Highly Pathogenic Porcine Reproductive and Respiratory Syndrome Virus Nsp4 Cleaves VISA to Impair Antiviral Responses Mediated by RIG-I-like Receptors

Porcine reproductive and respiratory syndrome virus (PRRSV) is one of the most significant etiological agents in the swine industry worldwide. It has been reported that PRRSV infection can modulate host immune responses, and innate immune evasion is thought to play a vital role in PRRSV pathogenesis. In this study, we demonstrated that highly pathogenic PRRSV (HP-PRRSV) infection specifically down-regulated virus-induced signaling adaptor (VISA), a unique adaptor molecule that is essential for retinoic acid induced gene-I (RIG-I) and melanoma differentiation associated gene 5 (MDA5) signal transduction. Moreover, we verified that nsp4 inhibited IRF3 activation induced by signaling molecules, including RIG-I, MDA5, VISA, and TBK1, but not IRF3. Subsequently, we demonstrated that HP-PRRSV nsp4 down-regulated VISA and suppressed type I IFN induction. Importantly, VISA was cleaved by nsp4 and released from mitochondrial membrane, which interrupted the downstream signaling of VISA. However, catalytically inactive mutant of nsp4 abolished its ability to cleave VISA. Interestingly, nsp4 of typical PRRSV strain CH-1a had no effect on VISA. Taken together, these findings reveal a strategy evolved by HP-PRRSV to counteract anti-viral innate immune signaling, which complements the known PRRSV-mediated immune-evasion mechanisms.


IFNβ induction and signaling is inhibited by HP-PRRSV infection.
In the previous study, we verified that HP-PRRSV inhibited the production of IFNα and IFNβ in PAMs by qPCR 30 . To identify the transcription factors which participate in the effect of HP-PRRSV on IFNβ induction and signaling, Marc-145 cells were transfected with IFNβ , IRFs, NF-κ B or ISRE response elements luciferase reporter plasmid. Six hours later, Marc-145 cells were infected with HP-PRRSV strain JXwn06, and then transfected with or without poly(I:C). At 8 h post-transfection, luciferase activities were examined. As expected, HP-PRRSV infection reduced IFNβ promoter activity induced by poly(I:C) (Fig. 1A). In addition, IRFs, NF-κ B and ISRE responsive promoter activities induced by poly(I:C) were also affected by HP-PRRSV infection with the down-regulation of 57%, 45% and 43% compared to that of mock infection, respectively (Fig. 1B-D). We also infected porcine alveolar macrophages (PAMs) with HP-PRRSV, and analyzed the mRNA levels of ISGs using qPCR. As shown in Fig. 1E and F, HP-PRRSV significantly suppressed the expression of myxovirus resistance 1 and 2′ ,5′ -oligoadenylate synthetase to 12% (Mx1) (Fig. 1E) and 18% (OAS) (Fig. 1F) of that induced by poly(I:C). These data suggest that HP-PRRSV infection antagonizes IFNβ induction and downstream signaling pathway.
HP-PRRSV nsp4 blocks IRF3 signaling pathway. IRF3 is the most significant transcription factor in IFNβ induction. To confirm and clarify whether and how nsp4 inhibits IRF3 signaling pathway, we analyzed several key steps of the IRF3 activation process, such as IRF3 phosphorylation and nuclear translocation, upon poly(I:C) treatment. As shown in Fig. 2A, treatment with poly(I:C) activated IRF3 phosphorylation in 3D4/21 cells. However, addition of nsp4 reduced the phosphorylation level of IRF3 in a dose-dependent manner, suggesting that nsp4 inhibits IFNβ induction upstream of IRF3 activation. To confirm this result, nuclei protein (N.P) and cytoplasma protein (C.P) fractions were separated for analyzing the distribution of IRF3 (Fig. 2B). The results showed that nsp4 expression led to a pronounced, dose-dependent decrease in poly(I:C)-induced IRF3 nuclear translocation.
