MOV10 has emerged as an important host antiviral factor. MOV10 not only inhibits various viruses, including human immunodeficiency virus type 1, hepatitis C virus and vesicular stomatitis virus, but also restricts the activity of retroelements long interspersed nucleotide element-1, Alu, SVA and intracisternal A particles. Here, we report that MOV10 suppresses influenza A virus infection through interacting with viral nucleoprotein (NP), sequestering viral RNP in the cytoplasm and causing the degradation of viral vRNA. The antiviral activity of MOV10 depends on the integrity of P-bodies. We also found that the antiviral activity of MOV10 is partially countered by viral NS1 protein that interferes with the interaction of MOV10 with viral NP and causes MOV10 degradation through the lysosomal pathway. Moreover, NS1-defective influenza A virus is more susceptible to MOV10 restriction. Our data not only expand the antiviral spectrum of MOV10 but also reveal the NS1 protein as the first viral antagonist of MOV10.
Mov10 gene was first discovered as the genetic locus into which Moloney murine leukemia virus provirus is integrated [1,2]. Human MOV10 has 1003 amino acids, it belongs to the UPF-1 like helicase superfamily-1 (SF1) , and exhibits ATP-dependent 5′ to 3′ helicase activity. MOV10 has been shown to facilitate UPF-1-mediated RNA degradation through binding to and translocating along the 3′-untranslated region of mRNA. MOV10 homologs, including SDE3 in Arabidopsis, armitage in Drosophila and ERI-6/7 in Caenorhabditis elegans, participate in small-interfering RNA-mediated gene silencing [4–6]. MOV10 itself associates with RNA-induced silencing complex (RISC) and it is found in P-bodies and stress granules, and has a role in small RNA-mediated gene silencing. Indeed, knocking down MOV10 results in the impairment of RNA interference [7,8].
MOV10 modulates viral replication and the activity of retroelements. Studies have shown that MOV10 facilitates HDV (hepatitis delta virus) replication, but restricts HCV replication [9,10]. The activities of retroelements intracisternal A particles, LINE-1 (long interspersed nucleotide element-1), Alu and SVA are inhibited by MOV10 [11–14]. The effects of MOV10 on HIV-1 replication manifest at several levels. Overexpression of MOV10 reduces HIV-1 Gag expression and virus production. MOV10 is also found in HIV-1 particles and impairs virus infectivity [15–18]. However, depletion of endogenous MOV10 moderately diminishes HIV-1 production , which may result from a cofactor role of MOV10 for Rev . MOV10 may thus suppress some steps of HIV-1 replication while enhancing others.
MOV10 has been reported to be associated with proteins of influenza A virus (IAV). The interaction of MOV10 with viral nucleoprotein (NP) was first reported in the analysis of NP-associated cellular proteins by mass spectrometry . A nonstructural viral protein, NS1, was also shown to associate with MOV10 . We have now investigated the effects of these interactions on IAV infection. We found that MOV10 bound NP of IAV and sequestered viral RNP in P-bodies probably for further degradation. Moreover, we found that the NS1 protein antagonized MOV10 restriction by either diminishing the interactions between NP and MOV10 or inducing the degradation of MOV10 through the lysosome-dependent pathway.
MOV10 cDNA was cloned in pcDNA3.1 as previously described . Truncations of MOV10 were generated using a PCR-based strategy. The primers are presented in Supplementary Table S1. Amino acid substitutions in MOV10 helicase mutants were generated using the QuickChange site-directed mutagenesis kit (Stratagene). Specifically, seven MOV10 helicase motif mutants were generated by substituting conserved amino acids within helicase motifs, i.e. movMIa 548AHILACAPSNSGADL562 to 15As, movMI 524GPPGTGKT531 to 8As, movMII 645DEAG648 to 4As, movMIII 676VLAGDPRQ683 to 8As, movMIV 728LLRNYR733 to 6As, movMV 864SVEEFQGQ871 to 8As, movMVI907VAVTR911 to 5As, as descripted previously . WSN NP expressing plasmids were created by inserting WSN NP ORF into pcDNA 3.1. WSN NS1 expressing plasmid was generated by inserting the codon-optimized WSN NS1 sequence into pcDNA3.1.
