Gene-Specific Inhibition of Reovirus Replication by RNA Interference

Mammalian reoviruses contain a genome of 10 segments of double-strandedRNA (dsRNA). Reovirus replication and assembly occur withindistinct structures called viral inclusions, which form in thecytoplasm of infected cells. Viral nonstructural proteins µNSand ${sigma}$ NS and core protein µ2 play key roles in forming viralinclusions and recruiting other viral proteins and RNA to thesestructures for replication and assembly. However, the precisefunctions of these proteins in viral replication are poorlydefined. Therefore, to better understand the functions of reovirusproteins associated with formation of viral inclusions, we usedplasmid-based vectors to establish 293T cell lines stably expressingsmall interfering RNAs (siRNAs) specific for transcripts encodingthe µ2, µNS, and ${sigma}$ NS proteins of strain type 3 Dearing(T3D). Infectivity assays revealed that yields of T3D, but notthose of strain type 1 Lang, were significantly decreased in293T cells stably expressing µ2, µNS, or ${sigma}$ NS siRNA.Stable expression of siRNAs specific for any one of these proteinssubstantially diminished viral dsRNA, protein synthesis, andinclusion formation, indicating that each is a critical componentof the viral replication machinery. Using cell lines stablyexpressing µNS siRNA, we developed a complementation systemto rescue viral replication by transient transfection with recombinantT3D µNS in which silent mutations were introduced intothe sequence targeted by the µNS siRNA. Furthermore, wedemonstrated that µNSC, which lacks the first 40 aminoresidues of µNS, is incapable of restoring reovirus growthin the complementation system. These results reveal interdependentfunctions for viral inclusion proteins and indicate that celllines stably expressing reovirus siRNAs are useful tools forthe study of viral protein structure-function relationships.

INTRODUCTION

Top

Introduction

Viral replication and assembly often take place in intracellularcompartments called viral inclusions, where viral componentsconcentrate. For several viruses, formation of viral inclusionsoccurs at distinct cytoplasmic sites, such as the perinucleararea, and involves complex interactions between viral and cellularfactors. Studies of the composition and organization of viralinclusions have provided insight into essential steps in viralreplication.

Mammalian orthoreoviruses (reoviruses) are members of the familyReoviridae and contain a genome of 10 segments of double-strandedRNA (dsRNA) (43). The genome is encapsidated by two proteinshells, termed outer capsid and core. Following internalizationby receptor-mediated endocytosis (3, 6, 51, 59), the core isreleased into the cytoplasm and begins to synthesize capped,single-stranded RNA (ssRNA) copies of the 10 dsRNA genome segments(6, 19, 23, 56). These mRNAs are competent for translation andserve as templates for minus strand synthesis, resulting information of nascent genomic dsRNA (33, 52, 55). Synthesis ofthe complementary strand appears to be concomitant with assortmentof the 10-genome segments into progeny particles (1). Reovirusassembly is completed by the addition of outer-capsid proteins,resulting in the formation of mature, double-shelled virions(40).

Reovirus replication and assembly are thought to occur withinviral inclusions that form in the cytoplasm of infected cells(21). Viral inclusions contain dsRNA (57), viral proteins (21),and both complete and incomplete particles (21). These structuresare devoid of ribosomes (54) and cellular membranes (27, 46),suggesting that they offer an environment uniquely suited toefficient viral replication and assembly. However, it is notknown how viral inclusions are organized with respect to structureor function. Core protein µ2 and viral nonstructural proteinsµNS and ${sigma}$ NS play important roles in forming viral inclusionsand recruiting other viral proteins and RNA to inclusion structuresfor replication and assembly (4, 5, 7, 9, 36, 39, 45, 65).

The reovirus µ2 protein is encoded by the M1 genome segmentand forms a structurally minor component of the viral core (14,37, 41). The M1 genome segment is associated with viral-strain-specificdifferences in the in vitro transcriptase and nucleoside triphosphatase(NTPase) activities of viral particles (44, 64). M1 is alsothe genetic determinant of strain-specific differences in themorphology—globular or filamentous—of viral inclusions(36, 45). The µ2 proteins of some reovirus strains interactwith and stabilize microtubules, which are properties responsiblefor the filamentous inclusion morphology exhibited by prototypestrains type 1 Lang (T1L) and type 2 Jones (45, 65). Differencesin inclusion morphology are correlated with a single-amino-acidpolymorphism in µ2 at position 208 (45).

