MOV10 Provides Antiviral Activity against RNA Viruses by Enhancing RIG-I–MAVS-Independent IFN Induction

Cellular innate immunity is the first line of defense mounted by the host upon pathogen invasion. For virus infection this immunity is primarily mediated by type I IFNs and IFN-stimulated genes (ISGs), and provides immediate protection and shapes the subsequent adaptive immune response. Upon virus infection, viral nucleic acids are sensed by various innate immune receptors, such as retinoic acid–inducible gene I (RIG-I)-like receptors (RLRs) and TLRs to initiate IFN and ISG induction (1). The cytosolic DExD/H-box family helicases RIG-I and melanoma differentiation–associated gene 5 (MDA5) are crucial components for sensing cytoplasmic viral RNA resulting from RNA virus infection (2). Upon binding to viral RNA, RIG-I and MDA5, through their common adaptor protein mitochondrial antiviral-signaling protein (MAVS), trigger a signaling cascade. This signaling cascade leads to the activation of transcription factors such as IFN regulatory factor 3 (IRF3) and NF-κB, resulting in transcriptional induction of IFN (1, 3, 4). Although RLR signaling is primarily responsible for viral RNA sensing and IFN induction, there is still residual IFN induction upon specific RNA virus infection in the absence of the critical adaptor MAVS (5), indicating the existence of additional viral RNA sensor signaling.

DExD/H-box family helicases are involved in various cellular processes, such as nucleic acid metabolism, RNA interference (RNAi), and innate immunity. Besides RIG-I and MDA5, a number of helicases such as DDX3 and DHX9 have been implicated in the regulation of RLR signaling (6, 7). In contrast, DDX41, DDX60, DHX9, and DHX36 have been proposed to be involved in cytoplasmic DNA and DNA virus sensing (8). Some of these helicases have been shown to be targeted by various viral proteins (7), which supports their role in antiviral innate immunity. Because of their similarities with RIG-I, various DExD/H-box helicases have been examined for their importance in viral RNA sensing. However, none of the helicases with antiviral activity seemed to function independent of RIG-I–MAVS signaling. Further, for most of these helicases, mechanistic details regarding their involvement in antiviral innate immunity and knockout (KO) studies have yet to be described.

Moloney leukemia virus 10, homolog (MOV10) (mouse), a putative member of SF1 family helicase (9), was first identified as a novel Argonaute-associated protein and was shown to be involved in RNAi (10, 11). MOV10 orthologs Armitage in Drosophila, and SDE3 in plants, are also involved in RNAi (12, 13). MOV10 is known to bind a broad variety of RNA (14), and through its helicase activity participates in mRNA degradation and translation inhibition (15). In addition, MOV10 has been shown to inhibit HIV replication through several different mechanisms (16–19). It has also been shown to inhibit retroviral transpositions via its association with L1 ribonucleoprotein particle in processing (P-) bodies and stress granules (20, 21). However, given that MOV10 is induced by IFN, its role in regulating the replication of other RNA viruses besides HIV has been less clear and in some cases conflicting. MOV10 provides antiviral activity against hepatitis C virus in an ISG expression library screen (22). In contrast, MOV10 interacts with hepatitis delta virus Ag and enhances viral RNA replication (23).

In this article, we describe antiviral activity of MOV10 against RNA viruses, which is mediated through enhanced IFN induction. The physiological significance of the MOV10 antiviral activity is demonstrated by our finding that MOV10 is specifically targeted by viruses to evade the innate immune response. Mechanistically, this antiviral activity of MOV10 is independent of the known intracellular RNA-sensor signaling through RIG-I and MAVS, and required inhibitor of κB kinase (IKK) ε. Taken together, our findings provide a mechanistic basis for the antiviral properties of MOV10.

