Mitochondrial damage activates the NLRP10 inflammasome

Upon detecting pathogens or cell stress, several NOD-like receptors (NLRs) form inflammasome complexes with the adapter ASC and caspase-1, inducing gasdermin D (GSDMD)-dependent cell death and maturation and release of IL-1β and IL-18. The triggers and activation mechanisms of several inflammasome-forming sensors are not well understood. Here we show that mitochondrial damage activates the NLRP10 inflammasome, leading to ASC speck formation and caspase-1-dependent cytokine release. While the AIM2 inflammasome can also sense mitochondrial demise by detecting mitochondrial DNA (mtDNA) in the cytosol, NLRP10 monitors mitochondrial integrity in an mtDNA-independent manner, suggesting the recognition of distinct molecular entities displayed by the damaged organelles. NLRP10 is highly expressed in differentiated human keratinocytes, in which it can also assemble an inflammasome. Our study shows that this inflammasome surveils mitochondrial integrity. These findings might also lead to a better understanding of mitochondria-linked inflammatory diseases. NLRP10 has been considered as an inflammasome inhibitor. Here the authors show that upon mitochondrial rupture, NLRP10 assembles a canonical inflammasome and is highly expressed in differentiated keratinocytes, possibly supporting skin homeostasis.

Upon detecting pathogens or cell stress, several NOD-like receptors (NLRs) form inflammasome complexes with the adapter ASC and caspase-1, inducing gasdermin D (GSDMD)-dependent cell death and maturation and release of IL-1β and IL-18. The triggers and activation mechanisms of several inflammasome-forming sensors are not well understood. Here we show that mitochondrial damage activates the NLRP10 inflammasome, leading to ASC speck formation and caspase-1-dependent cytokine release. While the AIM2 inflammasome can also sense mitochondrial demise by detecting mitochondrial DNA (mtDNA) in the cytosol, NLRP10 monitors mitochondrial integrity in an mtDNA-independent manner, suggesting the recognition of distinct molecular entities displayed by the damaged organelles. NLRP10 is highly expressed in differentiated human keratinocytes, in which it can also assemble an inflammasome. Our study shows that this inflammasome surveils mitochondrial integrity. These findings might also lead to a better understanding of mitochondria-linked inflammatory diseases.
Inflammasomes are protein complexes bridging the recognition of pathogens and sterile damage to GSDMD-mediated pyroptotic cell death and IL-1β-driven inflammation 1 . Genetic factors, aging and metabolic dysfunction promote chronic inflammasome activation, contributing to multiple inflammatory diseases 2 . Therefore, members of this class of innate immune receptors represent possible pharmacological intervention targets 3 . Many inflammasome-forming sensors belong to the NACHT, leucine-rich repeat (LRR) and pyrin domain (PYD)-containing (NLRP) protein family. Their LRR domains participate in recognizing perturbations of key cellular processes and microbeor stress-associated ligands, whereas the PYDs typically recruit the adapter ASC, providing a platform for pro-caspase-1 activation 4 . The roles of several NLRP family members and their activating signals are incompletely characterized, impeding our understanding of pathological processes and drug development.
To identify which inflammasome sensor responds to m-3M3FBS, we performed an overexpression screen in HEK cells with a fluorescent ASC (ASC TagBFP ) reporter. Expression of NLRP10, but not other NLRPs (Fig. 1c), or other inflammasome-forming proteins (Extended Data Fig. 1h), enabled ASC speck formation in m-3M3FBS-treated HEK cells. This NLRP10-mediated inflammasome response was specific to m-3M3FBS, as nigericin and poly(dA:dT) did not cause ASC speck formation in NLRP10 reporter cells (Extended Data Fig. 1i).
To determine the upstream signal for NLRP10 activation, we tracked the subcellular localization of fluorescently tagged NLRP10 and protein-lipid interactions 9,10 . Our initial investigation of the putative NLRP3 activator m-3M3FBS, a reported phospholipase C agonist [11][12][13] , unexpectedly revealed that this molecule, in fact, triggers NLRP3-independent inflammasome activation.
Here, we discovered that m-3M3FBS and other molecules (thapsigargin and SMBA1) trigger mitochondrial damage leading to NLRP10 and AIM2 inflammasome activation in keratinocytes and macrophages, respectively. While AIM2 detected release of mitochondrial DNA, the novel NLRP10 inflammasome sensed DNA-independent factors in damaged mitochondria.
We next examined the inflammasome response to mitochondrial rupture in mouse macrophages. These cells do not express NLRP10 (refs. [24][25][26][27] but nevertheless responded to m-3M3FBS (Fig. 2a,b), suggesting that another inflammasome-forming sensor is engaged in this cell type. We had already excluded a contribution of NLRP3 in the response to m-3M3FBS ( Fig. 1b and Extended Data Fig. 1c-g), which was confirmed for thapsigargin and SMBA1 (Extended Data Fig. 4a). As AIM2 is expressed in macrophages and was identified as a sensor for DNA leaking from damaged mitochondria 28 , we tested whether IL-1β secretion induced by the mitochondria-damaging agents was inhibited in AIM2-deficient macrophages. Indeed, neither m-3M3FBS, thapsigargin (Fig. 2c) nor SMBA1 (Extended Data Fig. 4b) triggered IL-1β release from AIM2-deficient macrophages. Similar to m-3M3FBS (Fig. 2a,b), thapsigargin and SMBA1 induced ASC speck formation in macrophages (Extended Data Fig. 4c). These results suggest that these mitochondria-damaging stimuli activate the AIM2 inflammasome in macrophages.
To identify the mechanism of AIM2 activation by m-3M3FBS and thapsigargin, we depleted mtDNA using 2′,3′-dideoxycytidine (ddC) in AIM2-deficient immortalized macrophages overexpressing AIM2. The mtDNA depletion efficiency was confirmed by quantitative PCR (qPCR) (Extended Data Fig. 4d). We found that m-3M3FBS-and thapsigargin-mediated AIM2 activation depended on mtDNA (  Letter https://doi.org/10.1038/s41590-023-01451-y 2d). Since AIM2-deficient immortalized macrophages no longer responded to mitochondrial damage, we overexpressed NLRP10 in these cells and found that this was sufficient to enable IL-1β release upon treatment with m-3M3FBS and thapsigargin (Fig. 2e). Notably, mtDNA depletion in these cells did not prevent NLRP10 activation, suggesting that alternative ligands must be exposed by damaged mitochondria to trigger NLRP10 inflammasome formation (Fig.  2e). These data were corroborated in NLRP10 reporter HEK cells (Extended Data Fig. 4e,f). We next assessed the mechanism by which m-3M3FBS, thapsigargin and SMBA1 damage mitochondria. We determined that the mPT inhibitor cyclosporin A (CsA) 22,29-31 selectively blocked mitochondrial rupture and NLRP10 inflammasome activation in thapsigargin-but not m-3M3FBS-treated HEK cells (Fig. 3a,b and Extended Data Fig. 5a). Similarly, CsA blocked mitochondrial damage, and AIM2-mediated ASC speck formation and IL-1β release in macrophages stimulated with thapsigargin but not with m-3M3FBS (Fig. 3c-e and Extended Data Fig. 5b). SMBA1 showed the same CsA sensitivity profile as thapsigargin, with CsA treatment entirely preventing NLRP10-mediated ASC speck formation in HEK cells (Extended Data Fig. 6a,b) and AIM2-mediated ASC speck formation in macrophages (Extended Data Fig. 6c,d).
CsA inhibits both mPT and calcineurin/nuclear factor of activated T cells 32 . Nonimmunosuppressive analogs of CsA, Debio025 and NIM811 are, in contrast, selective mPT blockers 33,34 . We observed that, similar to CsA, Debio025 and NIM811 inhibited NLRP10 and AIM2 inflammasome responses to thapsigargin but not to m-3M3FBS (Extended Data Fig. 7a-f). Together, these data confirm that mitochondrial damage is upstream of NLRP10 and AIM2, but the activation mechanisms of these inflammasomes likely differ.
Whereas some models suggest that mPT is initiated by mitochondrial Ca 2+ fluxes 35,36 , we did not observe a strong link between Ca 2+ signaling and NLRP10 activation. m-3M3FBS still elicited ASC speck formation in the presence of the intracellular Ca 2+ chelator BAPTA-AM (Extended Data Fig. 8a) and of the inositol 1,4,5-trisphosphate receptor blocker 2-APB (Extended Data Fig. 8b). Administration of ionomycin, globally increasing the cytosolic Ca 2+ concentration, did not damage the mitochondria, as measured by the mito-mCherry reporter (Extended Data Fig. 8c,d), and did not cause NLRP10-driven ASC speck formation (Extended Data Fig. 8c-e). Conversely, m-3M3FBS remained capable of eliciting ASC speck formation in NLRP10 reporter HEK cells pretreated with ionomycin (Extended Data Fig.  8f), indicating that cytosolic Ca 2+ does not inhibit NLRP10 under these conditions.
Since our work suggested that NLRP10 functions as an inflammasome sensor, we tested the requirement for the different NLRP10 domains in inflammasome formation. We expressed full-length NLRP10 or the individual NLRP10 PYD and NACHT domains (NLRP-10 PYD , NLRP10 NACHT ) in AIM2-deficient macrophages and found that neither individually expressed NLRP10 PYD nor NLRP10 NACHT enabled inflammasome activation (Fig. 3f), demonstrating the requirement for the full-length protein. Furthermore, we profiled Walker A (K179M) and B (D249N) NLRP10 mutants 37 and found that they cannot assemble inflammasomes (Fig. 3f), which is in accordance with the reports that NLRP3 Walker A/B mutants are inactive 38,39 and that NLRP10 Walker A mutant is deficient in NF-κB induction in response to Shigella 25 . These data show that full-length NLRP10 with active Walker A/B sites is required for sensing mitochondrial damage and inflammasome function.
