Hyperacetylated microtubules assist porcine deltacoronavirus nsp8 to degrade MDA5 via SQSTM1/p62-dependent selective autophagy

ABSTRACT

The microtubule (MT) is a highly dynamic polymer that functions in various cellular processes through MT hyperacetylation. Thus, many viruses have evolved mechanisms to hijack the MT network of the cytoskeleton to allow intracellular replication of viral genomic material. Coronavirus non-structural protein 8 (nsp8), a component of the viral replication transcriptional complex, is essential for viral survival. Here, we found that nsp8 of porcine deltacoronavirus (PDCoV), an emerging enteropathogenic coronavirus with a zoonotic potential, inhibits interferon (IFN)-β production by targeting melanoma differentiation gene 5 (MDA5), the main pattern recognition receptor for coronaviruses in the cytoplasm. Mechanistically, PDCoV nsp8 interacted with MDA5 and induced autophagy to degrade MDA5 in wild-type cells, but not in autophagy-related (ATG)5 or ATG7 knockout cells. Further screening for autophagic degradation receptors revealed that nsp8 interacts with sequestosome 1/p62 and promotes p62-mediated selective autophagy to degrade MDA5. Importantly, PDCoV nsp8 induced hyperacetylation of MTs, which in turn triggered selective autophagic degradation of MDA5 and subsequent inhibition of IFN-β production. Overall, our study uncovers a novel mechanism employed by PDCoV nsp8 to evade host innate immune defenses. These findings offer new insights into the interplay among viruses, IFNs, and MTs, providing a promising target to develop anti-viral drugs against PDCoV.

IMPORTANCE

Coronavirus nsp8, a component of the viral replication transcriptional complex, is well conserved and plays a crucial role in viral replication. Exploration of the role mechanism of nsp8 is conducive to the understanding of viral pathogenesis and development of anti-viral strategies against coronavirus. Here, we found that nsp8 of PDCoV, an emerging enteropathogenic coronavirus with a zoonotic potential, is an interferon antagonist. Further studies showed that PDCoV nsp8 interacted with MDA5 and sequestosome 1/p62, promoting p62-mediated selective autophagy to degrade MDA5. We further found that PDCoV nsp8 could induce hyperacetylation of MT, therefore triggering selective autophagic degradation of MDA5 and inhibiting IFN-β production. These findings reveal a novel immune evasion strategy used by PDCoV nsp8 and provide insights into potential therapeutic interventions.

INTRODUCTION

Porcine deltacoronavirus (PDCoV) is an emerging enteropathogenic coronavirus that causes acute watery diarrhea, vomiting, dehydration, and even death in lactating piglets (1, 2). It was first identified in Hong Kong by molecular surveillance studies in 2012 (3), and the first clinical outbreak of PDCoV in pig farms was announced in the United States in 2014. To date, PDCoV has been reported in some countries and regions, including mainland China, Korea, Thailand, Japan, Lao People’s Democratic Republic, Vietnam, Mexico, and Peru, causing significant economic losses (4–10). In addition to piglets, PDCoV infects calves, turkeys, chickens, and mice, suggesting that PDCoV possesses cross-species transmission (1, 11–15). Importantly, three cases of PDCoV infection in Haitian children were reported recently, highlighting the potential threat of PDCoV to humans. Therefore, PDCoV has been proposed as the eighth coronavirus capable of infecting humans, following human coronaviruses 229E, OC43, NL63, and HKU1, severe acute respiratory syndrome coronavirus, Middle East respiratory syndrome coronavirus, and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (16).

PDCoV is a single-stranded, positive-sense RNA virus belonging to the newly discovered genus Deltacoronavirus in the Coronaviridae family. Its genome length is approximately 25.4 kb (3, 17) and encodes two large polyproteins, pp1a and pp1ab, as well as four structural proteins (S, E, M, and N) and three accessory proteins (NS6, NS7, and NS7a). Papain protease nsp3 and 3C-like protease nsp5 cleave pp1a and pp1ab into 15 mature non-structural proteins (nsp2–nsp16) (18, 19). Among them, non-structural protein 8 (nsp8) is a 189 amino acid polypeptide that is relatively conserved in the Coronaviridae family. It binds to nsp7 and nsp12 to form viral replication and transcriptional complexes and plays an important role in viral replication (20, 21). However, the precise functions of nsp8 in coronavirus infection and pathogenesis are not fully understood compared with other viral proteins. Recently, SARS-CoV-2 nsp8 was reported to be involved in negative regulation of type I interferon (IFN) (22). However, we found that histone deacetylase 6 (HDAC6) inhibits PDCoV proliferation by degrading nsp8 and promoting type I IFN production (23).

IFNs, together with their induced cellular anti-viral responses, serve as the primary defense mechanism against viral infections. When a cell is infected by a virus, viral components and replication intermediates are recognized by host pattern recognition receptors such as cytoplasmic retinoic acid-inducible gene I (RIG-I) and melanoma differentiation gene 5 (MDA5). These viral components are known as pathogen-associated molecular patterns. Activation of RIG-I and MDA5 triggers the formation of mitochondria anti-viral signaling protein (MAVS)-dependent signaling vesicles, which induce the expression of type I IFNs and IFN-stimulated genes, and the subsequent anti-viral state in the cell (24).

The microtubule (MT) is a highly dynamic polymer composed of α- and β-tubulin heterodimers that play multiple roles in cellular functions, including maintaining the cellular structure, cell polarity, and organelle transport. The function of MTs is subject to regulation by various post-translational modifications, including acetylation, de-tyrosination, phosphorylation, methylation, and polyglutamylation (25). Among them, acetylation of α-tubulin occurring mainly at position 40 of lysine (K40) is important to regulate many cellular functions (26), which is mainly modulated by α-tubulin acetyltransferase 1 (ATAT1), HDAC6, and sirtuin 2 (27, 28). Lei et al. demonstrated that HDAC6 reduces MT acetylation to inhibit acidic pH-mediated autophagic vesicle formation in rat cardiomyocytes (29). Many studies have indicated that MTs play important roles in coordinating and organizing many important steps in the functions of autophagosomes. For example, acetylation of α-tubulin enhances endoplasmic reticulum-mitochondrial linkage, thereby promoting the formation of autophagic vesicles (30, 31). MT-associated protein 1A/B (MAP1A/B) and 1S (C19ORF5) interact with microtubule-associated proteins 1 light chain 3B (LC3) and facilitate their association with MTs, suggesting that MTs are involved in the biogenesis and degradation of autophagy (32, 33). Similar to most membrane-bound organelles and vesicles, aspects of autophagic vesicle dynamics depend in part on their interactions with the cytoskeleton, particularly MTs (34). The MT network is extensively exploited by many viruses to facilitate efficient viral infection (35). For example, human immunodeficiency virus type 1 (HIV-1) capsid protein p24 interacts with kinesin to facilitate transport of the HIV-1 capsid to the perinuclear region or MT tissue center (36). Influenza A virus (IAV) infection induces acetylation of α-tubulin in epithelial cells, which increases the release of viral particles from infected cells (37). Additionally, hepatitis E virus open reading frame 3 product interacts with MTs and interferes with their kinetics to establish viral infection (38).

