Cytoplasmic Cargo Receptor p62 Inhibits Avibirnavirus Replication by Mediating Autophagic Degradation of Viral Protein VP2

Avibirnavirus causes severe immunosuppression and mortality in young chickens. VP2, the capsid protein of avibirnavirus, is responsible for virus assembly, maturation, and replication. Previous study showed that avibirnavirus particles could be engulfed into the autophagosome and degradation of virus particles took apart. Selective autophagy is a highly specific and regulated degradation pathway for the clearance of damaged or unwanted cytosolic components and superfluous organelles as well as invading microbes. However, whether and how selective autophagy removes avibirnavirus capsids is largely unknown. Here, we have shown that selective autophagy specifically clears ubiquitinated avibirnavirus protein VP2 by p62 recognition and that p62 is an inhibitor of avibirnavirus replication, highlighting the role of p62 as a potential drug target for mediating the removal of ubiquitinated virus components from cells.

of viruses for degradation is a critical step. This step requires cargo receptors that mediate selective autophagy (7). The modular compositions of the interacting domains or motifs in cargo receptors enable them to link the selected materials to the appropriate sites of developing phagophores and autophagosomes. Several cargo receptors have been extensively characterized, such as p62, NBR1, and NDP52 (8)(9)(10)(11). Most of these receptors recognize ubiquitinated substrates through their ubiquitin-associated domain (UBA) and then link to LC3 on nascent autophagosomes via the LC3-interacting region (LIR) (12)(13)(14).
The modification of cargos with ubiquitin is a well-known selective signal for autophagy. Ubiquitin, often labeled as simply Ub, is a highly conserved small protein with a molecular weight of 7 kDa that was first studied in the ubiquitin-proteasome system (UPS) (15). As a protein modifier, ubiquitin marks a target protein on lysine (Lys) residues in different forms under the action of three enzymes: E1, ubiquitin-activating enzyme; E2, ubiquitin-conjugating enzyme; and E3, ubiquitin protein ligase (16,17). A single ubiquitin molecule (monoubiquitination) or many ubiquitin molecules (polyubiquitination) can be ligated to a Lys residue of the target protein (13,18). Ubiquitin itself has several Lys residues that can be further ubiquitinated and then form polyubiquitin chains. Among the Lys residues available for further ubiquitination, K48 and K63 are the most extensively studied (19). Ubiquitination regulates both proteasomal and autophagy pathways. Generally, K63 chains seem to have a preferential affinity for autophagy cargo receptors, while proteins decorated with K48-, K27-, and K11-linked ubiquitin chains undergo proteasomal degradation (20,21).
Avibirnavirus infectious bursal disease virus (IBDV) belongs to the Birnaviridae family (22,23) and causes a highly contagious and immunosuppressive disease in young chickens. IBDV is a nonenveloped virus and contains two segments of double-stranded RNAs, including segment A and segment B (24,25). The smaller segment B encodes VP1, the RNA-dependent RNA polymerase (26,27). The larger segment A encodes VP5 and the polyprotein pVP2-VP4-VP3, which is autoproteolytically processed to generate precursor VP2 (pVP2), the viral protease VP4, and a major structural protein, VP3, through the proteolytic activity of VP4 (28,29). Subsequently, pVP2 is further cleaved at its C-terminal end to the mature VP2 along with four short peptides (30,31).
The IBDV capsid is a single-layer capsid composed of VP2 on the outer surface and VP3 on the inner surface (24,32). VP2, as an avibirnavirus capsid, plays a key role in virus adsorption, virus invasion, and the induction of host immune response (33). Critical amino acid mutations in VP2 result in immune escape from vaccination against avibirnavirus (34)(35)(36).
The first link between autophagy and avibirnavirus was discovered by electron microscopy visualization in 1976, when the degradation of IBDV within autophagic vacuoles was observed (37). Since then, increasing evidence has demonstrated that avibirnavirus protein VP2 induces autophagy upon virus entry (33). However, whether autophagic induction degrades VP2 is still unclear.
Accordingly, in the present study we demonstrate that the avibirnavirus capsid protein VP2 is efficiently ubiquitinated at lysine residue 411 (K411). Thereafter, p62 recognizes the K411-ubiquitinated VP2 and delivers it to autophagosomes for degradation. Thus, our study highlights the role of p62 in mediating selective autophagy as a means to remove cytoplasmic pathogens.