The reovirus µNS protein is encoded by the M3 genome segment(37, 41) and associates with viral mRNAs (1) and viral cores(8) but does not inhibit viral transcription or capping activities(8). Transiently expressed µNS forms inclusion-like structures,which are similar in appearance and localization to the globularinclusions observed in cells infected with prototype straintype 3 Dearing (T3D) (9). In addition, µNS associateswith µ2 and recruits viral core proteins ${lambda}$ 1, ${lambda}$ 2, and ${sigma}$ 2 intoviral inclusion-like structures when these proteins are coexpressedin transiently transfected cells (7, 9). These findings suggestthat µ2 and µNS together are essential for the formationand morphology of viral inclusions. µNSC, also producednaturally during reovirus infection (32), lacks the 40 amino-terminalresidues of µNS and has been proposed to be a productof translation initiation from an alternative site in M3 mRNA(63). The 40 amino-terminal residues of µNS contain interactingdomains for viral proteins µ2 and ${sigma}$ NS (9, 39). However,the role of µNSC in reovirus replication has not beenelucidated.

The reovirus ${sigma}$ NS protein, encoded by the S3 genome segment, alsoplays a role in viral inclusion formation. Temperature-sensitive(ts) reovirus mutant tsE320, which contains a mutation in theS3 genome segment responsible for the ts phenotype, does notform inclusions when grown at nonpermissive temperature (4).The ${sigma}$ NS protein binds ssRNA, including reovirus mRNAs, and formshigher-order structures (24-26, 30, 47). Large complexes containing ${sigma}$ NS prepared from infected cells are dissociated by treatmentwith RNase A (24, 30), suggesting that RNA stabilizes thesestructures. Furthermore, ${sigma}$ NS colocalizes with µNS in viralinclusions in reovirus-infected cells (5, 39). Although thefunction of ${sigma}$ NS in viral replication has not been determined,it is possible that ${sigma}$ NS is involved in recruitment of viral RNAsinto or retention of viral RNAs within inclusions.

RNA interference (RNAi) is a sequence-specific gene-silencingmechanism that employs dsRNA to produce short interfering RNAs(siRNAs) approximately 21 nucleotides (nt) in length (12, 18).siRNAs assemble into endoribonuclease-containing complexes knownas RNA-induced silencing complexes and subsequently guide thesecomplexes to degrade mRNAs containing complementary sequences(29, 53). Since its discovery, RNAi has been developed intoa widely used gene-silencing technique for studies of gene function.Indeed, numerous studies in which 21-nt synthetic siRNAs (12,18) or siRNAs transcribed from an RNA polymerase III promoter,such as U6 (60, 66) or H1 (10, 38), were used to specificallysuppress the expression of endogenous genes in mammalian cellswithout activating nonspecific responses to dsRNA have beenreported. Similarly, siRNAs have been used to interfere withthe replication of a number of viruses, including rotaviruses(2, 11, 17, 28, 34, 35, 58), which are closely related to reoviruses.Thus, the technology of siRNA-mediated gene silencing has thecapability to provide information necessary to understand mechanismsof viral replication and functions of individual viral genes.

To understand the function of reovirus proteins associated withthe formation of viral inclusions, we established cell linesstably expressing siRNAs specific for transcripts encoding theµ2, µNS, and ${sigma}$ NS proteins of T3D by using plasmid-basedvectors. Our results reveal that the effects of siRNAs on reovirusinfection are strain specific and that µ2, µNS,and ${sigma}$ NS proteins are each involved in inclusion formation andmaturation as indispensable components of viral replication.Furthermore, we have established a complementation system tostudy reovirus replication by using a cell line stably expressingreovirus M3 siRNA. This system was used to demonstrate thatµNS but not µNSC is required for reovirus replication.

MATERIALS AND METHODS

Top

Materials and Methods

Cells and viruses. Murine L929 (L) cells were grown in Joklik's modified Eagle'sminimal essential medium (Irvine Scientific, Santa Ana, Calif.)supplemented to contain 5% fetal calf serum (Gibco-BRL, Gaithersburg,Md.), 2 mM L-glutamine, 100 U of penicillin G per ml, 100 µgof streptomycin per ml, and 250 ng of amphotericin B per ml(Gibco-BRL). Human embryonic kidney 293T cells were maintainedin Dulbecco's modified Eagle's medium (Gibco-BRL) supplementedto contain 10% fetal calf serum, 2 mM L-glutamine, 100 U ofpenicillin G per ml, 100 µg of streptomycin per ml, and250 ng of amphotericin B per ml. To establish cell lines stablyexpressing reovirus siRNAs, 293T cells transfected with siRNAexpression vectors to suppress reovirus M1, M3, or S3 gene expressionwere selected using 5 µg per ml of puromycin (CALBIOCHEM,San Diego, Calif.).

