Infectious hematopoietic necrosis virus N protein suppresses fish IFN1 production by targeting the MITA

A B S T R A C T
Interferon (IFN) is a vital antiviral factor in host in the early stages after the viral invasion. Meanwhile, viruses have to survive by taking advantage of the cellular machinery and complete their replication. As a result, viruses evolved several immune escape mechanisms to inhibit host IFN expression. However, the mechanisms used to escape the host’s IFN system are still unclear for infectious hematopoietic necrosis virus (IHNV). In this study, we report that the N protein of IHNV inhibits IFN1 production in rainbow trout by degrading the MITA. Firstly, the upregulation of IFN1 promoter activity stimulated by poly I:C was suppressed by IHNV infection. Consistent with this result, the overexpression of the N protein of IHNV blocked the IFN1 transcription that was activated by poly I:C and MITA. Secondly, MITA was remarkably decreased by the overexpression of N protein at the protein level. Further analysis demonstrated that the N protein targeted MITA and promoted the ubiquitination of MITA. Taken together, these data suggested that the production of rainbow trout IFN1 could be suppressed by the N protein of IHNV via degrading MITA.

1.Introduction
Interferons (IFNs) are the first line of the host defense against viral invasion. IFNs initiate the transcription of more than 300 IFN-stimu- lated genes (ISGs) and set up an antiviral immunity in the host. When the host cells are infected by a virus, the cellular innate immune sensors induce the synthesis of IFNs. Retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) including RIG-I, melanoma differentiation-associated gene 5 (MDA5) and laboratory of genetics and physiology 2 (LGP2) are critical in sensing RNA virus invasion. Subsequently, mitochondrial adaptor protein (MAVS), which is recruited and activated by RIG-I and MDA5, transduces a signal to mediator of IRF3 activation (MITA) and TANK-binding kinase 1 (TBK1). Finally, MITA recruits TBK1 to phos- phorylate IFN regulatory factor 3/7 (IRF3/7) and initiates the expression of IFNs [1–3]. This evolutionary conserved RLR-based in- duction pathway exists in both mammals and fish [4]. It has been perceived that over-production of the genes involved in the RLR-based
inductive pathway showed activation of immunity. The retrograde regulation protein 2 (RTG2) cells transfected with MDA5 showed ele- vated resistance towards VHSV infection, similar to the resistance that has been observed in Atlantic salmon transfected with MAVS against infectious hematopoietic necrosis virus (IHNV), Viral hemorrhagic septicemia virus (VHSV) and spring viremia of carp virus (SVCV) [5–7].

MITA is the key component of the RLR signaling pathway. To date, MITA has been identified in many species of fish and its expression profile is similar to mammalian counterparts. Such as MITA homolog from grouper also contains transmembrane motifs in the N terminal and anchors in the endoplasmic reticulum [8]. In grass carp, MITA is necessary for IFN activation through ZDHHC1 that supplements evi- dence to previous mammal studies [9]. Although the roles of MITA in innate immune responses is conservative from fish to mammals, there are also species unique features in fish. IRF10 is present in non- mammals including birds and fish. The IRF association domain of zebrafish IRF10 inhibited the activation of IFN1 and IFN3 promoters by interacting with MITA [10]. Virus-induced IFN genes in fish are not the true orthologs of mammalian IFN-α/β, they are regulated through the MITA pathway but dependent on distinct transcription factors [11]. MITA from goldfish interacted with TBK1 and IRF3 to form a complex, but the presence of MAVS was not found. It is not clear whether MITA and MAVS can independently transduce the RIG-I signal or need to interact with each other to be activated and recruit TBK1 and IRF3 in fish [12].IHNV is a linear non-segmented and negative-sense ssRNA virus which, belongs to the genus Novirhabdovirus of the family Rhabdoviridae, and has caused high mortality in salmonid fish. The genome of IHNV, which is approXimately 11 kb, encodes nucleoprotein (N), phosphoprotein (P), matriX protein (M), glycoprotein (G), non- virion (NV) protein, and RNA-dependent RNA polymerase (large pro- tein, L) [13,14]. The availability of sequence information on the entire IHNV genome and homology analysis has resulted in the accurate functional deduction of the IHNV proteins. In non-segmented negative- strand RNA viruses, N, P and L make up the minimal set of proteins required for the transcription and replication of Rhabdoviridae [15]. Apart from their functions as structural components, M and P proteins have also displayed the ability to facilitate host-cell shutoff or inhibit IFNs. The NV protein which only presenting in some fish rhabdoviruses, appears to be related to the viral growth [16,17].Generally, the function of N protein is to encapsidate genomic RNA to form the RNase A-resistant helical ribonucleoprotein complex. Nevertheless, some studies have also revealed its additional role as a suppressor of IFN1 transcription by targeting the RLR signaling adaptor protein MAVS during the SVCV infection [18]. To elucidate the me- chanisms of IHNV infection, we report that the N protein of IHNV suppresses the fish IFN1 production by degrading MITA in an ubiquitin- proteasome manner. This finding would better illuminate the essential roles of N protein in the immune evasion strategies of IHNV.

