key: cord-1041123-00mqmpzw
authors: Qian, Wei; Wei, Xiaoqin; Guo, Kelei; Li, Yongtao; Lin, Xian; Zou, Zhong; Zhou, Hongbo; Jin, Meilin
title: The C-Terminal Effector Domain of Non-Structural Protein 1 of Influenza A Virus Blocks IFN-β Production by Targeting TNF Receptor-Associated Factor 3
date: 2017-07-03
journal: Front Immunol
DOI: 10.3389/fimmu.2017.00779
sha: 2b6524fc0ce858d2db326b94ff9bf48cc14cd9a9
doc_id: 1041123
cord_uid: 00mqmpzw

Influenza A virus non-structural protein 1 (NS1) antagonizes interferon response through diverse strategies, particularly by inhibiting the activation of interferon regulatory factor 3 (IRF3) and IFN-β transcription. However, the underlying mechanisms used by the NS1 C-terminal effector domain (ED) to inhibit the activation of IFN-β pathway are not well understood. In this study, we used influenza virus subtype of H5N1 to demonstrate that the NS1 C-terminal ED but not the N-terminal RNA-binding domain, binds TNF receptor-associated factor 3 (TRAF3). This results in an attenuation of the type I IFN signaling pathway. We found that the NS1 C-terminal ED (named NS1/126-225) inhibits the active caspase activation and recruitment domain-containing form of RIG-I [RIG-I(N)]-induced IFN-β reporter activity, the phosphorylation of IRF3, and the induction of IFN-β. Further analysis showed that NS1/126-225 binds to TRAF3 through the TRAF domain, subsequently decreasing TRAF3 K63-linked ubiquitination. NS1/126-225 binding also disrupted the formation of the mitochondrial antiviral signaling (MAVS)–TRAF3 complex, increasing the recruitment of IKKε to MAVS; ultimately shutting down the RIG-I(N)-mediated signal transduction and cellular antiviral responses. This attenuation of cellular antiviral responses leads to evasion of the innate immune response. Taken together, our findings offer an important insight into the interplay between the influenza virus and host innate immunity.

of caspase activation and recruitment domains (CARD). As the cascade continues, different residues in RIG-I are ubiquitinated by the E3 ligases TRIM25 (5) and RIPLET (6) , resulting in RIG-I oligomerization. Subsequently, RIG-I interacts with the adaptor protein mitochondrial antiviral signaling (MAVS) via their CARD-CARD association, which in turn results in MAVS oligomerization. Through TNF receptor-associated factor 3 (TRAF3) or TRAF6, this leads to activation of the kinase complexes containing TBK1 or IKKε and IKKα/β/γ. Through several final phosphorylation steps, these kinases ultimately elicit antiviral and pro-inflammatory responses through interferon regulatory factor 3 (IRF3) and nuclear factor κB (NF-κB), respectively (7, 8) .

Influenza A virus belongs to the orthomyxovirus family, containing eight negative-sense RNA segments in an enveloped viral particle encoding 14 or 17 proteins (9) . This array of proteins contributes to virulence; including the proteins associated with viral RNA-dependent RNA polymerase (10) and the nonstructural protein 1 (NS1). NS1 consists of 215-237 amino acids and comprises two functional domains: an N-terminal RNAbinding domain (RBD) (AA1 to 73) and a C-terminal effector domain (ED) (AA74-end) (11) . The NS1 protein plays a crucial role in regulating the host antiviral response through various mechanisms. One important function of the NS1 protein involves inhibition of IFN production. The mechanism of this inhibition includes activation of the transcription factors IRF3 (12) , NF-κB (13) , and AP-1 (14) , thus blocking IFN production. This efficient inhibitory effect is associated with an RIG-I signaling pathway through the NS1-RIG-I complex (15) (16) (17) . Previous studies have indicated that NS1 is also related to two positive factors of RIG-I, the E3 ligases TRIM25 (18) and RIPLET (19) . The residues E96/ E97 of NS1 mediate their interaction with the coiled-coil domain of TRIM25, thus blocking both TRIM25 multimerization and RIG-I CARD domain ubiquitination. This subsequently induces lower levels of IFN-β (18) . NS1 can also interact with RIPLET preventing the activation of RIG-I, although E96/E97 are not involved in that inhibition (19) . The dsRNA binding ability of NS1 could also be playing a role in the pre-transcriptional inhibition of the interferon pathway by sequestering the pathogenassociated molecular patterns (PAMPs) that RIG-I recognizes. Two residues, R38 and K41, are required for the dsRNA binding activity of NS1 (20) , thus highly impairing its ability to block interferon production.

