NLRP3 is a key component of caspase-activating macromolecular protein complexes called inflammasomes. It has been found that DHX33 is a cytosolic dsRNA sensor for the NLRP3 inflammasome, which induces caspase-1–dependent production of IL-1β and IL-18 upon activation. However, how the cytosolic dsRNAs induce the interaction between DHX33 and the NLRP3 inflammasome remains unknown. In this study, we report that TRIM33, a member of the tripartite motif (TRIM) family, can bind DHX33 directly and induce DHX33 ubiquitination via the lysine 218 upon dsRNA stimulation. Knocking down of TRIM33 abolished the dsRNA-induced NLRP3 inflammasome activation in both THP-1–derived macrophages and human monocyte-derived macrophages. The ubiquitination of DHX33 by TRIM33 is lysine 63 specific and is required for the formation of the DHX33–NLRP3 inflammasome complex.
Tripartite motif (TRIM) family proteins contain RBCC motifs, which consist of the RING (really interesting new gene 1) finger domain, B-box motif, and a coiled-coil domain (1, 2). Most of the TRIM family members are E3 ubiquitin (Ub) ligases. These proteins interact with Ub-conjugating enzymes (E2) via their RING domains and transfer Ub from Ub-activating enzymes (E1) to the target molecules (3–5). Many TRIM members are IFN-stimulated genes and play important roles in a broad range of immune responses including antimicrobial infection (6, 7). It has been reported that TRIM25 ubiquitinates the caspase recruitment domains of retinoic acid–inducible gene-I (RIG-I), and this ubiquitination activity is essential for the activation of downstream antiviral innate immune responses (8). TRIM5α has been intensively studied with its well-known retroviral restriction activity (9). TRIM21 negatively regulates an intracellular dsDNA sensing pathway by ubiquitinating and degrading DDX41 (10). TRIM30α induces TAB2 and TAB3 ubiquitination and degradation, and it inhibits TRAF6-induced NF-κB activation (11). Ubiquitination of stimulator of IFN gene (STING) by TRIM56 is essential for STING dimerization and IFN-β promoter activation (12).
TRIM33, previously known as transcriptional intermediary factor 1 γ (TIF1-γ), has been shown to function in transcriptional regulation during hematopoiesis (13). It is also reported to have tumor suppressor activity in multiple tissues (14, 15). A recent study reported that TRIM33 functions in DNA repair (16). It is unknown whether TRIM33 plays a role in the innate immune system.
Inflammasomes are caspase-activating multiprotein complexes that were identified in 2002 (17). NLRP3 is a member of Nod-like receptors. Upon activation, NLRP3 forms a macromolecular signaling complex with its adaptor protein ASC and procaspase-1 called the NLRP3 inflammasome (18, 19). This leads to the cleavage and activation of caspase-1, which in turn processes the proforms of IL-1β and IL-18 to generate biologically active cytokines (20). Multiple types of stimulatory signals can activate the NLRP3 inflammasome, including ATP, crystalline reagents and microbial toxin nigericin (19, 21, 22). It’s believed that these stimuli may activate the NLRP3 inflammasome via different pathways (23–26). Our laboratory has recently reported that DHX33, a member of DExD/H-box helicase family, is a cytosolic dsRNA sensor for the NLRP3 inflammasome (27). However, the mechanism of how the cytosolic RNA induces the activation of the DHX33-NLRP3 inflammasome is unclear. In this paper, we report that TRIM33 ubiquitinates DHX33 and is essential for the cytosolic RNA-induced NLRP3 inflammasome activation. When TRIM33 is knocked down in human macrophages, the dsRNA-induced NLRP3 inflammasome activation is blocked. TRIM33 binds DHX33 directly and induces lysine 63 (K63)–specific ubiquitination of DHX33, which is essential for the formation of the DHX33–NLRP3 complex.
