The Membrane Protein of Severe Acute Respiratory Syndrome Coronavirus Functions as a Novel Cytosolic Pathogen-Associated Molecular Pattern To Promote Beta Interferon Induction via a Toll-Like-Receptor-Related TRAF3-Independent Mechanism

Wang, Yi; Liu, Li

Reports indicate that SARS-CoV M protein alone can form virus-like particles (VLPs) that can be secreted extracellularly. Therefore, the secreted M protein might be sensed by its own or other cell PRRs on the cell surface to activate IFN-I responses. To rule out this possibility, OxPAPC (a TLR2 and TLR4 inhibitor) was used to block the function of the accessory proteins CD14, LBP, and MD2 that are required for TLR2 and TLR4 signaling (32) . Figure 6C shows that addition of OxPAPC could effectively inhibit lipopolysaccharide (LPS)-mediated IFN-␤ production (right panel) but did not reverse the M-mediated IFN-␤ induction (left panel), indicating that the extracellular stimulation by M pro- SARS-CoV M protein promotes IFN-␤ induction independently of TRAF3. It has been well known that TRAF3 plays a critical role in TLR-mediated IFN-␤ induction, especially through TLR3 and TLR4 pathways (27, 29, 33) . Since M protein could activate the TLR pathway from inside the cells, it remained unclear whether or not this activation is in a TRAF3-dependent or -independent manner. To address this issue directly, a specific siRNA against TRAF3 was successfully constructed (Fig. 7A) . A dual-luciferase assay was conducted to assay the effect of siTRAF3 on M-mediated IFN-␤ induction in HEK293T cells. Figure 7B clearly shows that the increased delivery of siTRAF3 did not reverse the M-mediated IFN-␤ induction, indicating that TRAF3 might not be essential for this activation. To further confirm the above result, Western blot analysis was performed to detect IFN-␤ expression in responding to the increased delivery of siTRAF3 into HEK293T cells. The increased delivery of siTRAF3 into HEK293T Co-IP was conducted as shown in the left panel. About 10% input from each lysate preparation was subjected to Western blot analysis using anti-HA, anti-Flag, and anti-Myc antibodies as probes. The rest of the lysate was first immunoprecipitated with anti-Flag antibody conjugated with an affinity gel. Then, the reaction products were probed with anti-HA and anti-Myc antibodies. A reverse co-IP experiment was also conducted as shown in the right panel. The lysates were first immunoprecipitated with anti-Myc antibody. Then, the reaction products were individually probed with anti-Myc, anti-Flag, and anti-HA antibodies and subsequently subjected to Western blotting. The result is representative of at least 2 identical experiments. IB, immunoblotting; ns, nonspecific. (H) The M gene product does not suppress poly(I:C)-mediated IFN-␤ induction. About 2 g/ml of poly(I:C) was cotransfected with 2 g of either M or M(V68A) plasmid along with pGL3-IFN-␤luciferase reporter (100 ng) plus pRL-TK (10 ng) into HEK293T cells. After 48 h of transfection, dual-luciferase assays were performed to detect IFN-␤ expression (left panel). Each value represents the mean Ϯ standard deviation from three independent tests. The statistical difference was considered to be significant at a P value of Յ0.05. The IFN-␤ expression was also subjected to Western blotting (right panel). The relative band intensity was quantitated with the Image J program in comparison with the ␤-actin control. cells failed to inactivate the activation of IRF3 and NF-B p65 and did not affect the M-mediated IFN-␤ induction (Fig. 7C) , while a result was obtained consistent with the increased delivery of si-TRAF3 into J2-M cells (Fig. 7D) . In addition, a stable HEK293 cell line with TRAF3 knocked down by siTRAF3 was also establis