) or caspase-11 (mouse) [45, 46]. The binding of LPS to the CARD of these inflammatory caspases triggers oligomerization independently of TLR4 and ASC [47, 48]. These activated caspases directly cleave GSDMD to induce pyroptosis [49]. Crucially, the resulting pore-induced potassium (K⁺) efflux stimulates secondary activation of the NLRP3 canonical inflammasome, linking the two pathways to amplify the inflammatory response [50, 51].

GSDMD activity is finely tuned by various PTMs. For instance, AMPK phosphorylates the N-terminal domain at Ser46 to inhibit oligomerization, whereas TRAF1-mediated ubiquitination facilitates its membrane translocation [38, 39]. Similarly, the neddylation of the caspase-1 CARD domain by Nedd8 is essential for efficient self-cleavage [52, 53]. Negative regulators, including the E3 ubiquitin ligase Nedd4, Irgm2, and GATE-16, act as checkpoints to prevent excessive inflammation and sepsis [54, 55]. Furthermore, metabolic signals like cyclic AMP (cAMP) and proteases such as Cathepsin B modulate caspase-11 activation, while inhibitory proteins like SERPINB1 prevent spontaneous caspase oligomerization [56–58].

Dysregulated GSDMD activation contributes to several pathologies, including myocardial hypertrophy and fibrosis [59]. Beyond acute inflammation, caspase-4 activation by cytoplasmic LPS is linked to cellular senescence and the senescence-associated secretory phenotype (SASP) via the GSDMD and p53 pathways [60]. Pharmacological targeting of these pathways—using agents such as Korean Red Ginseng saponins—offers potential therapeutic avenues for managing chronic inflammation and lethal sepsis [61].

The NLRP3 inflammasome is a central component of the innate immune system, serving as an elaborate sensor for numerous environmental and endogenous danger signals. The NLRP3 protein consists of an N-terminal pyrin domain (PYD), a central nucleotide-binding and oligomerization domain (NOD, also known as NACHT domain), and a C-terminal leucine-rich repeat (LRR) domain, classifying it as a member of the NLR family [62, 63]. The NACHT domain is required for ATP-dependent oligomerization, while the LRR domain acts as a regulatory scaffold that maintains the protein in an autoinhibited conformation during resting states [64] (Figure 3). Activation follows a stringent two-step mechanism:

Priming (Signal 1): Triggered by TLR, NLR, or cytokine receptors (e.g., TNFR), leading to NF-κB-dependent transcriptional upregulation of NLRP3 and pro-IL-1β [65–67]. PTMs, such as deubiquitination by BRCC3, are critical during this phase to license NLRP3 for assembly [68, 69].

Unlike the versatile sensing of NLRP3, the AIM2 inflammasome specifically detects double-stranded DNA (dsDNA) in the cytosol. AIM2 is composed of a C-terminal HIN-200 domain that binds dsDNA and an N-terminal PYD domain that initiates downstream signaling [75–77].

Upon detecting dsDNA—whether from microbial pathogens (e.g., Francisella, Listeria) or host mitochondrial damage—AIM2 undergoes oligomerization [78, 79]. This process is regulated by type I interferons and guanylate-binding proteins (GBPs), which facilitate the liberation of DNA from intracellular vesicles [80]. Additionally, the stability of AIM2 is managed by the deubiquitinase USP21, which prevents premature proteasomal degradation [81].

Modern immunology recognizes that inflammasomes do not operate in isolation but interact through complex signaling networks.

Non-canonical to canonical bridge: Caspase-11/4/5 activation by LPS triggers GSDMD pores, which subsequently induce K⁺ efflux, thereby serving as a secondary signal for NLRP3 activation [82].

To prevent chronic inflammation and autoimmune pathologies, several regulatory “brakes” are employed:

Autophagy and mitophagy: These processes act as critical negative regulators by clearing assembled inflammasome complexes and damaged mitochondria, reducing ROS-mediated activation [69].

Dysregulation of NLRP3 and AIM2 is implicated in a spectrum of diseases, ranging from neurodegenerative disorders (Alzheimer’s, Multiple sclerosis) to metabolic syndromes (type 2 diabetes, gout) [81, 88, 89]. In oncology, inflammasome activity influences the tumor microenvironment by modulating cytokine milieus (IL-1β/IL-18) and adaptive immune recruitment [90]. Targeted inhibitors, like MCC950 and glyburide, are promising therapeutic approaches to reduce sterile inflammation in disorders like ischemia/reperfusion (I/R) injury and autoimmune myopathies [91].

SSc and atherosclerosis are increasingly recognized as related pathologies within the CVD spectrum. They share a common mechanistic substrate characterized by chronic inflammation, immune dysregulation, and progressive vascular injury [92]. The NLRP3-caspase-1-GSDMD axis is central to atherosclerosis; cleaved GSDMD assembles into membrane pores, inducing pyroptosis and releasing pro-inflammatory cytoplasmic contents, which exacerbate vascular inflammation [93]. Similarly, SSc is characterized by widespread endothelial dysfunction and vasculopathy, typically initiated by anti-endothelial cell (EC) antibodies and oxidative stress [94]. This injury impairs vascular relaxation, enhances leukocyte adhesion, and increases the risk of both organ fibrosis and accelerated atherosclerosis in patients with SSc [95, 96] (Figure 4).