In another similar pathway, NS1 has been shown to inhibit host mRNA synthesis by binding a cellular 3′ end-processing factor, the 30 kDa subunit of the cleavage and polyadenylation specificity factor (CPSF30), thus attenuating type I interferon (IFN-α/β) and other interferon stimulated gene (ISG) mRNAs that are involved in the antiviral response (21) . The NS1 proteins encoded by the seasonal H1N1, H2N2, H3N2, and avian H5N1 viral subtypes strongly bind to CPSF30 (22) , whereas PR8, 2009 pandemic H1N1, and novel H7N9 virus do not efficiently bind CPSF30 (23) . It is noteworthy that cells infected with viruses expressing NS1 proteins in seasonal H3N2 and H2N2 viruses do not inhibit IRF3 activation. However, activation is blocked in cells infected with viruses expressing NS1 proteins in some, but not all, seasonal H1N1 viruses, 2009 pandemic H1N1, and avian H5N1 viruses. TRIM25 was previously reported to interact with each of these NS1 proteins, whether or not they block IRF3 activation, indicating that binding of TRIM25 by the NS1 protein does not necessarily lead to blocking of IRF3 activation (22) . Hence, binding of the NS1 protein to dsRNA, RIG-I, and TRIM25 has not established that these NS1 interactions are responsible for inhibiting the activation of IRF3 and IFN transcription. In this case, one or more host factors may participate in the NS1 blocking of IRF3 activation.

In view of several yet undetermined roles of NS1 in the inhibition of interferon, we conducted a study aimed to determine the importance of the function of the N-and/or C-terminal domains of the NS1 protein in immune evasion. By using the luciferase reporter assay, we were able to demonstrate that the C-terminal ED (AA126 to 225, named NS1/126-225) of the NS1 protein was sufficient to inhibit the production of IFN-β driven by RIG-I(N). Mechanistically, NS1/126-225 was found specifically to interact with TRAF3, to dissociate MAVS-TRAF3 complex, and to decrease K63-linked polyubiquitination of TRAF3. This was shown to result in reduced IRF3-dependent production of IFN-β, with subsequent enhancement of virus replication. These data reveal a novel mechanism for how the influenza A virus NS1 protein induces inhibition of the host IFN production and may provide a potential target for antiviral drug development.

The HPAI H5N1 virus strain, A/duck/Hubei/hangmei01/2006 (H5N1; designated H5N1/HM) was isolated from a duck. Influenza A virus (strain A/Puerto Rico/8/1934 H1N1), A/ PR/8/34, was grown in our laboratories and stored until use. rNS1-SD30 was constructed as previously described (24) . Influenza virus stocks of H5N1/HM, PR8, and rNS1-SD30 strains were amplified using 10-day-old embryonic chicken eggs and then titrated by determining log10 TCID50/ml values in MDCK cells. All cell experiments with H5N1 virus were performed in an Animal Biosafety Level 3 laboratory (BSL-3). This study was carried out in accordance with the recommendations of BSL-3, Huazhong Agricultural University (HZAU). The protocol was approved by the BSL-3 of HZAU. The recombinant vesicular stomatitis virus (VSV) encoding green fluorescence protein (VSV-GFP) was a gift from the Harbin Veterinary Research Institute (Harbin, China). Sendai virus (Sev) was grown in 10-day-old embryonic chicken eggs and titrated using a hemagglutination assay as previously described (25) .

Human embryonic kidney 293T cells and HeLa cells were purchased from ATCC (Manassas, VA, USA) and cultured at 37°C with 5% CO2 in Roswell Park Memorial Institute-1640 medium (HyClone, China) supplemented with 10% fetal bovine serum (FBS) (PAN-Biotech, Germany), containing 100 U/ml penicillin, and 100 mg/ml streptomycin (GNM15140 Cell lysates and the immunoprecipitates were resolved by 10-12% SDS-PAGE and transferred to pure nitrocellulose membranes (GE). The membranes were blocked in 1% bovine serum albumin (BSA) in TBST buffer for 1 h at room temperature and probed with indicated primary antibodies for 1-2 h at room temperature. After hybridizing with either goat anti-rabbit or goat anti-mouse secondary antibodies at a dilution of 1:10,000 in TBST buffer, the membranes were washed with TBST buffer for four times (10 min each) before visualized with ECL reagents (Advansta).

HeLa cells were plated onto coverslips in 24-well plates and transfected with the indicated plasmids. At 24 h post transfection, cells were washed once with phosphate-buffered saline (PBS) and fixed in 4% paraformaldehyde for 15 min. Cells were permeabilized with 0.1% Triton X-100 for 15 min and blocked for 1 h at room temperature with 1% BSA in PBS, followed by incubation with primary antibody for 1 h. After three washes with PBS containing 0.1% Tween 20, cells were incubated with FITC or Cy3-conjugated secondary antibodies for 1 h at room temperature and then incubated with 4′,6-DAPI for 10 min. Finally, the coverslips were washed extensively and fixed onto slides. Images were taken under a Zeiss LSM510 Meta confocal microscope (Carl Zeiss, Zena, Germany). Quantitative real-time Pcr (qrT-Pcr)

Total RNA was isolated from cells using TRIzol reagent (Invitrogen) following manufacturer's instructions and cDNA was prepared by using avian myeloblastosis virus reverse transcriptase (TaKaRa). cDNA was used for quantification of the indicated mRNA copy number on an ABI ViiA 7 PCR system (Applied Biosystems, USA) by using SYBR Green Master Mix (Rox). To detect and validate the specific amplification of PCR products, dissociation curve analysis of the products was conducted at the end of each PCR. Transcript levels of each gene were normalized with the expression of β-actin, and the 2 −∆∆C t method was used to analyze gene expression in the samples (26) . The primers used in qRT-PCR are listed in Table 1 .