Human embryonic kidney (HEK) 293, 293T, HT1080 cells, and primary human foreskin fibroblasts (HFF) were cultured in DMEM (Lonza, Rockland, ME) supplemented with 10% FBS (Atlanta Biologicals, Lawrenceville, GA) and penicillin/streptomycin. Plasmids pcDNA3-FLAG-MOV10 and G681A/D682A (MOV10-GD) were generously provided by Dr. Vinay K. Pathak (Center for Cancer Research, National Cancer Institute). Plasmids pcDNA3-FLAG-MOV10-Q129A, Q869A, G529A/K530A/T531A (MOV10-GKT), and D645A/E646A (MOV10-DE) were created by site-directed mutagenesis (Genewiz, South Plainfield, NJ). pcDNA3-FLAG-MOV10-Q129A/Q869A (MOV10-DM) was generated using Q129A and Q869A plasmids as backbone. V5-MOV10 plasmid was generously provided by Dr. Yong-Hui Zheng (Michigan State University) (17). FLAG-MAVS, Myc coxsackievirus B (CVB)-3C^pro, and 3C^pro-C147A have been described previously (24). Encephalomyocarditis virus (EMCV) 3C^pro plasmid was generously provided by Dr. Takashi Fujita (25). Small interfering RNA (siRNA) against MOV10 (23), IKKε, and RIG-I were synthesized by Sigma-Aldrich (St. Louis, MO) and have been described previously (26). Control siRNA SMARTpool were obtained from Thermo Scientific (Waltham, MA). Lipofectamine RNAiMAX was obtained from Invitrogen (Grand Island, NY). Low m.w. polyinosinic:polycytidylic acid [p(I):p(C)] was from Invivogen (San Diego, CA). Transfection reagent XtremeGENE 9, protease inhibitors, and GFP Ab were purchased from Roche (Indianapolis, IN). FLAG Ab and anti-FLAG beads were from Sigma-Aldrich (St. Louis, MO). MOV10 Ab was purchased from Abcam. RIG-I and IKKε Abs were from Cell Signaling Technology (Danvers, MA). ISG56, ISG60, and actin Abs; GFP-tagged vesicular stomatitis virus (VSV); and Sendai virus C protein Abs have been described previously (27). Sendai virus Cantell strain (SeV) was purchased from Charles River Laboratories (Wilmington, MA). EMCV was from American Type Culture Collection (Manassas, VA). Respiratory syncytial virus (RSV) strain A2 was described previously (28). IFN-β in cellular supernatant was detected by Verikine-HS Human IFN β ELISA Kit (PBL Assay Science, Piscataway, NJ) according to the manufacturer’s protocol.

HEK293, 293T, or HT1080 cells were cotransfected with pcDNA3-FLAG-MOV10 or pcDNA3 vector control, selected with G418 (800 μg/ml), and collected as pool stable cells. The MOV10 stable expression was confirmed by immunoblotting using anti-FLAG Ab.

293T cells deficient in RIG-I (293T-RIG-I-KO) were previously described (27). All other KO cell lines were generated as follows: HEK293T cells were plated at a density of 2 × 10⁴ cells/well in a 96-well plate. The next day, clustered regularly interspaced short palindromic repeat (CRISPR) or transcription activator–like effector nuclease (TALEN) plasmids were transfected using GeneJuice transfection reagent (Merck Millipore) according to the manufacturer’s protocol. pRZ-mCherry-Cas9 and pLenti–guide RNA constructs were transfected at a ratio of 3:1 (i.e., 150:50 ng), whereas TALEN pair plasmids were transfected at a ratio of 1:1 (100 ng each). Critical exons of the following genes were targeted using the following guide RNA constructs (PAM region in bold letters): IRF3, 5′-GGGGGTCCCGGATCTGGGAGTGG-3′; IFN receptor (IFNAR1), 5′-ACAGGAGCGATGA GTCTGTCGGG-3′; MOV10, 5′-GTTCTTCAGACTCGACCGCTGGG-3′; IKBKE, 5′-GCACAATGCCGTTCTCCCGCAGG-3′. TANK-binding kinase 1 (TBK1)–deficient cells were created using a TALEN pair targeting the following region: 5′-TTCTAATCATCTGTGGCTTTTATCTGATATTTTAGGCCAAGGAGCTACTGCAA-3′ (29). Subsequently, limiting dilution cloning was performed and after 10 d, growing monoclones were selected by bright-field microscopy. Thus, identified clones were trypsinized and expanded in two separate wells. One well was used to recover gDNA as previously described (30), and subsequently the target region of interest was amplified in a two-step PCR and subjected to deep sequencing. KO cell clones were identified as cell clones harboring all-allelic frameshift mutations using OutKnocker (31). Genotypes of the respective KO cell lines are available upon request.

