SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Hepatitis C virus (HCV) infection blocks cellular interferon (IFN)-mediated antiviral signaling through cleavage of Cardif by HCV-NS3/4A serine protease. Like NS3/4A, NS4B protein strongly blocks IFN-β production signaling mediated by retinoic acid–inducible gene I (RIG-I); however, the underlying molecular mechanisms are not well understood. Recently, the stimulator of interferon genes (STING) was identified as an activator of RIG-I signaling. STING possesses a structural homology domain with flaviviral NS4B, which suggests a direct protein-protein interaction. In the present study, we investigated the molecular mechanisms by which NS4B targets RIG-I–induced and STING-mediated IFN-β production signaling. IFN-β promoter reporter assay showed that IFN-β promoter activation induced by RIG-I or Cardif was significantly suppressed by both NS4B and NS3/4A, whereas STING-induced IFN-β activation was suppressed by NS4B but not by NS3/4A, suggesting that NS4B had a distinct point of interaction. Immunostaining showed that STING colocalized with NS4B in the endoplasmic reticulum. Immunoprecipitation and bimolecular fluorescence complementation (BiFC) assays demonstrated that NS4B specifically bound STING. Intriguingly, NS4B expression blocked the protein interaction between STING and Cardif, which is required for robust IFN-β activation. NS4B truncation assays showed that its N terminus, containing the STING homology domain, was necessary for the suppression of IFN-β promoter activation. NS4B suppressed residual IFN-β activation by an NS3/4A-cleaved Cardif (Cardif1-508), suggesting that NS3/4A and NS4B may cooperate in the blockade of IFN-β production. Conclusion: NS4B suppresses RIG-I–mediated IFN-β production signaling through a direct protein interaction with STING. Disruption of that interaction may restore cellular antiviral responses and may constitute a novel therapeutic strategy for the eradication of HCV. (HEPATOLOGY 2013)

Type I interferon (IFN) plays a central role in eliminating hepatitis C virus (HCV) both under physiological conditions and when used as a therapeutic intervention.1-3 In experimental acute-resolving HCV infection in chimpanzees, numerous IFN-related genes are expressed during clinical course of infection.4 Viruses are recognized by cellular innate immune receptors, such as toll-like receptors, and a family of RIG-I–like receptors, such as retinoic-acid-inducible gene I (RIG-I) and melanoma-differentiation-associated gene 5 (MDA-5); host antiviral responses are then activated, resulting in the production of cytokines such as type I and type III IFNs.5 RIG-I is activated through recognition of short double-strand RNA (dsRNA) or triphosphate at the 5′ end of dsRNA as pathogen-associated molecular patterns,6, 7 forming a homo-oligomer that binds with the caspase recruitment domain (CARD) of Cardif (also known as MAVS, VISA, or IPS-1).8-11 Cardif subsequently recruits TANK binding kinase 1 (TBK1) and IκB kinase ϵ (IKKϵ) kinases, which catalyze phosphorylation and activation of IFN regulatory factor-3 (IRF-3).12 Activation of TBK1 and IKKϵ results in the phosphorylation of IRF-3 or IRF-7, translocation to the nucleus, and induction of IFN-β mRNA transcription.

Several HCV proteins can block host cellular antiviral responses. HCV core protein blocks IFN signaling by interacting with signal transducer and activator of transcription protein-1 (STAT1).13 The core protein also induces expression of suppressor of cytokine signaling-1 (SOCS1) and SOCS3, and blocks Janus kinase–STAT signaling.14, 15 A well-elucidated immune evasion strategy of HCV involves NS3/4A serine protease and its ability to inhibit host IFN signal pathways. Gale and colleagues11, 16, 17 revealed that NS3/4A protease cleaves Cardif at Cys-508 resulting in dislocation of Cardif from mitochondria, and blocks downstream signaling of IFN-β production. On the other hand, Baril et al.18 reported that Cardif was still able to form a homo-oligomer and to activate downstream IFN production signaling despite delocalization from the mitochondria. These reports suggest that homo-oligomerization of Cardif, and not mitochondrial anchorage, is essential for the activation of downstream IFN signaling and that other virus-derived molecules may cooperate with NS3/4A to abrogate the signaling of IFN production.

We reported previously that HCV-NS4B, as well as NS3/4A, inhibited RIG-I and Cardif-mediated interferon-stimulated response element (ISRE) activation, while TBK1- and IKKϵ-mediated ISRE activation were not suppressed.19 These results indicate that NS4B suppresses IFN production signaling by targeting Cardif or other unknown signaling molecules between the level of Cardif and TBK1/IKKϵ.

