DEF cells were obtained from 9 to 11-day-old specific pathogen-free duck embryos as described previously (Jacolot et al., 2008) and cultured in Dulbecco's modified Eagle's medium (Gibco) supplemented with 5% FBS (Gibco) and antibiotics (0.1 mg/ml streptomycin and 0.1 mg/ml penicillin) at 37°C in 5% CO₂. DEV CSC strain was purchased from China Institute of Veterinary Drug Control. Mouse monoclonal antibody against glycoprotein B (gB) was kept in our laboratory.

To construct a GFP-LC3 recombination plasmid, the duck LC3B gene was amplified from DEF cells with primers LC3F 5′-ATGCAACCGCCTCTG-3′ and LC3R 5′-TCGCGTTGGAAGGCAAATC-3′, according to the GenBank sequence for duck LC3B (NW_004676873.1), and cloned into pEGFP-C1 plasmid prepared in our laboratory, to express LC3 with the EGFP protein at its N terminus.

DEV of m.o.i 1 was removed after 2 h infection at 37°C. DEF cells were washed three times with sterile PBS (pH7.4) and maintained in 2% FBS in culture medium until samples were harvested. DEF cells were then cultured in fresh media in the absence or presence of the same drug as for pretreatment for the indicated times. Chemicals and their optimal concentrations used in this experiment included 10 mM MP (Sigma), and 5 μM Compound C (Merck–Millipore). At 48, 60, and 72 hpi, DEF cells were collected for subsequent experiments.

Drug-treated, siRNA-transfected and virus-infected cellular proteins were extracted as follows: cell protein was extracted using IP lysis buffer (Beyotime, Jiangsu, China) and protease inhibitor PMSF (Beyotime). Protein samples with 5 × loading buffer were boiled for 10 min, analyzed by 12% SDS-PAGE, and transferred onto nitrocellulose membranes (GE Healthcare Life Sciences, Little Chalfont, UK). The membranes were blocked with 3% BSA (Sigma) for 2 h at room temperature, and then incubated with primary antibody for 2 h at room temperature: rabbit anti-LC3B (L7543), rabbit anti-p-mTOR (SAB4504476), mouse anti-β -actin (A1978), rabbit anti-mTOR (SAB2701842), rabbit anti-pTSC2 (SAB4504003), rabbit anti-TSC2 (SAB4503037) (Sigma–Aldrich), rabbit anti-p-AMPK (44-1150G) or rabbit anti-AMPK (AHO1332) antibody (Thermo Scientific). The membranes were incubated with IRDye 800 CW goat anti-mouse IgG or goat anti-rabbit IgG as secondary antibodies (LI-COR Biosciences) for 1 h at room temperature. Detection was carried out using an Odyssey Infrared Fluorescence Scanning Imaging System (LI-COR Biosciences).

TEM observation of autophagy was carried out as described previously (Alexander et al., 2007). DEF cells in 25-cm² flasks were collected at 48 h after DEV infection, with mock-infected cells as controls. Ultrathin sections were viewed under an H-7650 transmission electron microscope (Hitachi).

For the detection of autophagosomes, DEF cells were transfected with 2.5 μg GFP-LC3 plasmid using the Calcium Phosphate Transfection Kit (Invitrogen) when cells were grown to 70–80% confluence in culture dishes, at 24 hpi. Drug-treated, siRNA-transfected or virus-infected DEF cells were fixed in absolute ethanol for 30 min, and nuclei were stained by DAPI (Sigma). The green fluorescence dots of GFP-LC3 were observed using a Leica SP2 confocal microscopy system (Leica Microsystems).

To study the effects of cell autophagy on viral replication, siRNAs targeting AMPK and TSC2 were synthesized (Shanghai GenePharma). The sequence of siRNA targeting AMPK was: AMPK-1#, GCAGGUCCAGAAGUAGAUATT(sense) and UAUCUACUUCUGGACCUGCTT(antisense); AMPK-2#, GCCAUUCUUGGUAGCCGAATT (sense), UUCGGCUACCAAGAAUGGCTT(antisense). AMPK-3#, GCACAUUAGGCUUCAUAUATT(sense), UAUAUGAAGCCUAAUGUGCTT(antisense). The sequence of siRNA targeting TSC2 was: TSC2-1#, GCUGCUAUCUGGAAGACUATT(sense), and UAGUCUUCCAGAUAGCAGCTT(antisense); TSC2-2#, GCCACUAUAUGUACUCGUATT(sense), and UACGAGUACAUAUAGUGGCTT(antisense); TSC2-3#, GCAAAUGGCAGAGAAGUAATT(sense), UUACUUCUCUGCCAUUUGCTT(antisense). Six-well plates were transfected with siRNAs and negative control RNA (siNC) using transfection reagents for 24 h and then infected with DEV. Cell samples were collected at 48 hpi to detect silencing effects and used in subsequent experiments.

