TBK1 Isoform Inhibits Grass Carp Reovirus Infection by Targeting the Degradation of Viral Nonstructural Proteins NS80 and NS38

Grass carp reovirus (GCRV), a genus Aquareovirus of the Reoviridae family, causes extremely high mortality rate among grass carp (Ctenopharyngodon idella) and other fish (1, 2). At present, many strains of GCRV were isolated from different areas of China, which were divided into GCRV-I, GCRV-II, and GCRV-III, three distinct subtypes according to the genome homology, pathogenicity, and biological characteristics (3–5). GCRV-873 is the representative strain of GCRV-I, which contains a genome of 11 dsRNA segments enclosed in a core surrounded with a double-layered icosahedral capsid (6). The 11 genomic segments encode 5 nonstructural proteins (NS16, NS26, NS31, NS38, and NS80) and 7 structural proteins (VP1–VP7) (6, 7). Previous studies have demonstrated that the proteins encoded by GCRV play important roles in viral infection and replication. The VP2 protein of GCRV has the RNA polymerase activity, which promotes the transcription and replication of GCRV (8). The VP5 and VP7 proteins of GCRV are required for virus–cell interaction and mediate virus entry during infection (9, 10). The NS16 protein of GCRV is a fusion-associated small transmembrane protein, which induces cell–cell fusion during viral infection (11). The membrane fusion ability of NS16 could be enhanced by NS26 protein (12). The NS80 and NS38 proteins of GCRV could form viral inclusion bodies (VIBs) during GCRV infection and recruit viral and host factors into the VIBs to promote the replication and assembly of GCRV (13–16).

Fish innate immunity mediated by pattern recognition receptors constitutes the first line of defense against the invading GCRV (17). To survive within infected hosts, GCRV has evolved intricate strategies to antagonize the host’s antiviral immune response. For GCRV-106 from the GCRV-II subtype, the VP41 protein of GCRV inhibits the production of IFN by targeting STING/MITA at the cellular endoplasmic reticulum (18), the VP56 protein of GCRV for suppressing host IFN production by promoting the K48-linked ubiquitination and degradation of IFN regulatory factor 7 (IRF7) through the ubiquitin-proteasome (19), the NS79 protein for decreasing TANK-binding kinase 1 (TBK1)–mediated phosphorylation of IRF3 and IRF7, and the VP3 and VP35 proteins for degrading MAVS through an autophagosome-dependent pathway (20, 21). For GCRV-097 from the GCRV-II subtype, the capsid protein VP4 of GCRV interacts with RIG-I to induce the degradation of RIG-I through the ubiquitin-proteasome pathway (22) and the VP56 protein for enhancing the immune escape function of the VP4 protein (23). Furthermore, GCRV-097 could hijack the host TBK1 to escape IRF7-mediated IFN signaling activation by restraining the K63-linked ubiquitination of TBK1 and promoting its K48-linked ubiquitination (24). For GCRV-873 from the GCRV-I subtype, the NS80 and NS38 proteins of GCRV inhibit the translocation of IRF3 into the nucleus and hijack TBK1 and IRF3 into the VIBs, which restricts the antiviral immune response of grass carp (25).

Alternative splicing is a frequent posttranscriptional regulatory event in eukaryotic organisms that allows individual genes to generate numerous mRNA transcripts and encode functionally diverse proteins (26). These splicing isoforms could affect many biological processes, such as innate immunity (27, 28). The key molecules in RIG-I–like receptors (RLRs) and the cyclic GMP-AMP synthase-stimulator of interferon genes signaling pathway undergo alternative splicing, which has one or more splicing isoforms (29, 30). TBK1 is a key kinase in the innate immune signaling pathway that promotes the phosphorylation and nuclear entry of IRF3 to induce IFN production (31, 32). In mammals, TBK1 has a splice isoform that can bind to RIG-I and inhibit the interaction between RIG-I and MAVS (33). In zebrafish, TBK1 has at least five splicing isoforms. Zebrafish TBK1_tv1 and TBK1_tv2 can disrupt the formation of a functional TBK1-IRF3 complex and restrict the phosphorylation of IRF3 mediated by TBK1 (34). Zebrafish TBK1_tv3 negatively regulates the antiviral immunity signaling pathway by inducing TBK1 and IRF3 degradation through the ubiquitin-proteasome and autophagy-lysosomal pathways, respectively (35). However, the roles of TBK1 splicing isoforms in GCRV replication and infection remain unclear.