Cell culture, stable cell line generation and transfection
HEK293, HEK293T and A549 cells were cultured in complete DMEM (Gibco) containing 10% FBS (Gibco), penicillin (100 U/ml) and streptomycin (100 µg/ml). Control and MOV10 overexpression cell lines were generated by transduction with pQC-XIP (Clontech)-based pseudo-virus according to the manufacturer's instruction. Monoclonal cell lines were obtained by puromycin selection and followed by limited dilution. Plasmid DNA and siRNA were transfected into cells using PEI (Sigma) and RNAiMAX (Invitrogen), respectively, according to the manufacturer's instructions. siRNA sequences used in the present study are provided in Supplementary Table S2.
Generation of A549 knockout cell lines
A549 knockout cell lines were generated using the CRISPR/Cas9 system. Briefly, sgRNAs were designed online: http://www.e-crisp.org/E-CRISP/. sgRNA sequences were then inserted into the vector of LentiV2 (Addgene). The resultant plasmids were then used to generate lentiviruses. A549 cells were infected with lentiviruses and selected with puromycin (0.8 µg/ml) for 4 days. Cells were then amplified for later use.
Wild-type WSN (A/WSN/1933) was generated by transfecting 293T and MDCK cocultured cells , and supernatants were collected and used to infect MDCK with an MOI of 0.001 for 36 h. PR8 (A/PR/8/34) was propagated in embryonated chicken eggs according to classical virological techniques. NS1-defective PR8 viruses were generated by substituting wild-type NS1 segment with a mutated NS1 segment whose 13th codon of NS1 (Cys) was changed to a stop codon. A/H1N1pdm09 and A/H3N2 viruses were propagated in serum-free medium using MDCK cells as a host. A/PR/8/34 was provided kindly by Dr Yuelong Shu (China CDC), and A/H1N1pdm09 and A/H3N2 viruses were provided kindly by Dr Dayan Wang (China CDC).
MDCK cells (0.18 millions) were seeded in 12-well plate 1 day before infection. Culture medium of infected cells was serially diluted by 10 folds using virus maintenance medium (DMEM with 0.5% FBS, 1 µg/ml TPCK-trypsin, 1% penicillin and streptomycin). MDCK cells were washed twice with PBS and 250 µl of diluted viruses were added. Cells were infected at room temperature for 1 h (rotated the plate to prevent drying every 15 min). Supernatants were discarded, and cells were washed twice with PBS. Then, cells were coated with coating buffer (DMEM with 1% Agarose II, 0.075% BSA, 1 µg/ml TPCK-trypsin, 1% penicillin and streptomycin). Place the plate upside down in the 37°C incubator when the coatings became solid. Seventy-two hours later, count the plaque and calculate the PFU.
Briefly, five 10 cm dishes of cells were either transfected with vector or Flag-MOV10-expressing plasmids. Twenty-four hours later, cells were infected with WSN MOI = 0.01 for 24 h. The supernatants were collected and pre-clarified by centrifugation 1800×g 30 min and followed by 10 000 rpm 30 min. The supernatants were then loaded on 30% sucrose cushion and followed by centrifugation 25 000 rpm 2 h. The pellets were resuspended by TNE buffer and subjected to western blotting.
IAV minigenome assay
Briefly, we seeded HEK293T cells on six-well plates, and transfected the WSN minigenome system, which includes 100 ng each of 4P components (PA, PB1, PB2 and NP), 10 ng pPOLI-NP luciferase (NP ORF was substituted with luciferase ORF), and internal control SV40-Renila (Renila luciferase gene under control of SV40 promoter), along with up to 400 ng MOV10 expression plasmids. Twenty-four hours post-transfection, cells were harvested and measured luciferase activity according to the manufacturer's instruction.
