Mumps Virus Induces Protein-Kinase-R-Dependent Stress Granules, Partly Suppressing Type III Interferon Production

Stress granules (SGs) are cytoplasmic granular aggregations that are induced by cellular stress, including viral infection. SGs have opposing antiviral and proviral roles, which depend on virus species. The exact function of SGs during viral infection is not fully understood. Here, we showed that mumps virus (MuV) induced SGs depending on activation of protein kinase R (PKR). MuV infection strongly induced interferon (IFN)-λ1, 2 and 3, and IFN-β through activation of IFN regulatory factor 3 (IRF3) via retinoic acid inducible gene-I (RIG-I) and the mitochondrial antiviral signaling (MAVS) pathway. MuV-induced IFNs were strongly upregulated in PKR-knockdown cells. MuV-induced SG formation was suppressed by knockdown of PKR and SG marker proteins, Ras-GTPase-activating protein SH3-domain-binding protein 1 and T-cell-restricted intracellular antigen-1, and significantly increased the levels of MuV-induced IFN-λ1. However, viral titer was not altered by suppression of SG formation. PKR was required for induction of SGs by MuV infection and regulated type III IFN (IFN-λ1) mRNA stability. MuV-induced SGs partly suppressed type III IFN production by MuV; however, the limited suppression was not sufficient to inhibit MuV replication in cell culture. Our results provide insight into the relationship between SGs and IFN production induced by MuV infection.


Introduction
Mumps is an infectious disease caused by mumps virus (MuV) and is characterized by swelling of the parotid gland [1]. Mumps has severe characteristic complications such as aseptic meningitis, encephalitis, severe sensory hearing loss, pancreatitis and orchitis. The disease can be prevented by vaccination with attenuated live vaccine, which is used universally in many countries around the world. MuV is an enveloped single negative strand RNA virus that belongs to the R (HuR), and insulin-like growth factor 2 binding protein (IGF2BP)1 were purchased from Active Motif (Carlsbad, CA), Abnova (Taipei, Taiwan), Immuno-Biological Laboratories (Gunma, Japan), Medical and Biological Laboratory (Nagoya, Japan) and Abgent (San Diego, CA), respectively. Mouse mAbs against MAVS (E-3) and β-actin (AC-74) were purchased from Santa Cruz Biotechnology (Dallas, TX) and Sigma-Aldrich (St. Louis, MO), respectively. Rabbit PcAb against mumps P/V protein (T61), and mouse mAb against mumps F protein (44C) were generated as previously described [21][22][23]. Actinomycin D was purchased from Wako Pure Chemicals (Osaka, Japan) and dissolved with dimethyl sulfoxide.

Cells and Viruses
FL (human amnion) and Vero (African green monkey kidney) cells were obtained from American Type Culture Collection (ATCC; Manassas, VA), and were maintained in RPMI-1640 medium supplemented with 10% (v/v) fetal bovine serum, 100 U/ml penicillin and 10 μg/ml streptomycin. BEAS-2 (human bronchial epithelial) cells were obtained from ATCC, and maintained in LHC-9 medium supplemented with 100 U/ml penicillin and 10 μg/ml streptomycin. The MuV strain Torii was grown in Vero cells, and supernatant containing MuV was frozen in liquid nitrogen and stored at −80°C until use. Cells were infected with MuV at a multiplicity of infection (MOI) of 0.005-0.5 at 37°C for 1 h. After infection, cells were washed twice with prewarmed medium and then incubated for 24 h. The viral titers in supernatants were determined by plaque-forming assays in Vero cells as follows. Samples serially diluted 10-fold were adsorbed onto Vero cells in 24-well plates at 37°C for 1 h. After incubation, infected Vero cells were washed twice with phosphate-buffered saline (PBS) and incubated in culture medium supplemented with 1% carboxymethyl cellulose for 6 days. After incubation, cells were stained with staining solution [0.5% (w/v) crystal violet, 5% (v/v) formalin, 50% (v/v) ethanol, and 0.15 M NaCl]. Experimental methods and preparation of human respiratory syncytial virus (RSV) were as described previously [24].