Recently, a stimulator of interferon genes (STING, also known as MITA/ERIS/MPYS/TMEM173) was identified as a positive regulator of RIG-I–mediated IFN-β signaling.20-23 STING is a 42-kDa protein localized predominantly in the endoplasmic reticulum (ER) that binds RIG-I, Cardif, TBK1, and IKKϵ. STING is thought to act as a scaffold for Cardif/TBK1/IRF-3 complex upon viral infection.22 It has been reported that NS4B of yellow fever virus, which is a member of the flaviviridae family of viruses, inhibits STING activation probably through a direct molecular interaction.24 These reports have led us postulate that HCV-NS4B may also inhibit RIG-I dependent IFN signaling through association with STING.

In the present study, we further investigated the molecular mechanisms by which HCV-NS4B protein inhibits RIG-I–mediated IFN expression signaling. We demonstrated that HCV-NS4B specifically binds STING, blocks the molecular interaction between STING and Cardif, and suppresses the RIG-I–like receptor–induced activation of IFN-β production signaling.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Plasmids.

The ΔRIG-I and RIG-IKA plasmids express constitutively active and inactive RIG-I, respectively.5 Full-length Cardif (Cardif) and CARD-truncated Cardif (ΔCARD) plasmids were provided by J. Tschopp.11 Plasmids expressing STING were provided by G. N. Barber.20 Plasmids expressing HCV NS3/4A, NS4B, and truncated NS4B have been described.25 Plasmid pIFNβ-Fluc was provided by R. Lin.26

Cell Culture.

HEK293T and Huh7 cells were maintained in Dulbecco's modified minimal essential medium (Sigma) supplemented with 2 mM L-glutamine and 10% fetal calf serum at 37°C with 5% CO2.

HCV Replicon Constructs and HCV-JFH1 Cell Culture.

An HCV subgenomic replicon plasmid, pRep-Feo, expressed fusion protein of firefly luciferase and neomycin phosphotransferase.27, 28 Huh7 cells were transfected by Rep-Feo RNA, cultured in the presence of 500 μg/mL of G418, and a cell line that stably expressed Feo replicon was established. For HCV cell culture, the HCV-JFH1 strain was used.29, 30

Antibodies.

Antibodies used were anti–IRF-3 (FL-425, Santa Cruz Biotechnology), anti-HA (Invitrogen), anti-myc (Invitrogen), mouse anti-PDI (Abcam), rabbit anti-PDI (Enzo Life Science), anti-Flag (Sigma Aldrich), anti-Cardif (Enzo Life Science), anti-phospho–IRF-3 (Ser396, Millipore), anti-monomeric Kusabira-Green C- or N-terminal fragment (MBL), and anti-FACL4 (Abgent).

Luciferase Reporter Assay.

IFN-β reporter assays were performed as described.19, 31 The plasmids pIFN-β-Fluc and pRL-CMV were cotransfected with NS3/4A or NS4B, and ΔRIG-I, Cardif, STING or poly(deoxyadenylic-deoxythymidylic) acid [poly(dA:dT)] (Invivogen). RIG-IKA, ΔCARD, and pcDNA3.1, respectively, were used as controls. Luciferase assays were performed 24 hours after transfection by using a 1420 Multilabel Counter (ARVO MX PerkinElmer) and Dual Luciferase Assay System (Promega). Assays were performed in triplicate, and the results are expressed as the mean ± SD.

Immunoblotting.

Preparation of total cell lysates was performed as described.19, 28 Protein was separated using NuPAGE 4%-12% Bis/Tris gels (Invitrogen) and blotted onto an Immobilon polyvinylidene difluoride membrane. The membrane was immunoblotted with primary followed by secondary antibody, and protein was detected by chemiluminescence.

Immunoprecipitation Assay.

HEK-293T or Huh7 cells were transfected with plasmids as indicated. Twenty-four hours after transfection, cellular proteins were harvested and immunoprecipitation assays were performed using an Immunoprecipitation Kit according to the manufacturer's protocol (Roche Applied Science). The immunoprecipitated proteins were analyzed by immunoblotting.

Indirect Immunofluorescence Assay.

Cells seeded onto tissue culture chamber slides were transfected with plasmids as indicated. Twenty-four hours after transfection, the cells were fixed with cold acetone and incubated with primary antibody and subsequently with Alexa488- or Alexa568-labeled secondary antibodies. Mitochondria were stained by MitoTracker (Invitrogen). Cells were visualized using a confocal laser microscope (Fluoview FV10, Olympus).

