TNFR-associated factor 6 (TRAF6) not only recruits TBK1/IKKε to MAVS upon virus infection but also catalyzes K63-linked polyubiquitination on substrate or itself, which is critical for NEMO-dependent and -independent TBK1/IKKε activation, leading to the production of type I IFNs. The regulation at the TRAF6 level could affect the activation of antiviral innate immunity. In this study, we demonstrate that zebrafish prmt2, a type I arginine methyltransferase, attenuates traf6-mediated antiviral response. Prmt2 binds to the C terminus of traf6 to catalyze arginine asymmetric dimethylation of traf6 at arginine 100, preventing its K63-linked autoubiquitination, which results in the suppression of traf6 activation. In addition, it seems that the N terminus of prmt2 competes with mavs for traf6 binding and prevents the recruitment of tbk1/ikkε to mavs. By zebrafish model, we show that loss of prmt2 promotes the survival ratio of zebrafish larvae after challenge with spring viremia of carp virus. Therefore, we reveal, to our knowledge, a novel function of prmt2 in the negative regulation of antiviral innate immunity by targeting traf6.
Innate immunity is the host’s first line of defense against microbial infection. Germline-encoded cellular pattern recognition receptors), such as TLRs, retinoic acid-inducible gene 1 (RIG-I)–like receptors (RLRs) and NOD-like receptors, recognize conserved microbial components called pathogen-associated molecular patterns, such as virus-derived nucleic acids, to initiate host antimicrobial response (1, 2). Viral DNAs and RNAs are mainly sensed by the cytosolic nucleic acid sensors, the DNA sensor cyclic guanosine monophosphate-adenosine monophosphate synthase (cGAS) (3–5), and the RNA sensors RLRs, including RIG-I and melanoma differentiation-associated gene 5 (MDA5), respectively (6–9). After binding to viral nucleic acids, activated sensors recruit adaptor proteins, including TIR domain-containing adaptor molecular 1 (TRIF), mitochondrial antiviral signaling protein (MAVS) (also known as CAR-DIF, IPS-1, and VISA), or stimulator of IFN genes (STING) protein, which then activates the downstream signaling molecules TBK1/IKKε, as well as IKKα/β/γ (10–13). These kinases phosphorylate the transcription factor IFN regulatory factor (IRF) 3 and the inhibitor of NF-κB, leading to activation of IRF3 and NF-κB and eventual induction of type I IFNs, proinflammatory cytokines, and other downstream effector genes (1, 2). To efficiently protect the host from infections and concomitantly avoid excessive immunopathology, innate immunity needs to be tightly regulated (1, 14–17) (16, 17).
Although RIG-1 and MDA5 sense distinct types of viruses (18), they share a common adaptor, MAVS. In response to RNA virus infection, MAVS serves as a key hub that links virus recognition to downstream innate antiviral immune responses (2, 19–21). Upon sensing diverse cytosolic RNAs, RIG-I undergoes conformational changes, oligomerization, and exposure of the N-terminal tandem caspase activation and recruitment domains (2CARD) (22–24). RIG-I 2CARD then binds to the N-terminal CARD domain of MAVS and induces a conformational change, leading to the formation of prion-like MAVS aggregates (20, 25). MAVS aggregation serves as a signaling platform to form a MAVS signalosome for activating downstream components (20, 21, 25).
TNFR-associated factors (TRAFs), as E3 ubiquitin ligases, are required for the recruitment of the downstream kinases to MAVS (26) (20, 25). Among all of the TRAFs, TRAF6 is essential and could not be replaced by other TRAF proteins, which are preassociated with TBK1/IKKε and recruited to MAVS upon virus infection. In addition, its E3 ligase activity is required for NEMO activation by attaching ubiquitin chains to NEMO for NF-κB and TBK1/IKKε activation (13, 27–30). Therefore, the regulation of TRAF6 plays an important role in antiviral innate immunity.
