Inhibition of Beta Interferon Induction by Severe Acute Respiratory Syndrome Coronavirus Suggests a Two-Step Model for Activation of Interferon Regulatory Factor 3

Cells and viruses.

Simian VeroE6 cells and human 293 cells were maintained and grown as previously described (59). The FFM-1 isolate of SARS-CoV was kindly provided by Stephan Becker, University of Marburg, Germany. Bunyamwera delNSs virus (67) was used as an IFN-inducing control virus.

Generation of antibodies.

Monoclonal antibodies against human IRF-3 were generated by immunizing mice with a bacterially expressed glutathione S-transferase fusion protein containing the C-terminal 100 amino acids (328 to 427) of human IRF-3. Hybridomas were screened on 96-well plates coated with bacterially expressed full-length human IRF-3 protein, and two specific monoclonal antibodies (17C2 and 6C7) were obtained. To generate a polyclonal antiserum against SARS-CoV, the N cDNA sequence (Urbani strain) was amplified by reverse transcription-PCR, with RNA from infected Vero cells as the template. Primer sequences were 5′-CCGGGAATTCACCATGTCTGATAATGGACCCCAATCA-3′ and 5′-GCGCGCATGCTTATGCCTGAGTTGAATCAGCAGA-3′. The reverse transcription-PCR product was cloned into pET28a (Novagene) with EcoRI and XhoI restriction sites. pET28a/N was transformed into Escherichia coli strain BL21(DE3) to produce recombinant N protein. After purification by Ni-nitrilotriacetic acid-agarose beads (Qiagen), a rabbit was immunized and serum was collected after two boosts (Cocalico Biologicals Inc., Reamstown, Pa.). The specificity of all sera was tested by immunofluorescence and Western blot analysis.

IFN reporter assays.

Subconfluent 293 cells were transfected with 0.5 μg of the IFN-β promoter reporter construct p-125Luc (kindly provided by Takashi Fujita, Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan) and 0.05 μg of control plasmid pRL-SV40 (Promega) mixed with 2.5 μl of DAC-30 (Eurogentec) in 200 μl of Optimem (Gibco-BRL). After 6 h at 37°C, the liposome-DNA mixture was removed, and cells were infected at a multiplicity of infection of 5. At 18 h postinfection, cells were harvested and lysed in 100 μl of passive lysis buffer (Promega). An aliquot of 20 μl of lysate was used to measure firefly and Renilla luciferase activities according to the manufacturer's instructions. The firefly luciferase activities were normalized to the corresponding Renilla luciferase activities to determine induction.

Reverse transcription-PCR analyses.

Cells were infected for 18 h and total RNA was extracted. For reverse transcription, 1 μg of RNA was incubated with 200 U of Superscript II reverse transcriptase (Gibco-BRL) and 100 ng of random hexanucleotides in 20 μl of 1× reverse transcription buffer (Gibco-BRL) supplied with 1 mM each of the four deoxynucleotide triphosphates, 20 U of RNasin, and 10 mM dithiothreitol. The resulting cDNA was amplified by 35 cycles of PCR, with each cycle consisting of 30 s at 94°C, 1 min at 58°C, and 1 min at 72°C, followed by 10 min at 72°C. The upstream and downstream primers for amplifying mRNA sequences were 5′-GACGCCGCATTGACCATCTA-3′ and 5′-CCTTAGGATTTCCACTCTGACT-3′ (IFN-β), 5′-ATGTCTGATAATGGACCCCAATCAAACCAA-3′ and 5′-TTATGCCTGAGTTGAATCAGCAGAAGCTCCA-3′ (SARS-CoV N), 5′-ATGATTGAGTTGGAATTTCATGAT-3′ and 5′-CCCTGCTTACATGTTGATTCCGAATTTAGCAA-3′ (Bunyamwera virus N), and 5′-GCCGGTCGCAATGGAAGAAGA-3′ and 5′-CATGGCCGGGGTGTTGAAGGTC-3′ (γ-actin).

IRF-3 nuclear translocation assay.

