Inhibition of NLRP3 inflammasome activation by A20 through modulation of NEK7

Significance

Previous study suggested that A20 plays a crucial role in regulating NLRP3 inflammasome activation, yet the specific mechanism remains to be investigated. Here, we revealed that A20 notably diminishes the expression of NEK7, leveraging its role in moderating the NF-κB signaling pathway. A20 directly binds with NEK7, facilitating its proteasomal degradation by enhancing ubiquitination. Crucially, K189 and K293 residues on NEK7 are pivotal for supporting A20-mediated ubiquitination. A20 affects the binding of NEK7 to NLRP3 complex potentially through its OTU domain and/or synergistic effect of its ZnF4 and ZnF7 motifs. Utilizing A20-derived peptides and genetic approaches, we showed that interfering NEK7 function in macrophages can significantly dampen pyroptosis and alleviate the process of sepsis.

Abstract

The NLRP3 inflammasome, a pivotal component of innate immunity, has been implicated in various inflammatory disorders. The ubiquitin-editing enzyme A20 is well known to regulate inflammation and maintain homeostasis. However, the precise molecular mechanisms by which A20 modulates the NLRP3 inflammasome remain poorly understood. Here, our study revealed that macrophages deficient in A20 exhibit increased protein abundance and elevated mRNA level of NIMA-related kinase 7 (NEK7). Importantly, A20 directly binds with NEK7, mediating its K48-linked ubiquitination, thereby targeting NEK7 for proteasomal degradation. Our results demonstrate that A20 enhances the ubiquitination of NEK7 at K189 and K293 ubiquitinated sites, with K189 playing a crucial role in the binding of NEK7 to A20, albeit not significantly influencing the interaction between NEK7 and NLRP3. Furthermore, A20 disrupts the association of NEK7 with the NLRP3 complex, potentially through the OTU domain and/or synergistic effect of ZnF4 and ZnF7 motifs. Significantly, NEK7 deletion markedly attenuates the activation of the NLRP3 inflammasome in A20-deficient conditions, both in vitro and in vivo. This study uncovers a mechanism by which A20 inhibits the NLRP3 inflammasome.

As the most extensively characterized inflammasome, NLRP3 inflammasome can be activated by various environmental irritants, multiple pathogen-associated molecular patterns (PAMPs), including bacterial, viral, and fungal pathogens, or danger-associated molecular patterns (DAMPs) such as ATP and oxidized DNA (ox-DNA) (1). NLRP3 is increasingly recognized as a crucial therapeutic target for several inflammatory diseases, such as septic shock, peritonitis, and inflammatory bowel disease (IBD) (2, 3). Structurally, the NLRP3 inflammasome is a tripartite platform that contains the sensor NLRP3, an adaptor apoptosis-associated speck-like protein containing a CARD (ASC) and effector caspase-1 (1). Its activation is broadly delineated into two phases: priming and activation. The priming phase is marked by the upregulation of NLRP3 and pro-IL-1β expression, driven by the activation of nuclear factor (NF)-κB signaling pathways. NLRP3 activation, in turn, is facilitated by potassium efflux and/or mitochondrial changes, which enable subsequent signaling (4). Notably, another research has highlighted that 8-OH-dG, a component of ox-DNA, could activate the NLRP3 inflammasome, positioning ox-DNA as a potential ultimate NLRP3 ligand (5). This activation culminates in the assembly of inflammasomes with ASC specks, facilitating self-cleavage and the maturation of caspase-1. This process subsequently leads to the cleavage of gasdermin D (GSDMD), inducing pyroptosis (6).

The NLRP3 inflammasome is a fine-tuned multimeric complex, and how to regulate NLRP3 inflammasome has emerged as a critical focus in contemporary research. Recent advancements have identified the never in mitosis A (NIMA)-related kinase 7 (NEK7) as a component of the NLRP3 inflammasome (7). The catalytic domain of NEK7 is known to specifically interact with NLRP3, a key step for its activation and essential for the subsequent ASC speck formation and caspase-1 activation (8). This direct interaction between NLRP3 and NEK7 proves indispensable for the assembly and activation of the NLRP3 inflammasome. Consequently, NEK7 is increasingly recognized as a vital target for NLRP3-associated pyroptosis. Moreover, Shi et al. suggested that the cellular quantity of NEK7 plays a nonredundant role in regulating the NLRP3 inflammasome, a function absent during mitosis (9). The phosphorylation of NEK7 is observed to attenuate the NLRP3–NEK7 interaction, thus weakening NLRP3 inflammasome activation (10). Conversely, deglutathionylation of NEK7 may amplify its impact on NLRP3 (11). These findings collectively indicate that both the quantity and posttranscriptional modifications of NEK7 significantly influence its regulatory effect on NLRP3 activation.

A20 (TNFAIP3) is widely recognized as a pivotal regulator of inflammation and immunity, which is attributed to its ability to inhibit the NF-κB pathway and confer cellular protection from death (12). Polymorphisms in the TNFAIP3 gene have been linked to multiple inflammatory pathologies, including rheumatoid arthritis, Crohn’s disease, and systemic lupus erythematosus (13). In line with this, A20-deficient mice exhibit a severe and lethal inflammatory phenotype. Similarly, specific cell-type deficiencies of A20 in mice lead to diseases mirroring those in humans (14, 15). As a ubiquitin-editing enzyme, A20 comprises an amino-terminal ovarian tumor (OTU) domain and seven zinc finger (ZnF) domains (16). The OTU domain is responsible for deubiquitinating activities, facilitating the hydrolysis of K48- and K63-linked ubiquitin chains (17). ZnF4 is notable for its ubiquitin-binding capability, predominantly to K63-chains, and incorporates K48-specific E3 ligase activity, critical for mediating proteasomal degradation (18). The ZnF7 motif, known for its affinity to methionine 1 (M1)-chains, plays a noncatalytic but essential role in ubiquitin binding, significantly contributing to the downregulation of the NF-κB signaling pathway (19).

According to existing research, A20 plays a significant role in modulating the priming of the NLRP3 inflammasome. This modulation is achieved through the restriction of NLRP3’s transcriptional synthesis, contingent on the inhibition of NF-κB, subsequently reducing the basal expression of NLRP3 (20, 21). Additionally, A20 limits the K63-linked ubiquitination of the pro-IL-1β complex, thus curbing spontaneous activation of the NLRP3 (22). Intriguingly, the isolated functionality of the A20’s ZnF4 or ZnF7 domain alone proves ineffective in restraining this spontaneous activation. However, synergistic employment of ZnF4 and ZnF7 motifs exhibits a marked inhibitory effect (23). This collaborative mechanism between the ZnF4 and ZnF7 domains underscores the noncatalytic capability of A20, independent of its ubiquitinating activity (23). These findings point to yet-to-be-uncovered mechanisms of A20 in the regulation of NLRP3 inflammasome activation, warranting further investigation.

