NLRC3 protein inhibits inflammation by disrupting NALP3 inflammasome assembly via competition with the adaptor protein ASC for pro-caspase-1 binding

Inflammasomes are multiprotein complexes that sense pathogen-associated and danger-associated molecular patterns and induce inflammation in cells. The NALP3 inflammasome is tightly regulated by recently discovered control mechanisms, but other modulators still remain to be characterized. NLR family CARD-containing 3 (NLRC3) protein, a caspase recruitment domain (CARD)–containing member of the nucleotide oligomerization domain–like receptor (NLR) family, was found to down-regulate the NF-κB pathway and stimulator of interferon genes (STING)–dependent cytokine secretion. However, the effect of NLRC3 on the NALP3 inflammasome or other inflammasomes is still unknown. We hypothesized that NLRC3 might inhibit NALP3 inflammasome complex assembly. Toward this end, we tested whether NLRC3 overexpression or knockdown influences NALP3 activity in human monocyte and HEK293FT cells when the complex is ectopically reconstituted. We found that NLRC3 indeed decreases NALP3-induced IL-1β maturation and secretion, pro-caspase-1 cleavage, and speck formation by apoptosis-associated speck-like protein containing a CARD (ASC) protein in response to NALP3 activators. We also show that endogenous NLRC3 interacts with both ASC and pro-caspase-1 but not with NALP3, disrupts ASC speck formation through its CARD, and impairs the ASC and pro-caspase-1 interaction. Moreover, the NLRC3 CARD alone could dampen IL-1β secretion and ASC speck formation induced by NALP3 mutants associated with autoinflammatory diseases. In conclusion, we show here that, besides its role in the inhibition of the NF-κB pathway, NLRC3 interferes with the assembly and activity of the NALP3 inflammasome complex by competing with ASC for pro-caspase-1 binding.

Inflammasomes are multiprotein complexes responsible for the sensing of pathogen-associated molecular patterns and danger-associated molecular patterns and the induction of inflammation (1). Different types of inflammasomes have been identified, and most of them have in common the adaptor protein ASC, 2 the effector pro-caspase-1, and the downstream inflammatory mediator cytokine pro-IL-1␤ (2). The assembly of the inflammasome is mediated by the homotypic interaction of ASC and pro-caspase-1 through their CARDs (3) and the formation of ASC foci called "specks" (4). This interaction results in the cleavage of pro-caspase-1 and the maturation and secretion of IL-1␤ (4). The NALP3 (also called NLRP3 or cryopyrin) inflammasome is specifically activated by ionophores such as ATP and nigericin (5), crystals such as monosodium urate (MSU) (6) and many others, whereas the NLRC4 (or IPAF) inflammasome is stimulated by Pseudomonas aeruginosa infection (7). A two-step process is required to promote the secretion of IL-1␤ by the NALP3 inflammasome: the induction of the NFB pathway through Toll-like receptors (called the "priming step") (8) and the complex assembly (called the "activation step") mediated by potassium efflux (9,10), lysosomal rupture (11), and reactive oxygen species production (12).
Mutations in the NALP3 gene are associated with autoinflammatory syndromes known as cryopyrin-associated periodic syndromes (or CAPS) (13), and among these mutations, the NALP3 p.R260W, and p.D303N mutants enhance the ectopic assembly of ASC specks (14,15) and result in an increase in pro-caspase-1 cleavage (14) and IL-1␤ secretion (16,17). IL-1␤ blockers such as anakinra are used to treat CAPS (18,19), and identification of novel inhibitors of the NALP3 inflammasome that act upstream of IL-1␤, on the NALP3 mRNA level (20), on ASC oligomerization (21), or on caspase-1 inhibition (22), is presently ongoing, and these inhibitors may be used in therapy of autoinflammatory diseases in the future.
Although the most studied NOD-like receptor (NLR) family members belong to the pro-inflammatory group, more recently, NLR proteins with anti-inflammatory roles have been characterized. Especially the members of the NLRC subgroup, NLRC5, NLRX1, and NLRC3, were shown to be negative regulators of inflammation. NLRX1 inhibited the NFB pathway by binding to the active IB kinase in response to LPS stimulation (23,24). Furthermore, increased IFN-␤ secretion was observed in immune cells from NLRC5 knock-out mice when stimulated with vesicular stomatitis virus or poly(I:C) (24).
NLRC3, a less studied member of the NLR family, contains a C-terminal leucine-rich repeat (LRR) domain, a central NACHT domain and a N-terminal CARD (25). NLRC3 was shown to inhibit the NFB pathway (25,26) through the ubiquitination of TRAF6 (26). NLRC3 also inhibited stimulator of interferon genes-dependent IFN␤, TNF␣, and IL-6 secretions in response to cytoplasmic DNA stimulation. NLRC3 altered stimulator of interferon genes and TBK1 interaction by directly binding to these proteins, and NLRC3 KO mice responded more efficiently to HSV-1 virus infection because IFN-␤ levels were higher in the sera of these mice compared with WT NLRC3-expressing counterparts (27). We have also shown previously that the overexpressed NLRC3 protein co-localizes with the inflammasome components and decreases IL-1␤ processing and ASC speck assembly in HEK293FT cells (28).
In this study, the suppressive effect of the NLRC3 protein on inflammasomes is confirmed, and the molecular mechanism of this inhibition is elucidated. We determined that, besides its inhibitory role on the NFB pathway described previously, NLRC3 also acts on the activation step of the inflammasome by decreasing the efficiency of the complex assembly by competing with ASC for pro-caspase-1 binding via its N-terminal CARD domain. We propose to use the CARD domain of NLRC3 as a novel inhibitor of the NALP3 inflammasome.

