Middle East Respiratory Syndrome Coronavirus Nucleocapsid Protein Suppresses Type I and Type III Interferon Induction by Targeting RIG-I Signaling

MERS-CoV N protein suppresses the expression of IFN-α2, IFN-β, and IFN-λ1.

To study whether any MERS-CoV structural proteins play a role in IFN production, we transduced A549 cells with viral E, M, and N components separately and infected viral protein-expressing A549 cells with Sendai virus (SeV). The levels of IFN-α2, -β, and -λ1 mRNA were analyzed by quantitative real-time PCR (qRT-PCR). Figure 1A shows that SeV infection induced high levels of type I and III IFN mRNA in A549 cells. However, the SeV-induced IFN mRNA levels were affected in cells expressing viral structural proteins (Fig. 1B and C). While MERS-CoV E protein did not affect the level of SeV-induced IFN-α2, -β, or -λ1 mRNA, viral N protein significantly suppressed the mRNA levels of all three IFNs and M protein suppressed those of IFN-β and IFN-λ1. The expression of N protein dose-dependently inhibited the induction of IFN-β and IFN-λ1 mRNA (Fig. 1D).

We employed a luciferase assay to examine whether N protein also inhibited SeV-induced ISG expression. HEK293T cells were transfected with reporter plasmid pISRE-Luc or pBV-CXCL10(1200)-Luc, internal control pRL-TK, and titrated doses of pcDNA-MERS-N (expressing V5- and His-tagged N protein) before infection with SeV. The expression of C-X-C motif chemokine 10 (CXCL10; also known as interferon γ-induced protein 10, IP-10), an ISRE-dependent ISG was determined. As shown in Fig. 2A, expression of viral N protein dose-dependently reduced the luciferase activity of both ISRE-Luc and CXCL10-Luc. Transfected HEK293T cells were also treated with exogenous IFN-β to bypass the effect of viral N protein on IFN production and instead focus on the effect of N protein on downstream IFN receptor signaling. Results in Fig. 2B show that viral N protein failed to suppress IFN-β-induced ISG promoter activity. In addition, while N protein suppressed the expression of SeV-induced IFN-induced protein with tetratricopeptide repeats 3 (IFIT3), it did not affect IFN-β-induced IFIT3 expression (Fig. 2C and D). These data indicate that MERS-CoV N protein suppresses antiviral gene expression through inhibition of IFN production and its suppressive effect can be rescued by exogenous IFN-β.

MERS-CoV N protein suppresses IFN-β/IFN-λ1 promoter and NF-ĸB/IRF3 activities.

We further cotransfected HEK293T cells with luciferase reporters driven by the IFN-β/IFN-λ1 promoter and titrated doses of viral N protein plasmids to investigate the involvement of MERS-CoV N protein in transcriptional regulation of IFNs. Results show that viral N protein significantly reduced IFN-β promoter (nucleotides [nt] −113 to +20; IFNB-113) and IFN-λ1 promoter (nt −1200 to +1; IFNL-1200) activities (Fig. 3A and B). Consistent with previous reports that IFN-λ1 transcription is highly dependent on the distal NF-κB-binding sites within the promoter upstream region between −1.2 and −1.0 kb (4, 16, 33, 34), we found that IFN-λ1 promoters of lengths 303 nt (IFNL-303) and 123 nt (IFNL-123) were only weakly activated by SeV. Since both IFNB-113 and IFNL-1200 contain potential binding sites for NF-κB, we further examined whether NF-κB contributes to N protein-mediated suppression of IFN-β/IFN-λ1 promoter activities. Cells were cotransfected with titrated doses of N protein and pNF-ĸB-Luc plasmids. Results in Fig. 3C show that N protein dose-dependently suppressed NF-ĸB-Luc luciferase activity. In addition, N protein suppressed the level of SeV-induced phosphorylated p65 (Fig. 3D). Since IRF3 has been reported to be essential for the activation of IFN-β and IFN-λ1 genes (16, 35, 36), we also investigated the effect of N protein on IRF3 activity. As shown in Fig. 3E and F, both IRF3-Luc luciferase activity and the level of SeV-induced phosphorylated IRF3 were reduced in the presence of N protein. These data indicate that both NF-κB and IRF3 are involved in MERS-CoV N protein-mediated suppression of IFN.

