E3 Ubiquitin Ligase RNF114 Inhibits Innate Immune Response to Red-Spotted Grouper Nervous Necrosis Virus Infection in Sea Perch by Targeting MAVS and TRAF3 to Mediate Their Degradation

The innate immune system can recognize invading viruses and initiate host antiviral responses by pattern recognition receptors, including TLRs, retinoic acid–inducible gene I (RIG-I)–like receptors (RLRs), and NOD-like receptors (1, 2). During RNA virus infection, RLRs, as pivotal intracellular sensors for viral RNAs, recognize dsRNA and bind to its downstream mediator, mitochondrial antiviral signaling protein (MAVS), to transduce downstream antiviral innate immune signaling (3). Upon interacting with RLRs, MAVS undergoes polymerization and a series of posttranscriptional modification, such as ubiquitination and phosphorylation, and mediates the activation of NF-κB and recruits TNFR-associated factors (TRAF2/3/5/6) to form the MAVS-TRAF complexes, thereby promoting the phosphorylation and dimerization of the IFN regulatory factor (IRF) 3/7 (4, 5). Phosphorylated IRF3 and IRF7 translocate into the nucleus to induce the expression of type I IFNs, which activate the Janus tyrosine kinase and STAT signaling pathways to stimulate the production of IFN-stimulated genes (ISGs) (6).

Similar with other signaling pathways, it has become evident that RLR signaling pathway is tightly regulated to achieve the immune homeostasis between eliminate virus infections and avoid harmful immunopathology by a variety of posttranslational modifications, including ubiquitination and phosphorylation (7, 8). Among them, ubiquitination, including K27-, K63-, and K48-linked polyubiquitination, is important for the modulation of RLR signaling pathways. For instance, tripartite interaction motif (TRIM) 25, ring finger protein 135 (RNF135), RNF166, and TRIM31 mediated K63-linked ubiquitination of RIG-I, MAVS, TRAF3, and TRAF6 to facilitate RLR signaling pathways activation (9–13). In contrast to K63-linked polyubiquitination, several E3 ubiquitin ligases, such as RNF5, RNF125, RNF122, and TRIM40, mediated K48-linked ubiquitination and degradation of RIG-I, melanoma differentiation–associated gene 5 (MDA5), and MAVS, thereby exerting a negative regulation on RLRs signaling (14–17).

RNF114, belonging to the family of RING domain E3 ubiquitin ligases, was first reported as a psoriasis-susceptibility gene in humans (18). Mounting evidences have shown that RNF114 is widely participated in the regulation of dsRNA-induced type I IFN production, NF-κB activity, and T cell activation (19, 20). Lin and colleagues (21) suggested that RNF114 negatively regulated RLR signaling pathways through proteasome-mediated degradation of MAVS. However, contradictory results were reported that RNF114 was a positive regulator of RIG-I/MDA signaling (22). Therefore, the exact mechanisms of RNF114 regulating RLR signaling pathways remain unclear and it deserves further investigation.

Nervous necrosis virus (NNV), belonging to genus Betanodavirus, family Nodaviridae, is one of the most devastating pathogens infecting more than 120 species of cultured fresh and marine fishes worldwide (23). NNV infection causes more than 90% mortality in the larval stage of fish, leading to great economic losses of aquaculture industry in many countries. Previous studies demonstrated that RLRs could recognize viral RNA and activate antiviral response during NNV infection. For instance, in zebrafish, RIG-I played a crucial role in type I IFN induction and the inflammatory response to NNV infection (24). In sea perch, the key components of RLR signaling pathways, MDA5, MAVS, TRAF3, and IRF3, had drastic inhibitory effects on NNV (25–27). Sea perch (Lateolabrax japonicus) is a valuable commercial fish that is suffering from various pathogen infections due to the intensive farming, including NNV, Vibrio anguillarum, and Vibro harveyi (28–30). In previous studies, we reported the isolation of NNV from diseased sea perch and demonstrated that NNV infection could activate sea perch RLR signaling pathways (26, 30, 31), however, the exact mechanisms underlying sea perch RLR signaling pathways regulation during NNV infection remains poorly understood. To elucidate the role and molecular mechanism of sea perch RNF114 (LjRNF114) in regulating RLR signaling pathways, in this study, the relationship between LjRNF114 and RLR-regulated IFN antiviral responses was investigated. We reported that LjRNF114 functioned as a negative regulator of RLR signaling pathways by interacting with and promoting the K27- and K48-linked ubiquitination and degradation of MAVS and TRAF3. Overall, to our knowledge, our findings revealed a novel mechanism by which RNF114 negatively regulated RLR signaling pathways.