We next tested whether NLRP10 recruits ASC. We thus monitored the localization of NLRP10 mCitrine and ASC TagBFP in untreated and m-3M3FBS-or thapsigargin-stimulated HEK cells. Indeed, NLRP10 colocalized with ASC following m-3M3FBS and thapsigargin challenge but not under resting conditions, suggesting that NLRP10 recruits ASC upon sensing mitochondrial damage (Fig. 3g). Consistently, NLRP10 mCitrine coimmunoprecipitated with ASC TagBFP after stimulation with m-3M3FBS or thapsigargin but not under basal conditions (Fig. 3h), confirming the inducibility of the NLRP10-ASC interaction.
To interrogate the requirement of caspase-1 for the release of IL-1β by the mitochondrial damage-sensing inflammasomes, we used the caspase-1 inhibitor VX-765 (ref. 40 ) and the pan-caspase inhibitor emricasan 41 . VX-765 and emricasan blocked IL-1β secretion driven by both NLRP10 (Fig. 3i) and AIM2 ( Fig. 3j and Extended Data Fig. 9a), suggesting that these inflammasomes engage caspase-1. Importantly, VX-765 and emricasan did not block NLRP10-( Fig. 3k and Extended Data Fig. 9b) or AIM2-mediated ( Fig. 3l and Extended Data Fig. 9c) ASC speck formation, indicating that ASC speck assembly does not require upstream caspase activity. Instead, caspase-1 is likely engaged by the inflammasome sensors and ASC following mitochondrial damage.
Nature Immunology | Volume 24 | April 2023 | 595-603 599 Letter https://doi.org/10.1038/s41590-023-01451-y Since NLRP10 expression is enriched in terminally differentiated keratinocytes in the stratum granulosum of human epidermis (Fig. 4a), we next differentiated normal primary human epidermal keratinocytes (NHEKs) using CaCl 2 , which induced an increase in NLRP10 mRNA expression (Extended Data Fig. 10a). Correspondingly, immunoblotting analysis confirmed that N/TERT keratinocyte ****  Letter https://doi.org/10.1038/s41590-023-01451-y differentiation induced, along with the differentiation marker involucrin, NLRP10 protein expression from low but detectable levels in undifferentiated, to high levels in differentiated, N/TERT keratinocytes (Extended Data Fig. 10b). In these differentiated cells expressing endogenous NLRP10, m-3M3FBS treatment markedly increased secretion of mature IL-1β compared to undifferentiated cells (Fig. 4e,f). Importantly, in some experiments, low levels of NLRP10 were detected in keratinocyte populations without administration of the differentiating stimulus, possibly due to spontaneous differentiation of small cell subsets. These low levels of NLRP10 are likely responsible for IL-1β release from undifferentiated keratinocytes stimulated with m-3M3FBS. We next generated NLRP10-deficient N/TERT keratinocytes (Extended Data Fig. 10c) to assess whether endogenous NLRP10 was nonredundantly required for inflammasome activation by mitochondria-damaging agonists in differentiated N/TERT keratinocytes. This analysis showed that partial knockout of NLRP10 decreased IL-1β release following m-3M3FBS administration (Fig. 4g). Furthermore, NLRP10 was required to activate a caspase-1-CARD reporter in m-3M3FBS-stimulated N/TERT keratinocytes (Extended Data Fig. 10d). Using differentiated caspase-1-deficient N/TERT keratinocytes 45 , we demonstrated that caspase-1 was essential for NLRP10-driven IL-1β release (Fig. 4g).
We validated the specificity of an anti-human NLRP10 monoclonal antibody (Extended Data Fig. 10e) and found that NLRP10 is exclusively expressed in the uppermost layer of differentiated human keratinocyte cultures (Extended Data Fig. 10f). Using this monoclonal antibody, we assessed endogenous NLRP10 activation in NHEKs. Consistent with the observations from the overexpression system (Fig. 1d), m-3M3FBS elicited the formation of endogenous NLRP10 puncta in differentiated NHEKs ( Fig. 4h and Extended Data Fig. 10g). Together, these data confirm that endogenous NLRP10 drives inflammasome formation upon mitochondrial damage, leading to IL-1β maturation and release from differentiated keratinocytes.
NLRP10 variants have been linked with an increased risk of atopic dermatitis (AD) 46,47 . Intriguingly, one of these reported AD-associated variants produces a missense R243W mutation near the NLRP10 Walker B site. In the NLRP10 reporter system, we found that the R243W variant had a loss-of-function phenotype (Extended Data Fig. 10h), similar to the D249N NLRP10 Walker B mutant (Fig. 3f). Thus, the NLRP10 ability to assemble an inflammasome might protect against skin inflammation, suggesting that NLRP10 could liberate factors promoting skin homeostasis upon sensing physiological triggers in the stratum granulosum.
Our study demonstrates that NLRP10 nucleates inflammasomes in cells treated with m-3M3FBS, thapsigargin or SMBA1. In contrast to earlier reports suggesting an inflammasome-inhibitory role for NLRP10 (refs. 48,49 ), the only NLRP family member lacking the LRR domain 50 , we found that NLRP10 is a mitochondrial-damage sensor. Our results also confirm the capacity of AIM2 to recognize mitochondrial disruption and mtDNA cytosolic leakage, which was previously shown using other mitochondria-disrupting agents 28 . While this manuscript was under review, an unrelated study confirmed the mitochondria-rupturing m-3M3FBS activity 36 . Mitochondrial disruption described in this report was independent of most known factors mediating mitochondrial membrane permeabilization. In contrast to our observations, this mitochondrial permeabilization mechanism relied on intracellular Ca 2+ fluxes. Future studies may reconcile these models, as the apparent differences may be due to subtle divergences in inhibitor or activator concentrations, administration modes or analyzed cell types.
Physiologically, NLRP10 is expressed in keratinocytes and cardiomyocytes 25,51 . These cell types are exposed to different sterile and microbial stimuli, suggesting that diverse upstream signals could converge on a common proximal event activating NLRP10. In the skin, commensal and invasive microbes, ultraviolet radiation and environmental chemicals could provide signals for mitochondrial damage and NLRP10 activation. In contrast, cardiomyocytes are shielded from these agents, but cell-intrinsic processes, including oxidative stress, may regulate their mitochondrial integrity and NLRP10 activation status.
The AD-associated NLRP10 variant 46 was unable to initiate inflammasome activation. An essential future question is how the inflammasome-forming activity of NLRP10 is linked to its role in AD pathogenesis. As NLRP10-expressing keratinocytes in the stratum granulosum undergo physiological cell death producing a protective layer of cornified epithelium, it is tempting to speculate that this process involves NLRP10.
Further supporting the pathophysiological importance of NLRP10, a study by Zheng et al. 52 , cosubmitted with this study, uncovered a protective role of NLRP10 inflammasome in colitis. Loss of NLRP10 activity was linked with more severe disease. Collectively, it is plausible that NLRP10-mediated loss of damaged cells favors the epithelial barriers' protective function.
Given the reported inflammasome-independent involvement of NLRP10 in regulating innate immune signaling and inflammatory gene transcription 25,42,43 , it will be valuable to examine under which conditions NLRP10 behaves as an inflammasome and when it exerts other functions, and how this distinction is made. Understanding the partial redundancy between NLRP10 and AIM2 as mitochondrial damage sensors, especially in in vivo studies, will be vital for dissecting their roles in host defense against microbes or sterile inflammatory processes. Such studies will likely reveal disease mechanisms in which targeting NLRP10 could provide a therapeutic benefit. Note that positive RNAscope signal appears as pink puncta. Brown staining is caused by endogenous skin pigmentation and should not be confused with positive mRNA signals (n = 3 independent skin sections obtained from two donors). b, Representative phase-contrast micrographs of immortalized N/TERT human keratinocytes, with or without NLRP10 overexpression, with or without m-3M3FBS (50 μM) treatment. Cells with pyroptotic morphology are indicated by arrows (n = 3). c, Representative immunoblot analysis of NLRP10, ASC oligomers, GSDMD cleavage and pro-IL-1β expression in cell lysates, and the release of mature IL-1β to supernatants of immortalized N/TERT human keratinocytes, with or without NLRP10 overexpression, stimulated with the AIM2 agonist poly(dA:dT) dsDNA (2 μg ml −1 ), the NLRP10 activator m-3M3FBS (80 μM) or the NLRP1 activator Val-boroPro (VbP) (3 μM) (n = 3). d, Representative confocal micrographs of NLRP10, ASC and mitochondria localization in immortalized NLRP1-and ASC-deficient N/TERT human keratinocytes, expressing NLRP10 GFP and ASC mScarlet , and stained MitoTracker Deep Red, with or without m-3M3FBS (50 μM) treatment (n = 3). e, Representative immunoblot analysis of NLRP10 and pro-IL-1β expression in cell lysates, and the release of mature IL-1β to the supernatants of undifferentiated or differentiated immortalized N/TERT human keratinocytes, stimulated with poly(dA:dT) dsDNA (2 μg ml −1 ), m-3M3FBS (80 μM) or Val-boroPro (VbP) (3 μM) (n = 3). f, IL-1β secretion from undifferentiated or differentiated immortalized N/TERT human keratinocytes, stimulated with m-3M3FBS (80 μM) or Val-boroPro (VbP) (3 μM) (n = 2). g, IL-1β secretion from differentiated immortalized WT, NLRP10-or Casp1-deficient N/TERT human keratinocytes, stimulated as in e (n = 2). P values were calculated using two-way ANOVA with Šídák's (f) or Tukey's (g) multiple comparison test. *P = 0.0394, **P = 0.002-0.0029, ***P = 0.0007, ****P < 0.0001. h, Representative confocal micrographs of immunostained normal primary human keratinocytes, stimulated with m-3M3FBS (85 μM) (n = 3). Error bars represent s.d.

