The DHAV-1 H strain (GenBank: JQ301467.1) was provided by the Institute of Preventive Veterinary Medicine at Sichuan Agricultural University. The primary DEFs were described previously (Xie et al., 2019; Lai et al., 2020). Six-well cell culture plates were seeded 5 × 10⁶ cells per well and the cells were grown in minimum essential medium (MEM) containing 10% newborn calf serum (Gibco) and incubated at 37°C with 5% CO₂ in an incubator. Then, DEFs were infected with DHAV-1 for 2 h, and the unbound virus was removed by washing with phosphate-buffered saline (PBS) twice before the cells were overlaid with MEM containing 2% newborn calf serum. DEFs were transfected expression plasmids or poly(I:C) with TransIntro™ EL Transfection Reagent (TransGen Biotech). UV-DHAV-1 is obtained by irradiating DHAV-1 with UV light with a wavelength of 253.7 nm for 6 h.

Mouse anti-puromycin antibody was obtained from Millipore. Rabbit anti-phospho-eIF2α (S51) and mouse anti-Myc antibody were obtained from Cell Signaling Technology. Mouse anti-HA and mouse anti-Flag antibody were purchased from MBL. Mouse anti-GFP antibody was obtained from ABclonal. Rabbit anti-beta (β)-actin antibody was obtained from Proteintech. HRP-conjugated goat anti-mouse IgG and HRP-conjugated goat anti-rabbit IgG were purchased from Beyotime. Rabbit anti-VP3 antibody was prepared in our laboratory (Shen et al., 2016).

C16 (PKR Inhibitor) and GSK2606414 (PERK inhibitor) were purchased from APExBIO, and GCN2-IN-1 (GCN2 inhibitor) was purchased from MCE. These inhibitors were dissolved in dimethyl sulfoxide (Sigma) and configured to 10 mmol/L, which were diluted to working concentration with MEM when used. Sodium arsenite was purchased from Sigma. Poly(I:C) was purchased from Invivogen.

According to the manufacturer’s instructions, to construct plasmids expressing the viral protein, DHAV-1 RNA was isolated using RNAiso Plus Reagent (TaKaRa). According to the manufacturer’s instructions, genomic DNA was then removed, and reverse transcription was performed using a PrimeScriptTM RT Reagent Kit (TaKaRa). VP0, VP1, 2A, 2B, 2C, 3AB, and 3D sequences were amplified from DHAV-1 cDNA with PCR and primers (Table 1). VP0, VP1, 2A, 2B, 2C, and 3D were integrated into the pCAGGs vector with a one-step cloning kit (Vazyme). 3AB was cloned into the pCMV-Myc vector with the DNA Ligation Kit (TaKaRa). The eukaryotic expression vector pCAGGs was gifted by Shanghai Veterinary Research Institute, Chinese Academy of Agricultural Sciences. The eukaryotic expression vector pCMV-Myc was purchased from TaKaRa. pCAGGs-VP3-Flag and pEGFP-N1-3C were also stocks in our laboratory (Lai et al., 2019; Sun et al., 2019). Duck-derived G3BP1 gene was cloned into the pEGFP-C2 vector with the DNA Ligation Kit (TaKaRa). The pEGFP-C2 vector was purchased from YouBio.

The pGPU6/GFP/Neo-PERK#1/#2/#3 expression vectors and the pGPU6/GFP/Neo-GCN2#1/#2/#3 expression vectors were designed and constructed by GenePharma (Shanghai, China), and the sequences of shRNAs were as follows: PERK#1 5'-GCCAGTCATTAGAGGAAATTT-3', PERK#2 5'-GGCATGATAATGCAATTATTC-3', PERK#3 5'-GCAGGAAAGAGAACCTTAAAG-3', GCN2#1 5'-GCCTAAAGTTAGTGTTATAGC-3', GCN2#2 5'-GGACTATGATGAGTCAAATAT-3', GCN2#3 5'-GGTGCGAAATAAGCTTGATGG-3'. The DEFs were plated into six-well cell culture dishes and cultured overnight. The cells were then transfected with the recombinant plasmids using TransIntro™ EL Transfection Reagent (TransGen Biotech).