Sea perch (Lateolabrax japonicus) brain (LJB) cells were maintained in DMEM (Invitrogen) supplemented with 15% FBS (Life Technologies) at 28°C (32). Human embryonic kidney 293T (HEK 293T) cells were cultured in DMEM with 10% FBS at 37°C in 5% CO₂ incubator. Fathead minnow (FHM) cells were cultured in M199 medium (Life Technologies) with 10% FBS at 28°C. Red-spotted grouper NNV (RGNNV) was isolated from diseased sea perch and propagated in LJB cells (30). The titers of RGNNV were determined by 50% tissue culture infective dose (TCID₅₀).

Anti-Flag (M20008), anti-Myc (M20002), anti-GFP (M20004), anti-HA (M20003), anti-Actin (P30002), and anti-His Abs (M20001) were purchased from Abmart (Guangzhou, China); Anti-GFP Abs (G1544) were purchased from Sigma-Aldrich (St. Louis, MO). Anti-LjRNF114 Abs were purchased from Kebiochem (Shanghai, China). The specificity and sensitivity of anti-LjRNF114 Abs were assessed by Western blotting (Supplemental Fig. 1). The secondary Abs HRP-labeled goat anti-rabbit IgG (H + L) Abs (A0208) were purchased from Beyotime (Guangzhou, China), and anti-mouse IgG HRP-linked Abs (7076S) were from Cell Signaling Technology (Danvers, MA). Alexa Fluor 488–labeled donkey anti-rabbit IgG and Alexa Fluor 555–labeled goat anti-mouse IgG secondary Abs were purchased from Sigma-Aldrich.

DAPI stain solution (40728ES03) was purchased from Yeasen Biotechnology (Shanghai, China). PMSF (PMSF, ST506), cell lysis buffer (P0013), and BeyoECL Plus (P0018) were purchased from Beyotime. Polyinosinic-polycytidylic acid [poly (I:C)], DMSO, and Triton X-100 were obtained from Sigma-Aldrich. MG132 (S2619) was purchased from Selleckchem and dissolved in DMSO (25 mg/ml).

LJB cells were seeded into six-well plates (1 × 10⁶ cells per well). After 24 h, cells were infected with RGNNV (multiplicity of infection [MOI] = 2) or treated with poly (I:C) (5 μg/ml) at 28°C. Cells were harvested for RNA isolation at 0, 12, 24, 36, and 48 h posttreatment, respectively. The expression of LjRNF114 was detected by quantitative real-time PCR (qRT-PCR).

To analyze the role of LjRNF114 during RGNNV infection in vitro, LJB cells in six-well plates at 70–80% confluence were transfected with pCMV-Flag or pCMV-Flag-LjRNF114 plasmids using Lipofectamine 8000 (Beyotime) according to the manufacturer’s instruction, respectively. At 24 h posttransfection, LJB cells were infected with RGNNV (MOI = 2). Then cells were harvested for RNA isolation at 0, 12, and 24 h postinfection (hpi), respectively. The expressions of capsid protein (CP) and ISGs were detected by qRT-PCR.

The small interfering RNA (siRNA) targeting LjRNF114 and control siRNA (NC) were obtained from RiboBio (Guangzhou, China). The sequences were listed as following: si-LjRNF114-1, 5′-AGTGTTTGCGTCCACAGAA-3′; si-LjRNF114-2, 5′-TCAGATGAGAAGCCACACA-3′; and si-LjRNF114-3, 5′-CACCTGAAGATCAGACACA-3′. LJB cells were seeded in six-well plates at 70–80% confluence and transfected with LjRNF114 siRNA mixtures (100 nM) or NC (100 nM) by using Lipofectamine 8000 (Beyotime) according to the manufacturer’s protocol. After transfection for 24 h, LJB cells were infected with RGNNV (MOI = 2) and harvested at 0, 12, and 24 hpi, respectively. The expressions of CP and ISGs were detected by qRT-PCR. The expression of LjRNF114 protein was detected by immunoblotting with anti-LjRNF114 Abs at 24 hpi.