Endogenous IRF-3 can be detected in Vero cells by enzymatic amplification of the immunofluorescence signal (32). Cells were grown on coverslips and infected with SARS-CoV at a multiplicity of infection of 5. After an incubation period of 8 h or 16 h, cells were fixed with 3% paraformaldehyde and permeabilized with 0.5% Triton X-100 dissolved in phosphate-buffered saline (PBS). Cells were washed three times with PBS and incubated with the primary antibodies, polyclonal rabbit anti-SARS-CoV N protein and monoclonal mouse anti-IRF-3 antibody 17C2, diluted 1:1,000 and 1:500, respectively, in TNB blocking buffer (Perkin Elmer). After incubation at room temperature for 1 h, the coverslips were washed three times in PBS and then treated with the secondary antibodies, Cy3-conjugated goat anti-rabbit immunoglobulin G (Alexis) and biotin-conjugated anti-mouse immunoglobulin (Perkin Elmer) at a dilution of 1:200 each. Cells were again washed three times in PBS and incubated for 30 min with streptavidin-conjugated horseradish peroxidase (Perkin Elmer) at a dilution of 1:100. After washing in PBS, cells were incubated for 5 min in Fluorophore Tyramide amplification reagent (Perkin Elmer), then washed in PBS, and mounted with Fluorsave solution (Calbiochem). Stained cell samples were examined with a Leica confocal laser scanning microscope with an ×63 NA1.4 objective.

IRF-3 hyperphosphorylation assay.

Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris, pH 7.6, 150 mM NaCl, 1% NP-40) containing protease inhibitors (Complete protease inhibitor; Roche) and phosphatase inhibitors (phosphatase inhibitor cocktail II; Calbiochem). A total of 50 μg of protein was separated by electrophoresis on a 5% to 12% polyacrylamide gradient gel and blotted onto an Immobilon-P membrane (Millipore) following the manufacturer's instructions (Bio-Rad). The membrane was incubated with a 1:500 dilution of polyclonal anti-IRF-3 antibody FL-425 (Santa Cruz Biotechnology). Protein bands were visualized with the ECL Plus method (Amersham).

IRF-3 dimerization assay.

Cells were lysed in buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% NP-40, protease inhibitors, and phosphatase inhibitors, vortexed, incubated on ice for 10 min, and then centrifuged at 4°C for 5 min at 10,000 × g. Proteins were then separated by electrophoresis in a 10% nondenaturing polyacrylamide gel, with 1% deoxycholate in the cathode buffer (29). IRF-3 monomers and dimers were detected by Western blot analysis with polyclonal anti-IRF-3 antibody FL-425 (Santa Cruz Biotechnology) diluted 1:500.

Coimmunoprecipitation assay.

Human 293 cells (2 × 10⁶) were grown to 80% confluency and infected with SARS-CoV at a multiplicity of infection of 5. After an incubation period, cells were washed in PBS and incubated in 200 μl of RIPA buffer containing protease and phosphatase inhibitors. In parallel, 30 μl of a 50% slurry of protein G-Sepharose in RIPA buffer were preadsorbed with 2 μg of polyclonal anti-CBP antibody A-22 (Santa Cruz Biotechnology) and incubated for 2 h at 4°C. Then, the Sepharose beads were washed three times with RIPA buffer and incubated with the cell lysate for 2 h at 4°C. The immunoprecipitates were washed three times with RIPA buffer, and bound proteins were subjected to Western blot analysis with the mouse monoclonal anti-IRF-3 antibody SL12 (Pharmingen), diluted 1:500, or the anti-CBP antibody A-22, diluted 1:200. As infection controls, supernatants from the first precipitation step were analyzed by Western blot with rabbit polyclonal serum against SARS-CoV N protein (1:2,000) or Bunyamwera virus N (1:500).

Nuclear localization assays.

The human IRF-2 cDNA was amplified from HeLa cell total RNA by reverse transcription-PCR with oligo(dT) primers for reverse transcription and primers hIRF-2/ClaI/5′ (CGCGATCGATACCATGCCGGTGGAAAGGATGCGC) and hIRF-2/KpnI/3′ (CGCGGGTACCTTAACAGCTCTTGACGCGGGC) for PCR. The PCR product was cloned into the pEGFP-C1 expression plasmid (Clontech) with ClaI and KpnI restriction enzymes. The resulting plasmid, termed pEGFP-C1-hIRF-2, expresses the green fluorescent protein (GFP) fused to IRF-2. Vero cells were transfected with 2 μg of pEGFP-C1-hIRF-2 with DAC-30 liposomes as described above. After 24 h at 37°C, cells were infected with a multiplicity of infection of 5, and 16 h postinfection, cells were immunostained for SARS-CoV N protein as indicated. Cell samples were examined with a Leica confocal laser scanning microscope with an ×63 NA1.4 objective.