Subsequently, we examined IRF3 phosphorylation during the course of HP-PRRSV infection. As expected, poly(I:C) activated IRF3 phosphorylation in mock infected cells. However, the activation was remarkably impaired in cells infected with HP-PRRSV (Fig. 2C). In consistent with this result, nuclear translocation of IRF3 was also affected. As shown in Fig. 2D, the level of IRF3 presented in the nuclear fraction increased in the presence of poly(I:C), whereas nuclear translocation of IRF3 was down-regulated in PRRSV-infected PAMs. Taken together, these data show that HP-PRRSV and its protein nsp4 interfere with poly(I:C) activated IRF3 signaling pathway by reducing IRF3 phosphorylation and translocation to the nucleus.
HP-PRRSV nsp4 disrupts RIG-I like receptor signaling pathway. RIG-I and RLR protein family are key cytoplasmic receptors that are implicated in pathogen sensing of virus infection to initiate and modulate antiviral immune responses 3,4 . Arteriviruses were shown to be primarily detected by MDA5, while RIG-I might also be involved under some circumstances 25 . Given the pivotal role of RLRs in mediating IFN-I production and ISGs expression, we further investigated whether over-expression of HP-PRRSV nsp4 inhibits RLRs-mediated signaling. Over-expression of RIG-I or MDA5 stimulated IRFs responsive promoter activity compared with that of the vector control (Fig. 3A,B). However, activation of IRFs responsive promoter by RIG-I or MDA5 was significantly down-regulated to 75% and 60% in the presence of nsp4, respectively. Similar results were obtained when IRFs responsive promoter was activated by overexpression of TBK1, the downstream kinase for RIG-I/MDA5 Scientific RepoRts | 6:28497 | DOI: 10.1038/srep28497 (Fig. 3C). IRFs responsive promoter luciferase activity induced by TBK1 was suppressed about 48% by nsp4. In contrast, activation of IRFs responsive promoter by over-expression IRF3 was not affected by nsp4 (Fig. 3D). These results suggest that HP-PRRSV nsp4 disrupts RLR signaling pathway at a step upstream of IRF3.

HP-PRRSV nsp4 inhibits IFN responses induced by VISA.
Sensing of virus by the RLRs engages a complex signaling cascade that utilizes VISA adapter protein to initiate the innate host antiviral and inflammatory responses against pathogen infection 31 . VISA is central to RLR signaling pathway and orchestrates the , and pRL-TK. pRL-TK was used as an internal control of transfection efficiency. Six hours later, cells were mock infected or infected with HP-PRRSV (JXwn06) at a multiplicity of infection (MOI) of 0.1 for 24 h, and then transfected with or without poly(I:C) for 8 h. Total cell lysates were assayed for luciferase activities. (E,F) Porcine alveolar macrophages (PAMs) were mock infected or infected with HP-PRRSV at an MOI of 0.1 for 24 h, and then treated with or without poly(I:C) (10 μ g/ml) for 6 h. Total RNAs were extracted and qPCR was performed for analyzing the expression of Mx1 (E) and OAS (F). GAPDH was used as an internal control. Data were representative of three independent experiments (mean ± SD). Statistical analysis was performed by Student's t test. * * P < 0.01; * * * P < 0.001.
Scientific RepoRts | 6:28497 | DOI: 10.1038/srep28497 ordered recruitment of various signaling molecules to create an antiviral platform 32,33 . To determine whether nsp4 inhibits IFN responses induced by VISA, we transfected expression vectors encoding nsp4 and VISA into 3D4/21 cells together with a luciferase reporter driven by the IFNβ promoter, IRFs responsive promoter or ISRE responsive promoter. As shown in Fig. 4A, over-expression of VISA significantly up-regulated IFNβ promoter activity to about 600-fold compared with that of the vector control. However, co-transfection of nsp4 could result in a 52% decrease of luciferase activity relative to transfected with VISA only. In consistent with this observation, IRF3 and ISRE responsive promoter activity was stimulated in cells transfected with VISA construct, while the luciferase activity was suppressed about 37% and 66% in the presence of nsp4, respectively (Fig. 4B,C). These results suggest that HP-PRRSV nsp4 blocks IFNβ induction and downstream signaling by VISA.