Reovirus prototype strains T1L and T3D and reassortant virusesG16, EB85, and EB113 derived from crosses of T1L and T3D (42)are laboratory stocks. Purified virion preparations of reoviruswere made using second-passage L-cell lysate stocks of twice-plaque-purifiedreovirus as previously described (22). Virus titers were determinedby plaque assay using L-cell monolayers as previously described(61).

Plasmid construction. To generate plasmid-based vectors expressing reovirus M1, M3,or S3 siRNAs, the pSUPER RNAi system (OligoEngine, Seattle,Wash.) was used in concert with a pair of custom oligonucleotidesthat contain a unique 19-nt sequence derived from the T3D M1,M3, or S3 transcripts targeted for suppression (Table 1). Oligonucleotidescontaining unique sequences in the sense and antisense orientationswere annealed and cloned into the pSUPER.puro vector betweenthe BglII and HindIII restriction sites 3' of the polymeraseIII H1-RNA promoter. Mammalian expression vectors pCMVM3wt andpCMVS3wt, containing the open reading frames of the T3D µNSand ${sigma}$ NS proteins, respectively, were described previously (5).To generate a mammalian expression vector encoding the T3D µ2protein (pcFM1T3) fused to a FLAG epitope tag at the amino terminus,an M1 cDNA fragment was amplified by reverse transcription andPCR using viral dsRNA extracted from purified reovirus virionsas a template. Amplified cDNA fragments were cloned into theXhoI-KpnI site of the pcFLAG vector, which was constructed byinsertion of the FLAG tag sequence into the multicloning siteof the pcDNA3 vector (Invitrogen, San Diego, Calif). To constructthe mammalian expression vector pCMVM3_1861m, in which threesilent mutations were introduced into the M3 1861 siRNA sequence,a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla,Calif.) was used with the primer set 5'-CGTATTTCTAAGGAAGCAGCaGCgAAgTGTCAAACTGTTATTGATGAC-3'and 5'-GTCATCAATAACAGTTTGACAcTTcGCtGCTGCTTCCTTAGAAATACG-3' (lowercaseletters indicate altered nucleotides in the T3D M3 gene at positions1866, 1869, and 1871) and pCMVM3wt as a template. Fragmentsof M3 cDNA containing site-directed substitutions were PCR amplifiedusing the primer set 5'-TAAGGTACCATGGCTTCATTCAAGGGATTC-3' and5'-TAACTCGAGTTACAACTCATCAGTTGGAAC-3' and pCMVM3_1861m as a template.Amplicons were cloned into the KpnI-XhoI site of pcDNA3 to generatepcM3_1861m. The µNSC cDNA was amplified with the specificprimer set 5'-TAAGGTACCATGTCTCAATCGCGTGAATTCC-3' and 5'-TAACTCGAGTTACAACTCATCAGTTGGAAC-3'and pcM3_1861m as a template. Amplified fragments were clonedinto the KpnI-XhoI site of pcDNA3 to generate pcM3_dN138. pcM3_1ATGmand pcM3_2ATGm, containing site-direct substitutions of ATG toGCG in initiation codons corresponding to µNS (nucleotidepositions 19 to 21 of M3 RNA) and µNSC (nucleotide positions139 to 141 of M3 RNA), respectively, were generated using aQuikChange site-directed mutagenesis kit with the primer sets5'-GAGACCCAAGCTTGGTACCgcGGCTTCATTCAAGG-3' and 5'-CCTTGAATGAAGCCgcGGTACCAAGCTTGGGTCTC-3'(pcM3_1ATGm) and 5'-CTCCGTCTGTGGATgcGTCTCAATCGCGTGAATTC-3' and5'-GAATTCACGCGATTGAGACgcATCCACAGACGGAG-3' (pcM3_2ATGm) (lowercaseletters indicate altered nucleotides in the T3D M3 gene) andpcM3_1861m as a template.

View this table:
[in this window]
[in a new window]
TABLE 1. Oligonucleotides for siRNA expression

Protein expression in mammalian cells. 293T cells were transfected using Lipofectamine 2000 transfectionreagent (Invitrogen) according to the manufacturer's instructions.Transfected cells were subjected to immunofluorescence assayor immunoblot analysis.