Total RNA was isolated using RNAiso Plus Reagent (TaKaRa) according to the manufacturer’s instructions. The number of viral copies in total RNA was measured using methods previously established in our laboratory (Hu et al., 2016). Three genes (PKE, PERK, and GCN2) and a housekeeping gene (β-actin) were analyzed by qPCR using primers designed with Primer Premier 5 (Table 1). The expression levels of four genes were determined by qPCR using a SYBR® Premix Ex Taq™ II (Tli RNaseH Plus) Kit (TaKaRa) and an Applied CFX96 Real-Time PCR Detection System (Bio-Rad). Amplification was performed in 10 μl reaction volumes containing 0.5 μl of each primer and 1 μl of cDNA. The following thermal cycling conditions were used: initial activation at 95°C for 30 s, 40 cycles of denaturation at 95°C for 5 s and annealing and extension at 56.9°C for 30 s, and a dissociation curve analysis step.

Refer to the previous method, DEFs were treated with 1 μg/ml of puromycin for 30 min to detect the protein translation efficiency (Wang et al., 2020). After DHAV-1 infection for 22 h, DEFs were treated with different concentrations of C16, GSK2606414, and GCN2-IN-1 for 2 h to screen the kinases that affect eIF2α phosphorylation. DEFs were treated with different concentrations of sodium arsenite for 30 min to explore the relationship between eIF2α phosphorylation and translation shutoff.

Cells were lysed in 200 μl cell lysis buffer (Beyotime) containing 1% PMSF. The cell lysate was centrifuged, and the supernatant was collected. Samples were fractionated by SDS-PAGE electrophoresis and then transferred to PVDF membrane, which was blocked with 5% non-fat dry milk at room temperature for 5–6 h. The membranes were incubated overnight at 4°C with primary antibodies diluted in blocking buffer. The membranes were washed three times with TBS-Tween and incubated for 1 h at 37°C with the respective secondary antibodies diluted in blocking buffer. The membranes were then washed three times with TBS-Tween, and bound proteins were detected using an ECL chromogenic kit (Beyotime).

The cells were rinsed (three times) with phosphate-buffered saline (PBS) and then fixed in 4% paraformaldehyde overnight. The cells were permeabilized (0.2% Triton X-100 for 25 min), incubated with blocking solution [5% bovine serum albumin (BSA) in PBS with Tween 20 (PBST) for 60 min at 37°C], and then rinsed with PBST. The cells were then treated with DAPI (Roche). Images were captured using an 80i upright microscope (Nikon) and a SPOT Flex camera.

The cell viability was measured using the Cell Counting Kit-8 (Biosharp) according to the manufacturer’s instructions.

After many viruses infect host cells, they hijack the cell’s translation-related factors and reduce the overall cellular translation efficiency. Puromycin is a protein synthesis inhibitor. It has a structure similar to the end of the tRNA molecule. It can replace the aminoacylated tRNA to enter the ribosome A site and bind to the elongating polypeptide chain. When the elongating polypeptide is transferred to this abnormal A position, the polypeptide synthesis is blocked, and the immature C-terminal polypeptide containing puromycin is released. Therefore, we use the puromycin antibody to detect labeled elongating polypeptides, which can dynamically detect protein translation efficiency in a real-time manner. And, this method has been used in our previous report (Wang et al., 2020).

DEFs were infected with DHAV-1 at MOI = 1, and samples were harvested at 6, 12, 24, and 48 h after infection, and control groups were set at each time point. Before harvesting the cells, puromycin (1 μg/ml) was added to the medium, and the cells were incubated at 37°C for 30 min (Wang et al., 2020). As shown in Figure 1A, after DHAV-1 infects DEFs, eIF2α phosphorylation increases with the infection time, while DEFs translation is gradually inhibited. Subsequently, to further prove the effectiveness of DHAV-1 on eIF2α phosphorylation and cellular translation, DEFs were infected with DHAV-1 at 0.2 MOI, 0.4 MOI, 0.6 MOI, 0.8 MOI, and 1.0 MOI, and samples were collected at 24 hpi, and the control group was set at the same time. As shown in Figure 1B, eIF2α phosphorylation increases with the increase of the infectious dose of DHAV-1, while the overall translation of DEFs is gradually inhibited. Interestingly, the viral protein VP3 can still be expressed (Figures 1A,B), indicating that the inhibition of DEFs translation does not inhibit the translation of DHAV-1. This is related to the ability of the internal ribosome entry site (IRES) element of DHAV-1 to initiate cap-independent translation (Liu et al., 2011), and this result is similar to poliovirus (PV; Kastan et al., 2020). The above results show that DHAV-1 infection can induce eIF2α phosphorylation and inhibit cellular translation in DEFs.