In the current study, we demonstrate that grass carp TBK1_tv3 inhibits the replication and infection of GCRV, which is different from the function of zebrafish TBK1_tv3 in spring viremia of carp virus infection (35). Interesting, gcTBK1_tv3 targets the NS80 and NS38 proteins for degradation through the ubiquitin-proteasome pathway. Furthermore, the necessary ubiquitination sites of the NS80 and NS38 proteins were identified. These findings, to our knowledge, shed light on a novel relevance that the TBK1 splicing isoform is involved in the degradation of nonstructural proteins NS80 and NS38, which thereby restricts the production of VIBs and ultimately inhibits GCRV replication and infection.

C. idellus kidney (CIK) cells were grown in MEM supplemented with 10% FBS. GCRV (GCRV-873) was propagated in CIK cells using MEM supplemented with 2% FBS.

Plasmids used in this study, including pTurboGFP vector (Evrogen), p3XFLAG-CMV-14 Expression Vector (Sigma-Aldrich), NS80-GFP, NS80-FLAG, and NS38-FLAG, were previously prepared and stored in our laboratory (25). The plasmids of pCI-neo vector, pCI-neo-VP3, pCI-neo-VP5, and pCI-neo-NS38 were previously constructed (13, 15). The plasmids Ub-hemagglutinin (HA) wild-type (WT), Ub-HA K48, and Ub-HA K63 were obtained from Yossi Yarden (Weizmann Institute of Science, Rehovot, Israel). The gcTBK1_tv3-GFP was obtained using the primer pairs gcTBK1_tv3F1/gcTBK1_tv3R1 and cloned into the pTurboGFP-N vector. YFP-FLAG, gcTBK1-FLAG, gcTBK1_tv3-FLAG, NS80(K279R)-FLAG, NS80(K503R)-FLAG, NS80(K610R)-FLAG, NS38(K197R)-FLAG, and NS38(K328R)-FLAG were obtained using the primer pairs YFPF/YFPR, gcTBK1F/gcTBK1R, gcTBK1_tv3F1/gcTBK1_tv3R1, NS80(K279R)F/NS80(K279R)R, NS80(K503R)F/NS80(K503R)R, NS80(K610R)F/NS80(K610R)R, NS38(K197R)F/NS38(K197R)R, and NS38(K328R)F/NS38(K328R)R and cloned into the p3XFLAG-CMV-14 vector, respectively. The primers used for plasmid constructs are listed in Supplemental Table I.

Healthy grass carps (mean weight 10 ± 1 g) were obtained from Sichuan Province, China. Fish were acclimatized in aerated freshwater with temperature maintained at 25 ± 2°C for 1 wk. The fish were fed a commercial pelleted diet at 3% body weight/d throughout the study. For tissue distribution experiments, seven tissue types, including liver, heart, intestine, spleen, brain, gill, and kidney, were collected from three random untreated grass carps and used for quantitative RT-PCR (qRT-PCR). For GCRV infection, grass carps were i.p. injected with 200 μl of GCRV-873 (1 × 10⁸ PFU/ml), whereas fish of the control group were injected with an equal volume of PBS. These fish were kept under the same conditions as mentioned above. Three individual fish were collected at the indicated times postinfection, including 6, 24, 48, and 72 h postinjection (hpi), respectively. The gill, spleen, and intestine were collected and used for qRT-PCR. All animal experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the Institute of Hydrobiology, Chinese Academy of Sciences.

The anti-FLAG (#F2555) rabbit mAb and anti-FLAG (#F3165) and anti-HA (#H3663) mouse mAbs were purchased from Sigma-Aldrich. The anti-pTurboGFP rabbit polyclonal Ab (#AB513) and anti-GAPDH mouse mAb (#60004-1-Ig) were purchased from Evrogen and Proteintech, respectively. Polyclonal Ab against zebrafish TBK1 was generated according to our previous method (34). The anti-NS38, anti-NS80, anti-VP3, and anti-VP5 polyclonal rabbit Abs against the GCRV-873 strain were produced by immunizing with the prokaryotic recombinant proteins in the Escherichia coli system according to our previous report (36). The specificity of the anti-NS80 and anti-VP5 polyclonal rabbit Abs against the GCRV-873 strain has been confirmed in the previous study (36). The specificity of the anti-NS38 and anti-VP3 polyclonal rabbit Abs against the GCRV-873 strain is provided in Supplemental Fig. 1, which is consistent with the specificity of the anti-NS38 and anti-VP3 polyclonal mouse Abs (36). Goat anti-mouse Ig-HRP conjugate secondary Ab, goat anti-rabbit Ig-HRP conjugate secondary Ab, Alexa Fluor 488–conjugated secondary Ab against mouse IgG, Alexa Fluor 594–conjugated secondary Ab against rabbit IgG, DAPI, Lipofectamine 2000, and protease inhibitor mixture were purchased from Thermo Fisher Scientific. DMSO and the FLAG Immunoprecipitation Kit were purchased from Sigma-Aldrich. MG132 (S2619), PS-341 (S1013), 3-methyladenine (3-MA; S2767), ammonium chloride (NH₄Cl; E0151), and cycloheximide (CHX; S7418) were purchased from Selleck Chemicals.