HP-PRRSV nsp4 reduces endogenous VISA expression. The data presented above show that HP-PRRSV nsp4 might inhibit RLR signaling at a step upstream of IRF3. Considering that VISA plays a pivotal role in RLR signaling and a variety of viruses evolve mechanisms to disrupt VISA, we then investigated whether HP-PRRSV infection affected VISA expression. PAMs were mock infected or infected with HP-PRRSV strain JXwn06. Thirty-six hours post-infection, cells were harvested and examined by Western blot analysis. As shown in Fig. 5A, the level of endogenous VISA protein decreased in a dose-dependent manner. In contrast, TBK1 and IRF3 exhibited no reduction in HP-PRRSV-infected PAMs. To examine at which level HP-PRRSV affects VISA expression, the mRNA level of VISA was analyzed using qPCR. Result showed that HP-PRRSV had no effect on the mRNA level of VISA (Fig. 5B), implicating that HP-PRRSV infection might affect the protein level of VISA.
Next, we analyzed the effect of HP-PRRSV nsp4 on VISA expression. Expression vector encoding nsp4 was transfected into porcine 3D4/21 cells and Western blot analysis was performed. As shown in Fig. 5C, nsp4 reduced the level of endogenous VISA in a dose-dependent manner when over-expressed in cells. This coincided well with the progress of HP-PRRSV infection. To determine whether the protease activity of nsp4 is required for this reduction, we constructed a series of nsp4 mutants, including mutation of the canonical catalytic triad of His39-Asp64-Ser118 (3A) and one of three domains located at amino acids 1-69 (N69), 89-153(MID), and 157-199 (C157), respectively 34 . Results showed that these protease-dead mutants completely lost its ability to down-regulate the protein level of VISA (Fig. 5D). Taken together, these results indicate that HP-PRRSV, as well as its protein nsp4, reduces endogenous VISA level. VISA is cleaved from mitochondria by HP-PRRSV nsp4. The above findings that the protease activity is critical for HP-PRRSV nsp4-mediated reduction of endogenous VISA raise the possibility that nsp4 might target VISA for cleavage. To verify this possibility, we cloned porcine VISA gene into a mammalian expression vector with N-Flag tag. The expression vector encoding VISA was co-transfected into HeLa cells with empty or HP-PRRSV nsp4-expressed vector. As shown in Fig. 6A, the protein abundance of full-length Flag-VISA was decreased as the level of HP-PRRSV nsp4 increased. This was accompanied by the appearance of a smaller anti-Flag-reactive band, presumably a VISA cleavage product.
VISA is primarily localized on the outer mitochondrial membrane, and this sub-cellular localization is essential for its function in RLR antiviral signaling 35 . We therefore investigated whether any changes to the cellular distribution of VISA occurred during nsp4 expression. Expression vector encoding Flag-VISA along with c-Myc-nsp4 or nsp4-3A was transfected into HeLa cells. Twenty-four hours post-transfection, cells were stained with the corresponding antibodies followed by imaging with a laser scanning confocal microscope. The results showed that VISA co-localized with a mitochondrial marker in control cells. When wild-type nsp4 was . A plasmid expressing pRL-TK was used as a control. Twenty-four hours later, cell lysates were assayed for luciferase activities. In addition, lysates of cells were subjected to Western blot analysis with antibodies against Flag, c-Myc and β -actin to show the expression of these signaling molecules and nsp4. Data were representative of three independent experiments with triplicate samples (mean ± SD). Statistical analysis was performed by Student's t test. * P < 0.1; * * P < 0.01; NS, not significant.