Arginine methylation is a posttranslational modification (PTM) in histone and nonhistone proteins catalyzed by a group of nine protein arginine methyltransferases (PRMT1–9) (31). These enzymes transfer a methyl group from S-adenosyl-methionine to an arginine guanidine nitrogen, which are classified into three categories, depending on the modification they catalyze (31). Type I (PRMT1, PRMT2, PRMT3, CARM1/PRMT4, PRMT6, and PRMT8) catalyzes to form ω-NG-monomethyl arginine (MMA) and ω-NG, NG-asymmetric dimethyl arginine (ADMA); type II (PRMT5 and PRMT9) catalyzes the formation of MMA and ω-NG, N′G symmetric dimethyl arginine (SDMA); type III (PRMT7) catalyzes the formation of MMA only (31). Arginine methylation can affect numerous cellular activities, including transcription and chromatin regulation, cell signaling, DNA damage response, RNA expression, and cellular metabolism (31–33). Arginine methylation also plays a critical role in inflammatory responses as well as establishment and maintenance of the lymphoid and myeloid progenitors (31, 34–36). Some PRMTs have been shown to impact antimicrobial responses (37–41), linking arginine methylation to the modulation of innate immunity. However, the underlying mechanisms about how arginine methylation regulates innate immunity are still largely unknown. As a type I PRMT, PRMT2 has been described as a transcriptional coactivator for nuclear receptors and its repressive function in E2F-mediated G1 to S phase cell-cycle progression and STAT3 signaling (34, 42–45). PRMT2 has also been found to inhibit NF-κB signaling and is even considered as a new member of the NF-κB pathway controlling the LPS-induced inflammatory and lung responses (46, 47), implicating a role for PRMT2 in innate immunity. Recently, we have shown that prmt3 and prmt7 negatively regulate antiviral innate immunity and PRMT7 catalyzes arginine monomethylation of MAVS to control its activity, extending the role of arginine methylation in antiviral responses (38, 40, 48).
In this study, we reveal that zebrafish prmt2 negatively regulates RLR signaling both in vitro and in vivo. Prmt2 associates with traf6 to catalyze arginine methylation of traf6 at arginine residue 100 (R100), inhibiting traf6 autoubiquitination, leading to the repression of RNA virus-induced IFN activation. Moreover, it seems that the N terminus of prmt2 competes with mavs for traf6 binding and prevents tbk1/ikkε being recruited to mavs, resulting in the inhibition of tbk1/ikkε activation. Therefore, we uncover a novel, to our knowledge, function of prmt2 in RLR signaling by targeting traf6.
Materials and Methods
Cells, viruses, and fish
We cultured epithelioma papulosum cyprini (EPC) cells (derived from skin of cyprinid fathead minnow Pimephales promelas; American Type Culture Collection, Manassas, VA) supplemented with 10% FBS and that were maintained at 28°C in a humidified incubator containing 5% CO2. HEK293T cells were cultured in DMEM (HyClone Laboratories) with 10% FBS and were grown at 37°C in a humidified incubator containing 5% CO2. THP1 cells, originally obtained from American Type Culture Collection, were maintained in RPMI 1640 supplemented with 10% FBS and were grown at 37°C in a humidified incubator containing 5% CO2. We propagated spring viremia of carp virus (SVCV) in EPC cells until the cytopathic effect was complete. We collected the culture medium and stored it at −80°C until use. Sendai virus (SeV) was provided by Dr. Bo Zhong (Wuhan University, Wuhan, China), VSV-GFP was provided by Dr. Mingzhou Chen (Wuhan University, Wuhan, China). For viral injection of zebrafish, 2-mo-old zebrafish were i.p. injected with SVCV (∼2.51 × 107 50% tissue culture-infective dose [TCID50]/ml) at 10 μl/individual. An i.p. injection with cell culture medium was used as the control.