Nuclear export assays.

Vero cells were transfected with 850 ng of the Rev expression construct pcRev, 100 ng of the reporter plasmid pDM128 (19), and 50 ng of the control plasmid pRL-SV40 (Promega) expressing Renilla luciferase. pDM128 generates a chloramphenicol acetyltransferase (CAT) mRNA which is only functional if exported out of the nucleus by Rev in an unspliced form. After 4 h at 37°C, transfected cells were infected with a multiplicity of infection of 5, and 16 h postinfection, cells were harvested and lysed in 100 μl of passive lysis buffer (Promega). An aliquot of 20 μl of lysate was used to measure CAT (Roche) and Renilla luciferase (Promega) activities according to the manufacturers' instructions. The CAT activities were normalized to the corresponding Renilla luciferase activities to determine induction.

SARS-CoV does not induce beta interferon.

To investigate induction of IFN-β, we took advantage of a human 293 cell clone which can be productively infected with SARS-CoV (data not shown). Transactivation of the IFN-β promoter was first evaluated in a luciferase reporter assay. Cells were transfected with the reporter plasmid and were then infected either with SARS-CoV or the IFN-inducing control virus Bunyamwera delNSs (67). Mock-infected cells served as a negative control. Surprisingly, SARS-CoV was unable to activate the transfected IFN-β promoter, whereas the control virus was a strong inducer, as expected (Fig. 1a).

To investigate the inducibility of the endogenous IFN-β promoter, we analyzed infected cells for the presence of IFN-β mRNA with reverse transcription-PCR. In control-infected cells, a clear IFN-β mRNA-specific signal was detected (Fig. 1b, panel 1). In contrast, no signal was detectable in cells infected with SARS-CoV, although both the control virus and SARS-CoV replicated well, as assessed by reverse transcription-PCR analysis detecting viral nucleocapsid protein (N) mRNAs (Fig. 1b, panels 2 and 3). The reverse transcription-PCR signal specific for the cellular γ-actin mRNA indicated that all preparations contained similar amounts of RNA (Fig. 1b, panel 4). Taken together, our results suggest that SARS-CoV has an intrinsic ability to prevent induction of IFN in infected cells.

Transient activation of IRF-3 nuclear transport in SARS-CoV-infected cells.

The transcription factor IRF-3 normally resides in the cytoplasm. Immediately after virus infection, it becomes activated and moves to the nucleus to induce IFN-β gene transcription (35, 72). We investigated the subcellular localization of IRF-3 in cells that were infected with SARS-CoV with immunofluorescence analysis. At 8 h after infection, IRF-3 and viral antigens were detected by double labeling with specific antibodies. In uninfected cells, no specific nuclear accumulation of IRF-3 was observed, as expected (Fig. 2, panels 1 to 3). In cells infected with SARS-CoV, IRF-3 was found in the nucleus (Fig. 2, panels 4 to 6), exactly as in cells infected with the control virus (Fig. 2, panels 7 to 9). Obviously, infection with SARS-CoV induced the nuclear transport of IRF-3 but did not lead to IFN synthesis. We therefore investigated the fate of IRF-3 at a later time point of infection. Surprisingly, in cells infected for 16 h, IRF-3 was found in the cytoplasm again (Fig. 3, panels 4 to 6), similar to uninfected cells (Fig. 3, panels 1 to 3). In contrast, IRF-3 was still in the nucleus in cells infected with the control virus (Fig. 3, panels 7 to 9). Thus, SARS-CoV infection seems to result in an early but transient nuclear accumulation of IRF-3 which is reverted at later time points.

IRF-3 hyperphosphorylation, homodimerization, and recruitment of CBP.

Since SARS-CoV can block IFN-β transcription despite the observed early induction of IRF-3 nuclear translocation, we investigated whether other steps in IRF-3 activation were affected, such as hyperphosphorylation, homodimer formation, or recruitment of the coactivator protein CBP. IRF-3 is phosphorylated at several serine residues in response to virus infection (45, 56, 72). Since hyperphosphorylated IRF-3 migrates slower in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, it can be distinguished from the unphosphorylated form (40, 66, 72).