Materials and Methods
For reconstitution of TRIM33, TRIM33 cDNA was subcloned into pCMV vectors coding for HA- and Myc-tagged proteins (BD Clontech). Various primers were designed and used for the generation of truncations using HA-tagged full-length TRIM33 as template. All of the PCRs were carried out according to a standard procedure. HA-tagged DHX33 lysine-to-arginine mutants were obtained using a site-directed mutagenesis kit (Agilent, Life Technologies), according to the manufacturer’s manual.
HEK293T cells were maintained in DMEM medium with 10% FBS. THP-1 cells, a human acute monocytic leukemia cell line, were maintained in RPMI-1640 medium containing 10% FBS, 2 mM l-glutamine, and 50 μM 2-ME. All of the FBS was heat inactivated before use.
Differentiation and stimulation of THP-1 macrophages
As previously described (27), THP-1 cells were differentiated to macrophages with 60 nM PMA (Sigma-Aldrich) for 16 h, and cells were cultured for an additional 48h without PMA. Differentiated cells were stimulated for 8 h in 96-well plates with one of the following conditions: 5 μg/ml high m.w. (HMW) polyinosinic-polycytidylic acid (poly I:C; InvivoGen) plus Lipofectamine 2000, 5 μg/ml low m.w. (LMW) poly I:C or 1 μg/ml poly(deoxyadenylic-deoxythymidylic) (poly dA:dT; InvivoGen) plus Lipofectamine 2000, 2.5 μg/ml reoviral genomic RNA plus Lipofectamine 2000, and 2.5 μg/ml bacterial RNA plus Lipofectamine 2000, or 2 μM nigericin (InvivoGen). Culture supernatants were collected to measure cytokines IL-1β and IL-18 using ELISA and cleavage of caspase-1 using immunoblot. Cells were harvested for real-time PCR or immunoblot analysis.
Lentivirus production and infection
The short hairpin RNA (shRNA) targeting sequences for human TRIM33, DHX33, NLRP3 and caspase-1 are the following (5′ to 3′): shTRIM33-1, AACTGGAAAGTAATCAGTCGC; shTRIM33-2, ATTAGGAGTATAACCAGGAGC; shDHX33-1, CATTTCCTTTAGAACCCAAAT; shDHX33-2, GTTGACACGGGCATGGTTAAA-3; shNLRP3, GCGTTAGAAACACTTCAAGAA; and shcaspase-1, CTACAACTCAATGCAATCTTT.
The scrambled or gene-specific targeting shRNAs (Open Biosystems) in PLKO.1 vector were transfected to HEK293T cells together with packaging plasmid psPAX2 and envelope-encoding plasmid pMD2G using Lipofectamine 2000 (Life Technologies) for production of lentiviral particles. The supernatants of transfected HEK293T cells were harvested 24 and 48 h posttransfection and then were centrifuged at 1500 rpm for 15 min before infection. THP-1 cells were infected with the lentiviral particles containing supernatants in the presence of 8 μg/ml polybrene (Sigma-Aldrich), followed by centrifugation at 3500 rpm for 1 h. The medium was changed to fresh complete medium 6 h postinfection. The infected THP-1 cells were cultured for another 48 h before antibiotic selection using 5 μg/ml puromycin (InvivoGen). The knockdown efficiency of each shRNA was examined by real-time PCR and/or immunoblot analysis after 3 d of puromycin selection.
Isolation of reoviral genomic RNA
As previously described (27), Vero cells were infected with reovirus serotype 3 (VR-824; American Type Culture Collection) at a multiplicity of infection of 0.1 PFU/cell. The supernatant was collected 48 h post viral infection, centrifuged at 2500 rpm for 30 min, and then passed through 0.45 μm filters. The filtered supernatant was centrifuged at 21,000 rpm for 180 min. The viral pellet was resuspended in RLT Buffer (RNeasy kit; Qiagen), which was followed by RNA extraction according to the manufacturer’s manual.
