Severe Acute Respiratory Syndrome Coronavirus Protein 6 Accelerates Murine Hepatitis Virus Infections by More than One Mechanism^▿

Cells.

Human epithelial kidney cells (HEK293), baby hamster kidney cells (BHK), and BHK cells stably expressing T7 polymerase gene (BSR-T7) (5) were grown and maintained in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 10 mM Na-HEPES, 2 mM l-glutamine, and penicillin (100 U/ml)-streptomycin (100 μg/ml). Mouse fibroblast cells (17Cl-1) were grown and maintained in DMEM supplemented with 5% tryptose phosphate broth, 5% FBS, and penicillin-streptomycin.

Plasmid constructions.

Expression plasmids encoding full-length protein 6 [pCAGGS-6 and pBMN-6 (pBMN-6-IRES-eGFP)] and C-terminal deletion mutant [pBMN-6ΔC (pBMN-6ΔC-IRES-eGFP), lacking C-terminal residues 53 to 63] were constructed using standard PCR techniques. The retroviral vector pBMN-IRES-eGFP was kindly provided by Garry P. Nolan, Stanford University, Stanford, CA (30). All clones carry a hemagglutinin (HA) tag (sequence, YPYDVPDYA) at C termini for easy detection of expressed protein by Western blot and immunofluorescence assays. Renilla luciferase reporter plasmid (pRL-TK), which contains Renilla luciferase cDNA under the herpes simplex virus thymidine kinase (TK) promoter, was purchased from Promega (Madison, WI), and firefly luciferase reporter plasmid pECMVT7Luc, containing the firefly luciferase gene under control of a bacteriophage T7 promoter, has been described previously (3).

The plasmids encoding the enhanced green fluorescent protein (EGFP) trimer with or without the nuclear localization signal (NLS) were created using methods developed by Genove et al. with some modifications (18). Briefly, an EGFP cassette was PCR amplified using forward primer 5′-ATCTCGAGTGAGCAAGGGCGAGGAGC-3′ and reverse primer 5′-GAGAATTCAGCAAGGGCGAGGAGCTG-3′ and ligated into pEGFP-C1 (Clontech, Inc.) to generate a construct we designate p2xEGFP. A third EGFP cassette was PCR amplified using forward primer 5′-CTGAATTCTCCGGACTTGTACAGCAGG-3′ and reverse primer 5′-TTGTCGACGTCCGGACTTGTACAGCTCG-3′ and cloned into p2xEGFP to yield p3xEGFP. Subsequently, oligonucleotides encoding two tandem copies of the simian virus 40 (SV40) large T antigen NLS and human heterogeneous ribonucleoprotein (hnRNPA1) nuclear localization signal (M9 NLS) were subcloned at the 3′ end of the third GFP open reading frame to yield p3xEGFP-SV40 NLS and p3xEGFP-M9 NLS, respectively. The p3xEGFP-SV40 NLS encodes an ∼84-kDa fluorescent protein with a C-terminal-appended SV40 NLS (ELYKSGRRAQDPKKKRKVDPKKKRKV; the C-terminal GFP residues are shown in bold, linker residues are in italics, and the NLS is underlined). The p3xEGFP-M9 NLS encodes a fluorescent protein with C-terminal-appended M9 NLS (ELYKSGRRAQGNYNNQSSNFGPMKGGNFGGRSSGPYGGGGQYFAKPRNQGGY, with C-terminal GFP residues in bold, linker residues in italics, and the NLS underlined).

Transient transfection and infection.

293T cells were grown in 10-cm² dishes and transiently transfected with 2.0 μg pcDNA-mCEACAM1a (MHV receptor; J. Lacny and T. M. Gallagher, unpublished data) and 1 μg of p6-encoding plasmid (pCAGGS-6/pCAGGS-6-UT [untagged]/pBMN-6/pBMN-6ΔC) or vector control, using a calcium phosphate transfection method (47). After 24 h, transfection media were removed and cells were infected with recombinant MHV strain JHM (rJ.2.2) or rJ2.2 expressing SARS-CoV protein 6 (rJ2.2.6) at a multiplicity of infection of 0.2 PFU/cell in serum-free DMEM. After a 1-h adsorption period, the inocula were removed and infections were allowed to proceed overnight in DMEM-10% FBS.