BiFC Assay.

Expression plasmids of NS4B, Cardif, or STING that was fused with N- or C-terminally truncated monomeric Kusabira-Green (mKG) were constructed by inserting polymerase chain reaction–amplified fragments encoding NS4B, Cardif, or STING, respectively, inserted into fragmented mKG vector (Coral Hue Fluo-Chase Kit; MBL). HEK293T cells were transfected with a complementary pair of mKG fusion plasmids. Twenty-four hours after transfection, fluorescence-positive cells were detected and counted by flow cytometry, or observed by confocal laser microscopy.

Small Interfering RNA Assay.

Nucleotide sequences of STING-targeted small interfering RNAs (siRNAs) were as follows: (1) 5′-gcaacagcatctatgagcttctggagaac-3′, (2) 5′- gtgcagtgagccagcggctgtatattctc;-3′, (3) 5′-gctggcatggtcatattacatcggatatc-3′.22 Stealth RNAi Negative Control Duplex (Medium GC Duplex, Invitrogen) was used. Forty-eight hours after siRNA transfection, expression levels of STING were detected by immunoblotting.

Statistical Analyses.

Statistical analyses were performed using unpaired, two-tailed Student t test. P < 0.05 were considered to be statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

NS4B Suppressed RIG-I, Cardif, and STING-Mediated Activation of IFN-β Expression Signaling.

First, we performed a reporter assay using a luciferase reporter plasmid regulated by native IFN-β promoter. Consistent with our previous study,19 overexpression of NS4B, as well as NS3/4A, inhibited the IFN-β promoter activation that was induced by ΔRIG-I and Cardif, respectively (Fig. 1A). We next studied whether NS4B targets STING and inhibits RIG-I pathway–mediated activation of IFN-β production. Expression of NS4B protein significantly suppressed STING-mediated activation of the IFN-β promoter reporter, whereas expression of NS3/4A showed no effect on STING-induced IFN-β promoter activity (Fig. 1A). To study whether NS4B blocks the STING-mediated DNA-sensing pathway, we performed a reporter assay using a luciferase reporter plasmid cotransfection with poly(dA:dT), which is a synthetic analog of B-DNA and has been reported to induce STING-mediated IFN-β production and NS4B. NS4B significantly blocked poly(dA:dT)-induced IFN-β promoter activation, suggesting that NS4B may block STING signaling in the DNA-sensing pathway (Fig. 1A).

thumbnail image

Figure 1. NS4B suppressed IFN-β signaling mediated by RIG-I, Cardif, or STING. (A) Plasmids expressing ΔRIG-I, Cardif, or STING or poly(dA:dT) as well as NS3/4A or NS4B were cotransfected with pIFN-β-Fluc and pRL-CMV into HEK293T cells. After 24 hours, dual luciferase assays were performed. Plasmids expressing RIG-IKA, ΔCARD, or an empty plasmid (pcDNA) were used as a corresponding negative control. The experiments were performed more than three times and yielded consistent results. The y axis indicates relative IFN-β–Fluc activity. Assays were performed in triplicate and error bars indicate mean ± SD. *P < 0.05. (B) HEK293T cells were cotransfected with indicated plasmids. On the day after transfection, the cells were lysed and immunoblot analyses were performed. No Tf, transfection-negative controls. pIRF-3 and IRF-3, phosphorylated and total IRF-3, respectively.

Download figure to PowerPoint

Activation of RIG-I signaling induces phosphorylation of IRF-3, which is a hallmark of IRF-3 activation.32 Thus, we examined the effects of NS3/4A and NS4B expression on phosphorylation of IRF-3 by immunoblotting analysis. As shown in Fig. 1B, overexpression of ΔRIG-I, Cardif, or STING in HEK293T cells increased levels of phosphorylated IRF-3 (pIRF-3). Expression of NS4B impaired the IRF-3 phosphorylation that was induced by ΔRIG-I, Cardif, or STING. NS3/4A also blocked production of pIRF-3 induced by ΔRIG-I or Cardif. Intriguingly, NS3/4A did not block STING-induced pIRF-3 production. These results demonstrate that both NS3/4A and NS4B suppress RIG-I–mediated IFN-β production, but they do so by targeting different molecules in the signaling pathway.

Subcellular Localization of NS4B, Cardif, and STING.