2.Material and methods
2.1.Cell culture and virus propagation
EPC cells were bought from China center for type culture collection (CCTCC, Wuhan, China) and maintained in medium 199 (Invitrogen) with 10% fetal bovine serum (FBS, Invitrogen) under 28 °C and 5% CO2. Human embryonic kidney (HEK 293T) cells were cultured at 37 °C, 5% CO2 in DMEM medium (Invitrogen) supplemented with 10% FBS. RTG2 cells were grown in L-15 medium (Gibico) with 10% FBS at 18 °C. IHNV was propagated in EPC at 18 °C, then the cultured media with cells were stored at −80 °C and prepared for further use.

2.2.Plasmid construction and reagents
Using the cDNA of RTG2 cells as template, MITA open reading frames (ORFs) was cloned and inserted into the NheI and BamHI double-digested sites of pcDNA3.1 (+) vector (Invitrogen) to form pcDNA3.1 (+)-MITA. The pCMV-HA/Myc-MITA plasmid was con- structed by inserting the DNA fragment encoding MITA into the XhoⅠ and NotⅠ double-digested sites of pCMV-HA vector (BD Clontech) or pCMV-Myc vector (BD Clontech), respectively. The ORF of N from IHNV was cloned and inserted into the pcDNA3.1 (+), pCMV-HA, and pCMV-Myc to form plasmid expressing N protein with different tags. The ORF of P from IHNV was cloned and inserted into the pCMV-HA to form plasmid expressing P protein. For promoter activity analysis, gene promoter of rainbow trout IFN1 was cloned and inserted into the KpnⅠ and XhoⅠ sites of pGL3-basic luciferase reporter vector (Promega). The primers including the restriction enzyme cutting sites used for plasmids construction are listed in Table 1. MG132, a potent and cell-permeable proteasome inhibitor, and polyinosinic:polycytidylic acid (poly I:C), an immunostimulant, were purchased from Sigma-Aldrich and used at a final concentration of 20 μM and 1 μg/ml respectively.

2.3.Luciferase activity assay
Luciferase activity analysis was performed using Lu’s method [18]. 250 ng luciferase reporter plasmid (IFN1pro-Luc) and 25 ng Renilla luciferase control vector (pRL-TK) (Promega) were used to cotransfect RTG2 cells or EPC cells 24 h after subculture in 24-well plates. Mean- while, empty vector pcDNA3.1 (+) was used to equate the DNA amount in each well. Poly I:C with a concentration of 1 mg/ml were used to activate IFN1 expression 24 h before cell harvest. Luciferase activity analysis was performed using the Dual-Luciferase Reporter Assay System (Promega) 48 h after transfection.

2.4.Quantitative real-time PCR (qPCR)
Total RNA was extracted with the Trizol reagent (Invitrogen, USA), cDNAs used for real-time PCR were reverse transcribed with the RNA using GoScript reverse transcription system (Promega). Primers used in this study are shown in Table 1. Reactions were carried out on CFX96 Touch™ Deep Well Real-Time PCR Detection System (Bio-Rad) using Fast SYBR Green PCR Master MiX (Bio-Rad). The thermocycler was programmed as follows: 95 °C denaturation for 5 min, 40 cycles of 95 °C denaturation for 20 s, 60 °C annealing for 20 s, and 72 °C for 20 s, fi- nally end with 5 min at was 72 °C for the extension. Each reaction system included Fast SYBR Green PCR Master miX 5 μl, cDNA (0.5 μg/ μl) 2 μl, reverse primer (10 μM) 0.5 μl, forward primer (10 μM) 0.5 μl and ddH2O 2 μl. The relative fold changes were calculated by com- parison with the corresponding controls the 2−ΔΔCt method. Three in- dependent experiments were conducted for statistical analysis purposes.