CIK cells grown in 12-well plates were transfected with 1000 ng FLAG empty plasmid, gcTBK1-FLAG, or gcTBK1_tv3-FLAG, respectively. After 24 h posttransfection, cells were infected with GCRV at a multiplicity of infection (MOI) of 5, 1, or 0.1 in serum-free MEM medium at 25°C for 1 h. Following adsorption, cells were washed with PBS to remove nonadsorbed virions. Then, the infected cells were maintained in 2% FBS MEM at 25°C for 24 h. The supernatants in 12-well plates were collected to measure viral titers using standard plaque assays. Twelve-well plates fixed in 4% paraformaldehyde (PFA) for 1 h were stained with 1% crystal violet and photographed.

For confirming the endogenous protein expressions of gcTBK1 and gcTBK1_tv3, 1 × 10⁶ CIK cells seeded in six-well plates were infected with GCRV at an MOI of 1. At 6 and 12 hpi, the infected CIK cells were collected for protein extraction and Western blotting.

The total RNA was extracted from the samples of healthy or GCRV-infected grass carps using TRIzol reagent according to the manufacturer’s instructions. The concentration of total RNA was determined using the spectrophotometer (NanoDrop 2000; Thermo Fisher Scientific). RNase-free DNase I (Thermo Fisher Scientific) was used to remove genomic DNA remnants at 37°C for 30 min. The cDNA was synthesized using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). qRT-PCR was performed on a Bio-Rad CFX96 C1000 thermal cycler using iQ SYBR Green Supermix (Bio-Rad, Singapore) under the following conditions: 3 min at 95°C, followed by 45 cycles of 10 s at 95°C, 15 s at 60°C, and 10 s at 72°C. All reactions were performed in triplicate and the mean value recorded. The specificity of primers was confirmed by sequencing and melting curves. The housekeeping genes, including β-actin, EF-1α, and 18S rRNA, were used to normalize cDNA amounts. The fold changes relative of target genes to the corresponding control group were calculated using the 2^−ΔΔCt method. All primers used for qRT-PCR are shown in Supplemental Table I.

To investigate the effects of gcTBK1 or gcTBK1_tv3 on the expressions of GCRV proteins, 1 × 10⁶ CIK cells seeded in six-well plates were transfected with indicated plasmids. After 48 h, the transfected CIK cells were collected for protein extraction. To investigate the effect of gcTBK1_tv3 on the endogenous expressions of NS38 and NS80 proteins, 1 × 10⁶ CIK cells seeded in six-well plates were transfected with FLAG or gcTBK1_tv3-FLAG for 24 h and then infected with GCRV at an MOI of 0.5. At 24 hpi, the infected CIK cells were collected for protein extraction. To investigate whether gcTBK1_tv3 could affect the degradation of NS80 and NS38 proteins, 1 × 10⁶ CIK cells seeded in six-well plates were transfected with indicated plasmids for 24 h and then treated with 100 μg/ml CHX, an inhibitor of protein synthesis (37). After culturing for 0, 4, 8, and 12 h, the treated CIK cells were collected for protein extraction.

To investigate the impacts of MG132 and PS-341 on the NS38 and NS80 protein levels, 1 × 10⁶ CIK cells seeded in six-well plates were transfected with NS80-GFP or NS38-FLAG for 24 h and then treated with MG132 or PS-341 at the indicated concentration. After 6 h, the treated cells were collected for protein extraction. To determine the mechanism of gcTBK1_tv3 on the protein degradations of GCRV proteins, 1 × 10⁶ CIK cells seeded in six-well plates were transfected with indicated plasmids for 24 h and then treated with 40 μM MG132, 40 μM PS-341, 40 mM NH₄Cl, or 10 mM 3-MA. After 6 h, the treated cells were collected for protein extraction.

To investigate the possible interaction between gcTBK1_tv3 and the exogenous proteins of GCRV, CIK cells seeded in 10-cm² dishes were transfected with various indicated plasmids for 24 h and then treated with 20 µM PS-341 or MG132 for 6 h or left untreated. To investigate the possible interaction between gcTBK1_tv3 and the endogenous proteins of GCRV, CIK cells seeded in 10-cm² dishes were transfected with YFP-FLAG or gcTBK1_tv3-FLAG for 24 h and then infected with GCRV at an MOI of 1 or left untreated. At 18 hpi, these cells were treated with 20 µM PS-341 for 6 h. At 24 hpi, these cells were collected and used for protein extraction using 600 μl immunoprecipitation (IP) lysis buffer containing protease inhibitor mixture. Cellular debris was removed by centrifugation at 12,000 × g for 10 min at 4°C. Coimmunoprecipitation (Co-IP) was performed using the FLAG-tagged Protein Immunoprecipitation Kit according to the manufacturer’s manual. The agarose was washed six times with ice-cold washing solution, and protein was eluted with elution buffer.