We next studied the subcellular localization of NS4B following its overexpression and measured the colocalization of NS4B with Cardif and STING in both HEK293T cells and Huh7 cells by indirect immunofluorescence microscopy. NS4B was localized predominantly in the ER, which is consistent with previous reports33 (Fig. 2A). Cardif was localized in mitochondria but did not colocalize with the ER-resident host protein disulphide-isomerase (PDI). Interestingly, Cardif and NS4B colocalized partly at the boundary of the two proteins, although their original localization was different (Fig. 2A,C). STING was localized predominantly in the ER20, 21 (Fig. 2B,D). STING colocalized partly with Cardif, which is consistent with a previous report by Ishikawa and Barber20 (Fig. 2B,D). In cells cotransfected with NS4B and STING expression plasmids, NS4B colocalized precisely with STING (Fig. 2B,D). To examine the region of NS4B-STING interaction, we next observed the two proteins by performing staining for them along with mitochondria-associated ER membrane (MAM), which is a physical association with mitochondria34 and has been reported the site of Cardif-STING association.24 Both NS4B and STING were adjacent to and partially colocalized with fatty acid-CoA ligase long chain 4 (FACL4), which is a MAM marker protein35, 36 (Fig. 2E). These findings suggest that NS4B might interact with STING on MAM more strongly than with Cardif.

thumbnail image

Figure 2. Subcellular localization of NS4B, Cardif, and STING. (A-D) Subcellular localization of NS4B, Cardif, and STING in 293T (A,C) and Huh7 (B,D) cells. (A,C) NS4B-myc (first, second, and fifth panels of A and third panel of C) was transfected, and 24 hours later the cells were fixed and immunostained with anti-myc. In the third, fourth, and fifth panels of A, and the first and second panels of C, endogenous Cardif was detected with anti-Cardif antibody. ER was immunostained with anti-PDI antibody (first and third panels of A and first panel of C). Mitochondria were stained using Mitotracker (second and fourth panels of A and second panel of C). Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI). (B,D) STING-HA (all panels) and NS4B-myc (third panels) were transfected, and after 24 hours the cells were fixed and immunostained with anti-HA or anti-myc, respectively. In the fourth panels, endogenous Cardif was detected with anti-Cardif antibody. ER was immunostained with anti-PDI antibody (first panels). Mitochondria were stained using Mitotracker (second panels). Nuclei were stained with DAPI. (E) NS4B-myc and STING-HA were transfected into Huh7 cells and after 24 hours the cells were fixed and immunostained with anti-HA, anti-myc, and anti-FACL4 (MAM) antibody. Cells were visualized by confocal microscopy. Scale bars indicate 5 μm. In each microscopic image, the grade of protein colocalization in a single cell was quantified and is shown in the graphs at the bottom of each panel. Values are shown as the average colocalization ratio in 8 cells. Error bars indicate the mean + SD.

Download figure to PowerPoint

Protein-Protein Interaction Between NS4B, Cardif, and STING.

Knowing that NS4B was colocalized strongly with STING and only partly with Cardif, we next analyzed direct protein-protein interactions between NS4B, Cardif, and STING. To detect those interactions in living cells, we performed BiFC assays.37, 38 We constructed NS4B, Cardif, and STING expression plasmids that were N- or C-terminally fused with truncated mKG proteins, respectively. First, we cotransfected several different pairs of NS4B and STING expression plasmids that were fused with complementary pairs of N- or C-terminally truncated mKG. Strong fluorescence by mKG complexes (BiFC signal) was detected in all pairs of cotransfections, suggesting significant molecular interaction (Fig. 3A). In flow cytometry, all pairs of NS4B- and STING-mKG fusion proteins were positive for strong BiFC signal (Fig. 3B). The percentages of cells positive for BiFC signal were significantly higher in STING-mKG and NS4B-mKG fusion complexes than in corresponding controls (Fig. 3C). These results demonstrate that HCV-NS4B and STING proteins interact with each other strongly and specifically in cells. Fluorescence microscopy indicated that N- and C-terminal fusion of mKG onto NS4B and STING did not affect subcellular localization (Fig. 3D).

thumbnail image

Figure 3. BiFC assays of STING and NS4B. The complementary pairs of N- or C-terminally mKG-fused NS4B and STING expression plasmids were cotransfected in HEK293T cells. After 24 hours, the cells were fixed and observed by confocal microscopy (A) or subjected to flow cytometry to measure mKG-emitted fluorescence (BiFC signal) and to count BiFC signal-positive cells (B,C). Plasmids expressing p65-mKGN and p50-mKGC individually were used as a BiFC-positive control and plasmids expressing N- or C-terminally mKG fused Rluc were used as a negative control. The letters N and C denote complimentary N- and C-terminal fragments of mKG, respectively. Assays were performed in triplicate and error bars indicate the mean ± SD. Scale bars indicate 10 μm (A). *P < 0.05 compared with corresponding negative controls. (D) Plasmids expressing mKG fragment-fused STING or NS4B were transfected in HEK293T cells. After 24 hours, the cells were fixed and immunostained with anti-mKG and anti-PDI (ER) antibody. Nuclei were stained with DAPI. Cells were observed by confocal microscopy. Scale bars = 5 μm.