2.5.Co-immunoprecipitation (Co-IP) assay
For transient-transfection and Co-IP experiments, HEK 293T cells and EPC cells were used instead of RTG2 due to higher transfection efficiency of HEK 293T cells and EPC cells. Cells were seeded into 10 cm2 dishes overnight and transfected with indicated plasmids. 24 h after transfection, the HEK 293T cells were washed with 10 ml cold PBS buffer two times and then lysed in 1 ml RIPA buffer (Beyotime, Shanghai, China) adding with protease inhibitor cocktail (Sigma-Aldrich) for 1 h at 4 °C following Lu’s method [19]. After centrifugation at 12000g for 15 min at 4 °C, the supernatant was incubated with 30 μl anti-HA-agarose or anti-Myc affinity gel (Sigma-Aldrich) at 4 °C for 12 h with constant agitation. After incubation, immunoprecipitated proteins were harvested by centrifugation at 5000g at 4 °C for 1 min. Then, the collected proteins were rinsed with lysis buffer and suspended in 50 μl SDS sample buffer. Finally, the immunoprecipitates were analyzed using western blotting method.

2.6.Immunoblotting assay
The expressions of HA-N, Myc-MITA and actin, were measured with Western blotting method, their quantification was accomplished en- suing the subsequent processes with relative antibodies as following: (a) immunoprecipitated proteins were separated using SDS-PAGE method and then transferred to PVDF membrane; (b) the membranes were blocked in TBST buffer with 5% skim milk for 1 h under room tem-
perature; (c) the blocked membranes were incubated with diluted pri- mary antibody (1:1000 for β-actin; 1:3000 for HA; 1:2000 for Myc) for 1 h; (d) conjugated with horseradish peroXidase labeled second anti- body anti-mouse IgG at a dilution rate of 1:5000 (Thermo scientific) for 1 h; (e) Finally, protein was presented with ECL detection reagent.

2.7.In vivo ubiquitination assay
HEK 293T cells and EPC cells were transiently transfected with 5 μg HA-N, 4 μg Myc-MITA, 1 μg HA-Ub expression plasmids. After 18 h transfection, the cells were treated with 20 μM MG132. HEK 293T cells were harvested at 24 h posttransfection and EPC cells were harvested at 48 h posttransfection. Samples were lysed with RIPA buffer (Beyotime,Shanghai, China) adding with protease inhibitor cocktail (Sigma- Aldrich) and then denatured by heating 10 min adding with 0.1% SDS. After centrifugation at 12000g for 15 min at 4 °C, the supernatants were then immunoprecipitated overnight at 4 °C with constant agitation with 30 μl anti-Myc agarose conjugate (Sigma-Aldrich). The im-munoprecipitated protein was washed three times with lysis buffer and suspended in 50 μl SDS sample buffer. Finally, the immunoprecipitated protein was analyzed using western blotting method.

2.8.Statistics
Data were expressed as means ± SE (Standard Error of Mean) and analyzed with SPSS program (IBM Inc, USA). Comparisons were ac- complished with Duncan’s multiple range tests, p < 0.05 was con- sidered as statistically significant. In this study, each experiment was repeated three times.

3.Results
3.1.N protein suppresses IFN1 expression induced by poly I:C during IHNV infection
Poly I:C is a mimic of viral RNAs and can be used to induce IFN expression. Whether the expression of fish IFN could be inhibited by IHNV was explored by the luciferase reporter gene assay. As shown in Fig. 1A, Poly I:C induced a nearly 12-fold increase in luciferase activity, and the activation declined with increasing doses of IHNV inoculation. To identify the mechanism by which IHNV suppresses the IFN1 acti- vation, and based on the results of preliminary experiments, the N protein of IHNV was chosen for further study. As shown in Fig. 1B and C, N protein suppressed the activation of IFN1 induced by IHNV, poly I:C and MITA, and it was consistent with the tendency of whole IHNV. However, NV protein had the opposite effect, increasing the activation of IFN1 induced by poly I:C and MITA (Fig. 1C). These results show that the N protein plays an essential role in suppressing IFN1 production.