Isolation and stimulation of human monocyte-derived macrophages
As previously described (27), buffy coats of blood samples from healthy donors were used for isolation of peripheral blood monocytes. Monocytes were isolated using RosetteSep Human Monocyte Enrichment Cocktail (StemCell Technologies) and were differentiated into macrophages in RPMI 1640 medium containing 10% FBS, 2 mM l-glutamine, and 10 ng/ml M-CSF (Life Technologies) for 10 d. Scramble or gene-specific small interfering RNAs (siRNAs; Dharmacon) were transfected into macrophages using HiPerFect Transfection Reagent (Qiagen). At 48 h posttransfection, cells were stimulated with the indicated dsRNA plus Lipofectamine 2000 for 8 h before the culture supernatants were harvested for cytokine expression analysis. For respiratory syncytial virus (RSV) infection, we first incubated the RSV (B strain; from Advanced Biotechnologies) with the macrophages at a multiplicity of infection of 1 in Opti-MEM reduced-serum medium (Life Technologies) for 2 h as described previously (27), followed by washing and continued culturing with complete culture medium for 8 h before the supernatants were collected for ELISA.
The concentration of secreted human IL-1β and IL-18 in culture supernatants was measured by ELISA kits (IL-1β, BD Biosciences; IL-18, MBL International), according to the manufacturers’ instructions.
Proteins were probed with a primary Ab as follows: polyclonal rabbit anti-TRIM33 (8972S; Cell Signaling Technology), polyclonal rabbit anti-DHX33 (sc-137424; Santa Cruz Biotechnology), monoclonal mouse anti-NLRP3 (Cryo-2) (ALX-804-881-C100; Enzo Life Sciences), monoclonal rabbit anti–caspase-1 (D7F10) (3866; Cell Signaling Technology) and monoclonal mouse anti-GAPDH (G9295; Sigma-Aldrich).
Expression and purification of recombinant proteins in HEK293T cells
PCMV-HA plasmid encoding human DHX33 was transfected to HEK293T cells. At 36 h posttransfection, cells were harvested, and the expressed HA-tagged human DHX33 protein was purified using anti-HA beads (Sigma-Aldrich). After washing, the recombinant protein was eluted from the beads using acidic elution buffer (0.1 M glycine-HCl [pH 2.5]) and then the pH was adjusted to ∼7.5 using 1 M Tris-HCl (pH 9).
Coimmunoprecipitation assays using transfected HEK293T cells
pCMV vectors encoding Myc-tagged and HA-tagged expression plasmids were cotransfected to HEK293T cells using Lipofectamine 2000. Cells were lysed 36 h after transfection in the lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1% Nonidet-P40, 5 mM EDTA, and 1 mM DTT). Cell lysates were immunoprecipitated by anti-HA beads (Sigma-Aldrich).
Endogenous immunoprecipitation assays using stimulated THP-1 macrophages
THP-1 cells were differentiated with 60 nM PMA for 16 h and then treated with 5 μg/ml HMW poly I:C plus Lipofectamine 2000 for 2 h or 2 μM nigericin for 45 min. For overexpression in THP-1 cells, HA-tagged DHX33 wild-type or K218R mutant was transfected to THP-1 cells for 24 h before stimulation. Stimulated cells were washed with 1× PBS once and then resuspended in 1× lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.3% Nonidet-P40, 1 mM EDTA, and 1 mM DTT). Cell lysates were immunoprecipitated with isotype control rabbit IgG, anti-DHX33 Ab (NB100-2581; Novus Biologicals) or anti-TRIM33 (8972S; Cell Signaling Technology) for 4 h, followed by 1 h incubation with protein G–Sepharose (Thermo Scientific).