Download figure to PowerPoint

We next studied the molecular interaction between NS4B and Cardif by BiFC assay using NS4B and Cardif fusion plasmids that were tagged with complementary pairs of truncated mKG. Weak fluorescence was detected in cells transfected with the pairs N-Cardif and NS4B-C, N-Cardif and C-NS4B, C-Cardif and NS4B-N, and C-Cardif and N-NS4B (Fig. 4A,B). The percentage of cells positive for BiFC signal increased with the combination of N-Cardif and NS4B-C, and C-Cardif and NS4B-N (Fig. 4C). Fluorescence microscopy indicated that mKG-Cardif, but not Cardif-mKG, was partially colocalized with mitochondria, possibly due to disruption of mitochondria anchor domain by C-terminal fusion with mKG (Fig. 4D). These results indicate the lack of significant molecular interactions between NS4B and Cardif.

thumbnail image

Figure 4. BiFC assays of Cardif and NS4B. The complementary pairs of N- or C-terminally mKG-fused NS4B and Cardif expression plasmids were cotransfected in HEK293T cells. After 24 hours, the cells were fixed and observed by confocal microscopy (A) or subjected to flow cytometry to measure mKG-emitted fluorescence (BiFC signal) and to count BiFC signal-positive cells (B,C). Plasmids expressing p65-mKGN and p50-mKGC individually were used as a BiFC-positive control and plasmids expressing N- or C-terminally mKG-fused Rluc were used as a negative control. The letters N and C denote complimentary N- and C-terminal fragments of mKG, respectively. Assays were performed in triplicate, and error bars indicate the mean ± SD. Scale bars indicate 10 μm (A). *P < 0.05 compared with corresponding negative controls. (D) Plasmids expressing mKG fragment-fused STING or NS4B were transfected in HEK293T cells. After 24 hours, the cells were fixed and immunostained with anti-mKG antibody. Mitochondria were stained using Mitotracker, and nuclei were stained with DAPI. Cells were observed by confocal microscopy. Scale bars = 5 μm.

Download figure to PowerPoint

Binding of NS4B to STING Blocks Molecular Interaction Between Cardif and STING.

It has been reported that STING binds Cardif directly.20, 22 Thus, we hypothesized that NS4B, through a competitive interaction with STING, may hinder the direct molecular interaction between Cardif and STING. To verify this hypothesis, we performed immunoprecipitation assays. First, we transfected plasmids that expressed NS4B and Cardif, or NS4B and STING, in HEK293T cells or Huh7 cells, and performed immunoprecipitation. NS4B strongly bound to STING in both HEK293T cells and Huh7 cells, suggesting specific molecular interactions, whereas NS4B and Cardif did not show any obvious interaction (Fig. 5A,C). Consistent with previous reports, STING and Cardif showed significant interaction (Fig. 5B,D). Interestingly, those interactions were decreased by coexpression of NS4B, depending on its input amount, and finally blocked completely in both HEK293T and Huh7 cells (Fig. 5B,D). Collectively, the results above demonstrate that NS4B disrupts the interaction between Cardif and STING possibly through competitive binding to STING.

thumbnail image

Figure 5. Binding of NS4B to STING blocks molecular the interaction between Cardif and STING. (A,C) NS4B expression plasmid was cotransfected with STING or Cardif expression plasmid into HEK293T cells (A) or Huh7 cells (C). After 24 hours, cell lysates were subjected to immunoprecipitation using anti-HA or anti-Flag and were immunoblotted with anti-myc. (B,D) Cardif and STING expression plasmids were cotransfected with various amounts of NS4B plasmid in HEK293T cells (B) or Huh7 cells (D). After 24 hours, cells lysates were subjected to immunoprecipitation using anti-Flag and were immunoblotted with anti-HA.

Download figure to PowerPoint

Effects on HCV Infection and Replication Levels by STING Knockdown and NS4B Overexpression.