3.2.MITA is degraded by overexpression of N protein
MITA is the crucial component for IFN induction and it was speculated that N protein suppressed IFN1 expression by targeting MITA. To investigate the function of the N protein on MITA, the modulation pattern of MITA during the expression of the N protein was also analyzed. Immunoblot analysis showed that the expression of MITA decreased when the N protein was over-expressed; furthermore, this inhibitory effect was shown to be correlated with the level of N protein expression. The data suggested that increased N protein ex- pression suppressed MITA expression in a dose-dependent manner (Fig. 2A and B). The P protein of IHNV, used as a control, displayed a weak tendency to increase MITA expression (Fig. 2C). These results suggest that the N protein of IHNV can degrade MITA at the protein level.

3.3.N protein degrades MITA by the proteasome-dependent manner
Based on the above results, and to identify the mechanism by which the N protein of IHNV induced the degradation of MITA, EPC cells were transfected with HA-MITA and Myc-N and were cultured in the pre- sence of MG132. DMSO was used as a control, as MG132 was dissolved in DMSO before using in these experiments. As shown in Fig. 3A, the band of MITA in EPC cells coexpressing N protein was weaker than that in the control cells. However, the degradation of MITA was significantly rescued by MG132. Furthermore, this blocking effect of MG132 in- creased in a dose-dependent manner (Fig. 3B). These results confirmed

Fig. 1. N protein of IHNV blocks the ex- pression of IFN1. (A) IHNV inhibited poly I:C-mediated activation of IFN1 promoter in a dose-dependent manner. RTG2 cells seeded in 24-well plates overnight were transfected with 250 ng luciferase reporter plasmid (IFN1pro-Luc) and 25 ng Renilla luciferase control vector (pRL-TK) (Promega). At 24 h posttransfection, cells were transfected with poly I:C for 4 h and then infected IHNV (MOI = 1, 10, 100, 1000). The luciferase assays were measured 48 h after stimulation, and fold activation was determined by comparing the luciferase activity in control cells. (B) N protein in- hibited IHNV induced IFN1 activation. RTG2 cells seeded in 24-well plates over- night were transfected with 250 ng IFN1pro-Luc, 25 ng pRL-TK and 250 ng of empty or pcDNA3.1-N. At 24 h post- transfection, cells were infected with IHNV (MOI of 10). The luciferase activities were monitored at 24 h after stimulation. (C) N protein inhibited poly I:C induced IFN1 ac- tivation. EPC cells were seeded in 24-well plates and transfected the next day 250 ng pcDNA3.1 or pcDNA3.1-N or NV, at 24 h posttransfection, cells were treated with poly I:C. The luciferase activities were monitored at 24 h after stimulation. (D) N protein inhibited MITA induced IFN1 acti- vation. EPC cells were seeded in 24-well plates and transfected the next day 250 ng pcDNC3.1 or pcDNA3.1-N or NV, plus 250 ng pcDNA3.1-MITA. The luciferase ac- tivities were monitored at 48 h after sti- mulation. Error bars are the SDs obtained by measuring each sample in triplicate. Asterisks indicate significant differences from control (*p < 0.05) that the N protein of IHNV suppressed MITA expression via the pro- teasome pathway. Additionally, in terms of the relationship between the N protein and MITA, as shown in Fig. 3C, MITA could be co-im- munoprecipitated using the anti-HA-N antibody, suggesting that MITA could interact with the N protein.

3.4.N protein promotes ubiquitination of MITA
To further clarify how MITA could be degraded by the N protein of IHNV in a proteasome-dependent manner, HA-N, Myc-MITA and HA-Ub were cotransfected in the presence or absence of MG132. As shown in Fig. 4A and B, immunoblot analysis revealed that the cells coexpressing N protein and containing MG132 had more Ub than the others, and this was consistent in both HEK 293T and EPC cells. These results indicate that the N protein promotes the ubiquitination of MITA.