Three siRNA oligonucleotides against TRAF3 and the corresponding negative control siRNA were obtained from GenePharma. Sequences are as follows: si-1, 5′-CCACUGGAGAG AUGAAUAU-3′; si-2, 5′-GUUGUGCAGAGCAGUUAAU-3′; and si-3, 5′-CUGGUUACUUUGGCUAUAA-3′. Transfection of siRNA into 293T cells was performed by Lipofectamine 2000 according to manufacturer's instructions.

293T cells were seeded into 60-mm dishes and transiently transfected with the indicated plasmids. Thirty-six hours after transfection, cells were harvested and the lysates were prepared in a 1% NP-40 lysis buffer supplemented with a commercially available 0.1% protease inhibitor cocktail and a 10 mM deubiquitinase inhibitor N-ethylmaleimide (Sigma-Aldrich). Samples were immunoprecipitated with 1 µg anti-Flag antibodies along with 30 µl Protein A/G PLUS-Agarose. Polyubiquitination was detected using anti-HA antibodies.

influenza a Virus infection of a549 cells A549 cells were transfected with the indicated plasmids for 24 h at 37°C. Cells were washed twice with F12 medium and then infected with H5N1/HM or PR8 at an indicated MOI (2 or 0.001, 

The results are expressed as means ± SD. Statistical analyses were performed on data from triplicate experiments by using twotailed Student's t-test. A P-value of less than 0.05 was considered significant and a P-value of less than 0.01 was considered highly significant.

The h5n1 ns1 Protein inhibits the rigi(n)-Mediated activation of iFn-β via its c-Terminal eD in an rna Bindingindependent Manner

With the aim of elucidating the mechanism by which IAV NS1 protein counteracts the host innate immune responses, we generated four truncated H5N1 NS1 proteins, NS1/1-73, NS1/74-225, NS1/1-125, and NS1/126-225 ( Figure 1A) . In order to investigate the function of wtNS1, and its truncated peptides, we assessed its effect on IFN-β promoter activity using a luciferase reporter assay in 293T cells. Our results showed that wtNS1, NS1/74-225, and NS1/126-225 significantly decreased the IFN-β reporter activities driven by RIG-I or RIG-I(N). Conversely, NS1/1-73 did not change the activity of IFN-β reporter, and NS1/1-125 only slightly increased the activity compared to an empty vector control ( Figure 1B) . In addition, we found that wtNS1 and all truncated peptides had inhibitory effects on IFN-β reporter activities induced by Sev or rNS1-SD30 virus infection or transfection of poly(I:C) ( Figure 1B) . This suggests that NS1 N-terminal RBD alone is sufficient to inhibit the activation of IFN-β only in the presence of dsRNA; the C-terminal ED of NS1 could inhibit the activity of IFN-β reporter in all tested conditions. Driven by RIG-I(N), NS1/126-225 caused a dose-dependent inhibition of IFN-β promoter activity and IFN-β transcription ( Figure 1C) . Previous studies indicated that IAV NS1 sequesters dsRNA and binds RIG-I at its RBD, subsequently inhibiting the activation of IRF3 and preventing the induction of IFN-β (11, 16) . Our findings reveal that C-terminal ED of NS1 (NS1/126-225) blocked RIG-I(N)-mediated IFN-β induction in an RNA binding-independent manner.

Transcription factor IRF3 is a key innate immune system component that mediates IFN-β induction. Once IRF3 is phosphorylated, it forms a dimer, translocates into the nucleus from the cytoplasm, and induces the expression of IFN-β and ISGs through specifically binding to their promoter regions (27) . In order to further investigate how NS1/126-225 inhibits the signaling that mediates type I IFN production, we used an IRF3-luciferase reporter plasmid, allowing for the measurement of IRF3 activation. As shown in Figure 2A , IRF3-luciferase reporter activation by RIG-I(N) was blocked in 293T cells overexpressing NS1/126-225 or wtNS1. We next addressed whether NS1/126-225 affected the dimerization and nuclear localization of endogenous IRF3 mediated by RIG-I(N). Non-reduced SDS-PAGE and immunoblot analysis of cell lysates after RIG-I(N) transfection showed that NS1/126-225 or wtNS1 expression produced a considerable reduction in the activated dimer form of IRF3 (Figure 2B) . Similarly, RIG-I(N)induced phosphorylation of IRF3 was strongly repressed by NS1/126-225 or wtNS1 and that phosphorylated IRF3 was largely distributed in nuclear fractions ( Figure 2C) . ELISA assays of IFN-β in the medium of cells transfected with plasmids encoding NS1/126-225 or wtNS1 along with RIG-I(N) showed that both NS1/126-225 and wtNS1 inhibited the production of IFN-β ( Figure 2D) . Together, these data indicate that NS1/126-225 inhibits the expression of type I IFN induced by RIG-I(N) through blocking the phosphorylation of IRF3. Previous studies have shown that a recombinant VSV-GFP system can be used as a strategy to screen proteins possessing IFN-antagonizing activity (28) . In the present study, we employed recombinant VSV-GFP to investigate if NS1/126-225 serves as an antagonist of IFN production. When 293T cells expressed NS1/126-225, a high level of VSV-GFP replication was present, consistent with wtNS1 protein (Figure 2E) , suggesting that the inhibitory effect of NS1/126-225 on IFN production is also present during actual viral infection.