Total RNA was isolated from transfected and/or stimulated cells by TRIzol (Life Technologies) and treated with DNase I at 37°C for 1 h (DNA Free kit; Ambion). cDNA was synthesized using iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA). One part (1/20) of the cDNA synthesized from 1 μg RNA was subjected to real-time PCR using EvaGreen Supermix in a CFX96 Real Time System (Bio-Rad) according to the manufacturer’s instructions. All PCR amplification was normalized to ribosomal protein L32 (RPL32) and expressed as fold change with respect to untreated vector control cells, indicated with black diamonds (♦).

Cells were seed in 24-well plates and subsequently infected with EMCV- or VSV-expressing GFP at different multiplicity of infection (m.o.i.) depending on the assay. SeV infection was carried out at different doses ranging from 0.5 to 240 hemagglutinin unit (HAU)/ml. Postinfection, the cells were lysed and subjected to immunoblotting using Ab against SeV C protein or GFP protein. For one-step growth curve analysis, MOV10 and control cells were infected with different viruses including VSV or EMCV; cell-free medium was collected postinfection, and virus titers were determined by plaque assay using BHK21, and Vero cells, in 24-well plates. Each virus infection was performed in triplicate. The VSV, EMCV, and SeV replication were measured as viral RNA expression using quantitative RT-PCR (qRT-PCR) with the following specific primers: VSV, forward, 5′-GAGGAGTCACCTGGACAATCACT-3′, reverse, 5′-TGCAA GGAAAGCATTGAACAA-3′; EMCV, forward, 5′-TGCAGTGGTTGCTCCCCTGA-3′, reverse, 5′-TGACCGGAATGGGCGACTGT-3′; SeV, forward, 5′-GCTGCCGACAAGGTGAGAGC-3′, reverse, 5′-GCCCGCCATGCCTCTCTCTA-3; RPL32, forward, 5′-GCCAGATCTT GATGCCCAAC-3′, reverse, 5′-CGTGCACATGAGCTGCCTAC-3′; RSV forward, 5′- GCTCTTAGCAAAGTCAAGTTGAATGA-3′, reverse, 5′ TGCTCCGTTGGA TGGTGTATT-3′. IFIT1, IFT3, ISG15, Cig5, and IFN-β primers were described previously (27). All primers were custom-synthesized by Integrated DNA Technologies (Coralville, IA).

HEK293-MOV10 or vector cells were plated in eight-well chamber slides (4 × 10⁴ cells/well). Cells were infected with VSV for up to 16 h and fixed with paraformaldehyde (4% v/v), after permeabilization with Triton X-100 0.1%. Nuclei were stained with DAPI Vectashield (Vector Laboratories, Burlingame, CA). Human primary foreskin fibroblasts were plated in eight-well chamber slides (1 × 10⁴ cells/well) and transfected with siRNA using Lipofectamine RNAiMAX with either siRNA 10 nM or control siRNA for 72 h. Subsequently, cells were infected with EMCV for 16 h and fixed with paraformaldehyde (4% v/v), after permeabilization with Triton X-100 1%. Infected cells were stained using anti-dsRNA sera (J2), as previously described (27), for 16 h at 4°C. Immunofluorescence detection was carried out with conjugated anti-rabbit Alexa Fluor 488 and conjugated anti-mouse Alexa Fluor 594 (Invitrogen). Representative micrographs were obtained followed by quantitation of the percentage of infected cells by counting >500 DAPI⁺ cells for each condition.