Total RNA was isolated with TRIzol reagent (Invitrogen) and reverse transcribed into cDNA by Primescript First Strand cDNA Synthesis Kit (Takara Bio) according to the manufacturer’s instruction. qRT-PCR was performed as described previously (27). The expression levels of target genes were normalized with β-actin of sea perch by the 2^−ΔΔCT methods. Data from each sample were shown as mean ± SD from three independent experiments in triplicates. Primers for qRT-PCR are listed in Supplemental Table I.

Various expression plasmids for LjRNF114, MAVS, TRAF3, TANK-binding kinase 1 (TBK1) and IRF3 were constructed by standard molecular biology techniques. Series truncated mutants of LjRNF114 with Myc-tag and GFP-tag, including pCMV-Myc-LjRNF114-ΔRING, pCMV-Myc-LjRNF114-ΔC2HC, pCMV-Myc-LjRNF114-ΔC2H2a, pCMV-Myc-LjRNF114-ΔC2H2b, pCMV-Myc-LjRNF114-ΔUIM, pEGFP-LjRNF114-ΔRING, pEGFP-LjRNF114-ΔC2HC, pEGFP-LjRNF114-ΔC2H2a, pEGFP-LjRNF114-ΔC2H2b, pEGFP-LjRNF114-ΔUIM, pEGFP-LjRNF114-ΔRU, pEGFP-LjRNF114-1-139, and pEGFP-LjRNF114-140-227 were generated using the pCMV-Myc-LjRNF114 plasmid as a template. All constructs were confirmed by DNA sequencing. Primers are listed in Supplemental Table I.

HA-Ub, HA-K27, HA-K48, and HA-K63 plasmids were purchased from YouBia (Guangzhou, China).

pGL3-LjIFNh-pro-Luc plasmid was prepared in our laboratory (K. Lu, J.W. Zeng, K.T. Jia, and M.S. Yi, unpublished observations).

FHM cells were transiently transfected with pGL3-LjIFNh-pro-Luc, pRL-TK (Promega), and indicated plasmids or empty vector for 24 h. Luciferase activity assay was carried out as described previously in the presence or absence of RGNNV (33). The activity of firefly luciferase was normalized by that of Renilla luciferase. Data were expressed as mean ± SD from three independent experiments performed in triplicates.

Coimmunoprecipitation (Co-IP) experiments were performed as previously described with some modification (34). HEK 293T cells seeded in 25-cm² cell culture flasks were cotransfected with indicated plasmids using Lipofectamine 8000. At 48 h posttransfection, cells were collected and lysed in 500 μl of radio immunoprecipitation lysis buffer containing PMSF (Beyotime) for 30 min. Then, the supernatants were collected by centrifugation for 15 min at 12,000 × g at 4°C and incubated with specific Abs and protein G agarose overnight at 4°C with constant agitation. The immunoprecipitates were washed five times with radio immunoprecipitation buffer and subjected to immunoblot analysis.

For immunoblot analysis, immunoprecipitates or whole-cell lysates were subjected to SDS-PAGE, transferred onto PVDF membranes (Millipore, Sigma), blocked, and then incubated with the appropriate Ab as described previously (35).

HEK 293T cells in 12-well plates at 70–80% confluence were cotransfected with indicated plasmids using Lipofectamine 8000 according to the manufacturer’s instruction. At 24 h posttransfection, cells were fixed with 4% paraformaldehyde for 10 min, permeabilized with 0.1% Triton X-100, and blocked with 1% BSA. Then cells were incubated with indicated primary Abs for 1 h, washed, and incubated with secondary Abs for 1 h at room temperature. After nuclei were stained with DAPI, cells were observed with a confocal microscope (Zeiss).