In order to test the effect of NS1/126-225 on various components of the RLR pathway, NS1/126-225 and expression (Figure 3A) . By contrast, wtNS1 significantly decreased the RLR adaptor-mediated IFN-β promoter activity. As expected, NS1/126-225 exhibited a decrease in the IFN-β mRNA level induced by RIG-I(N) or MAVS ( Figure 3A ). In addition, the secretion of IFN-β and the transcription level of ISGs triggered by TBK1 or IRF3(5D) in the presence of NS1/126-225 were tested. NS1/126-225 induced nearly a complete loss of the inhibition of IFN-β and ISGs, including OASL, PKR, and Mx1, whereas wtNS1 strongly blocked the production of IFN-β and ISGs mRNA expression (Figures 3B,C) . Together, these data indicate that NS1/126-225 significantly inhibits the cellular antiviral response at the level between MAVS and TBK1. In the RLR-mediated signaling pathway, TRAF3 serves as a critical link between the adaptor MAVS and downstream regulatory kinases that are essential for IRF3 activation (29, 30) . Thus, we hypothesized that TRAF3 is the target of NS1/126-225. Co-IP experiments revealed that NS1/126-225 interacted selectively with TRAF3 but not other components (Figures 4A,B) . In another experiment, wtNS1 and NS1/74-225 also interacted most potently with TRAF3 ( Figure 4C ). This association was confirmed under physiological conditions in an experiment that detected this interaction by overexpression of Flag-TRAF3 in infected A549 cells, where RIG-I served as a positive control ( Figure 4D) . Furthermore, PR8 NS1 also interacted with TRAF3 in infected A549 cells (Figure 4E ). To address whether the wtNS1 protein physically interacts with endogenous TRAF3, we performed endogenous IP assays on H5N1/HM or PR8-infected cell lysates. Our results showed that TRAF3 could be co-precipitated by the NS1 antibody ( Figure 4F ). In addition, the NS1 proteins of the strain A/Shanghai/02/2013(H7N9) and other avian The promoter activities were detected by the dual-luciferase assay system. All luciferase assays were repeated at least three times, and the data shown are mean ± SD from one representative experiment. Significance was analyzed with a two-tailed Student's t-test (*P < 0.05 or **P < 0.01, ***P < 0.001). H9N2 strain did not bind TRAF3 ( Figure S1 in Supplementary Material). These results demonstrate that NS1/126-225 or wtNS1 interacts with TRAF3 in a strain-specific manner. Based on the findings that NS1/126-225 or wtNS1 interacted with TRAF3, we next asked whether the two molecules co-localize in cells. Confocal microscopy revealed the co-staining of NS1/ 126-225 or wtNS1 and TRAF3 in cells, suggesting the co-localization of the two proteins ( Figure 4G) . Normally, expression of wtNS1 in HeLa cells resulted in a nuclear localization with minor cytoplasmic staining. TRAF3 overexpression led to a marked increase in the cytoplasmic localization of wtNS1 or NS1/126-225. These results indicate that NS1/126-225 or wtNS1 colocalize with TRAF3 in cells and TRAF3 overexpression led to a marked increase of NS1/126-225 or wtNS1 cytoplasmic localization.

TraF3 is essential for ns1/126-225 to Downregulate iFn-β

The above data showed that NS1/126-225 blocks IFN-β induction and interacts with TRAF3. Hence, we used TRAF3 siRNA to determine whether TRAF3 is essential for the regulatory function of NS1/126-225 in 293T cells. Knockdown of TRAF3 by siRNA diminished IFN-β promoter activity triggered by RIG-I(N) (Figures 5A,B) . In cells with silenced TRAF3, after transfection of RIG-I(N), the induction of IFN-β was greatly reduced and the inhibitory effect of NS1/126-225 was markedly attenuated. Nevertheless, this inhibitory effect was rescued successfully with a TRAF3 expression plasmid ( Figure 5C) . These data suggest that TRAF3 is necessary for NS1/126-225 to decrease IFN-β activity.