RNAs from ∼0.2 million cells were isolated using the RNeasy Micro Kit (Qiagen) and used for synthesis of cDNA with iScript cDNA Synthesis Kit (Bio-Rad). The iTaq SYBR Green Supermix with ROX (Bio-Rad) was used for real-time PCR. The housekeeping gene GAPDH was used as internal control to normalize the amount of cDNA among different samples.
Ubiquitination assay in HEK293T overexpression system
HEK293T cells were cotransfected with expression plasmids encoding HA empty vector, HA-tagged full-length wild-type or mutant DHX33 together with Myc-tagged TRIM33, using Lipofectamine 2000. The protease inhibitor MG132 (10 μM; Sigma-Aldrich) was added 10 h before harvest. At 36 h posttransfection, cells were harvested and lysed for immunoprecipitation with anti-HA beads and then followed by immunoblotting of HA.
In vitro ubiquitination assay
An in vitro ubiquitination assay was performed as described previously (10). Briefly, 2 ng mammalian-derived HA-tagged DHX33 and Myc-tagged TRIM33 proteins were incubated with E1 (100 ng), E2 (500 ng), and Ub (2.5 μg) in 50 μl assay buffer (50 mM Tris-HCl [pH 7.5], 2.5 mM MgCl2, 0.5 mM DTT, and 2 mM ATP). Reactions were incubated at 30°C for 2 h. Then, samples were loaded to SDS-PAGE gel and analyzed by immunoblotting.
TRIM33 plays a critical role in cytosolic poly I:C–induced NLRP3 inflammasome activation in THP-1–derived macrophages
Helicase DHX33 is a critical cytosolic dsRNA sensor for the NLRP3 inflammasome (27). To further investigate how DHX33 is regulated in dsRNA-induced the NLRP3 inflammasome, we identified DHX33-interacting proteins by immunoprecipitation with Ab to DHX33 (anti-DHX33) in THP-1 cells, followed by protein sequencing by liquid chromatography–mass spectrometry. We found that the E3 Ub ligase TRIM33 was in a DHX33-interacting protein complex. We then examined whether TRIM33 also functions in NLRP3 inflammasome activation. We checked this function by stable knockdown of TRIM33 using shRNA in THP-1–derived macrophages. Two distinct TRIM33-targeting shRNAs that showed good knockdown efficiency were selected. We then stimulated those THP-1–derived macrophages with HMW poly I:C delivered by Lipofectamine 2000 and measured IL-1β and IL-18 in the culture supernatants by ELISA. Consistent with published data, knockdown of DHX33 or NLRP3 abrogated the production of IL-1β and IL-18 (Fig. 1A) as well as the cleavage of caspase-1 (Fig. 1B) that was induced by intracellular poly I:C (27–29). Interestingly, knockdown of TRIM33 also reduced the production of IL-1β and IL-18 and the cleavage of caspase-1 by THP-1–derived macrophages in response to intracellular HMW poly I:C (Fig. 1A, 1B). DHX33 is specifically involved in a cytosolic dsRNA-induced inflammasome activation pathway but does not respond to other stimuli such as nigericin. Similar to the knocking down of DHX33, knocking down of TRIM33 in THP-1–derived macrophages did not affect the production of IL-1β and IL-18 (Fig. 1C) or the cleavage of caspase-1 (Fig. 1D) in response to nigericin. Both caspase-1 and NLRP3 are dispensable for the nigericin pathway. We also examined the TRIM33 function in THP-1–derived macrophages in response to LMW poly I:C and poly dA:dT. Similar to HMW poly I:C, the LWM poly I:C–induced IL-1β production decreased dramatically in TRIM33-knockdown cells (Fig. 1E). By contrast, knockdown of TRIM33 or DHX33 did not affect poly dA:dT induced IL-1β production (Fig. 1E). The knockdown efficiency of the shRNAs targeting TRIM33, DHX33, caspase-1 and NLRP3 were shown in Fig. 1F. Taken together, these data indicate that TRIM33 is specifically involved in poly I:C–induced DHX33-dependent NLRP3 inflammasome activation but not in nigericin-induced NLRP3 inflammasome activation.
