Host Protein Moloney Leukemia Virus 10 (MOV10) Acts as a Restriction Factor of Influenza A Virus by Inhibiting the Nuclear Import of the Viral Nucleoprotein

ABSTRACT

The viral ribonucleoprotein (vRNP) complex of influenza A viruses (IAVs) contains an RNA-dependent RNA polymerase complex (RdRp) and nucleoprotein (NP) and is the functional unit for viral RNA transcription and replication. The vRNP complex is an important determinant of virus pathogenicity and host adaptation, implying that its function can be affected by host factors. In our study, we identified host protein Moloney leukemia virus 10 (MOV10) as an inhibitor of IAV replication, since depletion of MOV10 resulted in a significant increase in virus yield. MOV10 inhibited the polymerase activity in a minigenome system through RNA-mediated interaction with the NP subunit of vRNP complex. Importantly, we found that the interaction between MOV10 and NP prevented the binding of NP to importin-α, resulting in the retention of NP in the cytoplasm. Both the binding of MOV10 to NP and its inhibitory effect on polymerase activity were independent of its helicase activity. These results suggest that MOV10 acts as an anti-influenza virus factor through specifically inhibiting the nuclear transportation of NP and subsequently inhibiting the function of the vRNP complex.

IMPORTANCE The interaction between the influenza virus vRNP complex and host factors is a major determinant of viral tropism and pathogenicity. Our study identified MOV10 as a novel host restriction factor for the influenza virus life cycle since it inhibited the viral growth rate. Conversely, importin-α has been shown as a determinant for influenza tropism and a positive regulator for viral polymerase activity in mammalian cells but not in avian cells. MOV10 disrupted the interaction between NP and importin-α, suggesting that MOV10 could also be an important host factor for influenza virus transmission and pathogenicity. Importantly, as an interferon (IFN)-inducible protein, MOV10 exerted a novel mechanism for IFNs to inhibit the replication of influenza viruses. Furthermore, our study potentially provides a new drug design strategy, the use of molecules that mimic the antiviral mechanism of MOV10.

INTRODUCTION

Influenza A viruses (IAVs), which belong to Orthomyxoviridae family, cause acute respiratory disease in humans and are responsible for periodic human pandemics as well as seasonal influenza (1–3). The genome of IAV consists of eight negative single-stranded RNA segments, each of which is coiled by multiple nucleoprotein (NP) molecules and associated with one RNA-dependent RNA polymerase (RdRp) complex (4). Three viral proteins—polymerase basic protein 1 (PB1), PB2, and polymerase acidic protein (PA)—constitute a heterotrimeric complex to form the core components of viral polymerase complex (4, 5). The RdRp complex is further associated with NP and viral RNA to form the viral ribonucleoprotein (vRNP) complex, which is the minimal functional unit for mediating viral mRNA transcription and RNA-dependent RNA replication. Especially, the primary transcription is primer dependent, and the primers for mRNA transcription of IAV are obtained via “cap-snatching” of host cellular mRNAs (6–8). During the transcription and replication of IAV, PB1, PB2, and PA are responsible for cap-snatching, transcription initiation, and elongation (9–13). After IAV infection, vRNP complexes are released from endosomes and transported into the nucleus, where viral transcription and replication take place. The vRNP complex utilizes the nuclear localization signals (NLSs) on NP for nuclear import through the cellular importin-α/β-dependent nuclear import pathway (14–17). It has been reported that vRNP complexes are shuttled between the nucleus and cytoplasm during viral replication (18, 19). During the late stage of infection, the newly assembled vRNP complexes are transported out of the nucleus in the guise of nuclear export protein (NEP), matrix protein (M1), and NP (20).

The successful completion of the viral life cycle relies on host factors and processes (21–25). Proper function of vRNPs is key for the IAV life cycle and important for viral pathogenicity and host range determinants. Numerous host proteins have been reported to be potential interacting partners of the vRNP complex (4, 26–30). Some host factors have already been elucidated to inhibit influenza virus replication at multiple levels. The interferon (IFN)-inducible protein Mx1 was reported to impede vRNP complex assembly (29), and RNA helicase DDX21 was shown to restrict influenza virus replication at several stages, including its inhibitory effect on vRNP assembly (27). However, with the advent of more advanced technologies, it is important to uncover other host proteins that interact with vRNP, which could ultimately lead to the development of novel anti-IAV strategies. In this study, we purified vRNP complexes by tandem affinity purification using PB2 as bait and identified the associated host proteins by mass spectrometry (MS). We found that Moloney leukemia virus 10 (MOV10) was one of the vRNP-associated proteins. Knockdown of MOV10 increased the growth rate and polymerase activity of IAV, while overexpression of MOV10 reduced the polymerase activity. Further studies indicated that MOV10 bound to NP, led to the delay of nuclear import of NP, and subsequently resulted in the inhibition of viral replication. Our results strongly suggest that MOV10 acts as a novel IAV restriction factor by hindering the nuclear import of NP.

MATERIALS AND METHODS

Cells and viruses.

Human lung carcinoma A549 cells, human embryonic kidney 293T cells, and the Madin-Darby canine kidney (MDCK) cells were maintained in Dulbecco's modified Eagle medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% penicillin-streptomycin at 37°C with 5% CO₂.

Influenza A/PR/8/34 (H1N1) virus and A/Guangdong/1/2009 (H1N1) virus were used. Influenza viruses were rescued from cDNAs and grown in 10-day-old specific-pathogen-free (SPF) embryonated chicken eggs at 37°C for 48 h. Allantoic fluid containing viruses was collected, aliquoted, and stocked at −80°C until used. Virus titers were determined by plaque assays in MDCK cells.

Plasmids and antibodies.

The expression plasmids for full-length hemagglutinin (HA)-tagged or FLAG-tagged MOV10, the MOV10 DEAG-box mutant (MOV10-E/Q-FLAG)-expressing plasmid, and a series of deletion mutants of MOV10 were constructed as described previously (31, 32). The coding sequences for four HA-tagged importin-α isoforms were amplified from a cDNA library of human 293T cells by PCR and cloned into pcDNA3.1 vector at BamHI and XhoI sites.