It is well-known that TRAF3 is modified with a polyubiquitin chain to provide a scaffold for complex formation. Thus, in order to study the effect of NS1/126-225 on TRAF3 ubiquitination, Co-IP experiments were conducted. Flag-TRAF3, pUb-HA, and NS1/126-225 along with RIG-I(N) were cotransfected into 293T cells and the ubiquitination level of TRAF3 was monitored. Compared to the control, NS1/126-225 reduced the RIG-I(N)-induced ubiquitination of TRAF3 (Figure 6, left) . To explore the type of TRAF3 ubiquitin chains, Flag-TRAF3 was transfected into 293T cells with ubiquitin mutants, including the pUb-K48-HA or pUb-K63-HA expression plasmid. The results showed that NS1/126-225 could decrease the K63-linked ubiquitination of TRAF3 but not the K48-linked ubiquitination of TRAF3 (Figure 6, right) . In summary, these results demonstrate that NS1/126-225 suppresses the K63-linked ubiquitination of TRAF3 that is important for the recruitment of the TBK1-IKKε kinase complex. expressing AA1 to 346 and containing the RING domain, the Zinc-fingers domain, the Isoleucine Zipper domain, and Flag-TRAF3-TD, expressing AA347 to 568 and containing the TRAF domain ( Figure 7A ). Co-IP assays showed that the TRAF domain of TRAF3 was found to be the crucial region responsible for the association of TRAF3 with NS1/126-225 ( Figure 7B ) as well as with wtNS1 (data not shown).

Previously, it has been reported that the TIM domain of MAVS interacts with amino acid residues Y440 and Q442 within the TRAF domain of TRAF3 (31) . To examine whether NS1/126-225 affected IFN signaling at the level of MAVS-TRAF3 interaction, MAVS-TRAF3 association was determined in the presence of NS1/126-225. Expression of MAVS led to an interaction with TRAF3 and was increased by RIG-I(N) transfection. However, NS1/126-225 and wtNS1 markedly disrupted this interaction by 4.6-and 8.5-fold, respectively ( Figure 7C) . In H5N1/HM infection of A549 cells, the MAVS-TRAF3 complex was decreased by 3.3-fold compared to the control ( Figure 7D) .

As reported previously, IKKε is recruited to the C-terminal region of MAVS following Sev or VSV infection, mediated by Lys63-linked polyubiquitination of MAVS at Lys500, resulting in the inhibition of downstream IFN signaling (31, 32) . Therefore, it was necessary to test whether wtNS1 or NS1/126-225 affect this process to accomplish its negative regulatory role in IFN-β production after Sev infection. The interaction of MAVS and IKKε was readily detected by Co-IP, while Sev infection resulted in a significant decrease in the MAVS-IKKε interaction. Interestingly, the interaction of MAVS and IKKε was remarkably increased when cotransfected with NS1/126-225 or wtNS1 (Figure 7E) , indicating that NS1/126-225 or NS1 promotes the recruitment of IKKε to MAVS. Furthermore, we measured the secretion of IFN-β in cell supernatants. Results showed that NS1/126-225 or NS1 inhibited the production of IFN-β mediated by MAVS-TRAF3 or MAVS-IKKε (Figure 7F) , implying that NS1/126-225 functions in downregulating IFN expression. Taken together, these results indicate that the association of the MAVS-TRAF3 complex is disrupted by NS1/126-225, which, in turn, blocks IFN-β production.

To determine whether the replication of IAV is enhanced by the NS1/126-225 protein, A549 cells were transfected with NS1/126-225, or an empty vector, then infected with different titers of H5N1/HM or PR8 virus. Upon infection with H5N1/HM or PR8 virus, NS1/126-225 could facilitate transcription of NP of both viruses (Figure 8A) , resulting in an increase in NP and HA proteins observed in the NS1/126-225 group (Figure 8B ). This was confirmed by the titers of H5N1/HM or PR8 virus, which significantly increased by 40-and 31-fold, respectively, in NS1/126-225 overexpressing cells compared with control cells (Figure 8C) . These results demonstrate that NS1/126-225 enhances the capacity of IAV to replicate in cells.

PRRs of host cells recognize a PAMP and subsequently initiate a series of signaling cascades. The final step is activation of IRF3 and NF-κB, thus inducing the transcription of IFN-β (33) . Given the cascade of responses triggered by the host in response to infection, influenza viruses adapted different strategies to escape the IFN response. This survival tactic has proven successful in order for virus proliferation and infection. Both PB2 and PB1-F2 limit IFN production by associating with MAVS (10, 34, 35) . Other structural proteins, such as PB1, PA, NP, and even the genomic RNA itself, also contribute to impairing RIG-I-mediated antiviral responses (36) . Moreover, HA (HA1) was recently shown to drive the degradation of the IFN receptor chain IFNAR1, thereby suppressing IFN-triggered JAK/STAT signaling (37) . The most effective weapon influenza A viruses have at their disposal is NS1 protein.

