Japanese Encephalitis Virus NS5 Inhibits Type I Interferon (IFN) Production by Blocking the Nuclear Translocation of IFN Regulatory Factor 3 and NF-κB

ABSTRACT

The type I interferon (IFN) response is part of the first-line defense against viral infection. To initiate replication, viruses have developed powerful evasion strategies to counteract host IFN responses. In the present study, we found that the Japanese encephalitis virus (JEV) NS5 protein could inhibit double-stranded RNA (dsRNA)-induced IFN-β expression in a dose-dependent manner. Our data further demonstrated that JEV NS5 suppressed the activation of the IFN transcriptional factors IFN regulatory factor 3 (IRF3) and NF-κB. However, there was no defect in the phosphorylation of IRF3 and degradation of IκB, an upstream inhibitor of NF-κB, upon NS5 expression, indicating a direct inhibition of the nuclear localization of IRF3 and NF-κB by NS5. Mechanistically, NS5 was shown to interact with the nuclear transport proteins KPNA2, KPNA3, and KPNA4, which competitively blocked the interaction of KPNA3 and KPNA4 with their cargo molecules, IRF3 and p65, a subunit of NF-κB, and thus inhibited the nuclear translocation of IRF3 and NF-κB. Furthermore, overexpression of KPNA3 and KPNA4 restored the activity of IRF3 and NF-κB and increased the production of IFN-β in NS5-expressing or JEV-infected cells. Additionally, an upregulated replication level of JEV was shown upon KPNA3 or KPNA4 overexpression. These results suggest that JEV NS5 inhibits the induction of type I IFN by targeting KPNA3 and KPNA4.

IMPORTANCE JEV is the major cause of viral encephalitis in South and Southeast Asia, with high mortality. However, the molecular mechanisms contributing to the severe pathogenesis are poorly understood. The ability of JEV to counteract the host innate immune response is potentially one of the mechanisms responsible for JEV virulence. Here we demonstrate the ability of JEV NS5 to interfere with the dsRNA-induced nuclear translocation of IRF3 and NF-κB by competitively inhibiting the interaction of IRF3 and NF-κB with nuclear transport proteins. Via this mechanism, JEV NS5 suppresses the induction of type I IFN and the antiviral response in host cells. These findings reveal a novel strategy for JEV to escape the host innate immune response and provide new insights into the pathogenesis of JEV.

INTRODUCTION

Japanese encephalitis virus (JEV) is a mosquito-borne flavivirus in the family Flaviviridae, which includes several other important human pathogens, such as Zika virus (ZIKV), West Nile virus (WNV), dengue virus (DENV), yellow fever virus (YFV), tick-borne encephalitis virus (TBEV), and hepatitis C virus (HCV) (1). The JEV genome comprises a single-stranded, positive-sense RNA that encodes a single polyprotein that is cleaved by both host and viral proteases to produce three structural (C, prM-M, and E) and seven nonstructural (NS1, -2A, -2B, -3, -4A, -4B, and -5) proteins. The core, M, and E proteins are components of mature, extracellular virus particles. NS proteins are not incorporated into particles and are thought to be involved in viral replication, which occurs in close association with endoplasmic reticulum-derived membranes (2). JEV is the major cause of viral encephalitis in South and Southeast Asia. It leads to more than 50,000 cases, with 20 to 30% mortality annually (3). However, the molecular mechanisms contributing to the severe pathogenesis are poorly understood. The ability of JEV to counteract the host innate immune response may be one of the mechanisms responsible for JEV virulence.

Type I interferons (IFN-α/β) are the earliest innate immune mediators against viral infection (4, 5). Recognition of viral components by membrane-associated and/or cytosolic pattern recognition receptors (PRRs) triggers type I IFN production in infected cells (6). As reported in our previous study, both RIG-I and Toll-like receptor 3 (TLR3), which recognize double-stranded RNA (dsRNA), have been demonstrated to mediate the production of type I IFN during JEV infection (7). For other flaviviruses, MDA5 and TLR7 have also been implicated in infection (8–10). After viral recognition, activation of the transcription factors IFN regulatory factor 3 (IRF3), IRF7, NF-κB, and activating transcription factor 2 (ATF-2)/c-Jun is induced by different pathways, leading to the expression of IFN-α/β (11). Binding of secreted IFN-α/β to the IFN receptor triggers the activation of Jak1 and Tyk2 through tyrosine phosphorylation, which stimulates the phosphorylation of STAT1 and STAT2 (12). This in turn results in the stimulation of hundreds of promoters containing IFN-α/β-stimulated regulatory elements (ISREs), thus driving the expression of the wide variety of interferon-stimulated genes (ISGs) that are responsible for establishing the antiviral state within the cell (13, 14).

To establish infection and replication in their hosts, viruses have developed powerful strategies to counteract IFN induction and IFN-stimulated antiviral responses. In the case of WNV infection, the NS1 protein was found to inhibit TLR3-induced transcriptional activation of IFN-β through a failure of nuclear translocation of IRF3 and NF-κB (15). Studies with Kunjin virus (KUN), a less pathogenic lineage I WNV variant, identified NS2A as an inhibitor of IFN-β gene transcription (16, 17). Nevertheless, the exact molecular mechanism of inhibition remains unknown. Alternatively, the WNV E protein may block the production of IFN-β by interfering with ubiquitination of RIP-1, a signaling kinase common to both the RIG-I and TLR3 pathways (18). In addition, downstream signaling of type I IFN is also targeted by viruses. For KUN, five of the seven nonstructural proteins (NS2A, NS2B, NS3, NS4A, and NS4B) were shown to inhibit STAT translocation to the nucleus and subsequent IFN-dependent reporter activity when they were expressed individually in plasmid-transfected cells (19). Similarly, DENV NS2A, NS4A, or NS4B was shown to impair the JAK/STAT signaling pathway by reducing the phosphorylation and nuclear translocation of STAT1 (20). Among the nonstructural proteins, NS5 of flaviviruses is known to play crucial roles in innate immune evasion; for instance, NS5 mediates STAT2 degradation and inhibits IFN-dependent signaling during DENV infection, and ZIKV NS5 was recently reported to inhibit both type I IFN production and its downstream signaling (21–24). In the case of JEV infection, the strategy for NS5 to suppress the downstream signaling of type I IFN has also been well studied. The JEV NS5 protein could reduce the phosphorylation of Tyk2 and STAT1 through a PTP-dependent mechanism, and thus could inhibit STAT1 nuclear localization (25, 26). However, the interaction of JEV NS5 with the steps of IFN induction has been less well studied.