RNA was purified from the virions of H1N1pdm strain A/Guangdong/1/2009 to generate cDNAs for eight fragments (PB1, PB2, PA, NP, HA, NA, M, and NS). The eight fragments of influenza A/PR/8/34 (H1N1) virus were synthesized by Invitrogen. All fragments from both H1N1pdm virus and A/PR/8/34 virus were cloned into pHW2000 vector separately with restriction enzyme BsmBI or BsaI (NEB) (33).

The dually tagged pcDNA3.1 plasmid encoding PB2 and singly tagged or nontagged pcDNA3.1 encoding PB1, PB2, PA, and NP were derived from influenza A/PR/8/34 virus and cloned into pcDNA3.1 vector separately. pPolI-Luci-T was constructed by placing the coding sequence of firefly luciferase in a negative-sense orientation flanked with influenza virus cRNA promoter ends (33, 34). pTK-RL, expressing Renilla luciferase under the control of a cytomegalovirus (CMV) promoter, was used as internal control to normalize the transfection efficiency (31). pPolI-GFP-T was generated by replacing the luciferase sequence of pPolI-Luci-T with the coding sequence of gfp.

The antibodies used in our study included anti-MOV10 (rabbit polyclonal; Abcam), anti-HA (mouse monoclonal; MBL), anti-FLAG (rabbit polyclonal; MBL), anti-His (mouse monoclonal; CST), anti-glyceraldehyde-3-phosphate dehydrogenase (anti-GAPDH; rabbit polyclonal; Proteintech), anti-NP (rabbit polyclonal; Abcam), anti-lamin B1 (rabbit polyclonal; Proteintech), anti-M1 (rabbit polyclonal; Sino Biological, China), and anti-p65 (mouse monoclonal, Cell Signaling).

Generation of recombinant influenza A viruses.

The influenza A/PR/8/34 (H1N1) virus and A/Guangdong/1/2009 (H1N1) virus were rescued using an eight-plasmid-based reverse-genetics system (35–37). Plasmids of PB1, PB2, PA, NP, HA, NA, M, and NS on the pHW2000 vector were used for transfection. Briefly, 293T cells were seeded in 6-well plates and transfected with 0.5 μg each of the 8 plasmids of the virus rescue system. At 6 h posttransfection (p.t.), the medium was replaced with fresh DMEM containing 1 μg/ml of tosylsulfonyl phenylalanyl chloromethyl ketone (TPCK)-treated trypsin (Sigma) and 0.2% bovine serum albumin (BSA; Sigma). After the cells were further grown for 48 h, the medium containing the recombinant viruses was collected and centrifuged at 5,000 × g for 5 min. The supernatant was inoculated into 10-day-old SPF chicken eggs and cultured for 48 h at 37°C to prepare virus stock. The recombinant viruses were confirmed by a hemagglutination inhibition (HI) test and sequencing.

Recombinant firefly luciferase influenza virus (PR8-Luc virus).

The coding region for the viral HA protein was replaced with that of firefly luciferase together with the packaging signals for the HA segment. The recombinant PR8-Luc virus was generated by reverse genetics in the presence of complementing HA as previously described (21).

RNA interference (RNAi).

The sequences of MOV10-specific small interfering RNAs (siRNAs) (SMARTpool) were described previously (31). MOV10-specific siRNAs and control siRNA were synthesized by Ribobio (Guangzhou, China). The desired siRNAs were transfected into A549 or 293T cells using Lipofectamine RNAiMAX (Invitrogen) for 24 to 36 h, followed by virus infection.

Depletion of 7SL RNA.

Human 293T cells were transfected with SRP14-short hairpin RNA (shRNA) or gfp-shRNA, followed by puromycin selection. The cells were placed into 60-mm plates and further transfected with MOV10-HA- and NP-FLAG-expressing plasmids. Procedures were followed as previously described (31).

Influenza A virus minigenome system for polymerase activity.

Human 293T cells were seeded onto 24-well plates at 5 × 10⁴ cells per well. After 12 to 24 h, cells were transfected with the expression plasmids pHW2000 or pcDNA3.1-PB1, -PB2, -PA, and -NP (50 ng each), luciferase reporter pPolI-Luci-T (10 ng), and pTK-RL (5 ng) using Lipofectamine 2000 (Invitrogen). Cells were lysed and analyzed with a dual-luciferase reporter assay kit according to the instructions of the manufacturer (Promega).

Growth curve analysis and plaque assay.

After transfected with siRNAs or plasmids, cells were infected with influenza A/PR/8/34 (H1N1) virus at a multiplicity of infection (MOI) of 0.001 PFU (A549 cells) or 0.1 PFU (293T cells) (37). At 1 h postinfection (p.i.), the medium was replaced with DMEM containing 0.5% BSA and 1 μg/ml of TPCK-trypsin. Supernatants were collected at 6, 12, 24, 48, and 60 h p.i. for measuring virus titers. The virus titers were determined by plaque assay on MDCK cells. MDCK cells were seeded in 12-well plates and used for infection when the cells were grown to 100% confluence. The cells were washed with phosphate-buffered saline (PBS) once and infected with a series of dilutions of viruses for 1 h at 37°C with 5% CO₂. After the virus inocula were removed, the cells were washed with PBS. The cells were overlaid with agarose medium (DMEM containing 0.6% BSA, 2 μg/ml of TPCK-trypsin, and 1% low-melting-point agarose [Sigma]). The plates were settled at 4°C for 5 to 10 min until the agarose medium became solid, followed by culture upside-down at 37°C, and then the cells were cultured for 48 to 72 h. Visible plaques were counted and the virus titers were determined. The data are shown as means ± standard deviations (SD) from three independent experiments.

Sample preparation for mass spectrometry analysis.

To purify vRNP complexes, tandem affinity purification (TAP) was applied by following liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis as described previously (31). The TAP assay was performed with a FLAG HA tandem affinity purification kit (Sigma; TP0010) according to the instructions of the manufacturer.

Mass spectrometry analysis.