Generation of prmt2-null zebrafish
CRISPR/Cas9 was used to knockout prmt2 in zebrafish. prmt2 single guide RNA (sgRNA) was designed using an online CRISPR design tool (http://crispr.mit.edu). The zebrafish codon-optimized Cas9 plasmid was digested with XbaI, then purified and transcribed using the T7 mMessage mMachine Kit (Ambion). pUC19-gRNA vector was used for amplifying prmt2 sgRNA. The primers for amplifying prmt2 sgRNA template were as follows: 5′-TGTAATACGACTCACTATAGGAGTTCAGACGTGTGGTGGGTTTTAGAG CTAGAAATAG-3′and 5′- TGTAATACGACTCACTATAGGATTTGGGCTGT GGGACAGGTTTTAGAGCTAGAAATAG-3′. The TranscriptAid T7 High Yield Transcription Kit (Fermentas) was used to synthesize sgRNA. Zebrafish (Danio rerio) strain AB was raised, maintained, reproduced, and staged according to the standard protocol. Cas9 RNA (0.75–1.25 ng/embryo) and sgRNA (0.075 ng/embryo) were mixed and injected into embryos at the one-cell stage. The mutants were initially detected using a heteroduplex mobility assay (HMA) as previously described. If the HMA results were positive, the remaining embryos were raised to adulthood as the F0 generation and then backcrossed with wild-type (WT) zebrafish (strain AB) to generate the F1 generation, which were genotyped by HMA initially and confirmed by sequencing of target sites. Heterozygous F1s were backcrossed with WT (disallowing offspring–parent mating) to generate the F2 generation. F2 adults carrying the target mutation (+/−) were intercrossed to generate F3 offspring. The F3 generation contained WT (+/+), heterozygous (+/−), and homozygous (−/−) individuals. The zebrafish prmt2 mutants were named following the zebrafish nomenclature guidelines (http://wiki.zn.org/display/general/ ZFIN+Zebrash+Nomenclature+Guidelines). Two mutants were obtained, namely ihbp2d11 (https://zfin.org/action/feature/view/ZDB-ALT-200707-10#) (mutant 1) and ihbp2d26 (https://zfin.org/action/feature/view/ZDB-ALT-200707-11#) (mutant 2). All experiments with zebrafish were approved by the Institutional Animal Care and Use Committee of Institute of Hydrobiology, Chinese Academy of Sciences under protocol number 2017-001. Virus infection of zebrafish larvae was performed as previously described (49).
CRISPR-Cas9 knockout cell lines
To generate PRMT2-, MAVS-, and TRAF6-knockout cell lines, double-stranded oligonucleotides corresponding to the target sequences were cloned into LentiCRISPRv2 plasmid and then cotransfected with viral packaging plasmids into HEK293T cells. Two days after transfection, the viruses were harvested, ultrafiltrated (0.22-mm filter; Millipore), and used to infect HEK293T cells in the presence of polybrene (8 μg/ml). The infected cells were selected with puromycin (1 μg/ml) for 2 wk and then pooled for further experiments. The efficiency of CRISPR/Cas9-mediated gene knockout was confirmed by Western blot analysis. The target sequences of guide RNAs are listed as the following: human PRMT2-sgRNA1, 5′-GAGGCCGGTCTCCTGCAGGA-3′; human PRMT2-sgRNA2, 5′-GCAGGAGGGAGTACAGCCAG-3′; human MAVS-sgRNA, 5′-CTGTGAGCTAGTTGATCTCG-3′; and human TRAF6-sgRNA, 5′-CTAAACTGTGAAAACAGCTG-3′.
Western blot analysis
We extracted the total protein of EPC, THP1, or HEK293T cells with RIPA buffer containing 50 mM Tris (pH 7.4), 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM EDTA (pH 8), 150 mM NaCl, 1 mM NaF, 1 mM PMSF, 1 mM Na3VO4, and a 1:100 dilution of protease inhibitor mixture (Sigma-Aldrich). Coimmunoprecipitation was performed as previously described. We used the Fujifilm LA S4000 Immunoluminescent Image Analyzer to image the blots. The following Abs were used for immunoblotting: anti–c-Myc Ab (9E10, 1:1000, sc-40; Santa Cruz Biotechnology), anti-HA Ab (1:5000, MMS-101R; Covance), Flag (1:5000, F1804-5MG; Sigma-Aldrich), Flag (Ra) (1:1000, 14793; Cell Signaling Technology), anti–β-actin Ab (1:20,000, AC026; ABclonal Technology), GAPDH (1:2000, sc-47724; Santa Cruz Biotechnology), PRMT2 (1:1000, ARP40196_T100; Aviva Systems Biology), TRAF6 for immunoblotting (1:1000, ab33915, Abcam), TRAF6 for IP (1:200, R1311-2, HUABIO).
Quantitative real-time PCR
Total RNA was extracted using RNAiso Plus (TaKaRa Bio, Beijing, China) following the protocol provided by the manufacturer. cDNAs were synthesized using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA). MonAmp SYBR Green qPCR Mix (High ROX) (Monad Biotech, Shanghai, China) was used for quantitative real-time PCR (qRT-PCR) assays. The primers for qRT- PCR are listed in Supplemental Table I.
