CoV PLPs associate with STING

One possible mechanism for HCoV PLPs antagonism of STING-mediated activation of IFN is to associate with STING, either directly or as part of a multi-protein complex. Co-immunoprecipitation experiments were performed to determine if CoV PLPs associate with STING. HEK-293T cells were co-transfected with plasmid DNA expressing an epitope tagged version of STING (STING-Flag) in the presence or absence of PLP2-TM and cell lysates were subjected to immunoprecipitation with anti-Flag antibody. The products of the immunoprecipitation were separated by SDS-PAGE and visualized by immunoblotting (Fig. 2A). The results show that both NL63 wild-type and catalytic mutant PLPs are detected in association with STING. Similar co-immunoprecipitation results were obtained using PLpro-TM (Fig. S2) Next, we wanted to determine if STING is sequestered in HCoV-NL63-infected cells. HEK293-ACE2 cells, which express angiotensin-converting enzyme 2 (ACE2), a key receptor for SARS and NL63 coronaviruses, were transfected with STING-V5, infected with HCoV-NL63 and analyzed by confocal microscopy at 24 hrs postinfection. HCoV-NL63 replicase protein nsp3 which contains the PLP2-TM region is detected as punctate, perinuclear staining in virus-infected cells [35], [38]. Interestingly, we detected partial co-localization of STING and nsp3 in virus-infected cells suggesting that STING may be sequestered in the viral replication complex and unable to mediate signaling (Fig. 2B). These results indicate that these CoV PLPs associate with STING either directly or as part of a multi-protein complex.

Open in a separate window

Figure 2

Coronavirus NL63 PLP2-TM associates with STING and nsp3 co-localizes with STING in virus-infected cells.

(A) HEK293T cells were cotransfected with plasmid DNAs expressing STING-Flag and either wild type or catalytic mutants of NL63-PLP2-TM-V5. Cell lysates were prepared at 28 hrs post-transfection and subjected to immunoprecipitation (IP) with anti-Flag antibody. The products of the immunoprecipitation were separated by SDS-PAGE and subjected to immunoblotting (IB). STING-Flag, PLP2-TM-V5 and the catalytic mutant expression were selectively detected from whole cell lysates (WCL) using anti-Flag and anti-V5 antibodies. (B) HEK293-ACE2 cells were transfected with STING-V5 for 4 hours and then infected with HCoV-NL63 for 24 hrs and evaluated for expression of and localization of replicase product nsp3 (anti-nsp3, red) and STING-V5 (anti-V5, green).

STING dimerization is reduced in the presence of CoV PLPs

Recent studies indicate that activation of the innate immune response signaling pathway induces dimerization and phosphorylation of STING, which are required for activation of the IFN response [9]. STING dimers can be visualized as a band at 80 kDa when resolved on SDS-PAGE [9]. We hypothesize that PLPs inhibit STING-mediated signaling through the disruption of assembly or stability of STING dimers. To test this hypothesis, cells were co-transfected with plasmid DNA expressing STING-Flag in the presence or absence of PLP2-TM and Sendai virus (SeV), and cell lysates were evaluated for STING dimers by immunoblotting with anti-Flag (Figure 3A). We detected STING dimers in STING-transfected and SeV-infected cells (Fig. 3A, lanes 2 and 4). In contrast, STING dimers were reduced in cells co-expressing PLP2-TM (Fig. 3A, lanes 3 and 5). Similar results were obtained when we evaluated PLpro-TM for disruption of STING dimers (Fig. S3), indicating that these CoV PLPs either prevent assembly or promote dissociation of STING dimers. A similar reduction in STING dimers was seen in cells transfected with STING-HA and infected with SARS-CoV (Fig. 3B, lane 2). In contract, STING dimers were not reduced when infected with SeV (Fig. 3B, lane 3). Collectively, these results indicate that STING dimerization was reduced in the presence of CoV PLPs, and was also substantially reduced in cells infected with SARS-CoV.

Open in a separate window

Figure 3

NL63 PLP2-TM interacts with STING and disrupts STING dimers.