In this study, we aimed to assess the ability of JEV NS5 to modulate the induction of type I IFN. We found that the JEV NS5 protein impaired the capability of cells to produce IFN-β via inhibition of nuclear translocation of IRF3 and NF-κB. NS5 was demonstrated to bind to nuclear transport proteins, i.e., karyopherin α3 (KPNA3; also known as importin α4) and karyopherin α4 (KPNA4; also known as importin α3), which competitively blocked the interactions of KPNA3 and KPNA4 with their cargos, namely, p65, a subunit of NF-κB, and IRF3. These findings suggest a novel strategy by which JEV subverts cellular innate immunity and evades host antiviral responses.

RESULTS

JEV NS5 suppresses poly(I·C)-induced IFN-β production.

The expression of type I IFN can be induced by various stimuli, including poly(I·C), a double-stranded homopolymer used as a model dsRNA. In order to investigate whether JEV NS5 can modulate the induction of type I IFN, increasing amounts of a plasmid expressing JEV NS5 were cotransfected with IFN-β promoter-dependent luciferase reporter constructs, followed by poly(I·C) stimulation. The cells were harvested at 12 h poststimulation, and the IFN-β promoter activity was determined by a dual-luciferase reporter assay. The results revealed that expression of NS5 suppressed the IFN-β promoter activation induced by poly(I·C) in a dose-dependent manner (Fig. 1A). To further verify this result, mRNA levels of IFN-β were determined by quantitative reverse transcription-PCR (qRT-PCR) analysis of cells expressing increasing amounts of NS5 and transfected with poly(I·C). As expected, JEV NS5 reduced the IFN-β mRNA expression induced by poly(I·C) in a dose-dependent manner (Fig. 1A). These results indicate that the JEV NS5 protein can suppress dsRNA-induced IFN-β expression. The expression level and intracellular compartmentalization of NS5 in plasmid-transfected cells were further determined and compared to those in JEV-infected cells. The expression levels of 500 ng and 1,000 ng of transfected plasmid were similar to those detected for NS5 in JEV-infected cells at 24 and 48 h postinfection (Fig. 1B), and most of the NS5 protein in both plasmid-transfected and JEV-infected cells was detected in the cytoplasm, while only some was observed in the nucleus (Fig. 1C).

FIG 1.

JEV NS5 inhibits poly(I·C)-induced IFN-β expression. (A) (Top) Increasing amounts (250 ng, 500 ng, and 1,000 ng) of plasmid encoding JEV NS5 or empty vector (Vec.) were cotransfected into HeLa cells with the IFN-β promoter-dependent reporter plasmid p125-Luc and with phRLTK for normalization. After 24 h, cells were left untreated or transfected with poly(I·C) and incubated for another 12 h. Luciferase activity was determined by a dual-luciferase assay. (Middle) HeLa cells were transfected with increasing amounts (250 ng, 500 ng, and 1,000 ng) of plasmid encoding JEV NS5 or empty vector (Vec.), followed by transfection with poly(I·C) 12 h prior to cell lysis. The mRNA level of IFN-β was determined by qRT-PCR. Expression data were normalized to the expression of β-actin. ***, P < 0.005; **, P < 0.01 (compared to cells transfected with empty vector [n = 3]). (Bottom) The protein levels of NS5 were determined by Western blotting. (B) HeLa cells were transfected with increasing amounts (250 ng, 500 ng, and 1,000 ng) of plasmid encoding JEV NS5 or infected with JEV at an MOI of 1.0. The expression levels of NS5 were determined at 36 h posttransfection and 24 h and 48 h postinfection by Western blotting. Con, control. (C) HeLa cells were transfected with pFlag-NS5 (500 ng) or infected with JEV at an MOI of 1.0. JEV NS5 was detected by an immunofluorescence assay using a monoclonal antibody against JEV NS5 (green), and nuclei were stained with DAPI (blue). The images of the cells were acquired with a fluorescence microscope (Zeiss). Magnification, ×40.

JEV NS5 interferes with the nuclear translocation of IRF3 and NF-κB induced by poly(I·C).

In order to clarify the molecular mechanism of suppression of IFN-β expression by JEV NS5, we analyzed the activity of the IFN-β transcription factors IRF3, NF-κB, and AP-1 by using a dual-luciferase reporter assay. Cells were cotransfected with increasing amounts of pFlag-NS5 and an IRF3-, NF-κB-, or AP-1-dependent luciferase reporter plasmid, followed by poly(I·C) treatment. The results of the dual-luciferase reporter assay showed that JEV NS5 reduced the expression of the IRF3- and NF-κB-dependent reporter genes induced by poly(I·C) in a dose-dependent manner but did not alter AP-1-dependent luciferase activity (Fig. 2A), suggesting that JEV NS5 inhibits the activation of IRF3 and NF-κB but not AP-1.

FIG 2.