In this study, our findings elucidate a mechanism by which A20 regulates NLRP3 inflammasome activation. This was investigated using A20 conditional knockout mice (A20^myel-KO), Nek7 conditional knockout mice (Nek7^myel-KO), and Nek7 conditional knockin mice (Nek7^myel-KI). Regarding the specific mechanism, our results demonstrate that A20 directly interacts with the NEK7 protein, mediating ubiquitination of NEK7 to accelerate its proteasomal degradation. Furthermore, A20 disrupts recruitment of NEK7 to the NLRP3 complex, thus attenuating inflammation induced by the NEK7–NLRP3–pyroptosis axis. Consequently, this study will offer a theoretical framework for the clinical management of inflammatory disease like sepsis.

Results

A20 Affects the Expression of NEK7 by NF-κB Signaling Pathway.

A previous study indicated that A20 deficiency can trigger spontaneous activation of the NLRP3 inflammasome in macrophage (22). To explore the specific mechanism, we generated myeloid-specific A20 knockout mice (A20^myel-KO) (SI Appendix, Fig. S1A) and verified a marked reduction in A20 expression in myeloid cells (SI Appendix, Fig. S1 B and C). Notably, A20-deficient BMDMs exhibited increased mRNA and protein levels of NEK7, both in the presence and absence of LPS stimulation (Fig. 1 A and B and SI Appendix, Fig. S1 D and E). To further elucidate the relationship between A20 and NEK7, we depleted A20 through the RNP system in iBMDMs and employed small interfering RNA (siRNA) targeting A20 in THP-1 (SI Appendix, Fig. S1F). Similarly, knockout of A20 in iBMDMs (A20^KO) resulted in spontaneously elevated expression level of NEK7, with mRNA and protein levels further amplified upon LPS stimulation (Fig. 1 C and D). A similar pattern was observed in THP-1 cells with attenuate expression of A20 (Fig. 1 E and F). Furthermore, our RNA-sequencing (RNA-seq) data further confirmed the regulation of NEK7 by A20 (SI Appendix, Fig. S1G).

Fig. 1.

A20 affects the expression of NEK7 at the transcription level. (A) BMDMs were extracted from A20^f/f or A20^myel-KO mice. qPCR analysis of Nek7 mRNA in A20^f/f BMDMs and A20^myel-KO BMDMs with or without treatment of LPS. (n = 3, mean ± SD). (B) Protein levels of A20 and NEK7 in cells described in (A) were determined by immunoblotting. (C) qPCR analysis of Nek7 mRNA in WT iBMDMs and A20^KO iBMDMs with or without LPS treatment. (n = 3, mean ± SD). (D) Protein levels of A20 and NEK7 in cells described in (C). (E) A20 expression in THP-1 cells was knocked down by A20 siRNA (siA20-2) or control siRNA (siCTRL), with or without LPS treatment. Nek7 mRNA was measured by qPCR analysis. (n = 3, mean ± SD). (F) Protein levels of A20 and NEK7 in cells described in (E). (G-L) Cells were all treated by LPS (100 ng/mL, 4 h). (G) qPCR analysis of Nek7 mRNA in A20^f/f BMDMs and A20^myel-KO BMDMs with or without treatment of BAY11-7082. (n = 3, mean ± SD). (H) Protein levels of A20 and NEK7 in cells described in (G). (I) qPCR analysis of Nek7 mRNA in WT iBMDMs and A20^KO iBMDMs with or without treatment of BAY11-7082. (J) Protein levels of A20 and NEK7 in iBMDMs mentioned in (I). (K) qPCR analysis of Nek7 mRNA in WT THP-1 and A20-knockdown THP-1 with or without treatment of BAY11-7082. (L) Protein levels of A20 and NEK7 in cells described in (K). *P < 0.05, **P < 0.01, ***P < 0.001.

Previous research has revealed that p65 could activate NEK7 expression through targeting its promoter (24). Thus, we next inquired whether the restricted effect of A20 depends on its properties in dampening the NF-κB signaling pathway. As expected, BAY11-7082 (specific inhibitor of NF-κB signaling pathway) was capable of restricting the expression level of NEK7 in both A20^myel-KO BMDMs and A20^KO iBMDMs (Fig. 1 G–J). Similar results were observed in A20-knockdown THP-1 (Fig. 1 K and L). We noted that in addition to the NF-κB signaling pathway, wolf-Hirschhorn syndrome candidate 1 (WHSC1) was shown to directly regulate the expression of NEK7 through H3K36me2 (25). However, our result suggested that protein levels of WHSC1 and H3K36me2 were equivalent in iBMDMs and A20^KO iBMDMs (SI Appendix, Fig. S1H). In addition, our data suggested that A20 was incapable of destabilizing NEK7 mRNA (SI Appendix, Fig. S1I). Together, these data indicated that A20 affected the transcriptional level of NEK7 through restricting the NF-κB signaling pathway.

A20 Is a Protein That Binds to NEK7.

Given A20’s character as a ubiquitin-editing enzyme, we further wondered whether A20 affects the turnover of NEK7 protein. Initially, A20 immunoprecipitates from ox-DNA-treated BMDMs were subjected to mass spectrometry, revealing NEK7 as a protein binding to A20 (Fig. 2A). Immunofluorescence staining in BMDMs demonstrated colocalization of A20 with NEK7 (Fig. 2B), and coimmunoprecipitation (Co-IP) assays further confirmed their interaction (Fig. 2C and SI Appendix, Fig. S2A). To investigate whether NEK7 associates with A20’s binding partners, NEK7 immunoprecipitates from BMDMs were also analyzed by mass spectrometry. We searched for several classic known binding partners of A20, including RIP1, NEMO, TRAF6, and TRAF2. While RIP1 and NEMO did not bind with NEK7, TRAF6, and TRAF2 were present in the NEK7 immunoprecipitates (SI Appendix, Fig. S2B). Furthermore, siRNA-mediated knockdown of TRAF6 or TRAF2 did not affect the binding of A20 and NEK7 (SI Appendix, Fig. S2C). Thus, these results suggested the association of A20 and NEK7 is independent of the potential binding partners of A20.

Fig. 2.