Online content
Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41590-023-01451-y.   Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Retroviral transduction and puromycin selection
To generate cell lines stably transduced with constructs of interest, retroviruses were used as a nucleic acid vector. HEK293T cells were used for the production of retroviruses. Briefly, ~5 × 10 5 HEK293T cells per well were seeded in 6-well plates in complete DMEM (~3 ml). After overnight incubation (37 °C, 5% CO 2 ), the cells were transfected with combinations of three vectors: the vector of interest (carrying the insert to be stably integrated into the genome of the target cell line; 2 μg per well), the packaging vector (pCMV-gag-pol; 1 μg per well) and the entry vector (pCMV-vsv-g; 100 ng per well). To prepare transfection mixes (per one transfection well) GeneJuice transfection reagent (8 μl; or approx. 2.6 μl of GeneJuice per 1 μg of DNA) was combined with 100 μl of FBS-and antibiotic-free DMEM, and, in a separate tube, the three plasmids were mixed and filled up to 20 μl with FBS-and antibiotic-free DMEM. After 5 min, GeneJuice DMEM and plasmids DMEM were mixed and left for 20 min to allow for the formation of transfection complexes. After this time, transfection complexes were transferred to the HEK cells, followed by an 18-20-h incubation (37 °C, 5% CO 2 ). After this time, the tissue culture media were removed, discarded and replaced with ~2 ml of DMEM supplemented with 30% FBS. After ~24 h of incubation under the high-FBS conditions (37 °C, 5% CO 2 ), the retrovirus-containing tissue culture supernatants were collected. Briefly, the supernatants were collected in a Luer-Lok syringe using a blunt 18G needle. Then, the supernatants were filtered through a 0.45-μm membrane filter. Such supernatants were either directly used for stable transgene delivery into target cells or cryopreserved at −80 °C. For retrovirus-mediated transgene delivery (retroviral transduction), target cells were plated at ~5 × 10 4 cells per well (one day before transduction) or ~10 5 cells per well (on the day of transduction) in 24-well plates. Next, the cells were subjected to several doses of retrovirus-containing supernatants (typically 5-500 μl). After 24-48 h of retroviral transduction (37 °C, 5% CO 2 ), the transduction efficiency was assessed using an epifluorescence microscope and, if necessary, the wells containing populations of positive cells were subjected to antibiotic selection with puromycin (~10 μg ml −1 ; in complete DMEM, at 37 °C and in the presence of 5% CO 2 ). Puromycin was from ThermoFisher Scientific (catalog no. A11138). After successful transduction, the resulting cell lines were expanded and used for experiments at passages 4-20 following retroviral transduction and/ or cryopreserved at −150 °C.