His pull-down assays were performed as described previously (13). pET32a-His or pET32a-His-LjRNF114 plasmids were transformed into Escherichia coli BL21(DE3) and cultured in 50 ml Luria-Bertani medium with 0.5 mM isopropyl-1-thio-β-d-galactopyranoside (IPTG) at 18°C overnight with agitation at 120 rpm. Cells were collected by centrifugation at 1000 × g for 15 min and lysed in lysis buffer (100 mM sodium-phosphate [pH 8], 600 mM NaCl, 0.02% Tween-20) through sonication. Following centrifugation at 13,000 × g at 4°C for 20 min, the supernatant containing pET32a-His or pET32a-His-LjRNF114 fusion protein was affinity-purified with Dynabead His-Tag magnetic beads (Invitrogen). Then purified protein was mixed with lysates from cells transfected with pCMV-Flag-MAVS, pCMV-Flag-TRAF3, pCMV-Flag-IRF3, or pCMV-Flag-MAVS and pCMV-Flag-TRAF3 at 4°C overnight. The beads were washed and analyzed by immunoblotting using indicated Abs.

The ubiquitination assay was performed as described previously (13). HEK 293T cells were cotransfected with the indicated plasmids for 48 h. Cells were lysed and immunoprecipitated with the indicated Abs as described above. The expression of related proteins was examined by immunoblotting with indicated Abs.

HEK 293T cells in 25-cm² culture flasks were cotransfected with pCMV-Myc-LjRNF114 (2 μg) and pCMV-Flag-MAVS (3 μg) or pCMV-Flag-TRAF3 (3 μg). At 18 h posttransfection, cells were treated with proteasomal inhibitor MG132 (20 or 50 μM) or DMSO for 6 h, respectively. The levels of ubiquitinated protein were tested by Western blotting with β-actin as the loading control.

All statistics were calculated using SPSS, version 19. Differences between control and treatment groups were assessed by one-way ANOVA. The p values < 0.05 (*) and < 0.01 (**) were considered to be statistically significant and very significant, respectively.

To investigate the relationship between LjRNF114 and RGNNV infection, we detected the expression level of LjRNF114 during RGNNV infection. As shown in Fig. 1A and 1C, the mRNA and protein levels of LjRNF114 were significantly upregulated after RGNNV infection in LJB cells. Additionally, the expression levels of LjRNF114 mRNA and protein were also induced in LJB cells after stimulation with poly (I:C) (Fig. 1B, 1D). All these results indicate a potential role of LjRNF114 in response to viral infection.

To further confirm the role of LjRNF114 during RGNNV infection, we examined the effect of LjRNF114 on RGNNV replication. The expressions of IFN h and its downstream genes (MX, ISG15, and Viperin) were significantly increased at 12 and 24 hpi compared with that of control group (0 h) (Fig. 2). The transcription level of CP was significantly upregulated in LjRNF114-overexpressing LJB cells compared with that in LJB cells transfected with pCMV-Flag vector at 12 and 24 hpi (Fig. 2A, 2B). Accordingly, the expressions of IFN h, MX, ISG15, and Viperin were also significantly decreased at 12 and 24 hpi (Fig. 2C–F). In addition, overexpression of LjRNF114 observably inhibited the promoter activity of sea perch IFN h in comparison with the control in the presence or absence of RGNNV (Fig. 2G).

Specific LjRNF114 siRNA was used to knock down the endogenous expression of LjRNF114 in LJB cells. qRT-PCR and immunoblot analysis showed that endogenous LjRNF114 was obviously decreased at the mRNA and protein levels in LJB cells transfected with LjRNF114 siRNA, compared with that transfected with NC (Fig. 2H, 2I). Silencing of LjRNF114 could significantly reduce the viral RNA level of RGNNV and increase IFN h, MX, ISG15, and Viperin mRNA expression in the presence or absence of RGNNV (Fig. 2J–N). These data indicate that LjRNF114 may inhibit type I IFN activation, thereby promoting RGNNV replication.

It has been known that human RNF114 exhibits E3 ubiquitin ligase function. Consistent with previous reports, LjRNF114 also displayed strong E3 ubiquitin ligase activity (Fig. 3A). Meanwhile, RING and UIM domains, but not C2HC, C2H2a, and C2H2b domains were important for its E3 ubiquitin ligase activity as suggested by the fact that LjRNF114 mutants without RING or UIM suffered some degree of reduced ubiquitin ligase activity compared with LjRNF114 and other LjRNF114 mutants (Fig. 3A).