Cells plated in 12-well plates (3 × 10⁵ cells/well), transfected, and/or treated with SeV were lysed in lysis buffer (Triton X-100 1%, HEPES [pH 7.4] 20 mM, NaCl 150 mM, MgCl₂ 1.5 mM, EGTA 2 mM, DTT 2 mM, NaF 10 mM, β-glycerophosphate 12.5 mM, Na₃VO₄ 1 mM, PMSF 1 mM, and protease inhibitor). The cleared cell lysates were incubated at 4°C with Ab and protein A/G agarose beads or anti-FLAG beads overnight, washed five times with lysis buffer, and boiled in 2× SDS-PAGE loading buffer for elution. Cell lysates boiled in 1× SDS-PAGE loading buffer and immunoprecipitated samples were subjected to SDS-PAGE and immunoblotting.

Cells were seeded in 12-well plates, and 6–24 h poststimulation, samples were harvested in 50 μl native lysis buffer [Tris-HCl 50 mM (pH 7.5), NaCl 75 mM, EDTA 1 mM, Nonidet P-40 1%]. Native gels (8% without SDS) were prerun with Tris 25 mM, glycine 192 mM, and deoxycholate 1% in the cathode chamber for 1 h at 100 V and 4°C. Before loading, equal volume of 2× sample buffer (Tris-HCl 125 mM [pH 6.8], glycerol 20%, BPB dye 0.1 mg/ml) was added to the samples. Subsequently, samples containing equal amounts of total proteins (20 μg) were electrophoresed for 180 min (100 V at 4°C) and immunoblotted for IRF3.

HEK293-MOV10, 293T, 293T-MAVS-KO, or 293T-RIG-I-KO (1.5 × 10⁵ cells/well) were seeded in 24-well plates and transfected using X-tremeGENE 9 as indicated. Genome-edited 293T were transfected with MOV10 or vector control plasmids (1 μg each) together with ISRE-luciferase (0.4 μg) and β-actin Renilla luciferase reporter (0.012 μg). Twenty four hours later, the cells from each well were collected by trypsin-EDTA digestion and seeded into 6 wells in 96-well plate. Forty-eight hours posttransfection, the cells were stimulated with SeV (10 HAU/ml for 16 h) or low m.w. p(I):p(C) transfection (1 μg/ml for 16 h), and luciferase activities were measured using the Dual-Glo luciferase assay system from Promega (Madison, WI). The results were expressed as fold induction of firefly luciferase relative to that of nonstimulated control-transfected cells after normalizing to Renilla luciferase, indicated with black diamonds (♦).

HEK293-MOV10 or vector cells stimulated with different concentrations of SeV (10–240 HAU/ml) and 16 h postinfection were washed and cell pellets were suspended in hypotonic buffer (HEPES 20 mM [pH 8.0], KCl 10 mM, MgCl₂ 1 mM, glycerol 20%, Triton-X 100 0.1%) with protease inhibitors (Roche). The cell suspensions (100 μl) were vortexed for 30 s, incubated on ice for 15 min, and centrifuged (16,000 × g for 10 min at 4°C). The supernatants were collected as soluble cytoplasmic fractions. The remaining nuclear pellets were thoroughly washed in 10 vol hypotonic buffer and then resuspended in 100 μl buffer [Tris-HCl 50 mM (pH 7.4), NaCl 150 mM, Nonidet P-40 1%, sodium deoxycholate 0.25%, EDTA 1 mM, PMSF 1 mM, protease inhibitor] and incubated in ice for 30 min before SDS-PAGE analysis.

Data were analyzed using two-tailed paired Student t test. Values were considered significant at p < 0.05, indicated by an asterisk symbol (*), and at p < 0.01, indicated by a double asterisk symbol (**).

We examined the antiviral activity of MOV10 against RNA viruses from two different families. HEK293 cells stably expressing MOV10 showed substantial suppression of the replication kinetics of VSV, a negative-strand RNA virus, compared with vector control cells as measured through viral RNA quantitation (Fig. 1A) and plaque assay (Fig. 1B). MOV10 expression also inhibited the replication of another negative-strand RNA virus, SeV (Fig. 1C). In addition to negative-strand RNA viruses, MOV10-expressing cells markedly restricted replication of positive-strand RNA viruses from the picornavirus family such as EMCV (Supplemental Fig. 1A, 1B) and CVB (Supplemental Fig. 1C). Taken together these results suggest that MOV10 has antiviral activity against certain RNA viruses.

