2.1 Vector construction

The vector with a U6 promoter (pBS/U6, a gift from Dr. Yang Shi at Harvard Medical School, Boson, MA, USA) was used to construct a 22 bp siRNAs. The hairpin cDNAs were generated through annealing of the complementary oligos synthesized, where ApaI and EcoRI sites were constructed. The hairpin cDNA insert was subcloned into the pBS/U6 vector through ApaI and EcoRI sites. The clones were verified by XhoI digestion. Finally the clones were sequence-confirmed.

The full length cDNA of the Spike protein was subcloned into a pCMV-Myc expression vector (Invitrogen) with BamHI and EcoRI through polymerase chain reaction (PCR) generating restrictive enzymes in the 5′ and 3′ ends of the insert. The subcloned vector, pCMV-Myc/HA-Spike, was sequence-confirmed for the correct frame in the expression vector. pEGFP-N1 was from Clontech Laboratories.

2.2 Cell culture and transfection

Human embryonic kidney cell line 293T and African green monkey kidney cell line Vero E6 were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin. 293T cells were plated onto 6 well plates at a density of 4×10⁵ cells per well and cultured at 37°C with 5% CO₂ overnight for transfection. A total amount of 5 μg DNA per well was used for transient transfection with CalPhos (Clontech) according to the manufacturer's protocol. Vero E6 cells were plated onto 12 well plates at a density of 2×10⁵ cells per well and cultured at 37°C with 5% CO₂ overnight. The transfection was performed at 90% cell confluence with a total amount of 4 μg DNA per well using Lipofectamine 2000 (Invitrogen) following the manufacturer's protocol.

2.3 SARS-CoV culture and infection

SARS-CoV (strain BJ01, GenBank accession number AY278488), provided by the Institute of Microbiology and Epidemiology, AMMS, China, was propagated on Vero E6 cells. The virus was released from the infected cells by three freeze–thaw cycles. The titration of the virus was determined by plaque assay on Vero E6 cell monolayer. For the test of inhibition of the virus by siRNA, 2×10⁵ transfected Vero E6 cells (with siRNA vectors) were infected with 5×10⁵ multiplicity of infection (MOI) of SARS-CoV (BJ01), in a final volume of 1 ml of DMEM with 2% FBS for 2 h at 37°C. Cells were then washed once with phosphate-buffered saline (PBS, pH 7.4), before adding 2 ml complete medium to allow growth for 48 h at 37°C with 5% CO₂. All those operations were performed in a bio-safety P3 level lab.

2.4 Western blot

Cells overexpressing hemagglutinin (HA)-tagged Spike protein were harvested 48 h post transfection and lysed with 0.3 ml of cold lysis buffer (20 mM Tris–HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM NaVO₃, 2.5 mM sodium pyrophosphate, 1 mM β-glycerol phosphate, 1 mM phenylmethylsulfonyl fluoride, 5 μg/ml aprotinin, and 5 μg/ml leupeptin, pH 7.5). 30 μg of total protein from cell lysate was separated on 6% sodium dodecyl sulfate–polyacrylamide gel electrophoresis, electro-transferred to nitrocellulose membrane (Hyclone ECI, Amersham Pharmacia Biotech). Antibodies directly to HA, enhanced green fluorescent protein (EGFP) and β-actin, obtained from Santa Cruz, were used to blot the membranes according to the conditions the manufacturer provided. The signals were detected by enhanced chemiluminescence (ECL, Amersham).

2.5 Immunostaining

Transfected cells were incubated for about 24 h. Cells were fixed in 4% paraformaldehyde for 20 min at room temperature, and permeabilized with PBS containing 0.2% Triton X-100 for 10 min. The cells were blocked with FDB (PBS with 1 mM CaCl₂, 1 mM MgCl₂, 5% normal goat serum, 10% fetal bovine serum and 2% bovine serum albumin) for 1 h, and were incubated with primary antibodies in FDB for 3 h. After washing with PBS containing 0.1% Triton X-100, the cells were incubated with suitable secondary antibodies conjugated with fluorescent isothiocyanate (TRITC) for 1 h. Cells were washed with PBS three times before observation. The results were analyzed using a Nikon TE300 fluorescence microscope. Pictures were taken with a CCD camera (SPOT, Diagnostic Instrument, USA) equipped with a computer-based image acquisition system.