Antibodies. To generate polyclonal antiserum against µNS, the T3DM3 gene was cloned 3' of sequences encoding glutathione S-transferase(GST) in the pGEX-4T-3 vector (Amersham Biosciences, Piscataway,NJ). The GST-µNS fusion protein expressed in BL21-DE3cells (Novagen, Madison, WI) was purified from the soluble fractionaccording to the manufacturer's instructions by using GSTrapaffinity chromatography (Amersham Biosciences). The GST tagwas removed by treating eluted protein with 20 units of thrombin(Amersham Biosciences) at 25°C for 2 h. Rabbits were immunizedand boosted with precipitated µNS protein to generateµNS-specific serum (Cocalico Biologicals Inc., Reamstown,PA). ${sigma}$ 3-specific monoclonal antibody (MAb) 4F2 (62) and antiseraspecific for ${sigma}$ NS (5) and µ2 (67) have previously been described.

Immunofluorescence staining. Cells were infected with reovirus at a multiplicity of infection(MOI) of 10 PFU per cell and seeded onto collagen-coated glasscoverslips (Fisher Scientific, Pittsburgh, PA). After incubationfor 20 h, cells were fixed with 1:1 methanol-acetone, washedwith phosphate-buffered saline (PBS), and incubated with antiseraspecific for ${sigma}$ NS, µNS, or µ2. After two washes withPBS, cells were incubated with Alexa Fluor 488 goat anti-rabbitimmunoglobulin G (IgG) or Alexa Fluor 546 goat anti-guinea pigIgG (Molecular Probes, Inc., Eugene, Oreg.) at a dilution of1:1,000. Cells also were incubated with TO-PRO3 (Molecular Probes,Inc.) to label nuclei and then washed twice with PBS. Infectedcells were visualized using a Zeiss inverted LSM510 confocalmicroscope (Carl Zeiss, New York, N.Y.).

Immunoblotting. Cells infected with reovirus were lysed in buffer consistingof 50 mM Tris-HCl (pH 7.6), 1% deoxycholic acid, 1% IGEPAL CA-630(MP Biomedicals, LLC, Aurora, Ohio), 0.1% sodium dodecyl sulfate(SDS), and 150 mM NaCl. After centrifugation, proteins in thesoluble fraction were size fractionated by SDS-polyacrylamidegel electrophoresis (PAGE) and electroblotted onto nitrocellulosemembranes. Viral proteins were detected using enhanced chemiluminescence(Amersham Biosciences) following incubation with ${sigma}$ 3-specificMAb 4F2 or antisera specific for ${sigma}$ NS, µNS, µ2, orT1L virions and appropriate secondary antibodies.

Electrophoretic analysis of cytoplasmic viral dsRNA. Cells were infected with reovirus at an MOI of 10 PFU per cell,harvested at 18 h postinfection, and lysed in buffer consistingof 100 mM Tris-HCl (pH 7.4), 0.5% IGEPAL CA-630, 1.5 mM MgCl₂,and 140 mM NaCl. After removal of nuclei by low-speed centrifugation,RNAs were collected by ethanol precipitation. RNA precipitateswere analyzed by 10% SDS-PAGE, followed by ethidium bromidestaining.

Reovirus mRNA transcription and translation. Linearized pcM3_1861m, pcM3_1ATGm, pcM3_2ATGm, and pcM3_dN138 wereprepared by digestion with XhoI. For generation of capped reovirusmRNA, linearized plasmids were transcribed in vitro using aMEGAscript high-yield transcription T7 kit and cap analog (AmbionInc., Austin, Texas). Reovirus mRNAs were translated in vitrousing rabbit reticulocyte lysates (Flexi Rabbit; Promega, Madison,Wis.) and Easy Tag Express [³⁵S]methionine protein-labelingmix (PerkinElmer, Boston, Mass.). Labeled translation productswere subjected to 10% SDS-PAGE, followed by autoradiography.

Complementation of reovirus replication in cells expressing reovirus siRNA. 293T cells (4 x 10⁴) stably expressing M3 siRNA were seededinto 12-well plates (Costar, Cambridge, Mass.) approximately24 h prior to transfection. Cells were transfected with recombinantplasmid DNA expressing either wild-type µNS or µNSmutants. After 4 h of incubation, transfected cells were infectedwith T3D at an MOI of 10 PFU per cell. Viral cultures were harvestedfollowing 24 h of incubation, and viral titers in cell lysateswere determined by plaque assay.