RESULTS
The avibirnavirus capsid protein VP2 undergoes autophagic degradation. A previous study demonstrated that the degradation of IBDV particles in autophagic vacuoles and the binding of viral protein VP2 to HSP90AA1 activate autophagy by the AKT-MTOR pathway in the early stage of infection (33). Accordingly, we investigated whether VP2, the major capsid protein of avibirnavirus, is subject to autophagic degradation via autophagic induction. The VP2 gene was inserted into lentivirus vector PCDH-CMV-MCS-EF1 to construct a DF-1 cell line that stably expresses VP2. Immunoblot assay and immunofluorescence assay (IFA) using anti-VP2 monoclonal antibody revealed that viral protein VP2 is successfully expressed in the VP2 stably expressed DF-1 cells (Fig. 1A and B). Cell viability assays showed that the activity of the VP2 stably expressed cells is not influenced significantly compared to that of wild-type (WT) DF-1 cells (Fig. 1C). Subsequently, we tested the levels of VP2 in VP2 stably expressed DF-1 cells and 293T cells transiently transfected with Flag-VP2 in the presence or absence of an autophagy inducer or inhibitor. Western blot assays (WB) showed that the autophagy inhibitors wortmannin (Wort) and chloroquine (CQ) significantly increase the level of VP2, while the autophagy activators rapamycin (Rapa) and Earle's balanced salt solution (EBSS) markedly decrease the level of VP2 (Fig. 1D to G). Similarly, the level of VP2 was also measured in avibirnavirus-infected cells treated with autophagy inducer or inhibitor. The results showed that CQ or Wort treatment inhibited the degradation of VP2, while EBSS or Rapa treatment promoted the degradation of VP2 ( Fig. 1H and I), confirming that autophagy regulated the degradation of VP2 produced by avibirnavi- rus. In addition, we further investigated the level of VP2 in 293T cells with ATG5 knockdown (KD), which is required for autophagy (38). The level of VP2 was shown to increase greatly in ATG5 KD 293T cells compared with that in WT 293T cells (Fig. 1J). These data indicate that autophagy plays a critical role in the degradation of the IBDV capsid protein VP2.
Autophagic induction enhances VP2 interaction with the cargo receptor p62. Increasing evidence supports the idea that cargo receptors, especially p62, deliver viral proteins to autophagosomes for selective degradation during infection (39)(40)(41). Accordingly, to investigate the relationship between cargo receptor p62 and viral protein VP2, 293T cells were cotransfected with vectors expressing Myc-VP2 and Flag-p62 for 36 h (or transfected with Flag-p62 or Myc-VP2 alone as negative controls). Coimmunoprecipitation (co-IP) was then performed with anti-Myc or anti-Flag antibody, respectively. Subsequent co-IP assays showed that Flag-p62 interacts with Myc-VP2 ( Fig. 2A and B).
We also investigated viral protein VP2 interaction with p62 during IBDV infection. DF-1 cells transfected with Flag-p62 for 24 h were infected with avibirnavirus for 0, 6, or 12 h. As shown in Fig. 2C, viral protein VP2 in avibirnavirus-infected cells also interacts with p62. Moreover, confocal and structured illumination microscopy (SIM) indicated that VP2 colocalized with endogenous p62 in VP2 transfected cells or avibirnavirus-infected cells ( Fig. 2D and E). Since the autophagy inducer promotes the degradation of VP2, we investigated whether the autophagy inducer affects the interaction of viral protein VP2 and p62. Anti-Flag IP and immunoblot analysis showed that Rapa and EBSS treatment increase the affinity of VP2 for p62 ( Fig. 2F and G). To further verify the effect of autophagy on the VP2 interaction with p62, Flag-VP2 and Myc-p62 were cotransfected into WT or ATG5 KD 293T cells with or without Rapa treatment. Anti-Flag IP showed that the knockdown of ATG5 blunted the effect of Rapa on the affinity of VP2 to p62, suggesting that autophagy itself indeed regulated the VP2 interaction with p62 ( Fig. 2H). Thus, the data described above collectively demonstrate that autophagy induction promotes the binding of p62 to VP2.
p62 is involved in the selective autophagic degradation of VP2. We next investigated whether p62 functions as a cargo receptor to mediate the autophagic degradation of VP2. Therefore, p62 knockdown DF-1 (p62 KD) and p62 knockout (p62 KO) 293T cell lines were generated using Cas9 enzyme ( Fig. 3A and B), and the distribution of VP2 in the p62 KD and KO cells was determined. As shown in Fig. 3C and D, viral protein VP2 did not overlap GFP-LC3 in p62 KD cell lines, while the expression of p62 promoted colocalization between VP2 and GFP-LC3.
