The RNA Helicase Lgp2 Inhibits TLR-Independent Sensing of Viral Replication by Retinoic Acid-Inducible Gene-I

Mice, viruses, and reagents

MyD88^−/−, TRIF^−/−, TRAM^−/−, Mal^−/−, and TLR3^−/− mice were obtained from S. Akira (Osaka University, Osaka, Japan). TRAF6^−/− mice were purchased from Tak Mak. Studies with these mice have been reviewed and approved by the University of Massachusetts Medical School institutional animal care and use committee. TLR3-expressing HEK293 cells were previously described (21). SV (300 hemagglutination units (HAU)/ml; Cantrell strain) and NDV (240 HAU/ml; strain VR-699) were purchased from Charles River Laboratories and American Type Culture Collection, respectively. The HSV-KOS strain was obtained from D. Knipe (Harvard Medical School, Boston, MA). Influenza (strain X:31, A/Aichi/68 (H3N2)) was obtained from D. Libraty (University of Massachusetts Medical School, Worcester, MA). CpG-A (oligodeoxynucleotide 2216) was purchased from Coley Pharmaceuticals. Polyinosinic-polycytidylic acid (poly(I-C)) was purchased from Amersham Biosciences. Re-extracted LPS was previously described (26). R-848 was obtained from GL Synthesis.

Plasmid constructs

The pEF-BOS-RIG-I, RIG-I-C, RIG-I-K270A, eIF-4A, IFN-β-luciferase (p125Luc) constructs were obtained from T. Fujita (Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan) (19), pCR-3-flag-murine-Mda5/Helicard was obtained from J. Tschopp (University of Lausanne, Epalinges, Switzerland) (27), and pcDNA3-Nod1 and Nod2 were obtained from G. Nunez (University of Michigan Medical School, Ann Harbor, MI). The NF-κB luciferase reporter was as previously described (28). The ISG54-IFN-stimulated regulatory element (ISRE) was purchased from Stratagene. pCDNA3-RIG-I-CFP (C-terminal fusion protein with cyano-fluorescent protein) was generated by PCR. Lgp2 was amplified by PCR from a human spleen cDNA library and cloned into pEF-BOS or as a C-terminal fusion protein with yellow-fluorescent protein (YFP) into pcDNA3. Lgp2-K30A was generated by site-directed mutagenesis. All PCR-generated constructs were confirmed by sequence analysis.

UV and heat inactivation of the Cantrell strain of SV

SV was inactivated by heating to 55°C for 30 min or through successive cycles of UV cross-linking. To confirm the inability to replicate, fractions of UV-inactivated virus were injected into 12-day-old fertilized eggs. Two days after incubation, the allantoic fluid was harvested and checked for hemagglutination of chicken RBC.

Reporter assays and ELISAs

For reporter assays, HEK293, 293T, or HEK293-TLR3 cells (2 × 10⁴ cells/100 μl/well in 96-well plates) were transfected with 40 ng of the indicated luciferase reporter gene together with 40 ng of thymidine kinase Renilla-luciferase reporter gene (Promega) and the indicated amount of expression plasmids using Genejuice (Novagen). Where indicated virus or poly(I-C) (25 μg/ml) was added for 16 h, and luciferase activity was measured as previously described (21). Murine bone marrow-derived DCs were generated in the presence of Flt3 ligand (Flt3-L) as previously described (29). Murine bone marrow-derived macrophages were generated in the presence of recombinant M-CSF (10 ng/ml) as previously described (30). Spleens were isolated from C57BL/6 or TRAF6^−/− mice, and splenocytes were harvested by mechanical disruption using 40-μm pore size nylon cell strainers (BD Biosciences), followed by lysis of RBC using Tris-NH₄Cl. DCs, macrophages, or splenocytes (1 × 10⁵/100 μl/well) were stimulated as indicated. Murine IFN-α was measured by ELISA as previously described (29). Using ELISA, we measured murine RANTES and IL-12-p40 according to the manufacturer’s recommendations (R&D Systems). Murine IFN-β was measured by ELISA using the following Ab pairs: coating Ab, rabbit polyclonal anti-mouse IFN-β (PBL Biomedical); and detection Ab, rat monoclonal anti-mouse IFN-β (Yamasa). In HEK293 cells, RANTES expression was measured by ELISA using a human RANTES ELISA Duoset (R&D Systems)