RESULTS

Top

Results

Viral protein expression in 293T cells transiently transfected with µ2, µNS, or ${sigma}$ NS siRNA expression vectors. To better understand the function of reovirus core protein µ2and nonstructural proteins µNS and ${sigma}$ NS in viral replication,we selectively reduced the level of each protein by using RNAi(29, 53). We first designed several 19-nt siRNA sequences correspondingto different regions of the T3D M1 (µ2), M3 (µNS),and S3 ( ${sigma}$ NS) genes (Table 1) and cloned them into the siRNA-expressingvector pSUPER.puro under the control of the H1-RNA promoter.These constructs were cotransfected with T3D µ2 (pcFM1T3),µNS (pCMVM3wt), or ${sigma}$ NS (pCMVS3wt) expression vectors into293T cells. Viral protein expression in transfected cells wasassayed by immunoblotting using anti-µ2, -µNS, or- ${sigma}$ NS antibodies at 24 h posttransfection (Fig. 1A to C). In comparisonto cells expressing green fluorescence protein (GFP) siRNA (Table1), cells expressing M1, M3, and S3 siRNAs exhibited diminishedsynthesis of µ2, µNS, and ${sigma}$ NS, respectively (Fig.1A to C). The effects of specific siRNAs were variable, withsome sequences exhibiting substantially more inhibition thanothers. Viral protein expression was not affected by mock transfectionor transfection of cells with the empty pSUPER.puro vector (datanot shown). These results demonstrate that siRNAs targeted todifferent regions of the T3D M1, M3, and S3 genes possess variouscapacities for suppression of reovirus protein synthesis andthat M1 895, M3 1861, and S3 630 siRNAs, in particular, werehighly effective at inhibiting the protein expression of theT3D µ2, µNS, and ${sigma}$ NS proteins, respectively.

View larger version (35K):
[in this window]
[in a new window]
FIG. 1. Viral protein expression in 293T cells transiently transfected with reovirus siRNA expression vectors. siRNA expression vectors encoding a 19-nt sequence from the genes encoding the T3D µ2 (A), µNS (B), or ${sigma}$ NS (C) proteins were cotransfected with mammalian expression vector pcFM1T3 (µ2), pCMVM3wt (µNS), or pCMVS3wt ( ${sigma}$ NS) into 293T cells. Viral protein expression in transfected cells was assessed by immunoblotting with anti-µ2, -µNS, or - ${sigma}$ NS antibodies at 24 h posttransfection. GFP siRNA-expressing vectors (GFP) were used as controls. Band intensities were quantitated using Scion Image software and are expressed relative to the GFP siRNA control at the bottom of each gel. The results shown are representative of one experiment out of at least two performed.

Kinetics of reovirus growth in cells stably expressing reovirus-specific siRNAs. To test whether virus production is inhibited in cells expressingreovirus-specific siRNAs, we established 293T cells stably expressingsiRNAs that inhibit M1, M3, or S3 gene expression. siRNA-expressingpSUPER.puro vectors encoding reovirus-specific siRNAs M1 895,M3 1861, and S3 630 were transfected into 293T cells. Transfectedcells were selected using puromycin to establish stable lines.To test the effect of reovirus siRNAs on viral growth, thesecells were infected with either T1L or T3D at an MOI of 2 PFUper cell. Virus production was assessed by plaque assay at 0,12, 24, 36, and 48 h postinfection (Fig. 2). Growth of T1L wasnot diminished at any time point by siRNAs directed to the T3DM1, M3, or S3 transcripts (Fig. 2A). In contrast, growth ofT3D was reduced 2,160-, 1,482-, or 97-fold in cells expressingM1 895, M3 1861, or S3 630 siRNAs, respectively, in comparisonto growth in untransfected 293T cells (Fig. 2B). Titers of bothT1L and T3D continued to rise through 48 h postinfection, consistentwith the secondary and tertiary rounds of viral replicationthat would be expected for subtotal infection of cells at theMOI used (2 PFU per cell). Complete lysis of cultures was observedabout 5 days after viral adsorption. To control for nonspecificeffects of siRNA on T3D replication, growth of T3D in cellsexpressing GFP siRNA was not affected (Fig. 2B). The T1L M1,M3, and S3 genes differ from the T3D homologues at the siRNAtarget sites (Table 2). Thus, siRNAs directed to reovirus replicationproteins cause marked reductions in viral growth. Furthermore,the effect appears to be specific, since growth of T1L was notaltered by the T3D-based siRNAs.