JEV NS5 inhibits poly(I·C)-induced activation of IRF3 and NF-κB. (A) (Top) Increasing amounts (250 ng, 500 ng, and 1,000 ng) of an expression plasmid encoding JEV NS5 were cotransfected into HeLa cells with phRL-TK and pIRF3-Luc, pNF-κB-Luc, or pAP-1-Luc. Twenty-four hours later, cells were transfected with or without poly(I·C). Luciferase activities were measured 12 h after transfection. ***, P < 0.005; **, P < 0.01 (compared to cells transfected with empty vector [n = 3]). (Bottom) The protein levels of NS5 were determined by Western blotting. (B) HeLa cells were transfected with a plasmid encoding JEV NS5 (+) or with empty vector (−). Twenty-four hours later, cells were transfected with or without poly(I·C) and incubated for another 12 h, followed by lysis. Cell lysates were separated into cytoplasmic and nuclear extracts, and the protein levels of NS5, IRF3, and p65 in the whole-cell lysate (WCL), cytoplasmic extract, and nuclear extract were analyzed by Western blotting. Protein levels of IRF3 and p65 were quantified by immunoblot scanning and normalized to the amount of GAPDH or lamin A/C expression. ***, P < 0.005; **, P < 0.01 [compared to cells transfected with poly(I·C) and empty vector (n = 3)].

Since IRF3 and NF-κB are cytoplasmic proteins which migrate to the nucleus and bind to the PRDI to PRDIII sites of the IFN-β promoter to initiate IFN-β transcription upon viral infection, we next examined the possibility that JEV NS5 alters the ability of IRF3 and NF-κB to enter the nucleus. HeLa cells were transfected with a plasmid encoding Flag-NS5 or with an empty vector. At 24 h posttransfection, cells were either transfected with poly(I·C) or left untreated, followed by lysis. Western blotting was performed to analyze the levels of IRF3 and p65, a subunit of NF-κB, in whole-cell lysates and cytoplasmic and nuclear extracts. Expression of NS5 did not affect the levels of IRF3 and p65 in whole-cell lysate; however, it obviously increased the levels of IRF3 and p65 in the cytoplasmic fraction and reduced their levels in the nuclear fraction for cells treated with poly(I·C) (Fig. 2B). These observations reveal the capacity of NS5 to inhibit the nuclear localization of IRF3 and NF-κB.

JEV NS5 does not interrupt the dsRNA signaling pathway upstream of IRF3 and NF-κB.

IRF3 and NF-κB activation requires the association of numerous signaling molecules. Phosphorylation by IKBKE and TBK1 kinases is critical for IRF3 activation. This induces a conformational change leading to IRF3 dimerization and nuclear localization, while activation of NF-κB is initiated by the signal-induced degradation of IκB proteins which mask the nuclear localization signals (NLSs) of NF-κB proteins and keep them sequestered in an inactive state in the cytoplasm of unstimulated cells. To investigate if IRF3 and NF-κB were directly affected by NS5 or if an upstream signaling molecule was inhibited, thus preventing the activation of IRF3 and NF-κB, we examined the protein level of IκBα and the phosphorylation of IRF3 in the presence of NS5. Western blot results revealed that poly(I·C) induced the phosphorylation of IRF3 and the degradation of IκBα (Fig. 3A). However, expression of NS5 affected neither the level of IκBα nor the phosphorylation of IRF3, suggesting that inhibition of IRF3 and NF-κB activation was not due to alterations in upstream signaling molecules but directly due to the inhibition of IRF3 and NF-κB nuclear translocation. To further validate the step at which JEV NS5 inhibits NF-κB activation, NF-κB-dependent luciferase activity was induced by coexpression of p65 and p50. In the presence of NS5, the NF-κB activity induced by p65/p50 was reduced in a dose-dependent manner (Fig. 3B), which confirms that NS5 inhibits NF-κB nuclear localization directly.

FIG 3.

JEV NS5 does not interrupt the dsRNA signaling pathway upstream of IRF3 and NF-κB. (A) HeLa cells were transfected with a plasmid encoding JEV NS5 (+) or with empty vector (−). Twenty-four hours later, cells were transfected with or without poly(I·C). After incubation for another 12 h, cells were harvested, and the phosphorylation level of IRF3 and expression level of IκBα were analyzed by Western blotting. (B) (Top) Increasing amounts of plasmid encoding JEV NS5 or empty vector were cotransfected into HeLa cells with plasmids expressing p50 and p65 and the NF-κB-dependent reporter system (pNF-κB-Luc and phRL-TK). At 24 h posttransfection, cells were lysed and NF-κB activity was determined by a dual-luciferase assay. ***, P < 0.005; **, P < 0.01; *, P < 0.05 [compared to cells transfected with poly(I·C) and empty vector (n = 3)]. (Bottom) NS5 expression levels were analyzed by Western blotting.

JEV NS5 competitively inhibits the interaction of IRF3 and p65 with KPNA3 and KPNA4.

Nuclear-cytoplasmic transport of large macromolecules (>40 kDa) between the nucleus and the cytoplasm occurs through nuclear pore complexes (NPCs) via signal-dependent, carrier-mediated processes. Cargo proteins containing classical NLSs are generally recognized by importin α in conjunction with importin β. The cargo-importin complex translocates to the nucleus via transient interaction with the phenylalanine-glycine repeat region of nucleoporins, which are components of NPCs. It has been reported that nuclear import of both IRF3 and p65 is mediated by importin α3 (KPNA4) and importin α4 (KPNA3) (27, 28). To explore whether NS5 may interrupt the nuclear translocation of IRF3 and NF-κB by affecting the nuclear transport proteins, levels of importin α and importin β were determined for cells transfected with the NS5 plasmid or empty vector. It was shown that expression of NS5 did not significantly influence the levels of importin α (KPNA1, KPNA2, KPNA3, KPNA4, KPNA5, and KPNA6) and importin β (KPNB1 and KPNB2) in host cells (Fig. 4A).

FIG 4.

