Decoding NLRP3: Phase separation enters the scene

Activation of the innate immune sensor protein NLRP3 leads to the assembly of a multiprotein complex called the inflammasome, causing cell death and inflammation. In a recent paper in Cell Research, Zou et al. now provide evidence that palmitoylation of NLRP3 promotes its liquid–liquid phase separation, driving inflammasome activation.

Inflammation is the body’s natural response to infection, driven by the innate immune system to eliminate pathogens and restore tissue integrity. However, it can also be initiated through sterile triggers such as tissue damage or metabolic stress. While essential for defense and repair, persistent inflammation plays a central role in the pathogenesis of many non-communicable diseases, including gout, atherosclerosis, and Alzheimer’s disease.

A key mechanism by which the immune system initiates inflammation is the assembly of a large cytosolic multiprotein complex called the inflammasome.¹ Inflammasomes are composed of a sensor protein that nucleates the complex through polymerization, the effector protease caspase-1 and, in many cases, an adaptor protein called ASC that links the sensor to caspase-1. Upon inflammasome assembly, caspase-1 becomes activated and proceeds to cleave key substrates, most notably the pore-forming protein Gasdermin D and the pro-inflammatory cytokine IL-1β. Cleavage of Gasdermin D induces pyroptosis, a lytic form of cell death that facilitates the release of IL-1β and other inflammatory mediators, amplifying the immune response.¹ A variety of sensor proteins capable of initiating inflammasome formation in response to different stimuli have been identified. Among them, NLRP3 has emerged as a particularly prominent and clinically relevant sensor, due to its involvement in a broad range of pathological contexts — from infectious diseases to sterile inflammatory conditions such as gout, atherosclerosis, and neurodegeneration.²

The reason that NLRP3 is implicated in such a wide range of conditions lies in its ability to be activated by a remarkably diverse array of stimuli, a property that has remained enigmatic. It became apparent early on that NLRP3 does not recognize a specific ligand, but instead senses disturbances to cellular integrity — distinguishing it from classical pattern recognition receptors. A widely accepted model posits that NLRP3 is activated in response to K⁺ efflux, often resulting from plasma membrane damage. This model explains how a broad spectrum of cell-damaging stimuli — both pathogenic and sterile — can converge on NLRP3 activation.^3,4 To mitigate the risk of unintended activation and its pro-inflammatory consequence, the system includes a safeguard known as “priming” (or “signal 1”). Before NLRP3 can be activated, it must be licensed into an ill-defined responsive state by upstream signals, typically provided by pattern recognition receptors. Priming is traditionally linked to increased NLRP3 expression at the transcriptional level, along with post-translational modifications that enable its activation. More recently, attention has turned to the link between NLRP3’s subcellular localization and its ability to get activated. Various organelles including the trans-Golgi network (TGN),^5,6 mitochondria,⁷ and endosomes⁸ have been proposed as activation platforms or hubs for inflammasome assembly. However, delineating the distinction between priming and activation remains a central challenge in the field.

To shed light on this process, Zou et al. set out to identify signals required for NLRP3 priming using an elegant approach that decoupled priming from activation.⁹ They employed a gain-of-function NLRP3 mutant causing the autoinflammatory disorder Muckle-Wells syndrome, which renders NLRP3 activation- competent without the need for a second signal. A human myeloid cell line (THP-1) expressing this mutant form of NLRP3 requires only a priming stimulus to trigger inflammasome activation. Using this system, the authors conducted a forward genetic screen and identified multiple genes whose loss conferred a survival advantage following signal 1 provision. In addition to known components of NF-κB signaling, the TGN-resident palmitoyl acyltransferase ZDHHC7 emerged as a regulator of NLRP3 activation. In agreement with prior work,^10,11 they confirmed that NLRP3 is palmitoylated by ZDHHC7 at Cys126 in mice (Cys130 in humans) and further identified Cys261 as an additional palmitoylation site in human NLRP3. Disruption of these cysteines impaired NLRP3 palmitoylation and abolished inflammasome activation. The authors also re-examined a conserved polybasic motif previously thought to be essential for intracellular membrane recruitment and NLRP3 activation adjacent to the palmitoylation site. Intriguingly, while mutations in this motif disrupted vesicle localization, they did not impair inflammasome activation if palmitoylation was intact. However, some mutations within the motif also impaired palmitoylation.