RNA interference

RNA interference vectors (piGENE hU6 vector; iGENE Therapeutics) targeting human and mouse RIG-I were obtained from T. Fujita (19). The p53 targeting sequence in pSuper was previously described (21). To determine the efficiency of gene silencing, 5 × 10⁴ 293T cells were seeded into 12-well plates and transfected with either 0.5 μg of RIG-I-CFP or 0.5 μg of RIG-I-Flag together with 0.5 μg of the small hairpin RNA constructs. After 48 h, the expression of RIG-I-CFP was measured by flow cytometry. Whole-cell lysates of RIG-I-Flag-transfected cells were also analyzed after silencing by immunoblotting with anti-Flag-M2 (Sigma-Aldrich) and an Ab to caveolin (Transduction Laboratories) as a control. For reporter assays, 2 × 10⁴ HEK 293 or murine L929 cells were transfected with 100 ng/well of the indicated shRNA constructs together with 10 ng of the IFN-β-luciferase reporter/well. After 72 h, the transfected cells were infected with SV, and 9 h later, luciferase activity was measured.

RNA binding analysis of RIG-I and Lgp2

To generate poly(I-C)-coated agarose beads, poly(C)-coated beads (Sigma-Aldrich) were resuspended in 2 vol of 2 mg/ml poly(I) (Sigma-Aldrich) prepared in 50 mM Tris (pH 7.0)-150 mM NaCl. The mixture was then rocked gently overnight at 4°C, collected by centrifugation at 1000 × g, washed with 50 mM Tris (pH 7.0)-150 mM NaCl, resuspended in the same buffer as a 10% final slurry, and stored at 4°C for use.

For poly(C) and poly(I-C) pull-down assays, poly(C)- or poly(I-C)-coated beads were equilibrated in binding buffer (50 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, and 1% Nonidet P-40) as a 10% slurry and combined with an equal volume of whole-cell extract from cells transiently transfected with the indicated YFP/CFP-tagged constructs or from an Lgp2-YFP-expressing stable cell line (lysis buffer: 50 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, and 10% glycerol). The cell extracts were supplemented with protease and phosphatase inhibitors and 25 U of RNase inhibitor/ml. The mixtures were incubated with gentle agitation for 2 h at 4°C. For competition experiments, pull-down reactions were supplemented with 10 μg of the competitor RNA/ml. Beads were then centrifuged at 1,000 × g, rinsed three times with lysis buffer, and resuspended in 2× SDS-PAGE sample buffer. Samples were boiled at 100°C for 5 min, centrifuged at 13,000 × g for 30 s, loaded immediately onto SDS-PAGE gels, and processed for immunoblot analysis with anti-GFP Abs as indicated.

Quantitative real-time PCR

Raw 264.7 macrophages were infected in six-well plates (1 × 10⁶/well) with SV, and RNA was isolated using RNeasy (Qiagen). cDNA was synthesized, and quantitative RT-PCR analysis was performed on a DNA engine Opticon 2 cycler (MJ Research) using the SuperScript III Two-Step qRT-PCR Kit with SYBR Green (Invitrogen Life Technologies) and the following primers: β-actin: forward, 5′-TTGAACATGGCATTGTTACCAA-3′; reverse, 5′-TGGCATAGAGGTCTTTACGGA-3′; IFN-β: forward, 5′-TGCTCTCCTGTTGTGCTTCTCC-3′; reverse, 5′-CATCTCATAGATGGTCAATGCGG-3′; Lgp2: forward, 5′-TCATCTGTACGGCAGAGTTGT-3′; reverse, 5′-TGTTGTAGACGGTGTCCTTGT-3′; and RIG-I: forward, 5′-AGAGAATTCGGCACCCAGAA-3′; reverse, 5′-AGCTCTCGCTCGGTCTCATC-3′.

The specificity of amplification was assessed for each sample by melting curve analysis, and the size of the amplicon was checked by electrophoresis. Relative quantification was performed using standard curve analysis. All gene expression data were normalized with β-actin and are presented as a ratio of gene copy number per 100 copies of β-actin ± SD.