(A) HEK293T cells were co-transfected with plasmid DNA expressing STING-Flag, and/or PLP2-TM and/or infected with Sendai virus (SeV) as indicated above. The cell lysates were separated by SDS-PAGE and subjected to immunoblotting with antibodies as indicated on the left. (B) HEK293-ACE2 cells were transfected with plasmid DNA expressing STING-HA and infected with SARS-CoV as indicated and cell lysates were subjected to immunoprecipitation with anti-HA. The immunoprecipitated products were analyzed by SDS-PAGE and immunoblotted to access STING monomers and dimers. Whole cell lysates were immunoblotted to detected SARS-CoV replicase protein nsp3 and STING-HA. (C) Cells were co-transfected with STING-HA and STING-Flag with either wild-type or the indicated catalytic mutant of PLP2-TM and lysates were immunoprecipitated (IP) and immunoblotted (IB) to detect expression of each product.

To determine if PLP catalytic activity is important for disruption of STING dimers, cells were co-transfected with plasmid DNA expressing STING-HA and STING-Flag and plasmid DNA expressing wt or catalytic mutants of PLP2-TM. Disruption of dimers was assessed by immunoprecipitation and immunoblotting (Fig. 3C). If STING-HA and STING-Flag form heterodimers, then immunoprecipitation with anti-HA and immunoblotting with anti-Flag will allow detection of these heterodimers, as shown in Fig. 3C, lane 2. A reduction in STING dimers was detected in cells expressing wt PLP2-TM and the D1849A mutant which retains DUB activity (lanes 3 and 6), but not in cells expressing the C1678A or H1836A mutants which do not possess DUB activity [33]. These results indicate that PLP2 DUB activity is important for disruption of STING dimers. Similar results were obtained when we evaluated PLpro-TM for disruption of STING dimers (Fig. S3) A previous report indicates that only the dimer form of STING is ubiquitinated [9], which coupled with our observation that PLPs with DUB activity reduce the accumulation of STING dimers, supports a role for DUB activity in negatively regulating STING. Thus, these results are consistent with a role for viral DUB activity in antagonizing either the assembly or stability of STING dimers.

PLP2-TM disrupts MAVS-STING-IKKε interaction

Next, we wanted to determine if PLP2-TM altered the assembly of complexes required for activation of IRF-3 and the IFN response. MAVS, a mitochondrial-associated adaptor protein is a critical player in viral activation of the IFN response. Activation of MAVS mediates the assembly of a multi-protein complex that activates TBK-1/IKKε to phosphorylate IRF-3 [39]. Recent studies revealed that STING associates with MAVS to recruit TBK-1/IKKε and IRF-3 to a complex [8], and that activation of STING is critical for activation of IRF-3 [11], [12]. Therefore, we investigated if PLP2-TM had any effect on assembly of these signaling complexes. HEK-293T cells were co-transfected with plasmid DNAs expressing STING-HA along with Flag-RIG-I, Flag-MAVS or Flag-IKKε in the presence or absence of PLP2-TM. Cell lysates were harvested and evaluated for co-immunoprecipitation of complexes by immunoblotting. We found that expression of PLP2-TM had no effect on co-immunoprecipitation of RIG-I with STING (Fig. 4A), but that co-immunoprecipitation of MAVS and IKKε was disrupted by expression of PLP2-TM (Fig. 4B, lane 4 and Fig. 4C, lane 4). Overall, these results are consistent with an important role for CoV PLPs in blocking activation of IFN by disrupting STING-mediated activation and complex formation.

Open in a separate window

Figure 4

NL63 PLP2-TM disrupts signaling complex formation.

HEK293T cells were co-transfected with STING-HA together with either Flag-tagged RIG-I (A), Flag-tagged MAVS (B) or Flag-tagged IKKε (C), and PLP2-TM-V5. At 28 h after transfection, cell lysates were prepared and subjected to immunoprecipitate (IP) and immunoblot (IB) with the indicated antibodies. The asterisk indicates the nonspecific band.