Luciferase reporter assays
EPC cells were grown in 24-well plates and transfected with various amounts of plasmids by VigoFect (Vigorous Biotechnology, Beijing, China), as well as with CMV-Renilla used as control. After the cells were transfected for the indicated time, the luciferase activity was determined by the Dual-Luciferase Reporter Assay System (Promega). Data were normalized to Renilla luciferase. Data are representative of three independent experiments (mean ± SD)
Flow cytometry assay
HEK293T cells were challenged with VSV-eGFP for 12 h at a multiplicity of infection (MOI) of 1, then harvested and washed with PBS. For each condition, 10,000 cells were counted, and the analyses that followed were performed on a Beckman Coulter CytoFLEX S. The data were analyzed and generated with FlowJo software.
Identification of traf6 methylarginine site(s) by mass spectrometry
HEK293T cells were cotransfected with Flag-traf6 and HA-prmt2 plasmid. Cell lysate was immunoprecipitated with anti-Flag Ab-conjugated agarose beads overnight. Immunoprecipitated traf6 proteins were subjected to 8% SDS-PAGE gel, and traf6 bands were excised from the gel and analyzed by mass spectrometry at Protein Gene Biotech (Wuhan, China).
Generation of zebrafish anti-prmt2 and anti-traf6-R100me2a Abs
Zebrafish prmt2 Ab were generated by using a zebrafish prmt2 peptide (QRTEDMEDAWQDDEYFGN) as an Ag; traf6 site-specific arginine asymmetric dimethyl arginine Abs (anti-traf6-R100me2a) were generated by using a zebrafish traf6 asymmetric dimethyl arginine peptide (RKSI(R-me2a) DTGQK-C) as an Ag. After purifying the Abs with excess unmodified traf6 peptide (RKSIRDTGQK-C), Abs recognizing zebrafish traf6 arginine asymmetric dimethyl arginine were enriched by biotin-labeled traf6 asymmetric dimethyl arginine peptide.
For analysis of the ubiquitination of traf6 in HEK293T cells, HEK293T cells were transfected with plasmids expressing Myc-traf6, Flag-traf6, His-ubiquitin (WT), His-ubiquitin (K48 only), His-ubiquitin (K63 only), and Flag-prmt2, Ubiquitination assays with His-ubiquitin or His-ubiquitin-mutants were performed by affinity purification on Ni2+-NTA Resin (Novagen) as described previously (50). For analysis of the ubiquitination of TRAF6 in HEK293T cells, HEK293T cells were infected with SeV, then whole cells were re-extracted in NP-40 lysis buffer containing 1% SDS and denatured by heating for 5 min. The supernatants were diluted with regular lysis buffer until the concentration of SDS was decreased to 0.1%, followed by reimmunoprecipitation with the indicated Abs, and the immunoprecipitants were analyzed by immunoblotting with the ubiquitin Ab.
An unpaired Student t test and log-rank (Mantel–Cox) test were used for statistical analysis with GraphPad Prism6 Software. All data are representative of three independent experiments performed in triplicates; p < 0.05 was considered significant; *p < 0.05, **p < 0.01 and ***p < 0.001.
Zebrafish prmt2 negatively regulates RLR signaling
We have shown that zebrafish prmt3 and prmt7 are involved in the regulation of antiviral innate immunity (38, 40). To get a full picture about the function of PRMTs in innate immunity, we further examined whether zebrafish prmt2 is also involved in antiviral innate immunity. Overexpression of zebrafish prmt2 suppressed IFN stimulation response element (ISRE), and Dr-IFNφ1, Dr-IFNφ3, and EPC-IFN (38, 40, 51) reporter activity was induced by poly(I:C) transfection or SVCV infection in EPC cells (Fig. 1A). Consistently, overexpression of zebrafish prmt2 reduced the expression of ifn, a typical irf3/irf7 downstream gene, and two typical IFN-stimulated genes (isg15 and viperin) in EPC cells after being transfected with poly(I:C) or challenged with SVCV via qRT-PCR assays (51) (Fig. 1B). In agreement with this notion, overexpression of prmt2 in EPC cells caused an enhanced cytopathic effect compared with the empty vector control upon challenge with different titers of SVCV (from an MOI of 1–1000) (Fig. 1C). The copy numbers of SVCV genes, including the G gene, N gene, and P gene, were increased in prmt2-overexpressed EPC cells, as revealed by qRT-PCR assays (Fig. 1D, 1E). Of note, ectopic expression of prmt2 in cells had no overt effect on cell viability. These data suggest that zebrafish prmt2 negatively regulates the RLR signaling pathway and attenuates the antiviral responses.