Overexpression of NLRC3 inhibits NALP3-induced IL-1␤ secretion
Previous studies showed that overexpression of the inflammasome components was sufficient to induce activation of the complex and maturation and secretion of caspase-1 and IL-1␤ in a NALP3-and ASC-dependent manner. To assess whether NLRC3 has an effect on the activation of the complex, we ectopically expressed NALP3, ASC, pro-caspase-1, and pro-IL-1␤ in HEK293FT cells in the presence or the absence of NLRC3 and measured the resulting IL-1␤ secretion. Although negligible levels of IL-1␤ were secreted from ASC, pro-caspase-1 and pro-IL-1␤ co-transfected HEK293FT cells, co-expression of NALP3 with these proteins induced a dose-dependent increase in IL-1␤ secretion ( Fig. 1A; 496 Ϯ 42.7 pg/ml for 250 ng, 1400 Ϯ 427.5 pg/ml for 500 ng, and 2903.7 Ϯ 229.81 pg/ml for 1000 ng of NALP3 plasmid), thus confirming the specificity of the experimental setting. Interestingly, when NLRC3 was co-expressed with NALP3 and the other inflammasome-forming proteins, a significant reduction in IL-1␤ secretion was observed ( Fig.  1B; 2903.7 Ϯ 229.81 pg/ml for 1000 ng of NALP3 plasmid alone versus 1403.5 Ϯ 48.1 pg/ml for 1000 ng of NALP3 and 1000 ng of NLRC3 plasmids, p ϭ 0.0001). Moreover, the more NLRC3 was expressed, the higher decrease in IL-1␤ release that was obtained (2925.5 Ϯ 151.2 pg/ml for 250 ng, 1949.3 Ϯ 171.9 pg/ml for 500 ng, and 1403.5 Ϯ 48.1 pg/ml for 1000 ng of NLRC3 plasmid co-transfected with 1000 ng of NALP3 plasmid). NLRC3 not only dampened NALP3-in-duced IL-1␤ secretion (Fig. 1, A and B) but also inhibited pro-IL-1␤ maturation (Fig. 1C) when ectopically expressed in HEK293FT cells.
To further confirm our results, we generated THP-1 stable lines that overexpressed the NLRC3 protein (named OvNLRC3) and confirmed its expression both at the mRNA (Fig. 1D) and the protein levels (Fig. 1E). Induction of the NALP3 inflammasome by stimulating stable THP-1 cells with nigericin ( Fig. 1F) or MSU (Fig. 1G) triggered a significantly lower IL-1␤ release from THP-1 OvNLRC3 cells compared with their OvControl counterpart (3084.8 Ϯ 64.1 pg/ml for OvNLRC3 versus 6385.7 Ϯ 287.2 pg/ml for OvCont when treated with nigericin, p ϭ 0.0039), suggesting that NLRC3 acts as an inhibitor of NALP3-induced IL-1␤ secretion in both cell types.
Because the shD9 cell line gave the more significant decrease in NLRC3 level, we used it in the rest of our experiments and named it shNLRC3. Similarly, although basal IL-1␤ secretion was not significantly different between shNLRC3 and shcont cells, stimulation of the NALP3 inflammasome with ATP (  2F) were observed in the shNLRC3 line compared with shcont in response to nigericin treatment. Also, the shNLRC3 line induced higher cell death compared with its control shcont counterpart when stimulated with nigericin ( Fig. 2G; 42.2 Ϯ 3.3% versus 18.9 Ϯ 2.3% of cell death, p ϭ 10 Ϫ5 , and supplemental Fig. 2). Thus, we concluded that the endogenous NLRC3 protein suppresses the activation of the NALP3 inflammasome induced by different types of ligand and the resulting pyroptotic cell death.