In this study, we initially identified PDCoV nsp8 as an IFN antagonist by interacting with and degrading MDA5. When we dissected how nsp8 degraded MDA5, we unexpectedly found that nsp8 used the MT network by inducing MT hyperacetylation to promote autophagic degradation of MDA5, and this degradation process was dependent on p62-mediated selective autophagy. By hijacking the MT network and inducing MDA5 autophagic degradation, PDCoV nsp8 evades host immune responses.

RESULTS

PDCoV nsp8 is an IFN antagonist

To investigate whether PDCoV nsp8 inhibited IFN-β production, HEK-293T cells were transfected with reporter plasmids IFN-β-Luc and pRL-TK and increasing amounts of pCAGGS-hemagglutinin (HA)-nsp8 or the empty vector. The expression plasmid encoding HA-tagged PDCoV nsp15, which inhibits IFN-β production (39), was used as a positive control. After co-transfection for 24 h, cells were left untreated, infected with Sendai virus (SeV), or transfected with poly(I:C) for 12 h and then lysed to determine IFN-β promoter-driven luciferase activity. SeV and poly(I:C) significantly induced activation of the IFN-β promoter in HEK-293T cells. However, such induction was significantly inhibited by nsp8 expression in a dose-dependent manner (Fig. 1A and B), suggesting that PDCoV nsp8 is an IFN antagonist.

Fig 1.

PDCoV nsp8 is an IFN antagonist. (A and B) HEK-293T cells were co-transfected with IFN-β-Luc and pRL-TK plasmids for 24 h, together with pCAGGS-HA-nsp15 and increasing amounts of pCAGGS-HA-nsp8 (0.1, 0.2, and 0.4 µg/well, 24-well culture plates) or the empty vector, followed by infection with SeV (A) or transfection with poly(I:C) for 12 h (B). Cells were lysed and subjected to dual luciferase assays. Expression of indicated proteins was confirmed by Western blotting with an anti-HA antibody. β-actin served as a protein loading control. (C–F) HEK-293T cells were co-transfected with NF-κB-Luc or IRF3-Luc and pRL-TK plasmids, together with pCAGGS-HA-nsp15 and increasing amounts of pCAGGS-HA-nsp8 (0.1, 0.2, and 0.4 µg/well, 24-well culture plates) or the empty vector, followed by infection with SeV (C and E) or transfection with poly(I:C) for 12 h (D and F). Cells were lysed and subjected to dual luciferase assays. Expression of indicated proteins was confirmed by Western blotting with anti-HA and β-actin antibodies. (G) HEK-293T cells were transfected with pCAGGS-HA-nsp8 or the empty vector and then infected with SeV for 12 h, followed by IFN-β detection using an IFN-β enzyme-linked immunosorbent assay. Results are representative of data from three independent experiments. ***P < 0.001.

Activation of transcription factors IRF3 and nuclear factor-kappa B (NF-κB) and their binding to the IFN-β promoter are critical steps for transcriptional induction of IFN-β (40, 41). Thus, we next explored the effect of nsp8 on IRF3 and NF-κB activation. To this end, HEK-293T cells were co-transfected with luciferase reporter plasmids IRF3-Luc, NF-κB-Luc, and pRL-TK, together with increasing amounts of pCAGGS-HA-nsp8 or the empty vector and then infected with SeV or transfected with poly(I:C) for 12 h. SeV and poly(I:C) significantly induced activation of IRF3 and NF-κB. However, SeV- and poly(I:C)-induced promoter activity of IRF3 and NF-κB was dose-dependently inhibited by nsp8 (Fig. 1C through F).

Next, we determined the effect of nsp8 on IFN-β production at the protein level in the culture supernatant from cells transfected with pCAGGS-HA-nsp8 or the empty vector and then infected with SeV by an enzyme-linked immunosorbent assay (ELISA). Overexpression of nsp8 significantly inhibited SeV-induced IFN-β expression compared with the control (Fig. 1G). Taken together, these results support the notion that PDCoV nsp8 inhibits IFN-β production by blocking activation of IRF3 and NF-κB.

PDCoV nsp8 targets MDA5 and RIG-I to antagonize IFN

PDCoV nsp8 counteracted SeV-induced IFN-β production as well as IRF3 and NF-κB activation, and therefore may target one or several molecules of the RIG-I-like receptor (RLR) pathway to hinder IFN-β induction. To test this hypothesis, we evaluated the effects of nsp8 on important components associated with the RLR signaling pathway, including MDA5-, RIG-I-, RIG-IN- (constitutively active form of RIG-I), MAVS-, TBK1-, IKKε-, IRF3-, and IRF3-5D (constitutively active form of IRF3)-mediated IFN-β production. All tested components of the RLR signaling pathway significantly induced activation of the IFN-β promoter. However, ectopic expression of nsp8 notably inhibited MDA5- and RIG-I-mediated IFN-β promoter activity, but not MAVS-, TBK1-, IKKε-, or IRF3-induced promoter activity (Fig. 2A through E). We also investigated whether nsp8 affected the cGAS/stimulator of interferon genes (STING)-induced IFN response. As shown in Fig. 2F, ectopic expression of nsp8 had no obvious inhibitory effect on STING-mediated IFN-β promoter activity. These results suggested that PDCoV nsp8 disrupts IFN-β production by regulating the expression of MDA5 and RIG-I or the upstream molecules of MDA5/RIG-I.

Fig 2.

PDCoV nsp8 inhibits IFN-β promoter activity mediated by MDA5 and RIG-I. HEK-293T cells were co-transfected with IFN-β-Luc, pRL-TK, and Flag-tagged MDA5 (A), RIG-I, RIG-IN (B), MAVS (C), TBK1, IKKε (D), IRF3, IRF3-5D (E), or STING (F) with or without pCAGGS-HA-nsp8 for 24 h. Cells were lysed and subjected to dual luciferase assays. Firefly luciferase activity is relative to that of the empty vector control with normalization to the Renilla luciferase activity. Expression of the indicated proteins was confirmed by Western blotting with anti-HA and β-actin antibodies. Results are representative of data from three independent experiments. **P < 0.01, ***P < 0.001.