co-expressed with VISA, the majority of VISA became cytosolic, as revealed by Flag-antibody that detects the N terminus of VISA. In sharp contrast, when nsp4 mutant (3A) and VISA were co-expressed, VISA co-localized  on the mitochondrial membrane and revealed an extensive overlapping staining pattern (Fig. 6B). These data indicate that HP-PRRSV nsp4 mediates VISA cleavage from mitochondrial membrane. nsp4 of typical PRRSV strain CH-1a has no effect on VISA. Recent study reported that nsp4 of different pathogenic PRRSV isolates exhibited differential inhibitory effect on IFNβ transcription activation 36 . To confirm these results, vectors encoding JXwn06 nsp4 or CH-1a nsp4 were transfected into 3D4/21 cells along with a luciferase reporter driven by the IFNβ promoter, IRFs responsive promoter or NF-κ B responsive promoter. Compared to nsp4 of typical PRRSV strain CH-1a, nsp4 of HP-PRRSV strain JXwn06 had a greater ability to suppress IFNβ , IRFs and NF-κ B luciferase activities induced by poly(I:C) (Fig. 7A-C). We then tried to verify whether the two PRRSV isolates have different effects on VISA. As shown in Fig. 7D, JXwn06 infection significantly down-regulated VISA expression, while CH-1a had no effect on VISA expression. In consistent with this observation, over-expression JXwn06 nsp4 reduced VISA expression, whereas over-expression of CH-1a nsp4 had no impact on the level of endogenous VISA protein (Fig. 7E). In addition, confocal microscopy analysis revealed that VISA co-localized on the mitochondrial membrane in the presence of CH-1a nsp4, indicating that CH-1a nsp4 has no ability to mediate the cleavage of exogenous porcine VISA (Fig. 7F). Taken together, these data show that nsp4 of different pathogenic PRRSV strains display various effects on VISA.
Collectively, these data demonstrate that HP-PRRSV nsp4 mediates the cleavage of VISA and dislodges it from the mitochondria, thus impairing IFNβ induction and RLR signaling. However, CH-1a nsp4 has no effect on VISA.

Discussion
HP-PRRSV causes porcine high fever syndrome (PHFS), which is characterized by high transmission efficiency, high morbidity and high mortality. Modulated and hijacked host immune responses are thought to facilitate viral pathogenesis. Sensing cytoplasmic RNA to initiate the induction of IFN-I by RLRs is a vital part of the innate antiviral responses. Therefore, the RLR signaling pathway is targeted by numerous viruses. In this work, we show that HP-PRRSV, as well as its protein nsp4, disrupts IFN-I induction and downstream signaling by mediating the Figure 7. CH-1a nsp4 has no effect on VISA cleavage. (A-C) nsp4 of JXwn06 or CH-1a isolate expression plasmid was transfected into 3D4/21 cells along with pGL3-IFNβ -Luc (A), pGL3-IRFs-Luc (B) or pGL3-NF-κ B-Luc (C), and pRL-TK. pRL-TK was used as an internal control of transfection efficiency. Twenty-four hours later, cells were stimulated with or without poly(I:C) (10 μ g/ml) for 8 h and analyzed using dual-luciferase assay. (D) PAMs were mock infected or infected with JXwn06 or CH-1a at an MOI of 1 for 24 h, and cells were harvested and performed the expression of VISA and nsp4 by Western blot analysis. (E) 3D4/21 cells were transfected with JXwn06 nsp4 or CH-1a nsp4 expression plasmid for 24 h, and then cell lysates were analyzed by Western blotting with antibodies for VISA, nsp4 and β -actin. β -actin was set up as a loading control. (F) HeLa cell treatment and confocal microscopy were performed as described in the legend to Fig. 7B. Data are mean ± SD from three independent experiments. Differences were evaluated by Student's t test. * P < 0.05; * * P < 0.01; * * * P < 0.001.