SV induces IFN-αβ independently of TLRs or their early downstream signaling molecules

Our first goal was to determine whether SV-induced IFN-αβ was dependent on TLR signaling using mice with targeted deletions of TLR adapters. To date, four adapter molecules have been associated with TLR signaling. These include, MyD88, Mal, Toll/IL-1R domain-containing adaptor inducing IFN-β (TRIF), and TRIF-related adaptor molecule (TRAM). TLR7, -8, and -9 induce IFN via MyD88, whereas TLR3 signals via TRIF, and TLR4 via TRIF and the related adapter molecule, TRAM (4). Flt3 ligand-differentiated DCs from either wild-type or MyD88-deficient bone marrow (consisting of 40–50% plasmacytoid DC and 50–60% myeloid DC) were infected with SV or stimulated with CpG-A as a positive control (Fig. 1⇓A). Induction of IFN-α after CpG DNA treatment was abrogated in the absence of MyD88, in agreement with published reports (7). In contrast, SV induced normal levels of IFN-α in these cells. Similarly, the SV response was normal in Flt3-L-derived DCs from TRIF or TLR3-deficient mice (Fig. 1⇓B). As a control for these measurements, we monitored the production of RANTES, another IRF-3-regulated gene, after SV infection, or stimulation by ligands for the following TLRs: poly(I-C) (TLR3), LPS (TLR4), R-848 (TLR7), and two other viruses, HSV-1 (TLR9) and influenza (TLR7). As shown in Fig. 1⇓C, the induction of RANTES was impaired in TRIF-deficient cells after LPS and dsRNA signaling. TRAM-deficient cells were also defective after LPS treatment. MyD88-deficient cells were defective for CpG DNA, R-848, HSV-1, and influenza, consistent with published reports. Macrophages from C57BL/6, MyD88^−/−, Mal^−/−, TRIF^−/−, and TRAM^−/− mice produced similar levels of IFN-β after SV infection (Fig. 1⇓, D and E). For completeness we also examined TLR2-, TLR4-, and TLR9-deficient cells and in all cases SV induced comparable levels of IFN-α as seen in wild-type mice (data not shown). Recently, the E3 ubiquitin ligase TRAF6 has been implicated in the regulation of IRF-7 and IFN-α responses after CpG DNA treatment (31). We therefore monitored the induction of IFN-α in total splenocytes isolated from TRAF6^−/− mice and found normal IFN-α induction after SV infection (Fig. 1⇓F). Similar results were obtained in TRAF6, TRAF2, and TRAF2/5 double-knockout embryonic fibroblast cells (data not shown). Collectively, the data from Fig. 1⇓ (and the related data not shown) show that no single TLR or early downstream signaling molecule is required for the measured responses to SV infection.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

SV induces IFN-α via TLR-independent mechanisms. A and B, Flt-3L-derived DCs from C57BL/6, MyD88^−/−, TRIF^−/−, and TLR3^−/− mice were infected with SV (60 HAU) or were stimulated with CpG-A DNA (2 μg/ml) for 16 h as indicated, and IFN-α was measured in the supernatants by ELISA. C, Flt-3L-derived DCs from C57BL/6, MyD88^−/−, TRIF^−/−, and TLR3^−/− mice were infected with SV (60 HAU) and stimulated with LPS (20 ng/ml), poly(I-C) (20 μg/ml), R-848 (10 nM), CpG DNA 1668 (2 μg/ml), HSV1-KOS (2 × 10⁵ PFU/ml), or influenza (2 × 10⁴ PFU/ml) for 16 h, and RANTES was measured by ELISA. D, Splenocytes from C57BL/6 or the indicated knockout mice were mock infected or infected with SV (60 HAU), and IFN-α or IFN-β was measured by ELISA. Data are expressed as the mean ± SD. All experiments described were performed a minimum of three times and produced similar results.

To determine whether the induction of IFN-β after SV infection required active viral replication, we monitored SV-induced IFN-β reporter gene activation in HEK293 cells and compared live or replication-deficient SV. The Cantrell strain of Sendai virus was inactivated either by heating to 55°C for 30 min or through successive cycles of UV cross-linking, and failure to replicate was controlled as described in Materials and Methods. Either method of inactivation abolished the ability of the virus to replicate together with the stimulatory activity and its ability to induce an IFN-β reporter gene (Fig. 2⇓A). In addition, we monitored NF-κB reporter gene induction in our HEK cell line. Although SV is a much poorer inducer of this reporter than either the ISRE or the IFN-β reporter, we also found that the UV or heat-inactivated virus failed to induce this response. Furthermore, purified plasmacytoid DCs, which are extremely sensitive to SV, were stimulated with either the live or UV-inactivated virus and IL-12-p40 was measured. There was no IL12-p40 induced with the inactivated virus. Collectively, these observations demonstrate that live virus is required to induce an IFN-β or NF-κB response, suggesting the requirement for active viral replication.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