To determine which domain is required for LjRNF114-mediated inhibition of IFN activation, we constructed multiple LjRNF114 deletion mutants and assessed their abilities to inhibit IFN activation (Fig. 3B). In reporter assays, LjRNF114 mutants △C2HC, △C2H2a, and △C2H2b significantly suppressed IFN promotor activation similarly with full-length LjRNF114, whereas mutants △RING and △UIM lost their inhibitory abilities in FHM cells without RGNNV infection (Fig. 3C). Similar results were observed in the presence of RGNNV infection (Fig. 3D). Above results suggest that the E3 ligase activity of LjRNF114 is required for its abilities to negatively regulate IFN activation.

Previous studies have reported that RNF114 can positively or negatively regulate RLR signaling pathways (20, 21). To investigate the effect of LjRNF114 on sea perch RLR signaling pathways, we examined the role of LjRNF114 in the activation of IFN promoter induced by the key components of RLR signaling pathways. As expected, ectopic expression of MAVS, TRAF3, TBK1, and IRF3 alone successfully activates the IFN h promoter; however, LjRNF114 overexpression significantly inhibited MAVS and TRAF3 but not TBK1- and IRF3-mediated IFN activation (Fig. 4). These results demonstrate that LjRNF114 negatively regulates RLR signaling pathways, thereby suppressing IFN activation.

To identify the key factors in RLR signaling pathways that bind to LjRNF114, we coexpress pEGFP-LjRNF114 with Flag-tagged signaling components of RLRs pathway (MAVS, TRAF3, TBK1, and IRF3) in HEK 293T cells, respectively. Co-IP experiments revealed that LjRNF114 interacted with MAVS and TRAF3 (Fig. 5A, 5B) but not TBK1 and IRF3 (Supplemental Fig. 2). Furthermore, immunofluorescence imaging showed the obvious colocalization between LjRNF114 and MAVS or TRAF3 in the cytoplasm of HEK 293T cells (Fig. 5C). His pull-down analysis showed that LjRNF114 was directly bound to TRAF3 but not MAVS (Fig. 5D). Furthermore, we found that the addition of TRAF3 could not lead to the direct interaction between LjRNF114 and MAVS (Fig. 5E). These data indicate that LjRNF114 physically interacts with MAVS and TRAF3.

To identify the domain of LjRNF114 which is responsible for its interaction with MAVS and TRAF3, HEK 293T cells were cotransfected with pCMV-Flag-TRAF3 or pCMV-Flag-MAVS and GFP-tagged series truncated mutants of LjRNF114 as shown in Fig. 3B. Interestingly, similarly with full-length LjRNF114, all truncated mutants of LjRNF114 could coprecipitate with MAVS and TRAF3, respectively (Fig. 6A, 6B). To further determine the key regions responsible for LjRNF114 and MAVS/TRAF3 interaction, we generated three other LjRNF114 mutants (Fig. 6C). The results showed that only the truncation mutant LjRNF114-ΔRU, which lacks both the RING and UIM domains, failed to interact with MAVS and TRAF3 (Fig. 6D, 6E). These results suggest that both RING and UIM domains of LjRNF114 are sufficient for its interaction with MAVS and TRAF3.

To elucidate the mechanism by which LjRNF114 inhibits RLR signaling pathways, we assessed the effect of LjRNF114 on MAVS and TRAF3 expression. Immunoblot analysis showed that MAVS and TRAF3 were downregulated in the presence of LjRNF114 in a dose-dependent manner (Fig. 7A, 7B). Furthermore, we found that overexpressed LjRNF114 had no influence on the transcription levels of MAVS and TRAF3 (Supplemental Fig. 3). These results indicate that LjRNF114 reduces the expression of MAVS and TRAF3 at protein levels.

Considering LjRNF114 is an E3 ubiquitin ligase and associated with MAVS and TRAF3, we determined whether LjRNF114 catalyzed MAVS and TRAF3 for ubiquitination and degradation. pEGFP-LjRNF114 and pCMV-Flag-MAVS or pCMV-Flag-TRAF3 were cotransfected into HEK 293T cells in the presence or absence of the proteasomal inhibitor MG132. As shown in Fig. 7C and 7D, the degradations of MAVS and TRAF3 caused by LjRNF114 were reversed with the increasing concentration of MG132. In contrast, mutants LjRNF114△RING and LjRNF114△UIM lost their ability to degrade MAVS and TRAF3 to different degrees (Fig. 7E, 7F), indicating that the E3 ligase activity of LjRNF114 is required for LjRNF114-mediated proteasomal degradation of TRAF3 and MAVS. These results clearly demonstrate that LjRNF114 acts as an E3 ubiquitin ligase promoting the proteasomal degradation of both MAVS and TRAF3.