The stained bands of interest were excised into gel slices with a clean scalpel. Samples were digested with trypsin using in-gel digestion. Each gel piece was diced into small (1-mm³) pieces and washed with acetonitrile. Gel pieces were dehydrated and incubated with 10 mM dithiothreitol (DTT) in 25 mM NH₄HCO₃ for 1 h at 56°C. The supernatant was removed and the gel pieces were incubated with 55 mM iodoacetamide in 25 mM NH₄HCO₃ for 45 min at room temperature. Gel pieces were washed with 25 mM NH₄HCO₃ and then with 25 mM NH₄HCO₃–50% acetonitrile and were then dehydrated. The gel pieces were incubated with 0.5 μg of trypsin (Promega) in 50 mM NH₄HCO₃ at 37°C for 16 h. The digestion was stopped by adding 1% trifluoroacetic acid (TFA). Peptides were extracted with 50% acetonitrile–5% formic acid, lyophilized in a SpeedVac (Thermo Savant), and then desalted using u-C18 Ziptip (Millipore). Samples were lyophilized and stored at −20°C prior to analysis by LC-MS/MS or dissolved in 0.1% (vol/vol) formic acid-water. All samples were analyzed on a Thermo Scientific Q EXACTIVE mass spectrometer coupled with an EASY n-LC 1000 liquid chromatography (ThermoFisher) system and a nanoelectrospray source (38).

Western blotting.

Cells were lysed with ice-cold lysis buffer for 30 min at 4°C. The lysates were collected and separated by SDS-PAGE. The bands were immunoblotted with the desired antibodies and IRDye secondary antibodies (LI-COR) and visualized with the Odyssey infrared imaging system (LI-COR). The relative protein expression level was analyzed using the software of the Odyssey system.

Coimmunoprecipitation (co-IP) assay.

Human 293T cells were seeded in 6-cm dishes and transfected with various plasmids as indicated below. After 48 h, cells were lysed with 450 μl of ice-cold lysis buffer. About 10% (40 μl) of the lysates was taken as an input control. The remaining lysates were incubated with anti-HA agarose beads for 4 h at 4°C. The beads were then washed three times with 500 μl of ice-cold lysis buffer, followed by Western blotting (31, 39, 40). For the detection of endogenous MOV10, A549 cells were infected with influenza A/PR/8/34 (H1N1) virus (MOI = 1 PFU). At 24 h p.t., the cells were collected and lysed with 500 μl of ice-cold lysis buffer. The lysates were precleared with protein A/G agarose beads for 30 min and then incubated with anti-NP antibody or mouse normal IgG antibody and protein A/G agarose beads (Millipore) for 4 h at 4°C. The beads were then washed three times with 500 μl of ice-cold lysis buffer, followed by Western blotting with staining with anti-NP and anti-MOV10 antibodies.

FPLC analysis.

Fast protein liquid chromatography (FPLC) analysis was performed as previously described (41). In brief, cells were lysed and clarified, subsequently loaded onto a Superose 6 (GE Healthcare) column, and fractionated at a flow rate of 400 μl/min. The resulting fractions were assessed by immunoblotting after ethanol precipitation.

IFAs.

Immunofluorescence assays (IFAs) were performed as previously described (39). Briefly, 293T cells were fixed with 4% paraformaldehyde, permeabilized with 1% Triton X-100, blocked with 5% BSA blocking solution, and stained with primary antibodies and secondary antibodies. Cell nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) reagent (Invitrogen).

Nuclear and cytoplasmic protein fractionation.

Human 293T cells transfected with the desired plasmids were harvested and washed with PBS. Fractionation of cytoplasmic and nuclear components was performed according to the manufacturer's instructions (Millipore, Paris, France) (39).

Statistical analysis.

A two-tailed Student t test was used to determine the significance of statistical data.

RESULTS

Isolation of influenza A virus vRNP complex and identification of MOV10 as an inhibitor of influenza A virus replication.

In order to identify additional cellular proteins involved in the regulation of viral RNA synthesis, we employed tandem affinity purification (TAP) (31) followed by LS-MS/MS in our study. We used dually tagged PB2 (HA tag at the N terminus and FLAG tag at the C terminus; HA-PB2-FLAG) as bait in the cells cotransfected with plasmids for other vRNP subunits, including PB1, PA, and NP. As shown in Fig. 1A, dually tagged PB2 functioned similarly to plasmid-nontagged PB2 in the minigenome system. Additionally, dually tagged PB2 could be copurified with PB1, PA, and NP in an immunoprecipitation assay (Fig. 1B). These results implied that the presence of an HA tag at the N terminus and FLAG tag at the C terminus did not affect the function of PB2 and the efficiency of RNA synthesis in the minigenome system. Human 293T cells were then transfected with the plasmids for dually tagged PB2, PB1, PA, and NP. At 48 h posttransfection (p.t.), cells were lysed for the TAP assay (Fig. 1C). The LC-MS/MS analysis identified viral proteins PB2, PB1, PA, and NP and 15 additional host proteins (Fig. 1D). Most of the cellular proteins had been previously identified as influenza virus-related host proteins (4, 28, 42, 43). However, we noticed that MOV10, an RNA helicase, had not been previously implicated in the replication of IAV. Of note, RNA helicases are RNA-binding proteins involved in the regulation of RNA transcription, mRNA translocation, and RNA decay (44, 45). MOV10 had been reported to participate in the replication of various RNA viruses, in the RNAi pathway, and in mRNA decay (31, 46–50). Based on the important role of MOV10 in RNA processing and antiviral function, we were therefore interested in its potential role in the life cycle of IAV.

FIG 1.

Isolation and identification of influenza A virus vRNP-associated proteins. (A) Dually tagged PB2 functioned normally in the minigenome assay. Human 293T cells were transfected in triplicate with the plasmids for expression of PB1 (50 ng), PA (50 ng), and NP (50 ng), together with pPolI-Luci-T (10 ng) and pTK-RL (5 ng). Equal amounts (50 ng) of empty vector (pcDNA3.1), dually tagged PB2 (HA-PB2-FLAG), or nontagged PB2 were separately transfected into the cells. The cells were collected at 48 h p.t., followed by a dual-luciferase reporter assay. Data are shown as the means ± SD from three independent experiments. FL, firefly luciferase; RL, Renilla luciferase. (B) Dually tagged PB2 was copurified with FLAG-tagged PB1, 6×His-tagged PA, and FLAG-tagged NP. Human 293T cells were transfected with the indicated plasmids, followed by immunoprecipitation with anti-HA agarose and Western blotting. (C) Tandem affinity purification of vRNP complexes with dually tagged PB2 as bait. Human 293T cells were transfected with pcDNA3.1-HA-PB2-FLAG (dually tagged PB2) and plasmids for PB1, PA, and NP. At 48 h p.t., cells were lysed for the tandem affinity purification assay. The eluted samples were separated on an SDS-PAGE gel and visualized by silver staining. (D) Fifteen host proteins were identified from the vRNP complex with LC-MS/MS analysis.

