Small Interfering RNA Effectively Inhibits the Expression of SARS Coronavirus Membrane Gene at Two Novel Targeting Sites

Small interfering RNA (siRNA) is a class of duplex RNA molecules of 21-25 nt nucleotides in length functioning post-transcriptionally to downregulate targeted gene expression. The membrane (M) protein of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) is highly abundant during viral infections and is a critical element for viral assembly. Nucleotide substitution in the viral genome occurs frequently during SARS-CoV infection. In the current study, we analyzed the M gene sequences derived from 15 SARS-CoV isolates and uncovered six nucleotide substitutions among these isolates. Interestingly, these nucleotide substitutions are all located at the 5’ half of the M gene. Based on this information and previous reports, we created two novel siRNAs targeting two unexplored and well conserved regions in the M gene. The effects of these two siRNAs were tested by semi-quantitative RT-PCR and EGFP-M fusion gene expression. The results demonstrated that both siRNAs effectively and specifically blocked the targeted gene expression. Real time quantitative RT-PCR (qRT-PCR) revealed that siRNA targeting the 3’ half of the M gene (si-M2) induced more potent inhibition than that targeting the 5’ half (si-M1). Both si-M1 and si-M2 significantly downregulated M gene mediated upregulation of interferon β expression. Thus, our results indicate that SARS-CoV M gene specific siRNA might function in a sequence-dependent manner.

of current study may provide valuable information for the design of more effective siRNA against the M gene of SARS-CoV.

Cloning and expression analysis of M gene
The full length of SARS-CoV M gene was amplified from Vero E6 cells infected with the viral strain HKU-39849 ( Figure 2a). The complete coding sequence of M gene was then subcloned into eukaryotic expression vector pCMV-Myc to generate plasmid pCMV-Myc-M. Western blot analysis indicated that SARS-CoV M protein was insoluble. The result showed that no significant amount of M proteins was presented in the supernatant fraction, while a large amount of M protein could be readily detected in the cell pellets ( Figure 2b). The result is in agreement with a previous report that SARS-CoV M proteins can be thermally aggregated upon heat treatment [19].

Two novel siRNAs effectively inhibits M gene expression
We designed two siRNAs specifically targeting these two regions as shown in Figure 1, and named them si-M1 and si-M2, respectively. M gene was also fused with the EGFP gene to make the pEGFP-M construct. The effect of siRNA on M gene expression was monitored by both RT-PCR and EGFP expression. Figure 3a demonstrated a significant reduction in M gene expression as the ratio of M to si-M1 reaching 1:8. Quantitation of the band intensity revealed about a 23 fold reduction in M mRNAs as co-transfecting with higher doses of si-M1 ( Figure 3b). Moreover, a marked inhibition on EGFP-M, but not EGFP expression was observed when administrated with higher doses of si-M1 ( Figure 3c). The specificity of si-M1 on M gene expression was further confirmed by using a non-specific siRNA, si-IL17RE, as a negative control. Higher doses of si-M1 but not si-IL-17RE effectively inhibited EGFP-M gene expression (Figure 3d). Flow cytometric analysis indicates that the intensity of EGFP-M gene expression was significantly inhibited by si-M1 but not si-IL17RE as the delivered dose increased ( Figure 3e). Similarly, si-M2 which targeted at the 3' half of M gene also dramatically inhibited M mRNAs by about 78 fold when the molar ratio of si-M2:M was increased to 8:1 (Figures 4a and 4b). si-M2 mediated EGFP gene inhibition was also markedly induced when a higher dose of si-M2 was co-transfected with EGFP-M fusion gene but not with EGFP gene (Figure 4c).
Higher doses of si-M2 but not si-IL-17RE effectively inhibited EGFP-M gene expression ( Figure  4d). Flow cytometric analysis indicates that the intensity of EGFP-M gene expression was significantly inhibited by si-M2 but not si-IL17RE as the delivered dose increased (Figure 4e). Overall, the above results clearly demonstrated that both si-M1 and si-M2 are effective inhibitors to block SARS-CoV M gene expression.

siRNA targeting the 3 portion of M gene is a more potent inhibitor
Previous studies showed that specific siRNAs targeting at the terminal sequences of M coding region are more potent gene silencers [15,16]. To assess the strength of gene inhibition induced by si-M1 and si-M2, as well as those known siRNAs, real time qRT-PCR was employed to directly measure siRNA-mediated M gene repression that included the siRNA targeting the 3' terminus (637-657nt) of M gene (named as si-M3 in this study) described by Qin et al. [15] as a positive control.
Both si-M1 and si-M2 were effective to down-regulate M gene expression but with a less effect than that of si-M3 ( Figure 5). Targeting at the 3' portion of M gene by si-M2 and si-M3 generated more potent inhibition on M mRNA expression than that of si-M1 ( Figure 5). Our study indicates that siRNA duplex mediated gene silencing in SARS M gene expression might work in a sequence-dependent manner.