The NS1 protein acts as an antiviral antagonist protein capable of limiting IFN production. RIG-I recognizes and binds dsRNA structures with 5′-triphosphates upon infection to initiate the host antiviral response. During the course of viral infection, the NS1 protein of IAV inhibits host IFN responses either by sequestering viral dsRNA or by binding to RIG-I and TRIM25 or RIPLET proteins required for RIG-I activation and IFN signaling pathways (11, 16, 18, 19) . In this study, we found that the NS1 C-terminal ED (AA126 to 225) of H5N1 virus inhibits the activation of IFN-β pathway. To achieve a negative regulatory function in the cellular antiviral response, NS1/126-225 associates with TRAF3 to remove the Lys63-polyubiquitin chains on TRAF3 and to disrupt the MAVS-TRAF3 complex. NS1/126-225 also increases the recruitment of IKKε to MAVS, releasing TRAF3 from the mitochondria. This further decreases the level of K63linked ubiquitination of TRAF3, impairing IRF3 phosphorylation and reducing the production of IFN-β (Figure 9) .

Interestingly, our study has shown that the ED of NS1 protein possesses the ability to suppress IFN response in the absence of RNA. Typically, the RBD of NS1 mediates the inhibition of IFN synthesis, and the ED of NS1 induces the inhibition of gene expression, together with its known interactors (11) . In this study, we have found that the NS1 C-terminal ED (NS1/74-225 and NS1/126-225), but not the RBD (NS1/1-73 and NS1/1-125) block IFN-β reporter activity induced by RIG-I(N). Expression of NS1/126-225 resulted in the inhibition of IRF3 activation, indicating that the NS1 protein blocks IFN-β activation through an RNA-independent manner. It was demonstrated in previous studies that influenza A viruses TX/98 and A/Viet Nam/1203/04 expressing C-terminally truncated NS1 proteins of 73, 99, or 126 amino acids were attenuated. The resultant reduced growth correlated with a high level of IFN-α/β induced by these mutant viruses (28, 38) . In addition, in both influenza B and C viruses, the C-terminal domains of the NS1 proteins were found to possess IFN antagonist activity (39, 40) . More importantly, the N terminus-truncated NS1 proteins encoded by PR8, which was To elucidate the mechanism of how the NS1 ED inhibits IFN-β activation, we speculated that the NS1 ED contacts its counterparts in RIG-I signaling leading to inhibition. For this purpose, we examined a step within the signaling pathway that NS1/126-225 targets and found that NS1/126-225 acted downstream of MAVS and upstream of TBK1. The Co-IP assays showed unexpectedly that NS1/126-225 binds to TRAF3, which interacts with MAVS forming a platform for RNA virus signaling. We also tested the binding of TRAF3 to the full-length NS1 protein and found that this interaction exists, and co-localized in the cytoplasm. The NS1/126-225 protein mainly localized in the cytoplasm. TRAF3 expression led to a marked increase in the NS1/126-225 cytoplasmic localization, suggesting that NS1/126-225 inhibits the activation of the IFN-β pathway. Although all types of influenza virus NS1 proteins interact with TRIM25, only part of NS1 prevents IRF3 activation, indicating that TRIM25 is not required for the inhibition of IRF3 activation. In this study, we did not observe the interaction between NS1/126-225 and RIG-I, consistent with the results described in a previous publication (41) . Consequently, RIG-I seems to be non-essential for the optimal inhibition of IFN production in IAV-infected cells. However, our study demonstrated that the influenza A virus NS1 ED targets TRAF3, subsequently inhibits IFN production, implying that TRAF3 is a key factor involved for IAV to escape host innate immune responses.

The MAVS-TRAF3 complex is a focal point of RLR-directed signaling response (42, 43) . TRAF3 localizes to the endoplasmic reticulum (ER) and needs to be recruited to mitochondrial MAVS in order to activate TBK1 complexes (44) . Many viral proteins, accessory and non-structural proteins in particular, hijack TRAF3 or the TRAF3 complex to mediate immune evasion. SARS coronavirus M protein or Open Reading Frame-9b prevents the formation of TRAF3-TANK-TBK1/IKKε complex or MAVS-TRAF3/TRAF6 signalosome to evade host innate immunity (45, 46) . SARS-CoV papain-like protease interacts with and disrupts STING-TRAF3-TBK1 complex, it also inhibits the TLR7-mediated innate immunity through removing Lys63linked ubiquitin chains of TRAF3 and TRAF6 (47, 48) . Herpes simplex virus 1 ubiquitin-specific protease UL36 deubiquitinates TRAF3 then counteracts the IFN-β pathway (49) . Over the past 10 years, there have been major advances in understanding how influenza A viruses successfully escape the surveillance of the immune system. The current report furthers this research revealing the surprising finding that NS1/126-225 acts by targeting TRAF3; specifically, NS1/126-225 targets the TRAF domain of TRAF3. TRAF3 links the upstream IFN signaling responses of MAVS to TBK1 relying on the TRAF domain. This report also shows that a specific interaction between TRAF3 and MAVS was observed when TRAF3 and MAVS were co-expressed in 293T cells. However, the interaction between TRAF3 and MAVS was disrupted in the presence of NS1/126-225. Interestingly, the interaction between MAVS and IKKε was markedly increased in NS1/126-225-expressing cells. It has been previously demonstrated that, after Sev infection, K63-linked polyubiquitination at Lys500 of MAVS recruits IKKε to the mitochondria, functionally causes release of TRAF3 from MAVS initiating the signal to shutdown the IFN response (31) . The MAVS-IKKε complex was enhanced when NS1/126-225 was present, indicating that NS1/126-225 can utilize this process to shut down further activation of IFN pathway. Taken together, these data indicate that NS1/126-225 impedes the interactions between components of MAVS-TRAF3 complex, preventing the phosphorylation of IRF3, where it would activate the IFN-β response.