View larger version (22K):
[in this window]
[in a new window]
FIG. 2. Reovirus growth in cells stably expressing reovirus-specific siRNAs. 293T cells stably expressing M1 895, M3 1861, or S3 630 siRNAs were infected with either T1L (A) or T3D (B) at an MOI of 2 PFU per cell. Untransfected 293T cells (293T) and cell lines stably expressing GFP siRNA (GFP) were used as controls. Viral titers at the time points shown were determined by plaque assay. The results shown are representative of one experiment out of three performed.

View this table:
[in this window]
[in a new window]
TABLE 2. Comparison of siRNA sequences in T1L and T3D

Reovirus siRNAs are specific for T3D-derived targets. To confirm specific inhibition of T3D M1, M3, and S3 siRNAson virus production, we investigated growth of selected T1Lx T3D reassortant viruses G16, EB85, and EB113 in cells stablyexpressing reovirus siRNAs. G16 contains the T3D M1 and S2 genes,EB85 contains the T3D M3 and S2 genes, and EB113 contains theT3D M1 and S3 genes in an otherwise T1L background (42). Growthof G16 was inhibited up to 3,086-fold in cells expressing M1895 siRNA (Fig. 3A), growth of EB85 was inhibited 2,058-foldin cells expressing M3 1861 siRNA (Fig. 3B), and growth of EB113was inhibited 16-fold in cells expressing S3 630 siRNA (Fig.3C), each in comparison to growth in untransfected 293T cells.In contrast, titers of G16, EB85, or EB113 in cells expressingGFP siRNA were not diminished at any time point (Fig. 3A to C).These results indicate that the effects of siRNAs on reovirusinfection are strain and gene specific, and inhibition of virusgrowth caused by reovirus siRNAs does not result from nonspecificinduction of antiviral responses.

View larger version (16K):
[in this window]
[in a new window]
FIG. 3. Growth of reovirus reassortants following infection of cells stably expressing reovirus-specific siRNAs. 293T cells stably expressing M1 895, M3 1861, or S3 630 siRNAs were infected with reassortant virus G16 (A), EB85 (B), or EB113 (C), respectively, at an MOI of 2 PFU per cell. Untransfected 293T cells (293T) and cell lines stably expressing GFP siRNA (GFP) were used as controls. Virus titers at the time points shown were determined by plaque assay. The results shown are the means from two independent experiments. Error bars indicate standard deviations.

Viral protein expression in cells stably expressing reovirus-specific siRNAs. To investigate viral protein expression in 293T cells stablyexpressing reovirus-specific siRNAs, cells were infected witheither T1L or T3D at an MOI of 5 PFU per cell and incubatedfor 20 h. Lysates prepared from infected cells were analyzedby immunoblotting using polyclonal antisera specific for theµ2, µNS, and ${sigma}$ NS proteins, MAb 4F2 specific for theT3D ${sigma}$ 3 protein, and an antiserum raised against T1L virions todetect T1L ${sigma}$ 3 protein. Stable expression of siRNAs specific forT3D µ2 (M1 895), µNS (M3 1861), or ${sigma}$ NS (S3 630) proteinssubstantially diminished viral protein expression in cells infectedwith T3D, in contrast to that in untransfected 293T cells orcells expressing GFP siRNA (Fig. 4). Expression of viral proteinsin cells infected with T1L was not affected by siRNAs targetingreplication proteins of T3D (Fig. 4). These results indicatethat the effect of reovirus siRNAs on viral protein synthesisis strain specific and that µ2, µNS, and ${sigma}$ NS proteinsare each required for viral protein synthesis during reovirusreplication.

View larger version (59K):
[in this window]
[in a new window]
FIG. 4. Viral protein expression in cells stably expressing reovirus-specific siRNAs. 293T cells stably expressing S3 630, M1 895, or M3 1861 siRNAs were infected with either T1L or T3D at an MOI of 5 PFU per cell and incubated for 20 h. Lysates prepared from infected cells were analyzed by immunoblotting using polyclonal antisera specific for the µ2, µNS, and ${sigma}$ NS proteins. Polyclonal antiserum raised against T1L virions and MAb 4F2 were used to detect the T1L and T3D ${sigma}$ 3 proteins, respectively. An actin-specific antibody was used as a loading control. The results shown are representative of one experiment out of three performed.