To examine the effect of MOV10 during influenza virus infection, we investigated the virus growth rate in A549 cells after depletion of MOV10 by MOV10-specific siRNA. A549 cells were treated with siRNA pool oligonucleotides (SMARTpool from Dharmacon) that targeted four different sequence regions of MOV10 (31). The cells were collected after 36 h and lysed for Western blotting to determine knockdown efficiency. We found that the MOV10-specific siRNA significantly decreased the expression of MOV10 (Fig. 2A). Simultaneously, after transfection with MOV10-specific siRNA or control siRNA for 36 h, A549 cells were infected with A/PR/8/34 virus at an MOI of 0.001. At the desired time points postinfection (p.i.), virus titers in supernatant were determined by plaque assay, and viral growth kinetics was measured (Fig. 2A). We found that MOV10 depletion significantly increased the virus titers compared to that with control siRNA treatment ≥12 h p.i. In line with this, we found that the virus titers were decreased when 293T cells were overexpressed with MOV10-FLAG-expressing plasmid at various time points p.i. (Fig. 2B). Furthermore, we examined the effect of MOV10 on influenza virus in a single-cycle replication. The recombinant virus with the HA segment modified to express firefly luciferase but maintaining the HA packaging sequences can replicate only one round (21). After the depletion of MOV10, 293T cells were infected with PR8-Luc virus and subjected to the luciferase reporter assay at 24 h p.i. (Fig. 2C). We found that MOV10 depletion significantly increased the luciferase activity of PR8-Luc virus. Accordingly, we found that MOV10 overexpression inhibited the luciferase activity of PR8-Luc virus in a dose-dependent manner (Fig. 2D), further confirming that MOV10 inhibits the replication of influenza A virus.

FIG 2.

Effect of MOV10 on influenza virus replication. (A and B) MOV10 inhibits the replication of A/PR/8/34 virus. (A) Human A549 cells treated with MOV10-specific siRNA (50 nM) or control (Ctrl) siRNA (50 nM) were infected with A/PR/8/34 virus at an MOI of 0.001. (B) Human 293T cells transfected with MOV10-FLAG-expressing plasmid (200 ng) or empty vector (200 ng) were infected with A/PR/8/34 virus at an MOI of 0.1. At the indicated time points, the viral titers in supernatant were determined by plaque assays on MDCK cells. The values are means of triplicates. Western blot results represent protein levels of MOV10 in cells with the indicated siRNAs (A) or plasmids (B). (C) Knockdown of MOV10 increases the luciferase activity of PR8-Luc viruses. Human 293T cells were transfected with MOV10-specific siRNA (50 nM) or control siRNA (50 nM), followed by infection with PR8-Luc virus. Firefly luciferase activity was measured at 24 h p.i. (D) MOV10 overexpression impairs the luciferase activity of PR8-Luc virus. Increasing amounts (0 ng, 50 ng, 100 ng, and 200 ng) of pcDNA3.1-MOV10-FLAG were cotransfected. The total plasmid inputs were equalized using an empty vector. The cells were infected with PR8-Luc viruses, followed by the luciferase reporter assay. (E) Knockdown of MOV10 increases viral protein synthesis of A/PR/8/34 virus. A549 cells transfected with MOV10-specific siRNA (50 nM) or control siRNA (50 nM) were infected with A/PR/8/34 virus at an MOI of 1 and collected at the indicated times after infection. The cells were lysed for Western blotting. (F) Statistical analysis of M1 levels in panel C. The value for M1 was standardized to GAPDH levels and normalized to the level of M1 in cells transfected with control siRNA. (G) MOV10 depletion increases viral protein synthesis of H1N1pdm virus. A549 cells treated with MOV10-specific siRNA or control siRNA were infected with H1N1pdm virus (MOI = 1). At the indicated times p.i., cells were lysed for Western blotting. (H) Statistical analysis of M1 levels in panel E. The value for M1 was standardized to GAPDH levels and normalized to the level of M1 in cells transfected with control siRNA. (I) MOV10 overexpression decreases protein synthesis of A/PR/8/34 virus. Human 293T cells transfected with MOV10-FLAG (100 ng) or an empty vector (100 ng) were infected with A/PR/8/34 virus (MOI = 1) at 36 h p.t. and collected at the indicated times after infection for immunoblotting. (J) Statistical analysis of M1 levels in panel G. The value for M1 was standardized to the GAPDH level and normalized to the level of M1 in the cells transfected with the empty vector. Data are shown as the means ± SD from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Student t test).

In addition, we observed a significant increase in viral protein synthesis after MOV10 depletion (Fig. 2E and F). These data indicate that MOV10 plays an inhibitory role in the life cycle of IAV. In order to investigate whether MOV10 additionally influenced other influenza A virus strains, A549 cells were transfected with MOV10-specific siRNA or control siRNA and infected with H1N1pdm virus at an MOI of 1. Similar results were obtained when H1N1pdm virus was used to infect the siRNA-treated cells (Fig. 2G and H). In line with the results obtained by MOV10 depletion, we found that the levels of viral M1 proteins were reduced when 293T cells were overexpressed with MOV10-FLAG and infected with A/PR/8/34 virus (MOI = 1) at various time points p.i. (Fig. 2I and J). The effects of MOV10 on viral protein synthesis at 3 h p.i. were consistent with the effects produced by PR8-Luc virus, since they all represent a single-cycle replication (28). These results further support our initial observations that MOV10 acts as an inhibitor of influenza virus replication.

MOV10 inhibits the polymerase activity of influenza A virus by binding NP.