Disruption of PRMT2 in HEK293T cells enhances cellular antiviral responses
PRMT2 is evolutionarily conserved between zebrafish and mammal (Supplemental Fig. 2A); to determine whether the function of PRMT2 is also conserved between zebrafish and human, we used human cell lines and genes to perform further assays. We found that knockout of PRMT2 in HEK293T cells by CRISPR/Cas9 caused an increase of ISRE promoter activity induced by SeV infection or poly (I:C) transfection (Fig. 2A). The efficiency of PRMT2-sgRNA 1– and PRMT2-sgRNA 2–mediated knockout of PRMT2 was confirmed by Western blot assay (Supplemental Fig. 1A). Furthermore, upon SeV infection, knockout of PRMT2 enhanced expression of IFNB1, ISG56, CXCL10, and IL6 greatly (Fig. 2B). However, reconstitution of PRMT2 recovered the suppressive role of PRMT2 on IFNB1 expression (Supplemental Fig. 1B). Consistent with the induction of gene expression, knockout of PRMT2 enhanced phosphorylation of IRF3 and TBK1 in response to poly(I:C) transfection and SeV infection (Fig. 2C, 2D). As detected by flow cytometric analysis and fluoresce microscopy, knockout of PRMT2 caused a reduction of VSV-GFP+ HEK293T cells (Fig. 2E). Taken together, these data suggest that PRMT2 negatively regulates type I IFN signaling induced by RLRs.
Disruption of prmt2 in zebrafish facilitates antiviral responses
To determine the physiological function of prmt2 in vivo, we disrupted prmt2 in zebrafish via CRISPR/Cas9 and obtained two mutant lines (Supplemental Fig. 2B, 2C). By crossing prmt2+/−(♀) × prmt2+/−(♂), the offspring with prmt2+/+, prmt2+/−, and prmt2−/− genetic backgrounds were born at a Mendelian ratio (1:2:1), and no obvious defects in growth rate and reproduction capability were detected in prmt2−/− zebrafish under normal conditions. We challenged prmt2−/− larvae (n = 100 in total) and prmt2+/+ larvae (having WT allele of prmt2) (WT) (n = 100 in total) with high-titer SVCV and counted the number of dead larvae at different time points. The dead larvae exhibited no movement, no blood circulation, and a degenerated body (Fig. 3A). As shown in (Fig. 3B, prmt2−/− larvae had a higher survival ratio compared with the WT larvae. Consistently, the viral titer was lower, and the key antiviral genes, including ifn1, lta, and mxc were enhanced; the copy number of the G, N, and P genes of SVCV indicated by mRNA abundance was decreased in prmt2−/− larvae compared with the WT larvae after challenge with SVCV (Fig. 3C, 3D) (51, 52). Upon i.p. injection with SVCV, prmt2−/− adult zebrafish (2 mo postfertilization [mpf]) exhibited less swelling and fewer hemorrhagic symptoms in the abdomen compared with the WT adult zebrafish (2 mpf) (Fig. 3E, 3F). In agreement, the key antiviral genes, including ifn1, lta, and mxc, were enhanced, and the copy number of the G, N, and P genes of SVCV were decreased in prmt2−/− zebrafish compared with the WT zebrafish (Fig. 3G). Collectively, our finding suggests that zebrafish prmt2 negatively regulates antiviral response in vivo.
Zebrafish prmt2 interacts with traf6
To determine the mechanisms of zebrafish prmt2 in negative regulation of antiviral immunity, we examined which component in RLR signaling pathway, including rig-1, mavs, traf3, traf6, tbk1, and irf3 or irf7 could interact with prmt2. As shown in (Fig. 4A, prmt2 could only be pulled by traf6 (Fig 4A), which was further confirmed endogenously in lysate of zebrafish brain (Fig. 4B). Domain mapping indicated that the C terminus of traf6 (373–542 aa) and the N terminus of prmt2 (1–73 aa) were required for their interaction (Fig. 4C, 4D). Additionally, we confirmed that human PRMT2 interacts with human TRAF6 as well (Fig. 4E). Of note, in THP1 cells, human PRMT2 interacted with human TRAF6 endogenously with or without SeV infection (Fig. 4F), indicating a conserved behavior of PRMT2/TRAF6 interaction between zebrafish and mammal. These data suggest that prmt2 might affect RLR signaling through impacting on traf6.