TRIM33 plays a critical role for cytosolic dsRNA-induced NLRP3 inflammasome activation in human primary monocyte-derived macrophages
We next determined the role of TRIM33 in human primary monocyte–derived macrophages (hPMDM). We used siRNA to knock down TRIM33, DHX33, and NLRP3. As shown in Fig. 2A by real-time PCR, all of the three siRNAs showed >60% knockdown of mRNA levels. Knocking down TRIM33 in hPMDM reduced the secretion of IL-1β and IL-18 in response to cytosolic HMW poly I:C, as measured by ELISA (Fig. 2B). To determine the role of TRIM33 in response to microbial RNA stimulation, we isolated bacterial total RNA and reoviral RNA. In cells with scramble siRNA, large amounts of IL-1β and IL18 were produced upon stimulation with cytosolic bacterial and reoviral RNAs (Fig. 2C, 2D). These levels of IL-1β and IL-18 were reduced dramatically in DHX33-knockdown cells or in NLRP3-knockdown cells, which confirmed a previous study showing that DHX33 and NLRP3 play critical roles in sensing microbial RNA to activate inflammasomes (27). TRIM33 knockdown resulted in about a 60–70% reduction in IL-1β and IL18 secretion in response to cytosolic bacterial RNA or reoviral RNA (Fig. 2C, 2D). To further determine the role of TRIM33 in response to viral infection, siRNA-treated hPMDMs were infected with RSV, which generates dsRNA intermediates during the replication (30). The induced IL-1β and IL-18 expression levels in the RSV-infected hPMDM supernatants were measured by ELISA. Both the IL-1β and IL-18 production decreased dramatically when TRIM33, DHX33, or NLRP3 is absent as compared with scramble siRNA knockdown. These results indicated that TRIM33 play a critical role in cytosolic dsRNA–inducted NLRP3 inflammasome activation in hPMDM cells.
TRIM33 binds to DHX33 directly and induces K63-linked ubiquitination
Because we found that TRIM33 is in a DHX33-interacting protein complex and TRIM33 is an E3 Ub ligase, we next investigated whether DHX33 is the ubiquitination target of TRIM33. HA-tagged DHX33 or Myc-tagged TRIM33 was expressed and purified from 293T cells and were shown by Coomassie Brilliant Blue staining together with commercially obtained TRIM25 and RIG-I (Fig. 3A, left panel). In vitro pulldown assay experiments show that TRIM33 binds DHX33 in a dose-dependent manner (Fig. 3A, right panel). By contrast, TRIM25, an E3 ligase regulating the sensing of RNA viruses by RIG-I, does not bind DHX33. To determine whether TRIM33 modified DHX33 directly, we performed an in vitro ubiquitination assay with TRIM33 and DHX33. In the presence of E1, E3 ligase TRIM33, Ub substrate, and ATP, DHX33 is ubiquitinated strongly by a specific E2 enzyme (Ubc H13). Two other E2 enzymes, Ubc H5c and H6, also can marginally ubiquitinate DHX33 by TRIM33 (Fig. 3B). Using K48-specific or K63-specific Ub substrates, we observed that the ubiquitination of DHX33 was mediated by TRIM33 via K63 linkage but not K48 linkage (Fig. 3C). To determine whether RIG-I is the ubiquitination target of TRIM33, we performed the ubiquitination assay with RIG-I, TRIM33, or TRIM25. As shown in Fig. 3D, RIG-I is ubiquitinated strongly by TRIM25 but not by TRIM33. These results demonstrated that TRIM33 binds to DHX33 and induces the ubiquitination of DHX33 specifically by K63 linkage.