PDCoV nsp8 interacts with MDA5 and RIG-I

To further investigate whether nsp8 specifically targeted MDA5 or RIG-I to block their functions, we assessed the potential interaction of nsp8 with MDA5 or RIG-I. HEK-293T cells were co-transfected with pCAGGS-HA-nsp8 and pCAGGS-Flag-MDA5 or pCAGGS-Flag-RIG-I for 24 h, and then lysates were collected for co-immunoprecipitation (Co-IP) assays with anti-HA or anti-Flag polyclonal antibodies. The results showed that nsp8 interacted with both MDA5 and RIG-I (Fig. 3A through D). To further investigate whether nsp8 interacted with endogenous MDA5 and RIG-I in the context of PDCoV infection, IPI-2I cells were infected with PDCoV for 18 h, followed by Co-IP assays with anti-PDCoV nsp8, MDA5, or RIG-I antibodies. The results showed that PDCoV nsp8 interacted with endogenous MDA5 and RIG-I in PDCoV-infected cells (Fig. 3E through H). Collectively, these results confirmed that PDCoV nsp8 interacts with MDA5 and RIG-I.

Fig 3.

PDCoV nsp8 interacts with MDA5 and RIG-I. (A–D) HEK-293T cells were transfected with pCAGGS-Flag-MDA5 (A and C) or a plasmid carrying RIG-I (B and D), together with pCAGGS-HA-nsp8 or the empty vector for 24 h. Lysates of these cells were assessed by co-immunoprecipitation (Co-IP) assays with anti-HA or anti-Flag antibodies. Whole cell lysates (WCLs) and immunoprecipitation (IP) complexes were analyzed by Western blotting with antibodies against Flag, HA, and β-actin. IPI-2I cells were infected with PDCoV (multiplicity of infection = 0.5) for 18 h. Cells were lysed and subjected to Co-IP assays using anti-nsp8 (E and F), anti-MDA5 (G), or anti-RIG-I (H) antibodies. WCLs and IP complexes were analyzed by Western blotting with antibodies against nsp8, MDA5, RIG-I, and β-actin.

PDCoV nsp8 degrades MDA5 via the autophagic pathway

Figure 3A and C shows that the expression level of MDA5 was obviously decreased by co-expression with PDCoV nsp8. Thus, we focused on whether PDCoV nsp8 disrupted the function of MDA5 by affecting its stability. HEK-293T cells were co-transfected with pCAGGS-Flag-MDA5 or pCAGGS-RIG-I (as a control) and increasing amounts of pCAGGS-HA-nsp8, followed by Western blot analysis. Nsp8 dose-dependently inhibited the expression of MDA5 but not RIG-I (Fig. 4A and B). We further analyzed the protein expression level of endogenous MDA5 in the context of PDCoV infection. The results showed that PDCoV nsp8 expression increased gradually with the progress of viral infection. However, MDA5 expression decreased in a time-dependent manner, suggesting that PDCoV infection degraded MDA5 (Fig. 4C).

Fig 4.

PDCoV nsp8 degrades MDA5 via the autophagic pathway. HEK-293T cells were transfected with pCAGGS-Flag-MDA5 (A) or -RIG-I (B), together with the empty vector or increasing amounts of pCAGGS-HA-nsp8 for 24 h. Cells were harvested for Western blotting. (C) IPI-2I cells were infected with PDCoV [multiplicity of infection (MOI)  =  0.5]. At 6, 12, and 18 h post-infection (hpi), cells were collected for Western blotting assay to detect the expression of MDA5, PDCoV nsp8, and β-actin. (D) HEK-293T cells were co-transfected with pCAGGS-Flag-MDA5 and pCAGGS-HA-nsp8 or the empty vector, treated with or without MG132 (20 µM), Z-VAD (20 µM), 3-methyl adenine (3-MA) (5 mM), CQ (10 µM), NH₄Cl (10 mM), or rapamycin (10 µM) for 12 h, and then subjected to Western blotting. (E) IPI-2I cells were infected with PDCoV (MOI = 0.5) and then treated with MG132 (20 µM), Z-VAD (20 µM), 3-MA (5 mM), CQ (10 µM), NH₄Cl (10 mM), or rapamycin (10 µM). At 18 hpi, cells were harvested for Western blotting to detect MDA5 expression. (F) IPI-2I cells were transfected with increasing amounts of pCAGGS-HA-nsp8. Cells were harvested for Western blotting to measure the LC3-II level and p62 expression. (G) IPI-2I cells were co-transfected with pCAGGS-HA-nsp8 and GFP-LC3 expression plasmids for 24 h, followed by immunofluorescence assay using an anti-HA antibody. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Scale bar, 5 µm. (H and I) Wild type (WT), autophagy-related (ATG)5 knockout (KO) (H), and ATG5 KO HEK-293T (I) cells were co-transfected with pCAGGS-Flag-MDA5 and pCAGGS-HA-nsp8 for 24 h. Cell lysates were subjected to Western blotting with the indicated antibodies. (J and K) IPI-2I cells were transfected with siRNA against ATG5 (J) or ATG7 (K), or control siRNA for 36 h and then infected with PDCoV (MOI = 0.5) for 18 h. Cells were harvested and subjected to Western blotting. DMSO, dimethyl sulfoxide.

Apoptosis, autophagy, and the ubiquitin-proteasome system are the three main intracellular degradation pathways. To determine which pathway was involved in MDA5 degradation induced by PDCoV nsp8, HEK-293T cells were co-transfected with pCAGGS-Flag-MDA5, together with pCAGGS-HA-nsp8 or the empty vector and then treated with Z-VAD-FMK (apoptosis pathway inhibitor), 3-methyl adenine (3-MA), chloroquine and NH₄Cl (autophagy pathway inhibitor), or MG132 (ubiquitin-proteasome pathway inhibitor). Western blotting showed that 3-MA, CQ, or NH₄Cl treatments, to a large extent, reversed nsp8-mediated MDA5 degradation, whereas Z-VAD-FMK or MG132 did not obviously affect MDA5 expression compared with the control dimethyl sulfoxide treatment (Fig. 4D). Furthermore, treatment with rapamycin, an autophagy inducer, enhanced degradation of MDA5 induced by nsp8 (Fig. 4D). These results suggested that the autophagy pathway played a major role in nsp8-mediated MDA5 degradation. To confirm this result, IPI-2I cells were infected with PDCoV, followed by treatment with the above inhibitors as described in Fig. 4D. As expected, PDCoV infection reduced expression of MDA5. 3-MA, CQ, or NH₄Cl largely restored the expression level of MDA5 compared with the control, whereas Z-VAD-FMK or MG132 did not exhibit such an effect (Fig. 4E). Rapamycin treatment further promoted endogenous MDA5 degradation induced by PDCoV (Fig. 4E). Taken together, these results demonstrated that the autophagic pathway was required for PDCoV nsp8-mediated MDA5 degradation.