Zebrafish prmt2 inhibits traf6 autoubiquitination via its arginine methyltransferase activity
Given that Lys63-linked autoubiquitination of TRAF6 is a key activation step for TRAF6 (53), we sought to know whether prmt2 can affect Lys63-linked autoubiquitination of TRAF6. As shown in (Fig. 5A–5C, overexpression of prmt2 reduced ubiquitination of traf6 with coexpression of WT ubiquitin and K63-only ubiquitin, but not with coexpression of K48-only ubiquitin. Notably, in PRMT2-knockout HEK293T cells (PRMT2−/−), endogenous Lys63-linked autoubiquitination of TRAF6 was higher than that in PRMT2-sufficient HEK293T cells (PRMT2+/+) upon SeV infection (Fig. 5D).
As an E3 ligase, the human TRAF6 (L74H) mutant prevents interaction with E2-conjugating enzymes (28, 54); thereby, this mutant is devoid of E3 ligase activity. To evaluate whether prmt2 affects the enzymatic activity of traf6, we tested the effect of prmt2 on ubiquitination of WT traf6 and the traf6 (L75H) mutant (corresponds to human TRAF6 [L74H]) (Supplemental Fig. 3A). Overexpression of prmt2 only caused a reduction of WT traf6 ubiquitination but not of the traf6 (L75H) mutant, suggesting that prmt2 might influence the enzymatic activity of traf6 (Fig. 5E). Subsequently, we examined whether arginine methyltransferase activity of prmt2 is required for inhibiting traf6 ubiquitination. As shown in (Fig. 5F, compared with overexpression of WT prmt2, overexpression of an enzymatic-deficient mutant of prmt2 (prmt2-GG) (42) had no overt effect on the reduction of traf6 ubiquitination (Fig. 5F), indicating that the enzymatic activity of prmt2 is required for prmt2 to inhibit traf6 ubiquitination.
Taken together, these data suggest that zebrafish prmt2 inhibits traf6 autoubiquitination via its arginine methyltransferase activity.
Arginine methylation of traf6 at R100 mediated by prmt2 inhibits traf6 autoubiquitination and RNA virus-induced IFN activation
Because arginine methyltransferase activity is required for prmt2 to inhibit traf6 ubiquitination, we next examined whether prmt2 catalyzes arginine methylation of traf6. Through mass spectrometry analysis, we identified that traf6 could be methylated at arginine 100 (R100) (Fig. 6A), which is evolutionarily conserved among human, mouse, and zebrafish (Fig. 6B). To further confirm the methylated site in zebrafish traf6, we developed a specific Ab against the asymmetric dimethylated form (two methyl groups on the same arginine guanidino nitrogen) of zebrafish traf6 arginine 100 (anti-traf6-R100me2a) and we confirmed the Ab with Western blot assay (Supplemental Fig. 3B). Overexpression of prmt2 together with traf6 induced asymmetric arginine dimethylation of traf6 (Fig. 6C). Moreover, in prmt2+/+ zebrafish, arginine dimethylation of traf6 was readily detected but not in prmt2−/− zebrafish (Fig. 6D). In response to SVCV infection, the induction ability of the methyl-mimic mutant of traf6 (traf-R100F) on ifn expression in EPC cells was much lower than that of WT traf6 (Fig. 6E), further validating that arginine methylation of traf6 mediated by prmt2 suppressed traf6 function in antiviral innate immunity. Consistently, overexpression of prmt2 inhibited K63-linked ubiquitination of WT traf6 but not of the methyl-mimic mutant of traf6 (traf6-R100F) (Fig. 6F).
To further validate whether the function of asymmetric dimethylation of traf6 at R100 mediated by prmt2 on the suppression of antiviral immunity was evolutionarily conserved between zebrafish and mammal, we used human HEK293T cells to perform assays. Disruption of TRAF6 in HEK293T cells resulted in a significant reduction of IFNB1 expression in response to SeV infection (Fig. 6G). Similar to what was observed in EPC cells, in response to SeV infection, the induction of IFNB1 expression was much lower in TRAF6-null HEK293T cells (TRAF6 −/−) reconstructed with the methyl-mimic mutant of human TRAF6 (TRAF6-R99F, corresponding to zebrafish traf-R100F) than in those with WT TRAF6 (Fig. 6G). In agreement, in response to SeV infection, disruption of TRAF6 caused a reduction in phosphorylation of TBK1 and IRF3 compared with that in WT HEK293T cells (TRAF6 +/+). In addition, upon SeV infection, the induction of phosphorylation of TBK1 and IRF3 in TRAF6 −/− HEK293T cells reconstructed with TRAF-R99F was lower than in those with WT TRAF6 (Fig. 6H). Consistently, in TRAF6 −/− HEK293T cells, the K63-linked autoubiquitination of TRAF6 was also lower with the reconstruction of TRAF6-R99F compared with the reconstruction of WT TRAF6 (Fig. 6I). In addition, we confirmed that R100 residue of traf6 is asymmetrically dimethylated by prmt2, and the R100F mutant of traf6 still interacts with prmt2 (Supplemental Fig. 3C, 3D).