Cell transfection for protein overexpression
For transient protein overexpression, GeneJuice transfection reagent was used for DNA delivery into HEK cells. Typically, the transfection reagent was combined with plasmid DNA at the ratio of 2.3-2.8 μl of GeneJuice per 1 μg of plasmid DNA. GeneJuice was first mixed with serum-and antibiotic-free OptiMEM. In a separate tube, the plasmid of interest was mixed with the same volume of serum-and antibiotic-free OptiMEM. After a 5-10-min incubation at room temperature (RT), the contents of both tubes were combined, and the resulting mixture was incubated at RT for 15-20 min to allow for the formation of transfection complexes. Thus formed, transfection complexes were transferred to HEK cells plated 6-24 h in advance in complete DMEM medium in either uncoated 6-well tissue culture-treated plates or in poly-l-lysine-coated 96-well tissue culture-treated microscopy-adapted plates (Perkin Elmer, catalog no. 6055300). The typical plating densities were between 5 × 10 3 and 2.2 × 10 4 cells per well (or 1.6 × 10 4 -6.9 × 10 4 cells cm −2 ) in 96-well plates, or ~5 × 10 5 cells per well (or ~5.3 × 10 4 cells cm −2 ) in 6-well plates. The final volume of the transfection mix was generally kept under 100 μl per well for cells transfected in 96-well plates or under 200 μl per well for cells transfected in 6-well plates. The addition of DNA transfection complexes was followed by a centrifugation step (340g, 5 min at RT).
For the transfections performed in 96-well plates, the cells were subjected to wide-field or confocal fluorescence microscopy after 24-48 h (depending on the experiment), either after direct fixation (4% formaldehyde with the nuclear counterstain (5 μM DRAQ5)) or following treatment with inflammasome activators and subsequent fixation. For the transfections performed in 6-well plates, the cells were incubated with the transfection complexes for 24 h (37 °C, 5% CO 2 ). Next, the cells were detached from the growth surface using TrypLE Express Enzyme and replated in poly-l-lysine-coated tissue culture-treated microscopy-adapted 96-well plates. After a 24-h 'resting' period (37 °C, 5% CO 2 ), these cells were treated with inflammasome activators, fixed (4% formaldehyde with the nuclear counterstain (5 μM DRAQ5)) and inspected by fluorescence microscopy.

Molecular cloning
Cloning of all constructs was performed following standard procedures and according to manufacturers' instructions for all the reagents. The inserts were amplified by PCR using the PfuUltra II Hotstart PCR Master Mix. The PCR reactions were performed in 50-μl final volumes, using 0.5 μM forward and reverse primers and ~100-500 ng of DNA per reaction. For site-directed mutagenesis, the overlap-extension PCR protocol was used 54 . Generally, PCR was initiated with a 1-min denaturation step (94 °C), followed by 35 or 40 cycles of: (1) denaturation (94 °C, 30 s), (2) primer annealing (58-65 °C, 30 s) and (3) elongation (72 °C, 1 min per kilobase (kb) of DNA). At the end of the reaction, a 5-or 10-min final elongation step at 72 °C was included. The PCR products were electrophoretically separated in an agarose (Biozym) gel (1-1.5%) with green DNA-RNA dye (PEQ-green, Peqlab/WVR; ~1.5 × 10 4 dilution) and isolated using PureLink Quick Gel Extraction kit. Next, sticky ends were generated by digestion with a pair of restriction enzymes (ThermoFisher Scientific) according to the manufacturer's instructions in final volumes of 50 μl. Inserts generated in this fashion were re-run in an agarose gel (1-1.5%) and purified. The backbone vectors were linearized by restriction digest with the same pair of restriction enzymes as the insert and then isolated from a 1% agarose gel after electrophoretic digest.
Ligation (20 μl final volume) was performed according to the manufacturer's instructions. Briefly, the insert of interest was mixed with the backbone at the ratio of 3:1-9:1 and kept on ice (~4 °C) with T4 DNA ligase (ThermoFisher Scientific) in the presence of 1× ligation buffer. After preparation of this reaction mix, the tubes were shifted to 22 °C for 20 min. Then, the reaction mixes were returned to the ice bath. Vectors generated in this manner (1-5 μl) were used to transform chemically competent DH5a Escherichia coli cells (15 μl). The heat-shock transformation protocol was employed; briefly, the cells were incubated with the ligation products on ice (~4 °C) for 5 min, followed by a 45-s incubation at 42 °C and a 2-min 'regeneration' step on ice (~4 °C). Then, the transformed cells were diluted in LB broth without Nature Immunology Letter https://doi.org/10.1038/s41590-023-01451-y antibiotics (~150 μl) and incubated for 30 min at 37 °C with shaking (400-600 r.p.m.). After this, the cells were plated in LB-agar (Ther-moFisher Scientific) plates with ampicillin (100 μg ml −1 , Sigma/Merck) used as a selection antibiotic. Following overnight incubation at 37 °C, colonies were picked from agar plates. They were then used for inoculation of small bacterial cultures (6 ml, LB broth with 100 μg ml −1 ampicillin), which were grown for 16-18 h at 37 °C with shaking (~360 r.p.m.). After this time, bacterial cells from these cultures were collected, and plasmid DNA was isolated using the PureLink Quick Plasmid Miniprep kit, following the manufacturer's instructions. Complete sequences of the inserts from the purified plasmids were obtained by Sanger sequencing performed by GATC or Eurofins Genomics. Of the positive clones, one was selected to prepare a cryopreserved stock (bacterial suspension in LB broth with 100 μg ml −1 ampicillin supplemented with 25% glycerol (Sigma/Merck); stored at −80 °C). The glycerol stocks were used for inoculation of larger scale bacterial cultures (120-150 ml grown in LB broth with 100 μg ml −1 ampicillin for 16-18 h at 37 °C with shaking (~340 r.p.m.)), from which plasmid DNA was isolated using the PureLink Quick Plasmid Maxiprep kit. For all subsequent applications, plasmid DNA from those validated Maxiprep preparations was used.

Primary macrophage culture
Femurs and tibias were typically obtained from 6-24-week-old female mice (see Supplementary Table 7 for details regarding sex, strain and genetic background of the animals). After isolation, the bones were briefly washed in DPBS in a tissue culture dish, followed by a 30-s incubation in 70% ethanol aq . Then, the bones were transferred to DPBS in a fresh tissue culture dish and kept there until the bone marrow extraction step (< 30 min). The bones were opened using scissors and flushed (using a syringe) with FBS-and antibiotic-free DMEM (bones from one mouse were flushed ~10 ml of medium) in a fresh tissue culture dish. The bone marrow suspensions in DMEM were centrifuged (340g, 5 min) and either cryopreserved at −150 °C or resuspended in complete DMEM supplemented with 15-30% L929 cell-conditioned medium.
Bone marrow-derived macrophage (BMDM) differentiation was conducted for 7 d in a tissue culture incubator (37 °C, 5% CO 2 ). On the last day of differentiation, the L929 cell-conditioned media-containing differentiation media were discarded, and the cells were washed with DPBS and detached from the growth surfaces using a cell scraper. The BMDM suspensions in DPBS were spun down (340g, 8 min), the DPBS supernatants were discarded and BMDMs were resuspended in complete DMEM supplemented with 1-5% L929 cell-conditioned medium. For further experimentation, BMDMs were plated in 96-well plates at the density of 5 × 10 4 cells per well (or ~1.56 × 10 5 cells cm −2 ). All steps of this procedure were performed at RT. Experiments using BMDMs were performed after a 16-18-h resting period in a tissue culture incubator (37 °C, 5% CO 2 ), following cell plating.