JEV NS5 inhibits the interaction of IRF3 and p65 with KPNA3 and KPNA4. (A) HeLa cells were transfected with JEV NS5 plasmid or empty vector. Twenty-four and 48 h after transfection, cells were harvested, and the expression levels of endogenous KPNA1, KPNA2, KPNA3, KPNA4, KPNA5, KPNA6, KPNB1, and KPNB2 were analyzed by Western blotting. (B) HEK293T cells were cotransfected with Flag-NS5 or Flag-NS5ΔNLS plasmid and the KPNA1-Myc, KPNA2-Myc, KPNA3-Myc, KPNA4-Myc, KPNA5-Myc, or KPNA6-Myc construct. At 36 h posttransfection, immunoprecipitation (IP) was performed with whole-cell lysates by using anti-Flag antibody. Immunoprecipitates and whole-cell lysates were analyzed by Western blotting (WB). (C) HEK293T cells were cotransfected with the KPNA3-Myc or KPNA4-Myc construct and 1.0 or 2.0 μg/ml of JEV NS5 plasmid or empty vector. The cells were harvested and lysed at 36 h posttransfection. Immunoprecipitation (IP) was performed with c-Myc antibody, and Western blotting was performed with antibodies against c-Myc, Flag, IRF3, and p65.

Some studies have demonstrated that NLS-containing viral proteins, including VP24 of Ebola virus (EBOV), ORF6 of severe acute respiratory syndrome coronavirus (SARS-CoV), and N protein of hantavirus (HTNV), allow the recruitment of importin α and thus the inhibition of nuclear import pathways (29–31). Given that two NLSs (a/bNLS and bNLS; amino acids [aa] 320 to 405) identified in NS5 of DENV are highly conserved in other members of the flavivirus family (32, 33), we next determined whether JEV NS5 utilizes the same strategy for inhibiting the nuclear transport of host proteins. First, the interactions between NS5 and importins α were examined. HEK293 cells were cotransfected with a plasmid encoding c-Myc-tagged KPNA1, KPNA2, KPNA3, KPNA4, KPNA5, or KPNA6 and a plasmid encoding Flag-tagged NS5 or an NS5 mutant lacking the NLS (Flag-NS5ΔNLS). Immunoprecipitation with anti-Flag and subsequent Western blotting with anti-Myc revealed that JEV NS5 specifically interacted with KPNA2, KPNA3, and KPNA4 but not with KPNA1, KPNA5, or KPNA6 (Fig. 4B). However, no interaction between NS5ΔNLS and KPNAs was detected, suggesting that the interaction between NS5 and importins is NLS dependent. Subsequently, we tested whether the interactions of IRF3 and p65 with importins were affected in cells expressing NS5. HEK293 cells were cotransfected with increasing amounts of the NS5 plasmid and the KPNA3 or KPNA4 plasmid, followed by poly(I·C) treatment. Cell lysates were harvested for immunoprecipitation 48 h after transfection. Western blotting revealed that smaller amounts of IRF3 and p65 interacted with KPNA3 and KPNA4 in cells expressing larger amounts of NS5 (Fig. 4C), indicating that NS5 competitively inhibited the interaction of IRF3 and p65 with KPNA3 and KPNA4.

NS5 residues K³⁹¹ and R³⁹⁴ are critical for its interactions with KPNAs.

To identify the region of NS5 interacting with KPNAs, we coexpressed Flag-tagged full-length NS5 or truncated NS5 lacking either bNLS (aa 321 to 370) or a/bNLS (aa 371 to 405) (Fig. 5A) with c-Myc-tagged KPNA2, KPNA3, or KPNA4. The results showed that KPNA proteins were coimmunoprecipitated with full-length NS5 and the mutant containing aa 371 to 405 but not with the a/bNLS-deleted mutant (Fig. 5B). To identify the crucial amino acids among NS5 residues 371 to 405 that participate in this interaction, we created a series of NS5 mutants (A to I) and screened for the ability to bind to KPNA proteins by coimmunoprecipitation (co-IP) (Fig. 5C). NS5 mutant F, with residues ³⁹¹KRPR³⁹⁴ changed to ³⁹¹AAAA³⁹⁴, showed the most obvious reduction in binding ability with KPNA proteins (Fig. 5D). NS5 mutants with each amino acid in ³⁹¹KRPR³⁹⁴ changed to A were subsequently constructed, and their interactions with KPNA proteins were further examined. Each of the mutations attenuated the interactions between NS5 and KPNAs to a certain extent compared to the interactions observed with the wild type (Fig. 5E). Among them, NS5 mutants J (K³⁹¹A) and M (R³⁹⁴A) showed the most obvious reductions of binding ability, which suggests that NS5 residues K³⁹¹ and R³⁹⁴ are critical for its interactions with KPNAs.

FIG 5.

Identification of the JEV NS5 residues which are crucial for its interaction with KPNAs. (A) Diagram of JEV NS5 truncated mutants lacking bNLS (mut.1) or a/bNLS (mut.2). (B) HEK293T cells were cotransfected with full-length NS5 (WT) plasmid, an NS5 truncated mutant construct (mut.1 or mut.2), or vector plasmid (Vec.) and the KPNA2, KPNA3, or KPNA4 construct. The cells were harvested and lysed at 36 h posttransfection. Immunoprecipitation (IP) was performed with a Flag antibody, and Western blotting was performed with antibodies against c-Myc, Flag, and GAPDH. (C) Amino acid mutations in NS5 a/bNLS. (D) HEK293T cells were cotransfected with wild-type NS5 plasmid (WT), NS5 mutant constructs (A to I), or vector plasmid (Vec.) and the KPNA2, KPNA3, or KPNA4 construct. The cells were harvested and lysed at 36 h posttransfection. Immunoprecipitation was performed with a Flag antibody, and Western blotting was performed with antibodies against c-Myc, Flag, and GAPDH. (E) HEK293T cells were cotransfected with wild-type NS5 plasmid (WT), NS5 mutant constructs (F and J to M), or vector plasmid (Vec.) and the KPNA2, KPNA3, or KPNA4 construct. The cells were harvested and lysed at 36 h posttransfection. Immunoprecipitation was performed with a Flag antibody, and Western blotting was performed with antibodies against c-Myc, Flag, and GAPDH.