Since MOV10 was among the vRNP complex-associated proteins and inhibited influenza virus replication, we hypothesized that MOV10 could affect the polymerase activity. To investigate this, we employed the dual-luciferase minigenome system in our study (37, 51). The pPolI-Luci-T plasmid was constructed by inserting the coding sequence of firefly luciferase in a negative-sense orientation flanked with influenza virus cRNA promoter ends. The firefly luciferase reporter gene (pPolI-Luci-T) was a polymerase Ι-driven vector inserted with the coding sequence of firefly luciferase in a negative-sense orientation flanked with influenza virus cRNA promoter ends (34). Polymerase II-driven plasmids for expression of PB1, PB2, PA, and NP were cotransfected into the cells to start the RNA synthesis of firefly luciferase. In addition, Renilla luciferase-expressing plasmid pTK-RL was transfected to normalize the transfection efficiency. We found that MOV10 overexpression inhibited the expression of the luciferase reporter gene in a dose-dependent manner, indicating that MOV10 significantly inhibited influenza virus polymerase activity (Fig. 3A). In line with this, MOV10 depletion with MOV10-specific siRNA significantly enhanced the activity of viral polymerase (Fig. 3B). When we used pPolI-GFP-T (with the gfp sequence to replace the sequence of firefly luciferase of pPolI-Luci-T) in the minigenome system, we obtained similar results: MOV10 overexpression decreased the expression of green fluorescent protein (GFP) in a dose-dependent manner (Fig. 3C and D), and knockdown of MOV10 increased the expression of GFP (Fig. 3E and F). Moreover, the expression of neither PB2-FLAG nor NP-FLAG in the minigenome system was affected by MOV10 overexpression (Fig. 3C) or depletion (Fig. 3E), implying that MOV10 inhibited viral polymerase activity without affecting the expression of PB2 and NP in the minigenome system.

FIG 3.

MOV10 inhibits influenza virus polymerase activity. (A) MOV10 overexpression impairs polymerase activity in a dose-dependent manner in the minigenome system. Human 293T cells were transfected in triplicates with plasmids for the minigenome system. Increasing amounts (0 ng, 50 ng, 100 ng, 200 ng, 400 ng, and 600 ng) of pcDNA3.1-MOV10-FLAG were cotransfected. The total plasmid inputs were equalized using an empty vector. The cells were collected at 48 h p.t., followed by a dual-luciferase reporter assay. (B) Knockdown of MOV10 increases the polymerase activity in the minigenome assay. Human 293T cells were transfected with MOV10-specific siRNA (50 nM) or control siRNA (50 nM), together with plasmids for the minigenome system. Firefly luciferase and Renilla luciferase activities were measured at 48 h p.t. The firefly luciferase activity was normalized to Renilla luciferase activity. (C) MOV10 overexpression affects the reporter gene of the minigenome system but not the expression of PB2-FLAG or NP-FLAG. 293T cells were transfected with increasing amounts of pcDNA3.1-MOV10-FLAG and plasmids for expression of PB1 (50 ng), PB2-FLAG (200 ng), PA (50 ng), and NP-FLAG (50 ng), together with pHW2000-HA-GFP (100 ng). Cells were harvested for 48 h and lysed for Western blotting. (D) Statistical analysis of GFP levels in panel C. Values of GFP were standardized to GAPDH levels and normalized to the level of GFP in cells transfected with the empty vector (MOV10-FLAG as 0 ng). (E) MOV10 depletion increases the reporter gene of the minigenome system but not the expression of PB2-FLAG or NP-FLAG. 293T cells were transfected with MOV10-specific siRNA (or control siRNA) together with the indicated plasmids and collected at 48 h p.t., followed by Western blotting. (F) Statistical analysis of GFP levels in panel E. Values of GFP were standardized to GAPDH levels and normalized to the level of GFP in cells transfected with control siRNA. Data in panels A, B, D, and F are shown as the means ± SD from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Student t test).

To further elucidate the relationship between MOV10 and IAV, we validated the specific interaction of MOV10 with the vRNP complex. As described above, MOV10 was pulled down by dually tagged PB2 in the vRNP complex. We performed the immunoprecipitation experiment to confirm the result. We found that MOV10 was copurified with PB2-HA in the influenza virus vRNP complex, as expected (Fig. 4A). To further explore the possible interactions of MOV10 with the subunits of the vRNP complex, we transfected pcDNA3.1-MOV10-FLAG, together with plasmids for expression of GFP-HA (negative control), PB1-HA, PB2-HA, PA-HA, or NP-HA, separately into 293T cells. The cell extracts were immunoprecipitated with anti-HA agarose beads, followed by immunoblotting (Fig. 4B). Unexpectedly, MOV10-FLAG was specifically pulled down by NP-HA, but not other subunits. In addition, in the complementary immunoprecipitation assay, MOV10-HA interacted only with NP-FLAG (Fig. 4C) and not PB1, PB2, or PA (data not shown). The interaction of MOV10 with NP was mediated by RNA, since the interaction was eliminated with RNase A treatment (Fig. 4C, lane 5). MOV10 did not directly interact with PB1, PB2, or PA (Fig. 4B). Given that PB2 could interact with MOV10 in the presence of subunits, including PA, PB1, and NP (Fig. 4A), we assumed that MOV10-PB2 interaction was mediated by MOV10-NP interaction. When MOV10 and PB2 were cotransfected into 293T cells in the absence of NP, we found that MOV10 no longer interacted with PB2 (Fig. 4B). Therefore, it was quite convincing that MOV10 was copurified with PB2 in the vRNP complex through the interaction with NP. To further characterize MOV10 and NP interaction, we examined the associations between MOV10 and NP in the infected A549 cells. As shown in Fig. 4D, endogenous MOV10 specifically bound to NP proteins in A/PR/8/34 virus-infected cells, which further confirmed the interaction between MOV10 and NP. It has been reported that MOV10 is associated with various cellular RNAs, including 7SL RNA (31, 52). As our previous work showed that MOV10 interacts with APOBEC3G through 7SL RNA (31), we decided to examine whether 7SL RNA also affects the interaction between MOV10 and NP. The interaction between MOV10-HA and NP-FLAG was analyzed after depletion of 7SL RNA. Previous studies have reported that the expression level of 7SL RNA can be affected by SRP14 depletion (53, 54). To knock down 7SL RNAs, 293T cells were transfected with shRNA targeting SRP14 (SRP14-shRNA). As shown in Fig. 4E, we found that 7SL RNA level was significantly decreased after SRP14 depletion. Importantly, the interaction between MOV10-HA and NP-FLAG decreased after 7SL RNA depletion (Fig. 4F), indicating that 7SL RNA mediates the interaction between MOV10 and NP. In addition, when we tested the comigration between MOV10 and NP, we found that MOV10 existed in the same fractions as NP in the FPLC analysis (Fig. 4G, top and middle). The negative control, NF-κB p65, did not migrate with them (Fig. 4G, bottom). Together, these results demonstrate that MOV10 interaction with the NP subunit of the vRNP complex is mediated by 7SL RNA.