Differentiation of N/TERT keratinocyte cultures
N/TERT keratinocytes were cultured to 80-90% confluence (referred to as undifferentiated) before the media was switched to high-CaCl 2 media, which were replenished every 2 d. The cultures were collected at day 6 after the high-CaCl 2 switch (referred to as differentiated).

Generation of NLRP10 reporter N/TERT cell line
The NLRP10 reporter line described in the manuscript was created using N/TERT cells that had NLRP1 and ASC knocked out before constitutive expression of GFP-tagged NLRP10 and mScarlet-tagged ASC. Briefly, CRISPR-Cas9 NLRP1 knockout N/TERTs were generated as above using the lentiCRISPR-v2. ASC was knocked out using the Integrated DNA Technologies Alt-R CRISPR-Cas9 guide system according to the manufacturer's instructions, and then delivery of CRISPR-Cas9 ribonucleoprotein (RNP) complex into NLRP1 knockout (KO) N/TERT using electroporation with the Nucleofector system from Lonza. Constitutive lentiviral expression of GFP-tagged NLRP10 and mScarlet-tagged ASC was performed using pCDH vector constructs (System Biosciences) and packaged using third-generation packaging plasmids.

NLRP10-overexpressing and NLRP10-deficient N/TERT cells with caspase-1 reporter
The NLRP10-overexpressing reporter line described in the manuscript (Extended Data Fig. 10d,e) was generated by lentiviral transduction in N/TERT-1 cells constitutively expressing a caspase-1 reporter (Casp1-CARD eGFP ) 56 . Lentivirus was produced with packaging vectors psPax2 and pMD2.G (kind gifts from D. Trono, École Polytechnique Fédérale de Lausanne, Switzerland) and the cell line was generated by transduction with virus multiplicities of infection that permit multiple insertions, followed by antibiotic selection.

Primary keratinocyte culture
Normal human epidermal keratinocytes from pooled adult donors (PromoCell, catalog no. C-12006) were used. These cells were grown in PromoCell Keratinocyte Growth Medium 2 (catalog no. C-20111) and passaged until the passage number of 10 using PromoCell Detach Kit (catalog no. C-41220).
For the differentiation assay, cells were plated at the concentration of 2.5 × 10 5 cells per well in tissue culture-treated 24-well μ-plates (ibidi, catalog no. 82426) and cultured for 6 d in the PromoCell Keratinocyte Growth Medium 2 supplemented with 1 mM CaCl and 1% FBS to induce differentiation ( Fig. 4h and Extended Data Fig. 10f,g). Alternatively, cells were plated at 1 × 10 4 cells per well in 96-well plates with or without CaCl 2 supplementation (0.1, 0.25 or 1 mM) and grown for 6 d before isolation of RNA for the assessment of NLRP10 expression (Extended Data Fig. 10a). Total cell DNA from mtDNA-depleted and control cells was isolated using QIAmp DNA micro kit, following the manufacturer's instructions. Per condition, material from nine wells of a 96-well plate was collected. The total cellular mtDNA content was analyzed by qPCR using sequences from nuclear DNA as a reference ('housekeeping') whose level is not strongly affected by the addition of the mtDNA-depleting agent 57,58 (see Supplementary Table 8).
For the qPCR reaction mixes (20 μl), 100 ng DNA per sample was analyzed. The primers were used at the final concentrations of 500 nM per primer. The qPCR master mix (Maxima SYBR Green/ROX 2× qPCR Master Mix) was based on SYBR Green DNA detection, and it was supplied as a 2× concentrated working solution. A 10-μl portion of the qPCR master mix was used per reaction, and nuclease-free water was added to a final volume of 20 μl. qPCR was run in technical duplicates according to the standard two-step cycling protocol. Briefly, the DNA samples were denatured for 10 min at 95 °C, followed by 40 cycles of: (1) denaturation (95 °C, 15 s) and (2) annealing and elongation (60 °C, 1 min). The fluorescence signal was acquired during the annealing and elongation step, following the manufacturer's instruction. The fold changes in the mtDNA content values were calculated using the 2 −DDCt method 59 .

qPCR for NLRP10 in primary human keratinocytes
RNA of 2 × 10 3 cells seeded into a well of a 96-well plate was isolated using the RNeasy mini kit (Qiagen, catalog no. 74004) according to the manufacturer's instructions. Cells were lysed with 350 μl of RLT lysis buffer containing 1% β-mercaptoethanol. RNA was eluted in 30 μl of nuclease-free water and the RNA concentrations were assessed by measuring absorbance at 260 nm.
Complementary DNA was synthesized on the template of messenger RNA using oligo-dT primers and SuperScript III Reverse Transcriptase (ThermoFisher Scientific, catalog no. 18080093), following the manufacturer's instructions. Approximately 100-500 ng of RNA was mixed with nuclease-free water to a volume of 12.9 μl, and 1 μl of oligo-dT primer (50 μM) per sample was added. The mix was first incubated at 65 °C for 5 min, and then cooled on ice for 1 min. Next, 4 μl of 5× first-strand buffer, 1 μl of 10 mM dNTPs, 1 μl of 0.1 M dithiothreitol and 0.1 μl of SuperScript III Reverse Transcriptase per sample were added to the mix, which was subsequently incubated for 50 min at 50 °C and then for 5 min at 85 °C. Last, cDNA was diluted 1:5 by addition of nuclease-free water.
The qPCR reaction was performed as described in the previous section, with HPRT primers targeting the housekeeping control reference gene and NLRP10 qPCR primers to assess the relative NLRP10 mRNA levels.

RNAScope in situ hybridization
Waste surgical skin tissues from abdomen and breast were collected with appropriate informed consent of the patients and sent to the Asian Skin Biobank (ASB) at the Skin Research Institute of Singapore (SRIS) (under A*STAR IRB 2020-209). The microtome and blade were sprayed with 100% ethanol before sectioning FFPE-embedded human skin biopsies at 15 μm. Freshly cut sections were allowed to fully adhere onto the slide for 1 h before being dewaxed and then dehydrated through an increasing ethanol gradient: 50%, 70%, 100%. Sections were then dried and demarcated using a hydrophobic barrier pen. Hydrogen peroxide was applied to each section for 10 min at RT before incubation with protease IV for 30 min at RT. Sections were then processed according to the RNAscope 2.5 HD RED chromogenic assay (ACDbio, catalog no. 322350) protocol provided by the manufacturer. Briefly, sections were incubated for 2 h at 40 °C with the appropriate probe: NLRP10, PPIB (positive control) or dapB (negative control). Slides were then washed twice for 2 min at RT with RNAScope Wash Buffer. This was then followed by sequential hybridizations with six amplification probes, alternating between 30 min and 15 min of incubation at 40 °C. After washing in RNAScope Wash Buffer, sections were incubated for 10 min at RT with RNAScope RED Working solution. This was then washed off with tap water and the sections were counterstained with 50% Gill's haematoxylin I staining solution for 3 min at RT. Nuclei were then blued using 0.02% ammonia water before drying at 60 °C for 15 min.