IFN-I are a family of cytokines that represent one of the first lines of defense to limit viral propagation and spread 13 . Secreted IFN-I trigger the activation of JAK-STAT signaling and then induce a large array of ISGs, leading to a remarkable antiviral state in cells 37 . Many ISGs control viral infection by directly interfering with different stages of viral life cycles or modulating antiviral immune responses. For instance, Mx1 traps nucleocapsids and prevents endocytosis of incoming virus particles. OAS catalyzes the synthesis of 2′ , 5′ -oligoadenylates, resulting in the activation of RNaseL, which mediates host and viral RNA degradation. Degraded RNA can activate cytoplasmic PRRs, such as RIG-I and MDA5, to reinforce the innate antiviral immunity 38,39 . In addition, IFN-I have been shown to play a significant role in shaping the adaptive immunity by promoting the development of antigen-specific T and B lymphocyte responses and immunological memory 40 . In consideration of the formidable antiviral function of IFN-I, it is not surprising that many viruses have evolved mechanisms to either prevent IFN-I induction or block the downstream signaling. Previous studies have shown that PRRSV has evolved a variety of strategies to inhibit and manipulate IFN-I responses 41,42 . Kim et al. have reported that nsp1 mediates CREB-binding protein (CBP) degradation to inhibit IRF3 association with CBP in the nucleus, resulting in the inhibition of IFN-I transcriptional activation 43 . PRRSV nsp2 has the ability to inhibit NF-κ B activation by suppressing Iκ Bα degradation 44 . nsp2 also has the effect on interfering with ISG15 conjugation to cellular proteins, impairing function of the antiviral ISG 45 . In the present study, we further verified that HP-PRRSV antagonized the production of IFN-I by inhibition of IRF3 activation. HP-PRRSV infection reduced the protein level of VISA, resulting in the disruption of RLR downstream signal transduction. In addition, we showed that HP-PRRSV infection inhibited the expression of ISGs induced by poly(I:C) (Fig. 1E,F). Except the inhibition of IFN-I expression, another possible strategy used by HP-PRRSV could be to suppress IFN-I downstream signaling. To verify this hypothesis, we used IFNα to induce JAK-STAT signaling pathway, and found that ISGs transcription was suppressed by HP-PRRSV (data not shown). This finding is in consistent with a previous study, in which the typical PRRSV strain VR2385 is used 46 .
RLRs are essential cytoplasmic pathogen recognition receptors that impart recognition of RNA viruses across genera and virus families, including Paramyxoviridae, Picornaviridae, Coronaviridae, Arteriviridae, etc 24 . Anti-virus response by the RLRs relies on the VISA adaptor protein to engage a high-energy signalosome that initiates downstream activation of transcriptional responses, inducing expression of IFN-I and immune modulatory genes to control virus propagation and spread 32 . VISA plays a central role in the induction of antiviral and inflammatory immune responses. Additionally, VISA appears to be involved in the coordination of metabolic and apoptotic functions 47 . Previous study reveals that mice deficient in VISA fail to mount a robust IFN-I responses and are highly susceptible to viral infection 48 . Thus, targeting VISA is likely to be an evolutionarily conserved and effective mechanism to suppress IFN-I transcription by viruses. Hepatitis C virus (HCV) NS3-4A protease is shown to inhibit VISA by cleaving it from the mitochondrial membrane and preventing the induction of IFNβ 49 . Wei et al. demonstrates that Hepatitis B virus (HBV) X protein promotes polyubiquitin conjugation to VISA, resulting in its degradation 50 . Human Metapneumovirus M2-2 protein is shown to inhibit VISA signaling and IFNβ production by interacting with VISA, which prevents the recruitment of VISA to RIG-I 51 . Here, we demonstrated that HP-PRRSV nsp4 protease inhibited IFN-I induction and signaling induced by VISA over-expression (Fig. 4). Furthermore, we showed that nsp4 had the ability to inactivate VISA by cleaving and dislocating it from the outer mitochondrial membrane. This finding reveals a strategy used by HP-PRRSV to antagonize IFN-I antiviral responses. Our previous study has revealed that nsp4 mediates the cleavage of NEMO, and then blocks the NF-κ B signaling pathway 30 . Considering that VISA locates upstream of IRF3 and NF-κ B, targeting VISA might be more efficient for HP-PRRSV to suppress IFN induction and downstream signaling. Recently, Dong et al. reported that nsp4 of HP-PRRSV strain WUH3 cleaved human VISA at the residue Glu-268 52 . However, amino acid analysis shows that the homology between human VISA and porcine VISA is only about 55.4%, and the cleavage site on the human protein does not exist in porcine VISA. Since pigs are the only nature host of PRRSV, we investigated the effect of PRRSV on porcine VISA, which might reflect PRRSV-host interactions in vivo.