Immunocytochemistry
The cells infected with MuV were grown in a 35-mm glass-bottom dish coated with type I collagen (Matsunami Glass, Kishiwada, Japan), and fixed with 4% paraformaldehyde at room temperature for 15 min. After rinsing in PBS, the cells were treated with PBS containing 0.1% Triton X-100 for 5 min, and then with blocking buffer (PBS containing 2% bovine serum albumin and 0.02% Triton X-100) for 20 min at room temperature. Cells were incubated with primary antibodies for 2 h at room temperature. After rinsing in PBS, the specimens were incubated with secondary antibodies, Alexa Fluor 488 (green)-conjugated anti-rabbit IgG or Alexa Fluor 594 (red)-conjugated anti-mouse IgG (Thermo Fisher Scientific, Waltham, MA), for 2 h at room temperature. Nuclei were stained with 4 0 ,6-diamino-2-phenylindole (DAPI) for 2 h at room temperature. After the stained slides were washed with PBS, the cells were embedded in 2% 1,4-diazabicycle [2.2.2.] octane (Sigma-Aldrich) and 50% glycerol in PBS. The specimens were examined and photographed using confocal microscopy (ConfoCor 3 LSM 510 META; Carl Zeiss, Oberkochen, Germany).

RNA Interference
Silencer selected small-interfering RNAs (siRNAs) were obtained from Thermo Fisher Scientific. Targeting siRNA was transfected using Lipofectamine RNAimax Transfection Reagent (Thermo Fisher Scientific).

Western Blotting
Cells were lysed with lysis buffer (1% Triton X-100, 0.5% sodium deoxycholate, 0.15 M NaCl, and 2 mM EDTA in 50 mM HEPES-NaOH, pH 7.4) supplemented with protease inhibitor cocktail (Thermo Fisher Scientific). After centrifugation at 20,000 × g for 5 min, protein concentration in supernatant was determined using the bicinchoninic acid method (Thermo Fisher Scientific). Sodium dodecyl sulfate polyacrylamide gel electrophoresis and western blotting were performed as described previously [25].

Statistical Analysis
Digital images of western blotting were quantitated using public domain Image J program (National Institutes of Health, Bethesda, MD). The significance of differences between two groups was analyzed by two-tailed Student's t test.

MuV-Induced SG Formation
We examined whether MuV induced SGs. In MuV-infected cells, at 24 h post-infection, MuV P and V proteins were detectable at MOI >0.05, and mRNAs of P/V and F at MOI >0.005 ( Fig  1A and 1B). PKR and eIF2α are regulators of SGs. PKR was strongly phosphorylated by MuV infection, whereas eIF2α was only weakly phosphorylated ( Fig 1A). As a reference, we used sodium arsenite, which induces oxidative stress and activation of HRI [15,26]. Sodium arsenite induced phosphorylation of eIF2α after 15 min treatment, which was dramatically increased after 1 h treatment ( Fig 1C). Phosphorylation of eIF2α induced by MuV, which was mainly mediated by PKR, was less than that induced by sodium arsenite. These results suggested that MuV infection led to moderate shutoff of protein translation, whereas sodium arsenite showed the strong potency of shutoff of protein translation in FL cells ( Fig 1C).
We confirmed that foci (aggregations) containing SG marker proteins, namely G3BP1, IGF2BP1 and HuR (also known as ELAV-like protein) [10,27], were induced by sodium arsenite (Fig 2). Similar foci were observed in the cytosol of MuV-infected cells, while G3BP1 and IGF2BP1 were mainly localized in the cytosol and HuR in the nucleus in untreated cells (Fig 2). MuV P/V proteins were observed as inclusion bodies in the cytosol, and MuV F protein was in the cell membrane. MuV P/V and F proteins did not colocalize with the SG markers. We also observed that MuV-induced foci contained G3BP1 and upregulation of phosphorylation of PKR in human bronchial epithelium cell line BEAS-2B (data not shown). These results indicated that MuV induced formation of SGs.

