USP38 Inhibits Type I Interferon Signaling by Editing ... - Cell Press

Article

USP38 Inhibits Type I Interferon Signaling by Editing TBK1 Ubiquitination through NLRP4 Signalosome Graphical Abstract

Authors Meng Lin, Zhiyao Zhao, Zhifen Yang, ..., Yunfei Qin, Rong-Fu Wang, Jun Cui

Correspondence [email protected] (R.-F.W.), [email protected] (J.C.)

In Brief Lin et al. use overexpression, RNAi, and knockout systems to demonstrate that USP38 targets TBK1 for degradation through editing the ubiquitination of TBK1. The authors identify NLRP4USP38-DTX4/TRIP as an essential signalosome to control TBK1 stability as well as antiviral innate immunity.

Highlights d

USP38 negatively regulates type I interferon signaling in vitro and in vivo

d

USP38 interacts with the active form of TBK1 through NLRP4

d

USP38 cleaves K33-linked poly-ubiquitin chains of TBK1

d

NLRP4 signalosome mediates K48-linked ubiquitination of TBK1 for degradation

Lin et al., 2016, Molecular Cell 64, 267–281 October 20, 2016 ª 2016 Elsevier Inc. http://dx.doi.org/10.1016/j.molcel.2016.08.029

Molecular Cell

Article USP38 Inhibits Type I Interferon Signaling by Editing TBK1 Ubiquitination through NLRP4 Signalosome Meng Lin,1,4,7 Zhiyao Zhao,1,7 Zhifen Yang,3,7 Qingcai Meng,1 Peng Tan,4,6 Weihong Xie,3 Yunfei Qin,1 Rong-Fu Wang,4,5,6,8,* and Jun Cui1,2,* 1Key

Laboratory of Gene Engineering of the Ministry of Education, State Key Laboratory of Biocontrol, School of Life Sciences Innovation Center of Cancer Medicine 3Zhongshan School of Medicine Sun Yat-sen University, Guangzhou, China 4Center for Inflammation and Epigenetics, Houston Methodist Research Institute, Houston, TX 77030, USA 5Department of Microbiology and Immunology, Weill Cornell Medical College, Cornell University, New York, NY 10065, USA 6Institute of Biosciences and Technology, College of Medicine, Texas A & M University, Houston, TX 77030, USA 7Co-first author 8Lead Contact *Correspondence: [email protected] (R.-F.W.), [email protected] (J.C.) http://dx.doi.org/10.1016/j.molcel.2016.08.029 2Collaborative

SUMMARY

TBK1 is a component of the type I interferon (IFN) signaling pathway, yet the mechanisms controlling its activity and degradation remain poorly understood. Here we report that USP38 negatively regulates type I IFN signaling by targeting the active form of TBK1 for degradation in vitro and in vivo. USP38 specifically cleaves K33-linked poly-ubiquitin chains from TBK1 at Lys670, and it allows for subsequent K48-linked ubiquitination at the same position mediated by DTX4 and TRIP. Knockdown or knockout of USP38 increases K33-linked ubiquitination, but it abrogates K48-linked ubiquitination and degradation of TBK1, thus enhancing type I IFN signaling. Our findings identify an essential role for USP38 in negatively regulating type I IFN signaling, and they provide insights into the mechanisms by which USP38 regulates TBK1 ubiquitination through the NLRP4 signalosome.

INTRODUCTION The innate immune system, eliciting robust antiviral responses through the detection of pathogen-associated molecular patterns (PAMPs), serves as the first line of defense against viral invasion (Takeuchi and Akira, 2010). Viral components are primarily recognized by several classes of pattern-recognition receptors (PRRs), including Toll-like receptors (TLRs), RIG-I-like receptors (RLRs), Nod-like receptors (NLRs), and DNA sensors (Goubau et al., 2013; Gu¨rtler and Bowie, 2013; Paludan and Bowie, 2013; Takeuchi and Akira, 2010). Several TLRs, including TLR3 and TLR7/9, initiate type I interferon (IFN) production upon

recognition of viral double-stranded RNA (dsRNA) or singlestranded RNA (ssRNA) in endosomes (Goubau et al., 2013), whereas Nod2 and the RNA helicases RIG-I, MDA5, and LGP2 sense viral dsRNA or ssRNA in the cytosol (Goubau et al., 2014; Kato et al., 2008; Loo and Gale, 2011). Recently, several DNA sensors, including cyclic GMP-AMP synthase (cGAS), RNA polymerase III, IFI16 (p204), DDX41, and AIM2, have been identified for detecting cytosolic DNA of pathogens (Ablasser et al., 2009; Chiu et al., 2009; Ferguson et al., 2012; Kondo et al., 2013; Paludan and Bowie, 2013; Sun et al., 2013; Unterholzner et al., 2010). Upon recognition of viral RNA and DNA, these PRRs trigger the activation of NF-kB, IRF3/7, and inflammasome signaling pathways through distinct adaptor proteins, such as TRIF, MAVS, and STING, leading to the production of type I IFNs and pro-inflammatory cytokines and the induction of subsequent adaptive immune responses (Takeuchi and Akira, 2010). Although plasmacytoid dendritic cells (pDCs) predominantly use MyD88-IRF7-dependent type I IFN signaling after CpG stimulation or viral infection, most immune cells use TBK1-IRF3-dependent type I IFN signaling after viral infection (Theofilopoulos et al., 2005). TBK1 is an essential kinase for IRF3 activation and IFN-b production in type I IFN signaling, and its activity is tightly regulated by multiple posttranslational modifications (PTMs). It has been reported that SHIP1, PPM1B, and glucocorticoids can modulate TBK1 activity by modifying its phosphorylation state (Gabhann et al., 2010; McCoy et al., 2008; Zhao et al., 2012). GSK3b was shown to enhance TBK1 self-association and auto-phosphorylation at Ser172 after viral infection (Lei et al., 2010). In addition to phosphorylation, TBK1 also undergoes robust ubiquitination during viral infection. The E3 ligases Mind bomb (MIB) and Nrdp1 positively regulate K63-linked ubiquitination of TBK1, promoting type I IFN production in response to RLR or TLR ligands, respectively (Li et al., 2011; Wang et al., 2009). Recently, we and others found that NLRP4/DTX4 and another E3 ligase TRAF-interacting protein (TRIP) induce the K48-linked

Molecular Cell 64, 267–281, October 20, 2016 ª 2016 Elsevier Inc. 267

Figure 1. USP38 Negatively Regulates Type I IFN Signaling as well as Antiviral Responses (A and B) Luciferase activity in 293T cells or 293T-TLR3 cells, transfected with a luciferase reporter for IFN-b (A) (IFN-b-luc) or for ISRE (B) (ISRE-luc), together with empty vector (EV) (no wedge) or an increasing dose of USP38, followed by treatment with or without intracellular (IC) low-molecular-weight poly(I:C) (pIC) (legend continued on next page)

