| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
* Department of Microbiology and Immunology, Hokkaido University Graduate School of Medicine, Kita-ku, Sapporo, Japan;
Department of Molecular Immunology, Nara Institute for Science and Technology, Ikoma, Nara, Japan;
Department of Molecular Genetics, Institute for Virus Research, Kyoto University, Kyoto, Japan; and
Department of Immunology, Osaka Medical Center for Cancer, Higashinari-ku, Osaka, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
promoter. RIG-I/MDA5 recognizes dsRNA in the cytoplasm, whereas TLR3 resides in the cell surface membrane or endosomes to engage in extracytoplasmic recognition of dsRNA. Recent reports suggest that TLR3 induces cellular responses in epithelial cells in response to respiratory syncytial virus (RSV). The modus for TLR3 activation by RSV, however, remains unresolved. By small interference RNA gene-silencing technology and human cell transfectants, we have revealed that knockdown of NAK-associated protein 1 (NAP1) leads to the down-regulation of IFN- promoter activation >24 h after poly(I:C) or virus (RSV and vesicular stomatitis virus) treatment. NAP1 is located downstream of the adapter Toll-IL-1R homology domain-containing adapter molecule (TICAM)-1 (Toll/IL-1R domain-containing adapter-inducing IFN-) in the TLR3 pathway, but TICAM-1 and TLR3 did not participate in the IRF-3 and IFN- promoter activation by RSV infection. Virus-mediated activation of the IFN- promoter was largely abrogated by the gene silencing of IFN- promoter stimulator-1 (mitochondria antiviral signaling (MAVS), VISA, Cardif), the adapter of the RIG-I/MDA5 dsRNA-recognition proteins. In both the TLR and virus-mediated IFN-inducing pathways, I
B kinase-related kinase 
and TANK-binding kinase 1 participated in IFN- induction. Thus, RSV as well as other viruses induces replication-mediated activation of the IFN- promoter, which is intracellularly initiated by the RIG-I/MDA5 but not the TLR3 pathway. Both the cytoplasmic and TLR3-mediated dsRNA recognition pathways converge upon NAP1 for the activation of the IRF-3 and IFN- promoter. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
promoter (2). It remains unsettled whether TLR3 participates in viral infection-mediated IFN- induction by detecting the extrinsic dsRNA originating from infected cells. Some reports have suggested that influenza virus, respiratory syncytial virus (RSV), and other viruses induce TLR3-mediated cellular responses in bronchial epithelial cells (5). Cytokine/chemokine responses are reported in influenza virus and RSV infection (3, 4, 6). In particular, RSV has been reported to serve as an inefficient inducer of type I IFNs (7), thereby being in part resistant to host antiviral immunity. The identification of the two pathways initiated by TLR3 and RIG-I, which involve IRF-3 activation followed by production of IFN-, enabled an examination of the IFN response occurring in virus-infected cells (1).
TLR3 and RIG-I act as dsRNA-recognition pattern receptors, and are followed by adapter molecules named TICAM-1 (Toll-IL-1R homology domain-containing adapter molecule 1) (Toll/IL-1R domain-containing adapter-inducing IFN-; TRIF) and IFN- promoter stimulator (IPS)-1 (mitochondria antiviral signaling (MAVS)/VISA/Cardif), respectively, that indirectly link the common kinase complex IB kinase-related kinase (IKK) and TANK-binding kinase 1 (TBK1) (8, 9, 10, 11, 12, 13, 14). These kinases in turn activate IRF-3 and then the IFN- promoter (15, 16). Thus, the two pathways must converge upstream of the kinase complex. However, the primary molecule responsible for connecting the two pathways to the kinases remains undetermined. The distinctive role of the IFN-inducing pathways in mDC also remains to be elucidated.