Previous studies have demonstrated that PDCoV infection induces autophagy (42, 43). We next examined whether PDCoV nsp8 played a crucial role in induction of autophagy. As shown in Fig. 4F, ectopic expression of nsp8 increased the LC3-II level and reduced p62 expression (Fig. 4F), and promoted the formation of GFP-LC3 spots (Fig. 4G), suggesting that nsp8 induced autophagy. Therefore, we further investigated whether PDCoV nsp8-induced autophagy was involved in the degradation of MDA5. To this end, we knocked out essential autophagy genes autophagy-related (ATG) 5 and ATG7 in HEK-293T cells. Nsp8-mediated MDA5 degradation was detectable in wild-type (WT) HEK-293T cells, but not in ATG5 or ATG7 knockout (KO) cells (Fig. 4H and I), suggesting that PDCoV nsp8-mediated degradation of MDA5 was dependent on autophagy. To confirm this conclusion, specific small interfering RNAs (siRNAs) against ATG5 or ATG7 were transfected into IPI-2I cells to knockdown expression of ATG5 or ATG7, respectively, and then the cells were infected with PDCoV, followed by measurement of MDA5 expression. The results showed that MDA5 degradation mediated by PDCoV infection was largely reversed in cells with ATG5 and ATG7 knockdown compared with control cells (Fig. 4J and K). Taken together, these results strongly indicated that PDCoV nsp8 degrades MDA5 via the autophagic pathway.

PDCoV nsp8 degrades MDA5 in a p62-dependent manner

Accumulating evidence indicates that autophagic clearance of protein aggregates is a highly selective process dependent on substrate recognition by cargo receptors (44). To determine which cargo receptor(s) were required for PDCoV nsp8-mediated autophagic degradation of MDA5, we analyzed the effects of cargo receptors NDP52, p62, NBR1, OPTN, TOLLIP, and BNIP3L/NIX on MDA5 expression. Interestingly, overexpression of nsp8 only enhanced p62 and NIX-mediated degradation of MDA5, but not other cargo receptors (Fig. 5A). We next performed Co-IP assays to analyze whether PDCoV nsp8 interacted with p62 and NIX. The results showed that PDCoV nsp8 interacted with p62, but not NIX (Fig. 5B). An immunofluorescence assay (IFA) showed that p62 colocalized with nsp8 in cytoplasm (Fig. 5C). These results showed that PDCoV nsp8 may depend on p62 receptor to enhance autophagic degradation of MDA5.

Fig 5.

PDCoV nsp8 degrades MDA5 in a p62-dependent manner. (A) HEK-293T cells were co-transfected with pCAGGS-HA-nsp8 and pCAGGS-Flag-MDA5, together with constructs carrying individual cargo receptors or the empty vector, followed by Western blotting with anti-Flag, anti-HA, and β-actin antibodies. (B) HEK-293T cells were co-transfected with pCAGGS-HA-nsp8 and constructs carrying Flag-tagged cargo receptors, followed by Western blotting with anti-Flag and β-actin antibodies. (C) IPI-2I cells were co-transfected with pCAGGS-HA-nsp8 and Flag-tagged p62 plasmids for 24 h. Cells were fixed for IFA using anti-HA and anti-Flag antibodies. Nuclei were counterstained with DAPI. Scale bar, 5 µm. (D) WT and p62 KO HEK-293T cells were co-transfected with pCAGGS-Flag-MDA5 and pCAGGS-HA-nsp8 for 24 h. Cell lysates were subjected to Western blotting with the indicated antibodies. (E) WT and p62 KO HEK-293T cells were transfected with pCAGGS-HA-nsp8 or the empty vector for 24 h. Cell lysates were subjected to Western blotting with the indicated antibodies. (F) IPI-2I cells transfected with shp62 or control shRNA (shNC) were infected with PDCoV (MOI = 0.5) for 18 h. Cell lysates were subjected to Western blotting with the indicated antibodies.

To validate this result, we used p62 KO HEK-293T. Western blotting showed that PDCoV nsp8 increased the LC3-II level in WT cells, but not p62 KO cells. Moreover, nsp8-mediated degradation of both exogenous and endogenous MDA5 was observed in WT cells but was largely blocked in p62 KO cells (Fig. 5D and E). These data demonstrated that nsp8-mediated MDA5 degradation was dependent on p62. Additionally, IPI-2I cells were subjected to lentivirus-mediated transduction of short hairpin RNA (shRNA) against p62 (shp62) and then infected with PDCoV, followed by measurement of the endogenous MDA5 expression level. Shp62 transduction obviously downregulated p62 expression and greatly reversed MDA5 expression suppressed by PDCoV compared with the control (Fig. 5F). Taken together, these results demonstrated that nsp8 degrades MDA5 dependently on p62-mediated selective autophagy.

PDCoV nsp8 induces autophagy by promoting MT acetylation

PDCoV nsp8 degraded MDA5 through the autophagic pathway, and therefore we next explored how nsp8 triggered autophagy. Our previous study showed that PDCoV nsp8 interacts with HDAC6 (23), an important deacetylase that regulates MT protein polymerization via deacetylation of α-tubulin and is associated with autophagy (45, 46). Thus, we analyzed the effect of nsp8 on acetylation of α-tubulin, an important indicator of MT acetylation, by Western blotting. Ectopic expression of nsp8 dose-dependently increased the acetylation level of α-tubulin (Fig. 6A) but had no effect on its total protein level, suggesting that nsp8 induced MT hyperacetylation. A MT stabilizer, paclitaxel (Taxol) (47), was used to determine whether MT hyperacetylation induced autophagy in IPI-2I cells. We first performed the cell viability assay for Taxol in IPI-2I cells using a cell counting kit-8 and found that no obvious cytotoxicity could be observed when the concentration of Taxol was up to 1  µM (Fig. 6B). As shown in Fig. 6C, Taxol dose-dependently increased MT hyperacetylation and the LC3-II level and downregulated p62 expression. IFA further showed that Taxol promoted GFP-LC3 dot formation (Fig. 6D). These results demonstrated that MT hyperacetylation induced autophagy.

Fig 6.

PDCoV nsp8 induces autophagy by promoting MT acetylation. (A) IPI-2I cells were transfected with increasing amounts of pCAGGS-HA-nsp8. Cells were then harvested for Western blotting. (B) IPI-2I cells were treated with increasing concentrations of taxol or with DMSO as control for 12 h. The cell viability was measured by using a cell counting kit-8 in accordance with the manufacturer’s protocol. (C) IPI-2I cells were treated with various concentrations of taxol (0.1, 0.2, and 0.5 µM) for 12 h and then harvested for Western blotting. (D) IPI-2I cells were transfected with a GFP-LC3 plasmid for 24 h and then treated with taxol (0.2 µM) for 12 h. Cells were fixed for IFA using an anti-HA antibody. Nuclei were counterstained with DAPI. Scale bar, 5 µm. (E and F) HEK-293T cells were co-transfected with IFN-β-Luc and pRL-TK plasmids, together with Flag-tagged MDA5 for 24 h, treated with increasing amounts of taxol, and then subjected to dual luciferase assays (E) and Western blotting (F). (G and H) HEK-293T cells were co-transfected with IFN-β-Luc and pRL-TK plasmids, together with Flag-tagged MDA5 for 24 h, treated with DMSO, taxol, or nocodazole (1 µM), and then subjected to dual luciferase assays (G) and Western blotting (H). (I) IPI-2I cells were infected with PDCoV (MOI = 0.5) and then treated with DMSO, taxol, or nocodazole. At 18 hpi, the cells were harvested for Western blotting. Results are representative of data from three independent experiments. **P < 0.01, ***P < 0.001. ns, non-significant difference.