SV signaling is dependent on live virus. HEK293 cells expressing an IFN-β-luciferase reporter (A) or an NF-κB reporter (B) were infected with live, heat-inactivated, or UV-inactivated SV (6, 60, and 600 HAU); luciferase reporter gene activity was measured after 9 h and is depicted as fold induction relative to the reporter-only control. C, Purified plasmacytoid DCs (pDCs) from Flt-3L-derived C57BL/6 cultures were infected as indicated with 60 HAU of live or UV-inactivated SV for 16 h, and IL-12-p40 was measured by ELISA. Data are expressed as the mean ± SD. All experiments described were performed a minimum of three times and produced similar results.

Recognition of SV is mediated by RIG-I

The discovery of RIG-I, as a mediator of NDV signaling and intracellular dsRNA responses (19), prompted us to evaluate its contribution to the SV response. As shown in Fig. 3⇓A, infection of HEK293 cells with SV or NDV induces activation of an ISG54-ISRE, NF-κB, or IFN-β reporter gene. Transfection of a cDNA expressing full-length RIG-I synergistically enhanced the response to SV or NDV in each case (Fig. 3⇓A). Interestingly, in contrast to RIG-I, overexpression of Mda5 did not enhance SV-induced ISRE activation (Fig. 3⇓B), suggesting that Mda-5 is unable to mediate SV recognition. RIG-I contains tandem CARD domains in its N terminus and a COOH-terminal DExD/H box RNA helicase domain. Transfection of cells with RIG-IC (the isolated helicase domain) or RIG-I-K270A (a RIG-I construct that has a point mutation in the Walker-type ATP binding site, which is required for ATP-dependent RNA unwinding) dose-dependently inhibited SV- and NDV-dependent ISRE-reporter gene activation (Fig. 3⇓C). Importantly, RIG-IC and RIG-I-K270A had no effect on ISRE activation after TLR3 stimulation (Fig. 3⇓D). RIG-IC and RIG-I-K270A also inhibited SV-dependent activation of an NF-κB reporter gene (Fig. 3⇓E), suggesting that RIG-I controls both these arms of the innate immune response. Similar results were seen when SV-induced RANTES expression was measured by ELISA. Consistent with the reporter assays, full-length RIG-I weakly induced RANTES expression, an effect that was dramatically enhanced after SV infection. In contrast, both RIG-IC and RIG-I-K270A abolished SV-induced RANTES induction (Fig. 3⇓F).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

SV and NDV induce ISRE and NF-κB reporter activities via RIG-I. A–E, 293T cells or HEK293-TLR3 cells were transfected with the indicated reporter genes and cotransfected with full-length RIG-I, RIG-IC, or RIG-I-K270A. Cells were infected with SV (50 HAU) or NDV (50 HAU) or were treated with poly(I-C), where indicated, for 16 h, and luciferase reporter activities were measured. In all cases, data are expressed as fold induction relative to the reporter-only control and are the mean ± SD. F, 293T cells were transfected with full-length RIG-I, RIG-IC, or RIG-IK270A. Cells were infected with SV (50 HAU) for 16 h, and RANTES was measured by ELISA. All experiments described were performed a minimum of three times and produced similar results.

An RNA interference approach using shRNA vectors targeting human and mouse RIG-I also demonstrated a clear role for RIG-I in SV-induced signaling. The efficiency and specificity of silencing RIG-I were examined using a CFP-tagged fusion protein transfected into 293T cells. Efficient and specific silencing of RIG-I by the human shRNA targeting vector was seen by monitoring fluorescence by FACS analysis or immunoblotting for the Flag epitope (Fig. 4⇓B). Transfection of the shRNA targeting human RIG-I into HEK293 cells (human cells) greatly reduced SV-dependent IFN-β promoter activation, whereas shRNA targeting murine RIG-I, due to a lack of sufficient complementarity, had no effect (Fig. 4⇓C). The shRNA vector targeting murine RIG-I inhibited SV signaling in murine L929 cells, whereas the vector targeting human RIG-I was without effect. This species dependence emphasizes the specificity of the shRNAs. Collectively, the data from Figs. 3⇑ and 4⇓ clearly establish that the SV and NDV-dependent induction of IFN-β, ISRE, and NF-κB is mediated by RIG-I.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