IAV mRNA and vRNA quantification
IAV vRNA, mRNA and cRNA were quantified as described by Kawakami et al.  Briefly, A549 cells that stably express MOV10 were incubated with WSN (MOI = 0.2) at 4°C for 1 h, then washed with ice-cold PBS twice to remove the unbound viruses, followed by incubation with warm virus maintenance medium (DMEM with 0.5% FBS, 1 µg/ml TPCK-trypsin, 1% penicillin and streptomycin) at 37°C for 2 h. Cells were harvested by trypsin digestion. Total cellular RNA was extracted and reverse-transcribed using the Superscript III kit (Invitrogen) following hot start protocol and using IAV HA sequence-specific cRNA/mRNA/vRNA primers. Real-time PCR was conducted to determine the quantity of IAV mRNA, vRNA and cRNA using primers specific to each kind of RNA. Primer sequences are provided in Supplementary Table S1.
Inhibition of lysosomal pathways
A549 cells were infected with WSN (MOI = 10) at 4°C for 1 h. Cells were washed with cold PBS twice and cultured at 37°C for 1 h. NH4Cl or chloroquine was added into the culture for 5 h before cells were harvested and examined for MOV10 expression.
A549 cell lines were pretreated with 100 µg/ml cycloheximide (CHX) for half an hour, then washed with ice-cold PBS twice. Cells were then infected with WSN at 4°C for 1 h. Then, cells were again washed with ice-cold PBS twice, and either harvested as time 0, or cultured in virus maintenance media with a final CHX concentration of 100 µg/ml for indicated time until ready for harvest.
Western blotting and antibodies
Western blotting was performed as previously described . Antibodies used in this article are as follows: Flag (Sigma–Aldrich, Cat# F7425), Myc (Sigma–Aldrich, Cat# C3956,), NP (Millipore, Cat# MAB8251), H1N1 HA (Sino bio., Cat# 11055-M08), NS1, PA, PB1, PB2 (GeneTex, Cat# GTX125990, Cat# GTX125932, Cat# GTX125923, Cat# GTX125926, respectively), DCP1a (Abnova Novus, Cat# H00055802-M06), MOV10 (Proteintech Group, Cat# 10370-1-AP), and actin (Sigma–Aldrich, Cat# A1978).
Flow cytometry analysis was performed as previously described . In brief, cells were detached and resuspended in complete growth media. Cells were fixed with 1% paraformaldehyde for 10 min at room temperature followed by a 10 min permeabilization with 0.2% TX-100 at room temperature. Cells were then incubated for 1 h with indicated antibodies, FITC-conjugated anti-Influenza A NP antibody (Merck, Cat# MAB8257F) and APC-conjugated anti-Flag antibody (BioLegend, Cat# 637308). All cell samples were analyzed on a BD FACS Canto II using the BD Diva software.
Immunofluorescence assay and proximity ligation assay
Immunofluorescence assay was performed as previously described . Proximity ligation assay was performed by using the Duolink®-PLA kit, and the experiments were carried out strictly according to the manufacturer's instructions. Briefly, cells were seeded on glass cover slips. Twenty-four hours later, cells were treated with CHX (100 µg/ml) for 30 min before infected with WSN. After infected at 4°C for 1 h, cells were washed with ice-cold PBS twice before incubated at 37°C for 2 h. Cells were then fixed with 4% paraformaldehyde at room temperature for 15 min and washed with PBS three times. Samples were then preceded to blocking, incubation of the primary antibodies, incubation of the secondary antibodies, ligation, amplification, washing and mounting onto slide glasses by following the manufacturer's instructions.
Intensities of protein bands in Western blots were determined with ImageJ. Histograms showed means ± SD of data from three independent experiments. Statistical significance was determined using Student's t-test (*P < 0.05, **P < 0.01, ***P < 0.001).