Overexpression of KPNA3 or KPNA4 restores the transcriptional activity of IRF3 and NF-κB suppressed by JEV NS5.

To further establish that NS5 inhibits the activation of IRF3 and NF-κB by targeting KPNA3 and KPNA4, IRF3- and NF-κB-dependent luciferase activities were determined for cells overexpressing KPNA3 or KPNA4. As expected, overexpression of KPNA3 or KPNA4 restored the activity of IRF3 and NF-κB reduced by NS5 (Fig. 6A). We also found that NS5 mutant F (NS5-F) did not show a significant inhibitory effect on IRF3 and NF-κB activation. In addition, protein levels of IRF3 and p65 in the cytoplasm and the nucleus were analyzed by Western blotting. Reduced levels of IRF3 and p65 in the cytoplasmic fraction and increased levels of IRF3 and p65 in the nuclear fraction were observed for cells cotransfected with the KPNA3 or KPNA4 and NS5 plasmids compared to those for cells transfected with NS5 plasmid alone (Fig. 6B). To verify this result under JEV infection conditions, similar IRF3- and NF-κB-dependent luciferase assays and Western blotting were performed with JEV-infected cells. Our results demonstrated that overexpression of KPNA3 or KPNA4 accelerated the activation of IRF3 and NF-κB induced by JEV infection (Fig. 6C and D).

FIG 6.

Overexpression of KPNA3 or KPNA4 recruits IRF3 and NF-κB activity inhibited by JEV NS5. (A) A plasmid expressing NS5 or the NS5-F mutant was cotransfected into HeLa cells with vector plasmid (−) or a KPNA construct (+) and the IRF3- or NF-κB-dependent reporter system. At 24 h posttransfection, cells were left untreated or transfected with poly(I·C). The cells were collected after 12 h of incubation, and IRF3 (top) and NF-κB (middle) activities were determined by a dual-luciferase assay. ***, P < 0.005; ** P < 0.01 (n = 3). (Bottom) Protein levels of NS5, NS5-F, KPNA2, KPNA3, and KPNA4 were determined by Western blotting. (B) HeLa cells were transfected or cotransfected with a plasmid encoding JEV NS5 or the NS5-F mutant and/or a plasmid expressing KPNA2, KPNA3, or KPNA4. Twenty-four hours later, cells were transfected with poly(I·C) and incubated for another 12 h. Cells were harvested, and cytoplasmic and nuclear fractions were extracted. Protein levels of NS5, NS5-F, KPNA3, KPNA4, IRF3, and p65 in the whole-cell lysate (WCL), cytoplasmic extract, and nuclear extract were analyzed by Western blotting. Protein levels of IRF3 and p65 in poly(I·C)-treated cells were quantified by immunoblot scanning and normalized to the amount of GAPDH or lamin A/C expression. ***, P < 0.005; **, P < 0.01; *, P < 0.05 (n = 3). (C) A plasmid encoding KPNA3 or KPNA4 or empty vector (Vec.) was cotransfected into HeLa cells with the IRF3- or NF-κB-dependent reporter system. Twenty-four hours later, cells were mock infected or infected with JEV at an MOI of 1.0. After incubation for 12 h and 24 h, respectively, cells were collected, and IRF3 (top) and NF-κB (middle) activities were determined by a dual-luciferase assay. ***, P < 0.005; **, P < 0.01; *, P < 0.05 (n = 3) (compared to cells transfected with vector plasmid). (Bottom) The expression of NS5, KPNA3, and KPNA4 was determined by Western blotting. (D) HeLa cells were transfected with a plasmid encoding KPNA3 or KPNA4 or with empty vector and then infected with JEV at an MOI of 1.0 at 24 h posttransfection. After 24 h of incubation, cells were collected, and protein levels of NS5, KPNA3, KPNA4, IRF3, and p65 in the whole-cell lysate (WCL), cytoplasmic extract, and nuclear extract were analyzed by Western blotting. (E) Vero cells were cotransfected with a plasmid encoding JEV NS5 or the NS5-F mutant and/or a KPNA construct and the IRF3- or NF-κB-dependent reporter system. At 24 h posttransfection, cells were left untreated or transfected with poly(I·C). The cells were collected after 12 h of incubation, and IRF3 (top) and NF-κB (middle) activities were determined by a dual-luciferase assay. ***, P < 0.005; **, P < 0.01 (n = 3). (Bottom) Protein levels of NS5, NS5-F, KPNA2, KPNA3, and KPNA4 were determined by Western blotting. (F) Vero cells were cotransfected with a plasmid encoding KPNA3 or KPNA4 or an empty vector and the IRF3- or NF-κB-dependent reporter system. Twenty-four hours later, cells were mock infected or infected with JEV at an MOI of 1.0. After incubation for 12 h and 24 h, respectively, cells were collected, and IRF3 (top) and NF-κB (middle) activities were determined by a dual-luciferase assay. ***, P < 0.005; **, P < 0.01; *, P < 0.05 (n = 3) (compared to cells transfected with vector plasmid). (Bottom) The expression of NS5, KPNA3, and KPNA4 was determined by Western blotting.