A20 is a protein that binds to NEK7. (A) BMDMs were treated with ox-dsDNA90, then A20 was immunoprecipitated with M2 beads and the binding proteins were analyzed by mass spectrometry. (B) Immunofluorescence staining of A20 (green), NEK7 (red), and DAPI (blue) in PBS or LPS plus ATP-activated BMDMs assessed by confocal microscopy. (Scale bar, 10 μm.) (C) HA-NEK7 and FLAG-A20 were transfected into HEK293T cells. Lysates were immunoprecipitated and then immunoblotted with the indicated antibodies. (D) Diagrammatic drawing the structure of A20 and NEK7. (E) HEK293T were transiently transfected with HA-NEK7 and indicated FLAG tagged A20 fragment or vector. HA-NEK7 protein was immunoprecipitated with an HA antibody and then samples were immunoblotted with the indicated antibodies. (F) HEK293T were transiently transfected with HA-NEK7, along with transient transfection of vector, FLAG-A20 WT (1 to 775), or ZnF4 (540 to 640), and lysates were immunoprecipitated and detected by immunoblotting with the indicated antibodies. (G) HEK293T were transiently transfected with HA-NEK7, along with transient transfection of vector, FLAG-A20 WT (1 to 775), or mutant ZnF4 (R540-V640), and lysates were immunoprecipitated and detected by immunoblotting with the indicated antibodies. (H) FLAG-A20 WT (1 to 775) or ZnF7 (741 to 775) were transiently transfected into HEK293T with transient transfection of HA-NEK7, protein level of A20 in immunoprecipitates of NEK7 was detected by immunoblotting with the indicated antibodies. (I) HEK293T were transiently transfected with FLAG-A20 and indicated plasmids of NEK7. Protein level of indicated NEK7 in immunoprecipitates of A20 was detected by immunoblotting with the indicated antibodies. The arrow indicates the right band. (J) HEK293T were transiently transfected with indicated plasmid, including combined mutation of OTU, ZnF4, and ZnF7 (^*OTU/^*ZnF4/^*ZnF7), simultaneous mutation of ZnF4 and ZnF7 (^*ZnF4/^*ZnF7), and individual mutation of OTU, ZnF4, or ZnF7 domains (^*OTU, ^*ZnF4, ^*ZnF7). Protein level of A20 in immunoprecipitates of NEK7 was detected by immunoblotting.

A20 is characterized by an OTU domain and seven ZnF domains, while NEK7 consists of an N-terminal lobe and a C-terminal lobe (Fig. 2D). To delineate the interactional region of A20 and NEK7, gene truncation experiment was implemented based on their structural domains. The OTU domain and ZnF fragment were both capable of binding to NEK7 (Fig. 2E). Given the ubiquitin-associated characteristics of the ZnF4 motif, we further verified whether NEK7 specifically interacted with ZnF4 in the absence of OTU motif. However, the binding of NEK7 and A20 occurred irrespective of the presence of ZnF4 domain alone or employing ZnF fragment with deleting ZnF4 motif (Fig. 2 F and G). This suggests the potential involvement of other domains within the ZnF fragment in binding to NEK7. According to previous study, ZnF7 motif possesses a noncatalytic ubiquitin binding function, known to mitigate inflammation by inhibiting TNF-induced NF-κB activation (19). Consistently, the NEK7 protein was identified in immunoprecipitated of ZnF7 motif, highlighting NEK7 as a protein that could bond to ZnF7 domain (Fig. 2H). Thus, our results confirm that the OTU, ZnF4, and ZnF7 motifs of A20 were all bind to NEK7, with the binding site localized in the C-terminal lobe of NEK7 (Fig. 2I).

To elaborate the role of each domain, we further designed diverse point mutants (cysteine-to-alanine) of A20 to abolish the function of its corresponding domain, including individual mutation of *OTU (C103A), *ZnF4 (C609A, C612A), or *ZnF7 (C764A, C767A) domains, simultaneous mutation of ZnF4 and ZnF7 (*ZnF4/*ZnF7), and combined mutation of OTU, ZnF4, and ZnF7 (*OTU/*ZnF4/*ZnF7). Our result suggested that *OTU and *OTU/*ZnF4/*ZnF7 could significantly attenuate the binding of A20 and NEK7 (Fig. 2J). Surprisingly, individual mutation of ZnF4 or ZnF7 made no difference compared to nonmutant type, while simultaneously mutated the ZnF4 and ZnF7 domains significantly impacted the binding of A20 and NEK7 (Fig. 2J). Therefore, ZnF4 and ZnF7 domains play a synergistic role in regulating the binding of these two proteins.

A20 Facilitates Degradation of NEK7 by Promoting Its Ubiquitination.

Additionally, we investigated how A20 affected the protein levels of NEK7. A20 and NEK7 were overexpressed in HEK293T cells, followed by treatment with the translational inhibitor cycloheximide (CHX). The protein level of NEK7 was measured 0 h, 4 h, 8 h, and 12 h after treatment. As anticipated, overexpressed A20 led to accelerated degradation of NEK7 (Fig. 3A). This reduction in NEK7 stability under A20-overexpressing condition was reversed upon treatment with the proteasome inhibitor MG132 (Fig. 3B), suggesting that A20 facilitated degradation of NEK7 through the proteasome pathway. Notably, ubiquitination of NEK7 was enhanced in peritoneal macrophages of septic mice (Fig. 3C) as well as in LPS plus ATP or ox-DNA treated iBMDMs (SI Appendix, Fig. S2 D and E). Indeed, A20 was confirmed as the facilitator of NEK7’s ubiquitination in the presence or absence of LPS plus ATP or ox-DNA (Fig. 3 D and E and SI Appendix, Fig. S2 D and E). Notably, K63-linked ubiquitination of NEK7 was significantly increased in A20^KO iBMDMs under the conditions of stimulation, while its K48-linked ubiquitination demonstrated inconspicuous change (SI Appendix, Fig. S2 D and E). To determine whether A20 directly ubiquitinates NEK7 in vitro, NEK7 purified from HEK293T (overexpressing 3× HA tagged NEK7) was subjected to in vitro ubiquitination assays with recombinant A20. Then, NEK7 was immunoprecipitated again with its antibody, and NEK7 immunoprecipitates were subjected to immunoblotting. Our findings confirmed that A20 directly ubiquitinates NEK7, establishing NEK7 as a substrate of A20 (Fig. 3F and SI Appendix, Fig. S2F). Moreover, A20 reduced K63-linked ubiquitination of NEK7 (Fig. 3G and SI Appendix, Fig. S2 D and E), and our data indicated A20 also increased the K48-linked ubiquitination of NEK7 (Fig. 3H). Given that the ZnF4 domain of A20 mediates K48-specific E3 ligase activity, our result aligned with this property, suggesting a potential role of A20 in destabilizing NEK7 by replacing K63 ubiquitin chains with K48 ubiquitin chains.