268 Molecular Cell 64, 267–281, October 20, 2016

poly-ubiquitination of TBK1, leading to its proteasomal degradation (Cui et al., 2012; Zhang et al., 2012). Although several E3 ligases have been reported to regulate TBK1 activity through different types of ubiquitination, the functions of deubiquitinating enzymes on TBK1 activation remain to be elucidated. In this paper, we report the identification of ubiquitin-specific protease 38 (USP38) as a negative regulator of type I IFN signaling. USP38 strongly induced the degradation of the active form of TBK1 after viral infection. USP38 deficiency abrogated the degradation of TBK1 and increased the production of type I IFN. Interestingly, we found that all known TBK1 stability regulators, including USP38, DTX4, and TRIP, interacted with TBK1 in an NLRP4-dependent manner. NLRP4 recruited USP38 and the E3 ligases DTX4/TRIP to TBK1 at different stages after viral infection, and it allowed USP38 to specifically cleave K33-linked poly-ubiquitin chains from TBK1 at Lys670, increasing the availability of Lys670 for subsequent K48-linked poly-ubiquitination, mediated by DTX4 and TRIP. Our findings identified the NLRP4-USP38-DTX4/TRIP plexus as a signalosome, controlling TBK1 stability and maintaining immune homeostasis during antiviral innate immunity. RESULTS Identification of USP38 as an Essential Negative Regulator of Type I IFN Signaling To identify the role of deubiquitinases (DUBs) in antiviral immunity, we screened 81 genes encoding DUBs (Cui et al., 2014) and identified USP38 as a negative regulator of type I IFN signaling. We found that USP38 potently inhibited the activation of IFN-b by treatments of intracellular (IC) low-molecular-weight (LMW) poly(I:C) (a ligand for RIG-I), IC high-molecular-weight (HMW) poly(I:C) (a ligand for MDA5), poly(dA:dT), or vesicular stomatitis virus (VSV) (Figure 1A). As IFN-b activation requires cooperative activation of IRF3 and NF-kB, we used an IFNstimulated response element (ISRE) luciferase reporter to test whether USP38 directly affects IRF3 signaling, and we found that USP38 similarly inhibited ISRE-luc activity (Figure 1B). We next assessed phosphorylation of IRF3 in 293T cells, expressing USP38 together with RIG-I(N), MDA5, MAVS, STING-cGAS, or TRIF, and we found that USP38 inhibited the activation of IRF3 induced by all these innate immune receptors and adaptors (Figure S1A). Further qPCR analysis revealed that the mRNA levels of IFNa4, IFNb, IFIT1, IFIT2, and ISG15 in USP38 ectopic expression cells were markedly reduced during VSV infection

(Figure S1B), suggesting that USP38 inhibits the expression of IFN-stimulated genes (ISGs). To determine the role of USP38 under physiological conditions, we efficiently knocked down USP38 in 293T cells, THP-1 cells, or human peripheral blood mononuclear cells (PBMCs) (Figure S1C); luciferase assay analysis showed that USP38 knockdown could enhance the IC poly(I:C)-, poly(dA:dT)-, or VSV-EGFP-induced ISRE activation (Figure 1C). Consistent with this result, we found that USP38 knockdown markedly enhanced the phosphorylation of IRF3 upon these treatments (Figure 1D). To substantiate these findings, we knocked down USP38 in THP-1 cells, 293T cells, A549 cells, or PBMCs, and we found that USP38 knockdown significantly increased the IFN-a4 and IFN-b expression, as well as IFN-b secretion induced by VSV-EGFP or HSV-1 infection (Figures 1E–1G and S1D–S1F). Further we showed that USP38 knockdown in these cells rendered them resistant to VSV-EGFP infection (Figures 1H, 1I, S1G, and S1H). Collectively, these results suggest that USP38 negatively regulates type I IFN signaling as well as antiviral immunity in various human cell types. USP38 Deficiency Enhances Antiviral Immunity In Vivo To determine the function of USP38 in primary cells, we generated USP38-deficient (Usp38/) mice using a TAL effector nuclease (TALEN)-based system (Figure S2A), and we prepared bone marrow-derived macrophages (BMMs) from wild-type (WT) and Usp38/ mice (Figure S2B). Next we infected those BMMs for 16 or 24 hr with the RNA virus VSV or the DNA virus HSV-1. Usp38/ BMMs produced 2- to 4-fold more IFN-b in response to VSV or HSV-1 than WT BMMs (Figures 2A and 2B). Consistent with these results, deletion of USP38 had a strong effect on the expression of many ISGs, including IFIT1, IFIT3, ISG15, and CXCL10, as well as the pro-inflammatory cytokines TNF-a and IL-6, induced by VSV or HSV-1 (Figures 2C, 2D, and S2C). Next we evaluated the importance of USP38 in facilitating effective host defense against viral infection. We detected significantly less VSV or HSV-1 in Usp38/ BMMs than in WT BMMs (Figure 2E). These data indicate a negative role for USP38 in the sensing of both RNA and DNA viruses in mouse BMMs. To investigate the functional significance of USP38 in host antiviral response in vivo, we challenged Usp38/ mice with VSV (1 3 108 plaque-forming units [PFU]/g), and we found that Usp38/ mice produced higher levels of IFN-b, TNF-a, and IL-6 than WT mice in response to VSV infection (Figure 2F).

(5 mg/mL) (IC pIC L), IC high-molecular-weight poly(I:C) (5 mg/mL) (IC pIC H), poly(dA:dT) (pdA:dT) (5 mg/mL), poly(I:C) (pIC) (10 mg/mL), or VSV-EGFP (MOI, 0.01) (VSV). Results are expressed relative to renilla luciferase activity. (C) Luciferase activity in 293T cells transfected with control (Ctrl) siRNA or USP38-specific siRNAs, together with an ISRE-luc, then left untreated (UT) or treated with IC poly(I:C), poly(dA:dT), or VSV-eGFP is shown. (D) Immunoblot analysis of total and phosphorylated (p-) IRF3 in THP-1 cells transfected with Ctrl siRNA or USP38-specific siRNA, followed by treatment with IC poly(I:C) or infection with VSV-EGFP at different time points is shown. (E–G) Real-time PCR analysis of IFNa4 and IFNb (E) and ELISA of IFN-b protein (F) and IFIT1, IFIT2, and CCL5 mRNA in THP-1 cells and human peripheral blood mononuclear cells (PBMCs) (G) treated with Ctrl siRNA or USP38-specific siRNAs, followed by infection with VSV-eGFP or IC poly(I:C) and analyzed at the indicated time points, are shown. (H and I) Phase-contrast (PH) and fluorescence microscopy (H) and GFP intensity analyses (I) of A549 cells transfected with Ctrl siRNA or USP38-1 siRNA, and then infected with VSV-EGFP at the indicated MOI. Data in (A)–(C), (E)–(G), and (I) are expressed as means ± SEM of three independent experiments (*p < 0.05, **p < 0.01, and ***p < 0.001, versus cells with the same treatment in WT cells, Student’s t test). See also Figure S1.

Molecular Cell 64, 267–281, October 20, 2016 269

Figure 2. USP38 Deficiency Enhances Type I IFN Signaling as well as Antiviral Responses In Vivo (A and B) Real-time PCR analysis of IFNb (A) and ELISA of IFN- b protein (B) in BMMs from WT or Usp38/ mice after VSV-EGFP or HSV-1 infection, at the indicated time points, is shown. (C and D) Real-time PCR analysis of IFIT1, IFIT3, ISG15, and CXCL10 mRNA in BMMs from WT or Usp38/ mice after VSV (C) or HSV-1 (D) infection is shown. (E) Virus titers of WT or Usp38/ BMMs after VSV or HSV-1 infection are shown. (F) ELISA of cytokine production in sera from WT and Usp38/ mice (n = 5 per group), which were intravenously injected with VSV, is shown. (G) Determination of VSV loads in organs by TCID50 assay from WT and Usp38/ mice is shown. (H) Survival of 7-week-old WT and Usp38/ mice given intravenously injection of VSV (1 3 108 PFU/g) (n = 6 per group; p < 0.01). Data in (A)–(H) are expressed as means ± SEM of three independent experiments (*p < 0.05, **p < 0.01, and ***p < 0.001, versus cells with the same treatment in WT controls, Student’s t test). See also Figure S2.