As previous studies demonstrated the role of flavivirus NS5 in inhibiting JAK/STAT signaling upon interferon stimulation, JEV NS5 may disrupt the downstream feed-forward loop of IFN signaling in addition to the competition for importins. Therefore, to negate the potentially confounding effects, the IRF3- and NF-κB-dependent luciferase activity assay was repeated by using a Vero cell line with reported deficiencies in IFN expression, and similar results were observed for Vero cells and HeLa cells (Fig. 6E and F). These data together suggest that JEV NS5 inhibits nuclear localization of IRF3 and NF-κB by blocking their interactions with importins.

Overexpression of KPNA3 or KPNA4 increases IFN-β expression and inhibits JEV propagation.

As demonstrated above, overexpression of KPNA3 and KPNA4 restored the activity of IRF3 and NF-κB. Thus, we next tested the possibility that overexpression of KPNA3 and KPNA4 could prevent the inhibition of IFN-β production by JEV NS5. The plasmid encoding NS5 or NS5-F or an empty vector was cotransfected with the KPNA3 or KPNA4 construct, followed by poly(I·C) stimulation, and mRNA and protein levels of IFN-β were determined. It was shown that expression of NS5 significantly reduced the production of IFN-β, whereas transfection with either the NS5-F mutant or empty vector showed no inhibitory effect (Fig. 7A). Overexpression of KPNA3 or KPNA4 restored the expression of IFN-β inhibited by NS5. To confirm the effect of KPNA3 and KPNA4 overexpression on JEV-induced secretion of IFN-β, cells were transfected with an empty vector or the KPNA3 or KPNA4 plasmid, followed by JEV infection at different multiplicities of infection (MOIs) (1.0 and 2.0). As expected, overexpression of KPNA3 or KPNA4 enhanced the IFN-β production induced by JEV infection (Fig. 7B). These results further indicated that JEV NS5 inhibited the expression of IFN-β by targeting KPNA3 and KPNA4.

FIG 7.

Overexpression of KPNA3 or KPNA4 increases the production of IFN-β and inhibits JEV propagation. (A) HeLa cells were transfected or cotransfected with a plasmid encoding NS5 or NS5-F and/or the KPNA3 or KPNA4 construct. At 24 h posttransfection, cells were left untreated or transfected with poly(I·C). After 12 h, relative mRNA levels (top) and protein levels (middle) of IFN-β were determined by qRT-PCR and ELISA, respectively. N.D., not detected. ***, P < 0.005; **, P < 0.01; *, P < 0.05 (n = 3). (Bottom) The expression of NS5, NS5-F, KPNA3, and KPNA4 was determined by Western blotting. (B) HeLa cells were transfected with the KPNA3 or KPNA4 construct or vector plasmid. After 12 h, cells were mock infected or infected with JEV at an MOI of 1.0 or 2.0. The supernatants were collected at 24 h and 48 h postinfection, the concentrations of IFN-β were determined by ELISA (top row), and RNA levels and titers of JEV were determined by qRT-PCR (second row) and plaque assay (third row), respectively. ***, P < 0.005; **, P < 0.01; *, P < 0.05 (compared to cells transfected with vector plasmid [n = 3]). (Bottom) The expression of NS5, KPNA3, and KPNA4 was determined by Western blotting.

Since the induction of type I interferon is a powerful host defense mechanism against viral infection, we subsequently examined whether overexpression of KPNA3 and KPNA4 modulated the propagation of JEV. HeLa cells were transfected with an empty vector or the vector encoding KPNA3 or KPNA4, followed by JEV infection at MOIs of 1.0 and 2.0. Cells were harvested 24 and 48 h after infection, and viral RNA levels and virus titers were determined by plaque assay. An apparent reduction in viral titer was observed for cells expressing KPNA3 or KPNA4 (Fig. 7B), suggesting that overexpression of KPNA3 or KPNA4 inhibits JEV propagation, which further indicates that JEV NS5 inhibits the host antiviral response by targeting KPNA3 and KPNA4.

DISCUSSION

Proper nuclear-cytoplasmic protein transport is essential for maintaining cellular homeostasis. In the case of a viral infection, cellular transcription factors are transported into the nucleus, where they activate the transcription of antiviral genes. To escape the antiviral response of host cells, various strategies for inhibiting nuclear-cytoplasmic transport have evolved in viruses. Some viruses, such as Ebola virus (EBOV), hantavirus (HTNV), and human papillomavirus (HPV), employ the strategy of binding to importin or transportin family proteins to block the nuclear import of cellular cargo molecules (31, 34–37). Porcine reproductive and respiratory syndrome virus (PRRSV) Nsp1β blocks interferon-stimulated gene factor 3 (ISGF3) nuclear translocation by inducing importin α5 degradation (38). Vesicular stomatitis virus (VSV) inhibits bidirectional nuclear transport by interacting with NPC components or NPC-associated factors (39). The Venezuelan equine encephalitis virus (VEEV) capsid protein forms a tetrameric complex with importin α/β1 and CRM1, showing a unique inhibitory strategy against nuclear transport of host proteins (40). In this study, we demonstrated, for the first time, that JEV NS5 inhibits the nuclear import of IRF3 and NF-κB by recruitment of KPNA3 (importin α4) and KPNA4 ((importin α3), leading to suppression of IRF3 and NF-κB transcriptional activity and preventing the host antiviral response.