For Western blotting analysis, the whole-cell extracts were subjected to 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes, followed by blocking with 5% nonfat milk in TBST for 1 h. The membrane was washed and then incubated with primary Ab overnight at 4°C. The primary Abs, including anti-GAPDH (1:5000), anti-FLAG (1:5000), anti-pTurboGFP (1:5000), anti-NS80 (1:5000), anti-NS38 (1:5000), anti-VP3 (1:5000), or anti-VP5 (1:5000), were used. After washing with TBST, the membrane was next incubated with goat anti-mouse Ig-HRP conjugate secondary Ab (1:5000) or goat anti-rabbit Ig-HRP conjugate secondary Ab (1:5000) for 1 h at room temperature. The bands were detected using Pierce ECL Western blotting Substrate and ECL Western blot system (LAS 4000mini).

To determine the possible colocalization of gcTBK1_tv3 with NS38 or NS80 protein, CIK cells were plated onto coverslips in 24-well plates and then transfected with 1000 ng FLAG or gcTBK1_tv3-FLAG. After 24 h posttransfection, the cells were infected with GCRV for 1 h or left untreated. The cells were washed with PBS to remove nonadsorbed virions, and the infected cells were maintained in 2% FBS MEM. At 18 hpi, the cells were treated with 20 μM PS-341 at 25°C for 6 h or left untreated. Then, the cells were washed twice with PBS and fixed with 4% PFA for 1 h. After being washed three times with PBS, the cells were incubated with anti-FLAG (1:1000), anti-NS80 (1:500), or anti-NS38 Ab (1:500), followed by incubation with Alexa Fluor 488–conjugated secondary Ab against mouse IgG (1:400) and Alexa Fluor 594–conjugated secondary Ab against rabbit IgG (1:400).

To determine the role of gcTBK1_tv3 or gcTBK1 on the production of VIBs during GCRV infection, CIK cells were plated onto coverslips in 24-well plates and then transfected with 1000 ng FLAG vector, gcTBK1-FLAG, or gcTBK1_tv3-FLAG. After 24 h posttransfection, the cells were infected with GCRV for 1 h. The cells were washed with PBS to remove nonadsorbed virions, and then, the infected cells were maintained in 2% FBS MEM. At 18 hpi, the cells were treated with 20 μM PS-341 at 25°C for 6 h or left untreated. Then, the cells were washed twice with PBS and fixed with 4% PFA for 1 h. After being washed three times with PBS, the cells were incubated with anti-FLAG (1:1000), anti-NS80 Ab (1:500), or anti-NS38 Ab (1:500), followed by incubation with Alexa Fluor 488–conjugated secondary Ab against mouse IgG (1:400) and Alexa Fluor 594–conjugated secondary Ab against rabbit IgG (1:400). DAPI staining was applied to detect the cell nucleus. After each incubation step, cells were washed with PBS. Finally, the coverslips were washed, and the images were obtained using an SP8 Leica laser confocal microscopy imaging system.

To determine the possible self-ubiquitylation of gcTBK1 or gcTBK1_tv3, the YFP-FLAG, gcTBK1-FLAG, or gcTBK1_tv3-FLAG and Ub-HA WT plasmids were cotransfected into CIK cells. After 24 h posttransfection, the cells were washed with ice-cold PBS buffer and then lysed in IP lysis buffer containing Protease Inhibitor Cocktail. To determine the possible ubiquitylation of gcTBK1_tv3 on the NS80 or NS38, the FLAG and GFP empty plasmids, gcTBK1_tv3-GFP, NS80, NS80 mutant plasmids, NS38, NS38 mutant plasmids, Ub-HA WT, Ub-HA K63, and/or Ub-HA K48 plasmids were cotransfected into CIK cells with the indicated combination. At 24 h posttransfection, cells were treated with PS-341 (20 μM) or MG132 (20 μM) for 6 h. Following this, the cells were washed with ice-cold PBS buffer and then lysed in IP lysis buffer containing Protease Inhibitor Cocktail. Ubiquitination assays were performed using the FLAG-Tagged Protein Immunoprecipitation Kit according to the manufacturer’s manual. The agarose was washed four times with ice-cold wash solution, and protein was eluted with elution buffer. Total lysate and eluted proteins were analyzed by Western blot analysis using anti-GAPDH (1:5000), anti-FLAG (1:5000), anti-HA (1:5000), and anti-TurboGFP (1:5000) Abs.