Furthermore, we determined VSV loads in lung and liver homogenate supernatants, and we found that VSV loads in tissue samples from infected Usp38/ mice were significantly


lower than those from WT mice 24 hr post-infection (Figure 2G). Importantly, Usp38/ mice were significantly more resistant to VSV infection in overall survival compared with

Figure 3. USP38 Interacts with TBK1 after Viral Infection (A) Luciferase activity of 293T cells transfected with an ISRE- or IFN-b-luc, together with RIG-I(N), MDA5, MAVS, TBK1, IKKi, or IRF3, along with EV (no wedge) or with increasing amounts (wedge) of USP38, is shown. (B) Co-immunoprecipitation and immunoblot analysis of 293T cells transfected with Myc-USP38 together with FLAG-tagged MAVS, TBK1, IKKi, IRF3, or IRF7 are shown. (C) Confocal microscopy of USP38 and TBK1 in 293T cells transfected with GFP-USP38 and DsRedTBK1 is shown. DAPI, DNA-intercalating dye. (D–F) Co-immunoprecipitation and immunoblot analysis of 293T, THP-1, and PBMCs infected with VSV-EGFP for various times (above lanes) are shown. (G) Co-immunoprecipitation and immunoblot analysis of 293T cells transfected with HA-TBK1 and deletion mutants of FLAG-USP38 (FL, full-length; N, N-terminal without any conversed domain; C, C-terminal with UCH domain). Numbers above indicate amino acid position. Data from (A) are plotted as means ± SEM. Similar results were obtained in three independent experiments (*p < 0.05, **p < 0.01, and ***p < 0.001, versus controls).

WT mice (Figure 2H). These results suggest that Usp38/ mice develop more potent antiviral responses against VSV infection than WT mice. USP38 Interacts with TBK1 after Viral Infection To determine the molecular targets of USP38 in type I IFN signaling pathway, we co-transfected 293T cells with RIG-I, MDA5, MAVS, TBK1, IKKi, or IRF3 together with increasing amounts of USP38 plus the IFN-b or ISRE luciferase reporter, and we found that USP38 inhibited activation of both luciferase reporters induced by RIG-I, MDA5, MAVS, TBK1, but not IKKi or IRF3 (Figure 3A). Co-immunoprecipitation and immunoblot analyses revealed that USP38 specifically interacted with TBK1, but not MAVS, IKKi, IRF3, or IRF7 (Figure 3B). In addition, confocal microscopic analysis using 293T cells transfected with USP38-GFP and TBK1-dsRFP showed that USP38 co-localized with TBK1 in the cytosol (Figure 3C).

To examine the interaction between USP38 and TBK1 under physiological conditions, we infected 293T cells, THP-1 cells, or PBMCs with VSV-EGFP, and we found that the interaction between USP38 and TBK1 was barely detectable in unstimulated cells (Figures 3D–3F). After viral infection, the interaction between USP38 and TBK1 was significantly increased in all three cell types (Figures 3D–3F). To further determine which domain of USP38 is responsible for its interaction with TBK1, we constructed several deletion constructs of USP38. Co-immunoprecipitation and immunoblot analyses revealed that the N-terminal USP domain of USP38 interacted with TBK1, but its C-terminal domain showed no interaction with TBK1 (Figure 3G). These results suggest that USP38 inhibits type I IFN signaling by interacting with TBK1 after viral infection. USP38 Specifically Degrades Active TBK1 Next we sought to determine the molecular mechanisms of how USP38 inhibits type I IFN signaling activation through its interaction with TBK1. In 293T cells transfected with FLAG-tagged TBK1 and HA-tagged IRF3, together with increasing doses of USP38, we found that the concentration of TBK1 protein decreased considerably with increasing USP38 expression. However, the mRNA level of TBK1 remained unchanged (Figure 4A), suggesting that USP38 causes TBK1 protein degradation. To determine the specificity of the USP38-mediated TBK1 degradation, we used other USP family members USP3 and USP13 as a control, and we found that USP38, but not USP3


Figure 4. USP38 Mediates the Degradation of Activated TBK1 (A) Immunoblot analysis (top) of extracts of 293T cells transfected with FLAG-TBK1 and HA-IRF3 and increasing doses of USP38 (wedge). RT-PCR analysis (bottom) of TBK1 mRNA is shown; GAPDH mRNA (encoding glyceraldehyde phosphate dehydrogenase) serves as a loading control. (B) 293T cells were transfected with FLAG-tagged TBK1, IKKi, IKKa, IKKb, and increasing doses of USP38 (wedge). (C) Immunoassay of extracts of 293T cells transfected with FLAG-TBK1 and Myc-USP38 and treated with MG132, DMSO, or 3-Methyladenine (3-MA) is shown. (D and E) Immunoblot analysis (D) and luciferase activity (E) of 293T cells transfected with FLAG-TBK1 and Myc-USP38, as well as USP38-specific or control siRNAs, together with an ISRE luciferase reporter, are shown. (F and G) 293T cells were transfected with Myc-USP38 together with FLAG-TBK1 (WT), FLAG-TBK1 (S172A) or (S172E). After immunoprecipitation with antiFLAG beads, Myc-USP38 was analyzed by immunoblot with anti-Myc (F). TBK1, TBK1-S172A, and their phosphorylation levels were determined by immunoblot (G). SA, S172A; SE, S172E. (H) Immunoblot analysis of extracts of 293T cells transfected with EV or Myc-USP38, followed by infection with VSV-EGFP for the indicated time points, is shown. (I) Immunoblot analysis of extracts of THP-1 cells transfected with Ctrl siRNA or USP38-specific siRNA, followed by infection with VSV-EGFP for the indicated time points with the indicated antibodies, is shown. (J) Immunoblot analysis of extracts of mouse bone marrow macrophages (BMMs) from WT or Usp38/ mice, after VSV or HSV-1 infection for the indicated time points with the indicated antibodies. Data in (E) are expressed as means ± SEM of three independent experiments (*p < 0.05, **p < 0.01, and ***p < 0.001, versus cells with the same treatment in control cells, Student’s t test). See also Figure S3.


Figure 5. USP38 Mediates the K33-K48 Ubiquitination Transition of TBK1 on Lys 670 (A) Co-immunoprecipitation and immunoblot analysis of extracts of 293T cells transfected with various combinations of plasmid encoding FLAG-tagged TBK1, Myc-tagged USP38, and HA-tagged K48-linked or K63-linked ubiquitin, and treated with MG132, are shown. (legend continued on next page)


or USP13, induced TBK1 degradation (Figure S3A). To determine whether USP38 causes the degradation of other proteins, we performed a similar experiment and found that USP38 specifically degraded TBK1, but not IKKi, IKK-a, or IKK-b (Figure 4B). To determine whether USP38 degrades TBK1 through a ubiquitin-protease pathway, we performed experiments in the presence of the proteasome inhibitor MG-132 or 3-Methyladenine (3-MA), and we found that MG-132, but not 3-MA, blocked the degradation of TBK1 (Figure 4C), suggesting that USP38induced TBK1 degradation is dependent on a proteasome pathway. Moreover, USP38 knockdown blocked the USP38mediated degradation of TBK1 and restored ISRE-luc activity (Figures 4D and 4E). Murine USP38 (mUSP38) showed similar inhibitory function of human USP38 protein (Figure S3B), suggesting a conserved function of USP38 between mice and humans. To determine whether USP38 can mediate degradation of endogenous TBK1 under physiological conditions, we transfected 293T cells with USP38, and we found that endogenous TBK1 protein was unchanged (Figure S3C). We reasoned that USP38 may specifically target the activated form of TBK1 for degradation. Therefore, we generated an inactive mutant of TBK1 with a substitution of alanine for the serine at position 172 (S172A), which abrogates its auto-phosphorylation, and then we co-transfected the 293T cells with USP38, WT TBK1, or TBK1 (S172A). Indeed, we found that USP38 could not interact with mutant TBK1 (S172A) and failed to promote it for degradation (Figures 4F and 4G); however, TBK1 constitutive active mutant (S172E) showed an increasing binding ability to USP38 compared to WT TBK1 (Figure 4F). Furthermore, we found that USP38 expression reduced TBK1 protein, but not TBK1 mRNA, after VSV infection compared with 293T cells transfected with the empty vector (EV) (Figures 4H and S3D). Conversely, USP38 knockdown in THP-1 cells increased endogenous TBK1 protein levels in cells infected with VSVEGFP or transfected with IC poly(I:C), but not in uninfected cells (Figures 4I and S3E). Similar results were obtained with WT or Usp38/ BMMs with VSV or HSV-1 infection (Figure 4J). These

