Inhibition of Beta interferon induction by severe acute respiratory syndrome coronavirus suggests a two-step model for activation of interferon regulatory factor 3

Abstract

Severe acute respiratory syndrome (SARS) is caused by a novel coronavirus termed SARS-CoV. We and others have previously shown that the replication of SARS-CoV can be suppressed by exogenously added interferon (IFN), a cytokine which is normally synthesized by cells as a reaction to virus infection. Here, we demonstrate that SARS-CoV escapes IFN-mediated growth inhibition by preventing the induction of IFN-beta. In SARS-CoV-infected cells, no endogenous IFN-beta transcripts and no IFN-beta promoter activity were detected. Nevertheless, the transcription factor interferon regulatory factor 3 (IRF-3), which is essential for IFN-beta promoter activity, was transported from the cytoplasm to the nucleus early after infection with SARS-CoV. However, at a later time point in infection, IRF-3 was again localized in the cytoplasm. By contrast, IRF-3 remained in the nucleus of cells infected with the IFN-inducing control virus Bunyamwera delNSs. Other signs of IRF-3 activation such as hyperphosphorylation, homodimer formation, and recruitment of the coactivator CREB-binding protein (CBP) were found late after infection with the control virus but not with SARS-CoV. Our data suggest that nuclear transport of IRF-3 is an immediate-early reaction to virus infection and may precede its hyperphosphorylation, homodimer formation, and binding to CBP. In order to escape activation of the IFN system, SARS-CoV appears to block a step after the early nuclear transport of IRF-3.

FIG. 1.

Absence of IFN-β induction in SARS-CoV-infected cells. (a) IFN-β promoter activation. Human 293 cells were transfected with plasmid p-125Luc containing the firefly luciferase gene under control of the IFN-β promoter, along with a control plasmid encoding Renilla luciferase under control of the constitutively active simian virus 40 promoter. At 6 h posttransfection, cells were infected with SARS-CoV or the IFN-inducing control virus Bunyamwera delNSs (CTRL) or left uninfected (mock). At 18 h postinfection, luciferase activities were measured. Firefly luciferase values (reflecting IFN-β promoter activation) were normalized to the Renilla luciferase activities. (b) Reverse transcription-PCR analysis. Total RNA from 293 cells which were infected for 18 h was assayed by reverse transcription-PCR for the presence of IFN-β mRNA (panel 1), mRNA of the N gene of SARS-CoV (panel 2) or the control virus (panel 3), or cellular γ-actin mRNA (panel 4).

FIG. 2.

Subcellular localization of IRF-3 early in infection. Confocal double immunofluorescence pictures of cells which were either mock infected (panels 1 to 3) or infected for 8 h with SARS-CoV (panels 4 to 6) or the control virus (panels 7 to 9). Viral N proteins (panels 4 and 7) and IRF-3 (panels 2, 5, and 8) were detected with specific rabbit and mouse antisera, respectively.

FIG. 3.

Subcellular localization of IRF-3 late in infection. Cells were infected for 16 h and immunostained as indicated for Fig. 2.

FIG. 4.

Hyperphosphorylation of IRF-3. Extracts from cells infected for 8 h (a) or 16 h (b) were analyzed by sodium dodecyl sulfate gel electrophoresis, followed by immunoblot analyses to detect IRF-3 (panels 1), the N protein of SARS-CoV (panels 2), or the N protein of the control virus (panels 3). The two forms of unphosphorylated IRF-3 are indicated by arrowheads, and the hyperphosphorylated form of IRF-3 is indicated by an asterisk (*).

FIG. 5.

Homodimerization of IRF-3. Extracts from cells infected with SARS-CoV or the control virus for 8 h (a) or 16 h (b) were analyzed by nondenaturing gel electrophoresis followed by an immunoblot to detect IRF-3. The positions of IRF-3 monomers and dimers are indicated by arrowheads.

FIG. 6.

CBP recruitment by IRF-3. Cells were infected for 8 h (a) or 16 h (b), and anti-CBP immunocomplexes (IP) were analyzed for the presence of IRF-3 (panels 1) and CBP (panels 2). Infections were monitored by probing the postprecipitation supernatants (SN) for the viral N proteins (panels 3 and 4).

FIG. 7.

Nuclear import and export. (a) Nuclear localization of IRF-2. Cells were transfected with plasmid pEGFP-C1-hIRF-2 for 24 h and then superinfected with SARS-CoV for 16 h (right panel) or left uninfected (left panel). Viral N protein was detected with a specific rabbit antiserum and a Cy3-conjugated secondary antibody. (b) Nuclear export assays. Cells were transfected with the Rev expression construct pcRev, the Rev export-dependent CAT reporter plasmid pDM128 (19), and the Renilla luciferase control plasmid pRL-SV40. After 4 h at 37°C, cells were either infected with SARS-CoV or left uninfected (black columns). As a control, cells were treated in parallel with the CRM-1 inhibitor leptomycin B (LMB) at a concentration of 2 ng/ml (grey columns). At 16 h postinfection, reporter activities were measured. CAT activities were normalized to the corresponding Renilla luciferase activities to determine induction. Data from a representative experiment are shown.