NAK-associated protein 1 (NAP1) is initially characterized as an activator of IKK-related kinases and suggested to be involved in protection of cells from TNF-
-induced apoptosis (17). According to the study, the virus-activated kinases IKK and TBK1 assemble in the regulatory subunit NAP1, and NAP1 facilitates activation of NF-B by these kinases. In contrast, NAP1 coprecipitates with TICAM-1 by immunoprecipitation, suggesting the involvement of NAP1 in the TLR3-mediated IFN-inducing pathway (18). NAP1 forms a family with IKK
and TANK, which recruit kinases to relay the signal for cellular responses (19). We previously showed that NAP1 but not TANK interacts with TICAM-1 (18). Using the RNA interference (RNAi) technology, we have searched for a molecule linking the TLR3 TICAM-1 pathway and the virus-mediated IFN-inducing pathway by infecting HeLa or HEK293 cells with viruses (including RSV, which reportedly derives TLR3 responses; Ref. 20). The results showed the gene silencing of NAP1 and the downstream kinases, but not TLR3 or TICAM-1, led to a decrease in IFN- promoter activation. In the same system, exogenously added poly(I:C) required both TLR3 and TICAM-1 in addition to the NAP1-kinase complex for IFN- induction. From these studies, we conclude that NAP1 essentially participates in both the TLR3 and RIG-I/MDA5-mediated IRF-3 activation pathways.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
HeLa (a human cervical carcinoma cell line) cells were cultured in MEM with 2 mM L-glutamine and 10% heat-inactivated FCS (JRH Biosciences), and its subline Hep-2 cells (Japanese Cell Resource Bank) were cultured in DMEM (Invitrogen Life Technologies) with 10% FCS and antibiotics, 100 U/ml penicillin, and 100 µg/ml streptomycin (Invitrogen Life Technologies). In some experiments, we used alternative HeLa lines with stable siNAP1 vector for NAP1 silencing or siGFP vector for control. HEK293 cells (RIKEN) were cultured in DMEM as described previously (9).
Plasmid constructs and Ab
Poly(I:C) and anti-IRF-3 Ab were purchased from Amersham Biosciences and IBL. Rabbit-polyclonal Ab (pAb) against Flag (Sigma-Aldrich), against myc (Santa Cruz Biotechnology), and mouse mAb against myc (NeoMarkers or Covance) were purchased from the indicated suppliers. A mouse mAb against human NAP1 and a rabbit pAb against human RIG-I were prepared as reported previously (2, 17). The p-125 luc reporter was a gift from Dr. T. Taniguchi (Tokyo University, Tokyo, Japan) and contained the human IFN- promoter region (–125 to +19). The Gal4-IRF-3, Gal4-DBD, and p55 UASG-Luc for IRF-3 activation, pEF-BOS-Flag-RIG-I, RIG-IN, MDA5, and MDA5N were described in Refs. 21 and 22). Expression vectors for NAP1-full, NAP1-DN, and del-NAP1 (NAP1 deleted of the TBK1 binding site; 158–270 aa) in pcDNA3.1 were prepared as described previously (17). Expression vector for IPS-1 (pEF-BOS-Flag-IPS1) was amplified by PCR using KIAA1271 (Kazusa DNA Research Institute) as a template. The dominant-negative form of Mal (Toll/IL-1R domain-containing adaptor protein; TIRAP) was prepared as described previously (23).
Immunoprecipitation
Methods for immunoprecipitation and immunoblotting were described previously (23). Briefly, HEK293 cells were transiently transfected with expression vectors in a 6-well plate using LipofectAMINE 2000 reagent (Invitrogen Life Technologies) and allowed to stand for 24 h. Cells were lysed and proteins were immunoprecipitated with anti-myc mAb (NeoMarkers). Immunoprecipitants were washed, resolved on SDS-PAGE (7.5 or 10% gel), and visualized by immunoblotting using the anti-Flag pAb or anti- myc pAb.
HeLa cells were used for detection of the endogenous NAP1-RIG-I interaction. Cells were cultured in DMEM/10% FCS with or without recombinant human IFN- (1000 IU/ml) for 24 h. Cell lysates were immunoprecipitated by anti-NAP1 mAb or anti-mouse IgG Ab. Coimmunoprecipitation was detected by anti-RIG-I pAb.
Reporter gene assay
Activation of the IFN- promoter was measured by reporter assay. HEK293 cells were transfected in 24-well plates using LipofectAMINE 2000 reagent with a p-125 luc (IFN-) reporter plasmid together with the RIG-I, RIG-IN, MDA5, MDA5N or IPS-1, and NAP1 158–270 (NAP1-DN) plasmids. The properties of RIG-IN and MDA5N were published in a previous study (22). Briefly, the caspase activation and recruitment domain (CARD) of RIG-I (aa 1–284) acted as a constitutive active nature in the IFN- induction. Luciferase activity was measured by the Dual-Luciferase assay kit (Promega). The luciferase activity of firefly was normalized to that of Renilla, and relative activation was determined. All experiments were performed in triplicate. Data were expressed as the means ± SD.