Phosphorylation of eIF2α causes total translation stagnation in the cells. The stalled messenger ribonucleoprotein (mRNP) complex aggregates under viral infection stress to form stress granules (SGs). Ras GTPase-activating protein-binding protein 1 (G3BP1) can regulate the SG core network through positive or negative cooperativity with other G3BP1-binding factors (Yang et al., 2020). Therefore, G3BP1 is often used as a marker protein for SGs. Due to lack of duck-derived antibody against G3BP1, we cloned duck-derived G3BP1 gene into pEGFP-C2 vector. As shown in Figure 2, in the control group, EGFP-G3BP1 was uniformly distributed in the cytoplasm. However, EGFP-G3BP1 displayed an obvious granular pattern in the cytoplasm of DHAV-1-infected cells, which was similar to the Ars group. These results indicate that DHAV-1 induces SGs formation in infected cells.

It was previously reported that the foot-and-mouth disease virus VP2 could cause the phosphorylation of eIF2α (Sun et al., 2018), so we speculate whether a certain viral protein of DHAV-1 is the cause of phosphorylation of eIF2α. DEFs were transfected with recombinant plasmids of structural proteins (pCAGGs-VP0-Flag, pCAGGs-VP3-Flag, and pCAGGs-VP1-Flag) and non-structural proteins (pCAGGs-2A-HA, pCAGGs-2B-HA, pCAGGs-2C-HA, pCMV-Myc-3AB, pEGFP-N1-3C, and pCAGGs-3D-HA) of DHAV-1. As shown in Figure 3, compared to the empty vector group, the expression of VP0, VP1, 2A, 3AB, 3C, and 3D proteins did not increase eIF2α phosphorylation. The expression of VP3, 2B, and 2C proteins can slightly increase eIF2α phosphorylation, which may be related to the function of VP3, 2B, and 2C proteins in cells (Cong et al., 2016; Li et al., 2016; Lai et al., 2019). However, none of these viral proteins caused eIF2α phosphorylation as obvious as DHAV-1 infection. These results indicate that viral protein is not the cause of significant eIF2α phosphorylation.

The previous results have shown that the viral protein of DHAV-1 is not the cause of obvious phosphorylation of eIF2α. To further investigate whether the viral activity is the cause of eIF2α phosphorylation, we used UV to inactivate DHAV-1 and then infected DEFs at 1 MOI. As shown in Figures 4A–C, compared to the control group, the UV-inactivated DHAV-1 did not appear obvious CPEs (cells shrink, round, and fragment), no VP3 expression was detected, and no increase in virus copy number, indicating that the UV-inactivated DHAV-1 lost the ability to replicate. Compared with the control group, UV-inactivated DHAV-1 did not cause the phosphorylation of eIF2α and did not affect the translation of DEFs (Figure 4D). These results indicate that UV-inactivated DHAV-1 no longer phosphorylates eIF2α, which is similar to EV71 (Jheng et al., 2010).