Plaque assay and immunoblotting of virion proteins.

Culture media were collected at 16 h postinfection or at other times postinfection as indicated and clarified at 2,000 × g for 20 min. In plaque assays, aliquots of clarified media were serially diluted and applied to monolayers of 17Cl-1 cells. After 1 h of adsorption, inocula were removed and cells were washed with phosphate-buffered saline (PBS) and then overlaid with DMEM containing 1% FBS and 0.5% agar (Becton Dickinson, MD). The cells were fixed at 3 days postinfection and stained with 1% crystal violet, and plaques were enumerated. For isolation and detection of virus particles, clarified medium (0.7 ml) was overlaid on 0.6-ml cushions of 20% sucrose-25 mM Na-HEPES [pH 7.4]-100 mM NaCl-0.01% bovine serum albumin. Virions were pelleted by ultracentrifugation at 200,000 × g for 30 min. Virion-containing pellets were dissolved in sodium dodecyl sulfate (SDS) solubilizer (50 mM Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, 5% 2-mercaptoethanol, and 0.1% bromophenol blue), fractionated using SDS-polyacrylamide gel electrophoresis (PAGE), and immunoblotted with antibodies reacting with the HA epitope (Covance, Inc., Berkeley, CA), S proteins (monoclonal antibody [MAb] 10G; provided by Fumihiro Taguchi), M proteins (MAb J.1.3), and N proteins (MAb J.3.1; provided by John O. Fleming).

Transient transfection and luciferase assay.

To evaluate the influence of protein 6 on reporter gene expression, BHK, 293T, and BSR-T7 cells were cotransfected with 1 μg of reporter plasmid (pRL-TK or pECMVT7Luc or both) and 10-fold-increasing concentrations of pCAGGS-6. The total amount of DNA was kept the same for each transfection mixture by adding empty pCAGGS-MCS vector. In 293T cells, transfections were performed with calcium phosphate transfection reagents as described elsewhere (9, 47), whereas in BHK and BSR-T7 cells, transfections were performed with Lipofectamine 2000 as recommended by the manufacturer (Invitrogen Corporation). At 24 h posttransfection (hpt), cells were rinsed twice in PBS and lysed in passive reporter lysis buffer (Promega) after resuspension at 5 × 10⁶ cells per ml. Five-μl aliquots were assayed using luciferase assay reagent (Promega), and luminescence was recorded using a Veritas Microplate luminometer (Turner Biosystems, Sunnyvale, CA). All transfections were performed in triplicate, samples were measured in quadruplicate, and numerical uncertainties are shown by error bars.

Immunofluorescence and confocal microscopy.

To assess the nucleo-cytoplasmic distribution of NLS-containing GFP, 293T cells were grown on glass coverslips, placed into 10-cm² dishes, and transfected with plasmids encoding reporter genes (p3xGFP-M9 NLS, p3xGFP-SV40 NLS, or p3xGFP) along with p6 expression plasmids (pCAGGS-6, pBMN-6, and pBMN-6ΔC) or vector control plasmids. Cells were fixed at 24 hpt in 4% paraformaldehyde for 10 min and permeabilized in digitonin buffer (300 mM sucrose, 100 mM KCl, 2.5 mM MgCl₂, 1 mM EDTA, 10 mM HEPES [pH 6.9]) containing 5-μg/ml digitonin for 15 min at room temperature. The coverslips were incubated in blocking buffer (2% bovine serum albumin in PBS) for 15 min followed by an overnight incubation with primary antibody directed against HA tag at a dilution of 1:5,000. The coverslips were washed three times in ice-cold PBS at room temperature and incubated with fluorochrome-conjugated secondary antibodies (all from Molecular Probes, Inc., Eugene, OR) for 1 h at room temperature. The coverslips were washed three times in PBS and mounted with ProLong Gold antifade reagent (Invitrogen Corporation). Confocal images were captured using a Carl Zeiss model 510 laser-scanning confocal microscope. Digitized images were processed with Image J (National Institutes of Health).

siRNA transfection and MHV infections.