Statistical analysis was performed and graphs produced using GraphPad Prism 7.0 software. Data from qRT-PCR are presented as mean and SEM. The significance of results was analyzed by Student t test or ANOVA test (*p < 0.05 and **p < 0.01).

Splicing isoforms of TBK1 exist in both mammals and zebrafish (33–35). When cloning grass carp TBK1 (gcTBK1), an isoform of gcTBK1, which has a 1-aa difference with zebrafish TBK1_tv3, was identified and validated by sequencing. Therefore, we named the short isoform of gcTBK1 as gcTBK1_tv3 (GenBank accession number: ON746275; https://www.ncbi.nlm.nih.gov/nuccore/ON746275). The sequence alignment of amino acids revealed that gcTBK1_tv3 differed from gcTBK1 mainly at the C terminus (Fig. 1A). Compared with gcTBK1, the C-terminal T2K kinase domain is missing for gcTBK1_tv3 (Fig. 1B).

Previous study has investigated the expression patterns of gcTBK1 in different tissues and in the response to GCRV infection (38). In this study, the constitutive and inducible expressions of gcTBK1_tv3 at the mRNA level were examined by qRT-PCR. The mRNA levels of gcTBK1_tv3 were the highest in the liver, then in the heart and intestine, and the lowest in the kidney (Fig. 1C). To investigate the effects of GCRV on the regulation of gcTBK1_tv3 expression in immune tissues, the mRNA levels of gcTBK1_tv3 were examined in systemic (spleen) and mucosal (gill and intestine) immune tissues of grass carps infected with GCRV for 0, 6, 24, 48, and 72 h. GCRV infection significantly induced the expressions of gcTBK1_tv3 in the gill only at 24 hpi and in the spleen and intestine only at 6 hpi (Fig. 1D).

The sizes of gcTBK1 and gcTBK1_tv3 were confirmed by Western blotting in CIK cells transfected with gcTBK1-FLAG or gcTBK1_tv3-FLAG (Fig. 1E). Because sequence identity between zebrafish and grass carp TBK1 is >93%, the zebrafish TBK1 Ab was used to detect gcTBK1 and gcTBK1_tv3 proteins in CIK cells. The protein band located between 55 and 70 kDa corresponded to gcTBK1_tv3 protein (Fig. 1E, 1F). The bigger protein band located between 70 and 100 kDa corresponded to gcTBK1 protein. The smaller protein band located between 70 and 100 kDa may be corresponding to gcTBK1_tv1 protein (Fig. 1E, 1F).

Together, these results demonstrate that gcTBK1_tv3 is the splicing isoform of gcTBK1 and induced by GCRV infection.

Previous study revealed that gcTBK1 inhibited GCRV replication at the high infected titer; however, it promoted GCRV replication at the low infected titer (24). To investigate the role of gcTBK1_tv3 during GCRV infection, the gcTBK1-FLAG and gcTBK1_tv3-FLAG were constructed. When CIK cells were infected with GCRV with an MOI of 5, overexpression of gcTBK1 or gcTBK1_tv3 significantly enhanced cell resistance to GCRV infection. When CIK cells were infected with GCRV with an MOI of 1 or 0.1, overexpression of gcTBK1 weakly enhanced the sensitivity of cells to GCRV infection; however, overexpression of gcTBK1_tv3 still enhanced cell resistance to GCRV infection (Fig. 2A). Compared with the control group transfected with the FLAG empty plasmid, GCRV titers with an MOI of 5 were obviously decreased in CIK cells transfected with gcTBK1 or gcTBK1_tv3 (Fig. 2B). When CIK cells were infected with GCRV with an MOI of 1 or 0.1, GCRV titers in CIK cells transfected with gcTBK1 were significantly higher than those in the control group transfected with the FLAG empty plasmid; however, GCRV titers in CIK cells transfected with gcTBK1_tv3 were significantly lower than those in the control group transfected with the FLAG empty plasmid (Fig. 2C). Taken together, these results suggest that gcTBK1_tv3 plays a protective role on host defense against GCRV infection.

The above results demonstrated that gcTBK1_tv3 suppressed GCRV replication and infection at both the high and low infected titer; we next examined whether gcTBK1_tv3 might destabilize the proteins of GCRV. CIK cells were cotransfected with gcTBK1_tv3 or gcTBK1 and NS80 or NS38, two vital proteins for forming the VIBs. When the CIK cells were cotransfected with gcTBK1 and NS80 or NS38, gcTBK1 has no influence on the expressions of NS80 and NS38 proteins (Fig. 3A, 3B). Different from gcTBK1, gcTBK1_tv3 decreased the protein levels of NS80 and NS38 (Fig. 3C, 3D). We next investigated whether gcTBK1_tv3 also destabilized other GCRV proteins. CIK cells were cotransfected with gcTBK1_tv3 and VP3, VP5, or NS31. However, gcTBK1_tv3 failed to destabilize VP3, VP5, and NS31 proteins of GCRV (Fig. 3E–G). Furthermore, gcTBK1_tv3 decreased the protein levels of NS80 and NS38 in the case of GCRV infection (Fig. 3H).