Relationship between SG Formation and IFN Response
Virus-induced SGs are reported to affect IFN induction and viral replication [10]. We investigated the contribution of MuV-induced SGs to viral replication and IFN production, and their relationship with RLR-related proteins, such as RIG-I, MDA5 and MAVS. We examined SG formation in PKR-, RIG-I-, MDA5-and MAVS-knockdown (KD) cells. In all KD cells, SGs were formed by treatment with sodium arsenite (data not shown). MuV-induced SGs were not observed in PKR-KD cells, while they were observed in RIG-I-, MDA5-and MAVS-KD cells, similar to control cells ( Fig 3A). SGs containing G3BP1 were observed in~80% of MuVinfected control cells. Phosphorylation of PKR was stimulated by MuV infection in the MDA5-, RIG-I-, MAVS-KD as well as control cells (Fig 3B). eIF2α was phosphorylated by MuV infection in all KD cells. eIF2α is a substrate protein of PKR; however, it was phosphorylated by MuV infection in PKR-KD cells. The results suggested that MuV induced eIF2α phosphorylation in a PKR-independent manner.
We determined the production of IFNs in the KD cells ( Fig 3C). IFN production by MuV almost completely disappeared in RIG-I-and MAVS-KD cells. It was not altered in MDA5-KD cells. MuV-infection-induced production of IFN-β, -λ1, -λ2 and -λ3 was significantly higher in PKR-KD cells, as compared with the control cells ( Fig 3C). IRF3, which is the main transcription factor for production of IFN-β and IFN-λs, was clearly phosphorylated by MuV infection, but not in MAVS-and RIG-I-KD cells. MDA5 and RIG-I induced by MuV infection disappeared in MAVS-KD cells (Fig 3B). Virus titer in the culture supernatants of RIG-I-and MAVS-KD cells significantly increased in comparison to that in control cells, while viral titer was not altered in PKR-and MDA5-KD cells (Fig 3D). These results indicated that MuV was recognized by RIG-I but not MDA5, and produced type I and III IFNs through a MAVSand IRF3-dependent signal transduction pathway that suppressed viral replication. SG formation induced by MuV was dependent on PKR-dependent, but not RLR-dependent signal  transduction. There was evidence that PKR-dependent SG formation partly suppressed IFN production in MuV-infected cells.
We investigated induction of IFN mRNA and activation of transcription factors in the KD cells (Fig 4). mRNA of IFN-β and -λ1 was markedly induced at 12 h post-infection, but not in MAVS-and RIG-I-KD cells (Fig 4A). The ISG MxA was not stimulated by MuV infection. IFN and MxA mRNA expression in PKR-KD cells was not altered in comparison to control cells   ( Fig 4A). The activity of IRF and NF-κB was measured by reporter gene assay using a luciferase expression plasmid containing IFN-β p125 promoter, PRD III-I (an IRF-binding sequence), and ELAM promoter (mainly regulated by NF-κB [28]) (Fig 4B). Luciferase activity was stimulated in control and PKR-KD cells by MuV infection at comparable levels. In MAVS-and RIG-I-KD cells, MuV-stimulated IFN-β promoter and PRD III-I activities were almost completely suppressed (Fig 4B). MuV-stimulated NF-κB activity was markedly decreased in RIG-KD cells, and decreased~50% in MAVS-KD cells compared with control cells. These data suggest that PKR is not directly related to NF-κB and IRF3 activation via the RIG-I/MAVS/ IRF3 signaling cascade in MuV-infected cells.