RNA interference targeting RIG-I abrogates SV signaling. 293T cells were cotransfected with RIG-I-CFP together with the indicated shRNA vectors for 48 h (A). RIG-I-CFP silencing was quantified by FACS analysis. The percentage of CFP⁺ cells and the mean fluorescence intensity (MFI) of the CFP⁺ cells are indicated. B, RIG-I-Flag silencing was quantified by Western blot. Lane 1, 293T; lanes 2–7, 293T cotransfected with RIG-I-Flag (lane 2, control; lane 3, piGENE vector; lane 4, shRNA anti-mouse RIG-I; lane 5, shRNA anti-human RIG-I; lane 6, pSuper vector; lane 7, shRNA anti p53 HEK293). C, HEK 293 or L929 cells were transfected with an IFN-β-luciferase reporter gene and shRNA vectors targeting human or murine RIG-I; 72 h later, cells were infected with SV, and luciferase reporter gene activity was measured. In all cases, data are expressed as fold induction relative to the reporter-only control and are the mean ± SD. All experiments described were performed a minimum of three times and produced similar results.

Characterization of Lgp2, a RIG-I like molecule which negatively regulates SV and NDV signaling

A genome-wide search for proteins structurally related to RIG-I identified two additional DExD/H box RNA helicases: Mda-5 (melanoma differentiation-associated gene 5) (17) and Lgp2 (25). Interestingly, we had earlier detected both helicases in a genome-wide expression profiling study of LPS treated MyD88-deficient macrophages (32). Mda-5, originally characterized as a CARD containing protein inducible by IFN, contains two N-terminal CARD domains, followed by a DExD/H box-RNA helicase domain (17). Lgp2 contains a DExD/H box RNA helicase domain, highly homologous to that of RIG-I, but lacks CARD domains or indeed any other signaling domain (see schematic representation in Fig. 5⇓A). Notably, transfection of HEK293 cells with a cDNA expressing Lgp2 did not induce the NF-κB, ISRE, or the IFN-β reporter gene (Fig. 5⇓B). In contrast, transient expression of Mda-5, like RIG-I, activated all three reporters (even in the absence of virus infection). In contrast, NOD1 and NOD2, anti-bacterial CARD-containing proteins induced the NF-κB reporter, consistent with previous reports, but did not activate the ISRE or IFN-β reporters (Fig. 5⇓B). The isolated CARDs of RIG-I induced much higher levels of all three reporters than the full-length construct (data not shown) consistent with published reports (19). These data suggest that the CARDs of RIG-I and Mda-5 are functionally distinct from those of the NOD proteins and can initiate signaling to both ISRE- and NF-κB-dependent genes, whereas Lgp2, which lacks CARD domains, has no signaling capability.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Lgp2 acts as a negative regulator of SV-induced signaling. A, Schematic representation of Lgp2, RIG-I, Mda-5, NOD1, and NOD2, showing domain architecture. B, 293T cells were transfected with the indicated reporter genes and cotransfected with Lgp2, full-length RIG-I, Mda5, NOD1, or NOD2. Luciferase reporter activity was measured after 24 h. C and D, HEK-293-TLR3 cells were transfected with the indicated reporters and Lgp2 or eIF4A as indicated. Cells were infected with SV (50 HAU) or were treated with poly(I-C) (20 μg/ml) for 16 h, and luciferase activity was measured. E, 293T cells were transfected with Lgp2. Cells were infected with SV (50 HAU), and RANTES was measured in the supernatants by ELISA. In all cases, data are expressed as fold induction relative to the reporter-only control and are the mean ± SD. All experiments described were performed a minimum of three times and produced similar results.

We next examined the effect of Lgp2 on SV signaling. Transfection of cells with a cDNA expressing Lgp2 dose-dependently inhibited virus-induced ISRE activation (Fig. 5⇑C). Similar results were seen with NDV (not shown). This inhibitory effect was specific for the virus-induced pathway, because Lgp2 did not inhibit TLR3-dependent induction of the ISRE reporter. The distantly related DExD/H box RNA helicase, elF4A (which binds and unwinds RNA duplexes), did not affect either response, demonstrating the specificity for Lgp2. Lgp2 also dose-dependently inhibited virus-induced, but not TLR3-induced, NF-κB or IFN-β reporter activation (Fig. 5⇑D). Lgp2 also dose-dependently inhibited SV-induced RANTES production (Fig. 5⇑E). A catalytically inactive helicase mutant of Lgp2, in which the Walker-type ATP binding site was mutated to alanine (Lgp2-K30A), which, based on comparisons to other RNA helicases, should still bind, but not unwind, dsRNA, did not reverse the inhibitory effect of Lgp2 on either SV or NDV signaling (data not shown).