3.5.N protein causes RTG2 cells to become more vulnerable to IHNV infection
Since the N protein was identified as a negative regulator of IFN transcription, the function of N protein on regulating both the expres- sion of ISGs and the replication of IHNV were analyzed. RTG2 cells were transfected with pcDNA3.1-N or the empty vector, and the total RNAs were extracted and subjected to qPCR. Overexpression of the N protein inhibited several ISGs, such as mx2, mx3 and vig3 (Fig. 5A–C). Also, as shown in Fig. 5D–F, upon stimulation with IHNV, the transcript levels of the viral g, n, and l genes were elevated when the N protein was over-expressed. These data indicate that the N protein can suppress the cellular IFN response and enhance the replication of IHNV.

4.Discussion
IFN response is the first line of host defense against viral infection. The type and activation mechanism of mammalian IFN has been well characterized [20]. In 2003, the fish type I IFN gene was first identified

Fig. 2. N protein degrades MITA in a dose-dependent. (A and B) N protein overexpression blocked MITA expression in dose-dependent. EPC cells were seeded in siX-well plates and transfected with 2.5 μg Myc-MITA together with 2.5 μg empty vector, HA-N (A), HA-N (1.5, 2.5 or 3.5 μg) (B) for 24 h. Whole-cell lysates were subjected to immunoblotting with anti-Myc, anti-HA and anti-β-actin Abs. (C) P protein overexpression upregulated MITA expression. EPC cells were seeded in siX- well plates and transfected with 2.5 μg Myc-MITA together with 2.5 μg empty vector, HA-P, for 24 h. Whole-cell lysates were subjected to immunoblotting with anti-Myc, anti-HA and anti-β-actin Abs in zebrafish (Danio rerio) [21], Atlantic salmon (Salmo salar) and puf- ferfish (Tetraodon nigroviridis) [22,23]. Both the type Ⅰ and Ⅱ IFN system of fish have been extensively studied and are now known to play an essential role in antiviral defense. IFNγ is the only type Ⅱ IFN and a key cytokine in defining T helper cell 1 (Th1) [24]. Meanwhile, type I IFN
can be divided into two groups, group I has two cysteines and group II has four cysteines to form one or two pairs of putative intramolecular disulphide bonds, respectively [25]. Group I IFNs have been discovered in all the investigated teleost species, whereas group II IFNs are limited to the salmonid, siluriform and cyprinid species [26]. The two groups of IFNs can be further divided into four subgroups, termed IFN-a, -b, -c, and -d. The rainbow trout possesses not only all four subgroups but also another two subgroups (IFN-e and -f), whereas the other teleost species have one or more, but not all groups. IFN-a1, also known as IFN1, in the rainbow trout is constitutively expressed and has a marked induction by poly I:C stimulation after a 2 h stimulation [27]. The reason why so many type I IFNs have been retained in salmonids is still controversial [28].IHNV is an economically important pathogen causing infectious hematopoietic necrosis (IHN) in a wide variety of salmonid species, such as the Atlantic salmon and rainbow trout (Oncorhynchus mykiss) [29]. Viral components can be recognized by pathogen-associated molecular patterns (PAMPs) and subsequently initiate innate antiviral responses. The host's body is an ideal breeding ground for viruses being obligatory intracellular parasitic entities. To seize the cellular ma- chinery and complete its replication, viruses have evolved several im- mune evasion strategies to counteract the IFN-mediated host antiviral defense [30]. Previous studies demonstrated that the M protein could down-regulate host transcription and induce programmed cell death [31]. Among the other proteins of IHNV, the NV protein had been re- ported to associate with cell rounding and was necessary for optimal viral growth [32]. The soluble nonglycosylated recombinant G protein