results suggest that USP38 specifically degrades the active form (p-TBK1) of TBK1 after viral infection. USP38 Mediates the K33-K48 Ubiquitination Transition of TBK1 on Lys 670 It has been documented that TBK1 undergoes both K48-linked and K63-linked ubiquitination after viral infection (Cui et al., 2012; Li et al., 2011; Wang et al., 2009; Zhang et al., 2012, 2014b). To determine how USP38 regulates TBK1 ubiquitination, we transfected 293T cells with FLAG-TBK1, Myc-USP38, HAtagged K48-linked ubiquitin (K48-Ub), or K63-linked ubiquitin (K63-Ub), and we found that USP38 markedly increased K48linked, but not K63-linked, ubiquitination of TBK1 (Figure 5A). Consistent with these results, USP38 knockdown remarkably attenuated K48-linked ubiquitination of TBK1 (Figure 5B). Since the K48-linked ubiquitination of TBK1 is regulated by NLRP4, DTX4, or TRIP proteins (Cui et al., 2012; Zhang et al., 2012), we first tested whether USP38 promotes TBK1 degradation by stabilizing these regulators. However, we did not observe any apparent differences in the protein abundance of NLRP4, DTX4, or TRIP in the presence or absence of USP38 after viral infection (Figures S4A–S4C), suggesting that USP38 regulates K48-linked ubiquitination of TBK1 via a distinct mechanism. Since USP38 is a DUB, we investigated how USP38 increases rather than decreases TBK1 K48-linked ubiquitination via its deubiquitinating activity. To this end, we generated DUB-inactive mutants of USP38 C545A, H857A, and double-point mutation USP38 (C545A/H857A). Interestingly, we found that the USP38 (C545A/H857A) mutant failed to degrade TBK1, indicating that the deubiquitinating activity of USP38 is required to degrade TBK1 (Figure 5C). Next we sought to determine what kind of ubiquitin chain is removed by USP38. We performed experiments using 293T cells expressing FLAG-TBK1, Myc-USP38, and different types of HAtagged ubiquitin, and we found that USP38 specifically removed K33-linked ubiquitination (Figures S4D–S4F). Consistent with our observation, the USP38 (C545A/H857A) mutant failed to remove K33-linked ubiquitination of TBK1, indicating that the

(B) Immunoassay of extracts of 293T cells transfected with FLAG-tagged TBK1 and HA-tagged K48-linked ubiquitin, together with control siRNA or USP38specific siRNA, assessed as in (A), is shown. (C) Immunoblot analysis of 293T cells transfected with FLAG-TBK1 and Myc-USP38 (WT), Myc-USP38 (CA), Myc-USP38 (HA), or Myc-USP18 (CA/HA) is shown. (D) Co-immunoprecipitation and immunoblot analysis of extracts of 293T cells transfected with various combinations of plasmid for FLAG-tagged TBK1, HA-tagged K33-linked ubiquitin, Myc-tagged USP38, or USP38 (CA/HA), and treated with MG132, are shown. (E) Co-immunoprecipitation and immunoblot analysis of extracts of 293T cells transfected with various combinations of plasmid encoding Myc-tagged USP38, and HA-tagged K33-linked or HA-tagged K48-linked ubiquitin, then infected with VSV-EGFP for the indicated time points, are shown. (F) Mass spectrometry analysis of a peptide derived from ubiquitinated TBK1 shows ubiquitin conjugation at K33 and K48 residues of ubiquitin. (G) Ratio of K33 and K48 ubiquitin linkage in TBK1 immunoprecipitates, in WT and USP38/ cells after VSV infection, is shown. (H) Di-K33-linked ubiquitin (left) or HA-K33-linked ubiquitinated TBK1 (right) was incubated with immuno-purified FLAG-USP38 in vitro in deubiquitinating buffer. The immunoblot was probed with anti-HA. (I) Confocal microscopy analysis of USP38 and K33-linked ubiquitin chains in 293T cells transfected with GFP-USP38 and DsRed-K33-linked ubiquitin by HSV-1 infection or UT is shown. (J) Immunoblot analysis of extracts of 293T cells transfected with EV or Myc-USP38, together with FLAG-tagged WT TBK, K661R mutant, or K670R mutant of TBK1, is shown. (K) Immunoprecipitation and immunoblot analysis of 293T cells transfected with FLAG-tagged WT TBK1 or K670R mutant of TBK1 and HA-tagged K33-linked or HA-tagged K48-linked ubiquitin are shown. (L) Immunoprecipitation and immunoblot analysis of 293T cells transfected with various combinations (above lanes) of vectors encoding Myc-tagged USP38, HA-tagged K33-linked or K0-linked ubiquitin, and FLAG-tagged WT TBK1 or the K670R mutant of TBK1, and treated with MG132, are shown. (M) Ratio of K33 and K48 ubiquitin linkage in WT FLAG-TBK1 and FLAG-TBK1(K670R) immunoprecipitates by mass spectrometry analysis. See also Figure S4.


Figure 6. USP38 Promotes TBK1 Degradation in an NLRP4-Dependent Manner (A) Co-immunoprecipitation and immunoblot analysis of 293T cells transfected with HA-USP38 together with FLAG-tagged NLRP4 or NLRP3 are shown. (B and C) Immunoassay of extracts of THP-1 cells (B) and PBMCs (C) infected with VSV-EGFP for the indicated time points, followed by immunoprecipitation with anti-NLRP4 and immunoblot analysis with anti-USP38, is shown. (D) Co-immunoprecipitation and immunoblot analysis of 293T cells transfected with Myc-USP38 and deletion mutants of FLAG-NLRP4 are shown. (E) Co-immunoprecipitation and immunoblot analysis of 293T cells transfected with HA-NLRP4 and deletion mutants of FLAG-USP38 are shown. (legend continued on next page)