Next, we validated the antiviral activity of MOV10 by silencing its expression in primary HFFs and subsequently infecting these cells with RNA viruses. Transfection of MOV10-specific siRNA enhanced VSV (Fig. 2A) and EMCV (Supplemental Fig. 1E) replication at least 2-fold compared with control siRNA-transfected cells. The extent of MOV10 silencing in HFFs by siRNA transfection is shown in Supplemental Fig. 1D. To further prove the antiviral function of human MOV10, we generated MOV10-deficient 293T cells (293T-MOV10-KO) through CRISPR-mediated genome editing. Compared with their wild-type (Wt) counterparts, MOV10-KO cells showed significantly higher levels of virus replication as measured by GFP expression (VSV, Fig. 2B, lanes 2 and 4, 2C). However, the extent of virus replication enhancement in MOV10-KO cells was not as remarkable for EMCV (at best 5-fold; Supplemental Fig. 1F) as the replication inhibition caused by the ectopic MOV10 expression (well >10-fold; Supplemental Fig. 1A). This indicated a possible amplification of the antiviral activity exerted by MOV10 expression, possibly through IFN signaling (see later).

MOV10 has in vitro directional helicase activity, which has been shown to be necessary for its ability to recruit nonsense-mediated mRNA decay factor UPF1 and promote cellular mRNA degradation (15). However, its RNA-binding property is independent of this helicase activity. To distinguish between different functional properties that were necessary for the MOV10 antiviral activity, we transiently expressed Wt and various helicase domain mutants of MOV10 in MOV10-KO cells and measured their antiviral activities against EMCV and VSV. Two mutants targeting the conserved residues in helicase Motif I, G529A/K530A/T531A (MOV10-GKT), and Motif II, D645A/E646A (MOV10-DE), showed similar antiviral activity compared with the Wt (Fig. 2D). These mutations are known to inactivate the catalytic activity of MOV10 without affecting its RNA binding (15). Therefore, the helicase activity of MOV10 was not necessary for the antiviral activity against VSV. It also indicates that the ability of MOV10 to promote UPF1-mediated mRNA decay is not important to inhibit the replication of these RNA viruses. Another mutant of MOV10, known to be defective in inhibiting HIV (G681A/D682A, MOV10-GD) (16), also showed similar antiviral activity to Wt (Fig. 2E), indicating that the antiviral activity of MOV10 against VSV is independent of its anti-HIV activity. Taken together these results indicate that similar to a number of ISGs, MOV10 provides antiviral activity against a number of RNA viruses, and that this antiviral activity is independent of the helicase and anti-HIV activity of MOV10.

As stated earlier, the strong antiviral activity observed in MOV10-expressing cells (Fig. 1A) prompted us to investigate the role of IFN signaling, which may work in a feed-forward manner to amplify the antiviral activity. We stably expressed MOV10 in human HT1080-derived U3A cells, which were defective in STAT1 expression and IFN signaling (32). As expected, U3A cells showed loss of IFIT1 (also known as ISG56) induction after IFN treatment compared with control parental 2fTGH cells (Fig. 3A, lanes 3 and 4). MOV10 expression did not affect IFIT1 induction, indicating that MOV10 did not influence IFN signaling downstream of STAT1. However, MOV10 expression in U3A cells failed to provide protection against SeV infection (Fig. 3B). To further confirm these results using a different system, we created IFNAR1-deficient 293T cells (293T- IFNAR1-KO) that were defective in IFN signaling (Supplemental Fig. 2A), using CRISPR-mediated genome editing. Transient expression of MOV10 in 293T-IFNAR1-KO cells showed no reduction in subsequent EMCV replication, whereas the control 293T cells showed reduced virus replication (Fig. 3C). These results, in two different systems, indicated that an intact JAK-STAT–mediated IFN signaling was necessary for the inhibition of virus replication by MOV10.