Viral dsRNA synthesis in cells stably expressing reovirus-specific siRNAs. To define the importance of µ2, µNS, and ${sigma}$ NS proteinsin viral dsRNA synthesis, we assessed viral dsRNA productionin cells expressing reovirus siRNAs. Cells were infected witheither T1L or T3D at an MOI of 10 PFU per cell, harvested at18 h postinfection, and lysed. After removal of nuclei by low-speedcentrifugation, RNA precipitates were analyzed by SDS-PAGE andethidium bromide staining (Fig. 5). Production of dsRNA by T1Lin cells expressing T3D-specific M1, M3, or S3 siRNAs was notdecreased in comparison to that in untransfected 293T cellsor a cell line stably expressing GFP siRNA. In contrast, productionof dsRNA by T3D was significantly decreased in cell lines expressingT3D-specific M1, M3, or S3 siRNAs (Fig. 5). These results indicatethat µ2, µNS, and ${sigma}$ NS proteins are each requiredfor viral dsRNA synthesis.

View larger version (96K):
[in this window]
[in a new window]
FIG. 5. Viral dsRNA expression in cells stably expressing reovirus-specific siRNAs. Untransfected 293T (293T) cells and cell lines stably expressing S3 630, M1 895, M3 1861, or GFP siRNA (GFP) were infected with either T1L or T3D at an MOI of 10 PFU per cell and incubated for 18 h. The cytoplasmic fraction containing viral dsRNA from infected cells was resolved by electrophoresis in 10% polyacrylamide gels and analyzed by ethidium bromide staining. Size classes of viral dsRNA segments (large [L], medium [M], and small [S]) are indicated. Viral dsRNA extracted from purified T1L and T3D virions was used as a control. The results shown are representative of one experiment out of two performed.

Subcellular localization and expression of reovirus proteins in cells expressing reovirus-specific siRNAs. The results described thus far indicate that the µ2, µNS,and ${sigma}$ NS proteins are required for viral protein and dsRNA synthesis.These proteins are also components of viral inclusions in reovirus-infectedcells. Thus, we investigated the effect of siRNAs on viral inclusionformation in cells infected with either T1L or T3D. 293T cellsinfected with reovirus at an MOI of 10 PFU per cell were seededonto collagen-coated glass coverslips, incubated for 20 h, andfixed and stained with primary antibodies specific for reovirusproteins. In previous studies, it was observed that T1L producesfilamentous inclusions, whereas T3D forms globular inclusionsin several cell lines (45). The M1 gene is the genetic determinantof this difference in inclusion morphology (36, 45, 65). Inour experiments, µNS and ${sigma}$ NS proteins were detected inglobular inclusions in untransfected and GFP siRNA-expressing293T cells infected with T3D. Both proteins were detected infilamentous inclusions in cells infected with T1L (Fig. 6).In T3D-infected cells expressing M3 1861 siRNA, µNS proteinwas not detected, and ${sigma}$ NS protein was diffusely distributed inthe cytoplasm. µNS was detected in small globular inclusionsin T3D-infected cells not expressing ${sigma}$ NS protein (S3 630 siRNAcells) (Fig. 6). Both µNS and ${sigma}$ NS proteins were evidentin filamentous inclusions in T1L-infected cells stably expressingM3 1861 or S3 630 siRNAs (Fig. 6).

View larger version (66K):
[in this window]
[in a new window]
FIG. 6. Detection of protein expression in cells stably expressing M1, M3, or S3 siRNAs by immunofluorescence microscopy. 293T cells and cell lines stably expressing GFP, M1 895, M3 1861, or S3 630 siRNAs were infected with either T1L or T3D at an MOI of 10 PFU per cell and incubated for 20 h. Infected cells were fixed and stained using anti-µNS (rabbit) or anti- ${sigma}$ NS (guinea pig) antiserum, followed by Alexa Fluor 488 goat anti-rabbit IgG (green) or Alexa Fluor 546 goat anti-guinea pig IgG (red), respectively. Cells were stained with TO-PRO3 (blue) to label nuclei.