2.6 RNA isolation and reverse transcription (RT) PCR

Total RNA from the transfected cells or infected cells was extracted using Trizol (Invitrogen). The PCR reaction was performed using the One Step RNA PCR kit (Takara, Japan) with 1 μg of total RNA for each sample. 1 μM of forward primer (5′-TATAGGATCCCACCATGTTTATTTTCTTATTATTTCTTAC-3′) and reverse primer (5′-TATAGAATTCTGTGTAATGTAATTTGACACC-3′) was used for the PCR reaction with 25 cycles of denaturation (94°C, 30 s), annealing (55°C, 1 min) and extension (68°C, 4 min). β-Actin (forward primer, 5′-CAAGAGATGGCCACGGCTGCT-3′; reverse primer, 5′-TCCTTCTGCATCCTGTCGGCA-3′) was used as an internal control and the reaction conditions were 25 cycles of denaturation (94°C, 30 s), annealing (55°C, 50 s) and extension (68°C, 1 min).

2.7 Northern blot

Total RNA (10 μg) from SARS-CoV-infected cells was fractionated on a 1.3% formaldehyde-agarose gel and transferred to a nitrocellulose membrane. Probe S-gene (1350–1950 bp from ATG from the CDS of the S gene) was labeled with ³²P using the Prime-a-Gene kit (Promega) according to the manufacturer's protocol. Hybridization was performed at 68°C for 2 h in hybridization buffer (Promega). Membranes were washed three times in 2×SSC with 0.1% sodium dodecyl sulfate at 50°C and autoradiographed by PhosphorImager (DNA Storm, Amersham Pharmacia Biotech).

2.8 Virus titration

Vero E6 cells were infected with 4×10³ MOI SARS-CoV (BJ01) per well 48 h post transfection with plasmids including pBS/U6, pBS/U6/GFPi and pBS/U6/S-RNAi1/2. The virus was allowed to infect the cells for 2 h and then was completely withdrawn by washing with PBS. After incubation for 72 h, the cells were lysed and assayed for the level of virus (MOI) in 96 well plates.

3.1 Selection of siRNA target sequences in mRNA of Spike protein

The coronavirus Spike protein plays a critical role in virus infection. Therefore, we chose the S gene as a target for inhibition of SARS-CoV infection. To locate the target site of siRNA in the SARS-CoV genome, we selected those sequences of 22 bp with the structure of AGN₁₈TT. As an empirical design, we preferably chose the sites with GC/AT ratios between 30% and 60%. Also we used the selected sequences to search GenBank to confirm their specificity. In matching for those criteria, two specific target sequences were determined and noted, S-RNAi1 and S-RNAi2. A schematic presentation of the sequences is shown in Fig. 1A . Those two targeted sequences in the Spike protein were at 1358–1376 and 3081–3099 bp from the first ATG of the cDNA of the gene, which were located in the S1 and S2 region of the protein, respectively.

3.2 Construction of the vectors generating siRNA

Although siRNA could be chemically synthesized directly, the cDNA vector driven by a promoter is used in most cases [14] for convenience and economy reasons. In this study, we adopted the U6 promoter vector [29] and constructed a 22 bp siRNA through the hairpin, which could be produced by the DNA vector. The hairpin cDNAs were generated through annealing of the complementary oligos synthesized, where ApaI and EcoRI sites were constructed. The hairpin cDNA as an insert was subcloned into pBS/U6 vector through ApaI and EcoRI sites. The correct clones were verified by XhoI digestion, which was replaced by the insert. Finally, the clones were sequence-confirmed. The single-strand hairpin RNAs generated from the vectors were predicted as in Fig. 1C.

3.3 Expression of Spike protein in cultured cells

To generate an in vitro expression system of the Spike protein from SARS-CoV, we adopted the mammalian cell culture method. The cDNA encoding the Spike protein from SARS-CoV was constructed into pCMV-Myc, a mammalian expression vector with a CMV promoter, which could drive the expression of the Spike protein in cultured cells. To detect the protein expression, the cDNA of Spike protein was tagged with HA at the C-terminus (Fig. 2A , top panel). This construct was transfected into 293T cells and the expression of Spike protein was measured by Western blot. The data showed that the vector could drive the expression of Spike protein of SARS-CoV in the cells at a high level (Fig. 2, top panel, lane 1) although the molecular weight of the expressed protein (>90 kDa) was smaller than the predicted one (105 kDa). To avoid different transfection efficiencies, we balanced the total amount of DNA with an EGFP expression vector. The data showed that the EGFP expressions were very even (Fig. 2, bottom panel, lanes 1, 3, and 4) and β-actin level showed equal loading of samples (Fig. 2, middle panel), suggesting that the transfection experiments were controlled under identical conditions. Taken together, we overexpressed the Spike protein from SARS-CoV in 293T cells, which could be used for the study of the gene silencing in the cultured cells.