NLRC3 is also an inhibitor of the NLRC4 (IPAF) inflammasome
To determine whether the inhibitory effect of NLRC3 is specific to the NALP3 complex or whether it could also modulate the activity of other inflammasomes, the NLRC4 (or IPAF) inflammasome was stimulated by live P. aeruginosa infection. Knockdown of the NLRC3 protein in THP-1 cells significantly increased IL-1␤ secretion in response to P. aeruginosa infection compared with NLRC3-expressing cells (Fig. 2H, 9493.9 Ϯ 1256.8 pg/ml for shNLRC3 versus 3099.96 Ϯ 1256.7 pg/ml for shcont, p ϭ 0.04). Because NLRC3 inhibits both the NALP3 and NLRC4 inflammasomes, we concluded that it may be acting downstream of the receptor proteins.

NLRC3 inhibits NALP3 inflammasome-induced inflammation NLRC3 does not regulate the protein levels of the inflammasome components
NLRC3 was shown to inhibit the NFB pathway by inducing the ubiquitination of TRAF6 and targeting it for proteasomal degradation (26). We investigated whether NLRC3 exert its inhibitory effect through induction of proteasomal degradation of the inflammasome components by co-transfecting constant NALP3, ASC, and pro-IL-1␤ concentrations and increasing the amount of NLRC3. No difference was observed in transfected NALP3, ASC, or pro-IL-1␤ protein levels at different NLRC3 concentrations (Fig. 3, A-C), suggesting that NLRC3 inhibits the inflammasome by a mechanism independent of protein degradation or synthesis.

NLRC3 interacts endogenously with pro-caspase-1 via its CARD and with ASC but not with NALP3
We have shown previously that NLRC3 co-localizes and interacts with ASC and caspase-1 when overexpressed in HEK293FT cells (28). Here we investigated whether NLRC3 could interact with NALP3. When the two proteins were overexpressed in HEK293FT cells, immunoprecipitation of either NALP3 (Fig. 3D) or NLRC3 (Fig. 3E) did not pull down NLRC3 or NALP3, respectively. Furthermore, the endogenous NLRC3 and NALP3 proteins did not interact with each other in any of the untreated or nigericin-and MSU treated THP-1 cells (Fig. 3F).
As a next step, to verify that the interactions that were seen previously between NLRC3 and ASC or caspase-1 are also occurring physiologically, the endogenous proteins were immunoprecipitated from THP-1 cells. Endogenous NLRC3 interacted with both endogenous ASC (Fig. 3, F and G) and pro-caspase-1 (Fig. 3H). In this study, we used a commercial anti-caspase-1 antibody that is directed against the p10 subunit of caspase-1. Although immunoprecipitation with the anti-caspase-1 p10 antibody pulled down both pro-caspase-1 and the mature, CARD-lacking caspase-1 p20 peptide, the anti-NLRC3 antibody immunoprecipitated only the CARD-containing pro form of caspase-1 (Fig. 3H). Thus, we concluded that NLRC3 interacts with the CARD of caspase-1.