Predesigned small interfering RNAs (siRNAs; catalog number AM16704; siRNA ID 11218, 11125, and 145041) corresponding to exons 3 and 9 of human importin-β mRNA (accession number NM_002265), negative control siRNA (catalog number 4611), and scrambled control siRNA (GGCACAAUAUCAGCAGCGG; catalog number AM16104) were purchased from Ambion (Austin, TX). These siRNAs were transfected into 239T cells, in graded doses as indicated, along with pcDNA-mCEACAM1a, using Lipofectamine 2000 (Invitrogen). Typical experiments involved transfecting ∼2 × 10⁵ 293T cells with 100 nmol of siRNA and 1.5 μg of pcDNA-mCEACAM1a in 1 ml of OPTI-MEM (Life Technologies). Controls included mock transfection (transfection reagent lacking siRNA), negative control siRNA, and scrambled control siRNA. At 2 days posttransfection, cells were infected with rJ.2.2. After 16 h, supernatants were collected and infectivities were evaluated by plaque assay. Transfections were done in triplicate, and each experiment was repeated two to three times.

Plasmid DNAs encoding protein 6 prime cells for accelerated MHV infections.

In our previous studies, we used recombinant MHVs with SARS-CoV open reading frames integrated in the genome. Using these chimeric recombinant viruses, we showed that the SARS-CoV protein 6 was a virulence factor, increasing both MHV yields in tissue culture and neurovirulence in mice (41, 55). In the present study, we employed a related but distinct approach. Here, cultured cells were transiently transfected with expression plasmids encoding wild-type and mutant protein 6 and then subsequently infected with different MHV strains. This approach was simpler than our previous approach because it did not require development of recombinant viruses and was also able to address whether protein 6 might independently modify the cell in ways that influence the subsequent infection process. 293 cells were transfected with the desired plasmids and subsequently infected with recombinant MHV (rJ2.2) or recombinant MHV expressing SARS-CoV protein 6 (rJ2.2.6). Relative to vector-transfected cells, we observed robust viral egress from p6-expressing cells, as registered by abundant Western blot signals corresponding to virion proteins S, N, and M (Fig. 1A). These secreted particles were infectious, as demonstrated by plaque assay results, revealing ∼10-fold-more infectivity in medium from p6-expressing cells relative to vector controls (Fig. 1B). Comparative analysis of HA-tagged and untagged protein 6 revealed that the HA epitope had no influence on p6 activity (Fig. 1A and B). This augmenting effect of p6 was evident as early as 8 to 10 h postinfection (Fig. 1C and D).

In a parallel experiment, 293 cells were transfected with plasmids encoding the SARS accessory proteins 3a, 3b, and 7a (24, 26, 29, 50, 69, 71) and then infected with MHV. Immunoblot analyses confirmed that the plasmids were expressed at levels comparable to protein 6; however, none had any significant impact on MHV infection kinetics (Fig. 1A and B and data not shown). These data unequivocally indicated that protein 6 creates cellular environments that benefit MHV growth; however, the underlying mechanism responsible for this accelerated infection remained unclear.

Protein 6 reduces expression from plasmids requiring transcription in the nucleus.

Clues to the mechanism by which protein 6 increases MHV growth came from the unexpected observation that this protein suppressed the expression of cotransfecting plasmid-based genes. An example of this phenomenon is illustrated in Fig. 2A, in which 293 cells were cotransfected with plasmids encoding Renilla luciferase and p6. p6 levels correlated with significant (∼5- to 10-fold) reductions in luciferase activities when samples were evaluated at 16 to 20 h posttransfection. Similar suppressive effects of p6 were recorded in sequential transfections in which 293 cells were first transfected with p6 expression vectors and then with luciferase reporter plasmids 18 h later (Fig. 2B). However, the reverse transfection order did not result in any p6-mediated effects on luciferase accumulations (Fig. 2C), suggesting that while p6 could inhibit expression from other plasmid DNA, it could not do so if the other plasmid had already become established in the cell nucleus.

We considered that these p6-mediated reductions in luciferase accumulation might arise from generalized suppression of transcription or translation but found no evidence to support this suggestion. Approximately 50 to 70% of 293 cells were routinely positive for p6 accumulation after plasmid transfection, as measured by immunofluorescence with anti-HA epitope antibodies, yet these cells grew normally out to 3 days posttransfection, retaining viabilities similar to control transfectants as measured by trypan blue exclusion. Pulse-labeling with [³H]uridine, in either the presence or absence of actinomycin D and with [³⁵S]methionine, revealed that p6 effected no change in the incorporation kinetics for these radioactive RNA and protein precursors (data not shown). These findings led us to suggest that p6 specifically interfered with the cotransfected or subsequently transfected DNAs, perhaps by preventing their entrance into the cell nucleus.