Fig. 3. N protein degrades MITA by proteasome-dependent manner. (A) N protein-induced MITA degradation was rescued in the presence of MG132. EPC cells were transfected with 2.5 μg HA-MITA and 2.5 μg Myc-N or empty vector. At 24 h posttransfection, the cells were treated with DMSO or MG132 for 6 h. Then, the cells were harvested for immunoblotting with anti-Myc, anti-HA and anti-β-actin. (B) N protein-induced MITA degradation was rescued by MG132 in a dose- dependent manner. EPC cells were transfected with 2.5 μg Myc-MITA and 2.5 μg HA-N or empty vector. At 24 h posttransfection, the cells were treated with DMSO or MG132 (10, 20, or 40 μM) for 6 h. Then, the cells were harvested for immunoblotting with anti-Myc, anti-HA and anti-β-actin. (C) N protein was associated with MITA. HEK 293T cells seeded into 10 cm2 dishes were transfected with 5 μg HA-MITA and 5 μg Myc-N or empty vector. After 24 h, cell lysates were im- munoprecipitated (IP) with anti-Myc affinity gel. The immunoprecipitates and cell lysates were then analyzed by immunoblotting with anti-HA, anti-Myc and anti-β- actin Abs respectively.

Fig. 4. N protein mediates ubiquitination of MITA. (A and B) N protein promoted the ubiquitination of MITA. HEK 293T cells or EPC cells were transfected with 5 μg HA-N/empty vector, 5 μg MITA-Myc, 1 μg Ub-HA. At 18 h posttransfection, the cells were treated with DMSO or MG132 for 6 h. Cell lysates were im- munoprecipitated with anti-Myc affinity gel and immunoblotted with anti-HA and anti-Myc Abs. The whole-cell lysates were immunoblotted with the Abs indicatedcould induce expression of IFNs and protected the rainbow trout from subsequent IHNV infection [33]. In the current study, the N protein, but not the P protein, inhibited IFN1 expression by degrading MITA, pro- viding another strategy for IHNV to evade the host innate immune defense system. Because of the significant fish mortality, the patho- genesis of different proteins of IHNV should be further studied.One virus can evolve several strategies to evade host defense me- chanisms, and one host immune molecule can be the target of several viruses. MITA is a signaling molecule central to the innate immune responses to cytosolic nucleic acids, and it recruits the TBK1 to MAVS- associated complex and is phosphorylated by TBK1 to mediate the ac- tivation of IRF3 [34]. The amino-terminal domain of MITA, which is composed of four transmembrane domains, is responsible for its loca- tion on the outer membrane of mitochondria, whereas the carboXy- terminal domain of MITA resides in the cytosol and its structure has been previously resolved [35]. Viruses invade the host's cell and release nucleic acids into the cytosol to stimulate the IFN system. Therefore, MITA is always treated as a repressive target to evade the host immune response. MITA is also avital protein targeted by fish viruses. Previous studies demonstrated that the phosphorylation of MITA was sig- nificantly decreased by the protein VP41 of the GCRV virus, and the miR-210 expression induced by SCRV infection could negatively reg- ulate virus-triggered type I IFN by targeting MITA [19,36]. In this study, N could degrade MITA to disrupt RLR-mediated immune sig- naling.

The ubiquitin-proteasome system and autophagy are two mechanisms of protein degradation [37]. Several viral proteins target sig- nificant molecular structures of PRR-signaling for degradation. For example, the Npro protein of the classical swine fever virus (CSFV) induced the proteasomal degradation of IRF3. Vpr and Vif, two encoded accessory proteins of HIV-1 virus that were encapsulated in large amounts in virions, were found to be associated with the ubiquitin-proteasomal degradation [38]. Similarly, the host has also evolved several mechanisms to degrade excessive signaling proteins or un- necessary activation. It has been reported that RNF5 (E3 ubiquitin li- gase ring finger protein 5) ubiquitinated and degraded MITA to inhibit virus-induced IRF3 excessive activation [39]. In the current study, N protein of IHNV could degrade MITA by the ubiquitin-proteasome pathway. Previous studies reported that the N protein of SVCV de- graded MAVS, which underwent K48-linked ubiquitination. Hence, the mechanism of N protein-triggered ubiquitination should be further studied.

In conclusion, the present study revealed a potential mechanism by which the N protein of IHNV negatively regulates MITA via the ubi- quitin-proteasome degradation pathway. These results provide more information on viral evasion to understand further that different viruses have the same evasion modes, while homologous proteins from dif- ferent viruses have different Polyinosinic acid-polycytidylic acid targets in the innate immune system. Further studies should focus on the interaction between hosts and aquatic viruses.