FIG 4.

Analysis of the interaction between MOV10 and the influenza virus vRNP. (A) MOV10-FLAG can be copurified with PB2-HA in the viral RdRp complex. Human 293T cells were transfected with the indicated plasmids and collected 48 h p.t. Cells were lysed for immunoprecipitation using anti-HA beads, followed by Western blotting. (B) MOV10 interacts with the NP subunit of RdRp. Cells were transfected with indicated plasmids and collected at 48 h p.t., followed by immunoprecipitation and a Western blot assay. (C) HA-tagged MOV10 binds to FLAG-tagged NP in an RNA-dependent manner. The cells transfected with the indicated plasmids were immunoprecipitated and treated or not with RNase A, followed by Western blotting. (D) MOV10 binds to NP in infected cells. A549 cells were infected with A/PR/8/34 virus (MOI = 1) for 24 h. The infected cells were lysed and precleared with protein A/G beads. The resulting extracts were aliquoted and immunoprecipitated with anti-NP or normal mouse IgG. The eluates were immunoblotted with anti-NP and anti-MOV10 antibodies. (E and F) The 7SL RNA mediates the interaction between MOV10 and NP. (E) The SRP14 mRNA and 7SL RNA in 293T cells expressing control shRNA and SRP14 shRNA were quantified using real-time PCR. Data are means ± SD. (F) 293T cells expressing control shRNA or SRP14 shRNA were cotransfected with MOV10-HA- and NP-FLAG-expressing plasmids and then lysed for immunoprecipitation using anti-HA beads, followed by Western blotting. Statistical analysis of NP-FLAG levels in the IP samples of each lane is shown. (G) MOV10 comigrates with NP in a Superose 6 FPLC column. Cells were infected with A/PR/8/34 virus (MOI = 1) for 24 h. The total cell lysates were fractionated on a Superose 6 FPLC column, and the resulting fractions were assessed by immunoblotting after ethanol precipitation. Fraction numbers are indicated, with molecular masses, in kilodaltons, listed above them.

MOV10 disrupts the nucleocytoplasmic distribution of NP by blocking the nuclear import of NP.

Influenza virus NP, containing both nuclear localization signals (NLSs) and nuclear export signals (NESs), is a nuclear shuttle protein. During viral infection, vRNPs utilize NLSs on NP to be transported into the nucleus, where viral replication and transcription begin (17, 18, 55, 56). Since MOV10 bound to NP, we further analyzed the colocalization of NP and MOV10 in the cells. Interestingly, we found that NP-FLAG was mainly detected in the nucleus without MOV10 overexpression (Fig. 5A, top, and B) but was mostly retained in the cytoplasm when MOV10 was overexpressed in the immunofluorescence assay (IFA) experiment (Fig. 5A, bottom, and B). When we repeated the assay using nuclear-cytoplasmic fractionation, a lower nuclear ratio of NP-FLAG was observed with MOV10 overexpression (Fig. 5C). In parallel, when we analyzed the location of PB2 in cells after MOV10 overexpression, we found that MOV10 did not affect the nuclear import of PB2 (Fig. 5D), implying that the effect of MOV10 is specific for NP.

FIG 5.

MOV10 affects the nucleocytoplasmic distribution of NP. (A) Overexpression of MOV10 inhibits NP-FLAG nucleocytoplasmic distribution. Human 293T cells transfected with NP-FLAG-expressing plasmid (100 ng) together with pcDNA3.1-MOV10 (100 ng) or an empty vector (EV; 100 ng) were fixed at 12 h p.t., followed by immunofluorescence using anti-NP (red) and anti-MOV10 (green) antibodies. The nucleus was stained with DAPI (blue). Scale bars, 10 μm. (B) Quantitative analysis of nucleocytoplasmic distribution of NP. At least 200 cells in each group from three independent assays were scored. N, predominantly nuclear; N+C, nuclear and cytoplasmic; C, predominantly cytoplasmic. Data are shown as the means ± SD from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Student t test). (C) The cells transfected with NP-FLAG and MOV10-FLAG (or empty vector) were separated into cytoplasmic (C) and nuclear (N) fractions. W, whole cell. Statistical analysis of NP-FLAG levels of each lane is shown. (D) MOV10 has no effect on the distribution of PB2 in cells. Human 293T cells transfected with PB2-HA-expressing plasmid (100 ng) together with pcDNA3.1-MOV10 (100 ng) or the empty vector (100 ng) were fixed at 12 h p.t., followed by immunofluorescence using anti-HA (red) and anti-MOV10 (green) antibodies.

In addition, we analyzed the effect of MOV10 on the localization of NP during influenza virus infection. The subcellular localizations of viral NP were compared in infected cells (A/PR/8/34 virus at an MOI of 1) with and without MOV10 overexpression and observed at 3, 6, and 9 h p.i. (Fig. 6A to C). At 3 h p.i. (Fig. 6A, left), NP in the cells transfected with empty vector (negative control) was located predominantly in the nucleus and then migrated into the cytoplasm soon afterwards. With MOV10 overexpression, most of NP was distributed and located in the cytoplasm (Fig. 6B and C). With high expression of NP, the subcellular localization of NP at 9 h p.i. was scored in each group in random fields of view with three replicates (Fig. 6D) and demonstrated that most NP was located in the cytoplasm after MOV10 overexpression. These results show that MOV10 affects the nucleocytoplasmic distribution of influenza virus NP during either transfection or infection.

FIG 6.

MOV10 affects the nucleocytoplasmic distribution of NP in infected cells. (A to C) Overexpression of MOV10 inhibits the NP nucleocytoplasmic distribution in influenza A virus-infected cells. 293T cells transfected with pcDNA3.1-MOV10-FLAG (100 ng) or empty vector (100 ng) were infected with A/PR/8/34 virus at an MOI of 1. At the indicated times p.i., cells were fixed and stained for NP (red) and nuclei (blue) by indirect immunofluorescence. Scale bars, 10 μm. (D) Quantitative analysis of nucleocytoplasmic distribution of NP at 9 h p.t. At least 200 cells in each group from three independent assays were scored. N, predominantly nuclear; N+C, nuclear and cytoplasmic; C, predominantly cytoplasmic. Data are shown as the means ± SD from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Student t test).