Transfected plasmid DNAs are inefficiently transported from the cell cytoplasm to the nucleus (7, 57). Numerous reports have indicated that this transport can be accomplished by DNA-binding proteins that contain nuclear localization signals. These proteins attach to plasmid DNAs and escort them through nuclear pores, using the nuclear import machinery of the cell (11-13, 59, 65). To address the suggestion that protein 6 inhibits luciferase by blocking nuclear import of its plasmid DNAs, we asked whether luciferase expressed from a cytoplasmic bacteriophage T7 promoter might be immune to p6-mediated inhibition when transfected into cells containing cytoplasmic DNA-dependent T7 RNA polymerase (6, 66). To this end, BHK cells containing a cytoplasmic T7 RNA polymerase (BSR-T7) were cotransfected with vector control or p6 plasmids along with reporter plasmids encoding firefly luciferase under T7 promoter control (3). Under these conditions, p6 was indeed unable to inhibit luciferase accumulation (Fig. 3A). Parallel transfections into parental BHK cells lacking T7 RNA polymerase only generated a background luciferase activity. These data, normalized to cotransfecting nuclear-based Renilla luciferase expression, clearly revealed that the nuclear Renilla but not cytoplasmic firefly luciferase accumulation was inhibited in BSR-T7 cells by protein 6 (Fig. 3B). Together these data suggest that the inhibitory effect of protein 6 is at the level of plasmid DNA transport from the cytoplasm to the nucleus.

Protein 6 impedes the nuclear import of NLS-containing proteins.

Interference with nuclear-based plasmid DNA expression represents an unconventional approach to evaluate nuclear import events. A more common technique involves the monitoring of reporter proteins with or without nuclear localization signals. Active, import factor-dependent translocation of NLS-containing proteins through nuclear pores is required for cargo larger than ∼60 kDa (19, 20). Therefore, we constructed reporter plasmids encoding three tandem copies of green fluorescent protein, with or without C-terminal-appended nuclear localization signals (SV40 NLS and M9 NLS) (27). These constructs, designated 3xGFP, 3xGFP-SV40 NLS, and 3xGFP-M9 NLS, respectively, were proteins of 84 to 90 kDa and therefore exceeded the size limit for passive diffusion into the nucleus (4). 293 cells were cotransfected with these reporter plasmids, along with vector control or p6-expressing plasmids, and then monitored 1 day later by fluorescent microscopy. As expected, the 3xGFP protein was restricted to the cytoplasm, with p6 having the expected suppressive effect on its abundance but no effect at all on its cytoplasmic distribution (Fig. 4A). In striking contrast, the 3xGFP-SV40 NLS efficiently accumulated in the nuclei of vector control-transfected cells, with p6 exerting a pronounced blockade of this nuclear accumulation (Fig. 4A). Indeed, the distribution of 3xGFP-SV40 NLS in p6-expressing cells was essentially identical to that of 3xGFP lacking an NLS, suggesting that p6 interfered with the nuclear import of classical NLS-bearing proteins. Quantitative data revealed that relative to vector-transfected cells, p6 effectively blocked the nuclear accumulation of 3xGFP-NLS in more than 85% of the cells (Fig. 4B). Notably, this p6-mediated inhibition of 3xGFP-SV40 NLS was transient. At 2 days posttransfection, significant proportions of the 3xGFP-SV40 NLS accumulated in the nuclei of p6-expressing cells (data not shown), suggesting that p6 operated to delay but not entirely eradicate the import of NLS-bearing proteins. We also found that nucleo-cytoplasmic distribution of 3xGFP-M9 NLS was not affected by p6, with clear evidence of nuclear accumulation in the presence and absence of p6 (Fig. 4A). Thus, we conclude that p6 interfered with a subset of the known nuclear import pathways (42, 63), leaving the transportin-dependent nuclear import pathway intact and perhaps also leaving slower alternate routes available for 3xGFP-SV40 NLS cargo to enter into nuclei. These observations are consistent with the absence of overt toxicity in cells accumulating protein 6.

The C terminus of protein 6 is required to retain 3xGFP-SV40 NLS in the cytoplasm but is not essential to augment MHV infections.