Ubiquitination has emerged as a key posttranslational modification that controls induction and shutdown of the interferon response. TRAF3, serving as a crucial functional link, is modified with a polyubiquitin chain providing a scaffold for complex formation, and, not surprisingly, many viruses encode proteins that inhibit ubiquitination processes to overcome host innate responses. Previous studies showed that nairoviruses and arteriviruses encode for ovarian tumor domain-containing proteases that hydrolyze ubiquitin chains from host proteins (50, 51) . In this report, we have shown that NS1/126-225 suppresses the K63linked ubiquitination of TRAF3. It is likely that NS1/126-225 works through recruiting a deubiquitinase to cleave the TRAF3 ubiquitin chain since it has been shown that NS1/126-225 does not belong to any known deubiquitinase family. For example, DUBA, a member of the Otubain (OTUB) family, has been shown to deubiquitinate TRAF3 and negatively regulate TLR3-and RIG-I/ MDA5-mediated IFN induction (52) . It was also shown that two OTUB deubiquitinating enzyme family members, OTUB1 and OTUB2, can deubiquitinate TRAF3 and TRAF6, leading to the inhibition of virus-induced IFN-β expression and cellular antiviral responses (53) . Therefore, whether these proteins, or other DUB proteins, are involved in this regulation, and the detailed regulatory mechanism of TRAF3 activity triggered by NS1/126-225 remains to be discovered.

Interestingly, strain-specific targeting of TRAF3 was demonstrated by specific interaction of NS1 proteins encoded by PR8 or avian H5N1 but not novel H7N9 or avian H9N2 viruses. This difference may be associated with strain-specific sequence variations. The NS1 protein most often occurs as a 230 residue peptide, including NS1 of seasonal H1N1 virus and avian H5N1 virus (80-84 residues have been deleted since 2000), which were used in this study. However, premature stop codons or, alternatively, suppression of the genuine stop codon (codon 231) resulted in length variations at NS1's C-terminus. Abdelwhab et al. analyzed NS1 protein sequences of all AIV subtypes in birds from 1902 to 2015 to study the prevalence and distribution of carboxyl terminal end truncation (ΔCTE). They found that NS217 proteins lacking amino acids 218-230 were the most prevalent form (88%). This truncation is prevalent in LPAIV of non-H5/H7 subtypes; particularly H9N2, H10, and H6 viruses that are known to be widespread and mostly (semi)endemic in land-based poultry (54) . Similar truncations have also been observed in swine influenza viruses, which harbor a C-terminally truncated NS1 and have also been found in human H1N1 viruses that have been in public circulation since the 2009 pandemic (55) . Hence, whether the interaction of NS1 and TRAF3 are associated with the ΔCTE requires further investigation.

In summary, the present study demonstrated that TRAF3 is a target of the C-terminal ED (AA126 to 225) of H5N1 NS1 protein, revealing a novel function of the NS1 protein in modulating host innate immunity and possibly facilitating IAV infection. The physiological significance of the NS1 ED in IAV replication and its pathological role in flu diseases warrant further investigation to probe the potential value of this molecule as a therapeutic and/or disease prevention target. 

Viral RNA detection by RIG-I-like receptors

Sensing viral invasion by RIG-I like receptors

RIG-I-mediated antiviral responses to single-stranded RNA bearing 5'-phosphates

RIG-I detects viral genomic RNA during negative-strand RNA virus infection

TRIM25 RINGfinger E3 ubiquitin ligase is essential for RIG-I-mediated antiviral activity

Riplet/RNF135, a RING finger protein, ubiquitinates RIG-I to promote interferon-beta induction during the early phase of viral infection

The roles of TLRs, RLRs and NLRs in pathogen recognition

Regulation of RIG-I-like receptor signaling by host and viral proteins

Molecular mechanisms enhancing the proteome of influenza A viruses: an overview of recently discovered proteins