Having established the necessity of an intact IFN signaling for the MOV10 antiviral activity, we examined whether the absence of MOV10 affected JAK-STAT–mediated IFN signaling by assaying the induction of a number of ISGs after IFN treatment in 293T-MOV10-KO cells. However, 293T-MOV10-KO cells did not exhibit any defect in ISG induction (Fig. 3D, Supplemental Fig. 2B), indicating that MOV10 did not directly affect JAK-STAT–mediated IFN signaling. Thus, we focused on the upstream events of this cascade, IRF3-mediated IFN induction, and examined the role of MOV10 on IRF3 signaling using 293T-IRF3-KO cells (Supplemental Fig. 2C). As shown in Fig. 3E, transient expression of MOV10 showed the expected inhibition of SeV replication in Wt 293T cells. However, this inhibition was absent in MOV10-expressing 293T-IRF3-KO cells, which suggested that the IRF3-mediated IFN induction was necessary for the inhibition of virus replication by MOV10. Altogether these results indicated that the antiviral activity of MOV10 is mediated through IRF3 signaling by possibly affecting IFN induction.

Because the antiviral activity of MOV10 was dependent on IRF3, we next examined the effect of MOV10 on IRF3 activation. We used SeV infection and low m.w. p(I):p(C) transfection, both of which causes RIG-I–dependent IRF3 activation (4). In response to both stimuli, MOV10 enhanced IRF3 dimerization kinetics in a dose-dependent manner (Fig. 4A, 4B, lanes 5 and 6 compared with lanes 2 and 3, respectively). The nuclear translocation of IRF3 after its dimerization was also increased in MOV10-expressing cells (Fig. 4C, lanes 11 and 12 compared with lanes 8 and 9). Conversely, silencing of MOV10 with siRNA reduced SeV-mediated IRF3 dimerization (Fig. 4D, lanes 8–10 compared with lanes 3–5). This enhanced IRF3 activation in the presence of MOV10 also resulted in enhanced IFN-β induction. MOV10-expressing cells showed significant increase in the kinetics and the amplitudes of endogenous IFN-β mRNA induction (Fig. 4E). Further, SeV-mediated IFN-β protein induction was reduced in 293T-MOV10-KO cells, whereas it was enhanced upon ectopic expression of MOV10 in these cells (Fig. 4F). These results suggested that MOV10 antiviral activity could be potentially mediated through IRF3 activation and IRF3-mediated IFN induction.

RLRs are the primary cytosolic receptors that sense viral RNA and initiate the signaling cascade leading to IRF3 activation. Our observation that MOV10 provided antiviral activity against both EMCV and SeV, which are sensed by MDA5 and RIG-I, respectively, indicated that it either worked in conjunction with or independent of RLR-mediated sensing. Thus, in the next series of experiments, we investigated the role of RIG-I and its adaptor MAVS in mediating MOV10 antiviral activity. First, we determined the effect of MOV10 expression on SeV replication in 293T-derived RIG-I-KO (also known as DDX58) cells created by genome targeting (29). As expected, loss of either RIG-I or MAVS markedly increased SeV replication in control vector-expressing cells (Fig. 5A). Surprisingly, transient expression of MOV10 inhibited SeV replication in control 293T, as well as in both RIG-I-KO and MAVS-KO cells (Fig. 5A). Further, the presence of MOV10 showed enhancement of IRF3 dimerization in MAVS-KO cells (Supplemental Fig. 3A, lanes 7 and 8 compared with lanes 5 and 6) indicating that MOV10 enables IRF3 activation even in the absence of MAVS. These results suggested that the antiviral activity of MOV10 is independent of RIG-I and MAVS. Indeed, MOV10 expression in MAVS-KO cells showed significant enhancement in IFN-β protein induction (Fig. 5B). However, this finding that antiviral activity of MOV10 is independent of RIG-I–MAVS and may operate in parallel would predict that the antiviral activities of these two pathways should have additive antiviral effects. We examined this prediction in primary human fibroblasts, where VSV replications were further enhanced when MOV10 and RIG-I were silenced together compared with their individual silencing (Fig. 5C, Supplemental Fig. 3B). Finally, we created double-deficient cells by genetically targeting MOV10 and RIG-I simultaneously (Fig. 5D). Compared with RIG-I-KO cells, RIG-I-KO/MOV10-KO cells were substantially more permissive to VSV replication (Fig. 5E). Altogether, these results indicated that the MOV10 antiviral activity operated independent of and in parallel to RIG-I and MAVS to provide additional host defense against RNA viruses through IFN induction.