The SARS protein 6 primed 293 cells for efficient MHV infections and also interfered with the nuclear import of a prototype NLS-bearing protein, 3xGFP-NLS. These observations prompted us to consider whether these two activities are causally related to each other. Previous findings, notably from Frieman et al. (15), might argue for such a causal connection, as their recent research has documented that STAT1 transcription factors are cytoplasmically retained by SARS protein 6. Given that STAT1 must be in nuclei to promote expression of antiviral effector proteins, the increased MHV replication might be related to a paucity of antiviral host proteins. Fortunately, these suggestions could be readily addressed because Frieman et al. (15) had also identified the hydrophilic C terminus of protein 6 as an essential portion interfering with transport of STAT1 into nuclei.

A series of protein 6 truncation mutants, generated by standard methods as described in Materials and Methods, were evaluated for their capacity to augment MHV infections and to prevent nuclear entry of 3xGFP-SV40 NLS. The most instructive p6 mutant was a C-terminal-truncated form lacking 11 amino acids, designated p6ΔC. In the MHV infection assays, we found that this mutant retained ∼80% of the complete p6 capacity to augment output of infectivity and virion particles, as measured by plaque assay and immunoblot analysis of secreted virion proteins (Fig. 5). In the transfected cells, the accumulations and subcellular localizations of these mutants were indistinguishable from the complete 6 proteins (Fig. 5 and data not shown), indicating that the incomplete augmenting activities of this mutant could not be explained by reduced abundance or position within the cells.

The C-terminal mutant was entirely incompetent in preventing the nuclear entry of 3xGFP-NLS (Fig. 6, lower panels). This finding confirms the data of Frieman et al. (15), and together these studies indicate that the protein 6 C terminus cytoplasmically retains both STAT1 as well as proteins with classical NLS motifs. More importantly, the collective findings suggest that a portion (∼20 to 30%) of protein 6-mediated MHV-augmenting activity is related to its ability to prevent NLS-containing proteins from entering the nucleus. The remaining portion (∼70 to 80%) of the augmenting activity presumably operates by a distinct mechanism.

siRNAs targeted to importin-β accelerate MHV infections.

It has been shown that p6 blocks nuclear import by sequestering import factors, especially importin-β, at the rough endoplasmic reticulum and Golgi membranes(15). To determine whether this mechanism at least partially contributes to the priming of cells for robust MHV growth, we attempted to reduce importin-β levels in cells without involving protein 6 at all. We reasoned that if nuclear import blockade were at least partially responsible for augmenting MHV infection, then siRNAs targeted to importin-β would exhibit MHV growth-promoting activities similar to protein 6. Thus, cells were first treated with siRNAs targeting importin-β mRNA and subsequently infected with MHV, in analogy with the experiments involving protein 6.

Three importin-β-specific and two different negative control siRNAs were evaluated in our experiments. Relative to the control siRNAs, 100 nM importin-β siRNAs 1, 2, and 3 reduced importin-β steady-state levels by 30 to 50% at 2 days posttransfection (Fig. 7A). These reductions did not have untoward toxic effects, as siRNA- and mock-treated cells were indistinguishable in integrity, adherence, and growth rate for up to 72 h posttransfection. Notably, these levels of importin-β reduction caused only a partial interference with the nuclear import of 3xGFP-SV40 NLS (data not shown), less stringently than that generated by protein 6. A combination of siRNAs 2 and 3, which target importin-β exons 3 and 9, was increasingly effective at preventing the nuclear localization of 3xGFP reporter protein (Fig. 7B).

MHV infections in these importin-β siRNA-treated cells did indeed exhibit high productivity (Fig. 7C). Interestingly, transfection with both of the control siRNAs also rendered cells more susceptible to MHV infections, but relative to these controls, the importin-β siRNAs improved MHV outputs in the range of four- to ninefold (Fig. 7C). Thus, our data revealed that reduced levels of importin-β correlated with improved MHV infections. These data suggest that protein 6 and importin-β siRNAs operate to foster MHV infections by a similar mechanism, namely, interference with protein import into the nucleus. However, protein 6 must have an additional operating mechanism to foster MHV growth, as the N-terminal p6 fragment (amino acids 1 to 52) augmented MHV growth without interfering with the NLS-GFP nuclear import.