Having identified palmitoylation as a prerequisite for NLRP3 activation, the authors examined how this lipidation promotes inflammasome formation. Using live cell microscopy, they observed that, upon stimulation with NLRP3 activators, NLRP3 rapidly formed cytoplasmic puncta exhibiting dynamic, liquid-like behavior, consistent with liquid–liquid phase separation (LLPS). This process was strictly dependent on palmitoylation; non-palmitoylated NLRP3 failed to form puncta or support inflammasome activation. To test whether vesicle localization was necessary for this behavior, the authors used human NLRP3 mutants lacking the polybasic motif required for membrane recruitment. Surprisingly, these mutants still formed condensates and supported inflammasome function, providing that palmitoylation and a short intrinsically disordered region (IDR) in the FISNA domain were intact. Within this IDR, three conserved hydrophobic residues were essential for phase separation. Their mutation disrupted NLRP3 aggregation and abolished activation, even though palmitoylation was unaffected — highlighting that both hydrophobic multivalent interactions and lipidation are required. To further explore the idea that NLRP3 phase separation is a direct and regulated event, the authors reconstituted the process in vitro using recombinant, palmitoylated NLRP3 purified from cells. Lowering the K⁺ concentration — a known activator of NLRP3 — was sufficient to trigger phase separation in solution, directly linking this ionic perturbation to the condensation of the protein. Finally, they demonstrated that amphiphilic small molecules, including diols and chemotherapeutic agents, could induce NLRP3 condensation by reducing its solubility even in the absence of palmitoylation. Together, these findings imply phase separation as a unifying mechanism of NLRP3 activation, integrating structurally diverse upstream signals into a biophysical response.

With this study, Zou et al. offer a fresh perspective on NLRP3 activation, proposing a model that shifts the focus from membrane association to post-translational regulation and phase behavior (Fig. 1). The authors suggest that membrane localization is not required for activation and that palmitoylation is sufficient to render NLRP3 responsive. In parallel, the authors suggest that LLPS, which was previously suggested to mediate activation of autoactive NLRP3 mutants,¹² may serve as a common endpoint for diverse NLRP3-activating signals. In this model, palmitoylation reduces NLRP3 solubility, enabling it to condense in response to triggers such as ionic efflux, mitochondrial lipids, or small molecules. An in vitro experiment that tentatively supports this model involved the recombinant reconstitution of NLRP3 LLPS, where a simple reduction in K⁺ concentration was sufficient to induce condensation. This observation suggests that K⁺ efflux might directly trigger NLRP3 condensation, although further validation is needed to confirm the specificity of this response — particularly by testing whether related NLR family members behave differently under comparable conditions. More broadly, it remains an open question how this phase separation model integrates with existing structural data, particularly the disc-like inflammasome complex visualized by cryo-EM. Understanding how phase-separated NLRP3 transitions into a defined multimeric assembly will be key to reconciling the dynamic, disordered nature of LLPS with the highly ordered structure of the activated inflammasome.

Fig. 1. NLRP3 activation according to Zou et al.

NLRP3 is palmitoylated by the acyltransferase ZDHHC7. This post-translational modification serves as a priming step that enables NLRP3 to undergo LLPS, either in response to K⁺ efflux or via LLPS-inducing stimuli. Inferring from previous work, it appears plausible that NLRP3 exists in a cage-like, autoinhibited conformation prior to palmitoylation and that it assembles into an inflammasome disc structure following LLPS. However, it should be noted that these conformational stages were not addressed in this study and remain to be explored in future research.

This study edges us closer to a unifying explanation for NLRP3 activation — but if past experience is any guide, NLRP3 is unlikely to give up all its secrets so easily.

Competing interests

The authors declare no competing interests.