We next studied the impact of STING-mediated IFN production and its regulation by NS4B on HCV infection and cellular replication. First, we transfected three STING-targeted siRNAs into Huh7/Feo cells (Fig. 6A). As shown in Fig. 6B, STING knockdown cells conferred significantly higher permissibility to HCV replication. We next transfected HCV-JFH1 RNA into Huh7 cells that were transiently transfected with NS4B. As shown in Fig. 6C, HCV core protein expression was significantly higher in NS4B-overexpressed cells. Furthermore, HCV replication was increased significantly in Huh7/Feo cells overexpressing NS4B (Fig. 6D). Taken together, the results above demonstrate that STING and NS4B may negatively or positively regulate cellular permissiveness to HCV replication.

thumbnail image

Figure 6. Effects on HCV replication levels by STING knockdown and NS4B overexpression. (A) Effects of siRNA knockdown of STING by siRNA. Huh7 cells were transfected with STING-targeted siRNAs (siRNA STING-1, -2, and -3, respectively) or negative control siRNA (siRNA NTC). Seventy-two hours after transfection, cells were harvested and expression levels of STING protein were detected by immunoblotting. (B) Huh7 cells expressing HCV-Feo subgenomic replicon (Huh7/Feo)27, 28 were transfected with STING-targeted siRNAs or negative control siRNA. Seventy-two hours after transfection, cells were harvested, and internal luciferase activities were measured. The y axis indicates luciferase activity shown as a ratio of transfection-negative control. Assays were performed in triplicate, and error bars indicate the mean + SD. *P < 0.05 compared with corresponding negative controls. (C) Empty plasmid or plasmid expressing NS4B was transfected into Huh7 cells. After 24 hours, HCV-JFH1 RNA was transfected into these cells. Seventy-two hours after virus transfection, HCV core antigen levels in culture medium were measured. Assays were performed in triplicate, and error bars indicate the mean + SD. *P < 0.05 compared with corresponding negative controls. tf(−), transfection-negative control. (D) Huh7 cells expressing HCV-Feo replicon (Huh7/Feo)27, 28 were transfected with NS4B expressing plasmid or empty plasmid (pcDNA). Forty-eight hours after transfection, internal luciferase activities were measured. The y axis indicates luciferase activity shown as a ratio of the transfection-negative control. Assays were performed in triplicate, and error bars indicate the mean + SD. *P < 0.05 compared with corresponding negative controls.

Download figure to PowerPoint

The N-terminal Domain of NS4B Is Essential for Suppressing IFN-β Promoter Activity Mediated by RIG-I, Cardif, and STING.

It has been reported that the N-terminal domain of several forms of flaviviral NS4B shows structural homology with STING.24 We therefore investigated whether the STING homology domain in NS4B is responsible for suppression of IFN-β production. We constructed two truncated NS4B expression plasmids, which covered the N terminus (NS4Bt1-84, amino acids 1 through 84) containing the STING homology domain and the C terminus (NS4Bt85-261, amino acids 85 through 261), respectively (Fig. 7A). Immunoblotting showed that NS4Bt1-84 and NS4Bt85-261 yielded protein bands of ∼9 kDa and ∼20 kDa, respectively. Aberrant bands in the truncated NS4B may be due to alternative posttranslational processing. HEK293T cells were transfected with ΔRIG-I, Cardif, or STING, and NS3/4A or the truncated NS4B, along with IFN-β-Fluc plasmid, and a reporter assay was performed. NS4Bt1-84 significantly suppressed RIG-I, Cardif, and STING-induced IFN-β promoter activity, whereas NS4Bt85-261 did not (Fig. 7B). These results suggest that the N-terminal domain of NS4B is responsible for association with STING. Fluorescent microscopy indicated that both NS4Bt1-84 and NS4Bt85-26 colocalized with ER and STING (Fig. 7C).

