(A–C) HeLa cells were mock-treated or infected with IAV (A) or IAVÄNS1 (B) for the indicated period, fixed, stained, and analyzed with confocal microscopy. The cells were stained with anti-RIG-I (RIG-I), anti-IAV nucleocapsid protein (NP), and anti-TIAR (TIAR) antibodies. Nuclei were stained with DAPI. At 9 h and 12 h after infection, the percentage of speckle-like distribution of RIG-I was 0.5% and 0.0% in IAV-infected cells, and 62.4% and 83.6% in IAVÄNS1-infected cells, respectively. The zoomed images correspond to the boxed region in each panel. The cells at 9 h post infection were stained with anti-RIG-I and anti-IRF-3 antibodies (C). (D and E) HeLa cells were infected with IAVÄNS1 for 9 h, and stained with anti-G3BP (G3BP), together with anti-RIG-I (RIG-I) (D) or anti-eIF3 (eIF3) (E). The zoomed images correspond to the boxed regions. (F) HeLa cells were mock-treated or infected with IAV or IAVÄNS1 for 12 h. Cell extracts were prepared and immunoprecipitated with anti-RIG-I antibody. The precipitates were analyzed by immunoblotting (IP:RIG-I) using antibody against G3BP, HuR, TIAR and RIG-I. Input: 1/50 of the extracts used for immunoprecipitation were analyzed similarly by immunoblotting.

(A–D) HeLa cells were mock-treated (mock), infected with IAVÄNS1 for 9 h, or treated with NaAsO₂ for 1 h. Cells were fixed and stained for G3BP and MDA5 (94.2% colocalization) (A), G3BP and LGP2 (97.6% colocalization) (B), RIG-I and RNase L (84.5% colocalization) (C), RIG-I and OAS (87.4% colocalization) (D) in IAVÄNS1-infected cells. The zoomed images correspond to the boxed regions.

(A) 293T cells were transfected with empty vector (empty) or the HA-PKR expression vector (HA-PKR) for 24 h, or treated with NaAsO₂ for 1 h and stained with anti-RIG-I, anti-HA (PKR) and anti-TIAR antibodies and DAPI. The zoomed images correspond to the boxed regions. (B) 293T cells were transfected with empty vector (empty), or the HA-PKR expression vector for 48 h, or treated with NaAsO₂ for 1 h, or infected with IAVÄNS1 for 12 h. Relative mRNA levels of endogenous IFN-â gene were determined by quantitative PCR (qPCR). Data are represented as the mean standard ± error of the mean (SEM). (C and D) HeLa cells were mock-treated or infected with IAVÄNS1 for 12 h. Viral RNA (vRNA) was detected by the FISH method using an RNA probe complementary to the segment 1 of the IAV, and NP (C) and RIG-I (D) were detected using anti-NP and anti-RIG-I antibodies (97.1%, and 98.2% colocalization of vRNA with NP and RIG-I, respectively). TO-PRO-3 was used for staining of nuclear DNA (DNA). (E) HeLa cell lines stably expressing FLAG-tagged IPS-1 were mock-treated or infected with IAV or IAVÄNS1 for 10 h. The cells were stained with anti-PMP70, anti-FLAG, and anti-TIAR antibodies. The white arrowheads indicated the contacts between FLAG-IPS-1 and TIAR. 67.8% of IAVÄNS1 infected cells exhibited contacts, whereas IAV infected cells hardly exhibited the contact (2.7%). The zoomed images of PMP70 (Green) and FLAG (Red), PMP70 (Green) and TIAR (Red), and FLAG (Green) and TIAR (Red) in IAVÄNS1-infected cells were shown in the bottom panel.

(A–E) HeLa cells were transfected with control siRNA (N.C) or siRNA targeting human G3BP (G3BPi). At 48 h after transfection, cells were harvested and G3BP and actin were detected by immunoblotting (A). Cells were mock-treated (mock) or infected with IAVÄNS1 for 12 h and fixed and stained with anti-RIG-I, anti-NP and anti-TIAR antibodies and DAPI (B). The percentage of cells containing foci of RIG-I (C) or TIAR (D) was determined. Relative mRNA level of IFN-â was determined by qPCR (E). Data are represented as the mean standard ± error of the mean (SEM).