The nuclear import of NP relies on molecular interactions with importin-α/β family members (14, 15, 57). Inefficient importin binding of NP will result in reduced nuclear localization of NP. Since MOV10 directly interacted with NP and retained it in the cytoplasm, we examined whether MOV10 could inhibit the interaction between NP and importins and subsequently inhibit nuclear import of NP. To test this hypothesis, we cloned four isoforms of HA-tagged importin-α, α1, α3, α5, and α7, separately and individually transfected them into 293T cells with NP-FLAG in the presence or absence of MOV10-FLAG-expressing plasmid. Through co-IP assay, we observed that NP bound to all four isoforms of importin-α, but the amount of NP binding to these proteins was significantly decreased when MOV10 was overexpressed (Fig. 7). These results suggest that MOV10 inhibits the nuclear import of NP by decreasing the interaction between NP and importin-α.

FIG 7.

MOV10 blocks the nuclear import of NP by weakening the binding of NP to importin-α. Shown is the interaction of NP with importin-α family members, importin-α1 (A), importin-α3 (B), importin-α5 (C), and importin-α7 (D), in the present or absence of MOV10-FLAG-expressing plasmid. Human 293T cells were transfected with the indicated plasmids and immunoprecipitated with anti-HA agarose, and the associated NP was determined with anti-FLAG antibody.

The NP-binding regions of MOV10 at both the N and C termini are responsible for the inhibition of viral RdRp activity.

MOV10 is an RNA helicase that contains seven RNA helicase domains (Fig. 8A) (58). Of these domains, the second helicase domain, containing the DEAG-box motif, has been shown to be essential for the RNA helicase activity of MOV10 (47). To study the possible role of RNA helicase activity of MOV10 in influenza virus replication, we introduced a point mutation in the DEAG-box motif of MOV10 (from DEAG to DQAG, yielding the E/Q mutant) (Fig. 8A) (47, 49). The E/Q mutant of MOV10 inhibited RdRp activity in the minigenome system as effectively as wild-type (WT) MOV10 did (Fig. 8B). In addition, the interaction between MOV10 and NP was unaffected by the E/Q mutant of MOV10 (Fig. 8C), implying that the inhibition did not require the RNA helicase activity of MOV10 and more likely relied on the binding of MOV10 to NP. To map the region(s) of binding of MOV10 to NP, we utilized the six sequence motif deletion mutants of MOV10 in our study (Fig. 8D) (31). Human 293T cells were transfected with full-length MOV10 or six truncation mutants separately, followed by infection with A/PR/8/34 virus (MOI = 1). At 12 h p.i., we found that the overexpression of full-length MOV10 reduced the expression of influenza virus M1 protein significantly (Fig. 8E). The middle regions of MOV10, from amino acid 162 to 669, appeared to be unnecessary for the inhibition, since the truncation of these regions did not affect the expression of viral M1 protein, while the N terminus (Δ1-170) and C terminus (Δ663-850 and Δ884-1003) of MOV10 were important for the inhibitory activity, since truncation of these regions failed to reduce the expression of M1 (Fig. 8E). In line with this, the deletion mutants of MOV10 sufficiently inhibited the RdRp activity in the minigenome system, except for the Δ1-170, Δ663-850, and Δ884-1003 mutants (Fig. 8F). In addition, NP-FLAG could be copurified with the MOV10 mutants with deletions in the middle region (from amino acid 162 to 669), while NP-FLAG was less or even not copurified with N-terminally (Δ1-170) or C-terminally (from amino acid 663 to 1003) truncated MOV10 mutants (Fig. 8G). It demonstrated that both the N-terminally and C-terminally truncated MOV10 mutants attenuated the interaction between MOV10 and NP, especially the region from amino acid 884 to 1003. Taken together, these data demonstrate that MOV10 binds to NP at both the N and C termini, and the binding regions of MOV10 are also required for the inhibition of viral polymerase activity.

FIG 8.

MOV10 interacts with NP through the N and C termini of MOV10. (A) Schematic diagram of DEAG-box mutant of MOV10 (E/Q mutant). (B) The DEAG-box mutant of MOV10 still inhibits the polymerase activity of influenza virus. Human 293T cells transfected with plasmids for the influenza virus minigenome system, together with plasmid for an empty vector (pcDNA3.1), MOV10, or MOV10 E/Q mutant (100 ng), were lysed for a dual-luciferase reporter assay. Data are shown as the means ± SD from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Student t test). (C) The DEAG-box of MOV10 is not necessary for interaction with NP. Cells transfected with indicated plasmids were lysed and immunoprecipitated with anti-HA agarose and analyzed by Western blotting assay. (D) Illustration of MOV10 deletion fragments used. Δ, deleted. The number indicated the amino acids of MOV10 truncated. (E and F) The N and C termini of MOV10 lose the functional inhibition of polymerase activity of influenza virus. (E) Human 293T cells were transfected with either the empty vector (pcDNA3.1) or plasmids for full-length MOV10 or its deletion mutants separately, followed by infection with A/PR/8/34 at an MOI of 1. Cells were collected at 12 h p.t. and lysed for Western blotting. (F) Cells were transfected with plasmids for the influenza virus minigenome system, along with full-length MOV10 or truncations or an empty vector. The cells were collected at 48 h p.t. and lysed for a dual-luciferase reporter assay. Data are shown as the means ± SD from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (Student t test). (G) Regions of binding of MOV10 to NP. Human 293T cells were transfected with plasmids for full-length MOV10 or deletion mutants, together with NP-FLAG-expressing plasmid. Cells were collected at 48 h p.t., followed by immunoprecipitation. Statistical analysis of NP-FLAG levels in the IP samples of each lane is shown.