Assay for IRF-3 activation
Two methods, reporter assay (21) and native gel assay (24), were used to determine the degree of IRF-3 activation. In some experiments, cells were stimulated with medium alone, poly(I:C) (10 µg/ml), or vesicular stomatitis virus (VSV). In some experiments of reporter assay, cells were transfected with poly(I:C) using DEAE-dextran at 24 h posttransfection. Six hours later, cells were harvested to measure the expression of luciferase using the dual luciferase assay kit (Promega). Data were expressed as the means ± SD. In native assay, cells were lysed after 8 h VSV infection (multiplicity of infection (MOI) = 10). The protein level of each sample was measured by the Protein assay Kit (Bio-Rad) and normalized 20 µg/each lane. Extracts were separated on 7.5% native gel and immunoblotting with anti-IRF-3 Ab (24).
Confocal microscope analysis
HeLa cells (1.5 x 105 cells/well) were plated onto coverslips in a 24-well plate. After cells adhered onto coverslips, cells were transiently transfected with Flag-tagged IPS-1, myc-tagged NAP1, and/or myc-tagged TANK using LipofectAMINE reagent (Invitrogen Life Technologies). Twenty-four hours later, in a typical experiment, cells were treated with 250 nM Mito Tracker Red (Molecular Probes) for 1 h. Cells on coverslips were washed twice with PBS and fixed in 3% formaldehyde in PBS for 30 min. Cells were permeabilized with staining buffer (3% BSA in PBS) containing 0.5% saponin for 30 min at room temperature. After three washes with staining buffer, cells were incubated with primary Ab diluted in staining buffer for 1 h. Cells were extensively washed, and then were treated with Alexa Fluor 488 monoclonal or polyclonal, or Alexa Fluor 594 pAbs (Molecular Probes) as secondary Ab for 30 min. After three washes, coverslips with the cells were mounted onto slide glasses using 2.3% DABCO in PBS. Imaging of the cells was curried out using Olympus Fluoview laser scanning confocal microscopy (25).
Virus preparation and infection
Human RSV field-isolate strain (RSV2177) in subgroup B was propagated with Hep-2 cells (provided by Dr. K. Imai, Wakayama Prefectural Center, Wakayama, Japan) (26). The titer of RSV2177 was determined by 50% tissue culture infective dose using Hep-2 cells. This RSV strain derives type I IFN from Hep-2 cells >24 h postinfection (M. Shingai and T. Seya, unpublished observation). HeLa cells were infected with RSV for small interference RNA (siRNA) knockdown studies at MOI = 1 or 2.5.
VSV was propagated as described previously (27). In brief, VSV (Indiana strain) was propagated with L929 cells. The titration of VSV was determined by plaque assay using L929 cells. HeLa cells were infected with VSV at MOI = 10.
RNAi
RNAi knockdown by siRNA-containing vectors was performed regarding NAP1 as follows. A DsRed2 fragment derived from pDsRed2 (BD Clontech) was cloned into pBluescript II (Stratagene). To paste into the XhoI and SpeI site of the vector, the PstI site of DsRed2 was replaced with CTGCAA by site-directed mutagenesis. Internal ribosome entry site (IRES) and puromycin resistance gene (Puro) fragments were amplified with an Expand High Fidelity PCR system (Roche Diagnostics) by using the following primer sets: AAACTAGTGCCCCTCTCCCTCCC and CTCACCATGGTTGTGGCCATATTATCATCGTGTTTTTCAAA, and AAACTCGAGTCCACCATGGGCACCGAGTACAAGCCCACGGT and AAAACTAGTGCGGCCGCTCAGGCACCGGGCTTGCGGG, respectively. pME-Puro vector was generated by cloning of the XhoI-SpeI fragment of Puro and placing it between the XhoI and XbaI sites of the pME vector. The XhoI-SpeI fragment of DsRed2 and the SpeI-NcoI fragment of IRES were cloned into the pME-Puro vector. pH1' vector was a gift from Drs. H. Hasuwa and M. Okabe (Osaka, Japan) (28). pH1'-DsRed-IRES-Puro was constructed by subcloning of the HindIII-KpnI fragment of the pME-DsRed2-IRES-Puro vector into the pH1' vector. Oligonucleotides were cloned into pH1 vector to express small interference hairpin-loop (sih)GFP and sihNAP1 (two sites) hairpins downstream of the human H1 RNA promoter as described previously (18, 28). 5'-AAGCTAATAGCTCGATTTGAAGA-3' and 5'-AAGTGATAATATGCAGCATGCAT-3' were the sequences targeted for sihNAP1-A and sihNAP1-B, respectively. The target sequence for sihGFP has been described earlier (28).
HeLa cells were transfected with pH1-GFP, pH1-NAP1-A, or pH1-NAP1-B (100 nM/4 x 105 cells) using PolyFect (Qiagen). Bulk cell populations in 1 mg/ml puromycin were selected from which single colonies were picked up for further analysis. To determine the efficiency of gene silencing, total RNA from each clone was isolated with RNeasy (Qiagen) and mRNA was estimated by RT-PCR. The primers used for PCR were GAPDH primers, and NAP1-F and NAP1-R for NAP1 (18).