PKR Regulated IFN-λ1 mRNA Stability
Compared with control cells, increased MuV-induced IFN protein production was observed only in PKR-KD cells (Fig 3C). The levels of MuV-induced IFN-β and -λ1 mRNAs were not significantly altered by PKR-KD (Fig 4A). IFN-β mRNA was induced by 337.8-and 430.1-fold in control and PKR-KD cells, respectively, and IFN-λ1 mRNA was induced by 3119-and 3384-fold in control and PKR-KD cells, respectively ( Fig 4A). SGs are known to regulate mRNA localization, stability and translation after transcription [29]. We therefore hypothesized that PKR controls the stability of IFN mRNA induced by MuV infection through SGs. The half-lives of mRNAs were determined by real-time RT-PCR after terminating transcription by actinomycin D, a transcription inhibitor, which does not alter and induce SGs [30]

SGs Partly Suppressed IFN-λ Production, but Did Not Affect Viral Replication
To investigate the contribution of MuV-induced SGs to IFN production and viral replication, we used cells with KD of the typical SG components, G3BP1 and TIA-1. SGs containing IGF2BP1 protein induced by MuV infection (24 h post-infection) were only reduced by~50% in TIA-1-or G3BP1-KD cells compared with control cells, although mRNA expression was decreased to a greater extent (data not shown). Therefore, we prepared double TIA-1/ G3BP1-KD cells, and MuV-infection induced SGs in 4.7% of these cells, compared with 45.6% of control cells (Fig 6A and 6B). We confirmed that SGs containing HuR protein were highly decreased by TIA-1 and G3BP1 double-KD (data not shown). Moreover, TIA-1 and G3BP1 mRNA levels were decreased to <10% (Fig 6C). MuV-induced phosphorylation of eIF2α and PKR was comparable in control and double-KD cells (Fig 6D). IFN-λ1, but not IFN-β, production induced by MuV infection was significantly higher in the double-KD cells than control cells ( Fig 6E); however, MuV-induced phosphorylation of IRF3 and viral replication were not significantly altered by double-KD (Fig 6D and 6F). These data indicated that MuV-induced SGs suppressed IFN-λ production to some extent; however, they did not influence viral replication in these particular experimental conditions. We examined whether SGs were related to induction of IFNs in RSV-infected BEAS-2 cells (Fig 7). RSV-induced SGs were formed in a PKR-dependent manner (Fig 7A). RSV infection markedly induced IFN-β and -λ1 production. The amount of IFNs induced was significantly higher in PKR-KD cells than control cells, but it was not altered in G3BP1 and TIA-1 double-KD cells compared with control cells (Fig 7B). Neither PKR-KD nor double-KD cells formed SGs in response to RSV infection (Fig 7A). Expression of viral proteins was not altered in these KD cells compared with control cells (Fig 7C). These results indicated that RSV-induced SGs did not interfere with IFN production and viral replication; however, PKR suppressed RSVinduced IFN protein production.