removal of K33-linked poly-ubiquitin chains on TBK1 depends on the deubiquitinating activity of USP38 (Figure 5D). We further showed that TBK1 underwent both K48- and K33-linked ubiquitination after viral infection (Figure 5E). However, we unexpectedly found that overexpression of USP38 decreased K33-linked ubiquitination of TBK1, but it increased K48-linked ubiquitination of TBK1 upon VSV infection (Figure 5E), suggesting that USP38 may remove K33-linked ubiquitin chains and add K48-linked ubiquitin chains to TBK1 after viral infection. To demonstrate the cleavage of K33-linked ubiquitin chains by USP38 under physiological condition, we performed mass spectrometry analysis of ubiquitinated TBK1 from 293T cells or USP38/ 293T cells by VSV infection, using a similar strategy as previously described (Yuan et al., 2014). We found that ubiquitinated TBK1 with K33 or K48 linkage was hardly detected without viral infection. By contrast, after VSV infection, the increased K33linked ubiquitinated TBK1 as well as decreased K48-linked ubiquitinated TBK1 in USP38/ cells were readily detected by mass spectrometry analysis, compared to WT cells, suggesting that USP38 is responsible for the cleavage of K33-linked ubiquitin chains of TBK1 (Figures 5F and 5G). To directly demonstrate the cleavage of K33-linked ubiquitin chains by USP38, we incubated purified USP38 and commercially available Di-K33 ubiquitin chain (K33-Ub)2 together and found that (K33-Ub)2 can be cleaved by USP38 in a dose-dependent manner (Figure 5H). We next performed in vitro deubiquitination assay using purified USP38 and K33-linked ubiquitinated TBK1, and we found that USP38 effectively removed K33-linked ubiquitin chains from the purified TBK1 (Figure 5H). To further support our findings, we performed confocal microscopic analysis and found that USP38 was colocalized with K33-linked ubiquitin chains inside the cells, while the co-localization between USP38 and K33-linked ubiquitin chains was enhanced after HSV-1 infection (Figure 5I). Taken together, these results clearly suggest that USP38 is responsible for cleaving K33-linked ubiquitin chains from TBK1. We previously reported that NLRP4/DTX4 induced the K48linked poly-ubiquitination of TBK1 at Lys670, leading to its proteasomal degradation (Cui et al., 2012). We further showed that neither USP38 nor TRIP can degrade TBK1 (K670R) mutants (Figures 5J and S4G), suggesting that Lys670 on TBK1 is critical for controlling TBK1 stability. Interestingly, we unexpectedly found that the TBK1 K670R mutant abrogates not only K48linked ubiquitination but also K33-linked ubiquitination of TBK1 (Figures 5K and S4H). These results suggest that K33- and K48-linked ubiquitin chains may conjugate to TBK1 at the same amino acid position. We further showed that USP38 could remove K33-linked ubiquitination from WT TBK1, but not from the TBK1 K670R mutant (Figure 5L). In addition, we pulled down the WT TBK1 and TBK1(K670R) mutant and checked the ratio of K33 and K48 poly-ubiquitin chains in the immunoprecipitates by mass spectrometry analysis, and we found that there was no detectable K33 and K48 poly-ubiquitin signal in

TBK1(K670R) immunoprecipitates (Figure 5M). Together these results suggest that USP38 may specifically remove K33-linked ubiquitin chains at Lys670 of TBK1. To determine whether the K33-linked ubiquitination of TBK1 is more stable than the K48-linked ubiquitinated TBK1, we performed the cycloheximide (CHX) assay and pulled down K33-, K63-, or K48-ubiquitinated TBK1, respectively; we found that K33-ubiquitinated TBK1 was more stable than K48-ubiquitinated TBK1 (Figure S4I). Taken together, these data suggest that USP38 inhibits type I IFN signaling by removing K33-linked and promoting K48-linked poly-ubiquitination chains of TBK1 at Lys670. USP38 Promotes TBK1 Degradation in an NLRP4Dependent Manner Since both NLRP4 and USP38 degraded TBK1 by a similar mechanism that promoted the K48-linked ubiquitination of TBK1 at Lys670, we reasoned that NLRP4 and USP38 work together to regulate TBK1 stability. To test this possibility, we performed immunoprecipitation and immunoblot analysis in 293T cells expressing USP38, NLRP3, or NLRP4, and we found that USP38 interacted with NLRP4, but not with NLRP3 (Figure 6A). To demonstrate the interaction between USP38 and NLRP4 in physiological conditions, we infected THP-1 cells or PBMCs with VSV-EGFP for the indicated time points, and we observed the enhanced interaction between USP38 and NLRP4 after viral infection (Figures 6B and 6C). In addition, we found that the PYD and NOD domains of NLRP4 interacted with USP38 (Figure 6D), whereas the N-terminal domain of USP38 was essential for binding to NLRP4 (Figure 6E). Further experiments showed that USP38 was co-localized with TBK1 and NLRP4 (Figure S5A). To substantiate these findings, we specifically knocked down NLRP4 and found that the interaction between USP38 and TBK1 was markedly attenuated (Figure 6F). Furthermore, we demonstrated that the dynamic interaction between USP38 and TBK1 was totally abrogated in NLRP4 small interfering RNA (siRNA)treated THP-1 cells after viral infection (Figure 6G). Interestingly, we also observed the reduced interaction between TBK1 and TRIP or DTX4 when NLRP4 was knocked down (Figures S5B and S5C). These results suggest that USP38, TRIP, and DTX4 may bind to TBK1 in an NLRP4-dependent manner. Consistent with these observations, we found neither USP38 nor TRIP degraded TBK1 in NLRP4-silencing cells (Figures 6H and S5D). Luciferase reporter assay further showed that the inhibition of ISRE-luc activity by USP38 could be completely relieved when NLRP4 was knocked down (Figures S5E and S5F). Taken together, these results suggest that NLRP4 is indispensable for the inhibitory function of USP38. Regulation of TBK1 Stability by the NLRP4 Signalosome It is well documented that NLRP3 plays a role as a core protein, recruiting ASC and pro-caspase-1 to initiate inflammasome

(F) Co-immunoprecipitation and immunoblot analysis of 293T cells transfected with HA-TBK1 and FLAG-USP38, as well as Ctrl or NLRP4-specific siRNA, are shown. (G) Co-immunoprecipitation and immunoblot analysis of THP-1 cells infected with VSV-EGFP and harvested at the indicated time points are shown. (H) Immunoblot analysis of 293T cells transfected with FLAG-TBK1 and Myc-USP38, as well as Ctrl or NLRP4-specific siRNA. See also Figure S5.


Figure 7. NLRP4 Signalosome Regulates TBK1 Stability (A and B) Immunoassay of extracts of 293T cells transfected with various combinations of plasmid encoding Myc-tagged TRIP and GFP-tagged DTX4, then infected with VSV-EGFP for the indicated time points, followed by immunoprecipitation with anti-TBK1 (A) or anti-NLRP4 (B) and immunoblotted with the indicated antibodies, is shown. (C) Co-immunoprecipitation and immunoblot analysis of extracts of WT or USP38/ 293T cells transfected with FLAG-TBK1, HA-NLRP4, HA-TRIP, and HADTX4 are shown. (D) Co-immunoprecipitation and immunoblot analysis of extracts of WT or NLRP4/ 293T cells transfected with FLAG-TBK1, HA-USP38, HA-TRIP and HA-DTX4 are shown. (legend continued on next page)