Fig. 3.

A20 promotes degradation of NEK7 by promoting NEK7 ubiquitination. (A) Immunoblots for HA and FLAG of HEK293T after transient transfection of HA-NEK7 with or without transient transfection of FLAG-A20, and the time-related quantification of HA was indicated with a line chart. Cells were all treated with protein translation inhibitor CHX (50 μg/mL). (B) Immunoblots for HA and FLAG of HEK293T after transient transfection of HA-NEK7 with or without transient transfection of FLAG-A20, and the time-related quantification of HA was indicated with a line chart. Cells were all treated with proteasome inhibitor MG132 (10 μM). (C) Co-IP of NEK7 was implemented in cell lysates from peritoneal macrophage (PM) in mice treated with intraperitoneal injection of PBS or LPS (5 mg/mL) for 24 h, and ubiquitinated proteins are detected by immunoblotting. (D and E) HA-NEK7, FLAG-A20, and His-Ub were transiently transfected into HEK293T. Immunoprecipitates were immunoblotted with the indicated antibodies. Cells were treated by MG132 without stimulation (D) or stimulated with ox-dsDNA90 (E). (F) In vitro ubiquitination assays. Co-IP was implemented with anti-NEK7 antibody after in vitro reaction and ubiquitinated proteins are detected by immunoblotting. (G and H) HEK293T was transiently transfected with HA-NEK7, FLAG-A20, and Ub-K63 (G) or Ub-K48 (H) to implement Co-IP, and K63-linked or K48-linked ubiquitinated proteins are detected by immunoblotting. Cells were all treated by MG132.

Two Ubiquitinated NEK7 Lysine Residues Are Critical for A20-Mediated NEK7 Ubiquitination and Degradation.

To unravel the specific molecular mechanism through which A20 regulated NEK7, we referred to the PhosphoSitePlus database and noted that NEK7 comprised 11 ubiquitinated lysines across both the N-terminal and C-terminal lobes (Fig. 4A). We then mutated these lysines (K) to arginines (R) and initially generated two NEK7 mutants: NEK7 (N-6KR) and NEK7 (C-5KR) (Fig. 4A). Significantly, the NEK7 (C-5KR) mutant demonstrated enhanced protein stability and greater resistance to degradation under the condition of A20 overexpression compared to wild-type NEK7 (WT) and NEK7 (N-6KR) (Fig. 4B). Moreover, both the K48- and K63-linked ubiquitination levels of NEK7 (C-5KR) mutant were markedly reduced, while mutations of all ubiquitinated sites in the N-terminal lobe resulted in minimal change in ubiquitination level (Fig. 4 C and D). These data suggested that the sites facilitating A20-mediated ubiquitination of NEK7 were localized in its C-terminal lobe.

Fig. 4.

Two ubiquitinated NEK7 lysine residues are critical for A20-mediated NEK7 ubiquitination. (A) Schematic depiction of the NEK7 point mutant. (B) Immunoblots for HA and FLAG from HEK293T cells after transient transfection of NEK7-HA WT, N-6KR, or C-5KR with or without transient transfection of FLAG-A20. Cells were all treated by CHX. (C and D) HEK293T cells were transiently transfected with indicated plasmids, and the cell lysates were immunoprecipitated, with detecting K48-linked ubiquitinated proteins (C) and K63-linked ubiquitinated proteins (D) by immunoblotting. Cells were all treated with MG132. (E) Immunoblots of HA in HEK293T cells after transiently transfected with indicated plasmid with or without transient transfection of FLAG-A20. Cells were all treated by CHX. (F and G) HEK293T cells were transiently transfected with FLAG-A20 and His-Ub, and respectively transfected with HA-NEK7 K189R, K248R/K249R, K281R, or K293R. HA-NEK7 protein was immunoprecipitated and then K48-linked ubiquitinated proteins (F) and K63-linked ubiquitinated proteins (G) were detected by immunoblotting. Cells were treated with MG132. (H) Immunoblots for HA in HEK293T cells after transient transfection of HA-NEK7 WT, K189R, or K293R, and the time-related quantification of HA was indicated as a line chart. Cells were all treated by CHX. (I) Lys189 and Lys293 of NEK7 are evolutionarily conserved in different species.

Subsequently, we generated individual K-to-R mutation of NEK7 and tested the stability of these mutants under A20 overexpression to ascertain the ubiquitination sites. Our data demonstrated that the protein levels of NEK7 were more stable under the condition of overexpress A20 when mutating the Lys189 and Lys 293 of NEK7 (Fig. 4E), accompanied with reduced K48-linked and K63-linked ubiquitination levels in these two types of mutants (Fig. 4 F and G). Furthermore, both K189R and K293R mutant were efficient in restricting the degradation of NEK7 by A20 (Fig. 4H), revealing that both K189R and K293R had influence on the ubiquitinated level and proteasomal degradation of NEK7. Importantly, we also found that Lys189 and Lys293 of NEK7 are evolutionarily conserved in other species (Fig. 4I), suggesting that Lys189 and Lys293 of NEK7 are potential targets of A20 and the regulatory mechanism may be all-purpose in diverse species.

A20 Affects the Binding of NEK7 to the NLRP3 Complex.

We next validated whether these ubiquitination sites affected the binding of A20 to NEK7. Consistent with the result mentioned above (Fig. 2I), the binding sites of A20 were localized to the C-terminal lobe of NEK7 and within the five ubiquitinated sites (Fig. 5A). Interestingly, individual mutation experiment suggested that only K189R mutant was efficient in restricting the binding of A20 to NEK7 (Fig. 5B), though both Lys189 and Lys293 are crucial to support A20’s ubiquitinated and degradative capability. As stated above, posttranscriptional modification of NEK7 affects its function of modulating NLRP3 activation. Therefore, we examined whether A20 affected the binding of NEK7 to the NLRP3 complex. Our data from Co-IP experiments indicated that mutation of Lys189 and Lys293 of NEK7 had almost no effect on the binding of NEK7 and NLRP3 (Fig. 5C), while the presence of A20 markedly reduced the binding of NEK7 to NLRP3 (Fig. 5D). Moreover, mutation of all the ubiquitinated sites in the N-terminal lobe or C-terminal lobe both made no difference in interfering the interaction of NEK7 and NLRP3 with the presence of A20 (Fig. 5E). These data suggest that A20 affects the binding of NEK7 to the NLRP3 complex, while this capability is independent the ubiquitinated sites of NEK7.