The method for gene silencing using siRNA oligonucleotides was described previously (23). The sequences of the siRNA for TICAM-1 (9), TBK1, IKK (15), RIG-I, and IPS-1 (MAVS) were reported previously (14). The target sequence for human MDA5 gene silencing is 5'- AAAUACCAUAAUGGAGCAAUA-3'. Knockdown was analyzed by RT-PCR.
Detection of human IFN- mRNA
RT-PCR and quantitative PCR (Q-PCR) were performed as described previously (23). Briefly, human cells were transfected with gene silencing or dominant-negative-expressing vectors (100 nM/4 x 105 cells) 24 h before infection. Cells were then infected with RSV (MOI = 1–2.5). Twenty-four hours later, total RNA was extracted from cells with RNeasy mini kit (Qiagen) and cDNA was generated by M-MLV-reverse transcriptase (Promega). cDNA was subjected to PCR using Ex-Taq DNA polymerase (Takara). For this study, previously described primer sets for IFN-, NAP1, and TICAM-1 for silencing, and GAPDH or -actin for internal control were used (9, 18, 23). Other primer sets for TBK1, IKK, RIG-I, MDA5, and IPS-1 were described in Table I. Q-PCR was performed with iQ SYBER Green Super mix and iCycler iQ real-time PCR analyzing system (Bio-Rad). The primers used for IFN- and -actin on the Q-PCR were described in Table I. The copy number of IFN- mRNA was normalized to -actin mRNA, and relative fold induction of the medium to control was determined.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
First, we tested whether RIG-I coprecipitates with NAP1 by immunoblotting using HEK293 cells. Cells were transfected with the indicated plasmids (Fig. 1A) and 24 h later solubilized with Nonidet P-40. NAP1 was immunoprecipitated with anti-myc Ab. Anti-Flag or anti-myc Ab was used as a probe for protein detection in immunoblotting. Flag-tagged RIG-I was shown to form a complex with myc-tagged NAP1 by analysis with SDS-PAGE followed by immunoblotting. RIG-I was identified together with NAP1 on the sheet, suggesting that RIG-I physically binds NAP1. Similar results were obtained with NAP1 and MDA5 (Fig. 1A). To confirm endogenous RIG-I physically binds NAP1, we used HeLa cells pretreated with IFN- (Fig. 1B). RIG-I was efficiently induced in HeLa cells 24 h after IFN treatment (inset of Fig. 1B) and pulled down with NAP1 by immunoprecipitation using the mAb against NAP1.
|
RIG-I and MDA5 participate in the NAP1-mediated signaling
RIG-I and MDA5 contain two CARDs. The CARDs of RIG-I recruits a CARD-containing adapter IPS-1. The CARD domains of RIG-I (RIG-IN) and MDA5 (MDA5N) have been used in addition to RIG-I and MDA5 for signaling studies. IFN promoter was activated in cells expressing RIG-IN and to a lesser extent in cells with full-length RIG-I (Fig. 2A). In this context, MDA5 and MDA5N elicited comparable activities. When increasing doses of NAP1 DN were added to the cells with RIG-I, MDA5, or their mutants, IFN- promoter activation was prohibited by NAP1 DN in a dose-dependent manner (Fig. 2A). Using the Gal4-IRF-3 activation system, RIG-I- and MDA5-mediated IRF-3 activation was impaired in cells expressing NAP1 DN dose dependently (Fig. 2B). In the two systems, RIG-IN more potentially activated the promoters than intact RIG-I, whereas MDA5 and MDA5N exhibited similar tendencies (Fig. 2). Thus, the CARD domain of RIG-I induces activation of IRF-3 and IFN- promoter, both of which are regulated by NAP1. It is notable, although consistent with a previous report (22), that the full-length MDA5 that contains the helicase domain exhibited high IFN-inducing activity compared with the case of RIG-I. A functional link between RIG-I/MDA5 and NAP1 is confirmed by this analysis in addition to the physical linkage.