assembly during inflammation (Sutterwala et al., 2014). Our results showed that USP38, DTX4, and TRIP interact with TBK1 in an NLRP4-dependent manner. Based on these findings, we assumed that NLRP4, together with USP38 and E3 ligases, might form the NLRP4 signalosome to control TBK1 stability and type I IFN signaling. To determine the sequential events of the interaction between TBK1 and the members of the NLRP4 signalosome under physiological conditions, we infected MycTRIP- and GFP-DTX4-expressing cells with VSV, and then we immunoprecipitated different complexes from cells, collected at different time points post-infection with either TBK1 or NLRP4 antibodies. Immunoblot analysis revealed that NLRP4 and USP38 targeted to TBK1 at 6 hr post-infection, whereas DTX4 and TRIP interacted with TBK1 at a later time point (at 8 hr) (Figure 7A). Importantly, USP38, but not DTX4 or TRIP, constitutively interacted with NLRP4 in unstimulated cells (Figure 7B). Interaction between NLRP4 and USP38 was further increased after viral infection. Interactions between NLRP4 and TRIP or DTX4 were detected at 6 or 8 hr, respectively, post-infection (Figure 7B). These results illustrate the dynamic assembly of the NLRP4 signalosome during viral infection. To identify the molecular mechanism by which the NLRP4 signalosome regulates TBK1 stability, we generated NLRP4/ and USP38/ HEK293T cells, respectively (Figures S6A and S6B). We found that USP38/ cells and NLRP4/ cells showed increased ISRE luciferase activity, ISG expression, and antiviral immunity after viral infection (Figures S6C–S6F). Next we determined the role of USP38 and NLRP4 in the interaction between TBK1 and the NLRP4 signalosome. We found that USP38 deficiency does not affect the interaction of TBK1 with NLRP4, TRIP, or DTX4 (Figure 7C). By contrast, in NLRP4/ cells, the interactions of TBK1 with USP38, TRIP, or DTX4 were completely abrogated (Figure 7D). These results suggest that NLRP4 is a core protein in the NLRP4 signalosome to bridge USP38, TRIP, or DTX4 to TBK1, whereas the interaction of TBK1 with NLRP4, TRIP, or DTX4 does not depend on USP38. Next we investigated how the members of the NLRP4 signalosome affected TBK1 ubiquitination. We observed increased levels of K33-linked ubiquitination and decreased levels of K48-linked ubiquitination of TBK1 in both USP38/ and NLRP4/ cells (Figures 7E and 7F). Moreover, TRIP or DTX4 deficiency only affected K48-linked ubiquitination, but not K33linked ubiquitination, of TBK1 (Figures S7A and S7B). These data suggest that both NLRP4 and USP38 are required for the removal of K33-linked ubiquitination on TBK1, whereas E3 ligases DTX4 and TRIP are required for adding K48-linked ubiquitination on TBK1. Interestingly, although NLRP4 or TRIP still interacted with TBK1 in USP38/ cells (Figure 7C), they could not cause TBK1 degradation in USP38/ cells even after VSV

infection or IC poly(I:C) treatment (Figures 7G, 7H, S7C, and S7D). Overexpression of WT USP38, but not the USP38 (CA/ HA) mutant, restored the function of NLRP4 to degrade TBK1 in USP38/ cells (Figure 7I), suggesting that the deubiquitinating activity of USP38 is a prerequisite for the whole NLRP4 signalosome to degrade TBK1. Furthermore, we showed that USP38 or TRIP could not degrade TBK1 in NLRP4/ cells (Figures 7J and S7E). Taken together, these data suggest that NLRP4, USP38, TRIP, and DTX4 worked cooperatively as a signalosome to control TBK1 ubiquitination transition and degradation, thus inhibiting IFN-b signaling. DISCUSSION Type I IFN pathways are crucial to set up the antiviral state in the cells against various types of viruses. Because aberrant immune responses can be harmful to the host, type I IFN signaling must be tightly regulated to avoid autoimmune or chronic inflammatory diseases (Eisena¨cher and Krug, 2012). TBK1 is a key point of convergence of type I IFN signaling activated by diverse PRRs in response to various DNA and RNA viruses, and it is precisely regulated in a spatiotemporal manner by various molecules through different PTMs, such as phosphorylation, ubiquitination, and deubiquitination (Zhao, 2013). Several regulators, including SHIP1, PPM1B, glucocorticoids, and GSK3b, have been reported to modulate TBK1 activity through phosphorylation or self-association of TBK1 (Gabhann et al., 2010; Lei et al., 2010; McCoy et al., 2008; Zhao et al., 2012). In addition, TBK1 can be activated by K63-linked poly-ubiquitination, mediated by the E3 ligases MIB1 or Nrdp1 in response to RNA virus or lipopolysaccharide (LPS), respectively (Cui et al., 2010; Xia et al., 2011). The K63-linked ubiquitination of TBK1 can be reversed by several DUBs, including A20, Tax1-binding protein 1 (TAX1BP1), and USP2b (Parvatiyar et al., 2010; Zhang et al., 2014b). Recently, we and others showed that NLRP4/DTX4, siglec1/ SHP2/TRIM27, as well as TRIP specifically regulate the activated form of TBK1 for its degradation through K48-linked ubiquitination of TBK1, thus inhibiting type I IFN signaling (Cui et al., 2012; Zhang et al., 2012; Zheng et al., 2015). Despite these significant progresses, mechanistic details regarding TBK1 ubiquitination/deubiquitination processing are still not clear. It appears that multiple protein complexes are involved in the control of TBK activation and degradation. Thus, it is critical to elucidate its molecular mechanisms for a better understanding of how dynamic ubiquitin editing shapes TBK1 function as well as antiviral responses. Our findings show that USP38 potently inhibits type I IFN signaling activated by RLRs and cGAS. Conversely, USP38 deficiency enhances type I IFN signaling, ISG expression, and

(E and F) Immunoassays of extracts of WT or USP38/ (E) or NLRP4/ (F) 293T cells transfected with HA-tagged K33-linked or HA-tagged K48-linked ubiquitin, then infected with VSV-EGFP for the indicated time points, followed by immunoprecipitation with anti-TBK1 and immunoblot analysis with anti-HA, are shown. (G and H) Immunoassays of extracts of WT or USP38/ 293T cells transfected with EV, HA-NLRP4 (G), or HA-TRIP (H), followed by VSV-EGFP infection for the indicated time points, are shown. (I) Immunoassay of extracts of USP38/ 293T cells transfected with HA-NLRP4 together with EV, Myc-USP38, or Myc-USP38 (CA/HA), followed by IC poly(I:C) stimulation for 6 hr, is shown. (J) Immunoassay of extracts of WT or NLRP4/ 293T cells transfected with EV or HA-USP38, followed by VSV infection at the indicated time points. See also Figures S6 and S7.


antiviral immunity against both DNA and RNA viral infection. Importantly, our results have been validated in Usp38/ mice in response to viral infection. To understand the molecular mechanisms of USP38 action, we show that USP38 specifically interacts with active TBK1 after viral infection and inhibits type I IFN signaling at the level of TBK1. Unlike A20 and USP2b, USP38 does not affect the K63-linked ubiquitination of TBK1, but rather it promotes the K48-linked ubiquitination of TBK1 for its degradation. Interestingly, our study indicates that, in addition to K48- and K63-linked ubiquitination, TBK1 undergoes robust K33-linked ubiquitination upon viral infection. USP38 specifically inhibits the K33-linked ubiquitination of TBK1 upon viral infection. Surprisingly, we found that both K33- and K48linked poly-ubiquitin chains conjugated to TBK1 on the same amino acid (Lys670). K33-linked poly-ubiquitin chains on Lys670 of TBK1 may prevent the degradation of TBK1 mediated by NLRP4, DTX4, or TRIP. Thus, our results have identified USP38 as a DUB responsible for cleaving K33-linked poly-ubiquitin chains and mediating the K33-K48 ubiquitination transition of TBK1, in concert with NLRP4, DTX4, and TRIP. Although NLR proteins were originally thought to positively regulate inflammatory responses (Petrilli et al., 2005)—for example, NLRP3 functions as a platform, recruiting ASC and pro-caspase1 to form large protein complexes termed inflammasomes, which cleave pro-IL-1b and pro-IL-18 (Sutterwala et al., 2014)—accumulating evidence suggests that several NLRs (also known as regulatory NLRs), including NLRX1, NLRC5, NLRP4, and NLRC3, negatively regulate innate immune signaling (Cui et al., 2010, 2012; Xia et al., 2011; Zhang et al., 2014a). Therefore, we propose whether NLRP4 also can form a large protein complex, which we termed the NLRP4 signalosome, to negatively regulate type I IFN signaling. Our results show that NLRP4 interacts with USP38, but not TRIP or DTX4, in unstimulated cells, but it binds to the activated form of TBK1 and recruits DTX4/ TRIP to TBK1/NLRP4/USP38 complexes in a time-dependent manner after viral infection. Using a CRISPR/Cas9 editing system, we generated NLRP4/, USP38/, DTX4/, and TRIP/ cells to analyze the assembly of the NLRP4 signalosome. NLRP4 functions as a core protein within the entire signalosome, which is essential for all of the regulators to target and degrade TBK1. USP38 deficiency could not block the interaction between TBK1 and NLRP4/DTX4/TRIP, but it abrogated the removal of K33-linked ubiquitin chains on TBK1, resulting in the inhibition of TBK1 degradation by NLRP4 or TRIP. It appears that DTX4 and TRIP play a redundant role as E3 ligases to prompt the K48-linked ubiquitination of TBK1. Deficiency of DTX4 or TRIP does not affect the K33-linked ubiquitination of TBK1, but it reduces the K48-linked ubiquitination of TBK1. On the basis of these findings in this study and published data, we propose a three-step working model to explain how the NLRP4 signalosome negatively regulates TBK1 and type I IFN signaling (Figure S7F). After VSV infection, TBK1 is activated by phosphorylation at Ser172 and undergoes both K63- and K33-linked ubiquitination, which triggers IRF3 phosphorylation and type I IFN signaling. The phosphorylation of TBK1 at Ser527 by DYRK2 (An et al., 2015) may facilitate the binding of NLRP4 to the activated TBK1 along with USP38 to remove the K33-linked ubiquitin chain on the Lys670 of TBK1. Subsequently,