After engagement of various TLR and RLR signaling, IRF3 is activated by IKK family kinases TBK1 and IKKε through Ser/Thr phosphorylation. Because IRF3 was essential for the MOV10 antiviral activity, we next investigated the roles of these two kinases in MOV10-mediated activation of IRF3 and antiviral activity. Once again, we used genome editing to target IKKε (IKBKE) and TBK1 in 293T cells. Comparison of the VSV replication in the vector-transfected samples from these three cell lines showed substantial increase in VSV replication only in TBK1-KO, but not in IKKε−KO cells (Fig. 6A), indicating the critical role of TBK1 in the antiviral activity. As expected, MOV10 expression inhibited VSV (Fig. 6A) replication in control 293T cells. Significant antiviral activities were also seen with MOV10 expression in TBK1-KO cells (Fig. 6A). However, MOV10 expression failed to protect the IKBKE-KO cells against VSV (Fig. 6A). This indicated that MOV10-mediated antiviral activity is most likely mediated through IKKε and not through TBK1. Involvement of IKKε was further established by examining the physical interaction of MOV10 and IKKε. In coimmunoprecipitation assays, IKKε coprecipitated with MOV10 in an SeV infection–dependent, but RNA-independent manner (Fig. 6B). In addition, we were also able to coprecipitate exogenously expressed (Supplemental Fig. 3C) as well as endogenous IKKε (Supplemental Fig. 3D) with endogenous MOV10 post SeV infection. Taken together, these results suggest involvement of IKKε as the downstream kinase for MOV10-induced IRF3 activation and antiviral activity.

Targeting of the host proteins involved in innate immune response is an evolutionarily conserved mechanism among many RNA viruses, which helps evade the host response (33, 34). Examination of this feature has allowed identification of physiological importance of a number of host proteins involved in the host protection (35). Picornaviruses accomplish this by using the virally encoded proteases to cleave crucial components of the RLR and IFN signaling pathways (36). To find out whether these viruses also target MOV10, we examined the amounts of endogenous MOV10 protein during two different types of picornavirus infection: EMCV and CVB. Endogenous MOV10 protein gradually decreased during EMCV infection (Fig. 7A). Similar reductions in MOV10 protein amounts were observed with CVB infection in a viral dose-dependent manner (Fig. 7B). To further investigate this viral antagonism of MOV10, we examined MOV10 degradation by viral proteases. Because CVB 3C^pro is known to target and degrade a variety of innate immune signaling proteins (24), we tested whether CVB 3C^pro expression also affected the steady-state levels of MOV10. Transfection of CVB 3C^pro in 293T cells induced a reduction in the total levels of both endogenous (Fig. 7C) and ectopically expressed (Supplemental Fig. 4A) MOV10 and MAVS, a known target of 3C^pro degradation (24). In contrast, expression of a catalytically inactive 3C^pro mutant (C147A) did not induce degradation (Fig. 7C). Expression of the 3C^pro from EMCV also showed a similar targeting of MOV10, signifying its importance in promoting innate immunity against these viruses (Supplemental Fig. 4B). Analysis of MOV10 primary sequence identified two glutamine residues (Q129 and Q169) that could serve as possible cleavage sites for the picornavirus 3C proteases (37, 38). To identify the 3C^pro target site(s) in MOV10, we created point mutants of MOV10 using site-directed mutagenesis. As shown in Supplemental Fig. 4C, both single mutants of MOV10 showed partial protection from 3C^pro-mediated loss of expression. However, coexpression of the double-mutant Q129A/Q869A (MOV10-DM) with both CVB (Fig. 7C) and EMCV (Supplemental Fig. 4B) 3C^pro showed almost complete protection from the 3C^pro-mediated degradation. Indeed, when we tested the functional activity of this mutant in MOV10-KO cells by ectopically expressing MOV10, the Q129A/Q869A mutant showed substantially higher antiviral activity against EMCV compared with the Wt MOV10 (Fig. 7E). Together, these results further support the notion that MOV10 is a physiologically important host protein that enhances in innate defense against certain RNA virus infections.