We further investigated whether Taxol-mediated MT hyperacetylation degraded MDA5 and inhibited MDA5-mediated IFN-β promoter activation. Dual luciferase and Western blot assays showed that Taxol promoted acetylation of α-tubulin and reduced MDA5 expression in a dose-dependent manner, resulting in the inhibition of MDA5-induced IFN-β promoter activity (Fig. 6E and F). In addition to Taxol, nocodazole (a microtubule polymerization inhibitor) was used to examine the effect of MT acetylation on nsp8-mediated autophagy and MDA5 degradation by dual luciferase and Western blot assays. Nsp8 inhibited MDA5-induced IFN-β promoter activity. The inhibitory effect was further increased after Taxol treatment but almost completely blocked by nocodazole treatment (Fig. 6G). Consistent with the dual luciferase assay results, Western blotting showed that nsp8 promoted α-tubulin acetylation and the LC3-II level, thereby reducing MDA5 expression. This effect was enhanced by Taxol treatment (Fig. 6H). Conversely, nocodazole downregulated α-tubulin acetylation and the LC3-II level, attenuating autophagic degradation of MDA5 induced by nsp8 (Fig. 6I). Similar results were obtained in the context of PDCoV infection. As shown in Fig. 6H, autophagy and MDA5 degradation induced by PDCoV were enhanced by Taxol but weakened by nocodazole. Taken together, these results demonstrated that PDCoV nsp8 induces autophagy by promoting MT acetylation.

PDCoV nsp8 requires MT hyperacetylation to block IFN

Previous studies have shown that the lysine 40 (K40) site of α-tubulin determines its acetylation and that expression of α-tubulin-K40A mutant (a non-acetylatable α-tubulin mutant) prevents hyperacetylation of the MT network but does not affect the overall structure of the MT network (48, 49). To investigate the association of MT hyperacetylation with nsp8-mediated inhibition of IFN-β, we conducted an IFN-β promoter-driven luciferase reporter assay in the presence of the α-tubulin-K40A mutant. HEK-293T cells were co-transfected with reporter plasmids IFN-β-Luc, pRL-TK, and pCAGGS-Flag-MDA5, together with pCAGGS-Flag-α-tubulin or pCAGGS-Flag-α-tubulin-K40A in the absence or presence of pCAGGS-HA-nsp8. Nsp8 significantly inhibited MDA5-mediated IFN-β promoter activity in cells expressing WT α-tubulin. Conversely, expression of α-tubulin K40A prevented this inhibitory effect (Fig. 7A). Consistent with the results from the dual luciferase assay, Western blotting showed that nsp8 increased the LC3-II level in cells expressing WT α-tubulin, resulting in MDA5 degradation compared with the empty vector control (Fig. 7B). Such an inhibitory effect on MDA5 expression was completely abolished in cells expressing α-tubulin K40A (Fig. 7C).

Fig 7.

MT hyperacetylation is required for nsp8 to block IFN. HEK-293T cells were co-transfected with IFN-β-Luc, pRL-TK, and pCAGGS-Flag-MDA5, together with pCAGGS-Flag-α-tubulin or pCAGGS-α-tubulin-K40A in the absence or presence of pCAGGS-HA-nsp8 for 24 h, and then subjected to dual luciferase assays (A) and Western blotting (B–E). (F) IPI-2I cells treated with shp62 or shNC were transfected with pCAGGS-Flag-α-tubulin or pCAGGS-Flag-α-tubulin-K40A and then infected with PDCoV (MOI = 0.5) for 12 h. Cell lysates were subjected to Western blotting with the indicated antibodies. (G) IPI-2I cells were co-transfected with pCAGGS-HA-nsp8 and pCAGGS-Flag-ATAT1 for 24 h. Cells were fixed for IFA using anti-HA and anti-Flag antibodies. Nuclei were counterstained with DAPI. Scale bar, 5 µm. (H) HEK-293T cells were co-transfected with pCAGGS-Flag-ATAT1 and pCAGGS-HA-nsp8 for 24 h. Cells lysates were assessed by Co-IP assays with an anti-HA antibody. WCLs and IP complexes were analyzed by Western blotting. (I, J) HEK-293T cells were co-transfected with IFN-β-Luc, pRL-TK, and pCAGGS-Flag-MDA5, together with pCAGGS-Flag-ATAT1, in absence or presence of pCAGGS-HA-nsp8 for 24 h, and then subjected to dual luciferase assays (I) and Western blotting (J). (K) IPI-2I cells treated with shp62 or shNC were transfected with pCAGGS-Flag-ATAT1 or the empty vector and then infected with PDCoV (MOI = 0.5) for 18 h. Cell lysates were subjected to Western blotting with the indicated antibodies. Results are representative of data from three independent experiments. **P < 0.01, ***P < 0.001. ns, non-significant difference.

Next, we investigated whether MT hyperacetylation was involved in IFN-β inhibition by assessing the effects of WT α-tubulin and its mutants on phosphorylation of the NF-κB p65 subunit and IRF3. As shown in Fig. 7D, nsp8 reduced the phosphorylation levels of p65 and IRF3 as well as MDA5 expression in cells expressing WT α-tubulin compared with the control. However, in cells expressing α-tubulin-K40A, nsp8 did not exert inhibitory effects on phosphorylation of p65 and IRF3 or MDA5 expression (Fig. 7E). Therefore, IFN-β inhibition by nsp8 inducing autophagy required MT hyperacetylation. To strengthen this conclusion, the effects of WT α-tubulin and α-tubulin-K40A on MDA5 expression were evaluated in PDCoV-infected cells with p62 knockdown. As shown in Fig. 7F, WT α-tubulin decreased MDA5 expression compared with α-tubulin-K40A in cells transduced with control shRNA. Such an inhibitory effect of WT α-tubulin on MDA5 expression was attenuated in p62 knockdown cells.