Four known kinases phosphorylate eIF2α in cells, namely, PERK, PKR, GCN2, and HRI. However, we only found three kinases PERK, PKR, and GCN2, in duck-derived cells from National Center for Biotechnology Information (NCBI). To better understand which kinase plays a role during DHAV-1 infection, we used PERK, PKR, and GCN2 kinase inhibitors for screening. Due to the lack of antibodies against PERK, PKR, and GCN2, we can only indirectly reflect the three kinases’ activation through the phosphorylation of eIF2α. DEFs were infected with DHAV-1 at 1 MOI. At 22 h after infection, PERK, PKR, and GCN2 kinase inhibitors were added, and cells were harvested after 2 h of treatment (Figure 5A). After adding PERK inhibitor (GSK2606414) and GCN2 inhibitor (GCN2-IN-1), eIF2α phosphorylation was inhibited, indicating that PERK and GCN2 are involved in DHAV-1 induced eIF2α phosphorylation (Figures 5B,C). However, PKR inhibitor (C16) did not inhibit eIF2α phosphorylation, which indicates that PKR is not involved in eIF2α phosphorylation induced by DHAV-1 (Figure 5D). To show whether the PKR kinase of DEFs can be activated, DEFs were transfected with poly(I:C), and then eIF2α phosphorylation and PKR transcription levels were detected. As shown in Figures 5E,F, poly(I:C) can cause the phosphorylation of eIF2α and increase PKR transcription.

In order to exclude the non-specific effects of inhibitors, we used shRNA to knock down PERK or GCN2. We designed three shRNAs against PERK or GCN2 and found that shRNAs could interfere with the transcription of PERK or GCN2 (Figures 6A,B). We also found that DHAV-1-induced the phosphorylation of eIF2α was reduced after PERK or GCN2 knockdown (Figures 6C,D). These results indicate that PERK and GCN2 are involved in eIF2α phosphorylation during DHAV-1 infection.

eIF2 plays an important role in eukaryotic cells, and its alpha subunit phosphorylation controls the cellular translation initiation. Many reports indicate that the virus causes cell translation inhibition through eIF2α phosphorylation, but recently dengue virus (DENV) and mouse norovirus (MNV) have found that cell translation inhibition and eIF2α phosphorylation are not coupled with each other (Roth et al., 2017; Brocard et al., 2020). Since the transfection efficiency of DEFs is not high, the knockdown effect of shRNA is not very good. In the Figures 6C,D, after we knocked down PERK or GCN2 using shRNA, eIF2α phosphorylation was weakened, but the cellular translation level did not increase. The main reason is that the level of the GTP conversion factor eIF2B in the cell is 10–20 times lower than that of eIF2; therefore, small changes in eIF2 phosphorylation can have a significant effect on protein translation (Montero et al., 2008).

To explore the direct relationship between eIF2α phosphorylation and DEFs translation inhibition, we used sodium arsenite (a reagent known to induce phosphorylation of eIF2α) as a positive control. As shown in the Figure 7A, sodium arsenite can induce phosphorylation of eIF2α very well. Using puromycin to label cells, we found that the translation efficiency of cells decreased. This phenomenon is consistent with the result in Figure 1, which indicates that DHAV-1 may inhibit DEFs translation through phosphorylation of eIF2α.

The results described above did not provide a clear answer to the role of eIF2α phosphorylation in translation shutoff. Therefore, we used inhibitors to treat DEFs and then labeled them with puromycin for 30 min to test whether the translation inhibition of DEFs was restored (Figure 5A). As shown in Figures 7B,C after treating DEFs with 10 μM PERK or 10 μM GCN2 inhibitors, the phosphorylation of eIF2α caused by DHAV-1 was reduced, consistent with the previous results (Figures 5B,C). And after the inhibitor treatment, the translation of the cells was restored, which was close to the mock group level. Besides, to ensure that the same inhibitor dose is added, we use 5 μM PERK and 5 μM GCN2 inhibitors to treat DEFs simultaneously. The results showed that the phosphorylation of eIF2α also recovered to the mock group level, and translation inhibition was restored. However, treatment of DEFs with PKR inhibitor did not restore eIF2α phosphorylation as in Figure 5D, nor did it restore the translation level. These results above show that DEFs translation shutoff caused by DHAV-1 is caused by phosphorylation of eIF2α.

Subsequently, we determined the effect of these kinase inhibitor treatments on DHAV-1 propagation. We treated DEFs with corresponding concentrations of kinase inhibitors for 2 h and then infected DEFs with 0.5 MOI or 1.0 MOI DHAV-1, respectively. The virus copy number was measured at 6, 12, and 18 h after infection, and the results showed that treatment with these kinase inhibitors would not affect DHAV-1 propagation (Figure 7D).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.