NLRC3 prevents the assembly of ASC specks through its CARD and LRR domain
Another informative way of measuring inflammasome activity is to count the number of ASC specks that are formed. To characterize which domain of NLRC3 disrupts ASC speck formation, we cloned its NLRC3 CARD (aa 1-60),
Although transfection of NALP3 into HEK293FT-ASC-EGFP cells triggered ASC speck assembly (Figs. 4 and 6), full-length NLRC3 (NLRC3 FL ) itself or its domains had no effect on ASC speck formation on their own (data not shown). However, when NALP3 was co-expressed with NLRC3 FL , NLRC3 CARD , NLRC3 LRR , NLRC3 CARD/NACHT , and NLRC3 NACHT/LRR , the number of ASC specks formed was significantly lower than the number of specks formed in the presence of NALP3 alone (Fig. 4, C and D). Thus, we concluded that the CARD and LRR domain of NLRC3 were responsible for the disruption of the NALP3-induced complex assembly and that the central NACHT domain had no effect.

NLRC3 inhibits NALP3 inflammasome-induced inflammation NLRC3 co-localizes with diffused ASC proteins and competes with ASC for pro-caspase-1 binding
To confirm that NLRC3 co-localizes with ASC endogenously, we stained PMA-differentiated and nigericin-treated THP-1 macrophages with anti-NLRC3 and anti-ASC antibodies. We found that NLRC3 and diffuse ASC stainings overlapped each other in the cytosol; however, an NLRC3 signal was not observed on ASC specks (Fig. 5A). Furthermore, we could not detect the NLRC3 protein by Western blotting of the lysate containing ASC specks isolated from ASC-and NLRC3-cotransfected cells (Fig. 5B). Taken together, these results demonstrate that NLRC3 interacts with diffused ASC proteins but is not embedded into ASC specks after inflammasome activation.
Because NLRC3 interacts with both ASC and pro-caspase-1, we determined whether NLRC3 might affect the interaction between these two proteins. Although ASC and pro-caspase-1 interacted with each other when they were overexpressed in HEK293FT cells, addition of NLRC3 resulted in a weaker interaction between ASC and pro-caspase-1 in both samples immunoprecipitated with anti-ASC (Fig. 5D) or anticaspase-1 (Fig. 5E) antibodies. Moreover, increasing the concentration of NLRC3 by keeping a constant amount of ASC and pro-caspase-1 proteins (Fig. 5C) decreased the interaction between ASC and pro-caspase-1 even more. Thus, NLRC3 may exert its inhibitory effect before complex formation by sequestering the pro-caspase-1 proteins and by

NLRC3 inhibits NALP3 inflammasome-induced inflammation
blocking the assembly of pro-caspase-1 and ASC required for inflammasome activity.

The CARD of NLRC3 is necessary and sufficient to inhibit inflammasome activation triggered by NALP3 mutants associated with autoinflammatory syndromes
Because mutations in the NALP3 gene are known to induce the constitutive activation of the NALP3 inflammasome, we tested whether full-length NLRC3 and NLRC3 CARD are able to suppress this aberrant inflammatory response. We found that both full-length-NLRC3 (269. 2 D303N NALP3, p ϭ 10 Ϫ8 ; all compared with the "no NLRC3" conditions) inhibit IL-1␤ secretion induced by the ectopically assembled inflammasome complex in HEK293FT cells, and NLRC3 CARD had an even more significant effect than the full-length protein (Fig. 6A, p ϭ 0.0002 for WT NALP3, p ϭ 10 Ϫ7 for NALP3 p.R260W, and p ϭ 0.0001 for NALP3 p.D303N). Furthermore, NLRC3 FL and NLRC3 CARD also impaired the ASC speck formation triggered by WT and NALP3 p.R260W proteins (Fig. 6, B and C). On the other hand, no effect was observed on the number of ASC specks formed by NALP3 p.D303N proteins, probably because the concentration of transfected NLRC3 was not high enough to block the excessive effect of the NALP3 p.D303N mutant. These data demonstrate that the CARD of NLRC3 is able and sufficient to repress the NALP3 inflammasome activity under normal or pathologic conditions.