We further investigated the role of ATAT1, the major acetylase of α-tubulin (50), in nsp8-induced MT hyperacetylation, autophagy, and MDA5 degradation. IFA (Fig. 7G) and Co-IP assays (Fig. 7H) showed that nsp8 colocalized and interacted with ATAT1, which suggested that nsp8 activated the acetyltransferase ATAT1, leading to MT hyperacetylation. To further examine whether ATAT1 was involved in MDA5-mediated inhibition of IFN-β by PDCoV nsp8, we performed a dual luciferase promoter activity assay. The results showed that ATAT1 expression promoted the inhibitory effect of nsp8 on MDA5-induced IFN-β promoter activity (Fig. 7I). Consistent with these results, Western blotting showed that ATAT1 enhanced MDA5 degradation induced by nsp8 (Fig. 7J). These data demonstrated that nsp8 induced MT hyperacetylation by linking the acetyltransferase ATAT1 to inhibit MDA5-mediated IFN-β production. Similar results were also obtained in PDCoV-infected cells. As shown in Fig. 7K, ATAT1 increased the LC3-II level and endogenous MDA5 degradation in the context of PDCoV infection. However, the ATAT1-mediated increases of the LC3-II level and MDA5 degradation were obviously attenuated in PDCoV-infected cells with p62 knockdown. Taken together, these results demonstrated that PDCoV nsp8-mediated inhibition of IFN-β via inducing p62-dependent selective autophagy requires MT hyperacetylation.

MDA5 is a common target of nsp8 of swine enteric coronaviruses

Nsp8 is well conserved among the Coronaviridae family and is considered to be a second RdRp encoded uniquely by coronaviruses (51). Thus, nsp8 may represent a potential anti-viral target against coronavirus infection. To determine whether nsp8 orthologs from various mammalian coronaviruses target MDA5 to inhibit IFN production, we selected three additional alpha-coronaviruses, porcine epidemic diarrhea virus (PEDV), transmissible gastroenteritis virus (TGEV), and swine acute diarrhea syndrome coronavirus (SADS-CoV), and a beta-coronavirus (SARS-CoV-2). The results showed that ectopic expression of nsp8 from all four mammalian coronaviruses tested (PEDV, TGEV, SADS-CoV, and SARS-CoV-2) dose-dependently inhibited SeV-induced activation of the IFN-β promoter (Fig. 8A through D), suggesting that all tested coronavirus nsp8s inhibited IFN-β production. Additionally, all tested coronavirus nsp8s interacted with MDA5 (Fig. 8E) and significantly inhibited MDA5-mediated activation of the IFN-β promoter (Fig. 8F), supporting the notion that antagonism of MDA5-mediated IFN production may be a conserved mechanism at least across swine enteric coronaviruses.

Fig 8.

MDA5 is a common target of nsp8 of swine enteric coronaviruses. (A–D) HEK-293T cells were co-transfected with the reporter plasmids IFN-β-Luc and pRL-TK, along with increasing amounts of pCAGGS-HA-nsp8 from one of several different coronaviruses, including SADS-CoV (A), TGEV (B), PEDV (C) or SARS-CoV-2 (D) or empty vector. At 24 h post-transfection, the cells were treated with or without SeV for 12 h and then subjected to dual luciferase assays. The expression of indicated protein was confirmed by Western blotting with anti-HA and β-actin antibodies. (E) HEK-293T cells were co-transfected with pCAGGS-Flag-MDA5 and pCAGGS-HA-nsp8 from one of several different coronaviruses or the empty vector. At 24 h post-transfection, cells lysates were assessed by Co-IP assay with anti-Flag antibodies and a subsequent Western blot analysis. (F) HEK-293T cells were co-transfected with the reporter plasmids IFN-β-Luc and pRL-TK, along with pCAGGS-Flag-MDA5 and pCAGGS-HA-nsp8 from one of several different coronaviruses. At 24 h post-transfection, the cells were harvested and subjected to dual luciferase assays. The expression of indicated protein was confirmed by Western blotting with anti-HA and β-actin antibodies. Results are representative of three independent experiments performed in triplicate. **P < 0.01, ***P < 0.001.

DISCUSSION

During viral infection, proper establishment of anti-viral immunity plays a crucial role in defending against viral replication and multiplication. Thus, viruses have developed multiple mechanisms to subvert innate anti-viral immunity. Our previous study showed that PDCoV antagonizes the IFN response (52). In this study, we found that PDCoV nsp8 promoted hyperacetylation of MTs to induce activation of p62-dependent selective autophagy, resulting in degradation of MDA5 and subsequent inhibition of IFN-β. Our results confirmed that PDCoV exploited hyperacetylation of MTs to escape innate immunity, providing new insights into the evasion strategies used by viruses to subvert the host anti-viral immune response.

Many viruses, including coronaviruses, generate numerous RNA products that are recognized by host cytoplasmic RNA sensors, such as RIG-I and MDA5 (24, 53–56), followed by recruitment of MAVS and subsequent activation of IRF3 and NF-κB. Although RIG-I and MDA5 possess similar domain structures, they employ distinct molecular mechanisms to recognize dsRNA, a replication intermediate of positive-strand RNA viruses (57). Coronaviruses have developed sophisticated strategies to target these molecules and interfere with IFN signaling, such as PDCoV NS6, PEDV nsp7, IAV NS1, and SARS-CoV-2 M (58–61). In the present study, we found that PDCoV nsp8 interacted with both RIG-I and MDA5, resulting in IFN-β inhibition. Interestingly, we observed obvious degradation of MDA5 after co-expression with PDCoV nsp8 (Fig. 3). Thus, we mainly focused on exploration of the potential mechanism of inhibiting the MDA5-mediated IFN response by PDCoV nsp8. Furthermore, we found that nsp8 activated the autophagic pathway and degraded MDA5 in a p62-dependent manner. During our investigations into the degradation pathway of MDA5 upon PDCoV infection, we observed that autophagy inhibitors or knockdown of ATG5 or ATG7 did not completely reverse MDA5 degradation (Fig. 4). These observations suggest that PDCoV employs alternative pathways to downregulate MDA5 expression for immune evasion.

We further determined the structural domains of MDA5 and RIG-I and found that PDCoV nsp8 interacted with the CARD and Hel domains of MDA5 (Fig. S1A and B), and with the Hel domain and CTD of RIG-I (Fig. S2A and B). Upon viral infection, MDA5 2CARD undergoes K63-linked polyubiquitin and recruits MAVS to form a signalosome (62). We hypothesized that PDCoV nsp8 may interrupt the process of K63-linked ubiquitination for MDA5. Indeed, our results showed that overexpression of nsp8 inhibited WT- and K63-linked polyubiquitination of MDA5 but had no obvious effect on its K48-linked polyubiquitination (Fig. S1C through E). A recent study (63) indicated that SARS-CoV-2 nsp8 inhibits activation of IFN responses by interacting with MDA5 and inhibiting its K63-linked polyubiquitination. The interaction of nsp8 with MDA5 CARDs may occupy or shield the epitope or space involved in K63-ubiquitination of MDA5, thereby interfering with formation of the MDA5 signalosome. However, the detailed mechanism and function of PDCoV nsp8 in inhibiting K63-linked polyubiquitination of MDA5 require further exploration.