TRIM33 binds the HA2–DUF region of DHX33 via its B-box C terminal domain
We next determined the interaction domains between TRIM33 and DHX33. We generated three truncations of HA-tagged DHX33 with a deletion of the C-terminal DUF domain (dDUF), the HA2 domain (dHA2) or the N-terminal DEXDc and HELICc domains (dHELICc) (Fig. 4A, top panel). We coexpressed the full-length DHX33 or the C terminus of DHX33 consisting of both HA2 domain and DUF domain with Myc-tagged TRIM33 and coimmunoprecipitated using anti-Myc beads. We found that either version of DHX33 could bind TRIM33 (Fig. 4A, bottom panel), indicating that DHX33 binds TRIM33 via the C-terminal HA2–DUF region. Likewise, we generated and expressed truncated forms of HA-TRIM33 in 293T cells along with Myc-DHX33. Coimmunoprecipitated experiments showed that the truncated version of TRIM33 that lacked the B-box C-terminal (BBC) domain was no longer pulled down by DHX33 (Fig. 4B), indicating that the BBC domain of TRIM33 binds DHX33.
TRIM33 ubiquitinates DHX33 at lysine 218
To determine whether TRIM33 is the main E3 ligase for the ubiquitination of DHX33, we performed an in vivo Ub assay on THP-1–derived macrophages. Upon stimulation with HMW poly I:C for 2 h, we detected the polyubiquitinated DHX33 (Ub-DHX33) in scramble knockdown cells, whereas nigericin stimulation did not generate detectable ubiquitinated DHX33 (Fig. 5A). When we knock down DHX33 in THP-1 macrophages using TRIM33-specific shRNA, the Ub-DHX33 is not detectable any more after HMW poly I:C stimulation (Fig. 5A). This result indicates that poly I:C–induced DHX33 ubiquitination in THP-1-derived macrophages is dependent on TRIM33.
To determine the residue on DHX33 that is ubiquitinated by TRIM33, we predicted the lysine sites of DHX33 that might be ubiquitinated using BDM-PUB online tool (http://bdmpub.biocuckoo.org/results.php). When the ubiquitination threshold was adjusted to 1.35, the tool predicted 11 putative ubiquitination target sites (Supplemental Table I). We therefore replaced each of the eleven lysine residues in DHX33 individually with arginine. We coexpressed the HA-tagged wild-type or mutations of DHX33 together with Myc-TRIM33 in HEK293T cells and stimulated the cells with HMW poly I:C for 2 h. Using the approach of an immunoprecipitation assay with anti-HA beads and followed by Ub immunoblotting, we were able to demonstrate that the K218R substitution blocked ubiquitination of DHX33 (Fig. 5B). These data indicated that TRIM33 is the main E3 ligase to induce the ubiquitination of DHX33, and this occurred on lysine 218.
TRIM33 is essential for DHX33-NLRP3 inflammasome formation upon dsRNA stimulation
Because DHX33 forms a complex with NLRP3 in response to cytosolic dsRNA (27), we next examined whether TRIM33 plays a role in the DHX33-NLRP3 complex formation. THP-1–derived macrophages with scramble and TRIM33-specific (shTRIM33) stable knockdown were stimulated with HMW poly I:C or nigericin for 2 h before harvesting and cell lysis. Immunoprecipitation was performed using anti-DHX33 followed by immunoblotting for DHX33, TRIM33, and NLRP3. As shown in Fig. 6A, both TRIM33 and NLRP3 were present with DHX33 only when stimulated with HMW poly I:C in scramble knockdown macrophages but not in nigericin-stimulated cells. Furthermore, when TRIM33 is absent by stable knockdown using shTRIM33, the interaction of NLRP3 with DHX33 is no longer detectable after HMW poly I:C stimulation; neither was it detectable after nigericin stimulation, which indicates that TRIM33 is essential and specific for poly I:C–induced DHX33-NLRP3 complex formation.