Immunoblotting
Protein concentrations in lysates were determined using the Bradford assay (Thermo Scientific, catalog no. 23200), and 20 μg of protein was loaded per lane of a gel, with the exception of cleaved GSDMD-NT visualization, where 40 μg of protein was used. All primary antibodies were used at 250 ng ml −1 . Visualization of ASC oligomerization was as previously described 55 . For analysis of IL-1β in the media by immunoblotting, samples were concentrated using filtered centrifugation (Merck, Amicon Ultra, catalog no. UFC500396). Protein samples were electrophoretically separated and immunoblotted, and then visualized using a ChemiDoc Imaging system (BioRad).

NLRP10-ASC coimmunoprecipitation
HEK293T cells overexpressing hASC-TagBFP and hNLRP10-mCitrine were plated in poly-l-lysine (0.01%)-coated 6-well tissue culture plates at 5 × 10 5 cells per well (three wells per condition) in DMEM supplemented with 10% FBS and 1:100 penicillin:streptomycin and rested overnight. On the next day, the culture medium was removed and discarded, and it was replaced with OptiMEM (900 μl per well). Then, 10× concentrated working solutions of the activators in OptiMEM were prepared (200 μM thapsigargin for the final concentration of final 20 μM; 850 μM m-3M3FBS for the final concentration of final 85 μM) and immediately used for stimulation of the cells. Untreated control was subjected to OptiMEM alone. The plates were swirled (figure of eight-shaped movements) to ensure equal distribution of the activator in the well and spun down at 340g for 3 min at RT. The plates were incubated for 45 min at 37 °C, 5% CO 2 . After this time the plates were transferred to ice. The stimulation media were discarded and the cells were lysed with 150 μl per well of lysis buffer (Tris-HCl, pH 7.4, 100 mM NaCl, 10% glycerol, 0.5% NP-40, cOmplete protease inhibitor cocktail). The cells were incubate with the lysis buffer for 5 min on ice, scraped and the resulting lysates were spun down at 1,000g for 5 min at 4 °C. The resulting supernatants (also serving as input controls) were transferred to fresh tubes and the pellets were discarded. Following the centrifugation, the lysates were incubated with GFP-trap magnetic particles M-270 (Chromotek), according to the manufacturer's instructions. At the end of this incubation, the beads were washed three times with the lysis buffer, after which the beads were resuspended in lithium dodecylsulfate lysis buffer with reducing agent (ThermoFisher Scientific) and boiled for 10 min at 95 °C. In parallel, the input samples were also mixed with the lysis buffer and the reducing agent and boiled. The resulting samples were separated by SDS-PAGE and blotted for NLRP10-mCitrine (anti-human NLRP10 clone 8H2, Merck Millipore) and ASC-TagBFP (anti-ASC AL177, Adipogen).

Microscopy of N/TERT immortalized keratinocytes
To assess pyroptotic cell morphology (Fig. 4b), N/TERT immortalized keratinocytes were seeded onto 12-well tissue culture plates and allowed to adhere overnight at 37 °C, 5% CO 2 . The next day, cells were pretreated with a caspase-1 inhibitor (VX-765, 10 μM), to prevent pyroptotic cell detachment from the plate. After 1 h of incubation with the caspase-1 inhibitor, the cells were further treated with the indicated stimuli. Three images per treatment were then acquired using the EVOS microscope.

Colocalization studies of NLRP10 with ASC and the mitochondria
Before treatment with m-3M3FBS, cells were stained using MitoTracker Deep Red (ThermoFisher Scientific, catalog no. M2246) to visualize the mitochondria. The 1 mM MitoTracker stock was diluted in cell culture media to 100 μM and incubated for 15 min at 37 °C, 5% CO 2 incubator. Following MitoTracker staining, cells were treated with m-3M3FBS (80 μM) and imaged immediately without fixation using a confocal laser scanning microscope (Olympus, FV3000).
On the day of the experiment, the cells were primed with LPS (200 ng ml −1 , 2 h at 37 °C, 5% CO 2 ) in complete DMEM, or left unprimed. Of note, NLRP3/ASC mCerulean reporter iMac cells do not require the priming step to mount the NLRP3 inflammasome responses due to their constitutive overexpression of NLRP3.
For inflammasome activation, the cells were shifted to OptiMEM (ThermoFisher Scientific, catalog no. 31985070) or to an extracellular buffer consisting of (in mM) 123 NaCl, 5 KCl, 2 MgCl 2 , 1 CaCl 2 , 10 glucose, 10 HEPES (pH 7.4). For inflammasome stimulations under increased KCl concentrations, the buffer composition was modified (the KCl concentration was raised at the cost of an equimolar decrease in the NaCl concentration), and for stimulations under Ca 2+ -free conditions, CaCl 2 was not added. Typically, 90 μl per well of the extracellular buffer was added, followed by the addition of 10 μl of the following inflammasome activators: nigericin (final concentration 10 μM), thapsigargin (final concentration ~20 μM), m-3M3FBS (final concentration ~85 μM) or poly(dA:dT) (model double stranded (ds) DNA transfection complexes, administered at 200 ng per well and complexed with 0.5 μl of Lipofectamine 2000 per well, which corresponds to 2 μg ml −1 poly(dA:dT) dsDNA complexed with 5 μl of Lipofectamine 2000). The addition of inflammasome agonists was followed by gently flicking the plate with a finger, a centrifugation step (340g, 5 min, at RT) and a 30-60-min incubation (37 °C, 5% CO 2 ). After this time, tissue culture supernatants were collected, or the cells were fixed (4% formaldehyde) and counterstained with a nuclear dye (5 μM DRAQ5) 60,61 . For inflammasome activation by SMBA1, due to its low water solubility, the cells were first shifted to 50 μl per well of OptiMEM or of the standard extracellular buffer, followed by the addition of 50 μl per well of SMBA1 (final concentration ~50 μM). Similar to other inflammasome activation protocols, the addition of SMBA1 was followed by gently flicking the plate with a finger, a centrifugation step (340g, 5 min, at RT) and a 30-60-min incubation (37 °C, 5% CO 2 ). After this time, tissue culture supernatants were collected, or the cells were fixed (4% formaldehyde) and counterstained with a nuclear dye (5 μM DRAQ5).

Small-molecule compound pretreatment and inflammasome inhibition
The low molecular weight (LMW) compound pretreatment experiments were performed in 96-well plates according to the following standard setup. For experiments in which the readout was the measurement of IL-1β concentration in tissue culture supernatants, the cells were primed with LPS or left unprimed (as a control). For experiments in which the readout was the imaging of ASC specks, the cells were generally left unprimed unless indicated otherwise in the figure legend. Next, the cells were shifted to an extracellular medium consisting of (in mM) 123 NaCl, 5 KCl, 2 MgCl 2 , 1 CaCl 2 , 10 glucose, 10 HEPES (pH 7.4), with the addition of LMW inhibitors/blockers or vehicle (solvent) controls (90 μl per well when the next step was inflammasome activation with nigericin, m-3M3FBS, thapsigargin or poly(dA:dT), or 50 μl when the next step was inflammasome activation with SMBA1). The tested inhibitor/blocker concentrations are indicated in the figure legends. Inhibitor addition was followed by ~10-min incubation at 37 °C, 5% CO 2 . Then, the cells were stimulated with inflammasome activators (typically 10 or 50 μl per well), followed by gently flicking the plate with a finger, a centrifugation step (340g, 5 min, at RT) and a 30-60-min incubation (37 °C, 5% CO 2 ). After this time, tissue culture supernatants were collected, or the cells were fixed (4% formaldehyde) and counterstained with a nuclear dye (5 μM DRAQ5).