Fig. 5.

A20 limits binding of NEK7 to NLRP3. (A) HEK293T cells were transiently transfected with FLAG-A20, HA-NEK7 WT, N-6KR, or C-5KR. Lysates were immunoprecipitated and the level of A20 were detect by immunoblot. (B) HEK293T cells were cotransfected FLAG-A20 with HA-NEK7 K189R, K248R/K249R, K281R, or K293R. Cell lysates were immunoprecipitated with an HA antibody and indicated proteins were detected by immunoblot. (C) HEK293T cells were cotransfected NLRP3 plasmid with HA-NEK7 WT, C-5KR, K189, or K293. Lysates were immunoprecipitated and implemented immunoblot to detect specific protein with indicated antibodies. (D) NLRP3 and HA-NEK7 plasmids were transiently transfected into HEK293T cells with or without transient transfection of FLAG-A20. Lysates were immunoprecipitated with an HA antibody and the level of specific proteins was detected by immunoblot. (E) HEK293T cells were transiently transfected with FLAG-A20 and NLRP3, along with respectively transient transfection of HA-NEK7 WT, C-5KR or N-6KR. Lysates were immunoprecipitated and immunoblotting was implemented to detect protein level of NLRP3 in NEK7 immunoprecipitates. (F) HEK293T cells with transient transfection of NLRP3 and HA-NEK7 were assigned to respectively transfected with indicated mutants of A20. Lysates were immunoprecipitated and specific protein was detected by immunoblotting. (G) A20-derived peptides (P-I and P-II) were dissolved with ddH₂O and diluted into indicated concentration, and the kinetic interaction of NEK7 and A20-derived peptides (P-I and P-II) were detected by Surface Plasmon Resonance (SPR) analyses. The fitting curve is shown as a red dotted line. (H) HEK293T were cotransfected NLRP3 plasmid and HA-NEK7 and treated with P-II (0 μM, 2.5 μM, 5 μM, 10 μM, or 20 μM). Lysates were immunoprecipitated and implemented immunoblot to detect protein level of NLRP3. (I and J) A20^KO iBMDMs were treated with PBS, LPS plus ATP, or LPS plus ATP with treatment peptide II (0 μM, 2.5 μM, 5 μM, 10 μM or 20 μM). Secretion of IL-1β (I) and IL-18 (J) into the cell culture medium was detected by ELISA. (n = 3, mean ± SD); (K) LDH release of A20^KO iBMDMs as described in (I). *P < 0.05, ***P < 0.001.

To delineate the specific mechanism, we used multiple mutants of A20 via cysteine-to-alanine point mutations (Fig. 2J). Our result suggested that individual mutation of *OTU (C103A) motif, simultaneous mutation of A20’s *OTU (C103A), *ZnF4 (C609A, C612A), and *ZnF7 (C764A, C767A) could completely restore the binding between NEK7 and NLRP3 (Fig. 5F). This underscored the critical role of the OTU domain in modulating the binding of NEK7 to NLRP3. Intriguingly, individual mutation of ZnF4 motif or ZnF7 motif barely abrogated the interfering capability of A20; however, concurrent mutation of A20’s ZnF4 and ZnF7 motifs effectively reinstated the NEK7–NLRP3 interaction (Fig. 5F). These findings corroborate previous research, demonstrating the synergistic role of the ZnF4 and ZnF7 motifs in constraining spontaneous activation of NLRP3 inflammasome (23).

In addition, we noted that catalytic Cys103 exists in a α-helix (helix α4) (26). Previous studies have suggested that α-helix are critical for protein–protein interactions (PPIs) and have high potential for therapeutic development. Therefore, we generated two peptides containing helix α4: P-I and P-II. The P-I exclusively consists of this helix, while the P-II encompasses loops on both sides of helix α4. Two types of peptides were synthesized and respectively dissolved in ddH₂O, and the solutions were diluted to various concentration, which were subsequently conducted on Surface plasmon resonance (SPR) analysis and implemented to LPS plus ATP-treated A20-deficient iBMDMs. Consequently, P-II displayed a high affinity for NEK7 (Fig. 5G) and impeded the NEK7 and NLRP3 interaction in a dose-dependent manner (Fig. 5H). Moreover, employment of P-II significantly decreased the secretion of IL-1β and IL-18, as well as limited LDH release under the condition of LPS plus ATP (Fig. 5 I–K). Thus, P-II could protect against the activation of NLRP3 inflammasome. To sum up, A20 is capable of restricting hyperactivation of NLRP3 by modulating NEK7.

NEK7 Is Critical for A20-Deficiency-Mediated Activation of NLRP3 Inflammasome.

Previous reports showed that deletion of A20 in macrophages gave rise to more significant pyroptosis after LPS plus ATP stimulation (21). To test the potential role of NEK7 in response to LPS plus ATP in A20-deficient macrophages, NEK7 was knocked down by utilizing siRNA and detected by immunoblotting (SI Appendix, Fig. S3A). Notably, LPS plus ATP was capable of inducing NLRP3 inflammasome activation and downstream pyroptosis in the absence of A20. Whereas, the simultaneous deletion of NEK7 and A20 in BMDMs markedly inhibited NLRP3-dependent caspase-1 processing, IL-1β and IL-18 secretion, and LDH release (Fig. 6 A–D), and results were similar in ox-DNA-stimulated BMDMs (SI Appendix, Fig. S3 B–E). Consistently, results of immunofluorescence suggested that formation of ASC speck was increased after knocking-out of A20 and under the condition of LPS plus ATP, and this elevation could be reversed by simultaneous knocking-down of NEK7 (SI Appendix, Fig. S3 F and G).

Fig. 6.

NEK7 are critical for A20-deficiency-mediated activation of NLRP3 inflammasome. (A) A20^f/for A20^myel-KO BMDMs were transfected with siCTRL or siNEK7-1 and then stimulated with LPS plus ATP. Cleaved and total Caspase-1, cleaved and total GSDMD, NEK7, and A20 were determined by immunoblotting. (B and C) The levels of IL-1β (B) and IL-18 (C) in cell culture medium of BMDMs as post in A were measured by ELISA. (n = 3, mean ± SD). (D) LDH release of BMDMs as described in (A). (n = 3, mean ± SD). (E) A20^f/for A20^myel-KO BMDMs were treated with berberine (10 μM/mL) or MCC950 (7.5 nM) and then stimulated with LPS plus ATP. Cleaved and total Caspase-1, cleaved and total GSDMD, NEK7, and A20 were determined by immunoblotting. (F) Secretion of IL-1β into the cell culture medium of BMDMs as post in (E). (n = 3, mean ± SD). (G) LDH release of BMDMs as mentioned in (E). (n = 3, mean ± SD). (H) THP-1 cells transfected with siCTRL or siA20-2 were treated with berberine or MCC950 and then stimulated with LPS followed by ATP. Cleaved and total Caspase-1, cleaved and total GSDMD, NEK7, and A20 were determined by immunoblotting. (I) The level of IL-1β in cell culture medium of THP-1 cells in (H). (n = 3, mean ± SD). (J) LDH release of THP-1 cells in (H). (n = 3, mean ± SD). *P < 0.05, **P < 0.01, ***P < 0.001.