Previous studies have shown that RIG-I CTD plays a dominant role in high-affinity binding and selectivity of dsRNA (64). The RIG-I CTD-nsp8 interaction raises the possibility that PDCoV nsp8 inhibits RLR-mediated IFN-β production by disrupting recognition or binding of dsRNA by RIG-I. Poly(I:C) pulldown experiments showed that PDCoV nsp8 was a dsRNA-binding protein (Fig. S2C). Furthermore, the competitive binding experiment demonstrated that PDCoV nsp8 competed with RIG-I for binding to dsRNA, supporting this possibility (Fig. S2D). Additionally, a recent study showed that SARS-CoV-2 nsp8 targets STING to inhibit type I IFN responses (22). However, in this study, we found that PDCoV nsp8 did not affect the STING-activated IFN response. These results suggest that PDCoV nsp8 employs multiple strategies to inhibit RLR-mediated IFN production.

MTs, dynamic cytoskeletal polymers composed of α- and β-tubulin heterodimers, play crucial roles in cellular processes, including intracellular transport and structural support (25, 28). Indeed, emerging evidence has shed light on the ability of viruses to manipulate MT dynamics by inducing MT acetylation, which has significant implications for viral replication, such as IAV, HIV, and respiratory syncytial virus (37, 65, 66). On the basis of our previous study showing that HDAC6 interacts with and degrades nsp8 via the proteasome pathway and that HDAC6 acts as a MT acetylation eraser, we assessed the effect of nsp8 on MT acetylation. Our results showed that acetylation of α-tubulin was increased after nsp8 overexpression or PDCoV infection (Fig. 6A and H), demonstrating that nsp8 induced MT acetylation. Several studies have demonstrated the importance of MT acetylation for induction of autophagic vesicle formation and maturation, as well as a direct correlation between MT acetylation levels and autophagic flux (49, 67, 68). p27, a tumor suppressor that promotes autophagy during glucose starvation, stimulates MT acetylation by binding to and stabilizing ATAT1, thereby allowing autophagic vesicles to pass near the centrosome and thus efficiently fuse with lysosomes (69). Increasing evidence has shown that MTs are involved in autophagosome formation (49, 70), autophagosome transport across the cytoplasm (71, 72), and autolysosome formation (68, 73). Using Co-IP assays, specific pharmacological inhibitors/agonists, and dominant negative mutation, we demonstrated that nsp8 interacted with ATAT1 and may promote its acetyltransferase activity to upregulate MT acetylation and that acetylated MTs are required for PDCoV nsp8 to induce autophagy, degrade MDA5, and inhibit IFN-β production (Fig. 6 and 7). Previous studies have indicated that Epstein-Barr virus and its encoded BHRF1 protein induce MT acetylation, modulate mitochondrial dynamics, and trigger mitochondrial autophagy, thereby inhibiting type I IFN production (74, 75). A recent study reported that SARS-CoV-2 nsp8 localizes to mitochondria and induces mitochondrial autophagy by damaging mitochondria (76). Our results also found that ectopically expressed nsp8 localized to mitochondria and may trigger mitochondrial autophagy (data not shown). However, whether PDCoV induces mitochondrial autophagy and, if so, the intricate interplay among PDCoV nsp8, MT acetylation, and mitochondrial autophagy requires further investigation.

Since coronavirus nsp8 is an important component of the viral replication complex (nsp7/nsp8/nsp12) and is absolutely essential for replication of PDCoV and other coronaviruses, the mutated virus with nsp8 deletion or deficiency is impossible for this study. Based on the published structure of SARS-CoV-2 nsp8, we predicted the secondary structure of PDCoV nsp8 via AlphaFold2. According to the predicted structure of PDCoV nsp8, we constructed a series of mutated nsp8 expression plasmids and examined the effects of different nsp8 mutants on the IFN-β and PDCoV replication. The results showed that these nsp8 mutants exhibited different degrees of inhibition on IFN production or of upregulation on viral RNA levels (data not shown). Among them, nsp8 mutants corresponding to the C-terminus of PDCoV nsp8 exhibited the strongest inhibitory effect on IFN production but did not exhibit the strongest increase in viral RNA level, suggesting that PDCoV nsp8 inhibited IFN production independent of its function in viral replication. Indeed, a recent study about the inhibitory role of SARS-CoV-2 nsp8 in IFN production reported a similar case (63).

In summary, our study revealed a novel mechanism through which PDCoV nsp8 dampens innate immunity by manipulating the MT network. We propose a model (Fig. 9) in which PDCoV nsp8 promotes acetylation of MTs, resulting in p62-dependent selective autophagic degradation of MDA5 and inhibition of IFN-β. These results provide novel insights into the complex interplay between viral proteins, the MT network, autophagy, and innate immune responses. Importantly, the identification of PDCoV nsp8 as a regulator of selective autophagy and its ability to modulate MT acetylation position it as a potential target to develop anti-viral drugs against PDCoV. These findings contribute to our understanding of viral immune evasion strategies and provide research directions for therapeutic interventions against PDCoV.

Fig 9.

Proposed model for autophagic degradation of MDA5 by PDCoV nsp8 to antagonize IFN-β production. PDCoV nsp8 acts an IFN antagonist by interacting with and degrading MDA5. Mechanistically, PDCoV nsp8 uses the MT network by inducing MT hyperacetylation, which in turn promotes autophagic degradation of MDA5 in a p62-dependent manner to inhibit IFN-β production.

MATERIALS AND METHODS

Cell culture and viruses

HEK-293T cells and IPI-2I cells were cultured in Dulbecco’s modified Eagle’s medium (HyClone) supplemented with 10% fetal bovine serum. PDCoV strain CHN-HN-2014 (GenBank KT336560), which was isolated from a suckling piglet with severe diarrhea in China in 2014 (77) was used in this study. SeV was obtained from the Centre of Virus Resource and Information, Wuhan Institute of Virology.

Antibodies and reagents

Rabbit anti-ATG5 and anti-ATG7 monoclonal antibodies (mAbs) were obtained from Hangzhou HuaAn Biotechnology. Rabbit anti-sequestosome 1, rabbit anti-p65, anti-p-p65, and anti-LC3B mAbs were obtained from Cell Signaling Technology. Rabbit anti-IRF3, anti-p-IRF3, and anti-β-actin polyclonal antibodies were obtained from ABclonal Technology. Rabbit anti-Flag and anti-HA pAb, and mouse anti-HA and anti-Flag mAbs were purchased from MBL International Corporation. A mouse anti-Ac-α-tubulin mAb was purchased from Santa Cruz Biotechnology. A rabbit anti-α-tubulin antibody was purchased from PTM BioLab. Rabbit anti-RIG-I and -MDA5 polyclonal antibodies were obtained from Proteintech. A mouse anti-PDCoV nucleocapsid (N) protein mAb and rabbit anti-PDCoV nsp8 polyclonal antibody were stored in our laboratory (23). Rapamycin was obtained from InvivoGen. MG132 and Z-VAD-FMK were purchased from Beyotime Biotechnology. Nocodazole and Taxol were purchased from MedChemExpress.