In comparison with the typical PRRS, the atypical PRRS caused by HP-PRRSV was characterized by quickly widespread, high fever, high morbidity, and high mortality in pigs of all ages 53 . The rapid molecular evolution of PRRSV results in a diverse composition of isolates with multifarious pathogenicity 54 . A series of studies have shown that different strains exhibit various effects on modulation of IFN-I antiviral responses. Most PRRSV strains have been shown to antagonize the induction of IFN-I 55,56 . However, a novel PRRSV isolate (A2MC2) induces the production of IFN-I and appears to have no inhibitory effect on antiviral response induced by IFNα 57 . Wang et al. reported that nsp1β of PRRSV strains VR-2332 and VR-2385 blocked ISGF3 nuclear translocation by inducing KPNA1 degradation, whereas Ingelvac PRRS modified live virus (MLV) had no effect on KPNA1 46 . Recent study revealed that HP-PRRSV nsp4 could display stronger inhibitory effect on IFNβ transcription than nsp4 of typical PRRSV isolates 36 . However, the precise molecule basis for this difference remains indistinct. In this work, we demonstrated that HP-PRRSV infection impaired VISA expression, while typical PRRSV strain CH-1a did not. In consistent with this observation, over-expression nsp4 of CH-1a exhibited no effect on VISA cleavage (Fig. 7D-F). These results indicated that the pathogenicity of PRRSV is likely related to its ability to evade host innate immunity. By comparing the amino acid sequences of HP-PRRSV strain JXwn06 nsp4 and the typical PRRSV strain CH-1a nsp4, we found there were 7 different amino acids, which might be responsible for the distinct effect on VISA. In addition, over-expression of CH-1a nsp4 in 3D4/21 cells suppressed IFNβ induction by blocking both NF-κ B and IRF3 signaling pathways (Fig. 7A-C), suggesting that other strategies exist for nsp4 to inhibit IFNβ expression.
In conclusion, we identified that HP-PRRSV infection could antagonize IFNβ induction and signaling by reducing the protein level of VISA in infected cells. In addition, our data showed that nsp4 protein interfered with RLR antiviral signaling and IFNβ transcriptional activation by mediating VISA cleavage. Interestingly, we Scientific RepoRts | 6:28497 | DOI: 10.1038/srep28497 found that typical PRRSV strain CH-1a and its protein nsp4 had no effect on VISA. These findings might help us understand the molecular mechanisms of PRRSV immune evasion and develop countermeasures to control HP-PRRSV infection in the future.
HP-PRRSV strain JXwn06 (a highly pathogenic PRRSV strain isolated in Jiangxi Province; GenBank accession, EF641008.1) was propagated in PAMs, and typical PRRSV strain CH-1a (the first type 2 PRRSV strain isolated in China; GenBank accession, AY032626.1) was propagated in Marc-145 cells. Virus preparations were titrated on PAMs and Marc-145 cells, and then stored at − 80°C until use.
Plasmid construction. Porcine IFNβ -luciferase reporter plasmid was constructed using pGL3-Basic vector as described elsewhere 30 . Transcription factor IRF3, NF-κ B and ISRE responding elements were synthesized and annealed to form double-stand DNA, and then separately cloned into pGL3-basic vector 58 . pRL-TK containing the Renilla luciferase was used as a normalization control. The protein expression plasmids pCMV-JXwn06-nsp4, pCMV-JXwn06-nsp4 mutants (3A, N69, MID, C157), pRK-Flag-RIG-I and pRK-Flag-IRF3 have been described previously 34,59,60 . To construct MDA5 and TBK1, cDNA fragments were amplified and cloned into pRK5-Flag at BamHI and SalI sites. To construct plasmid CH-1a-nsp4, fragment of PRRSV strain CH-1a cDNA was cloned into the pCMV-Myc vector at EcoRI and XhoI sites. cDNA encoding porcine VISA was amplified using reverse transcription PCR (RT-PCR) from total RNAs extracted from PAMs and cloned into pRK5-Flag at XbaI and HindIII sites. All the primers are listed in Table 1.