MOV10 inhibits IAV replication
Since MOV10 was reported to be a candidate interacting with NP , we first investigated whether MOV10 affects IAV infection. Three HEK293 cell clones stably expressing different levels of exogenous MOV10 were challenged with WSN (Figure 1A). The results show that MOV10 expression diminished the expression of both HA0 and NP in either one round (4 h) or multiple rounds of infections (24 h). Lower levels of HA0 and NP were expressed in cells that express higher levels of MOV10. Similar results were observed when MOV10 was overexpressed in A549 cells (Figure 1B). IFITM3, which has been shown to strongly inhibit the entry of IAV, was used here as a positive control (Figure 1B). We also confirmed these results using flow cytometry (Figure 1A,B, lower panel). When the production of IAV from infected A549 cells was monitored over prolonged period of time, expression of MOV10 significantly reduced viral titers at each time point tested (Figure 1C). Next, we knocked down the endogenous MOV10 in either 293T or A549 cells, and observed an increase in the expression of viral HA0 and NP proteins upon depletion of endogenous MOV10 (Figure 1D,E). Also, using flow cytometry, we obtained similar results (Figure 1D,E, lower panel). In addition, knockdown of MOV10 markedly increased the titer of IAV in the culture supernatants (Figure 1F). Furthermore, we also determined virus production in the A549-overexpressing cell line using two IAV strains (A/H1N1pdm09 and A/H3N2). The results showed that MOV10 overexpression inhibited replication of IAV (Supplementary Figure S1A,B). In addition, transient expression of MOV10 restricts influenza virus replication in Vero cells in which the interferon signaling pathway is defective (Supplementary Figure S1C). The results showed that MOV10 restricted the replication of IAV was independent of its enhancing interferon induction activity. Taken together, these data demonstrate an inhibitory role of MOV10 in IAV infection.
MOV10 restricts IAV replication.
MOV10 inhibits the early stage of IAV infection
MOV10 restricts HIV-1 infection by incorporating into HIV-1 particles . We therefore asked whether MOV10 is also present in IAV particles. To this end, we harvested IAV particles that were produced from cells that overexpressed MOV10 by ultracentrifuging the supernatants through a 30% sucrose cushion. The results did not reveal the presence of either the overexpressed or the endogenous MOV10 in the IAV particles (Figure 2A). Thus, we proposed that MOV10 may restrict IAV in the first round of infection. Indeed, we found a dramatic decrease in NP expression in the presence of MOV10 3 h post-infection as shown by immunostaining of NP (Figure 2B). We also observed that MOV10 and NP co-localized in the cytoplasm (Figure 2B, enlarged panel). We next measured the levels of three species of viral RNA, including complementary RNA (cRNA), messenger RNA (mRNA) and viral genomic RNA (vRNA), 2 h post-infection. The result showed that at 2 h post-infection, virus did not generate any cRNA and new vRNA. Interestingly, in MOV10-overexpressing cells, we observed an up to 70% of reduction in vRNA. This result suggested that MOV10 promoted the degradation of viral RNA. Moreover, we also detected the reduction in mRNA, which may result from the reduction in its template vRNA, although we could not exclude the possibility that MOV10 may restrict viral RNA transcription (Figure 2C). Since MOV10 is a cytoplasmic protein and is found in P-bodies, we next checked whether MOV10 restricted the IAV replicon which recapitulates IAV replication events except for virus entry, assembly and release. The results showed that MOV10, but not IFITM3, restricted the replication of IAV replicon (Figure 2D, the right panel showed the expression levels of replicon components, MOV10 and IFITM3, in the samples of luciferase assay). Taken together, these data suggest that MOV10 inhibits IAV replication at the early stage of virus infection.
MOV10 restricts the early stage of IAV replication.