During the course of infection with many DNA or RNA viruses, viral dsRNA is generated during transcription and/or replication. The dsRNA is a potent intracellular signal that stimulates the defense responses of cells. Signal transduction pathways activated by dsRNA lead to activation of IRF3, NF-κB, and AP-1, which mediates the transcriptional induction of type I IFN genes (11). To evade the antiviral response of host cells, many viruses have acquired sophisticated mechanisms to interfere with IRF3, NF-κB, or AP-1 signaling (41, 42). It has been shown that JEV infection activates IRF3, NF-κB, and AP-1 signaling (7, 42) and thus elicits the secretion of type I IFN. Recently, subgenomic flavivirus RNA (sfRNA, also known as small noncoding RNA of flavivirus) of JEV was found to inhibit the phosphorylation and nuclear localization of IRF3 (43). In the present study, our data demonstrated that JEV NS5 interfered with the activation of IRF3 and NF-κB induced by poly(I·C). IRF3 is present in the cytoplasm in an inactive form but is selectively phosphorylated upon viral infection (44). Phosphorylation induces dimerization and the subsequent translocation of IRF3 from the cytoplasm into the nucleus, which results in activation of the IFN-I promoter via interaction with the transcriptional cofactor (CREB)-binding protein (CBP)/p300. NF-κB transcription factors are dimers composed of five subunits, i.e., p65 (Rel A), Rel B, c-Rel, p50, and p52, and the p50/p65 heterodimer is the best-characterized and most abundant form of NF-κB (45, 46). NF-κB is rapidly activated after exposure to pathogens, pathogen-associated molecular patterns, and various cytokines (41). In its inactive state, IκB blocks the interaction of NF-κB with the transport molecule importin α (47, 48). Once IκB is degraded, the NF-κB complex interacts with importin α and is translocated to the nucleus, where it functions as a transcription factor for several genes, including type I IFN genes. Our findings suggested that JEV NS5 did not alter the phosphorylation level of IRF3 and the degradation of IκBα mediated by poly(I·C). In addition, NS5 suppressed not only poly(I·C)-stimulated NF-κB activity but also the activity of overexpressed NF-κB (p65 and p50). These results reveal a universal downstream inhibitory mechanism for NS5. Since it has been shown that both IRF3 and NF-κB are transported into the nucleus via interactions with KPNA3 and KPNA4 (27, 28), we hypothesized that JEV NS5 may inhibit the nuclear transport of IRF3 and NF-κB by targeting KPNA3 and KPNA4. A correlation between the inhibitory capacity of NS5 and its interaction with KPNA3 and KPNA4 was confirmed by a binding competition assay (Fig. 5B). As IRF3 and NF-κB are not unique cargo molecules for KPNA3 and KPNA4, we could not determine if the interaction of JEV NS5 with KPNA3 and KPNA4 contributes only to the inhibition of IRF3 and NF-κB transport. JEV NS5 may also interfere with the nuclear transport of other host molecules through the same mechanism.

NS5 of flavivirus plays multifunctional roles in viral replication and pathogenesis, making it an interesting target for antiviral drug development (49–51). Two classical NLSs, a/bNLS and bNLS, are present within the interdomain region of DENV NS5. Importin-NLS binding mediates the nuclear translocation of DENV-2 NS5, and the nuclear NS5 appears to partly antagonize the antiviral response, including modulation of interleukin-8 production by infected cells (29, 33, 52). However, the role of the conserved NLSs within the NS5 proteins of other flaviviruses is still unclear. Here we found that residues ³⁹¹KRPR³⁹⁴ in the JEV NS5 NLS are critical for the ability of NS5 to interact with importins. This finding is supported by a previous report which proposed the consensus sequence K-K/R-X-K/R for monopartite NLSs (53). A similar mechanism was also found during HTNV infection (37). The HTNV nucleocapsid protein interacts with importin α1, importin α2, and importin α3, leading to the inhibition of p65 nuclear localization.

Type I IFN production is the key part of the innate immune response against viral infection in mammalian cells (11). To replicate and spread, viruses have evolved powerful strategies to evade or suppress the host innate immune response (54–56). JEV NS5 was reported to inhibit the IFN-induced antiviral response by blocking the Jak/Stat signaling pathway (25, 26). In the present study, we demonstrated that JEV NS5 inhibits production of IFN-β by targeting nuclear transport proteins, revealing a novel immunosuppressive strategy for JEV NS5. In addition, our present data reveal a negative regulatory role for KPNA3 and KPNA4 in JEV replication, confirming that the inhibition of nuclear-cytoplasmic protein transport is a strategy for JEV to prevent the antiviral response of host cells and correlating this strategy with the pathogenesis of JEV. However, our assay could not determine whether the enhancement of IRF3 and NF-κB activation is the only mechanism regulating JEV replication, because KPNA3 and KPNA4 may also mediate the nuclear transport of other molecules or some undiscovered host pathways responsible for antiviral responses. Note that no significant difference was observed among the effects of KPNA3 and KPNA4 on cytokine production or JEV replication. This may be due to the similar interaction efficiencies between NS5 and both importins.

In summary, our data demonstrate that JEV NS5 interferes with the nuclear translocation of IRF3 and NF-κB induced by poly(I·C) by competitively inhibiting the interaction of IRF3 and p65 with nuclear transport proteins. Via this mechanism, JEV NS5 suppresses the induction of type I IFN in host cells. These findings reveal a novel strategy for JEV to evade the host innate immune response and provide us new insight into the pathogenesis of JEV.

MATERIALS AND METHODS

Cell culture and virus propagation.

HeLa, HEK293T, Vero, and BHK-21 (baby hamster kidney) cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (Gibco), 250 μg/ml amphotericin B (Gibco), and 100 IU penicillin-streptomycin (Gibco). JEV wild-type strain P3 was propagated in suckling mouse brains and titrated on BHK-21 cells.

Plasmid construction.

To generate pFlag-NS5, the Flag tag was fused in frame to the 5′ end of the NS5 sequence by PCR, using the genomic cDNA of JEV-infected HeLa cells as a template. Truncated mutant NS5 proteins lacking NLS (pFlag-NS5ΔNLS), bNLS (mut.1), and a/bNLS (mut.2) were generated by introducing deletions of amino acids 322 to 392, 321 to 370, and 371 to 392, respectively, using overlap extension PCR. The NS5 mutants A to M were constructed by changing the indicated amino acid(s) to alanine (A) by using overlap extension PCR. Plasmids pMyc-KPNA1, pMyc-KPNA2, pMyc-KPNA3, pMyc-KPNA4, pMyc-KPNA5, and pMyc-KPNA6 were generated by cloning the open reading frame of each KPNA gene into pCMV-Tag1. Plasmids encoding p50 and p65 were generated by cloning of the coding regions into pcDNA3.1. Luciferase reporter plasmids (p125-Luc for IFN-β, pIRF3-Luc for IRF3, pNF-κB-Luc for NF-κB, and pAP-1-Luc for AP-1) were kindly provided by S. Xiao (57).