The CHX chase experiment was used to investigate whether gcTBK1_tv3 could affect the synthesis or degradation of NS80 and NS38 proteins. The results of (Fig. 3I–L showed that overexpression of gcTBK1_tv3 significantly increased the degradation of NS80 (Fig. 3I, 3J) and NS38 (Fig. 3K, 3L) proteins of GCRV after using CHX to block protein synthesis.

Collectively, these results suggest that gcTBK1_tv3 specifically promotes the degradation of NS80 and NS38 proteins of GCRV.

To investigate the mechanisms involved in the gcTBK1_tv3-mediated degradation of NS80 and NS38 proteins during the posttranscriptional process, the effects of MG132 or PS-341 (inhibitor for the ubiquitin-proteasome) and the autophagosome-lysosome inhibitors 3-MA and NH₄Cl on the gcTBK1_tv3-mediated degradations of NS80 and NS38 were detected in CIK cells. Because MG132 could impair the expression of NS80 but PS-341 could maintain the stability of NS80 protein (Fig. 4A, 4B), PS-341 instead of MG132 was chosen for examining the effects of proteasome inhibitor on the gcTBK1_tv3-mediated degradations of NS80. Furthermore, both MG132 and PS-341 stabilized the NS38 protein (Fig. 4C, 4D). Moreover, the present results showed that the degradations of NS80 and NS38 proteins were not attenuated by 3-MA and NH₄Cl, but attenuated by PS-341 or MG132 (Fig. 4E, 4F). The inhibition of gcTBK1_tv3-mediated degradation of NS80 or NS38 proteins by PS-341 or MG132 was independent of the dose variation (Fig. 4G, 4H). Collectively, these results suggest that gcTBK1_tv3 promotes the degradation of NS80 and NS38 proteins via the ubiquitin-proteasome pathway.

To investigate whether gcTBK1_tv3 directly targeted the NS80 and NS38 proteins of GCRV to degrade them, the possible interactions between gcTBK1_tv3 and the proteins of GCRV were investigated. gcTBK1_tv3-FLAG and NS80-GFP plasmids were cotransfected into CIK cells. The anti-FLAG-M2 affinity gel-immunoprecipitated gcTBK1_tv3 could associate with the exogenous NS80 in the case or the absence of PS-341 (Fig. 5A, 5B). The interaction between YFP-FLAG and NS80 was examined as the negative control. As shown in (Fig. 5A and 5B, no NS80 band was observed. Furthermore, the interaction between gcTBK1_tv3 and the exogenous NS38 was observed in CIK cells cotransfected with gcTBK1_tv3-FLAG and pCI-neo-NS38 in the presence or the absence of MG132 (Fig. 5C, 5D). The interaction between gcTBK1_tv3 and the endogenous NS38 or NS80 was also confirmed in CIK cells transfected with gcTBK1_tv3-FLAG and infected with GCRV. Different from NS38 or NS80, no interaction between gcTBK1_tv3 and the endogenous VP3 or VP5 was observed in the GCRV-infected cells (Fig. 5E, 5F).

We next investigated whether gcTBK1_tv3 colocalized with NS80 or NS38 in CIK cells during GCRV infection. Under the degradation of NS80 or NS38, protein mediated by gcTBK1_tv3 was inhibited by the ubiquitin–proteasome inhibitor, and the obvious colocalizations were observed for gcTBK1_tv3 and NS80 or NS38 (Fig. 6).

Collectively, these results suggest that gcTBK1_tv3 specifically interacts with the NS80 and NS38 proteins of GCRV.

In mammals, TBK1 has been shown to be an E3 ubiquitin ligase that promotes self-ubiquitination and can degrade multiple picornavirus VP3 proteins via K63 ubiquitination (39). Similar to mammalian TBK1, the present results from Co-IP showed that gcTBK1 and gcTBK1_tv3 could also be self-ubiquitylated (Fig. 7A, 7B). As the aforementioned results suggest that gcTBK1_tv3 degrades NS80 and NS38 through the ubiquitin-proteasome pathway, we next performed in vitro ubiquitination assays to determine whether gcTBK1_tv3 promotes the ubiquitination of NS80 and NS38. NS80-FLAG or NS38-FLAG was cotransfected with Ub-HA WT and gcTBK1_tv3-GFP into CIK cells, respectively. In the presence of PS-341 or MG132, the ubiquitination of NS80 or NS38 was obviously increased by gcTBK1_tv3 (Fig. 7C, 7D). To determine the specific ubiquitination type of NS80 or NS38, the CIK cells were cotransfected with NS80-FLAG or NS38-FLAG with gcTBK1_tv3-GFP, HA-tagged K48-linked ubiquitin (Ub-HA K48), or K63-linked ubiquitin (Ub-HA K63). The K48-linked ubiquitination of NS80 or NS38 was increased by gcTBK1_tv3. However, the gcTBK1_tv3 has no obvious effect on the K63-linked ubiquitination of NS80 or NS38 (Fig. 7E, 7F). Together, these results demonstrate that gcTBK1_tv3-mediated K48-linked ubiquitination promotes proteasomal degradation of NS80 and NS38 proteins.