MOV10 sequesters IAV RNP in the cytoplasm
It is reported in a mass spectrometry study that MOV10 is associated with NP . We confirmed that NP was co-immunoprecipitated with myc-MOV10 in an RNA-dependent manner (Figure 3A). We further verified that endogenous MOV10 was associated with NP during WSN infection in an RNA-dependent manner (Figure 3B, left panel). We also demonstrated that the endogenous MOV10 was associated with NP of the PR8 IAV, which was also RNA-dependent (Figure 3B, right panel). Since MOV10 is located at P-bodies , we asked whether MOV10 co-localized with NP in P-bodies. Indeed, we detected the co-localization of NP and MOV10 at cytoplasmic foci at 1.5 h post-infection (Figure 3C, upper panel). The histograms on the right showed the percentage of co-localized foci. To determine whether MOV10 co-localizes with NP that is derived from the incoming virion or is newly synthesized, we treated cells with CHX during WSN infection to block the synthesis of both cellular and viral protein. Owing to the weak signal we detected after CHX treatment, we exploited proximity ligation assay (PLA), which allowed signal amplification when two proteins lay proximate to each other. We detected MOV10 and NP-interacting signal with Duolink®-PLA (proximity ligation assay) (Figure 3D, left and right panels), while CHX readily suppressed the synthesis of NP (Supplementary Figure S2A). Furthermore, we also detected co-localization of endogenous MOV10 with NP in the presence of CHX (Figure 3E and Supplementary Figure S2B). Since we observed that MOV10 co-localized with both P-body marker DCP1a (Figure 3C, lower panel) and NP which is a component of viral RNP, we hypothesized that MOV10 interacted with NP and further sequesters viral RNP in the cytoplasmic P-bodies. In support of this hypothesis, immunoprecipitation of endogenous MOV10 from IAV-infected cells led to co-precipitation of not only NP but also viral PA, PB1 and PB2 proteins as well as viral RNA (Figure 3F).
MOV10 sequesters IAV RNP in the cytoplasm.
MOV10 restriction of IAV depends on P-body integrity
P-bodies are the apparatus for host cells to process cellular mRNA or miRNA/siRNA-targeted RNAs. In light of the suggestion that MOV10 may recruit IAV RNP into P-bodies, we asked whether P-bodies were necessary for MOV10 antiviral activity. To this end, we disrupted P-bodies by knocking down P-body component DDX6 and observed the disappearance of MOV10 foci (Figure 4A). Next, we checked the anti-IAV activity of MOV10 when disrupting P-bodies. The results showed that MOV10 restriction was largely lost when cellular DDX6 was depleted (Figure 4B, left and middle panels). We also confirmed this result with flow cytometry. The result showed that IAV replication was impaired by MOV10 overexpression, this restriction was alleviated by knocking down DDX6 (Figure 4B, right panel). In addition, we depleted both P-body component DDX6 and AGO2 along with MOV10. We also depleted stress granule protein G3BP1 along with MOV10 as a control. The results showed that depleting both MOV10 and G3BP1 led to a greater promotion of IAV replication that was observed by the depletion of MOV10 or G3BP1 individually (equal to the addition of knocking down MOV10 and G3BP1 individually). In contrast, knockdown of both DDX6 and AGO2 along with MOV10 did not further elevate IAV infection compared with when each of these three factors was depleted individually (Figure 4C). These results suggest that MOV10 and the P-bodies exploit the same mechanism to inhibit IAV which is distinct with the stress granule pathway. Since P-bodies are involved in the processing of siRNA-targeted mRNAs, knocking down its components (i.e. MOV10, DDX6 and AGO2) may interfere with siRNA functions. Thus, we generated MOV10, DDX6 and AGO2 knockout cell lines using G3BP1 as a control. Similar observations were made with these knockout cells in regard to inhibiting WSN infection as with siRNA knockdown experiments (Supplementary Figure S3). These results suggest that P-bodies and MOV10 use the same mechanism to restrict IAV replication, which is distinct from the role of stress granule in controlling IAV infection. Taken together, our results indicate that the anti-IAV activity of MOV10 depends on the function of P-bodies.