|
We next examined the ability of NAP1 to recruit IPS-1 in HEK293 cell transfectants. By simple immunoprecipitation analysis, NAP1 managed to coprecipitate with IPS-1 (Fig. 3A). Because IPS-1 resides in the mitochondrial membrane (14), we homogenized the transfectants and collected the mitochondria-rich fraction. Similar blotting results were obtained using this fraction (data not shown). Because NAP1 is a cytoplasmic protein, we investigated the localization of NAP1 using myc-tagged NAP1 expressed in HeLa cells by confocal microscopy. NAP1 was distributed in cytoplasm, particularly located around mitochondria (Fig. 3B, upper panel). In contrast, TANK, which is a structural homologue of NAP1, barely merged with mitochondrial marker (Fig. 3B, lower panel). To test the localization of NAP1 and IPS-1, Flag-tagged IPS-1 and myc-tagged NAP1 were expressed in HeLa cells, and merging analysis by confocal microscopy was performed. We confirmed that IPS-1 is the mitochondrial protein using our construct (data not shown). IPS-1 recruited NAP1 in the vicinity of the membrane (Fig. 3C). The results indicate that NAP1, at least in part, forms a complex with IPS-1 to relay the IRF-3-activating signal, although the reason why IPS-1 and NAP1 marginally coprecipitate with each other on immunoblotting remains unknown (Fig. 3A).
|
promoter and IRF-3 activation (Fig. 4, A and B). Hence, there appear to be both physical and functional linkages between IPS-1 and NAP1. Thus, the results can be interpreted to most likely mean that the cytoplasmic (RIG-I/MDA5) and endosomal (TLR3) pathways converge on NAP1 to activate the IRF-3-activating kinases, leading to induction of IFN-.
|
Virus-mediated IRF-3 activation was examined with poly(I:C), VSV (Figs. 5 and 6), and RSV (Fig. 7), which are reported to activate cytoplasmic RIG-I and extracytoplasmic TLR3, respectively, to induce IFN promoter activation. Poly(I:C) is a reagent that activates the cytoplasmic IFN-inducing pathway in human cells in the presence of DEAE-dextran (Fig. 5A). HEK293 cells transfected with poly(I:C) using DEAE-dextran barely activated the IFN- promoter in the reporter assay (left-side bars in Fig. 5A). When RIG-I and MDA5 were transfected into the cells before poly(I:C) stimulation, IFN- promoter was activated by transfected poly(I:C), and NAP1-DN blocked those IFN- promoter activation (Fig. 5A) and IRF-3 activation (data not shown). MDA5 prominently activated the IFN promoter compared with RIG-I, and this activity was efficiently impaired by cotransfection of NAP1 DN (Fig. 5A). Similar NAP1 DN-mediated suppression of IFN- promoter activation was observed in cells without poly(I:C) (Fig. 5A), although the magnitude of the promoter activation was far less. A pervious study (29) demonstrated that VSV activates RIG-I in infected cells. VSV was subjected to the HEK293 cells instead of poly(I:C), and IRF-3 activation was tested in the reporter (data not shown) and native gel analyses (Fig. 5B). Under the conditions where IRF-3 is activated in response to VSV, NAP1 DN inhibited VSV-mediated IRF-3 activation in a dose-dependent manner (Fig. 5B). Thus, the results infer that RIG-I/MDA5 interacts with NAP1 in the virus-derived IFN-inducing pathway.
|
|
|
induction (Fig. 6A). IRF-3 dimer formation by VSV was apparently reduced in the NAP1-deficient cells (Fig. 6B). Concomitantly, the level of IFN- mRNA was decreased in the NAP1-deficient cells (Fig. 6C). In this experiment, HeLa cell clones with the vector containing GFP siRNA were used as controls. Two NAP1-depleting clones of one target’s site (site-A) and one depleting clone of another target site (site-B) showed similar low responses to VSV in IFN- induction compared with the control (Fig. 6C), which excludes the possibility of artificial clonal effect. These results, together with the fact that VSV replication allows human cells to produce IFN- in a RIG-I-dependent manner (29), suggest that NAP1 is indispensable to RIG-I-mediated IFN induction.