Various helicases have been known to participate in innate antiviral immunity, with RIG-I and MDA5 being the most important ones for RNA virus–mediated IFN induction. However, a number of these helicases with antiviral activity either positively or negatively regulate RLR signaling. In this study, using a series of genome-edited human cell lines, we provide evidence that MOV10 enhances IFN induction to inhibit viral replication in a unique RLR-independent pathway. The observation that VSV replication is significantly increased in RIG-I-KO, MOV10-KO double-deficient cells compared with only RIG-I–deficient, RIG-I-KO cells, provides further evidence for this independence and existence of a parallel MOV10-mediated antiviral signaling pathway against specific viruses. Similar results in primary fibroblasts using MOV10 and RIG-I silencing provided support for this notion in a physiological context. However, the role of MOV10 in modulating TLR-mediated IFN induction is not yet clear.

Our finding of MOV10 targeting by viruses further substantiates the physiological significance of this phenomenon. Both EMCV and CVB effectively targeted MOV10 for degradation presumably using the respective 3C^pro viral proteases. Similar to some of the previous studies, we did not detect specific cleavage fragments of MOV10 (39), but the protection of the MOV10 Q129A/Q869A mutants from cleavage and the resulting inhibition of virus replication suggest specific targeting of MOV10 by these viruses. Although our results indicate that for SeV and VSV, the RLR pathway may be dominant over the MOV10 pathway, it is possible that for other viruses that target RLR pathway, MOV10 might be important.

Similar to other helicase family members, MOV10 is a multifunctional protein and has been implicated in a diverse range of cellular functions including RNA silencing, mRNA translation, and tumor suppression (10, 11, 15, 23, 40, 41). It is also known to restrict HIV replication and retrotransposon mobility (17, 18, 20). Some of these functions are dependent on different functional properties of MOV10, such as the helicase activity and P-body localization (15, 16, 20). In this article, we describe another unique function of MOV10: antiviral activity against a number of RNA viruses that is independent of its helicase and anti-HIV activity. Although MOV10 has been shown to promote mRNA degradation by associating with UPF1, this activity requires the helicase activity of MOV10. Our demonstration that the helicase mutants of MOV10 provide antiviral activity indicates that this is independent of the UPF1-mediated RNA degradation activity of MOV10. The loss of this antiviral activity of MOV10 in either IRF3- or IFNAR1-deficient cells further supports its role in IFN-ISG–mediated inhibition of virus replication. Mechanistically, MOV10-signaling pathway specifically uses IKKε as the possible mediator kinase for IRF3 activation (Supplemental Fig. 4D).

Although we define how MOV10 can promote IFN induction, its activation mechanism remains unknown. Similar to MDA5 and RIG-I, MOV10 expression by itself could promote IFN induction, but the biochemical activation mechanism of MOV10 remains to be determined. The helicase mutants we used are known to destroy MOV10 RNA unwinding activity, but retain its RNA binding activity (15). Thus, the nature of the RNA that activates MOV10, either uniquely or in addition to RLR, is not yet clear. The presence of MOV10 in complexes with IFIT proteins, known for their role in the detection of 5′-triphosphate containing uncapped as well as capped viral RNA lacking 2′O methylation (42, 43), suggest that MOV10 can either directly or together with IFIT proteins bind to viral RNA to promote IFN induction. Although we did not detect any contribution of IFIT1 and IFIT3 in MOV10-mediated ISG induction by coexpression (data not shown), the biochemical basis of self nonself discrimination of RNA by MOV10 requires further clarifications. In this regard, the unique localization of MOV10 in cytoplasmic P-bodies may play an important role. In summary, this study establishes MOV10-mediated IFN induction as an antiviral signaling mechanism.

We thank Dr. Vinay K. Pathak (National Cancer Institute) and Dr. Yong-Hui Zheng (Michigan State University) for sharing crucial MOV10 Wt and mutant expression plasmids, Dr. Takashi Fujita (Kyoto University) for the EMCV 3C^pro plasmid, and Dr. Jennifer Bomberger for the RSV stock and primers.

The authors have no financial conflicts of interest.