DTX4 and TRIP are recruited to TBK1 by NLRP4, catalyzing K48linked ubiquitination at the Lys670 site of TBK1 for its proteasomal degradation. Importantly, in USP38/ cells, NLRP4/DTX4/ TRIP complexes can still bind to TBK1, but they fail to degrade TBK1 with K33 poly-ubiquitin chains due to a failure to replace it with K48-linked poly-ubiquitin chains. Besides its importance in innate immune signaling, TBK1 has been shown to play an important role in KRAS-mediated tumor development (Barbie et al., 2009). Thus, it is possible that the negative regulation of TBK1 by the NLRP4 signalosome may play an important role in cancer development. Although our previous work suggests that seven murine homologs of NLRP4 (NLRP4a–NLRP4g) may have redundant roles in inhibiting type I IFN signaling (Cui et al., 2012), only one USP38 gene exists in both human and mice; but its involvement in cancer requires further investigation. In summary, our study identified an essential role for USP38 in the negative regulation of type I IFN signaling and antiviral immunity in vitro and in vivo, and it provided molecular insights into the mechanisms by which USP38 plays a key role in removing K33linkaged poly-ubiquitin chains on TBK1, thus allowing the addition of K48-linked ubiquitin chains by DTX4/TRIP for its degradation through NLRP4, which forms the signalosome with DUBs and E3 ligases. Therefore, USP38 may play a critical role in the maintenance of the balance between innate immune responses and tolerance through dynamic ubiquitin editing of TBK1 in response to viral infection and perhaps other infectious diseases or cancer. EXPERIMENTAL PROCEDURES Mice Usp38-deficient mice were generated by TALEN technology (Cyagen). Primary bone marrow was collected from WT and USP38-deficient mice for BMMs. Animals were housed in specific pathogen-free barrier facilities. All experiments were performed according to the institutional guidelines at Sun Yat-Sen University. In Vitro Deubiquitination Assay Ubiquitinated TBK1 and recombinant FLAG-USP38 were purified from the cell extracts using an anti-FLAG affinity column. After extensive washing with lysis buffer, the proteins were eluted with FLAG peptides (Sigma). For in vitro deubiquitination assay, ubiquitinated TBK1 protein was incubated with recombinant USP38 in the deubiquitination buffer (50 mM Tris-HCl [pH 8.0], 50 mM NaCl, 1 mM EDTA, 10 mM DTT, and 5% glycerol) for 2 hr at 37 C, and it was immunoprecipitated with anti-FLAG antibody for immunoblot analysis. Cycloheximide Chase Assays Cells were treated with cycloheximide (100 mg/mL) for different time points at 24 hr after virus infection. Next the cells were harvested and cell lysates were analyzed by immunoblot. Immunofluorescence Staining Cells were fixed for 15 min with 4% paraformaldehyde and then permeabilized in methyl alcohol for 10 min at –20 C. After washing three times with PBS, fixed cells were blocked in 10% normal goat serum for 1 hr, incubated with primary antibody overnight, and incubated with a fluorescently labeled secondary antibody (goat anti-rat IgG [H+L] conjugated 43501A from Invitrogen). Nuclei were stained with DAPI (Invitrogen). Statistical Analysis Unless indicated otherwise, all data are plotted as means ± SEM. Significant differences between groups were determined by two-tailed Student’s t test.


SUPPLEMENTAL INFORMATION

Goubau, D., Deddouche, S., and Reis e Sousa, C. (2013). Cytosolic sensing of viruses. Immunity 38, 855–869.

Supplemental Information includes Supplemental Experimental Procedures and seven figures and can be found with this article online at http://dx.doi. org/10.1016/j.molcel.2016.08.029.

Goubau, D., Schlee, M., Deddouche, S., Pruijssers, A.J., Zillinger, T., Goldeck, M., Schuberth, C., Van der Veen, A.G., Fujimura, T., Rehwinkel, J., et al. (2014). Antiviral immunity via RIG-I-mediated recognition of RNA bearing 50 -diphosphates. Nature 514, 372–375.

AUTHOR CONTRIBUTIONS R.-F.W. and J.C. proposed the concept and conceived the entire study. M.L., Z.Z., and Z.Y. performed most of the experiments, with assistance from Q.M., P.T., W.X., and Y.Q. for some experiments. M.L., Z.Z., Z.Y., and J.C. collected data and performed analyses. M.L., J.C., and R.-F.W. wrote the manuscript with input from all the authors. ACKNOWLEDGMENTS This work was supported by National Key Basic Research Program of China (2014CB910800 and 2015CB859800), National Natural Science Foundation of China (31370869 and 31522018), Guangdong Natural Science Funds for Distinguished Young Scholar (S2013050014772), Guangdong Innovative Research Team Program (NO. 2011Y035 and 201001Y0104687244), and was in part supported by grants (CA101795 and DA030338) from National Cancer Institute (NCI) and National Institute on Drug Abuse (NIDA), NIH (to R.-F.W.). We thank Dr. Ning for the graphical abstract. We also thank Dr. Fan for the mass spectrometry analysis of TBK1 ubiquitination. Received: February 4, 2016 Revised: August 3, 2016 Accepted: August 24, 2016 Published: September 29, 2016 REFERENCES Ablasser, A., Bauernfeind, F., Hartmann, G., Latz, E., Fitzgerald, K.A., and Hornung, V. (2009). RIG-I-dependent sensing of poly(dA:dT) through the induction of an RNA polymerase III-transcribed RNA intermediate. Nat. Immunol. 10, 1065–1072. An, T., Li, S., Pan, W., Tien, P., Zhong, B., Shu, H.B., and Wu, S. (2015). DYRK2 negatively regulates type I interferon induction by promoting TBK1 degradation via Ser527 phosphorylation. PLoS Pathog. 11, e1005179. Barbie, D.A., Tamayo, P., Boehm, J.S., Kim, S.Y., Moody, S.E., Dunn, I.F., Schinzel, A.C., Sandy, P., Meylan, E., Scholl, C., et al. (2009). Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature 462, 108–112. Chiu, Y.H., Macmillan, J.B., and Chen, Z.J. (2009). RNA polymerase III detects cytosolic DNA and induces type I interferons through the RIG-I pathway. Cell 138, 576–591. Cui, J., Zhu, L., Xia, X., Wang, H.Y., Legras, X., Hong, J., Ji, J., Shen, P., Zheng, S., Chen, Z.J., and Wang, R.F. (2010). NLRC5 negatively regulates the NFkappaB and type I interferon signaling pathways. Cell 141, 483–496. Cui, J., Li, Y., Zhu, L., Liu, D., Songyang, Z., Wang, H.Y., and Wang, R.F. (2012). NLRP4 negatively regulates type I interferon signaling by targeting the kinase TBK1 for degradation via the ubiquitin ligase DTX4. Nat. Immunol. 13, 387–395. Cui, J., Song, Y., Li, Y., Zhu, Q., Tan, P., Qin, Y., Wang, H.Y., and Wang, R.F. (2014). USP3 inhibits type I interferon signaling by deubiquitinating RIG-I-like receptors. Cell Res. 24, 400–416. Eisena¨cher, K., and Krug, A. (2012). Regulation of RLR-mediated innate immune signaling–it is all about keeping the balance. Eur. J. Cell Biol. 91, 36–47. Ferguson, B.J., Mansur, D.S., Peters, N.E., Ren, H., and Smith, G.L. (2012). DNAPK is a DNA sensor for IRF-3-dependent innate immunity. eLife 1, e00047. Gabhann, J.N., Higgs, R., Brennan, K., Thomas, W., Damen, J.E., Ben Larbi, N., Krystal, G., and Jefferies, C.A. (2010). Absence of SHIP-1 results in constitutive phosphorylation of tank-binding kinase 1 and enhanced TLR3-dependent IFN-beta production. J. Immunol. 184, 2314–2320.