M gene specific siRNAs effectively inhibited M gene mediated upregulation of interferon  gene expression
The interferon  production is usually induced by dsRNA [20]. However, we found that overexpression of M gene alone in HEK293 cells was able to induce IFN production ( Figure 6, lane 1 and 2). The IFN induction by M gene could be reversed by M gene specific siRNAs. Either si-M1 or si-M2 could effectively downregulate M gene mediated IFN production ( Figure 6), indicating that both si-M1 and si-M2 were functionally effective to repress SARS-CoV M gene expression. . About 1 g of total RNA isolated from each reaction was subjected to RT-PCR analysis using either IFN or -actin primers.

Cell culture and antibodies
Human embryonic kidney cell line 293 (HEK293) and SV40 T antigen transformed HEK293 (HEK293T) were derived from the Cell Culture Center of Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences. Cells were cultured in Dulbecco's Modified Eagle Medium (HyClone, South Logan, UT) supplemented with 10% fetal calf serum and incubated in a 37 ºC incubator containing 5% CO 2 . Anti-Myc antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

Reverse transcription-polymerase chain reaction (RT-PCR) and real time quantitative RT-PCR (qRT-PCR)
Total RNAs were extracted from the cultured cells with TRIzol (Invitrogen, Carlsbad, CA, USA). All primers used in the RT-PCR reactions were listed in Table 1. One g of total RNAs was first reverse transcribed using AMV reverse transcriptase (Promega, Madison, WI, USA). About 2 L of the transcribed cDNAs was subjected to standard PCR reaction using M gene specific primers. One-step real-time quantitative RT-PCR (qRT-PCR) (Takara Biotechnology, Dalian, China) was also performed to monitor the targeted gene expression. Real time qRT-PCR was carried out with iQ5 real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA, USA ) at the following conditions: 42 ºC for 5 min and 95 ºC for 10 sec; 95 ºC for 5 seconds and 60 ºC for 10 seconds and repeated for 40 cycles. The dissociation of the reaction products was conducted from 55 ºC to 95 ºC as the temperature rose at 0.2 ºC per ten seconds.

Transient transfection
Cell cultured in 35-mm dishes were transiently transfected with the indicated plasmid DNAs using ProFection ® Mammalian Transfection Systems (Promega) according to the supplier's instructions. Briefly, transfected DNAs were first mixed with 2M CaCl 2 (37 L) and brought to a total volume of 300 L with sterile and deionized water. Then the DNA-CaCl 2 mixture was added into equal volume of 2×HBS drop by drop accompanying with gentle vortexing. After 15 minutes incubation, the reaction mixture was evenly distributed into the cell culture medium and incubated for 48 hours before harvesting.

Western blot analysis
The transfected cells were lysed with a lysis buffer containing 1% NP-40, 50 mM Tris-HCl (pH 7.5), 120 mM NaCl, 200 μM NaVO 4 , 1 g/mL leupeptin, 1 g/mL aprotinin, and 1 M PMSF. About 15 g of cell lysate for each sample was resolved onto 12% SDS-PAGE. After separation, the separated proteins were transferred onto Hybond nitrocellular membrane (Pharmacia, St. Louis, MO, USA). The transferred membrane was first probed with a primary antibody. Then, a secondary antibody labeled with horseradish peroxidase was added to the reaction and finally visualized with an ECL kit (Santa Cruz Biotechnology, Santa Cruz, CA, USA).

Flow cytometric analysis
HEK293 cells were transiently co-transfected with EGFP-M plus either si-M1 or si-M2, while the contransfection of EGFP-M plus si-IL-17RE#1 [22] was served as a negative control. After 48-h incubation, the transfected cells were released from the culture plates and resuspended into 1×PBS. The intensity of EGFP gene expression was measured using the FACSARIA flow cytometer (Becton Dickenson, San Jose, CA, USA).

Conclusions
The M protein of SARS-CoV is a key element required for viral assembly. Our data shows that the M protein was not present in the soluble fraction, but can be readily detected in cell pellets, indicating that M protein has the propensity to self-aggregate. SARS-CoV M genes were derived from 15 different viral isolates and it was found that nucleotide substitutions were exclusively located at the 5' half of the M gene (the first 360 nt). Two novel and conserved siRNA target sites were identified and tested. Both siRNAs (si-M1 and si-M2) could effectively inhibit M gene expression post-transcriptionally and functioned in a dose-dependent manner. Moreover, si-M2 targeting the 3' half of the M gene produced more potent inhibition than that of si-M1 that recognized a conserved sequence at the 5' half of the M gene. This observation is consistent with a previous report described by Qin et al. [15]. The less inhibitory effect induced by si-M1 was correlated with a higher chance of mutation in the targeted region nearby (such as nt203 and nt256). Moreover, our study indicates that both si-M1 and si-M2 were able to functionally counteract the M gene-mediated INF production. This study indicates that siRNA mediated M gene inhibition might be determined by the sequence content presented at or around the targeted site.