Previous studies showed that RNF114 could promote protein proteasomal degradation by facilitating K27- and K48-linked polyubiquitination (21, 36). Thus, we further determined whether LjRNF114 could catalyze K27 and K48 ubiquitination of MAVS and TRAF3. As shown in Fig. 8, overexpression of LjRNF114 promoted both K27- and K48-linked ubiquitination of MAVS and TRAF3. These data indicate that LjRNF114 induces the degradation of MAVS and TRAF3 via enhancing their K27- and K48-linked ubiquitination.

We further assessed the effect of LjRNF114-mediated MAVS and TRAF3 ubiquitination on their IFN-inducing activities. As shown in Fig. 9, ectopic expression of LjRNF114 and its mutants △C2HC, △C2H2a, and △C2H2b, but not △RING and △UIM, attenuated the IFN-inducing activities of MAVS and TRAF3. Furthermore, overexpression of K27- and K48-linked ubiquitin significantly enhanced the inhibition effect of LjRNF114 and mutants △C2HC, △C2H2a, and △C2H2b on MAVS- and TRAF3-mediated sea perch IFN h promoter activation. These data confirm the importance of LjRNF114-mediated K27- and K48-linked ubiquitination of MAVS and TRAF3 for IFN inhibition (Fig. 10).

Innate immunity is the first line of lower vertebrate in defense against pathogens. However, viruses have developed various elaborate strategies to directly or indirectly control innate immune response to facilitate their survival in host. For instance, peste des petits ruminants virus used its nucleocapsid protein to block type I IFN production through interfering with the formation of the TBK1-IRF3 complex (37). Rhabdovirus infection upregulated the expression of microRNA-3570, which inhibited the antiviral immune response by targeting MAVS (38). In this study, we found for the first time, to our knowledge, that LjRNF114, an E3 ubiquitin ligase, inhibited IFN signaling by functioning as a negative regulator of RLR signaling pathways during RGNNV infection.

A series of studies have revealed that RNF114 is a controversial gene that possesses both positive and negative regulatory functions in innate immune pathway. For example, Bijlmakers et al. (20) reported that RNF114 positively regulated IFN-β production, whereas Lin et al.(21) found that RNF114 played a negative role in RLR-induced IFN production. In the current study, we found that RGNNV infection induced the expression of LjRNF114 and overexpression of LjRNF114-suppressed type I IFN activation and facilitated virus replication, whereas knockdown of LjRNF114 by siRNA increased the IFN response, indicating LjRNF114 is a potential pro-RGNNV factor and may negatively regulate RGNNV-induced immune response. Similarly, with its paralogue RNF125, LjRNF114, as a member of the RING-UIM family, contains an N‐terminal RING finger domain, three zinc fingers (a C2HC domain and two C2H2-type zinc fingers) and a C-terminal UIM (39). In this study, we confirmed that LjRNF114 possessed ubiquitin E3 ligase activity, and UIM and RING domains were important for LjRNF114’s ubiquitin ligase activity, which is consistent with previous studies (18, 40). Lin et al. (21) demonstrated that the E3 ligase activity of RNF114 is responsible for its regulating effect on immune response. Similarly, we found that both UIM and RING domains were crucial for LjRNF114’s inhibitory effect on IFN promoter activation, indicating that the E3 ubiquitin ligase activity of LjRNF114 was required for its inhibitory effect on IFN activation. Thus, it is possible that RNF114 performs its E3 ubiquitin ligase function through RING and UIM domains.