Antibodies and reagents.
Rabbit antibodies directed against TBK1 (D1B4), IRF3 (D83B9) and pIRF3 (4D4G) were purchased from Cell Signaling Technology. Rabbit anti-VISA (also known as MAVS, polyclonal antibody raised against residues 1-13 of MAVS), mouse anti-mitochondria (MTC02, recognizes a 60 kD non-glycosylated protein component of mitochondria) and mouse anti-c-Myc antibodies were obtained from Abcam. Rabbit polyclonal antibody to Flag-tag and mouse monoclonal antibody to β -actin were purchased from Sigma. Antibodies against Histone3 and α -tubulin have been previously described 61 . The antisera of nsp4 and GP5 were prepared by our lab. HRP-conjugated goat anti-mouse or anti-rabbit secondary antibodies for Western blot were purchased from Santa Cruz. FITC-conjugated goat anti-rabbit and TRITC-conjugated goat anti-mouse secondary antibodies for confocal microscopy were purchased from Jackson ImmunoResearch. Poly(I:C) was purchased from InvivoGen.

RNA isolation and quantitative real-time PCR (qPCR).
RNA isolation and qPCR were performed as described previously 62 . Briefly, PAMs were infected with PRRSV at an MOI of 0.1 for 24 h, and then treated with or without poly(I:C) (10 μ g/ml) for 6 h. Total RNAs were extracted with TRIzol (Invitrogen) and used for cDNA synthesis using M-MLV reverse transcriptase according to the manufacturer's instructions (Takara). Quantitative RT-PCR (qPCR) analysis was performed using FastSYBR Mixture with ROX (Cwbiotech) on the ViiA TM 7 real-time PCR System (Applied Biosystems). Gene-specific primers for Mx1, OAS, VISA and GAPDH were designed and listed in Table 1. The expression of Mx1 and OAS was normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and presented as fold induction relative to the control. VISA expression was normalized to the mock infection.
Western blot analysis. Whole-cell extracts were lysed in RIPA lysis buffer (Cwbiotech) supplemented with 100 U proteinase cocktail (Cwbiotech) and 20 μ M phosphatase inhibitor. Cytosol and nuclear protein samples were prepared with Nuclear and Cytosol Fractionation Kit (Beyotime). Protein levels in each sample were quantified with BCA assay kit (Pierce Biotechnology, Inc.). Similar amounts of proteins from each extract were fractionated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore). After blocking with 5% milk in PBS with 0.05% Tween-20 (PBST), membranes were incubated for 2 h at room temperature with the primary antibodies at a suitable dilution as recommended (anti-TBK1, IRF3, p-IRF3, VISA and Histone3 at 1:1000; anti-β -actin, α -tubulin, Flag and c-Myc at 1:2000). The membranes were then incubated with the appropriate secondary antibodies for 1 h at a dilution of 1:10,000. The antibodies were visualized by use of the ECL reagent according to the manufacturer's protocols.
Confocal microscopy. HeLa cells cultured in microscope cover glass (Fisher Scientific) were washed with ice-cold phosphate buffered saline (PBS) and fixed with 4% paraformaldehyde. Cells were permeabilized with 0.2% Triton X-100 for 3 min and blocked with 1% bovine serum albumin (BSA) for 1 h at room temperature (RT). Cells were then incubated with the indicated primary antibodies (anti-mitochondria and Flag at 1:200 dilution) for 1 h at RT. Following washing, cells were incubated with proper secondary antibodies (FITC-conjugated goat anti-rabbit or TRITC-conjugated goat anti-mouse IgG) for 45 min at RT, washed and stained with 4, 6-diamidino-2-phenylindole (DAPI) to detect nuclei. Immunofluorescence was performed with a Nikon A1 confocal microscope.

Statistical analysis.
Results are presented as means ± SD for at least three independent experiments. Data were analyzed by GraphPad Prism software using Student's t test. Differences in data were considered to be statistically significant if the P value is less than 0.05. * P < 0.05; * * P < 0.01; * * * P < 0.001.