Discussion
We investigated, in this paper, the potential effect of the NLRC3 protein on the activation of the NALP3 inflammasome by using human THP-1 monocytes and HEK293FT cells when we had to ectopically express an exogenous protein. We found that NLRC3 decreases the number of ASC specks formed, the maturation of pro-IL-1␤ and pro-caspase-1 proteins, and the secretion of IL-1␤ in both overexpression and endogenous experimental settings. The suppressive effect of NLRC3 on the NALP3 inflammasome in THP-1 cells was stimulant-independent because complex activation in response to all ligands used (ATP, nigericin, and MSU) was affected by NLRC3.
Moreover, the CARD of NLRC3 was necessary to exert this inhibition. NLRC3 had no effect on protein levels, and its suppressive effect was independent of its role on the NFB pathway because, with the use of HEK293FT cells, overexpression of these proteins bypassed transcriptional regulation of the inflammasome components. Another advantage of HEK293FT cells that do not express the ASC adaptor protein at all and negligible levels of the other inflammasome components was the possibility to introduce mutant NALP3 proteins or different NLRC3 domains without possible interference of the WT NALP3 or full-length NLRC3 proteins, respectively.
Recruitment of pro-caspase-1 proteins by the ASC adaptor to foci is a very critical step in the activation of inflam-masomes. The CARD-CARD interaction between ASC and pro-caspase-1 was required for complex formation because mutagenesis of some residues in these domains abolished ASC/ pro-caspase-1 binding and inflammasome activation (29). Similarly, the self-oligomerization of the CARD of caspase-1 was also shown to prevent its interaction with ASC (30). Here we showed that the presence of NLRC3, which also contains a CARD, decreases the interaction between ASC and procaspase-1. The absence of an NLRC3 and NALP3 interaction and the ability of NLRC3 to inhibit both NALP3 and NLRC4 inflammasomes and different NALP3 mutant-induced IL-1␤ secretion suggest that NLRC3 acts on the ASC/pro-caspase-1 level. A similar phenomenon is observed in NFB pathway activation, where the ASC protein interferes with RIP2 and Figure 6. The CARD of NLRC3 is necessary and sufficient to inhibit inflammasome activation triggered by the NALP3 mutants associated with autoinflammatory syndromes. A, IL-1␤ secretion induced by WT NALP3 or NALP3 p.R260W and p.D303N mutants in the absence (black) or presence of full-length NLRC3 (gray) and NLRC3 CARD (light gray). B and C, ASC speck formation induced by WT NALP3 or NALP3 p.R260W and NALP3 p.D303N mutants in the presence or absence of NLRC3 and NLRC3 CARD . Shown are quantification of ASC specks (B) and representative images of three sets of independent experiments (C). Samples were compared with Student's t test.
caspase-1 assembly through its CARD (31). More recently, an antibody generated against the CARD of ASC was sufficient to inhibit inflammasome activation (32), again suggesting how important the interaction of ASC and caspase-1 through their CARD domain is.
Overall, we propose in this paper that, besides its previously defined suppressive role in the NFB pathway, NLRC3 is also a novel inhibitor of activation of the NALP3 and NLRC4 inflammasomes through disruption of ASC speck assembly by competing with the ASC protein for binding to the CARD of pro-caspase-1.