A key link between these conditions is the role of endothelial GSDMD in enhancing vascular damage. Recent studies have shown that GSDMD pores allow the cytosolic release of mtDNA, which triggers an immune signaling cascade via the cGAS-STING pathway [97]. This feed-forward loop leads to increased production of mitochondrial reactive oxygen species (ROS) and STING-dependent inflammatory signaling, which directly promotes atherogenesis [97]. Although this mechanism has not been specifically studied in SSc yet, it is highly probable, as clinical reports have documented augmented arterial stiffness and increased carotid intima-media thickness in SSc patients—both of which are hallmarks of accelerated vascular aging and subclinical atherosclerosis [98].

The NF-κB pathway largely governs inflammatory signaling in both SSc and atherosclerosis, driving the recruitment of macrophages and T cells to the vascular wall. This process is finely tuned by miRNAs and other noncoding RNAs that modulate cytokine production (TNF-α, IL-1β, IL-6) and vascular smooth muscle cell (VSMC) proliferation [99]. Given the central role of the NLRP3-GSDMD axis in executing pyroptosis, targeting this pathway represents a promising therapeutic avenue. Anti-inflammatory biologics, particularly those inhibiting TNF-α and IL-1β, effectively reduce vascular inflammation and plaque progression. These treatments potentially attenuate both the fibroproliferative changes in SSc and the associated accelerated cardiovascular risks [100].

Macrophage pyroptosis is a decisive factor in the progression and eventual rupture of atherosclerotic plaques. Within the lipid-rich microenvironment of a plaque, macrophages—the predominant immune cells—undergo GSDMD-mediated lysis, transforming stable lesions into vulnerable, pro-inflammatory sites [101].

The vascular endothelium consists of a monolayer of ECs that play a vital role in maintaining vascular tone, barrier integrity, and blood fluidity. Endothelial dysfunction involves the loss of these homeostatic functions and is characterized by impaired vasodilation—often due to decreased nitric oxide (NO) bioavailability—increased permeability, and pro-inflammatory activation [105–107].

EC injury is primarily driven by pyroptosis. The overproduction of ROS causes oxidative damage to mitochondria and cellular DNA, which triggers the assembly of the NLRP3 inflammasome. This assembly activates caspase-1, which cleaves GSDMD to form membrane pores, resulting in lytic cell death and the release of pro-inflammatory cytokines [106, 108].

In vitro and in vivo observations, this process disrupts the physical barrier of the vessel and recruits immune cells, thereby accelerating atherosclerotic plaque deposition and arterial stiffening [101]. Furthermore, chronic conditions such as diabetes and hypertension exacerbate oxidative stress, driving phenotypic shifts like the endothelial-to-mesenchymal transition (EndMT), which further destabilizes plaques.

Recent research highlights protective signaling networks that can mitigate this damage:

The melatonin/RORα/miR-223 axis: Melatonin signaling promotes miR-223 transcription, which suppresses pyroptotic proteins and adhesion molecules like ICAM-1, effectively restoring NO levels and vascular health [109].

GSDMD inhibitors have emerged as promising agents to block pyroptosis by targeting various stages of GSDMD activation and pore formation:

GI-Y1: This molecule selectively binds to the Arg7 residue of the GSDMD-N domain. It prevents mitochondrial binding and inhibits pore formation on the plasma membrane. GI-Y1 has demonstrated significant cardioprotective effects in MI/R models [137].

Caspase inhibitors prevent the proteolytic cleavage of full-length GSDMD into its active, pore-forming fragments.

VX-765: A potent caspase-1 inhibitor currently in clinical testing. It effectively reduces GSDMD cleavage in neurotoxicity and IBD models [139, 140].

Palmitoylation inhibitors: Targeting GSDMD palmitoylation at conserved cysteines (using agents like 2-bromopalmitate) prevents membrane localization and cytokine release [143].

The cargo of EVs reflects the physiological state of their origin cells, making them ideal candidates for biomarker development. In cardiovascular and inflammatory diseases, the detection of circulating EVs containing GSDMD-related fragments (such as the p30 N-terminal fragment) and inflammasome components provide non-invasive biomarkers for disease monitoring and personalized medicine [145].

EVs possess low immunogenicity and natural targeting capabilities, providing versatile platforms for the delivery of GSDMD inhibitors directly to fibrotic or inflamed vascular tissues.

Synergistic therapy: The co-delivery of classic anti-inflammatory agents (e.g., colchicine or IL-1β blockers) with GSDMD inhibitors can potentiate therapeutic efficacy [146].

Translating EVs-based therapies into clinical practice requires addressing several critical hurdles:

Standardization: There is a lack of standard protocols for EV isolation and characterization [147, 148].