Mouse IL-1β concentration assessment by HTRF
IL-1β concentrations in tissue culture supernatants were assessed by a homogenous time-resolved fluorescence (HTRF) 'sandwich' antibody-based assay, following the manufacturer's instructions. The supernatants were analyzed either directly upon completion of the experiment or stored at 4 °C for up to 24 h.
Briefly, the anti-mouse IL-1β solutions were mixed at a 1:1 ratio. A portion of 4 μl per well of this mixture was distributed in white low-volume medium-binding HTRF-adapted 384-well assay plates (Greiner Bio-One, catalog no. 784075). This was followed by the addition of the samples (tissue culture supernatants; 16 μl per well). The plates were centrifuged at RT, 1,000g for 5 min, followed by a 2-6-h incubation at RT or a 16-18-h incubation at 4 °C. After this time, the HTRF signals were measured using Spectramax i3 with an HTRF cartridge.

Human IL-1β ELISA
Secreted IL-1β cytokine levels were assessed by enzyme linked immunosorbent assay (ELISA) kit (BD), according to manufacturer's instructions.

Immunofluorescence staining and analysis of primary keratinocytes
Keratinocytes were fixed with 4% formaldehyde for 30 min at RT. This was followed by a quenching step (50 mM NH 4 Cl, 50 mM glycine in PBS) for 90 min, and a permeabilization/blocking step (10% goat serum and 0.1% Triton X-100 in PBS). After that, cells were incubated with the NLRP10 primary antibody (1:100) (clone 8H2, MABC293, Sigma) overnight at 4 °C and subsequently stained with the Alexa-488-labeled secondary antibody for 1 h at RT. The nuclei were stained with DAPI (1 μg ml −1 ) for 15 min, and the samples were stored in PBS at 4 °C for no more than 7 d until further processing.
The imaging was performed using a Leica SP8 point-scanning confocal microscope equipped with the ×40 water-immersion objective (numerical aperture (NA) 1.1). The images were acquired using LAS X software (v.3.5.5) as z-stacks at 2,048 × 2,048 resolution and with a line averaging of 4. A total of 6-8 stacks were acquired per condition.
The image processing and analysis were performed using Fiji (https://imagej.net/software/fiji/). The final images were obtained using maximum intensity projection of z-stack, and only the linear adjustment of the brightness and contrast were performed to aid visibility.

Endpoint wide-field fluorescence microscopy
Formaldehyde-fixed cell samples were analyzed by fluorescence microscopy no later than one week after the completion of the experiment. Until that time, the samples were stored at 4 °C. Samples were imaged at the Microscopy Core Facility (Medical Faculty, University of Bonn) using the Observer.Z1 fluorescence microscope (Zeiss) with a dry Nature Immunology Letter https://doi.org/10.1038/s41590-023-01451-y ×20 LD Plan Neo Fluor objective (NA 0.4) or a dry ×20 Plan Apochromat objective (NA 0.8). The microscope was operated using Zen 2.3 Pro software. Image acquisition was performed at RT.
All imaging assays were performed in 96-well plates. For most experiments, six images per well were acquired. The automated acquisition was set up in the microscope software; as all fixed samples were counterstained with the nuclear dye DRAQ5, the DRAQ5 channel was used as the reference channel for software autofocus. Generally, a phase-contrast micrograph was acquired for every condition, with the DAPI filter set (Zeiss filter set no. 49) in the light path. For the fluorescent proteins and dyes, the following filter sets were used: TagBFP, Zeiss filter set no. 49 (DAPI); mCerulean, Zeiss filter set no. 47 HE (CFP); mCitrine, Zeiss filter set no. 46 HE (YFP); mCherry, Zeiss filter set no. 43 HE (DsRed); DRAQ5, Zeiss filter set no. 50 (Cy5).
After the acquisition, where applicable, sample images were exported as TIFF files using Zen Lite software. Only linear adjustments were applied to the images (adjustments to the lookup table's lower and upper boundaries), and these adjustments were uniformly applied to all sample images within one experiment. No nonlinear adjustments were applied. TIFF files exported from Zen Lite software were directly imported into figures prepared using Ai Illustrator software. In Ai Illustrator, the images were cropped and, when necessary, the image dimensions were adjusted to the layout of the figure. For image quantification, raw imaging data were imported into Cell Profiler software.

Time-lapse wide-field fluorescence microscopy
Samples were imaged at the Microscopy Core Facility (Medical Faculty, University of Bonn) using the Observer.Z1 fluorescence microscope (Zeiss) with a dry ×20 LD Plan Neo Fluor objective (NA 0.4) or a dry ×20 Plan Apochromat objective (NA 0.8) and the 37 °C incubation chamber. The microscope was operated using Zen 2.3 Pro software. The cells were stimulated with inflammasome activators as detailed in 'Inflammasome stimulation'. The stimuli were added either directly to cells in the microscope incubation chamber just after the onset of imaging (<1 min) or, when a centrifugation step was necessary (thapsigargin and poly(dA:dT)), they were added outside the incubation chamber just before the onset of imaging (<2 min).
All imaging assays were performed in 96-well plates. For most experiments, 1-2 images per well were acquired. Automated acquisition with time-lapse recording was set up in the microscope software; the phase contrast was typically used as the reference channel for software autofocus, or the definite focus option was used. The time interval between consecutive acquisitions of a given imaging field was 1 min or 4 min, depending on the activator (nigericin, m-3M3FBS, thapsigargin, ionomycin and untreated controls, 1 min; poly(dA:dT), 4 min).
Generally, a phase-contrast micrograph was acquired for every condition, with the DAPI filter set (Zeiss filter set no. 49) in the light path. For the fluorescent proteins, the following filter sets were used: TagBFP, Zeiss filter set no. 49 (DAPI); mCerulean, Zeiss filter set no. 47 HE (CFP); mCitrine, Zeiss filter set no. 46 HE (YFP); mCherry, Zeiss filter set no. 43 HE (DsRed).
After the acquisition, where applicable, sample images were exported as TIFF files using Zen Lite software. Only linear adjustments were applied to the images (adjustments to the lookup table's lower and upper boundaries). No nonlinear adjustments were applied. TIFF files exported from Zen Lite software were directly imported into figures prepared using Ai Illustrator software. In Ai Illustrator, the images were cropped and, when necessary, the image dimensions were adjusted to the layout of the figure.