thumbnail image

Figure 7. The N-terminal domain of NS4B is essential for suppressing IFN-β promoter activity induced by RIG-I, Cardif, or STING. (A) Immunoblotting of NS4B and truncated NS4B, NS4B t1-84, and NS4Bt85-216. HEK293T cells were transfected with NS4B or truncated NS4B. After 24 hours, the cells were lysed and immunoblot assays were performed. The band indicated by the pound sign (#) is a truncated NS4B, probably generated via alternative posttranslational processing. (B) Plasmids expressing ΔRIG-I, Cardif, or STING as well as NS3/4A or the indicated truncated form of NS4B were cotransfected with pIFN-β-Fluc and pRL-CMV in HEK293T cells. Dual luciferase assays were performed 24 hours after transfection. Plasmids expressing RIG-IKA, ΔCARD, or pcDNA were used as negative controls. The y axis indicates IFN-β-Fluc activity shown as relative values. Assays were performed in triplicate, and error bars indicate the mean ± SD. *P < 0.05 compared with corresponding negative controls. (C) Plasmids expressing NS4Bt1-84-myc of NS4Bt85-261-myc were transfected with or without plasmids expressing HA-STING in HEK293T cells. After 24 hours, the cells were fixed and immunostained. Nuclei were stained with DAPI. Cells were observed by confocal microscopy. Scale bars indicate 5 μm.

Download figure to PowerPoint

NS4B Suppresses IFN Production Signaling Cooperatively with NS3/4A.

It has been reported that HCV NS3/4A serine protease cleaves Cardif between Cys-508 and His-509, releases Cardif from the mitochondrial membrane, and blocks RIG-I–induced IFN-β production. We next assessed whether NS4B suppresses IFN-β production in the presence of Cardif cleaved by NS3/4A protease (Cardif1-508, Fig. 8A). The truncation of Cardif–C-terminal residue abolished mitochondrial localization but still colocalized with STING (Fig. 8B). The reporter assay showed that Cardif1-508 induced weak IFN-β activation. Interestingly, NS4B completely blocked the residual function of the Cardif1-508 protein to activate IFN-β expression, suggesting an additive effect of NS3/4A and NS4B on the RIG-I–activating pathway (Fig. 8C).

thumbnail image

Figure 8. NS4B suppressed IFN-β production pathway independently of and cooperatively with NS3/4A. (A) Immunoblotting of Cardif and truncated Cardif (Cardif1-508). HEK293T cells were transfected with Cardif or truncated Cardif (Cardif1-508). After 24 hours, the cells were lysed and immunoblot assays were performed. (B) Subcellular localization of Cardif and truncated Cardif (Cardif1-508). HEK293T cells were immunostained with anti-Cardif antibody or HEK293T cells were transfected with myc-tagged truncated Cardif (Cardif1-508-myc), and after 24 hours the cells were immunostained with anti-myc. Mitochondria were stained with Mitotracker (red) and nuclei were stained with DAPI (blue). Plasmid expressing myc-tagged truncated Cardif (Cardif1-508) and plasmid expressing HA-tagged STING were transfected into HEK293T cells. The cells were immunostained with anti-myc and anti-HA antibodies and analyzed by confocal laser microscopy. Scale bars = 10 μm. (C) Plasmids expressing Cardif or truncated Cardif (Cardif1-508) and pIFN-β-Fluc and pRL-CMV were transfected with or without plasmid expressing NS3/4A or NS4B into HEK293T cells as indicated. Dual luciferase assays were performed 24 hours after transfection. Plasmid expressing ΔCARD or pcDNA was used as a negative control. The y axis indicates IFN-β-Fluc activity shown as relative values. Assays were performed in triplicate, and error bars indicate the mean ± SD. *P < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

It has been reported that viruses, including HCV, target IFN signaling to establish persistent replication in host cells.39 We have reported that NS4B blocks the transcriptional activation of ISRE induced by overexpression of RIG-I and Cardif, but not by TBK1 or IKKϵ.19 In the present study, we have shown that NS4B directly and specifically binds STING, an ER-residing scaffolding protein of Cardif and TBK1 and an inducer of IFN-β production (Figs. 3 and 5), and blocked the interaction between STING and Cardif (Fig. 5B,D) resulting in strong suppression of RIG-I–mediated phosphorylation of IRF-3 and expressional induction of IFN-β (Fig. 1). Furthermore, HCV replication was increased by knockdown of STING or overexpression of NS4B (Fig. 6). Taken together, our results demonstrate that HCV-NS4B strongly blocks virus-induced, RIG-I–mediated activation of IFN-β production signaling through targeting STING, which constitutes a novel mechanism of viral evasion from innate immune responses and establishment of persistent viral replication.

Our results also showed that the effects of NS4B on the RIG-I signaling were independent of NS3/4A-mediated cleavage of Cardif. Reporter assays showed that a cleaved form of Cardif (Cardif1-508) partially retained activity for the induction of IFN-β promoter activation. The residual IFN-β promoter activation was suppressed almost completely by NS4B but not by NS3/4A (Fig. 8C). These findings show that there are at least two mechanisms by which HCV can abrogate RIG-I–mediated IFN production signaling to accomplish abrogation of cellular antiviral responses.