P-bodies are required for MOV10 inhibition of IAV.
Interaction with NP rather than the RNA helicase activity of MOV10 determines the inhibition of IAV
We have shown that MOV10 binds to NP in an RNA-dependent manner. We asked whether the helicase activity or the interaction with NP is required for MOV10 to inhibit IAV. We accordingly generated MOV10 mutants K531R and E647Q that lost the helicase activity as a result of the impairment in ATP binding and ATP hydrolysis, respectively  (Figure 5A). Using the IAV minigenome reporter system, we observed that both mutations inhibited luciferase expression as efficiently as the wild-type MOV10 (Figure 5B, the right panel showed representative expression of replicon component PA, MOV10 and IFITM3). We further mutated each of the seven helicase motifs in MOV10 and observed that these mutants restricted IAV replicon as efficiently as the wild-type MOV10 (Figure 5A and Supplementary Figure S4A). These data support the conclusion that MOV10 does not need its helicase activity to inhibit IAV.
Identification of the domain in MOV10 that mediates interaction with IAV NP as well as the anti-IAV activity of MOV10.
We next generated a series of truncated versions of MOV10 to determine the minimal sequence of MOV10 that exhibits anti-IAV activity (Figure 5C). Results of co-transfection of these MOV10 mutants with the IAV minigenome system revealed that the MOV10 region of amino acid 93–497 was essential to inhibit IAV (Supplementary Figure S4B). We then tested the anti-IAV activity of these MOV10 mutants on IAV WSN and also found that MOV10 93–497 amino acids are important for its ability to restrict IAV replication (Figure 5D). Using FITC-conjugated anti-NP antibody, we determined the percentage of infected cells using flow cytometry and obtained similar results (Figure 5D, right panel). We also found that MOV10 truncations that were defective in inhibiting IAV also lost its NP-binding ability (Figure 5E). Furthermore, we generated the MOV10 (93–497) construct and found that this construct was capable of binding to NP and restricting IAV (Figure 5F,G). We therefore conclude that MOV10 uses the same domain to inhibit IAV and bind viral NP, and that this anti-IAV function of MOV10 does not require the helicase activity.
NS1 interacts with MOV10 and counters the anti-IAV activity of MOV10
Chen et al.  reported peptide fragments of MOV10 in the NS1 complex in their mass spectrometry study. We confirmed the interaction of NS1 and MOV10 by performing co-immunoprecipitation experiment (Figure 6A). Given that NS1 is an antagonist of host innate immunity, we suspected that NS1 may attenuate the anti-IAV activity of MOV10. First, we examined the MOV10 and NP interaction in the presence of NS1. The results of co-immunoprecipitation showed that NS1 moderately diminished the association of MOV10 with NP (Figure 6B). Second, when levels of MOV10 in IAV-infected cells were monitored, we observed that MOV10 expression was declined as the MOI of IAV infection increased (Figure 6C). To determine whether this decrease in MOV10 expression is caused by NS1, we cotranfected HEK293T cells with MOV10 and NS1 DNA, and observed a dose-dependent reduction in MOV10 expression with an increasing expression of NS1 (Figure 6D). IAV infection or NS1 alone diminished the level of endogenous MOV10 (Figure 6E,F). Previously, Wang et al.  reported that IAV NS1 protein induced lysosomal degradation of cellular protein eIF4B. We suspected that NS1 may diminish MOV10 expression by the same mechanism. Therefore, we treated cells with lysosome inhibitor CQ or NH4Cl and measured the levels of MOV10 in the presence of NS1 expression. CQ and NH4Cl are endosome acidification inhibitors, and can inhibit IAV release from the endosome. To avoid the influence of CQ and NH4Cl on IAV entry, we treated cells with CQ and NH4Cl 1 h post WSN infection. The results showed that both CQ and NH4Cl rescued MOV10 expression in cells that were infected by WSN (Figure 6G). These results indicate that NS1 interacts with MOV10 and causes MOV10 degradation through the lysosomal pathway.