Plasmid construction and transfection

Full-length cDNAs of ATAT1 and α-tubulin were amplified from IPI-2I cells using specific primers and then cloned into the pCAGGS vector with an N-terminal Flag tag to yield pCAGGS-Flag-ATAT1 and pCAGGS-Flag-α-tubulin, respectively. Expression plasmids encoding full-length cDNAs of autophagy receptors have been described previously (78). The eukaryotic expression plasmid encoding PDCoV nsp8 (pCAGGS-HA-nsp8) has also been described previously (23). The eukaryotic expression plasmid encoding PEDV nsp8 (pCAGGS-HA-PEDV-nsp8) was prepared in our laboratory. cDNAs of the nsp8 gene from TGEV, PEDV, and SADS-CoV were amplified using specific primers and cloned into pCAGGS with a HA tag, resulting in expression constructs pCAGGS-HA-TGEV-nsp8, pCAGGS-HA-PEDV-nsp8, and pCAGGS-HA-SADS-CoV-nsp8, respectively. The cDNA encoding SARS-CoV-2 nsp8 was synthesized by Tsingke Biotechnology (China) and cloned into the pCAGGS-HA vector, generating the expression construct pCAGGS-HA-SARS-CoV-2-nsp8. Luciferase reporter plasmids IFN-β-Luc, 4× PRDIII/I-Luc (referred to as IRF3-Luc), 4× PRDII Luc (referred to as NF-κB-Luc), pRL-TK (internal control), and expression constructs encoding Flag-tagged RIG-I, MDA5, MAVS, TBK1, IKKε, and IRF3 have been described previously (79). Expression plasmids encoding Flag-tagged MDA5 and RIG-I truncation mutants (pCAGGS-Flag-2CARD, pCAGGS-Flag-Hel, and pCAGGS-Flag-CTD) have also been described previously (58). GFP-LC3 was a gift from Dr. Hongbo Zhou (Huazhong Agricultural University, China). All expression constructs were validated by DNA sequencing. Primers used in this study are listed in Table S1. For the transfection assay, plasmids were transfected into cells using Jetprime transfection regent (Polyplus) in accordance with the manufacturer’s protocol.

Dual luciferase reporter assay

Cells cultured in 24-well plates were co-transfected with 0.4 µg of the indicated expression plasmids, 0.1-µg reporter plasmid (IFN-β-Luc, IRF3-Luc, or NF-κB-Luc), and 0.02-µg pRL-TK reference plasmid (Promega). At 24 h post-transfection, the cells were stimulated with or without SeV (10 hemagglutinating activity units/well) for 12 h. Subsequently, firefly and Renilla luciferase activities in lysed cells were evaluated by the dual-Luciferase reporter assay system (Promega) in accordance with the manufacturer’s instructions. Representative data from three independently conducted experiments were expressed as relative firefly luciferase activity normalized to Renilla luciferase activity.

RNA interference

siRNA targeting ATG5 (si-ATG5) or ATG7 (si-ATG7) and negative control siRNA (si-NC) were synthesized by Suzhou GenePharma Co., Ltd. shRNA targeting p62 was cloned into the pLKO.1 lentivirus expression vector. These sequences are shown in Table S2. At 60% confluence, cells were transfected with 50-pmol siRNA and then infected with PDCoV. At the indicated times post-infection, cells were collected for Western blotting.

IFA

Cells were seeded on microscope coverslips in 24-well plates. For examination by immunofluorescence microscopy, the cells were fixed with 4% paraformaldehyde for 10 min at room temperature, permeabilized with methanol for 15 min, blocked in 10% bovine serum albumin for 30 min, and then incubated with primary antibodies. After washing with phosphate-buffered saline (PBS), the cells were incubated with the Alexa Fluor 488-conjugated goat anti-mouse IgG (H + L) or Alexa Fluor 594-conjugated goat anti-rabbit IgG (H + L) secondary antibodies for 1 h at 37°C. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole. Fluorescence images were acquired under a confocal laser scanning microscope (Fluoview v.3.1; Olympus, Japan).

Western blotting

Cell lysates were prepared using lysis buffer containing a protease inhibitor cocktail (Beyotime). Equal amounts of proteins were separated by 10% SDS-PAGE and transferred to a polyvinylidene difluoride membrane. The membrane was blocked with 10% dry non-fat milk in 1× Tris-buffered saline with 0.1% Tween 20 (TBST) and then incubated with the specific primary antibodies overnight at 4°C. After washing three times with TBST, the membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG or goat anti-mouse IgG for 1 h at room temperature. After washing, the membranes were visualized by treatment with enhanced chemiluminescence reagents.

Co-IP assay

Cells were transfected with plasmids or infected with viruses for the indicated times and then lysed in lysis buffer. For immunoprecipitation, the lysates were rapidly rotated on a rotary shaker for 30 min, and a portion of each supernatant from the lysed cells was used in whole cell extract assays. The remaining portions of the supernatants from lysed cells were immunoprecipitated with the indicated antibodies overnight at 4°C and then treated with protein A + G agarose beads for 4 h at 4°C. Beads containing the immunoprecipitated samples were washed three times with 1-mL IP lysis buffer. Whole cell extracts and immunoprecipitated samples were resuspended in SDS-PAGE loading buffer, boiled at 95°C for 10 min, and then subjected to Western blotting.

Poly(I:C) pulldown assay

A poly(I:C)-agarose binding assay was performed to determine the RNA-binding ability of PDCoV nsp8 as described previously (80). Poly(I:C)-coated agarose was prepared by incubating poly(C)-coated agarose with poly(I) at 56°C for 30 min, followed by cooling to 4°C. Cell lysates were incubated with poly(I:C)-coated agarose for 4 h at 4°C. The beads were then washed three times with lysis buffer and subjected to immunoblot analysis.

IFN-β ELISA

Cells were transfected with the indicated plasmids and then infected with SeV for 12 h. Culture supernatants were collected and subjected to IFN-β detection using an IFN-β ELISA kit (Solarbio Science & Technology) in accordance with the manufacturer’s instructions.

Cell viability assay

The cytotoxic effects of the drugs on cells were measured using a cell counting kit-8 (Beyotime) according to the manufacturer’s instructions.

Statistical analysis

Significant differences were determined by Student’s t-test or one-way analysis of variance in GraphPad Prism v.8. Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001; ns denotes non-significant difference.

DATA AVAILABILITY

The data underlying this article are available in this article and in its online supplemental material.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jvi.00003-24.

Supplemental material. jvi.00003-24-s0001.pdf.

Fig. S1 and S2; Tables S1 and S2.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.