To further elucidate NEK7’s role in A20 regulating NLRP3 inflammasome, we employed MCC950, a specific inhibitor of NLRP3, and berberine, known to mitigate NLRP3 inflammasome mediated inflammation by targeting to NEK7 (27). We observed that the elevated cleavage of GSDMD and Caspase-1, resulting from A20 deletion, was attenuated by inhibiting NEK7 function using berberine. Remarkably, berberine’s effectiveness paralleled the direct inhibition of NLRP3 by MCC950 in BMDMs (Fig. 6E). Accordingly, the maturation of IL-1β and IL-18, as well as release of LDH, were decreased upon treatment with MCC950 and berberine (Fig. 6 F and G and SI Appendix, Fig. S3H). These findings were corroborated in siA20-treated THP-1 cells (Fig. 6 H–J and SI Appendix, Fig. S3I). Therefore, our result demonstrated that A20 restricts NLRP3 activation through its regulatory influence on NEK7.

Otherwise, study has shown that secretion of IL-1β was elevated in A20-deficient BMDMs when treated with LPS alone (22). Thus, we further investigated whether the elevation was involved with NEK7. Consistently, our data showed significantly elevated expression level of pro-IL-1β in A20-deficient iBMDMs, along with elevated secretion of IL-1β and release of LDH under the stimulation of LPS (SI Appendix, Fig. S4 A–C). Subsequently, A20-deficient iBMDMs were infected with lentiviruses containing small interfering RNA targeting NEK7 (Nek7^KD). Interestingly, A20^KO/Nek7^KDiBMDMs demonstrated significantly increased proliferative activity compared to A20^KO iBMDMs (SI Appendix, Fig. S4D). However, when stimulated with LPS alone, A20^KO/Nek7^KDiBMDMs showed analogous tendency of pro-IL-1β expression level and maturation of IL-1β compared to A20-deficient iBMDMs (SI Appendix, Fig. S4 E and F), while release of LDH was significantly decreased in A20^KO/Nek7^KDiBMDMs (SI Appendix, Fig. S4G). Therefore, these results indicated that secretion of IL-1β in A20-deficient cells was independent of the function of NEK7 when stimulated with LPS alone.

NEK7 Deficiency in Macrophage Alleviates the Process of Sepsis.

Analysis of a database from a latest representative published study (28) suggested NEK7 was significantly increased in cells associated with the progression of sepsis (SI Appendix, Fig. S5 A–D). Moreover, our data based on blood samples from 12 septic patients and 12 healthy donors demonstrated that expression level of NEK7 was elevated in myeloid cells from septic patients in comparison to that from healthy individuals (SI Appendix, Fig. S5 E and F). These data suggest that the expression of NEK7 has critical clinical relevance in sepsis. To further confirm whether NEK7 is the therapeutic target of pyroptosis mediated disease like sepsis, Nek7^myel-KO (SI Appendix, Fig. S6 A–C) and Nek7^myel-KI mice (SI Appendix, Fig. S6 D–F) were utilized, and then, the Nek7^myel-KO, Nek7^myel-KI, and Nek7^f/f mice were intraperitoneally injected with LPS. Notably, Nek7^f/f mice were all dead within 36 h, while Nek7^myel-KOmice entirely survived (Fig. 7A), along with efficaciously alleviated lung injury compared to Nek7^f/f mice (Fig. 7B). Accordingly, serum level of mature IL-1β, IL-18 were also significantly attenuated in Nek7^myel-KO mice (Fig. 7 C and D), though the secretion of TNF-α and IL-6 remained unchanged (SI Appendix, Fig. S7 A and B). On the contrary, Nek7^myel-KImice exhibited deteriorative overall survival and more severe lung injury after treated with LPS, along with excessive secretion of IL-1β, IL-18 (SI Appendix, Fig. S7 C–H).

Fig. 7.

Targeting NEK7 can alleviate the process of sepsis. (A) Survival curves of Nek7^f/f and NEK7^myel-KOmice treated with LPS (25 mg/kg). (n = 10). (B) Representative images of HE staining of the lung tissue sections from mice as described in (A). (Scale bar, 50 μm.) (n = 5). (C and D) Mice as described in (A) were treated with LPS (25 mg/kg) for 3 h. Serum levels of IL-1β (C) and IL-18 (D) were measured. (n = 5, mean ± SD). (E) Diagrammatic drawing of macrophage depletion and reconstitution procedure. (F) Survival curves of A20^mϕ-Ctrl, A20^mϕ-KD, NEK7^mϕ-KO/A20^mϕ-Ctrl or NEK7^mϕ-KO/A20^mϕ-KD mice treated with LPS (25 mg/kg). (n = 10). (G) Representative images of HE staining of the lung tissue sections from mice as described in (F). (Scale bar, 50 μm.) (n = 5). (H and I) Mice as described in (F) were treated with LPS (25 mg/kg) for 3 h. Serum levels of IL-1β (H) and IL-18 (I) were measured. (n = 5, mean ± SD). (J) A20-derived peptides were dissolved in PBS and diluted into 2 mg/mL. Mice were weighed and intraperitoneally injected with PBS, P-Ctrl, or P-II (15 mg/kg) 1 h before LPS challenge. Survival of mice was evaluated. (n = 10). (K) Representative images of HE staining of the lung tissue sections from mice as described in (J). (Scale bar, 50 μm.) (n = 5). (L and M) Mice as described in (J) were treated with LPS (25 mg/kg) for 3 h. Serum levels of IL-1β (L) and IL-18 (M) were measured. (n = 5, mean ± SD). *P < 0.05, ***P < 0.001.