Targeting the stimulator of IFN genes, STING

Recent studies have revealed the arsenal of proteins that viruses use to evade and subvert recognition by pattern-recognition receptors (PRRs) or activation of signaling molecules that are designed to respond to infectious agents [45]. STING is a key scaffolding protein that links the cytosolic viral RNA sensors RIG-I, rather than the MDA5, to the mitochondria protein MAVS [8], [11]. These cytosolic sensors have been shown to be important for recognition of coronavirus RNA in oligodendrocytes [21] and brain macrophages/microglia [20]. Activation of STING, either by expression of N-RIG, the constitutive active caspase recruitment domain of RIG-I or ectopic expression of STING itself, induces the formation of STING dimers, which are modified by phosphorylation and ubiquitylation [9]. The activation of STING facilitates the recruitment of IRF-3 and TBK-1 into a complex where IRF-3 is phosphorylated. Phosphorylated IRF-3 forms dimers and is transported to the nucleus to activate transcription of type I IFN genes. CoV PLPs target STING and prevent this key scaffolding protein from activating IRF-3. Previously, we showed that PLpro blocks NF-κB-dependent promoter activity and that antagonism is abrogated using protease inhibitors [14]. Thus, CoV PLPs interact with key signaling molecules and exploit both catalytic dependent and catalytic-independent mechanisms to block the innate immune response. CoV-PLPs disrupt signal transduction to both IRF-3 and NF-κB, the key transcription factors required for activation of IFN-β (Fig. 6). Interestingly, CoV PLPs seem to exploit both catalytic dependent and independent mechanisms to block STING activity and a catalytic-dependent mechanism to disrupt NF-κB activity.

Open in a separate window

Figure 6

Model depicting the mechanisms used by CoV PLPs to block STING from signaling the activation of the IFN-β induction pathway.

(A) Activation of sensors such as RIG-I induces interaction with the signaling complex including MAVS, STING, IRF-3 and TBK-1. Activated MAVS interacts with STING, which dimerizes, leading to the activation of IKK complex, TBK1 and IKKε [8]. The activation of this complex leads to the ubiquitination of RIG-I, STING, IRF-3 and TBK1 and the phosphorylatin of STING and IRF-3. Activated the transcription factor IRF-3 translocates to the nucleus inducing production of IFN. (B) Coronavirus papain-like protease domains (depicted here as PLP) interact with STING to block signaling by blocking assembly or stability of STING dimers and preventing the ubiquitination of signaling proteins, such as RIG-I, TBK1, and IRF-3.

Previous studies have shown a role for viral proteases in cleavage of key IFN signaling molecules [45]. For example, the NS3/4A protease of hepatitis C virus and GB virus B and the 3ABC precursor of hepatitis A virus cleave MAVS/IPS-1, the mitochondria-associated signaling molecule, which blocks activation of IFN synthesis [46]–[51]. Therefore, it seemed reasonable to hypothesize that CoV PLPs exploited their protease or DUB activity to antagonize the innate immune response. However, we had previously shown that catalytically defective PLP mutants were still capable of inhibiting IRF-3 activation [17]. Furthermore, addition of a protease inhibitor that blocks both protease and DUB activity [52] failed to abrogate the PLP inhibition on activation of IRF-3 dependent promoters [14]. This led us to suspect that CoV PLPs were interacting with a component in the IFN signaling pathway. The fact that CoV PLPs are expressed as a polyprotein that localizes to the ER [25], [26] led us to evaluate ER resident signaling molecules as targets of antagonism. The results presented in this study indicate that CoV PLPs are particularly potent antagonists because they can block: 1) STING dimerization; 2) the MAVS-STING-IKKε interaction required for signaling and 3) the ubiquitination of key signaling molecules such as RIG-I, STING, IRF-3 and TBK-1. Interestingly, catalytically inactive mutants of PLPs can interact, either directly or as part of a signaling complex, with STING and moderately inhibit IRF-3 activation for IFN induction, but wild type PLPs exhibit the most robust inhibition. We found that catalytic activity was important for blocking either the assembly or the stability of STING dimers. In addition, we found that PLP2-TM either actively deubiquitinates or blocks the ubiquitination of the signaling molecules. This is consistent with these reported works that STING dimers were modified by ubiquitin [9]. By associating with STING, either directly or as part of a signaling complex, CoV PLPs can target and either block or deubiquitinate this important signaling molecule. Of course the question still remains if the PLP-STING interaction is direct or indirect, and what region(s) of the PLP are required for this potential interaction. These studies are currently ongoing. Understanding how to negatively regulate STING-mediated innate antiviral signaling by CoV PLPs may lead to the development of novel antiviral therapies and new insights for regulating the IFN response during acute and chronic infections.