MERS-CoV N protein suppresses RIG-I-CARD-induced IFN promoter activity.

To understand the molecular mechanisms by which MERS-CoV N protein suppresses IFN expression, we first tested whether viral N protein interferes with the RLR pathway in which RIG-I and MDA5 bind to viral RNAs and activate MAVS to induce IFN expression. We found that the expression of N protein had little effect on the MAVS-induced IFN-β and IFN-λ1 promoter activities (Fig. 4A). Interestingly, however, expression of N protein suppressed RIG-I-CARDs-induced IFN-β and IFN-λ1 promoter activities (Fig. 4B). In contrast, N protein expression did not affect MDA5-CARD-induced IFN-β and IFN-λ1 promoter activities (Fig. 4C). Together, these data suggest that MERS-CoV N protein suppresses RIG-I pathway at a step prior to MAVS activation.

MERS-CoV N protein interferes with the ubiquitination of RIG-I.

TRIM25 E3 ligase is known to mediate K63-linked ubiquitination of RIG-I CARDs and is essential for the induction of IFN expression through the RIG-I signaling pathway (6, 37–39). We hypothesized that MERS-CoV N protein suppression of IFN promoter activity is the result of reduced RIG-I ubiquitination. HEK293T cells were cotransfected with RIG-I CARDs (FLAG-CARD-V5His)/full-length RIG-I (FLAG-RIG-I), HA-tagged ubiquitin, and titrated doses of N protein plasmids. As the expression of N protein increased, there was an N protein concentration-dependent decrease in ubiquitination on RIGI CARDs/full-length RIG-I (Fig. 5A and B). In addition, while ectopic overexpression of TRIM25 enhanced SeV-induced IFN-β/IFN-λ1 promoter activity, expression of N protein dose-dependently suppressed the promoter activity (Fig. 5C and D). We further cotransfected HEK293T cells with a luciferase reporter plasmid carrying the IFN-β/IFN-λ1 promoter, N-V5His plasmid, and titrated doses of FLAG-TRIM25 plasmids before infection with SeV. A luciferase assay demonstrated that ectopic overexpression of TRIM25 dose-dependently reversed N protein-mediated suppression of IFN promoter activity (Fig. 5E and F), indicating that N protein interferes with RIG-I signaling by sequestering functional TRIM25.

MERS-CoV N protein competes with RIG-I for interaction with TRIM25.

Since interaction of RIG-I with TRIM25 is a prerequisite for RIG-I ubiquitination, we investigated whether MERS-CoV N protein interferes with RIG-I signaling by interacting with TRIM25. A reciprocal immunoprecipitation assay showed that viral N protein coimmunoprecipitated with TRIM25 (Fig. 6A). Expression of viral N protein dose-dependently diminished RIG-I-TRIM25 interaction (Fig. 6B). MERS-CoV N protein contains three conserved regions, the N-terminal domain (NTD; 1 to 164 aa), the intrinsically disordered linker region (LKR; 165 to 236 aa), and the C-terminal domain (CTD; 237 to 413 aa). Domain mapping and a reporter assay further demonstrated the association of TRIM25 with the CTD fragment, N(237-413), which alone was sufficient to suppress IFN-β promoter activity (Fig. 6C and D). Taken together, our results indicate that MERS-CoV N protein competes with RIG-I to interact with TRIM25, thereby inhibiting RIG-I ubiquitination and IFN production.

MERS-CoV N protein functions as an IFN antagonist to rescue IFN-suppressed SINV-EGFP replication.