Since p62 promoted the localization of VP2 at autophagosomes, we next investigated whether p62 affects the level of VP2. Flag-VP2 was transfected into WT and p62 KO 293T cell lines for 24 h. As shown in Fig. 3E, the level of VP2 mRNA was not significantly different between WT and p62 KO 293T cells, but the level of VP2 protein was much higher in p62 KO cells.
In addition, we conducted a cycloheximide (CHX) chase assay to determine the degradation of VP2. WT and p62 KO 293T cells were transfected with Flag-VP2 for 24 h and treated with 100 ng/ml CHX for 0, 1, or 3 h. As shown in Fig. 3F, the Western blot assay shows that the degradation of VP2 is significantly inhibited in p62 KO cells. Furthermore, the p62 KO cell line was cotransfected with Flag-VP2 and Myc-p62 for 24 h, and a CHX chase assay was then performed. The results show that the expression of p62 rescues the degradation of VP2 (Fig. 3G). Collectively, these data demonstrate that p62 promotes the autophagic degradation of VP2.
Both the UBA and LIR domains are required for p62-mediated degradation of VP2. p62 utilizes the UBA domain to bind to ubiquitinated substrates and the LIR domain to interact with LC3, subsequently linking substrates to LC3-decorated autophagosomes (8,42). Therefore, we constructed the mutants Myc-p62oUBA and Myc-p62oLIR to investigate whether p62 interacts with VP2 through the UBA domain and links VP2 to LC3 through the LIR domain. Flag-VP2 was cotransfected with Myc-p62 or the mutants into 293T cells. Co-IP assays showed that the deletion of the UBA domain (Myc-p62oUBA) significantly decreases the interaction with VP2 (Fig. 4A), suggesting that the UBA domain of p62 is necessary for such interaction.
We also investigated whether the UBA domain or LIR domain is required for the p62-mediated degradation of VP2. We found that neither the UBA-deficient mutant nor the LIR-deleted mutant of p62 promotes the degradation of VP2 (Fig. 4B), suggesting that both UBA and LIR domains are required for p62-mediated degradation of VP2. Subsequently, we further conducted a CHX chase assay to test whether the UBA domain or LIR domain influences the degradation of VP2 in p62 KO 293T cells. As shown in Fig. 4C, compared to p62, the UBA-deleted mutant and the LIR-deleted mutant could not promote the degradation of VP2, confirming that both UBA and LIR of p62 were necessary for the degradation of VP2. Thus, the results described above indicate that binding by both the UBA and LIR domains of p62 to VP2 and LC3, respectively, is necessary for autophagic degradation of VP2, confirming that p62 acts as a cargo receptor linking VP2 to autophagosomes. The ubiquitination of VP2 at lysine 411 is required for p62-mediated autophagic degradation. Since the UBA domain is necessary for the interaction of p62 with VP2, we hypothesized that VP2 binding to p62 is involved in ubiquitination. Accordingly, we first investigated VP2 ubiquitination.
Myc-VP2 and hemagglutinin (HA)-Ub were transfected separately or together into 293T cells for 36 h, and then ubiquitination assays were conducted. Myc-62 and HA-Ub were cotransfected into cells as a positive control (43). Co-IP assays revealed that the viral protein VP2 and p62 were ubiquitinated (Fig. 5A). Further study showed that VP2 was ubiquitinated predominantly via K63-linked rather than K48-linked ubiquitination (Fig. 5B). To further identify the ubiquitination site of VP2, we predicted the sites online. As shown in Fig. 5C, there are six potential ubiquitination sites in VP2. These were each separately mutated from lysine (K) to arginine (R), and the mutants then were subjected to in vivo ubiquitination assays. As shown in Fig. 5D, only mutation of K411R abrogates the ubiquitination. Moreover, p62 no longer interacted with the K411R mutant of VP2 ( Fig. 5E and F). To investigate whether the ubiquitin chain affects the p62 interaction with viral protein VP2, Flag-VP2 or Flag-K411R and Myc-p62 were cotransfected together with or without HA-K48 or HA-K63 into 293T cells. Coimmunoprecipitation assay with anti-Flag MAb showed the K63 ubiquitin chain increased the affinity of wild-type VP2 rather than the K411R mutant to p62 (Fig. 5G). These data suggest that K411 with K63 ubiquitination is required for the interaction of VP2 with p62.