Plasmid DNAs

DNA constructs containing wild type and catalytic mutants of NL63 PLP2-TM SARS-CoV PLpro-TM and plasmids of IFN-β-Luc, PRD(III-I)4-Luc, ISRE-Luc and HA-tagged Ub were previously described [14], [17]. Flag-hIPS-1(MAVS), BOS-Myc-hIRF3, Flag-hTBK1, Flag-hRIG-I were kindly provided by Dr. Himanshu Kuma and Shizuo Akira (Immunology Frontier Research Center Osaka University, Osaka, Japan). pcDNA3.1-HA-ERIS (designated here as STING-HA) and pCMV14-Flag-ERIS (designated STING-Flag) were kindly provided by Dr. Zhengfan Jiang (School of Life Sciences, Peking University, Beijing, China).

Luciferase reporter gene assay

HEK293T cells were transfected with the indicated stimulator plasmid DNA (STING-HA), reporter plasmid DNA [pRL-TK, ISRE-Luc, IFN-β-Luc, or PRD(III-I)4-Luc] and either NL63 PLP2-TM or SARS-CoV PLpro-TM using either Lipofectamine 2000 or Mirus LT1 according to the manufacturer's protocol and incubated for 24 hours. Firefly luciferase and Renilla luciferase activities were assayed using the Dual Luciferase Reporter Assay Kit (Promega). Data were shown as mean relative luciferase (firefly luciferase activity divided by Renilla luciferase activity) with standard deviation from a representative experiment carried out in triplicate. For statistical analysis, the data between Vector (300 ng) and PLP (PLpro)-TM with various concentrations were subjected to unpaired, two-tailed Student's t test using the Microsoft SPSS 12.0 software, and P values of <0.05 were considered to indicate statistical significance.

Immunofluorescence assays

HEK293 or HEK293-ACE2 cells were plated on fibronectin treated glass coverslips in 12-well plates. To evaluate localization of STING and NL63 PLP2-TM, plasmid DNA expressing STING-HA (225 ng per well) was transfected in the presence or absence of 300 ng of PLP2-TM using Mirus LT1 according to the manufacturer's protocol. At 16 hours post transfection, cells were fixed with 3.7% formaldehyde for 10 min at room temperature. Cells were then incubated with 1200 dilution of rabbit anti-IRF-3 (Active Motif), mouse anti-HA (Covance), and goat anti-V5 (Genscript) in ADPS (PBS+0.1% Triton-×100+5% fetal calf serum) for 1 hour at room temperature. Cells were washed three times with PBS and incubated with 1200 dilution of chicken anti-rabbit Alexa Fluor (AF) 488, donkey anti-mouse AF568, and donkey anti-goat AF647 (Molecular Probes) in ADPS for 1 hour in the dark. Following the incubation, cells were washed three times with PBS, mounted, and imaged with the Zeiss LSM-510 confocal microscope.