NS4B and STING are ER proteins,20, 21, 40 whereas Cardif is localized on the outer mitochondrial membrane.9 Consistent with those reports, our immunostaining experiments demonstrated that most NS4B protein colocalized with STING (Fig. 2), and their association was localized on MAM (Fig. 2E). In addition to the significant colocalization of STING and NS4B, STING partially colocalized with Cardif at the boundary region of the two proteins (Fig. 2B). Furthermore, immunoprecipitation experiments showed that overexpression of NS4B completely blocked the interaction of STING with Cardif (Fig. 5B). Ishikawa et al.24 reported that STING could associate with Cardif by MAM interaction. Castanier et al.41 reported that Cardif-STING interaction was enhanced in cells with elongated mitochondria. In addition, Horner et al.42, 43 observed NS3/4A targeting of MAM-anchored synapse and cleavage of Cardif at MAM but not in mitochondria. These results led us to speculate that interaction between STING and Cardif was enhanced by altering their subcellular localization during viral infection and that NS4B inhibits Cardif activation by interfering with the association between STING and Cardif on MAM-like NS3/4A behavior against host innate immunity.

HCV-NS4B is an ER-localized 27-kDa protein with several functions in the HCV life cycle. Cellular expression of NS4B induces convolution of the ER membrane and formation of a membranous web that harbors HCV replicase complex.44, 45 NS4B also has RNA-binding capacity.46 In addition, several point mutations of NS4B were found to alter viral replication activity.33, 46, 47 The studies above indicate that NS4B provides an important protein-protein or protein-RNA interaction platform within the HCV replication complex and is essential for viral RNA replication. However, there are few reports on the involvement of NS4B with antiviral immune responses. Consistent with our previous study, Moriyama et al.48 reported that NS4B partially inhibited dsRNA-induced but not TRIF-induced activation of IFN-β. In NS4B-expressing cells, IFN-α induced activation of STAT1 was suppressed.49 The present study has demonstrated that NS4B functions against the host IFN response, such that NS4B directly interacts with STING and suppresses downstream signaling, resulting in the induction of IFN production.

STING contains a domain homologous to the N terminus of NS4B derived from several flaviviruses, including HCV. In our previous NS4B truncation assay, the NS4B N-terminal domain (amino acids 1-110) was important for suppression of RIG-I–induced IFN-β expression.19 Consistent with these results, N-terminally truncated NS4B (NS4Bt1-84) significantly suppressed STING and Cardif-induced IFN-β promoter activation, whereas the C terminus of NS4B (NS4Bt85-261) did not (Fig. 7). These results reinforce our hypothesis that NS4B binds STING at its homology domain and blocks the ability of STING to induce IFN-β production.

A small molecule inhibitor of NS4B has been developed and is under preliminary clinical trials.50 Einav et al.51 identified clemizole hydrochloride, an H1 histamine receptor antagonist, as an inhibitor of the RNA-binding function of NS4B and HCV RNA replication. A phase 1B clinical trial of clemizole in hepatitis C patients has been completed.52 Other two NS4B inhibitors which are a compound of amiloride analog and anguizole are under preclinical development.53, 54 The possibility remains that such NS4B inhibitors may suppress HCV replication partly through inhibiting the ability of NS4B to suppress IFN-β production and restore cellular antiviral responses.

In conclusion, IFN production signaling induced by HCV infection and mediated by RIG-I is suppressed by NS4B through a direct interaction with STING. These virus-host interactions help to elucidate the mechanisms of persistent HCV infection and constitute a potential target to block HCV infection.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors are indebted to J. Tcshopp for providing Cardif, ΔCARD, and CARD and to G. N. Barber for the STING plasmids. This study was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology, Japan; the Japan Society for the Promotion of Science; Ministry of Health, Labour and Welfare, Japan; and the Japan Health Sciences Foundation.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_26017_sm_SuppFig1.tif9482KSupporting Information Figure 1.
HEP_26017_sm_SuppFig2.tif4390KSupporting Information Figure 2.
HEP_26017_sm_SuppFig3.tif10152KSupporting Information Figure 3.
HEP_26017_sm_SuppFig4.tif9507KSupporting Information Figure 4.
HEP_26017_sm_SuppFig5.tif5352KSupporting Information Figure 5.
HEP_26017_sm_SuppFig6.tif11540KSupporting Information Figure 6.
HEP_26017_sm_SuppFig7.tif10094KSupporting Information Figure 7.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.