(A) HeLa cells were mock-treated or infected with IAVÄNS1 for 9 h or treated with NaAsO₂ for 1 h. Cells were fixed and stained with anti-RIG-I and anti-PKR antibodies (% of colocalization: 95.1% and 97.0% in IAVÄNS1-infected and NaAsO₂-treated cells, respectively). The zoomed images correspond to the boxed regions. (B) HeLa cells were mock-treated or infected with IAV or IAVÄNS1 for 9 h and stained with anti-phospho-eIF2á (Ser 51) (p-eIF2á) and G3BP (% colocalization: 0.0% and 46.5% in IAV and IAVÄNS1-infected cells, respectively). The zoomed images correspond to the boxed regions. (C) HeLa cells were infected with IAV or IAVÄNS1 for 12 h or treated with NaAsO₂ for 1 h. Cell extracts were prepared and subjected to SDS-PAGE, and immunoblotted using antibodies against RIG-I, PKR, phosphorylated PKR (Thr 446) (p-PKR), phosphorylated eIF2á (Ser 51) (p-eIF2á), IAV NP, and â-actin.

(A–C) MEFs derived from WT and PKR KO mice were mock-treated or infected with IAVÄNS1 for 12 h. The cells were stained with anti-RIG-I, anti-IAV NP and anti-TIAR antibodies and DAPI (A). The percentage of cells containing foci of RIG-I (B) or TIAR (C) was determined. (D–F) PKR WT and PKR KO MEFs were mock-treated or infected with IAVÄNS1. The IFN-â mRNA level at 9 h post-infection was determined by qPCR (D). The IFN-â protein levels in culture medium at 15 h post-infection were quantified by ELISA (E). Cell extracts were subjected to Native-PAGE and IRF-3 dimer was detected by immunoblotting using anti-IRF-3 antibody. IAV NP and actin were detected by SDS-PAGE followed by blotting using anti-NP and anti-actin antibodies (F). Data shown in B-E are represented as the mean standard ± error of the mean (SEM).

HeLa cells were transfected with control siRNA (N.C) or siRNA targeting human PKR mRNA or G3BP. At 48 h after transfection, cells were mock-treated or infected with IAVÄNS1 for 24 h. The expression levels of IAV RNA segment 3 and GAPDH mRNA were determined by RT-PCR (top). The IAV RNA expression patterns were quantified with LAS-1000 UV mini (Fujifilm, Japan) and normalized with GAPDH (bottom). Data are represented as the mean standard ± error of the mean (SEM).

(A–C) MEFs derived from WT and PKR KO mice were mock-treated or transfected with IAV genomic RNA, short poly I:C or long poly I:C for 9 h and stained with anti-RIG-I, anti-G3BP antibodies and DAPI (A). The percentage of cells containing foci of G3BP was shown in (B). Relative IFN-â mRNA levels were determined by qPCR (C). Data shown in B and C are represented as the mean standard ± error of the mean (SEM).

Viral infections generate RNA with non-self-signatures such as a 5′-tri-phosphate or double-stranded structure (a). In the case of IAVÄNS1, PKR is critical for the formation of avSGs (b). Wild-type IAV inhibits formation of avSGs by the actions of the NS1 protein (c). Some other viruses may activate different eIF2á kinases, such as GCN2, PERK, and HRI, to produce functional avSGs (d). avSGs are composed of SG markers and other RNA-binding proteins including RLRs, antiviral proteins (PKR, OAS, and RNase L) and viral RNA. Within avSGs, viral RNA could be sensed by RLRs to trigger antiviral signaling (e). Activated RLRs recruit mitochondrial IPS-1 via CARD-CARD interactions (f). IPS-1 serves as another platform for TRAFs and protein kinases, TBK-1, IKKi and IKK complex, to activate target genes (g). The antiviral proteins are activated by viral RNA to block viral RNA synthesis and translation (h). Moreover, OAS-RNase L system may produce dsRNA to amplify the RLR signaling (i).