To evaluate STING localization in NL63 infected cells, HEK293-ACE2 cells were transfected with 100 ng of a plasmid expressing human STING-V5 for 4 hours. Cells were subsequently infected with 200 µl of HCoV-NL63 (1×10⁴ pfu/ml). At 24 hours post infection, cells were fixed with 3.7% formaldehyde for 10 minutes at room temperature. Cells were then diluted with a 11000 dilution of rabbit anti-nsp3 [35] and 11000 dilution of mouse anti-V5 for 1 hour at room temperature in ADPS. Cells were washed three times with PBS and incubated with a 1200 dilution of goat anti-rabbit AF568 and chicken anti-mouse AF488, and 11000 dilution of DAPI for 30 minutes at room temperature in the dark. The cells were then washed three times with PBS, mounted, and imaged with the Zeiss LSM-510 confocal microscope.

Co-immunoprecipitation (Co-IP) analysis

HEK293T cells were seeded on 100-mm dishes at a density of 1×10⁶ cells/dish. 12 hours later, cells were transiently transfected with a total of 10 µg of empty plasmid or indicated expression plasmids using Lipofectamine 2000 (Invitrogen). At 28 hours after transfection, cells were lysed in buffer containing 0.5% Triton-X-100, 150 mM NaCl, 12.5 mM β-glycerolphosphate, 1.5 mM MgCl2, 2 mM EGTA, 10 mM NaF, 1 mM Na₃VO₄, 2 mM DTT plus protease inhibitor cocktail (Sigma). The cell extracts were spun down at 5000× g for 10 minutes at 4°C. The protein concentration of each lysate was determined using the BCA Protein Assay (Bio-Rad) and the concentration was adjusted to 1 µg/µl, with 500 µl of lysate used for each IP. The lysates were precleared by adding 20 µl protein A+G Agarose (Beyotime Institute of Biotechnology, China) and 1 µg of normal IgG and incubating for 2 hours at 4°C, followed by spinning down the beads. The precleared supernatant was incubated with the designated antibody [Anti-Flag (Sigma) or anti-HA (MBL)/anti-Myc (MBL)] with rocking overnight at 4°C. The beads-antibody-antigen complex was spun down and washed 3 times with 1 ml of lysis buffer. The proteins were eluted from the beads in 30 µl of 2× SDS-PAGE sample buffer subjected to boiling for 10 min. The sample was separated by SDS-PAGE and transferred to PVDF membrane for western blotting.

Assessing ubiquitination of signaling molecules in cultured cells

The effect of HCoV-NL63 PLP2-TM on ubiquitinated proteins in cultured cells was assessed as described previously [53], [55]. Briefly, Flag-tagged RIG-I, TBK1, STING, IRF-3 were co-transfected into HEK293T cells together with pcDNA3.1-HA-Ub, plus wild type or catalytic mutant PLP2-TM DNA using Lipofectamine 2000 according to the manufacturer's instructions. Empty vector pcDNA3.1/V5-HisB was used to standardize the total amount of DNA used for transfection. 24 post-transfection, cells were incubated with 25 µM MG132 for 4 hours, and then lysed in 300 µl of RIPA buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 2 mM EDTA, 1% NP-40, 0.1% SDS) containing protease inhibitor cocktail (1 mM, Roche) and 10 µM NEM, Non-covalently bound proteins were dissociated by boiling in 1% SDS, and samples diluted 110 in lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 2 mM EDTA, 1% NP-40) containing protease inhibitor cocktail and 10 µM NEM. The soluble lysates were then immunoprecipitated with anti-Flag antibody followed by washing with RIPA buffer for three times. Flag-tagged proteins were resolved by SDS-PAGE and sequentially blotted with anti-HA and anti-Flag antibodies. 150 µl of lysate was used for each immunoprecipitation reaction. To confirm the PLP and the catalytic mutant expression level, western blotting with anti-V5 antibody (Invitrogen) was used to detect wild type and catalytic mutant PLP2-TM-V5 protein expression.

We would like to thank Dr. Himanshu Kuma, Dr. Shizuo Akira, Dr. Zhengfan Jiang, Dr. Min-Jung Kim, Dr. Robert M. Krug and Dr. Edward W. Harhaj for kindly providing the reporter plasmids and expression constructs. We also would like to thank Dr. Zhengfan Jiang for the helpful suggestion of STING dimer detection.