The cell monolayers were infected with DEV serially diluted from 10⁻¹ to 10⁻⁸ in 96-well plates. Virus was removed after infection for 2 h at 37°C and cells were washed three times with sterile phosphate-buffered saline. DEF cells were maintained in 2% FBS in culture medium at 37°C for a further 72 h. At the final time point, we observed and recorded cytopathological changes, and virus titres were determined according to the Reed–Muench method.

Once the cell culture supernatant was discarded, we added 200 μl lysis buffer to the wells in a six-well plate. Cells were centrifuged 12,000 g for 5 min at 4°C and the supernatant was used for subsequent determination of ATP. We then added 100 μl ATP test liquid into inspection wells, which were then incubated for 3–5 min at room temperature, to ensure that all of the background ATP was removed before analysis; 20 μl of the sample was then added to the inspection well, rapid blending, then after at least 2 s, RLU readings were measured using an EnVision Multilabel Reader (Perkin Elmer).

All experimental results are expressed as the mean ± SD. All data were analyzed in three independent experiments. Tukey's test was used for statistical analysis. The difference between two group means is presented as P < 0.05 (^*) and P < 0.01 (^**).

We investigated whether energy metabolism was related to autophagy response during DEV infection. The supply of ATP is an important indicator of cell energy, so ATP levels were measured using the firefly-luciferase-based ATP assay. Changes in ATP levels were indicated by relative light unit (RLU) values, which were measured in three independent repeat tests. ATP levels were reduced in a dose- and time-dependent manner in DEV-infected DEF cells (Figures 1A,B). We also showed that DEV infection had no affect on cell viability (Figure S7). In addition, we counted cell numbers to confirm whether the samples had equal cell viability, to confirm that cellular ATP was indeed decreased in DEV-infected cells. In other words, DEV infection caused energy stress in DEF cells.

We used transmission electron microscopy (TEM) to establish whether the changes in mitochondrial energy metabolism were associated with changes in morphology. DEV-infected DEF cells had reduced electron density in the mitochondrial matrix, but in contrast, mock-infected cells showed mitochondria with characteristic electron-dense matrices (Figure 1C). The original pictures of Figure 1 were shown in Supplementary Materials (Figure S1).

To investigate whether this energy stress can contribute to DEV-induced autophagy, we explored whether consumption of ATP in DEV-infected cells was reversed by methyl pyruvate (MP). MP is a cell-permeable form of pyruvate and can be oxidized in the tricarboxylic acid cycle to produce NADH to fuel electron transport and ATP production (Bhutia et al., 2010). Mock- or DEV-infected DEF cells were treated with or without 10 mM MP for 48 h, and ATP levels were measured using a firefly-luciferase-based ATP assay kit. MP restored ATP production in DEV-infected cells (Figure 2A). The conversion of LC3I to LC3II was examined by western blotting after MP treatment. The changes in autophagy in MP-treated or DEV-infected cells were evaluated by the ratio of LC3II to LC3I. Conversion of LC3I to LC3II clearly increased upon DEV infection when compared with mock-infected cells, suggesting that autophagy was triggered upon DEV infection. LC3II expression was reduced after addition of MP to DEV-infected cells (Figure 2B), suggesting that relief of energy stress by ATP restoration reduced DEV-induced autophagy.

LC3II was also observed by confocal fluorescence microscopy as discrete puncta associated with autophagic vacuoles, and the increase in GFP–LC3 puncta after treatment may reflect increased autophagic activity. As expected, the number of GFP–LC3 puncta decreased dramatically when DEV-infected cells were treated with MP, suggesting that autophagy was inhibited (Figure 2C). Our data demonstrate that DEV-induced disruption of cellular energy contributes to autophagy.

Less DEV glycoprotein B (gB) was observed in MP-treated cells compared with mock-treated cells (Figure 2B). Consistent with this result, DEV titres were significantly reduced in MP-treated cells compared with control cells. TCID₅₀ was used to measure the viral titres (Figure 2D). These results also indicate that the MP-restored energy state inhibited DEV-induced autophagy and DEV replication. In addition, we also showed ATP production inhibitor could enhance DEV replication and autophagy (Figure S6). Our study confirms that DEV infection reduces ATP level, resulting in a type of energy stress, followed by induction of autophagy. The original pictures of Figure 2 were shown in Supplementary Materials (Figure S2).