We next investigated whether the K411 ubiquitination of VP2 is required for p62-mediated degradation. 293T cells were cotransfected with Flag-VP2, Flag-VP2-K411R and Myc-N, or Myc-p62. Western blot analysis revealed that the K411R mutant of VP2 is resistant to degradation, unlike WT VP2, in the case of p62 overexpression (Fig.  5H). Thus, the data suggest that the p62-mediated degradation of VP2 depends on its ubiquitination at K411. p62 inhibits avibirnavirus replication. Given that p62 mediates the degradation of avibirnavirus protein VP2, whether p62 inhibits avibirnavirus replication is unknown. To further investigate the effect of p62 on avibirnavirus replication, the rescue of ubiquitination-deficient VP2 IBDV (IBDV-K411R) was performed by the T7 expression system. Rescued virus was verified by observing cytopathic effects (CPEs) and the fluorescence signals of capsid protein VP2 under the microscope (Fig. 6A). We then determined the level of VP2 and the avibirnavirus titer in p62-overexpressing and p62 KD cells infected with IBDV-WT or IBDV-K411R. As shown in Fig. 6B, the VP2 level significantly increased in IBDV-infected p62 KD DF-1 cells, while the level of VP2 had no significant change in IBDV-K411R-infected p62 KD DF-1 cells. Furthermore, virus replication assays showed that, compared to wild-type IBDV, the replication ability of IBDV-K411R no longer remarkably increased in p62 KD DF-1 (Fig. 6C). Consistent with this result, we found that the overexpression of p62 in DF-1 cells decreased the wild-type IBDV replication rather than IBDV-K411R replication ( Fig. 6D and E). Collectively, these results demonstrate that p62 represses the replication of wild-type but not K411R mutant avibirnavirus, confirming that p62 is an avibirnavirus inhibitor via degrading capsid protein VP2.

DISCUSSION
Selective autophagy has an important function in the immune system, where it plays an antiviral role by removing viral components. Some evidence indicates that autophagy restricts viruses by removing cytoplasmic viral components, genomes, or viral particles (44)(45)(46). As selected examples, Smad ubiquitin regulatory factor 1 (SMURF1), an E3-ubiquitin ligase, is required for the p62-dependent autophagic degradation of Sindbis virus (SINV) capsid protein (47); Fanconi anemia group C protein has been reported to interact with the SINV capsid protein and to promote its autophagic degradation (41,48); Homo sapiens Shisa family member 5 restricts hepatitis C virus replication by targeting the viral nonstructural protein 5A (NS5A) for autophagic degradation (46); and histone deacetylase 6 inhibits human immunodeficiency virus 1 replication by forming a complex with apolipoprotein B mRNA-editing enzyme and Journal of Virology catalytic polypeptide-like 3G and promoting autophagy-dependent Vif degradation (49). In addition, galectin 8 suppresses viral infection by autophagic degradation of the viral RNA genome (45). Avibirnavirus VP2 is also reported to undergo lysosomal degradation (50).
In the present study, we revealed that p62-mediated selective autophagy degrades the avibirnavirus capsid protein VP2 in a ubiquitin-dependent manner. Thus, our findings provide the first evidence that autophagy recognizes and degrades virus capsid proteins in a ubiquitin-dependent manner.
p62 is a well-known autophagy cargo receptor and contains the UBA domain and an LC3 interaction region, both of which recognize ubiquitinated substrates and deliver them to autophagosomes for final degradation in the lysosome (8,51). Recently, the regulation of the selective autophagic degradation of the SIN virus was reported to involve p62 binding to its capsid. The same study demonstrated that the knockdown of p62 blocks GFP-LC3 colocalization with the SIN capsid and blunts the ability of autophagy to remove the SIN capsid from infected cells (47). However, whether recognition of the SIN capsid by p62 involves ubiquitination is unclear. However, in the current study, we have demonstrated that p62 binds to avibirnavirus capsid VP2 and that this binding is dependent on the UBA domain of p62 and the ubiquitination of VP2. Either the UBA or LIR domain, as well as the ubiquitination of VP2, is required for the p62-mediated degradation of VP2, suggesting that p62 acts as a bridge connecting VP2 to the autophagosome for degradation.