RSV is known to induce TLR3 responses in epithelial cells. RSV was used instead of VSV to confirm the effect on virus-activated IFN- promoter activation in HeLa cells (Fig. 7). NAP1 as well as TICAM-1, IKK, and TBK1 were silenced with siRNA, and the IFN- mRNA levels were then measured in RSV-infected cells. Cells stimulated with poly(I:C) were used as control for the TICAM-1 pathway. As expected and consistent with previous reports, poly(I:C)-dependent IFN- mRNA was found to be dependent on TICAM-1, IKK, and TBK1 (Fig. 7A). Notably, TICAM-1 silencing had no down-regulation effect on the IFN- mRNA level, whereas IKK and TBK1 were associated with RSV-mediated IFN induction (Fig. 7A). The IFN- mRNA levels were evaluated 48 h after RSV infection in cells silencing of RIG-I, MDA5, or IPS-1. The IFN- mRNA induction was most prominently impaired in cells depleted of RIG-I or IPS-1 (Fig. 7B). When NAP1 was silenced, 40% of the IFN- mRNA level was reduced in the cells stimulated with poly(I:C) within 6 h (cells die during long-term incubation with poly(I:C)) (data not shown). In this system, RSV infection resulted in a 50% decrease of the NAP1-mediated IFN- induction (Fig. 7C). To further confirm the involvement of NAP1 in the virus-mediated IFN--inducing pathway, we used the dominant-negative transfectants (Fig. 7D). No reduction of the IFN- mRNA level was observed with the Mal/TIRAP dominant-negative-expressing HEK cells, whereas the IFN- mRNA level was significantly decreased in cells expressing the NAP1 DN. Thus, RSV induces IFN- independent of TICAM-1 but dependent on RIG-I, IPS-1, virus-activated kinases, and NAP1.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
and TBK1. The kinases IKK and TBK1 are known to be virus-activated kinases (16) and are located downstream of TICAM-1 (also known as TRIF) (15). A previous study (18) revealed that NAP1 is also involved in a molecular complex containing TICAM-1. The N-terminal region out of the TIR domain of TICAM-1 participates in NAP1 recruitment. Thus, NAP1 works in the two different pathways for dsRNA-mediated IFN- induction, TLR3/TLR4 followed by the TICAM-1 pathway and RIG-I/MDA5 followed by the IPS-1 pathway.
The results were consolidated with virus infection studies. RSV as well as VSV induces IFN- in infected cells, and these viral IFN-inducing activities were blocked by NAP1 DN or siRNA, suggesting that NAP1 participates in virus-activated IRF-3 and IFN- induction. NAP1 is implicated in the intracytoplasmic IFN--inducing signal in physical proximity to the molecular complex containing RIG-I and MDA5. Hence, NAP1 connects RNA sensor proteins (TLR3, RIG-I, and MDA5) and kinases to activate IRF-3. This scenario would be suitable to the findings of previous studies on IFN- induction by cells with exogenously added poly(I:C) (18) and with replicated viruses (29).
NAP1 assembles IKK and TBK1 to form a kinase complex (17). This kinase complex participates in activation of not only IRF-3 but also NF-B. Actually, we have evidence that overexpression of RIG-I (particularly RIG-IN) or MDA5 in HEK cells results in activation of NF-B and, under this situation, NAP1 DN blocks the NF-B activation through RIG-I/MDA5 (data not shown). NF-B activation through TNF- receptors, however, revealed to involve two additional kinases, IKK and IKK (30). Furthermore, two additional subunits, TANK and IKK, may affect the level of NF-B activation (31). TLR4 as well as other TLRs principally activate NF-B through IKK and a kinase complex with IKK and IKK (32). In activation of NF-B via NAP1, what combinations of the four kinases preferentially join the NAP1 protein complex still remains undetermined. Accordingly, molecular configuration of the kinase complex for activation of NF-B remains to be defined downstream of IPS-1.
Using two species of viruses, VSV and RSV, RIG-I/MDA5 is evidently responsible for sensing viral infection to induce IFN-. Previous reports have indicated that VSV stimulates the RIG-I pathway (33) and RSV induces TLR3 up-regulation and responses including cytokine/chemokine secretion in airway epithelial cells (4, 5, 20) Our gene silencing studies using siRNA (Fig. 7) and the RSV strain (an IFN-inducible strain) suggest that RIG-I is a key molecule in RSV-mediated IFN- induction in HeLa cells. MDA5 or TLR3 minimally participates in RSV sensing, if any, in this in vitro study. VSV, a representative of the RIG-I-activating virus, confers a similar IFN-inducing profile to the RSV strain.
Hence, the involvement of TLR3 in the induction of type I IFN by RSV may reflect a secondary response resulting from TLR3 up-regulation by the initial production of IFN- in virtue of RIG-I. We previously showed that NAP1 but not TANK interacts with TICAM-1 in the TLR3 pathway (18). Preferential recognition of RSV replication by RIG-I with no involvement of TLR3 and TICAM-1 appears to be additional evidence that NAP1 plays a major role downstream of RIG-I and outcome of viral infection. Because the results were obtained with HeLa and HEK293 cells in an acute-phase infection, this result should be confirmed with a cell line of bronchial epithelial cells and in vivo animal models.