AMPK is a protein kinase, which is regulated by the energy metabolism in cells. This protein is a key regulator of autophagy (Meley et al., 2006). Lack of energy in cells and changes in AMP level are detected by AMPK, when the AMP/ATP ratio increases, AMPK is activated (Inoki et al., 2003). AMPK also is a key negative regulator of the MTOR pathway (Shang and Wang, 2011). We aimed to determine whether AMPK was involved in DEV-induced autophagy and inhibition of MTOR activity in DEF cells. DEV induced DEF cells autophagy at 48, 60, and 72 h post-infection (hpi), as demonstrated by significantly increased conversion of LC3I to LC3II. DEV increased expression of phosphorylated (p)-AMPK in DEFs cells and increased AMPK activity (Figures 3A,C). DEV also decreased the level of p-mTOR; however, the corresponding expression of LC3II increased (Figures 3B,D). Therefore, it can be speculated that AMPK was involved in mTOR signaling pathways. The original pictures of Figure 3 were shown in Supplementary Materials (Figure S3).

We verified the role of AMPK in DEV-induced autophagy. AMPK is activated in DEFcells infected with DEV. We used Compound C, a known AMPK inhibitor, to assess changes in p-mTOR corresponding to inhibition of AMPK. In DEV-infected cells, inhibition of AMPK partly reversed p-mTOR inhibition, and mTOR activity then inhibited autophagy induced by DEV, suggesting that AMPK is a negative regulator of mTOR (Figure 4A). Formation of GFP-LC3 puncta verified above speculation. Compared with uninfected cells, DEV-infected cells had a significant increase in GFP-LC3 puncta, while which was markedly decreased by Compound C (Figure 4B). We previously found that autophagy influenced DEV replication. Here, we found that Compound-C-treated cells decreased expression of DEV gB (Figure 4A), and reduced production of virus particles at 48 hpi (Figure 4C).

To eliminate nonspecific effects of chemical Compound C, three pairs of siRNAs targeting AMPK were designed, and the silencing effect was detected. The effective maximum of three siRNAs targeting AMPK was AMPK-3# (Figure S5). Then compared activity of cells treated with a negative control RNA (siNC) to the cells treated with the siRNA, and we observed a corroborating decrease in mTOR signaling when AMPK were targeted. Expression of LC3II was significantly reduced by AMPK (Figure S8) inhibition, which significantly reduced synthesis of DEV gB (Figure 4D). At the same time, the virus yield was reduced at 36 hpi (Figure 4E). Similar results were obtained in cells treated with another two siRNAs (siAMPK-1# and siAMPK-2#) (Figure S9). Our data suggests that AMPK regulates DEV-induced autophagy through mTOR, and is necessary for DEV replication. The original pictures of Figure 4 were shown in Supplementary Materials (Figure S4).

TSC2 is an essential player in the coordination of cellular energy levels and cell growth or survival (Inoki et al., 2003). It has been shown that TSC2 is a coordinator of energy levels in cells, AMPK is a sensor of cellular energy and positively regulates TSC2, while MTOR is a downstream target of the TSC2-regulated cell response (Huang and Manning, 2009). To illustrate whether TSC2 is an essential component of AMPK–mTOR in DEV-infected cells, three pairs of siRNAs targeting TSC2 were designed, and the silencing effect was detected. The effective maximum of three siRNAs targeting TSC2 was TSC2-3# (Figure S5). The siRNA was transfected and proved to reverse the inhibition of MTOR and down-regulated LC3II expression, suggesting that TSC2 regulates autophagy via the MTOR pathway. Activation of AMPK had no obvious effect on cells transfected with TSC2-specific siRNA infected with DEV (Figure 5A). Similar results were obtained in cells treated with another two siRNAs (siTSC2-1#and siTSC2-2#) (Figure S10). We hypothesized that AMPK is an upstream regulator of TSC2 in DEV-mediated autophagy. To verify this speculation, we transfected cells with siRNA targeting AMPK, which inhibited activation of AMPK and TSC2 (Figure 5B). GFP-LC3 aggregation was markedly decreased in DEV-infected cells treated with TSC2-specific siRNA compared to cells treated with control siRNA (Figure 5C), which suggested that TSC2 was required for DEV-induced autophagy. Knockdown of TSC2 reduced replication of DEV through the disturbance of autophagy (Figure 5D). These results imply that TSC2 is a pivotal role of the AMPK–mTOR signaling pathway, which contributed to DEV-induced autophagy.

In order to rule out case treatment with drugs or siRNAs could affect cell viability to inteference our results. The effects of the compounds used in this study on cell viability were detected using the WST-1 assay. The viability of the treated cells was almost equal to that of the untreated cells. So, we conclude that the drugs and siRNAs used in our study did not affect DEF viability.

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.