Autophagy is critical for antibody production (52,53). In fact, avibirnavirus VP2 induces the production of neutralizing antibodies (54,55). Currently, the IBDV subunit vaccine and recombinant vaccine against avibirnavirus are designed based on VP2 (56)(57)(58). In addition, the autophagic regulation of antigen processing and presentation is critical for CD4 ϩ and CD8 ϩ T cell activation (59)(60)(61)(62)(63)(64)(65)(66). Conjugating the human immunodeficiency virus 1 antigen to p62 enhances T-cell-mediated immunity (5). Viral protein VP2, the capsid of avibirnavirus, plays a key role in viral adsorption and viral core assembly and is one of its main immunogens. Our earlier study showed that autophagy can be induced by VP2 at the stage of avibirnavirus entry (33) and that autophagy is inhibited by viral VP3 at a later stage of infection (29). In our present study, we have found that lysine residue 411 of VP2 is modified in avibirnavirus-infected cells in a K63-linked ubiquitination manner and degraded by binding to p62. Thus, we hypothesize that p62-mediated autophagic degradation of VP2 participates in the presentation of VP2 peptide to initiate the adaptive immune response against avibirnavirus. Thus, pharmacological control of p62 expression represents a potential strategy for the treatment of avibirnavirus-induced diseases.
Coimmunoprecipitation, ubiquitination, and Western blot assays. For coimmunoprecipitation assays, 293T cells were prepared in six-well plates and cultured for 8 to 12 h until 70% to 90% confluence before transfection by Lipofectamine 3000 reagent. After 36 to 48 h of transfection, the resultant cells were harvested and lysed at 4°C for 2 h by NP-40 lysis buffer supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF). After centrifugation at 12,000 ϫ g for 10 min at 4°C, the supernatant was collected and treated with 20 l protein A/G plus-agarose at 4°C for 30 min to remove impurities. After centrifugation at 1,000 ϫ g for 5 min at 4°C, the supernatant was incubated with anti-Flag or anti-Myc mouse MAb at 4°C for 2 h. A volume of 80 l protein A/G plus-agarose then was added to these immune complexes, which were incubated at 4°C for 2 h. Finally, the beads were washed with NP-40 five times at 4°C and boiled in 4ϫ protein loading buffer (with sodium dodecyl sulfate, DL-dithiothreitol, and bromophenol blue) for 10 min. The samples were then subjected to Western blot assay.
For ubiquitination assays, Flag-VP2 and HA-Ub were cotransfected into 293T cells for 36 to 48 h. The cells were then harvested and lysed using NP-40 containing PMSF and 5 mM N-ethylmaleimide at 4°C for 2 h. Immunoprecipitation and Western blot assays were then performed.
For Western blotting, the cell samples were subjected to SDS-PAGE after centrifugation at 12,000 ϫ g for 10 min at room temperature and then transferred to a nitrocellulose membrane. After blocking with 5% skimmed milk, the membrane was incubated with the indicated primary antibodies at room temperature for 2 to 4 h or at 4°C overnight. The membrane then was washed with phosphate-buffered saline (PBS) containing 0.1% Tween 20 (PBST) five times and incubated with HRP-labeled secondary antibody at room temperature for 1 to 2 h. The immunoreactive protein bands were detected and imaged using enhanced chemiluminescence and a chemiluminescence imaging system, respectively.
Virus infection. DF-1 cells were cultured on 6-or 12-well plates and infected with IBDV diluted in DMEM at a multiplicity of infection (MOI) of 0.01 or 10. After 1 h of absorption at 37°C, the medium was removed. The infected cells then were washed with PBS three times and cultured in fresh DMEM containing 2% FBS at 37°C until the specified time.

FIG 6 Legend (Continued)
phase-contrast microscope, and the infected cells were incubated with anti-VP2 MAb and FITC antimouse IgG as the primary antibody and the secondary antibody, respectively. The fluorescence signals of VP2 were observed with an inverted fluorescence microscope (Nikon, Japan). Scale bar, 500 m. (B) WT or p62 KD DF-1 cells were infected with IBDV-WT or IBDV-K411R at an MOI of 0.1 for 0, 12, and 24 h and subjected to immunoblot analysis with the indicated antibodies. (C) One-step growth curve of IBDV. WT or p62KD DF-1 cells were infected with IBDV-WT or IBDV-K411R at an MOI of 0.1 at the indicated times, and the TCID 50 was detected as described in Materials and Methods. (D) DF-1 cells transfected with Myc-N or Myc-p62 were infected with IBDV-WT or IBDV-K411R at an MOI of 0.1 for 0, 12, and 24 h. The level of VP2 protein was determined by immunoblotting. (E) Myc-N or Myc-p62 was transfected into DF-1 cells, followed by IBDV-WT or IBDV-K411R infection at an MOI of 0.1 for the indicated times. TCID 50 assay was performed to draw the growth curve of IBDV-WT and IBDV-K411R. Results are presented as the means Ϯ SD from at least three independent experiments. ns, P Ͼ 0.05; *, P Ͻ 0.05; **, P Ͻ 0.01.