Endpoint confocal laser scanning fluorescence microscopy
All imaging assays were performed in 96-well plates. Formaldehyde-fixed cell samples were analyzed by confocal laser scanning microscopy no later than one week after the completion of the experiment. Until that time, the samples were stored at 4 °C.
Samples were imaged at the Microscopy Core Facility (Medical Faculty, University of Bonn) using the SP5 AOBS with SMD confocal microscope (Leica) with a water-immersion ×63 HCX PL APO objective (NA 1.2). The images were acquired at a 1,024 × 1,024 resolution with a line averaging of 8. The microscope was operated using LAS AF v.2.7.3 software. Image acquisition was performed at RT.
After the acquisition, where applicable, sample images were exported as TIFF files using LAS X Lite software. Only linear adjustments were applied to the images (adjustments to the lookup table's lower and upper boundaries). No nonlinear adjustments were applied. TIFF files exported from LAS X Lite software were directly imported into figures prepared using Ai Illustrator software. In Ai Illustrator, when necessary, the image dimensions were adjusted to the layout of the figure.

Image quantification and analysis
For image quantification (obtaining nuclear and ASC speck counts from images acquired on the Zeiss Observer.Z1 wide-field fluorescence microscope), raw imaging data were imported into Cell Profiler software [62][63][64] . To each image, an analysis pipeline was applied consisting of the following steps: first, the DNA (DRAQ5) and ASC (Tag-BFP or mCerulean) channels were extracted. Then, the illumination correction was calculated for each channel (using the background method with a block size of 60, with rescaling of the illumination function and using the fit polynomial smoothing method) and the illumination correction was applied. In these corrected images, the primary objects (nuclei and ASC specks) were identified and counted. For identification of the nuclei, a typical diameter of 10-80 pixels was used (this parameter was adjusted between image sets), and objects outside this diameter range were discarded. Objects touching the border of the image were retained. The thresholding strategy was global, and the thresholding method was robust background. The lower and upper outlier fractions were set to 0.02. The averaging method was mode, and the variance method was standard deviation. The number of deviations was set to 0, threshold smoothing scale to 1.3488, threshold correction factor to 1.86 and the lower and upper bounds on threshold were set between ~0.001 and ~1 (this parameter was adjusted between image sets, depending on the brightness of the nuclear signal). The methods to distinguish between clumped objects and to draw dividing lines between clumped objects were both set to intensity. The size of the smoothing filter for declumping and the minimum allowed distance between local maxima were both automatically calculated. Lower resolution images were used to find local maxima. Holes in identified objects were filled after both thresholding and declumping.
For identification of ASC specks, the background illuminationcorrected image was first enhanced using feature type 'speckles', feature size 14 and slow speed and accuracy. ASC specks were identified in these enhanced images by setting the typical diameter of objects to ~2-19 pixels (this parameter was adjusted between image sets), and objects outside this diameter range were discarded. Objects touching the border of the image were retained. The thresholding strategy was global, and the thresholding method was robust background. The lower and upper outlier fractions were set to 0.02. The averaging method was mode, and the variance method was standard deviation. The number of deviations was set to 0, threshold smoothing scale to 1.3488, threshold correction factor to 1.97 and the lower and upper bounds on threshold were set between ~0 and ~0.05 for HEK cells and between ~0.01 and ~1 for macrophages (this parameter was adjusted between image sets, depending on the brightness of the ASC signal). The method to distinguish between clumped objects was shape, and Nature Immunology Letter https://doi.org/10.1038/s41590-023-01451-y the method to draw dividing lines between clumped objects was intensity. The size of the smoothing filter for declumping and the minimum allowed distance between local maxima were both automatically calculated. Lower resolution images were used to find local maxima. Holes in identified objects were filled after both thresholding and declumping.
Raw imaging data from time-lapse recordings of cells overexpressing fluorescent protein-based markers targeted to the mitochondrial matrix were imported into Cell Profiler software. The granularity of the mitochondrial fluorescence signal was assessed in unprocessed images. The module used was 'MeasureGranularity', with the following parameter settings: measurement within objects was disabled, the subsampling factor for granularity measurements was 0.25, the subsampling factor for background reduction was 0.25, the radius of structuring element was 3 and the range of the granular spectrum was 1. To facilitate the comparisons between multiple recordings, the image granularity in the first frame of the recording was set to 100% and the values obtained from subsequently recorded frames were normalized to this initial value.

Data collection and statistical analysis
No statistical methods were used to predetermine sample sizes but our sample sizes are similar to those reported in previous publications (for examples, please see refs. 5,9,12,13,[15][16][17][18][19]28 ). Data collection was not randomized. Data collection and analysis were not performed blind to the conditions of the experiments. Data distribution was assumed to be normal but this was not formally tested. No samples and data points have been excluded from analysis.
Data are presented as mean + s.d.; the number of independent experiments is indicated for each panel in the figure legends. In most cases individual data points represent the mean of technical replicates in each independent experiment; this is specified in the figure legends, as are the numbers of technical replicates. Microsoft Excel 16 was used for data processing; GraphPad Prism 6, 7, 8 or 9 were used for plot preparation, s.d. calculation and statistical significance analysis. The specific tests used and P value ranges are indicated in the figure legends.

Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability
Representative images for all ASC speck formation quantifications are deposited in Mendeley Data using the following link: https://data. mendeley.com/datasets/42fsz64kn5/1. Source data are provided with this paper. All other data are available in the article Supplementary files or from the corresponding author upon reasonable request.
Correspondence and requests for materials should be addressed to Eicke Latz.
Peer review information Nature Immunology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: N. Bernard, in collaboration with the Nature Immunology editorial team.

Reporting Summary
Nature Portfolio wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Portfolio policies, see our Editorial Policies and the Editorial Policy Checklist.
Please do not complete any field with "not applicable" or n/a. Refer to the help text for what text to use if an item is not relevant to your study. For final submission: please carefully check your responses for accuracy; you will not be able to make changes later.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one-or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted

Software and code
Policy information about availability of computer code Data collection

Data analysis
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Portfolio guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A description of any restrictions on data availability -For clinical datasets or third party data, please ensure that the statement adheres to our policy 1 nature portfolio | reporting summary

Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study design
All studies must disclose on these points even when the disclosure is negative.
Sample size

Data exclusions
Replication All attempts at replication were successful. All experiments were performed two or more times to validate reproducibility.

Randomization Blinding
Reporting for specific materials, systems and methods We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response. In the instances where biological material availability was the limiting factor, we performed two biological repeats of the experiments.

Materials & experimental systems Methods
Randomization was not performed. No patient or experimental animal in vivo interventions were performed that would require randomization. In in vitro experiments on pools of cells (such as were performed in our study), randomization was achieved by stochastically dividing cells from the same initial population into distinct experimental conditions (different assay wells).
No blinding was performed. No patient or experimental animal in vivo interventions were performed that would require blinding on the subject side. The researchers were not blinded to the conditions, because data analysis involved only objective assessments (measurements of cytokine concentrations, counting objects, determining protein localization, comparing fluorescence signal intensities); as opposed to subjective assessments such as histopathologic scoring, which were not performed in our study.
No data points were excluded from analysis. No commonly misidentified cell lines were used in our study.

Animals and other organisms
Policy information about studies involving animals; ARRIVE guidelines recommended for reporting animal research Laboratory animals

Wild animals
Field-collected samples

Ethics oversight
No ethical approval was required as animals were only used for organ extraction.
Note that full information on the approval of the study protocol must also be provided in the manuscript.

Human research participants
Policy information about studies involving human research participants Note that full information on the approval of the study protocol must also be provided in the manuscript.
This checklist template is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/