However, it remains in question why IPS-1 only marginally coprecipitates with NAP1 in this study. According to the confocal analysis, NAP1 merges with IPS-1 on mitochondria. The discrepancy may be explained by the activation-induced mobility of IPS-1. IPS-1 overexpression leads to activation of the IFN- promoter (Fig. 4A; Refs. 12, 13) and shows a merging profile with NAP1. A previous study (14) suggests that IPS-1 moves from a detergent-soluble to detergent-insoluble fraction in mitochondria following virus infection. The transposition of IPS-1 on the mitochondrial membrane may affect the dynamics of the cytoplasmic protein NAP1. So far, we have not yet established the system to see the molecular interaction between endogenous NAP1 and IPS-1. Maybe, the question lies in the artificial overexpression system. We favor the idea that IPS-1 forms a complex with NAP1, but the complex is fragile depending on the conditions where IPS-1 is disposed or solubilized from the mitochondrial membrane. Alternatively, NAP1 may only temporally bind IPS-1 on the mitochondrial membrane in initiating the pathway that activates IRF-3 in human cells. In this context, it is of interest to see the effect of viral infection on the molecular association between NAP1 and IPS-1.
Two further points remain to be discussed. First, our previous immunoprecipitation studies suggested that TICAM-1 forms a complex with NAP1 (18), whereas IPS-1 barely joins the complex containing TICAM-1 (data not shown). NAP1 coprecipitates with TICAM-1 irrespective of poly(I:C) activation of TLR3 (18). However, NAP1 does not merge with intrinsic TICAM-1 (data not shown), but with IPS-1 (Fig. 3C) largely around the mitochondria as shown by confocal analysis. The localization pattern of NAP1 is mysterious but may reflect the differential properties between TICAM-1 and IPS-1. Human cells mostly have a low level of TICAM-1 and a high level of IPS-1, and the level of TICAM-1 protein expression is usually suppressed because of its apoptosis-inducing properties (34). TICAM-1 undergoes protein modifications in response to poly(I:C) stimulation (M. Sasai and T. Seya, unpublished data), which may be a prerequisite for the recruitment of NAP1.
Secondly, RIG-I and MDA5 have been shown to be responsible for intracellular viral dsRNA recognition (33). They are cytoplasmic proteins of the CARD-helicase-containing family (22). RIG-I without a helicase domain (RIG-IN) acts as a constitutively active IFN inducer in transfected cells, whereas MDA5 expresses full IFN-inducing activity regardless of its helicase domain. The manner of ligand recognition may be somewhat different in each. Why RIG-I mainly participated in recognition of VSV/RSV infection is unknown. After completing this study, a report (29) was published showing that MDA5 preferentially recognizes poly(I:C) and the replicated RNAs of picornaviruses, whereas RIG-I recognizes transcribed dsRNAs of many RNA viruses. Although the report did not mention the natural ligand of MDA5, it offers evidence that MDA5 and RIG-I discriminate between the dsRNA structures (29). Because RSV is a negative-strand RNA virus, it may hold a RNA structural motif inducing activation of RIG-I common to other negative-strand RNA viruses (35).
Pattern recognition molecules besides TLR3 and the CARD-helicase proteins are engaged in foreign RNA sensing in human cells. For example, TLR7 and TLR8 are expressed in human plasmacytoid DCs and mDCs, respectively (36). PKR is a receptor for dsRNA recognition (37). Generation of dsRNA may link to the pathway for inducing RNAi even in human cells. Why the host provides a variety of RNA pattern-recognition receptors and nonself RNA responses will be the issue to be clarified. Another point is that viral factors other than RNA have some host-immune-modulating functions. Measles nucleoproteins may modulate a Fc receptor-mediated immune response (38) and IRF-3 activation (39), which clearly occur independent of dsRNA or RNA replication. A recent report (40) suggests that negative-strand RNA viruses produce only undetectable amounts of dsRNA in infected cells. Although RIG-I is a key molecule in host protection against a number of RNA virus infections, other molecules and systems may be involved in host strategies against virus infection.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported in part by Core Research for Engineering, Science, and Technology, Japan Science and Technology Corporation, by Grants-in-Aid from the Ministry of Education, Science, and Culture (Specified Project for Advanced Research) and the Hepatitis C Virus project in National Institutes of Health of Japan, and by the Naito Memorial Foundation, Uehara Memorial Foundation, Mitsubishi Foundation, and Osaka Community Foundation. M.Sa. is supported by fellowships from the Japanese Society for the Promotion of Science. 