The effect of MERS-CoV N protein on IFN production was functionally analyzed using IFN-sensitive recombinant Sindbis virus expressing enhanced green fluorescence protein (SINV-EGFP). The expression of EGFP in SINV-EGFP-infected cells is closely correlated with viral RNA replication (40). HEK293T cells with/without MERS-CoV N protein coexpression were infected with SeV (Fig. 7A). Culture supernatants were collected and used as conditioned media to culture naive HEK293T cells. Conditioned medium was replaced with fresh culture medium before infection with SINV-EGFP the next day. EGFP expression in SINV-EGFP-infected cells was an indicator of IFN levels in the conditioned media. A lower level of IFNs in the conditioned medium allows better replication of SINV-EGFP, thus a higher level of EGFP expression. Immunofluorescence imaging (Fig. 7B) and Western blotting (Fig. 7C) results show that while SINV-EGFP-infected cells cultured in mock medium expressed high levels of EGFP, EGFP expression was attenuated in cells cultured in conditioned medium collected from SeV-infected HEK293T cells that did not express viral N protein. EGFP expression was partially rescued when cells were cultured in conditioned medium harvested from SeV-infected HEK293T cells that expressed viral N protein. These results demonstrate that MERS-CoV N protein functions as an antagonist to suppress the production of not only IFN mRNA but also bioactive IFN protein.

Cell lines.

Human embryonic kidney cell line HEK293T and adenocarcinomic human alveolar basal epithelial cell line A549 were maintained in Dulbecco’s modified Eagle’s medium (DMEM) and RPMI 1640 medium (GIBCO), respectively, supplemented with 8% heat-inactivated fetal bovine serum plus 100 units of penicillin and 100 μg of streptomycin per ml in a humidified 5% CO₂ atmosphere at 37°C.

Plasmids.

MERS-CoV structural protein expression plasmids pcDNA-MERS-N, pcDNA-MERS-M, and pcDNA-MERS-E encode V5- and His-tagged N, M, and E proteins, respectively, as previously described (60). Plasmids pcDNA-MERS-N(1-164), pcDNA-MERS-N(165-236), and pcDNA-MERS-N(237-413) encode V5- and His-tagged subdomains of N protein and were constructed by inserting their corresponding cDNA fragments into the EcoRI/XhoI sites of plasmid pcDNA-ctrl-V5HisTopo (60). Plasmids pLenti-MERS-X (where X represents N, M, or E) were constructed by replacing the mGFP gene sequence of pLenti-C-mGFP (OriGene) with the DNA fragments of the V5- and His-tagged MERS-CoV structural proteins (N, M, and E) following BamHI/PmeI restriction endonuclease digestion. These pLenti-MERS-X plasmids were used to produce pseudotyped lentivirus particles expressing individual MERS-CoV structural proteins as previously described (61).

Expression plasmids pEF-BOS-Flag-RIG-I (a gift from Curt Horvath at Northwestern University, Addgene plasmid number 52877; http://n2t.net/addgene:52877; RRID: Addgene_52877) (62) and pcDNA-Flag-RIG-I-N-V5HisTopo (kindly provided by Yi-Ling Lin at Academia Sinica, Taipei, Taiwan) encode the full-length RIG-I protein and the N-terminal CARDs, respectively. Plasmids pFlag-CMV2-TRIM25 (pFLAG-CMV2-EFP, a gift from Dong-Er Zhang at the University of California, San Diego, Addgene plasmid number 12449; http://n2t.net/addgene:12449; RRID: Addgene_12449) (63) and pcDNA-HA-TRIM25 (kindly provided by Rei-Lin Kuo at Chang Gung University, Taoyuan, Taiwan) encode Flag-tagged and HA-tagged TRIM25, respectively. Plasmid pCMV-HA-Ub (kindly provided by Show-Li Chen at National Taiwan University, Taipei, Taiwan) (64) encodes HA-tagged ubiquitin.