A20 plays a pivotal role in macrophages to protect mice from endotoxic shock while mice lacking A20 are sensitized to LPS challenge. To further confirm whether knockout of NEK7 was efficient in rescuing mice with A20-deficiency/A20-knockdown in myeloid cells, we generated NEK7^mϕ-KO/A20^mϕ-KDmice by employing shA20 to myeloid cells of Nek7^myel-KOmice, and A20^mϕ-Ctrl, A20^mϕ-KD, NEK7^mϕ-KO,or NEK7^mϕ-KO/A20^mϕ-KD macrophages were intravenously injected to the mice whose macrophages are precleared by clodronate (Fig. 7E and SI Appendix, Fig. S7I). Mice with A20^KDmacrophages were more susceptive to LPS stimulation and demonstrated rapid and overall death within 24 h (Fig. 7F). Whereas, mice injected with NEK7^mϕ-KO and NEK7^mϕ-KO/A20^mϕ-KDmacrophages showed significantly increased survival rate and survival time (Fig. 7F), along with efficaciously alleviated lung injury and decreased secretion of IL-1β and IL-18 resulted from the myeloid deficiency of A20 (Fig. 7 G–I). Consistently, the secretion of TNF-α and IL-6 remained unchanged (SI Appendix, Fig. S7 J and K).

Moreover, we verified the therapeutic significance of the peptide we designed. Mice were treated with P-II (15 mg/kg) or control peptide 1 h before LPS challenge. P-II markedly improved the survival of LPS-treated mice (Fig. 7J). And alleviative lung injury was observed in septic mice treated with P-II, along with decreased secretion of IL-1β and IL-18 to the serum (Fig. 7 K–M), while levels of TNF-α and IL-6 in serum were not altered (SI Appendix, Fig. S7 L and M). Thus, these data suggested NEK7 as a promising target for therapy of NLRP3-mediated disease like sepsis.

Discussion

This study conducted several investigations into A20, NEK7, and NLRP3, unveiling a previously unrecognized mechanism of A20 in regulating NLRP3 inflammasome activation, and establishing the interaction between A20 and NEK7. Innovatively, our results showed that NEK7 expression was significantly up-regulated in A20-deficient BMDMs and iBMDMs, with this inverse regulatory relationship hinging on A20’s ability to restrict the NF-κB signaling pathway. Verified through LC–MS, immunofluorescence staining, and coimmunoprecipitation, our findings identified A20 as a protein binding to NEK7, with this interaction dependent on OTU and /or collaboration of ZnF4 and ZnF7 domains. Additionally, our in vivo and in vitro ubiquitination assays revealed that NEK7 serves as a substrate for A20, which promotes proteasomal degradation of NEK7 via ubiquitination at sites K189 and K293. By generating various point mutations and truncated protein of A20 or NEK7, we further demonstrated A20’s capacity to disrupt the binding of NEK7 and NLRP3, a process reliant on the OTU domain. Finally, we confirmed that A20 deficiency-mediated activation of NLRP3 inflammasome was linked to modulation of NEK7. Targeting NEK7 is capable of attenuating NLRP3 inflammasome-mediated pyroptosis and ameliorating the survival of septic mice.

As a potent inhibitor of NF-κB signaling, A20 exerts a negative influence on NF-κB-induced expression of NLRP3 and pro-IL-1β, thereby restraining the assembly of the NLRP3 inflammasome. These findings highlight A20’s impact on the priming step of inflammasome activation (22, 29). Moreover, LPS-induced activation of NF-κB signaling pathway has been shown to increase NEK7 expression by affecting its promoter region (24). These observations suggested that A20 might impact expression of NEK7 at the transcriptional levels. Indeed, our data showed that basal mRNA and protein expression of NEK7 in A20-deficient macrophages spontaneously elevated without additional stimuli. In agreement, the inhibitor of NF-κB BAY11-7082 significantly reduced NEK7 levels in A20-deficient macrophages. While previous study indicated WHSC1 regulated the expression of NEK7 through H3K36me2 (25), our findings did not observe changes in WHSC1 and H3K36me2 in A20 deleted macrophages relative to normal macrophages. Furthermore, the mRNA stability of NEK7 is not affected in the absence of A20. Together, our findings revealed that A20 affects the transcriptional level of NEK7 by NF-κB signaling pathway.

As a ubiquitin-editing enzyme, A20 facilitates the hydrolysis of K48- and K63-linked ubiquitin chains, exerting its deubiquitinating activity via the OTU domain (30). Additionally, A20’s ZnF4 domain functions as a ubiquitin ligase, polyubiquitinating substrates with K48-linked chains that mediate proteasomal degradation (31). Previous study has shown that A20 restricts K63-linked ubiquitination of the pro-IL-1β complex, thereby limiting spontaneous activation of the NLRP3 inflammasome (22). In this study, we found that A20 decreases the K63-ubiquitination and increases the K48-ubiquitination of NEK7, leading to its proteasomal degradation. This process involves A20 directly binding to NEK7, as evidenced by the involvement of NEK7’s Lys189 and Lys293 sites. Consequently, the increase in NEK7 levels in A20-deficient macrophages can be attributed to both heightened transcription and inhibition of ubiquitination-mediated degradation. The NF-κB-regulatory capacity of A20 relies on the synergistic effect of OTU and ZnF4 motifs (32). Meanwhile, the presence of the OTU domain is also indispensable to the ubiquitin-ligase activity of ZnF4, ultimately promoting protein degradation via the proteasome (29). Thus, quantifying the relative contribution of these two mechanisms to the increased abundance of NEK7 remains challenging.

Additionally, we found that A20 can interfere with the binding of NEK7 to NLRP3. Recent studies have indicated that NEK7 is a crucial factor in NLRP3 activation, achieved through direct binding with NLRP3 (33). Posttranscriptional modification of NEK7, such as phosphorylation and deglutathionylation, can influence its interaction with NLRP3 (10, 11). Consequently, we investigated whether ubiquitination of NEK7 affects this interaction. Our data showed that mutation of ubiquitinated sites of NEK7 did not diminish its binding to NLRP3, while A20 overexpression significantly decreased the protein level of NLRP3 in NEK7 immunoprecipitates. Further experiments with truncated A20 proteins confirmed that this interference depended on the OTU domain and/or synergistic effect of A20’s ZnF4 and ZnF7 motifs. Given the OTU domain’s pivotal role in restrict NLRP3 inflammasome, we generated an α-helix peptide containing C103 amino acid residues. Intriguingly, although the peptide lacks catalytic activity (34), it disrupted the NEK7–NLRP3 interaction and limited pyroptosis. Published work suggested that A20 OTU domain restricted LPS signaling by limiting MAPK and NF-κB activation (35). Our data may offer an interesting perspective on this phenomenon. Additionally, our data also partially elucidate the mechanism behind synergistic function of A20’s ZnF4 and ZnF7 motifs in restricting the activation of NLRP3 inflammasome (23). However, the specific mechanism remains to be further explored.