To further determine the ubiquitination effect on DHX33-NLRP3 complex formation, we overexpressed HA-tagged DHX33 wild-type and the lysine 218 mutant (K218R) in THP-1 macrophages. The macrophages were stimulated with HMW poly I:C or nigericin, followed by HA immunoprecipitation. As shown in Fig. 6B, both NLRP3 and TRIM33 were pulled down by wild-type DHX33. However, the DHX33 K218R mutant, which abolished the DHX33 ubiquitination by TRIM33, only pulled down TRIM33 but not NLRP3. Neither the wild-type nor the K218R mutant DHX33 can pull down TRIM33 or NLRP3 after nigericin stimulation (Fig. 6B). These data indicate that the ubiquitination of DHX33 at lysine 218 by TRIM33 is required to form the DHX33–NLRP3 complex and thus is also required for NLRP3 inflammasome activation.
In this study, we have presented data showing that the E3 Ub ligase TRIM33 plays a critical role in the cytosolic dsRNA–induced NLRP3 inflammasome activation by targeting the dsRNA sensor DHX33. Knockdown of TRIM33 in THP-1–derived or hPMDMs results in reduced NLRP3 inflammasome activation in response to cytosolic dsRNA. TRIM33 binds DHX33 in macrophages and induces the K63-linked polyubiquitination of DHX33 upon cytosolic dsRNA stimulation. TRIM33 ubiquitinates DHX33 at lysine 218. DHX33 interacts with TRIM33 via the HA2-DUF region and the BBC domain. This interaction is essential for the activation and formation of the DHX33–NLRP3 inflammasome complex upon dsRNA stimulation. Various stimuli have been shown to activate the NLRP3 inflammasome, including the bacterial toxin nigericin, LPS plus ATP treatment, and cytosolic nucleic acids (26). DHX33 has been identified as the cytosolic dsRNA sensor that induces NLRP3 inflammasome activation.TRIM33 is the main E3 ligase of DHX33, specifically regulating the formation of the cytosolic dsRNA–induced NLRP3 inflammasome.
Most of the TRIM family members are E3 Ub ligases, and they have been reported to mediate both K48- and K63-linked ubiquitination. Two well-studied examples of TRIM protein ubiquitination of helicase family members are the reports that TRIM25 ubiquitinates RIG-I by K63-specific linkage and TRIM21 ubiquitinates DDX41 by K48-specific linkage (8, 10). Some other TRIM members that are important in regulating the innate immune response also have been reported. TRIM56 was reported to ubiquitinate and activate the STING molecule, which is a key adaptor for DNA sensors (12). TRIM38 was reported to target TRAF6 for degradation via K48-linked ubiquitination (31).
As important intracellular antimicrobial machinery (32), the inflammasome needs precise regulation of its activation. Our results showed that TRIM33-induced DHX33 ubiquitination is essential for the DHX33–NLRP3 inflammasome formation and function, as shown by ELISA detection of IL-1β and IL-18 production as well as our endogenous immunoprecipitation experiments. In contrast to cytosolic dsRNA, TRIM33-DHX33 is not involved in the NLRP3 activation that is induced by nigericin, which is believed to use a different activation pathway (33).
It has been reported that TRIM33 is involved in erythroid differentiation, tumor suppression, and the autoimmune disease dermatomyositis (34–37). Our study found another important function of TRIM33 in regulating NLRP3 activation in response to bacterial and viral infection. This finding will further shed light on the clinical application of TRIM33 for various diseases including microbial infections.
We thank Dr. Carson Harrod at Baylor Institute for Immunology Research for critical reading of the manuscript. We also thank all of the colleagues in our laboratory for helpful discussions.
This work was supported by National Institutes of Health Grant R37 AI091947.
The online version of this article contains supplemental material.
Abbreviations used in this article:
human primary monocyte–derived macrophage
- poly dA:dT
- poly I:C:
retinoic acid–inducible gene-I
respiratory syncytial virus
small interfering RNA
stimulator of IFN gene
The authors have no financial conflicts of interest.