We further investigated the specific sites of gcTBK1_tv3-mediated ubiquitinations for the NS80 and NS38 proteins of GCRV. Using the UbPred program (http://www.ubpred.org/) revealed that three residues (K279, K503, and K610) of NS80 and two residues (K197 and K328) of NS38 might be potential ubiquitination sites. Three single-site mutants (K279R, K503R, and K610R) were constructed for NS80 protein. Compared to WT NS80, only the K503R mutant completely abolished gcTBK1_tv3-mediated degradation of NS80 (Fig. 8A). Furthermore, gcTBK1_tv3-mediated ubiquitination of NS80 was completely abolished by the K503R site mutant of NS80 (Fig. 8B). For NS38 protein of GCRV, two single-site mutants (K197R and K328R) were created. Compared to WT NS38, the K328R mutant completely abolished gcTBK1_tv3-mediated degradation of NS38 (Fig. 8C). gcTBK1_tv3-mediated ubiquitination of NS38 was abolished by the K328R site mutant of NS38 (Fig. 8D). Together, these results demonstrate that the Lys⁵⁰³ site of NS80 and the Lys³²⁸ site of NS38 play important roles in gcTBK1_tv3-mediated degradation of NS80 and NS38, respectively.

During GCRV infection, NS80 and NS38 are the major proteins to form the VIBs (13, 15, 16). Because gcTBK1_tv3 could degrade the NS80 and NS38 proteins of GCRV, whether gcTBK1_tv3 impaired the production of VIBs during GCRV infection was further investigated. The gcTBK1 was used for the negative control because gcTBK1 had no effect on protein degradation of the NS80 or NS38 proteins. Using the anti-NS80 Ab, the results of immunofluorescence analysis showed that gcTBK1 did not affect the production of VIBs (Fig. 9A, 9B). Different from gcTBK1, the total amounts of VIBs were significantly decreased in CIK cells transfected with gcTBK1_tv3. The treatment of PS-341 restored the amounts of VIBs decreased by gcTBK1_tv3 (Fig. 9C, 9D). Together, these results demonstrate that gcTBK1_tv3 impairs the production of VIBs during GCRV infection, which is different from gcTBK1.

The innate immune system is the first line of host defense against pathogens, and IFN production induced by the RLR signaling pathway is vital for inhibiting viral infection (40). To successfully establish and maintain infection, GCRV uses a variety of strategies to escape the antiviral immune response mediated by the RLR signaling pathway. Previous studies demonstrated that GCRV proteins such as NS80, NS38, NS79, VP3, VP4, VP35, and VP56 can restrain IFN production by targeting RIG-I, MAVS, TBK1, IRF3, or IRF7 (19–23, 25). However, in the process of fighting against the virus, the host also uses different strategies to inhibit viral replication and infection. In this study, we found that the splicing isoform of grass carp TBK1, named gcTBK1_tv3, could restrict GCRV replication and infection by triggering the protein degradations of NS80 and NS38 via the ubiquitin-proteasome pathway.

Alternative splicing plays an important role in regulating the innate immune system of hosts. These molecules involved in the RLR signaling pathway, including RIG-I, MDA5, LGP2, MAVS, TBK1, and IRF3, undergo alternative splicing (30). TBK1, an IκB kinase–related serine/threonine kinase, has splicing isoforms found in both mammals and zebrafish (33–35). However, these reported splicing isoforms of TBK1 negatively regulated the host’s antiviral immune response. In mammals, the splicing isoform TBK1 disrupted the interaction of RIG-I with MAVS, thereby restricting the RIG-I–mediated production of IFN (33). In zebrafish, TBK1 isoforms TBK1_tv1 to TBK_tv3 used different strategies to suppress TBK1-mediated antiviral immune responses. TBK1_tv1 and TBK1_tv2 disrupted the formation of the TBK1-IRF3 complex and subsequently inhibited TBK1-mediated phosphorylation of IRF3 (34). TBK1_tv3 induced the degradations of TBK1 and IRF3 proteins to impede the antiviral immune response mediated by the RLR signaling pathway (35). Previous study demonstrated that grass carp TBK1 promoted GCRV replication and infection at a low infected titer, but inhibited GCRV replication and infection at a high infected titer (24). Different from previous studies, the current study reveals that gcTBK1_tv3 has the significant antiviral function both at high and low titers of GCRV infection, which suggests the positive role of TBK1 isoform in viral replication and infection.