Ectopic reconstruction of the NALP3 inflammasome in HEK293FT cells for the measurement of the IL-1␤ secretion
750,000 HEK293FT cells were transfected using the calcium phosphate protocol with 250 ng of pcDNA3-ASC, 250 ng of pcDNA3-FLAG-pro-caspase-1, and 500 ng of pro-IL-1␤-IRES-EGFP plasmids and the indicated amount of pcDNA3-FLAG-WT NALP3 or pcDNA3-MYC-NLRC3. Equal amounts of DNA were transfected under each condition by adding the empty pcDNA3 vector to compensate for differences. Cell supernatants were harvested 24 h after transfection and spun down to discard cells, and the IL-1␤ levels in 10% FBS containing DMEM were measured by ELISA as explained later. For analysis of CAPS mutants, besides the amount of ASC, pro-caspase-1, and pro-IL-1␤ indicated above, 1 g of WT, p.R260W, or p.D303N NALP3 was transfected with 1 g of empty vector, NLRC3 FL , or NLRC3 CARD .

Activation of the NALP3 and IPAF/NLRC4 inflammasomes in THP-1 cells
10 6 shcont or shNLRC3 or OvNLRC3 and OvControl THP-1 stable lines were seeded into a 6-well plate and differentiated into macrophages with 3 h of 0.5 M PMA treatment. The next day, cells were primed for 2 or 4 h with 100 ng/ml LPS and treated with 20 M nigericin, 5 mM ATP, or 150 g/ml MSU for 4 or 6 h, respectively, for NALP3 inflammasome activation and treated with 75 multiplicity of infection live P. aeruginosa for 4 h for IPAF/NLRC4 inflammasome activation. Supernatants were harvested and cleared from cells, and IL-1␤ levels were measured by ELISA.

Measurement of secreted IL-1␤ levels by ELISA
IL-1␤ levels were determined in 1 ml of cell supernatant by using the R&D Systems human IL-1␤ IL-1F2 kit according to the protocol of the manufacturer. Each treatment was performed in duplicate or triplicate, and the supernatants were measured in duplicate. Significance between samples was determined by two-tailed Student's t test.

ASC speck formation assay
The HEK293FT-ASC-EFGP stable cell lines generated in our laboratory were transfected with NALP3, full-length NLRC3, or the different domains of NLRC3 in a 6-well plate with 500 ng of NALP3, 500 ng of NLRC3, or 500 ng of the different domains of NLRC3 (Fig. 4) or 1 g of WT NALP3, NALP3 pR260W, or NALP3 p.D303N and 1 g of NLRC3 (Fig. 6). An equal amount NLRC3 inhibits NALP3 inflammasome-induced inflammation of DNA was transfected by using the pcDNA3-Myc empty vector. The number of ASC specks was counted 24 or 48 h after transfection under a fluorescence microscope. Each condition was transfected into two wells, and eight randomly chosen fields were counted for each well.

Isolation of the ASC specks
ASC specks were isolated from four 100-mm plates of 5 ϫ 10 6 HEK293FT cells each and transfected with 4 g of mCherry-ASC plasmid. 24 h after transfection, samples were sonicated, vortexed until filaments were formed, and spun down at 200 ϫ g for 1 h. The pellet was dissolved in 30 ml of PBS, passed through a 5-m filter, and spun down at 2400 ϫ g for 1 h. The final pellet containing ASC specks was dissolved in 500 l of PBS. The purity of isolated ASC specks was verified under the microscope and by Western blotting.

LDH assay
An LDH cytotoxicity detection kit from Biovision was used according to the protocol of the manufacturer. The percentage of cell death was determined with the formula (OD sample Ϫ OD LDH Ϫ ) ϫ 100 / (OD LDH ϩ Ϫ OD LDH Ϫ ). Untreated controls were used as LDH Ϫ samples and Triton X-treated control cells as LDH ϩ samples.
Author contributions-E. E. designed and performed the experiments and wrote the manuscript. M. B. contributed to the realization of some experiments. N. Ö. designed the experiments, wrote the manuscript, and provided financial resources.