IAV NS1 protein counters MOV10 restriction.
Next, we tested whether NS1 counters the anti-IAV activity of MOV10. First, we tested whether NS1 affects MOV10 restriction of IAV replicon. The results showed that MOV10 inhibited the activity of IAV minigenome system by 60%, and that NS1 rescued IAV replicon expression close to the control level (Figure 7A, left panel). The rescue function of NS1 is concomitant with the reduction in MOV10 expression (Figure 7A, right panel). Since we could not obtain NS1-defective WSN, we deleted NS1 in IAV PR8. PR8 NP has been shown to interact with MOV10 (Figure 3B, right panel). The 13th codon of NS1 was mutated to a stop codon, which did not affect the expression of NEP that is encoded in the same RNA segment. Since NS1 deletion caused severely impaired replication of IAV, to obtain comparable viral protein expression levels, we used different MOIs of the wild-type and the NS1-deleted viruses in infections (for Figure 7B, 0.0005 vs. 0.01; for Figure 7C, 0.0002 vs. 0.005). The results of infection assays showed that the delNS1 PR8 was more severely inhibited by MOV10 compared with the wild-type PR8 (Figure 7B). In support of these data, knocking down endogenous MOV10 led to a much more significant increase in delNS1 PR8 infection than the wild-type PR8 virus (Figure 7C). Taken together, these results suggest that NS1 counters the anti-IAV activity of MOV10 through interacting with this cellular helicase.
NS1-defective IAV is more sensitive to MOV10 inhibition.
Here, we report the anti-IAV function of MOV10. MOV10 is an RNA helicase. Among its multiple cellular functions are its involvement in RNA interference pathway and its localization to the P-bodies. Our results suggest that its localization to P-bodies allows MOV10 to sequester the incoming IAV RNP complex in the cytoplasm and subsequently causing the degradation of viral vRNA as a result of MOV10 interaction with NP, a key component of IAV RNP. The sequestration and diminished vRNA could lead to the reduction in nuclear entry of IAV RNP and thus diminished viral RNA replication. This mechanism of inhibition by MOV10 has also been reported to restrict the activity of retrotransposon LINE-1, in which MOV10 was found to co-localize with LINE-1 ORF1 protein and LINE-1 RNA in stress granules that store untranslated mRNA and often juxtapose and exchange components with P-bodies . Yet, in the case of IAV, MOV10 restriction requires the integrity of P-bodies, but not that of stress granules, probably because formation of stress granule is inhibited by NS1 . However, we did detect an increase in IAV replication by knocking down G3BP1, which indicates that mechanisms other than stress granule formation are involved.
It is reported that MOV10 restricted VSV proliferation through facilitating IFN production depending on IKKε and IRF3, but independent of RIG-I and MAVS . We also tested whether IFN production is involved in the restriction of IAV function of MOV10. Using Vero cell lines, which are deficient of IFN production due to spontaneous gene deletion, we found that in the absence of IFN, overexpression of MOV10 could still restrict IAV replication (Supplementary Figure S1C), suggesting that MOV10 restricted IAV independent of IFN production.
Li et al. recently reported that IAV genomic RNA is recognized by the cellular RNAi machinery and subsequently processed into anti-IAV vsiRNA, which may, in turn, restrict IAV replication through the cellular RNAi pathway. NS1 protein acts as a VSR (virus-encoded suppressors of RNAi) and abolishes the production of viral siRNA . Consistent with their findings, we also observed increased replication of IAV when AGO2 was depleted. Moreover, depleting another key component of P-bodies DDX6 also promoted IAV replication, suggesting an inhibitory role of P-bodies in IAV replication even in the presence of NS1. P-bodies are where cellular mRNA and siRNA/miRNA-bound RNA are degraded. MOV10 has been shown to function in cellular mRNA quality control pathway NMD (non-sense-mediated mRNA decay) . Given our finding that MOV10 binds to IAV RNPs and sequesters them in P-bodies, and subsequently causing the degradation of incoming IAV vRNA, MOV10 may act as an adaptor that links IAV genomic RNA to P-bodies for degradation.