Protein ubiquitination either positively or negatively regulates the activity and stability of target proteins (41). Ubiquitin has seven lysine residues, which can be attached to another ubiquitin and form a polyubiquitin chain. The K48-linked and K63-linked polyubiquitins are the best-characterized residues involved in polyubiquitylation (42). In most cases, K48-linked polyubiquitin chains target their substrates for proteasome-dependent degradation (43). Previous study demonstrated that the mammalian TBK1 was an E3 ubiquitin ligase, which could promote the degradation of picornavirus VP3 protein (39). Our previous study also showed that zebrafish TBK1_tv3 induced the K48-linked ubiquitination of TBK1 and promoted TBK1 proteasomal degradation (35). In the current study, gcTBK1_tv3 was found to specifically degrade the NS80 and NS38 proteins of GCRV and had no effect on other viral proteins, such as VP3 and VP5. Further data demonstrated that gcTBK1_tv3 promoted the polyubiquitination of itself and the degradation of NS80 or NS38 for the K48-linked ubiquitination by targeting Lys⁵⁰³ residue of NS80 or Lys³²⁸ residue of NS38, respectively. However, whether gcTBK1_tv3 is an E3 ubiquitin ligase needs further study.

The replication and assembly of many viruses often occur in specific intracellular compartments known as the VIBs, which are also called viral factories or viroplasms (44, 45). VIBs contain not only viral proteins, dsRNA, virus particles, but also host proteins and some membrane structures (46). VIBs promote efficient replication and assembly of virus by recruiting host and viral proteins and also hide viral nucleic acid to avoid being recognized by the host’s pattern recognition receptors. Moreover, VIBs also hijack the molecules of the host antiviral immune signaling pathway and restrict the production of IFN (47–49). These findings suggest that VIBs are vital for viral infection, replication and assembly, and immune escape. Therefore, host also evolve strategies to resist viral infection by impairing the formation of VIBs. In mammals, the IFN-induced gene C19orf66 was found to inhibit hepatitis C virus infection by restricting the formation of viral replication organelle (50). The compounds curcumin and (2E)-N-benzyl-3-(4-butoxyphenyl) prop-2-enamide (SBI-0090799) could repress human parainfluenza virus type 3 and Zika virus replication by affecting the VIBs formation, respectively (51, 52). During GCRV infection, NS80 and NS38 are the main proteins to form VIBs, which recruit all of the inner-capsid proteins (VP1–VP4 and VP6) and host factor into VIBs to promote viral replication and assembly (13, 15). Our recent report has shown that GCRV NS80 and NS38 proteins can hijack gcTBK1 into VIBs to escape host antiviral innate immunity via their interactions with gcTBK1 (25). Similar to gcTBK1, gcTBK1_tv3 is also recruited by GCRV NS80 and NS38 proteins into VIBs. However, different from gcTBK1, gcTBK1_tv3 impairs the production of VIBs of GCRV via the K48-linked ubiquitination degradations of NS80 and NS38 proteins. It is unclear the exact mechanism that dictates why gcTBK1 and gcTBK1_tv3 have distinct roles in the stability of GCRV NS80 and NS38 proteins at present. Analysis of gcTBK1 and gcTBK1_tv3 sequences revealed that obvious differences in the composition of 62 aa existed between gcTBK1 and gcTBK1_tv3, in addition to the C-terminal T2K domain of gcTBK1 being missing for gcTBK1_tv3. The C-terminal 29 aa of gcTBK1_tv3 are quite different from those of gcTBK1. We speculate the key residues required for degradation of NS80 and NS38 proteins only exist for gcTBK1_tv3 but not for gcTBK1, which needs to be further investigated.

In summary, our study reveals that gcTBK1_tv3 interacts with and degrades the NS80 and NS38 proteins of GCRV by the ubiquitin-proteasome pathway, which ultimately limits GCRV replication and infection. Mechanistically, gcTBK1_tv3 targets the K48-linked ubiquitination at the Lys⁵⁰³ residue of NS80 or Lys³²⁸ residue of NS38, which consequently leads to proteasomal degradations of NS80 and NS38. These findings provide insight into the function of the TBK1 isoform in the antiviral immune response and the exact mechanism by which TBK1 isoform targets the nonstructural proteins of GCRV for impairing the formation of VIBs.

The authors have no financial conflicts of interest.