DISCUSSION

The influenza virus vRNP complex is a major determinant of host range and regulator of viral pathogenesis (4). The important role of the vRNP complex in viral replication suggests that many host proteins are involved in the function of the complex. It is important to study the complicated relationship between host factors and the vRNP complex, which could potentially lead to better understanding and monitoring of viral replication and transmission. To date, several host proteins have been identified as capable of interacting with the influenza virus vRNP complex, but the mechanisms for how these proteins function remain poorly understood (22–24, 59, 60). In our study, we identified MOV10 as a vRNP-associated host protein through tandem affinity purification and mass spectrometry. We used dually tagged PB2 as bait in the presence of other vRNP subunits so that we could purify the vRNP complex and its associated cellular proteins. With further experiments, we demonstrated that MOV10 bound to vRNP subunit NP, inhibited the interaction between NP and importin-α, and subsequently blocked the nuclear import of NP. To the best of our knowledge, our study is the first to examine the participation of MOV10 in the influenza virus life cycle through inhibition of the nuclear import of NP.

MOV10 was discovered from a mouse strain infected with Moloney murine leukemia viruses (61, 62). As it contains seven putative RNA helicase motifs, MOV10 was recently classified into the UPF1-like RNA helicase superfamily (47, 58). MOV10 is a multifunctional protein involved in many cellular biological processes. It has been reported that MOV10 facilitates the assembly and maturation of the microRNA-inducing silencing complex (miRISC) and functions in UPF1-mediated mRNA degradation (31, 46, 47). In addition, MOV10 was reported to affect the replication of several kinds of viruses, including HIV-1, hepatitis C virus (HCV), and hepatitis B virus (HBV) (48–50, 63, 64). In the case of HIV-1, overexpression of MOV10 inhibits the production of viral particles (48–50). The molecular mechanisms for its inhibitory effect on HBV and HCV remain unknown. Although an affinity purification and mass spectrometry experiment found that MOV10 was among the host proteins binding with the vRNP complex of IAV, its function in IAV replication has not been clarified until our study (28). Full-length MOV10 has 1,003 amino acids and seven RNA helicase motifs (31, 47, 58). However, the activity of RNA helicases is not essential for some functions, but their N- and C-terminal regions seem to be important (27). Here, we showed that the RNA helicase activity of MOV10 was not required for inhibition of the influenza virus polymerase activity. With a co-IP assay, we demonstrated that MOV10 interacted with influenza virus NP through both the N and C termini. It was unexpected that MOV10 seemingly applied its two ends to bind to NP. This could be due to the N terminus and C terminus of MOV10 being in close proximity to each other in the three-dimensional (3D) structure of MOV10, such that they could be folded as one region to bind to NP. As the interaction between certain subregions of MOV10 and NP was required for the inhibition of influenza virus polymerase activity, MOV10 exerted its antiviral function by blocking the importin-α-binding sites of NP.

During viral RNA synthesis, NP maintains the RNA template sequentially to be suitable for transcription (4, 55). An important role of NP in the vRNP complex is to mediate the nuclear import of this complex during the early stage of influenza virus infection (56). Multiple NP subunits are present in the vRNP complex and can provide several NLS motifs to promote nuclear import. There are two NLSs and one nuclear accumulation signal (NAS) on NP (65, 66). After the vRNPs are released into the cytoplasm, they migrate into nucleus through the classical cellular importin-α/β-dependent nuclear import pathway. Importin-α can bind to the NLS of NP and then recruit importin-β to transport the complex to the nuclear pore and then translocate the vRNPs into nucleus (17, 56). To date, several importin-α isoforms—α1, α3, α5, and α7—have already been shown to bind to NP and mediate this process (4). The disruption of interaction between NP and importin-α results in the retention of NP and vRNPs in the cytoplasm and subsequently affects the polymerase activity. Though several host proteins have been identified as affecting influenza virus RNA transcription, replication, and vRNP assembly, few of them have been shown to affect the NP nuclear import except for the well-defined importin-α/β family. In our study, we elucidated a novel NP-interacting cellular protein, MOV10, which could inhibit NP binding to importin-α and block the nuclear import of NP. Both MOV10 and NP are RNA-binding proteins, and they bind to different kinds of RNAs (31, 47, 55, 67). Our observations indicate that the interaction between MOV10 and NP is dependent on RNAs. Among the MOV10-associated RNAs, 7SL RNA was reported to bind to MOV10 and affect its function (31, 52). Our current study also found that 7SL RNA mediates the interaction between MOV10 and viral NP. The N terminus of NP was responsible for RNA binding and importin-α binding, implying that the RNA binding deficiency of NP could affect the interaction with importin-α (55, 68). Also, the C terminus of MOV10 (from amino acid 884 to 1003) of MOV10 was one of the RNA-binding regions of MOV10 (31), which was also responsible for NP binding (Fig. 8G), further indicating that the RNA binding ability of MOV10 was required for MOV10-NP interaction. Our findings further extend our knowledge on the complicated interaction between host proteins and influenza viruses. Viral NP exploits the host nuclear transport machinery, while MOV10 attenuates this action by blocking the interaction between NP and importin-α. Importantly, importin-α1 and -α7 positively regulate viral polymerase activity in mammalian cells, but not in avian cells. Interestingly, it has been reported that importin-α7 contributes to the adaptation of avian viruses in mammals (57, 69). Given that MOV10 inhibits influenza virus replication by blocking importin-NP interaction, the effects of MOV10 on the interspecies transmission of avian or mammalian influenza A viruses merit further investigation.

Because the expression of MOV10 can be enhanced by treatment with type Ι IFNs, the MOV10 gene is considered one of the IFN-stimulated genes (ISGs) (63, 64). Type I IFNs have broad effects on the innate immune system against viruses (64). The ability of IFNs to restrict viral replication is largely dependent on the induction of ISGs (70). Many ISGs have been proved to inhibit influenza viruses, such as Mx family members and IFITMs (71, 72). Our study has identified one more ISG against influenza virus, demonstrating a novel mechanism, the inhibition by IFNs of the replication of influenza viruses. Considering that IFN treatment could cause serious immunopathic effects during acute infections with viruses such as influenza virus, ISGs may be better candidates for antiviral drug design. Treatments based on the induction of ISGs can also avoid the counteraction from viruses, such as the inhibitory effect of NS1 proteins encoded by influenza virus. Our current work could therefore provide a new rationale for anti-influenza virus drug design. For instance, the small-molecule compounds that mimic the function of MOV10 to inhibit the interaction between NP and importin-α could be screened out and sequentially developed.