Previous study reported that RNF114 negatively regulated RLR-mediated signaling activation (21). Similarly, LjRNF114 also blocked the type I IFN promoter activation induced by RLR signaling pathways, indicating LjRNF114 might target RLR signaling pathways to suppress IFN activation. Several viral or cellular factors hijacked the key components of RLR signaling pathways to inhibit IFN activation. For instance, severe acute respiratory syndrome coronavirus M protein prevents the formation of TRAF3/TANK/TBK1/IKKε complexes to inhibit IFN production (41). NLRP11 interacted with MAVS to negatively regulate type I IFN signaling (42). Hence, we further investigated the relationship between LjRNF114 and the key components of RLR signaling pathways. In this study, we found that LjRNF114 interacted with both MAVS and TRAF3. Differently, LjRNF114 directly interacted with TRAF3, but not MAVS. It has been known that MAVS is a molecular scaffold and can directly interact with multiple proteins to send signals to downstream adapter molecules, such as RIG-I, TRAF6, and so on. Previous study showed that the interaction between TRAF3 and MAVS needed TRAF3IP3 as a bridge (43). And our results indicated that LjRNF114 could not directly interact with MAVS in the presence of TRAF3. Thus, we speculated that LjRNF114 and MAVS might be bridged by other unknown protein, but this requires further study. Our further studies indicated that RING or UIM domain of LjRNF114 was sufficient for its interaction with TRAF3 and MAVS. Further studies are required to determine the exact amino acid sites of LjRNF114 for its interaction with MAVS and TRAF3.

It was well documented that RLR signaling pathways played crucial roles in initiating innate immune responses, meanwhile, RLR signaling pathways was also tightly regulated to ensure appropriate immune responses through posttranslation modifications or competitive combination. Ubiquitin modifications of key signaling molecules and regulators of RLR signaling pathways served pleiotropic functions in regulating RLR-induced type I IFN production (8). As reported previously, human RNF114 ubiquitinated MAVS and promoted its degradation (26), which promoted us to investigate whether LjRNF114 has analogous effects on MAVS and TRAF3. We found that overexpression of LjRNF114 significantly decreased TRAF3 and MAVS protein levels, which is dependent on LjRNF114’s E3 ubiquitin ligase activity, indicating that LjRNF114 targets MAVS and TRAF3 for ubiquitination and proteasome-dependent degradation. MAVS and TRAF3 were key factors of RLR signaling pathways. MAVS, as a signaling adaptor of the RLR signaling pathways, plays vital roles in transmitting upstream recognition signaling to downstream antiviral signaling (44). Many studies have reported that ubiquitination is involved in the regulation of MAVS activation during virus infection (45). For instance, TRIM29 negatively regulated the host innate immune response to RNA virus by targeting MAVS for K11-linked ubiquitination and degradation (46). RNF125 and RNF5 targeted MAVS for K48-linked ubiquitination and degradation to regulate virus-triggered induction of type I IFN and cellular antiviral response (14, 17). And the recent report showed that porcine RNF114 inhibited classical swine fever virus replication by targeting and promoting K27-linked polyubiquitination and degradation of classical swine fever virus nonstructural protein 4B (36). TRAF3 ubiquitination is essential for IRF3 activation and the IFN response, the K63-linked polyubiquitination of TRAF3 promotes type I IFN production, whereas the K48-linked polyubiquitination of TRAF3 results in the inhibition of IFN response (12, 47). Triad3A suppresses the RLR signaling pathways through K48-linked ubiquitin–mediated degradation of TRAF3 (48). In this study, we found that LjRNF114 promoted the K27- and K48-linked ubiquitination of MAVS and TRAF3, which is important for IFN inhibition. However, the exact lysine residues of MAVS and TRAF3 that K27-/K48-linked ubiquitination targets and whether other ubiquitin ligases are also involved in this complicated regulation remained to be further investigated. In addition, increasing evidence has shown that some viruses can evade host antiviral defenses by targeting RLR signaling pathways (49, 50). Thus, the relationship among RGNNV-encoded proteins, LjRNF114, and RLR signaling pathways is a question deserving further research.

In summary, we found that LjRNF114 negatively regulated RLR signaling pathways and promoted RGNNV replication. Mechanistically, LjRNF114 interacted with and promoted K27-/K48-linked polyubiquitination of MAVS and TRAF3, which in turn inhibited RLRs-mediated IFN activation and subsequent enhanced virus replication (Fig. 10). Our study demonstrated that LjRNF114 might be the potential target in the development of novel therapeutics against NNV infection.

We thank Dr. Yibing Zhang (Institute of Hydrobiology, Wuhan, China) for providing vector (pRL-TK).

The authors have no financial conflicts of interest.