Luciferase reporter plasmids were constructed as follows. Plasmid pIFNB(113)-Luc encodes firefly luciferase driven by the IFN-β promoter from nucleotide (nt) −113 to +20 (13). Plasmid pBV-IFNL(1200)-Luc encodes firefly luciferase driven by the IFN-λ1 promoter from nt −1200 to +1 (33) and was generated by inserting the corresponding DNA fragment into the KpnI/EcoRI sites of plasmid pBV-Luc-wt-MBS1-4 (a gift from Bert Vogelstein at Johns Hopkins University, Addgene plasmid number 16564; http://n2t.net/addgene:16564; RRID:Addgene_16564) (65). Plasmids pGL3-IFNL(303)-Luc and pGL3-IFNL(123)-Luc that encode firefly luciferase driven by IFN-λ1 promoters from nt −303 to +1 and from −123 to +1, respectively, were generated by inserting the corresponding DNA fragments into the KpnI/HindIII sites of plasmid pGL3-Basic (Promega). Plasmid pBV-CXCL10(1200)-Luc that expresses firefly luciferase driven by the CXCL10 promoter from nt −1200 to +1 was generated by inserting the corresponding DNA fragment into the KpnI/EcoRI sites of pBV-Luc-wt-MBS1-4. Cis-reporter plasmids pISRE-Luc (Agilent Technologies), pNF-ĸB-Luc (Agilent Technologies), and pIRF3-Luc (Nova Lifetech) contain five copies of ISRE, five copies of NF-ĸB binding sites, and four copies of IRF3 binding sites, respectively, to regulate the transcription of the luciferase gene. Plasmid pRL-TK that encodes Renilla luciferase driven by the HSV TK promoter (Promega) was used to serve as a transfection internal control in the luciferase assay system.

Antibodies.

Monoclonal anti-V5 antibody (R960-25) was purchased from Invitrogen, anti-His antibody (70796-3) was from Novagen, anti-FLAG antibody (F1804) was from Merck, and antibody against the HA tag (H3663) was from Sigma. Monoclonal antibodies against NF-κB p65 (number 8242), phospho-NF-κB p65 (number 3033), IRF3 (number 4302), and phospho-IRF3 (number 4947) were from Cell Signaling Technology. Polyclonal anti-GAPDH antibody (sc-25778) was from Santa Cruz Biotech, and anti-Sendai virus (ab33988) and anti-IFIT3 (ab76818) antibodies were from Abcam. Horseradish peroxidase-conjugated antibodies against mouse IgG (115-035-003), rabbit IgG (sc-2004), and chicken IgY (GTX224126-01) were from Jackson, Santa Cruz Biotech, and GeneTex, respectively.

Transient transfection and pseudotyped lentivirus transduction.

Transient transfection was performed in HEK293T cells with T-Pro NTR II transfection reagent (T-Pro Biotechnology) according to the manufacturer’s protocol. Briefly, expression plasmid was diluted into Opti-MEM (GIBCO) and mixed with transfection reagent at the ratio of 1 μg DNA to 2 μl transfection reagent. The cells were harvested 2 days posttransfection. For lentivirus transduction of A549 cells, pseudotyped lentivirus particles expressing individual MERS-CoV structural proteins were produced as previously described (61). The cells were harvested 3 days postransduction.

Sendai virus infection and IFN-β stimulation.

SeV activates RIG-I signaling and is often used as a model virus to stimulate IFN responses. SeV (8,000 to 16,000 HA unit/ml) was obtained following amplification in fertilized specific pathogen-free (SPF) eggs for 3 days at 37°C with 70% humidity. SeV infection (at 96 HA unit/ml) was conducted at 30 h posttransfection or 64 h after transduction to induce the IFN response in cultured cells. Alternatively, direct IFN-β stimulation was applied at 30 h posttransfection with exogenous human IFN-β at 1,000 units/ml.

Harvest of cell lysates and Western blot analysis.

For harvest of protein lysates, cells were lysed using a radioimmunoprecipitation assay (RIPA) buffer consisting of 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% sodium deoxycholate, 1% NP-40, 0.1% SDS, and 1% complete EDTA-free protease inhibitor (Roche). The lysates were resolved by polyacrylamide gel electrophoresis and electrotransferred onto an Immobilon-P membrane (Millipore) in transfer buffer (25 mM Tris-HCl, 192 mM glycine, and 20% methanol) at 300 mA for 90 min. The membrane was then blocked in blocking buffer (5% nonfat milk in phosphate-buffered saline [PBS] consisting of 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, and 1.8 mM KH₂PO₄) for 1 h and incubated with primary antibody at 4°C overnight. Following the incubation, the membrane was rinsed with PBST buffer (0.5% Tween 20 in PBS buffer) and incubated with horseradish peroxidase-conjugated secondary antibody at room temperature for 1 h. Specific interactions between antigens and antibodies were detected by the enhanced chemiluminescence system (Advansta).