Our findings are in agreement with a recent study by Zhang et al.  who showed that MOV10 inhibits IAV by sequestering viral NP within the cytoplasm. We have further monitored the subcellular localization of the incoming viral NP protein at the early time points of IAV infection and observed that viral NP was co-localized with MOV10 in the P-bodies, which illustrates where and how viral RNP is sequestered. But, this mechanism is not the only one by which MOV10 restricts IAV, since we found that MOV10 also restricted IAV minigenome replication, which corroborates the study by Zhang et al. Their work reported an alternative mechanism of restricting IAV, in which MOV10 impairs IAV replication by restricting the nuclear entry of newly synthesized NP. Consistent with their study, we also found that the interaction between NP and MOV10 is essential for MOV10 restriction of the replication of IAV minigenome as well as IAV itself. Our results further showed that in the case of IAV infection, MOV10 traps the incoming IAV vRNP by interacting with NP. We have further mapped this inhibitory activity of MOV10 to the N-terminal region from amino acid position 93–497 that does not have the RNA helicase domains. Our results support the data by Zhang et al., showing that the first 170 amino acids are important for MOV10 anti-IAV function.
DDX21, which is also an RNA helicase, has been reported to interact with IAV polymerase component PB1 and restrict IAV replication by disrupting the viral polymerase complex. This function of DDX21 is counteracted by NS1 by diminishing the association of PB1 with DDX21 . Similarly, we found that RNA helicase MOV10 binds to IAV polymerase component NP and sequesters IAV RNP in P-bodies. NS1 protein also directly associates with MOV10 and counteracts MOV10 restriction partially by abolishing the interaction between MOV10 and NP, which suggests a common mechanism that NS1 uses to antagonize host restriction.
In addition to illustrating the anti-IAV activity of MOV10 and its molecular mechanism of action, we have also discovered a viral countering strategy that is mediated by the viral antagonist NS1 [21,34–36]. Results of our study suggest that NS1 counters the anti-IAV cellular factor MOV10, which further illuminates the importance of NS1 in antagonizing host antiviral defense and promoting IAV infection.
Taken together, our results have revealed a new layer of host defense against IAV infection, which is mediated by MOV10 that acts through interacting with viral NP protein. This interaction leads to sequestration of the incoming viral RNP in the cytoplasmic P-bodies, thus inhibiting IAV replication. The anti-IAV activity of MOV10 is partially overcome by viral NS1 protein as a result of interaction of NS1 with MOV10.
F.G., C.L., J.L. and S.H. conceived and designed the study. J.L., S.H., F.X., and Z.X. carried out the experiments. F.G., C.L., J.W., J.L., and S.C. analyzed the data. F.G., C.L., J.L., and S.H. drafted the manuscript. J.L., S.H., F.X., S.M., X.L., L.Y., F.Z., X.Z., H.S., Z.X., D.Z., S.C., J.W., C.L., and F.G. read and approved the final manuscript.
The present study was supported by funds from the Ministry of Science and Technology of China [2018ZX10301408-003, 2018ZX10731101-001-018, 2012CB911103 and 2013ZX10001005-001-002], from the National Key Plan for Scientific Research and Development of China [2016YFD0500307], from the National Natural Science Foundation of China [81601771, 81371808, 81528012, 81702451 and 81401673], from CAMS Innovation Fund for Medical Sciences (CIFMS 2018-I2M-3-004 and CIFMS 2016-I2M-1-014), and the Canadian Institutes of Health Research [CCI-132561]. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
We are grateful to Dr Yuelong Shu and Dr Dayan Wang (Chinese Center for Disease Control and Prevention) for generously providing A/H1N1pdm09, A/H3N2 and A/PR/8/34 viruses.
The Authors declare that there are no competing interests associated with the manuscript.