These results suggest that prmt2 catalyzes traf6 to be dimethylated at R100, which inhibits traf6 autoubiquitination and RNA virus-induced IFN activation.
Prmt2 might also inhibit traf6 by competing with mavs for traf6 binding to prevent the recruitment of tbk1/ikkε to mavs
As we have found that prmt2 targets traf6 and inhibits traf6 autoubiquitination via its arginine methyltransferase activity, we sought to determine whether the inhibition of RLR signaling by prmt2 is completely relying on its arginine methyltransferase activity. By ISRE promoter assays in EPC cells, we noticed that, although compared with that of WT prmt2, the inhibitory role of the enzymatic-deficient mutant of prmt2 (prmt2-GG) on ISRE promoter activity was obviously recovered, it still suppressed the ISRE promoter activity significantly in response to SVCV infection (Fig. 7A). However, the mutant with the deletion of the N terminus (SH3 domain) in prmt2-GG (prmt2-ΔSH3-GG) completely lost the suppressive role on activating ISRE promoter activity (Fig. 7A). It is known that the C terminus of TRAF6 binds to MAVS and is required for recruiting TBK1/IKKε to MAVS for activating antiviral signaling (20, 30, 55, 56). In this study, we found that the C terminus of traf6 bound to the N terminus (SH3 domain) of prmt2 (Fig. 4C, 4D). Therefore, this result suggests that prmt2 can affect antiviral signaling, depending on protein–protein interaction in addition to its enzymatic activity. Similar results were obtained by examining ifn expression in EPC cells via qRT-PCR (Fig. 7B). Consistently, overexpression of WT prmt2, as well as prmt2-GG, prevented traf6 binding to mavs, but overexpression of prmt2-ΔSH3-GG did not (Fig. 7C, 7D).
To further validate that this behavior is evolutionarily conserved, we examined this phenomenon in human HEK293T cells. Similarly, in PRMT2−/− HEK293T cells upon SeV infection, compared with that of WT PRMT2, the inhibitory role of the enzymatic-deficient mutant of PRMT2 (PRMT2-GG), on the induction of IFNB1 expression was obviously recovered, but it still suppressed IFNB1 expression significantly in response to SeV infection (Fig. 7E). However, the mutant with the deletion of the N terminus (SH3 domain) in PRMT2-GG (PRMT2-ΔSH3-GG) completely lost the suppressive role on the induction of IFNB1 expression (Fig. 7E). In agreement with this notion, in PRMT2−/− HEK293T cells, reconstruction of WT PRMT2 (PRMT2-WT) dramatically suppressed phosphorylation of TBK1 and IRF3, but reconstruction of the enzymatic-deficient mutant of PRMT2 (PRMT2-GG) partially recovered phosphorylation of TBK1 and IRF3. Moreover, reconstruction of the mutant with the deletion of the N terminus (SH3 domain) in PRMT2-GG (prmt2-ΔSH3-GG) almost completely recovered phosphorylation of TBK1 and IRF3 (Fig. 7F).
Subsequently, we confirmed that the N terminus (SH3 domain) of PRMT2 inhibited the recruitment of TBK1 and IKKε to MAVS (Fig. 7G). Furthermore, we found that in the presence of WT MAVS, overexpression of PRMT2 caused a reduction of TBK1 and IKKε binding to MAVS, but in the presence of MAVS-d6 mutant [loss of interaction with TRAF6 (26)], overexpression of PRMT2 had no overt effect on the recruitment of TBK1 and IKKε to MAVS (Fig. 7H), further indicating that TRAF6 mediated the inhibitory role of prmt2 on the recruitment of TBK1 and IKKε to MAVS.