RNA isolation and quantitative real-time PCR.

Total RNA was isolated from culture cells with TRIzol reagent (Invitrogen). Reverse transcription was performed with the RNA templates, ProtoScript II First Strand cDNA Synthesis kit (New England Bioloabs), and oligo(dT) primer. The products were subjected to quantitative real-time PCR (qRT-PCR) with primer sets of specific genes and SensiFAST SYBR No-ROX kit (Bioline). The primer sets used were 5′-AAATACAGCCCTTGTGCCTGG-3′ and 5′-GGTGAGCTGGCATACGAATCA-3′ for IFN-α2, 5′-GTCTCCTCCAAATTGCTCTC-3′ and 5′-ACAGGAGCTTCTGACACTGA-3′ for IFN-β, and 5′-CGCCTTGGAAGAGTCACTCA-3′ and 5′-GAAGCCTCAGGTCCCAATTC-3′ for IFN-λ1. The primer set 5′-GAAGGTGAAGGTCGGAGTC-3′ and 5′-GAAGATGGTGATGGGATTTC-3′ for GAPDH was used in parallel as an internal control. The results were analyzed with the CFX Connect Real-Time PCR Detection System (Bio-Rad).

Luciferase reporter assay.

For analysis of promoter activity, HEK293T cells were cotransfected with the firefly luciferase reporter plasmid driven by the IFN-β/IFN-λ1 promoter and the Renilla luciferase-expressing control plasmid pRL-TK in the presence or absence of specific effector plasmid such as pcDNA-MERS-N. At 48 h posttransfection, the luciferase reporter assay was conducted using the Dual-Glo luciferase assay system (Promega) and measured by luminometer. Where applied, SeV infection and IFN-β stimulation were conducted at 30 h posttransfection. The relative luciferase activity was determined by normalization of firefly luciferase activity against Renilla luciferase activity.

Immunoprecipitation and ubiquitination assay.

For immunoprecipitation, cells were harvested and homogenized in PBS with 0.5% NP-40. Protein A beads (GE Healthcare) and specific antibodies were then mixed with the cell lysates and incubated at 4°C overnight with gentle rotation. Following four washes in PBS with 0.5% NP-40, protein complexes conjugated on the beads were eluted by boiling in sample buffer and subjected to SDS-polyacrylamide gel electrophoresis and Western blotting. For the ubiquitination assay, immunoprecipitation was applied prior to Western blot analysis to enrich the ubiquitinated target proteins and exclude potential interference of ubiquitinated nonspecific proteins.

Rescue of suppressed replication of Sindbis virus-EGFP.

Recombinant Sindbis virus expressing enhanced green fluorescence protein (SINV-EGFP; kindly provided by Lih-Hwa Hwang at National Yang-Ming University, Taipei, Taiwan) was used to functionally examine the effect of MERS-CoV N protein on IFN production. The SINV-EGFP virus is an IFN-sensitive virus (66) for which it was demonstrated that SINV RNA species are more abundant in representative cell lines that have lower levels of IFNs. In addition, the expression of EGFP in SINV-EGFP-infected cells is closely correlated with the viral RNA replication (40). To perform the study, cultured supernatants were collected from HEK293T cells that had been transfected with plasmid encoding MERS-CoV N protein or control vector and infected with SeV. The cultured supernatants were exposed to UV radiation for 2 h and used as conditioned media to culture naive HEK293T cells seeded on coverslips. Following replacement of the conditioned medium with fresh medium the next day, the HEK293T cells were infected with SINV-EGFP at 3 MOI. At 48 h after the infection, cells fixed on the coverslips were immersed in Hoechst 33342 staining dye (Abnova) for 1 h and subjected to microscopy. Meanwhile, Western blot analysis was performed to examine the expression levels of MERS-CoV N protein and EGFP.

Statistical analysis.

Each experiment was repeated at least three times. Analysis was performed using Microsoft Excel 2016. Statistic comparisons were analyzed by two-tailed Student’s t tests and presented as means ± standard deviations (SD) with a P value of <0.05 considered statistically significant.