Our data showed that A20 attenuated activation of NLRP3 inflammasome by reducing basal expression of NEK7, promoting NEK7 to proteasomal degradation and disturbing the interaction of NEK7 and NLRP3 (SI Appendix, Fig. S8). Yet, it is unclear which of the three mechanisms is the key determinant one on this A20–NEK7–NLRP3 axis inflammasome pathways. Previous study has indicated that abrogating A20’s ZnF4 motif did not impair A20’s ability to restrain IL-1β secretion following LPS treatment, both in vivo and vitro (23, 35). Similarly, while A20’s ZnF7 motif mutations are sensitive to TNF stimulation, they do not significantly respond to LPS stimulation (23). These results indicated the ZnF4 and ZnF7 motifs played a limited role in regulating the NLRP3 inflammasome. In contrast, OTU domain mutations in mice and cells have shown heightened sensitivity to LPS challenge, evidenced by increased lethality, prolonged MAPK and NF-κB activation, and elevated cytokines (35), highlighting the OTU domain’s importance in restraining NLRP3. Additionally, we observed that an A20-derived peptide significantly dampened pyroptosis of LPS plus ATP-treated iBMDMs and improved survival in the sepsis model, suggesting another aspect of the OTU domain’s function in NLRP3 regulation: its noncatalytic binding capability. In conclusion, the OTU domain is pivotal to regulation of A20–NEK7–NLRP3 axis, with binding function in disturbing the interaction of NEK7 and NLRP3 may be the key determinant in the NLRP3 inflammasome pathways.

Notably, we noted that NEK7 has significant clinical relevance in sepsis by analyzing a database of a latest representative published study (28). Especially, NEK7 significantly increased in MS1, MS2, and MS3, which were verified to associated with the progression of sepsis (28). According to our data, Nek7^myel-KO mice exhibited remarkable resistance to LPS treatment, while Nek7^myel-KI mice were more susceptible to such stimulation. Furthermore, A20-derived NEK7 targeting peptide significantly alleviates pyroptosis, as well as attenuates lung injury and improves survival of septic mice. These data indicate the potential role of NEK7 in serving as a therapeutic target of sepsis. Additionally, our data innovatively revealed that A20 is entwined with suppressing NLRP3 activation by interacting with NEK7. In vitro, our data verified that the activation of NLRP3 inflammasome in A20-deficient macrophages can be limited by depleted or pharmacological inhibition of NEK7. In vivo, macrophages-depleted mice showed improved survival and alleviated lung injury when reconstituting with Nek7 ^myel-KOmacrophages and A20^myel-KD/Nek7 ^myel-KOmacrophages. Whereas, mice injected with NEK7-deficient macrophages did not resist the stimulus as effectively as the Nek7^myel-KO mice. The observed incongruity might stem from the efficiency of clodronate in thoroughly depleting all macrophages in murine models, encompassing the anti-inflammatory phenotype that mitigates cellular demise and tissue impairment. However, intravenous administration of macrophages failed to restore the anti-inflammatory potential.

In summary, our data indicate mechanisms of A20 in regulating NEK7, which reiterates its crucial role in regulating the activation of the NLRP3 inflammasome and broadens the horizon of related research. Furthermore, NEK7 emerges as a promising target for potential therapeutic interventions in A20 and/or NLRP3-associated diseases, wherein its down-regulation or inhibition via genetic techniques or small molecules is presumed to hold therapeutic promise.

Materials and Methods

Mice and Cells.

Myeloid-specific A20^myel-KO, Nek7^myel-KO, and Nek7^myel-KI mice were developed by Cyagen Biosciences. C57BL/6 J mice were purchased from Vital River Laboratories (VRL) in Beijing. All mice were housed in the specific pathogen-free (SPF) facility within our laboratory, and all studies adhered to the ethical standards and regulations prescribed by the Animal Ethics Committee of Sichuan University. iBMDMs was generously provided by Feng Shao (National Institute of Biological Sciences, Beijing, China). L929, THP-1, and HEK293T were all purchased from ATCC.

Human Samples.

Human samples were collected from the blood of 12 patients with sepsis and 12 healthy donors for comparative analysis. The procurement and utilization of these samples were conducted under the approval of the Ethics Committee of West China Hospital of Stomatology, Sichuan University (Ethical code: WCHSIRB-D-2023-111).

Targeting of A20 by RNP-Mediated Genome Editing.

Gene editing mediated by RNP was implemented by GeneWiz company. To generate an A20 knockout cell line, the RNP system containing the sgRNAs and Cas9 protein was constructed. Then, electroporation was used to transfer the RNP system into iBMDMs. Three weeks later, single clones with A20 deficiency were identified by Sanger sequencing and immunoblotting with anti-A20 antibody.

Coimmunoprecipitation.

To perform the coimmunoprecipitation (Co-IP) assay, HEK293T cells were transiently transfected with indicated plasmids. After 36 h, cells were lysed in mild lysis buffer with protease inhibitor cocktail on ice for 10 min. Samples were centrifuged at 12,000 rpm at 4 °C for 10 min, and the supernatant was incubated with the corresponding primary antibody overnight at 4 °C. On the next day, magnetic beads were added and incubated for 1 h. The immunocoprecipitated protein was eluted and then heated with 4× Laemmli loading buffer at 95 °C for 5 min. Samples were then subjected to immunoblot analysis.

Mass Spectrometry Analysis.

Primary BMDMs were cultured to 90% density in 10 cm dishes and then treated with LPS followed by ATP or with ox-dsDNA90. Then, cells were lysed in a mild lysis buffer plus protease inhibitor cocktail. Immunoprecipitation was performed with A20/NEK7 antibody and Pierce Protein A/G Magnetic Beads, and MS assay was used for analysis.

In Vitro Ubiquitination Assays.

The 3× HA-NEK7 plasmids were transfected into HEK293T cells. Then, 36 h later, NEK7 protein was purified with an HA antibody via Co-IP and eluted with HA peptides. Recombinant A20 protein was purchased from Abcam. In vitro ubiquitination assays were carried out in 50 μL reaction volumes and reactions were incubated at 37 °C for 4 h. After the reaction was complete, samples were purified with an anti-NEK7 antibody via Co-IP. Then, the NEK7 immunoprecipitates were subjected to SDS-PAGE for the detection of ubiquitination.

All other materials and details of experimental procedures are provided in SI Appendix, Materials and Methods.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

The RNA-seq data reported in this paper have been deposited in the Gene Expression Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accession no. GSE253265) (36). All other data are included in the manuscript and/or SI Appendix.