2 Address correspondence and reprint requests to Dr. Tsukasa Seya, Department of Microbiology and Immunology, Graduate School of Medicine, Hokkaido University, Kita-ku, Sapporo 060-8638, Japan. E-mail address: seya-tu{at}pop.med.hokudai.ac.jp
3 Abbreviations used in this paper: RIG-I, retinoic acid-inducible gene I; mDC, myeloid dendritic cell; MDA5, melanoma differentiation-associated gene 5; IRF, IFN regulatory factor; RSV, respiratory syncytial virus; TICAM-1, Toll-IL-1R homology domain-containing adapter molecule 1; TRIF, Toll/IL-1R domain-containing adapter-inducing IFN-; IPS-1, IFN- promoter stimulator 1; MAVS, mitochondria antiviral signaling; IKK, IB kinase-related kinase ; TBK1, TANK-binding kinase 1; NAP1, NAK-associated protein 1; RNAi, RNA interference; pAb, polyclonal Ab; TIRAP, Toll/IL-1R domain-containing adaptor protein; CARD, caspase activation and recruitment domain; VSV, vesicular stomatitis virus; MOI, multiplicity of infection; siRNA, small interference RNA; IRES, internal ribosome entry site; sih, small interference hairpin-loop; Q-PCR, quantitative PCR; NAP1 DN, NAP1 dominant negative.
Received for publication May 3, 2006. Accepted for publication September 20, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
induction. Nat. Immunol. 4: 161-167. [Medline] signaling. Mol. Cell 19: 727-740. [Medline]B and IRF 3. Cell 122: 669-682. [Medline] and TBK1 are essential components of the IRF3 signaling pathway. Nat. Immunol. 4: 491-496. [Medline]B kinase-related kinases that potentiates NF-B signaling. Mol. Cell. Biol. 23: 7780-7793. B-activating kinase-associated protein 1 participates in TLR3/Toll-IL-1 homology domain-containing adapter molecule-1-mediated IFN regulatory factor 3 activation. J. Immunol. 174: 27-30. production in myeloid dendritic cells (mDC). J. Biochem. 139: 171-175. . J. Biol. Chem. 278: 49751-49762. B in response to double-stranded RNA and virus infection: common and unique pathways. Genes Cells 6: 375-388. [Abstract]B kinase kinases. Biochim. Biophys. Acta 1471: M57-M62. [Medline]B activation and apoptosis pathways. J. Biol. Chem. 279: 15652-15661. This article has been cited by other articles:
|
|
 |
|
  M. Shingai, M. Azuma, T. Ebihara, M. Sasai, K. Funami, M. Ayata, H. Ogura, H. Tsutsumi, M. Matsumoto, and T. Seya Soluble G protein of respiratory syncytial virus inhibits Toll-like receptor 3/4-mediated IFN-beta induction Int. Immunol., July 8, 2008; (2008) dxn074v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Funami, M. Sasai, H. Oshiumi, T. Seya, and M. Matsumoto Homo-oligomerization Is Essential for Toll/Interleukin-1 Receptor Domain-containing Adaptor Molecule-1-mediated NF-{kappa}B and Interferon Regulatory Factor-3 Activation J. Biol. Chem., June 27, 2008; 283(26): 18283 - 18291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Zhang, Y. K. Chan, B. Lu, M. S. Diamond, and R. S. Klein CXCR3 Mediates Region-Specific Antiviral T Cell Trafficking within the Central Nervous System during West Nile Virus Encephalitis J. Immunol., February 15, 2008; 180(4): 2641 - 2649. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. E. Randall and S. Goodbourn Interferons and viruses: an interplay between induction, signalling, antiviral responses and virus countermeasures J. Gen. Virol., January 1, 2008; 89(1): 1 - 47. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Shingai, T. Ebihara, N. A. Begum, A. Kato, T. Honma, K. Matsumoto, H. Saito, H. Ogura, M. Matsumoto, and T. Seya Differential Type I IFN-Inducing Abilities of Wild-Type versus Vaccine Strains of Measles Virus J. Immunol., November 1, 2007; 179(9): 6123 - 6133. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Matsumiya, S. M. Prescott, and D. M. Stafforini IFN-{epsilon} Mediates TNF-{alpha}-Induced STAT1 Phosphorylation and Induction of Retinoic Acid-Inducible Gene-I in Human Cervical Cancer Cells J. Immunol., October 1, 2007; 179(7): 4542 - 4549. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ebihara, H. Masuda, T. Akazawa, M. Shingai, H. Kikuta, T. Ariga, M. Matsumoto, and T. Seya Induction of NKG2D ligands on human dendritic cells by TLR ligand stimulation and RNA virus infection Int. Immunol., October 1, 2007; 19(10): 1145 - 1155. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||
, GFP-silencing cells; 
, NAP1-silencing cells. D, HEK293 cells were transiently transfected with the dominant-negative form of NAP1 or Mal/TIRAP. Twenty-four hours after transfection, cells were infected with RSV (MOI = 1 for 36 h). The mRNA levels of IFN-