The PB2 subunit of the influenza virus RNA polymerase affects virulence by interacting with the mitochondrial antiviral signaling protein and inhibiting expression of beta interferon

The multifunctional NS1 protein of influenza A viruses

Activation of interferon regulatory factor 3 is inhibited by the influenza A virus NS1 protein

Influenza A virus NS1 protein prevents activation of NF-kappaB and induction of alpha/beta interferon

The influenza A virus NS1 protein inhibits activation of Jun N-terminal kinase and AP-1 transcription factors

NS1 protein of influenza A virus inhibits the function of intracytoplasmic pathogen sensor, RIG-I

Inhibition of retinoic acid-inducible gene I-mediated induction of beta interferon by the NS1 protein of influenza A virus

IFNbeta induction by influenza A virus is mediated by RIG-I which is regulated by the viral NS1 protein

Influenza A virus NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral RNA sensor RIG-I

Species-specific inhibition of RIG-I ubiquitination and IFN induction by the influenza A virus NS1 protein

RNA binding by the novel helical domain of the influenza virus NS1 protein requires its dimer structure and a small number of specific basic amino acids

Influenza virus NS1 protein interacts with the cellular 30 kDa subunit of CPSF and inhibits 3' end formation of cellular pre-mRNAs

Influenza A virus strains that circulate in humans differ in the ability of their NS1 proteins to block the activation of IRF3 and interferon-beta transcription

A single amino acid substitution in the novel H7N9 influenza A virus NS1 protein increases CPSF30 binding and virulence

Effect on virulence and pathogenicity of H5N1 influenza A virus through truncations of NS1 eIF4GI binding domain

Multiple anti-interferon actions of the influenza A virus NS1 protein

Analysis of relative gene expression data using realtime quantitative PCR and the 2(-Delta Delta C(T)) Method

Immune signaling by RIG-I-like receptors

Mutations in the NS1 protein of swine influenza virus impair anti-interferon activity and confer attenuation in pigs

Critical role of TRAF3 in the Toll-like receptor-dependent and -independent antiviral response

TRAF3: uncovering the real but restricted role in human

A functional C-terminal TRAF3-binding site in MAVS participates in positive and negative regulation of the IFN antiviral response

Ubiquitinregulated recruitment of IkappaB kinase epsilon to the MAVS interferon signaling adapter

Pathogen recognition and innate immunity

The influenza virus protein PB1-F2 inhibits the induction of type I interferon at the level of the MAVS adaptor protein

Influenza virus protein PB1-F2 inhibits the induction of type I interferon by binding to MAVS and decreasing mitochondrial membrane potential

To conquer the host, influenza virus is packing it in: interferon-antagonistic strategies beyond NS1

Hemagglutinin of influenza A virus antagonizes type I interferon (IFN) responses by inducing degradation of type I IFN receptor 1

Live attenuated influenza viruses containing NS1 truncations as vaccine candidates against H5N1 highly pathogenic avian influenza

The N-and C-terminal domains of the NS1 protein of influenza B virus can independently inhibit IRF-3 and beta interferon promoter activation

Influenza C virus NS1 protein counteracts RIG-Imediated IFN signalling

Role of N terminus-truncated NS1 proteins of influenza A virus in inhibiting IRF3 activation

Triggering the innate antiviral response through IRF-3 activation

IKKepsilon and TBK1 are essential components of the IRF3 signaling pathway

Proteomic profiling of the TRAF3 interactome network reveals a new role for the ER-to-Golgi transport compartments in innate immunity

Severe acute respiratory syndrome coronavirus M protein inhibits type I interferon production by impeding the formation of TRAF3.TANK.TBK1/IKKepsilon complex

SARS-coronavirus open reading frame-9b suppresses innate immunity by targeting mitochondria and the MAVS/TRAF3/TRAF6 signalosome

SARS coronavirus papain-like protease inhibits the type I interferon signaling pathway through interaction with the STING-TRAF3-TBK1 complex

SARS coronavirus papain-like protease inhibits the TLR7 signaling pathway through removing Lys63-linked polyubiquitination of TRAF3 and TRAF6

Herpes simplex virus 1 ubiquitin-specific protease UL36 inhibits beta interferon production by deubiquitinating TRAF3

Viral evasion mechanisms of early antiviral responses involving regulation of ubiquitin pathways

Viral OTU deubiquitinases: a structural and functional comparison

DUBA: a deubiquitinase that regulates type I interferon production

Regulation of virustriggered signaling by OTUB1-and OTUB2-mediated deubiquitination of TRAF3 and TRAF6

Prevalence of the C-terminal truncations of NS1 in avian influenza A viruses and effect on virulence and replication of a highly pathogenic H7N1 virus in chickens

Stop-codon variations in non-structural protein NS1 of avian influenza viruses

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.The reviewer, BN, and handling editor declared their shared affiliation, and the handling editor states that the process nevertheless met the standards of a fair and objective review.

No use, distribution or reproduction is permitted which does not comply with these terms.