Consciously, given that all of above assays relied on ectopic expression of proteins, the enzyme-independent inhibitory role of prmt2 on traf6 might only represent an alternative mechanism, not a main mechanism. Of course, it might also be an overexpression artifact. Further addressing this issue by in vivo assays will get deep insights into the role and the underlying mechanism of prmt2 in antiviral immunity.
Based on the data we obtained in this study, we proposed a work model for prmt2 (Fig. 8). In general, prmt2 negatively regulates RLR signaling in both a prmt2 enzymatic activity-dependent and -independent manner. On one hand, arginine asymmetric dimethylation of traf6 catalyzed by prmt2 inhibits traf6 autoubiquitination, leading to the inactivation of the nemo-tank-tbk1/ikkε complex. On the other hand, the N terminus of prmt2 competes with mavs for traf6 binding, thereby preventing the recruitment of tbk1/ikkε to mavs mediated by traf6. By these two ways, prmt2 eventually inhibits phosphorylation of irf3/7, leading to the downregulation of type I IFNs.
RLR signaling is regulated by a series of PTMs, such as phosphorylation, polyubiquitination, succinylation, and O-GlcNAcylation (11, 16, 17, 57–66). Therefore, to identify PTMs of components in RLR signaling and understand their actions will give insights into the mechanisms of how RLR signaling is tightly regulated. As arginine methyltransferases for modifying histone and nonhistone proteins, PRMTs have been recently recognized to modulate antiviral signaling an enzymatic activity-dependent or -independent manner (37, 38, 40, 67). In comparison with other modifications, arginine methylation in components of RLR signaling is barely identified; therefore, the process and underlying mechanisms are still not well defined. In this study, we identify that prmt2 negatively regulates RLR signaling by catalyzing arginine asymmetric dimethylation of traf6 at R100, revealing a function of arginine methylation in antiviral innate immunity. Given that PRMT2 also negatively regulates the LPS-induced lung responses by targeting NF-κb signaling (47), PRMT2 might be involved in both bacterial and viral infection through targeting different signaling pathways.
We have previously shown that prmt3 and prmt7 could also negatively regulate RLR signaling (38, 40). Although how the mechanism of prmt3 acts in its role is still a puzzle, it is evident that prmt7 catalyzes arginine monomethylation of MAVS to play its suppressive role on RLR signaling (48). It seems that PRMTs act in their function through various mechanisms at different levels of the IRF signaling pathway. Further identifying whether other PRMTs are involved in RLR signaling and whether other molecules of RLR signaling are targeted by PRMTs will help us to get a full picture about the roles and underlying mechanisms of PRMTs in antiviral innate immunity.
As an important signaling scaffold downstream of multiple pattern recognition receptors, TRAF6 recruits IKKα/β/γ and/or TBK1 complex to adaptor proteins by attaching the K63-linked polyubiquitin chains on substrate proteins or itself (26, 30, 68). Therefore, the influences on autoubiquitination or protein stability of TRAF6 obviously affect antiviral innate immunity (69, 70). However, the effect of TRAF6 modification on antiviral innate immunity is barely understood. In this study, we find that prmt2 modifies traf6 at R100 to prevent its autoubiquitination, leading eventually to attenuation of antiviral innate immunity. It suggests that PTMs of TRAF6 might also contribute to tightly control RLR signaling as well as those of other components in the RLR signaling pathway.
By switching the zebrafish model and mammalian cell lines, we validate that the function of PRMT2 on RLR signaling is evolutionarily conserved between zebrafish and mammal. Given the sufficient tools in mammalian system, such as Abs and cell lines, further investigation of the precise process and the underlying mechanisms of PRMT2 in acting in its role will give insights in the action of PRMT2 in antiviral innate immunity.
We thank Drs. Hongbing Shu, Bo Zhong, and Mingzhou Chen for providing reagents. We thank Yan Wang at the Core Facility of the Institute of Hydrobiology for FACS.
This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences Grant XDA24010308, the National Natural Science Foundation of China Grants 31721005, 31830101, and the National Key R & D Program of China (2018YFD0900602).
The online version of this article contains supplemental material.
Abbreviations used in this article
epithelioma papulosum cyprinid
heteroduplex mobility assay
IFN regulatory factor
IFN stimulation response element
mitochondrial antiviral signaling protein
multiplicity of infection
quantitative real-time PCR
retinoic acid-inducible gene 1
single guide RNA
spring viremia of carp virus
50% tissue culture-infective dose
The authors have no financial conflicts of interest.