Gu¨rtler, C., and Bowie, A.G. (2013). Innate immune detection of microbial nucleic acids. Trends Microbiol. 21, 413–420. Kato, H., Takeuchi, O., Mikamo-Satoh, E., Hirai, R., Kawai, T., Matsushita, K., Hiiragi, A., Dermody, T.S., Fujita, T., and Akira, S. (2008). Length-dependent recognition of double-stranded ribonucleic acids by retinoic acid-inducible gene-I and melanoma differentiation-associated gene 5. J. Exp. Med. 205, 1601–1610. Kondo, T., Kobayashi, J., Saitoh, T., Maruyama, K., Ishii, K.J., Barber, G.N., Komatsu, K., Akira, S., and Kawai, T. (2013). DNA damage sensor MRE11 recognizes cytosolic double-stranded DNA and induces type I interferon by regulating STING trafficking. Proc. Natl. Acad. Sci. USA 110, 2969–2974. Lei, C.Q., Zhong, B., Zhang, Y., Zhang, J., Wang, S., and Shu, H.B. (2010). Glycogen synthase kinase 3b regulates IRF3 transcription factor-mediated antiviral response via activation of the kinase TBK1. Immunity 33, 878–889. Li, S., Wang, L., Berman, M., Kong, Y.Y., and Dorf, M.E. (2011). Mapping a dynamic innate immunity protein interaction network regulating type I interferon production. Immunity 35, 426–440. Loo, Y.M., and Gale, M., Jr. (2011). Immune signaling by RIG-I-like receptors. Immunity 34, 680–692. McCoy, C.E., Carpenter, S., Pa˚lsson-McDermott, E.M., Gearing, L.J., and O’Neill, L.A. (2008). Glucocorticoids inhibit IRF3 phosphorylation in response to Toll-like receptor-3 and -4 by targeting TBK1 activation. J. Biol. Chem. 283, 14277–14285. Paludan, S.R., and Bowie, A.G. (2013). Immune sensing of DNA. Immunity 38, 870–880. Parvatiyar, K., Barber, G.N., and Harhaj, E.W. (2010). TAX1BP1 and A20 inhibit antiviral signaling by targeting TBK1-IKKi kinases. J. Biol. Chem. 285, 14999– 15009. Petrilli, V., Papin, S., and Tschopp, J. (2005). The inflammasome. Curr. Biol. 15, R581. Sun, L., Wu, J., Du, F., Chen, X., and Chen, Z.J. (2013). Cyclic GMP-AMP synthase is a cytosolic DNA sensor that activates the type I interferon pathway. Science 339, 786–791. Sutterwala, F.S., Haasken, S., and Cassel, S.L. (2014). Mechanism of NLRP3 inflammasome activation. Ann. N Y Acad. Sci. 1319, 82–95. Takeuchi, O., and Akira, S. (2010). Pattern recognition receptors and inflammation. Cell 140, 805–820. Theofilopoulos, A.N., Baccala, R., Beutler, B., and Kono, D.H. (2005). Type I interferons (alpha/beta) in immunity and autoimmunity. Annu. Rev. Immunol. 23, 307–336. Unterholzner, L., Keating, S.E., Baran, M., Horan, K.A., Jensen, S.B., Sharma, S., Sirois, C.M., Jin, T., Latz, E., Xiao, T.S., et al. (2010). IFI16 is an innate immune sensor for intracellular DNA. Nat. Immunol. 11, 997–1004. Wang, C., Chen, T., Zhang, J., Yang, M., Li, N., Xu, X., and Cao, X. (2009). The E3 ubiquitin ligase Nrdp1 ‘preferentially’ promotes TLR-mediated production of type I interferon. Nat. Immunol. 10, 744–752. Xia, X., Cui, J., Wang, H.Y., Zhu, L., Matsueda, S., Wang, Q., Yang, X., Hong, J., Songyang, Z., Chen, Z.J., and Wang, R.F. (2011). NLRX1 negatively regulates TLR-induced NF-kB signaling by targeting TRAF6 and IKK. Immunity 34, 843–853. Yuan, W.C., Lee, Y.R., Lin, S.Y., Chang, L.Y., Tan, Y.P., Hung, C.C., Kuo, J.C., Liu, C.H., Lin, M.Y., Xu, M., et al. (2014). K33-linked polyubiquitination of Coronin 7 by Cul3-KLHL20 ubiquitin E3 ligase regulates protein trafficking. Mol. Cell 54, 586–600. Zhang, M., Wang, L., Zhao, X., Zhao, K., Meng, H., Zhao, W., and Gao, C. (2012). TRAF-interacting protein (TRIP) negatively regulates IFN-b production

and antiviral response by promoting proteasomal degradation of TANK-binding kinase 1. J. Exp. Med. 209, 1703–1711.

Zhao, W. (2013). Negative regulation of TBK1-mediated antiviral immunity. FEBS Lett. 587, 542–548.

Zhang, L., Mo, J., Swanson, K.V., Wen, H., Petrucelli, A., Gregory, S.M., Zhang, Z., Schneider, M., Jiang, Y., Fitzgerald, K.A., et al. (2014a). NLRC3, a member of the NLR family of proteins, is a negative regulator of innate immune signaling induced by the DNA sensor STING. Immunity 40, 329–341.

Zhao, Y., Liang, L., Fan, Y., Sun, S., An, L., Shi, Z., Cheng, J., Jia, W., Sun, W., Mori-Akiyama, Y., et al. (2012). PPM1B negatively regulates antiviral response via dephosphorylating TBK1. Cell. Signal. 24, 2197–2204.

Zhang, L., Zhao, X., Zhang, M., Zhao, W., and Gao, C. (2014b). Ubiquitin-specific protease 2b negatively regulates IFN-b production and antiviral activity by targeting TANK-binding kinase 1. J. Immunol. 193, 2230–2237.

Zheng, Q., Hou, J., Zhou, Y., Yang, Y., Xie, B., and Cao, X. (2015). Siglec1 suppresses antiviral innate immune response by inducing TBK1 degradation via the ubiquitin ligase TRIM27. Cell Res. 25, 1121–1136.