Discussion
We demonstrated that MuV induced SGs in a PKR-dependent manner and strongly induced IFN-β (type I) and IFN-λs (type III). MuV was recognized by RIG-I and activated IRF3 via MAVS, resulting in IFN production. IFN-λs were strongly stimulated by MuV infection, more than IFN-β.
We found that protein and mRNA levels of IFN-λ1 induced by MuV were markedly higher than those of IFN-β (Fig 4). IFN-λ1 promoter contains transcription factor binding motifs of GATA binding protein 1 (GATA-1), IRF3 and NF-κB. IRFs and NF-κB are necessary for maximum expression induced by poly I:C, a mimic compound of viral dsRNA [37]. IFN-β promoter contains binding motifs of activator protein-1 (AP-1), NF-κB, and two IRFs, and IRFs and NF-κB are necessary for maximum expression, as well as IFN-λ [38]. IRFs and NF-κB were significantly stimulated by MuV infection (Fig 4B). The difference in expression levels of IFN-λ1 and -β proteins might be explained by mRNA stability. Stability of mRNA is regulated by RNAbinding proteins [39]. IL-6 mRNA is stabilized and degraded by AT-rich interactive domaincontaining protein 5A and Regnase-1, respectively [40,41]. IFN-β mRNA is stabilized by HuR protein, which binds to A/U-rich elements on 3 0 untranslated regions [18,42]. Prokunia-Olsson et al. reported that IFN-λ mRNA was still detectable 24 h after poly I:C stimulation [43]. In contrast, type I IFN mRNA expression seemed to be transient, and was decreased within 10 h after viral infection [44]. In MuV-infected cells, half-life of IFN-λ1 mRNA was 2-fold longer than that of IFN-β mRNA. The higher production of IFN-λ1 compared with IFN-β might be related to the prolonged half-life of its mRNA (Fig 5). In addition, the half-life of IFN-λ1 mRNA was shorter in PKR-KD cells. We anticipated as a working hypothesis that PKR-dependent SGs act to sequester IFN-λ1 mRNA in MuV-infected cells. Although it is conceivable that the half-life was regulated by PKR-related RNA-binding proteins, this was complicated by the fact that PKR-KD cells expressed higher levels of IFN-λ1 proteins following induction by MuV infection. However, the mRNAs could be sequestered and stabilized in PKR-dependent SGs and not be used for translation in MuV-infected cells.
There are three types of SG formation depending on virus species: stable formation, transient formation, and no formation [10]. MuV infection induces SGs, which were observed iñ 85% of MuV-infected cells. Thus, MuV might induce stable formation of SGs. MeV, SeV and IAV do not form SGs, whereas mutant viruses with deleted nonstructural proteins do form SGs. Their nonstructural proteins prevent PKR activation [18-20, 45, 46]. C protein deletion mutant of MeV forms SGs, and the SGs stimulate IFN-β production and suppress viral replication [19]. NS1 deletion mutant of IAV also induces SGs and produces excess IFN-β, resulting in suppression of viral RNA replication and protein synthesis [18,46]. These SGs interfere with IFN production and play an antiviral role. However, it is still not completely understood whether SG formation interferes antivirally or provirally. Hepatitis C virus (HCV) induces SGs in a PKR-dependent manner. HCV-induced SGs suppress translation of ISG proteins, such as MxA and ubiquitin-specific protease 18, and are required for assembly and egress of viral particles [47]. Suppression of RSV-induced SG formation by PKR-KD suggests that SGs do not interfere directly with viral replication, while IFN-β production is suppressed [48,49]. However, suppression of RSV-induced SG formation by G3BP1-KD suggests that SGs facilitate viral replication [50]. These SGs play a proviral role. In the present study, RSV-induced PKRdependent SGs suppressed IFN-λ1 and -β production, whereas suppression of SG formation by KD of G3BP1 and TIA-1 did not alter IFN production (Fig 7B). The relationship between SGs and IFN production/viral replication seems to be complicated. MuV infection induced strong PKR phosphorylation and weak eIF2α phosphorylation. Although production of IFN, especially IFN-λ, was potentially suppressed by SGs induced by MuV infection in a PKR-dependent manner, the SGs did not promote viral replication. There are two possible reasons why SG-dependent suppression of IFNs had no effect on viral replication. (i) PKR induces SG formation in viral infection, and has antiviral activity, as an ISG [51,52]. For example, PKR phosphorylates eIF2α, which prevents the generation of viral proteins by transient stalling of translation. (ii) MuV-infected cells do not respond to IFNs, because MuV V protein strongly suppresses IFN signal transduction through degradation of STAT1 [53,54]. Nevertheless, MuV-induced PKR activation and SG formation might partly prevent IFN production; however, suppression of IFNs might be insufficient for suppression of viral replication in the experimental infection system used in this study. The role of PKR and its relationship with IFN production in MuV infection are still not fully understood; therefore, we need to elucidate this mechanism in a future study.
In conclusion, MuV induces SGs in a PKR-dependent manner and partly suppresses MuVinduced IFN production. MuV infection induces type I and III IFNs, and RIG-I and MAVS are important molecules for IFN production. Our results promote understanding of the relationship between virus-induced SGs and the innate immune response; however, further studies are needed to understand the role of SGs and innate immune responses against viral infection, including mumps.