Antibodies.

Monoclonal mouse anti-JEV NS5 was prepared in our laboratory (58). Commercially available antibodies used included mouse monoclonal antibodies against Flag, c-Myc, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Abclonal Technology); rabbit polyclonal antibodies against KPNA1, KPNA2, KPNA3, KPNA4, KPNA5, KPNA6, KPNB1, KPNB2, p50, p65, IRF3, IκBα, and lamin A/C (Abclonal Technology); a mouse monoclonal antibody against phosphorylated IRF3 (IRF3-p; Cell Signaling Technology); and horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgGs (Boster, China).

Transfection and reporter assay.

HeLa cells were seeded in 24-well plates at a density of 2 × 10⁵ to 4 × 10⁵ cells per well. When cells were grown to 70 to 80% confluence, 100 ng of luciferase reporter plasmid (p125-Luc for IFN-β, pIRF3-Luc for IRF3, pNF-κB-Luc for NF-κB, and pAP-1-Luc for AP-1) and 10 ng of the Renilla luciferase construct phRL-TK (Promega), which served as an internal control, were cotransfected with empty vector or a plasmid encoding the indicated JEV or KPNA protein. Twenty-four hours later, cells were transfected with poly(I·C) (1.0 μg) or infected with JEV. Luciferase activities were measured 12 h after poly(I·C) transfection or at 12 h and 24 h postinfection. Cells were harvested in passive lysis buffer (Promega), and the reporter gene assay was performed by using a dual-luciferase reporter system (Promega). Luciferase activity was measured with a luminometer and expressed as relative luciferase activity by normalizing firefly luciferase activity to Renilla luciferase activity.

Immunofluorescence assay.

After the cells were grown to 80% confluence, nonadherent cells were removed by washing with medium before virus inoculation. The cells were transfected with 500 ng pFlag-NS5 or infected with JEV at an MOI of 1.0. At 24 h posttransfection/postinfection, cells were fixed and blocked with 10% bovine serum albumin (BSA) in phosphate-buffered saline (PBS; pH 7.2) for 30 min. The cells were then stained with the monoclonal antibody recognizing JEV NS5 for 1 h. After washing three times with PBS, cells were incubated with an Alexa Fluor 488-conjugated secondary antibody (Invitrogen) for 30 min. Cell nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole; Invitrogen). Staining was observed using a fluorescence microscope (Zeiss) at a magnification of ×40.

Coimmunoprecipitation and Western blotting.

HEK293T cells (1 × 10⁷) were transfected with the plasmids indicated in the figures. At 36 h posttransfection, cell extracts were prepared using RIPA buffer (Sigma-Aldrich) containing protease inhibitor cocktail (Roche). The cell lysate was incubated with the indicated antibodies at 4°C overnight. Protein A+G agarose beads (25 μl; Beyotime) were added and incubated for another 3 h. The agarose beads were subsequently washed three times with wash buffer (0.05 M Tris-HCl with 0.15 M NaCl). The bound proteins were eluted by boiling in SDS-PAGE loading buffer for 5 min and used for Western blotting with the indicated antibodies.

For Western blotting, whole-cell lysates were generated by lysing cells in RIPA buffer containing proteinase and phosphatase inhibitors (Roche). The cytoplasmic and nuclear proteins were extracted using an NE-PER nuclear and cytoplasmic extraction kit (Thermo Scientific). Protein concentrations were measured with a bicinchoninic acid (BCA) protein assay kit (Thermo Scientific). The proteins were separated by SDS-PAGE (10% polyacrylamide), transferred to polyvinylidene difluoride membranes, incubated with primary and secondary antibodies, and visualized with an enhanced chemiluminescence system (Bio-Rad).

qRT-PCR.

Total RNA was extracted from cells by use of TRIzol reagent (Invitrogen). Reverse transcription was performed using a ReverTra Ace qPCR RT kit (Toyobo). The level of each cDNA was determined by quantitative PCR using SYBR green real-time PCR master mix (Toyobo), and fluorescence intensity was analyzed with the ABI StepOne Plus system (Applied Biosystems). Expression data were normalized to the expression of β-actin.

Enzyme-linked immunosorbent assay (ELISA).

Kits from eBioscience were used to determine the levels of human IFN-β in cell cultures according to the manufacturer's instructions.

Plaque assay.

HeLa cells transfected with empty vector or a plasmid encoding KPNA3 or KPNA4 were infected with JEV at MOIs of 1.0 and 2.0. Twenty-four and 48 h after infection, cells were harvested, and virus titers were determined with a plaque assay on BHK cells. Briefly, viruses were serially diluted and inoculated onto monolayers of cells. After 1 h of absorption, cells were washed with serum-free DMEM and cultured in DMEM containing 3% fetal bovine serum and 1.5% sodium carboxymethyl cellulose (Sigma-Aldrich). Visible plaques were counted, and viral titers were calculated after 3 days of incubation.

Statistical analysis.

All experiments were repeated at least three times. Statistical analyses were performed using Prism 5 (GraphPad Software). Data represent the means ± standard errors of the means. Statistical differences among the experimental groups were determined using two-way analysis of variance (ANOVA), with subsequent t tests with the Bonferroni posttest used for multiple comparisons. P values of <0.05 were considered significant.