We examined the phosphorylation status of IRF-3 in response to SARS-CoV with gradient gel electrophoresis coupled to Western blot analysis. Cells were infected with either SARS-CoV or the control IFN-inducing virus or left uninfected and lysed at 8 h and 16 h postinfection. Surprisingly, no hyperphosphorylation was detectable early after infection (Fig. 4a, panel 1), although both viruses replicated to detectable levels (Fig. 4a, panels 2 and 3). At the late time point, no hyperphosphorylation of IRF-3 was detectable in SARS-CoV-infected cells, whereas the control virus induced IRF-3 hyperphosphorylation, as expected (Fig. 4b). Additional treatment with calf intestinal phosphatase (40) revealed that phosphorylation was indeed responsible for the slower-migrating form of IRF-3 (data not shown).

To examine IRF-3 homodimerization, we used nondenaturing polyacrylamide gel electrophoresis coupled to Western blot analysis (29). At the early time point of infection, neither the control virus nor SARS-CoV induced the appearance of a high-molecular-weight IRF-3 band which represents the homodimer (Fig. 5a). At the late time point, the control virus but not SARS-CoV induced IRF-3 homodimer formation (Fig. 5b).

Recruitment of the coactivator protein CBP is important for transcriptional activity of IRF-3. With a coimmunoprecipitation assay (40), we investigated the CBP-binding capacity of IRF-3 at early and late time points of infection. Figure 6a (panel 1) shows that, early in infection, neither the control virus nor SARS-CoV induced the binding of IRF-3 to CBP. At the late time point, the CBP-IRF-3 complex was formed in cells infected with the control virus, whereas with SARS-CoV no complex formation was detectable (Fig. 6b, panel 1). Nevertheless, comparable amounts of CBP were precipitated from the lysates (Fig. 6a and b, panels 2) and infection efficiencies were comparable, as demonstrated by Western blot analyses with N-specific antibodies (Fig. 6a and b, panels 3 and 4).

Taken together, our data indicate that SARS-CoV as well as an IFN-inducing control virus activate the nuclear transport of IRF-3 at early times postinfection, before detectable IRF-3 hyperphosphorylation, homodimerization, or recruitment of the coactivator CBP occurs. Subsequent to IRF-3 nuclear translocation, IRF-3 becomes fully activated and transcriptionally active in cells infected with the IFN-inducing virus but not in SARS-CoV-infected cells. Consequently, the IFN-β gene is not transcribed after SARS-CoV infection, and IRF-3, being a shuttling protein (35), localizes back to the cytoplasm later in infection.

SARS-CoV does not affect nuclear import or export.

We wondered whether SARS-CoV generally affects the subcellular localization of cellular proteins. To investigate the influence on nuclear import, we transfected cells with a GFP fusion construct expressing IRF-2, a nuclear homolog of IRF-3, and infected them with SARS-CoV. In Fig. 7a (left panel) it is shown that in uninfected cells, GFP-IRF-2 localized to the nucleus, as expected. Similarly, GFP-IRF-2 was still located exclusively in the nucleus when the transfected cells were superinfected with SARS-CoV (Fig. 7a, right panel). Similar results were obtained with IRF-1, another IRF family member which localizes to the nucleus, and NLS-MxA, a cytoplasmic protein artificially fused to a nuclear localization sequence (data not shown). This indicates that SARS-CoV does not manipulate the nuclear import of cellular proteins, even if they are related to IRF-3.

The nuclear localization of IRF-3 during the early phase of infection may be caused by a block in nuclear export, trapping the shuttling protein IRF-3 in the nucleus. Nuclear export of IRF-3 is mediated by the adaptor protein CRM1 (35). To test the influence of SARS-CoV on CRM1-dependent nuclear export, we used a reporter system which measures the capacity of the human immunodeficiency virus type 1 Rev protein to export CAT mRNAs out of the nucleus (19). A second reporter construct, generating an mRNA for Renilla luciferase, which is exported independently of Rev/CRM1, served as an internal control. Cells were transfected with the reporter plasmids and the Rev expression construct and infected with SARS-CoV, and reporter activities were measured at 16 h postinfection. The normalized data shown in Fig. 7b demonstrate that SARS-CoV has no specific influence on CRM1-dependent nuclear export, whereas the specific CRM1 inhibitor leptomycin B strongly downregulated reporter expression, as expected. Thus, taken together, our data suggest that the transient and incomplete activation of IRF-3 by SARS-CoV in not based on a general inhibition of nuclear transport events.