RIG-I and Mda5 bind dsRNA (poly(I-C)) via their RNA helicase domains and are activated by dsRNA (19, 20). CARDs, like other structurally related protein interaction modules found in innate immune adaptor molecules, such as death domains, TIR domains, and death effector domains, mediate homotypic interactions to promote signaling events (33). It is therefore possible that a RIG-I dimer/oligomer might be necessary for initiating downstream signaling. One possible explanation for the inhibitory activity of Lgp2 might relate to its ability to bind RIG-I via its helicase domain and prevent the formation or CARD-CARD interactions (because Lgp2 lacks a CARD domain). To test this hypothesis, we first examined the ability of Lgp2 to bind to RIG-I in coimmunoprecipitation experiments in the absence and the presence of SV infection. Lgp2 did not bind RIG-I or Mda-5 under these conditions (data not shown).

We next took advantage of the fact that full-length RIG-I can elicit a spontaneous induction of downstream signaling pathways (even in the absence of virus infection). We therefore investigated whether Lgp2 could block this constitutive response. Full-length RIG-I induced the ISRE reporter gene normally even in the presence of Lgp2 (Fig. 6⇓A). Treating the cells with SV in the presence of Lgp2 inhibited the response, consistent with Fig. 5⇑, C–E. These results suggest that Lgp2 blocks at a level upstream of RIG-I in SV signaling, most likely by sequestering dsRNA.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Lgp2 binds dsRNA. A, HEK293 cells were transfected with an ISRE-luciferase reporter gene together with full-length RIG-I and increasing amounts of Lgp2. Where indicated, cells were infected with SV in the absence of RIG-I as a control. B, Cell lysates from HEK293 cells transfected with RIG-I-YFP, Mal-CFP, or Lgp2-YFP were subjected to poly(I-C) pull-down assays as indicated. Poly(I-C)-agarose-associated complexes were subjected to SDS-PAGE and immunoblot analysis with anti-GFP. The expression of each protein before pulldown is shown by immunoblot with anti-GFP in the input lysate. C, Lysates from a HEK-Lgp2-YFP stable cell line were subjected to poly(I-C) or poly(C) pull-down. Where indicated, soluble poly(I-C) was added to cell lysates for competition analysis. In each case, immunoprecipitated complexes were subjected to SDS-PAGE and immunoblot analysis with anti-GFP (BD Clontech).

We therefore examined the ability of Lgp2 to bind dsRNA, as has been shown for RIG-I. Consistent with published reports, RIG-I binds dsRNA (Fig. 6⇑B). Under similar conditions, Lgp2 also bound dsRNA. As a control for these experiments, the TIR domain adapter, Mal, did not bind dsRNA. We also generated a stable cell line expressing Lgp2-YFP to further examine the ability of Lgp2 to bind dsRNA. As shown in Fig. 6⇑C, Lgp2 bound to poly(I-C), but not poly(C), beads. Preincubation of cell lysates with soluble poly(I-C) inhibited Lgp2 binding to dsRNA. Collectively, these data suggest that like RIG-I, Lgp2 is a dsRNA-binding protein.

Because Lgp2 appears to function as a negative regulator of virus-induced signaling, we wanted to compare expression levels of RIG-I and Lgp2 before and after infection with SV. Lgp2 and RIG-I mRNA levels were monitored by quantitative real-time PCR in murine macrophages (RAW264.7) in the first 12 h after infection (Fig. 7⇓). Notably, RIG-I was expressed at a much higher level than Lgp2 in resting cells. The expression of both helicases reached similar levels 4–6 h after infection. IFN-β levels also peaked ∼4 h after infection and dropped rapidly thereafter. This expression profile is consistent with a model in which Lgp2 functions as a postinduction repressor that provides a negative feedback signal to RIG-I/Mda5-mediated IFN induction.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Lgp2 and RIG-I are virus-inducible genes. Raw264.7 macrophages were infected with SV (60 HAU), and mRNA expression of RIG-I, Lgp2, IFN-β, and β-actin was measured by quantitative PCR. Data were normalized with β-actin and are presented as a ratio of gene copy number per 100 copies of β-actin.