3.4 Silencing of S gene expression in cells by siRNA

To determine whether the siRNA we generated could effectively reduce the expression of Spike protein in cultured cells, we first transfected the siRNA vectors pBS/U6/S-RNAi1 and pBS/U6/S-RNAi2 into 293T cells where Spike protein was overexpressed. To show the specificity of the siRNA targeting, we used EGFP as a control. From the data, it was clearly demonstrated that pBS/U6/EGFP-RNAi could specifically silence the expression of EGFP protein in cells (Fig. 2, bottom panel, lane 2). However, pBS/U6/S-RNAi1, pBS/U6/S-RNAi2 or the vector alone (pBS/U6) had no effect on EGFP expression (Fig. 2, bottom panel, lanes 1, 3 and 4). Interestingly, both pBS/U6/S-RNAi1 and pBS/U6/S-RNAi2 could specifically impair the expression of Spike protein in cultured cells (Fig. 2, top panel, lanes 3 and 4), while pBS/U6/EGFP-RNAi or the vector did not (Fig. 2, top panel, lanes 1 and 2). Meanwhile, all of the constructs had no effect on β-actin expression (Fig. 2, middle panel). All these data suggest that siRNA targeting of the Spike protein in SARS-CoV could block or silence the expression of Spike protein in mammalian cells.

To show the silencing of the Spike protein in transfected cells, we also performed an immunostaining experiment using anti-HA antibody. Similarly, we assessed EGFP expression by using fluorescence microscopy. The results showed that EGFP proteins were highly expressed in cells transfected with the expression vector in the presence of the co-transfected pBS/U6 vector (Fig. 3a ), pBS/U6/S-RNAi1 (Fig. 3e) and pBS/U6/S-RNAi2 (Fig. 3g), but rarely expressed in cells in the presence of the co-transfected pBS/U6/EGFP-RNAi (Fig. 3c). In contrast, Spike protein could be detected at levels similar to the EGFP protein in the control panel (Fig. 3a,b). However, when pBS/U6/S-RNAi1 (Fig. 3f) and pBS/U6/S-RNAi2 (Fig. 3h) respectively were transfected into cells, Spike proteins were shown to be gradually reduced. These data suggest that pBS/U6/S-RNAi1 and pBS/U6/S-RNAi2 significantly blocked the expression of Spike protein in the cells, indicating that the siRNAs we designed functioned in our mammalian system.

To verify the inhibition of Spike protein expression by siRNAs through the inhibition of the S gene, semi-quantitative RT-PCR was performed. The data show that pBS/U6/S-RNAi1 and pBS/U6/S-RNAi2 reduced the accumulation of mRNA S gene (Fig. 4 , lanes 3 and 4 compared to lane 2) while it had no effect on β-actin mRNA. These data indicate that siRNAs inhibited Spike protein expression through blocking mRNA accumulation.

3.5 Inhibition of SARS-CoV infection by siRNA in cultured cells

To determine whether the vector-based siRNA could efficiently prevent SARS-CoV infection of cells, we transfected 2, 3 and 4 μg of pBS/U6/S-RNAi1 (Fig. 5 , lanes 3–5), and pBS/U6/S-RNAi2 (Fig. 5, lanes 6–8) into Vero E6 cells in a 12 well plate. As controls, the pBS/U6 vector alone (Fig. 5, lane 1) and pBS/U6/EGFP-RNAi (as a non-related gene RNAi, Fig. 5, lane 2) were also transfected into parallel cells. After incubation for 1 day, the transfected cells were infected with SARS-CoV (BJ01) at 5×10⁵ MOI per well with 2×10⁵ cells. The cells were harvested and the total RNA was isolated for RT-PCR to measure S gene expression. The data indicate that both pBS/U6/S-RNAi1 and pBS/U6/S-RNAi2 could effectively inhibit gene expression of the Spike protein in the SARS-CoV-infected cells (compare lanes 3 and 6 to lanes 1 and 2 in Fig. 5). As expected, with the increasing amounts of siRNA transfected, there was increasing inhibition of S gene expression (compare lanes 3–5 and lanes 6–8 in Fig. 5). Those data were reconfirmed by Northern blots (Fig. 6 , same design as Fig. 5), suggesting that the siRNA we generated could block S gene expression of SARS-CoV in the infected cells.

3.6 siRNA inhibited the replication of SARS-CoV from the infected cells

Since SARS-CoV can replicate rapidly in cells we tested the ability of siRNA to inhibit virus production. After 72 h of culture, the virus in the medium of the cells should represent the newly replicated virus in the cells. To validate the siRNA effects on replication of SARS-CoV, we performed virus titration assays by assaying for extracellular virus in infected cells 72 h post infection. The data showed that the control plate had high titers of virus (Fig. 7 , lane 1), suggesting the virus replicated in the cells during the 72 h culture period. In contrast, the virus titer was dramatically decreased in the plate where the cells were transfected with siRNA (Fig. 7, lanes 2 and 3), demonstrating that the siRNA inhibited virus replication. Both pBS/U6 and pBS/U6/EGFP-RNAi treatments had similar levels of virus to that of the control, suggesting the RNAi effect was specific.