Following infection of cells expressing M1 895 siRNA with T3D,µNS and ${sigma}$ NS were contained in small globular inclusions(Fig. 6). The µ2 protein was localized to cytoplasmicviral inclusions and nuclei of untransfected 293T cells infectedwith T3D (Fig. 7); µ2 was not detected in cells expressingM1 895 siRNA following infection with T3D (Fig. 7). The µ2,µNS, and ${sigma}$ NS proteins were detected in filamentous cytoplasmicinclusions and nuclei of M1 895 siRNA-expressing cells infectedwith T1L (Fig. 6 and 7). These results indicate that the µNSprotein plays an essential role in the formation of viral inclusionsand recruitment of other viral proteins, including ${sigma}$ NS, in reovirus-infectedcells. Furthermore, the effects of reducing levels of µ2and ${sigma}$ NS proteins in infected cells reveal that both proteinsserve essential functions in the formation and maturation ofviral inclusions.

View larger version (37K):
[in this window]
[in a new window]
FIG. 7. Detection of protein expression in cells stably expressing M1 siRNA by immunofluorescence microscopy. 293T cells and cell lines stably expressing M1 895 siRNA were infected with either T1L or T3D at an MOI of 10 PFU per cell and incubated for 20 h. Infected cells were fixed and stained using anti-µ2 (rabbit) or anti- ${sigma}$ NS (guinea pig) antiserum, followed by Alexa Fluor 488 goat anti-rabbit IgG (green) or Alexa Fluor 546 goat anti-guinea pig IgG (red). Cells were stained with TO-PRO3 (blue) to label nuclei.

Complementation of reovirus replication defects in cells stably expressing T3D M3 siRNA. To develop a complementation system for studies of reovirusreplication, we first constructed a T3D µNS expressionplasmid containing three silent point mutations within the siRNAtarget sequence (pcM3_1861m). Cells stably expressing M3 1861siRNA were transfected with pcM3_1861m followed by infectionwith T3D at an MOI of 10 PFU per cell at 4 h posttransfection.Viral titers were determined by plaque assay at 24 h postinfection.Yields of T3D were decreased approximately 3,000-fold in siRNA-expressingcells transfected with mock plasmid compared to an approximately75-fold reduction in cells transfected with pcM3_1861m plasmid(Fig. 8A). Thus, provision of µNS in trans rescues a substantiallevel ( $~$ 40-fold) of reovirus growth in cells expressing M3 siRNA.

View larger version (29K):
[in this window]
[in a new window]
FIG. 8. Complementation of reovirus replication in cells expressing M3 siRNA. (A) 293T cells or cells stably expressing M3 1861 siRNA were transfected with pcDNA3, pcM3_1861m, pcM3_1ATGm, pcM3_2ATGm, or pcM3_dN138 and infected with T3D at an MOI of 10 PFU per cell. At 24 h postinfection, viral titers were determined by plaque assay. The results are means from three independent experiments. Error bars indicate standard deviations. (B) RNA transcripts of pcM3_1861m, pcM3_1ATGm, pcM3_2ATGm, and pcM3_dN138 were translated in rabbit reticulocyte lysates containing [³⁵S]methionine. Labeled products were resolved by 10% SDS-PAGE, followed by autoradiography.

To determine whether expression of µNSC, which lacks the40 amino-terminal residues of µNS, can complement theviral growth defect in cells expressing M3 siRNA, pcM3_dN138,which encodes µNSC but not µNS (Fig. 8B), and pcM3_1ATGm,which encodes µNS with a mutated start codon (¹⁹ATG²¹ $->$ ¹⁹GCG²¹) such that only µNSC is expressed (Fig. 8B),were introduced into M3 1861 cells prior to infection with T3D.In these experiments, neither construct was capable of restoringreovirus replication (Fig. 8A). In contrast, transfection ofpcM3_2ATGm, which expresses µNS but is incapable of expressingµNSC due to a mutated µNSC start codon (¹³⁹ATG¹⁴¹ $->$ ¹³⁹GCG¹⁴¹), mediated the rescue of viral replication in cellsstably expressing M3 siRNA. Complementation efficiency obtainedusing pcM3_2ATGm was equivalent to that effected with pcM3_1861m(Fig. 8A), which expresses both µNS and µNSC (Fig.8B). These results demonstrate that the 40 amino-terminal residuesof µNS are indispensable for its native function in reovirus-infectedcells and that µNSC is not required for reovirus replication.

DISCUSSION

Top

Discussion

Reovirus mutant strains obtained by selection, screening, orpassage have served as powerful tools to facilitate an understandingof a broad range of viral processes (15). However, there aresome reovirus genes and proteins for which informative mutantviruses have not been isolated. Notably, a reovirus mutant bearinga lesion in the M3 gene has not been described. Although a reversegenetics system that allows incorporation of nonviral sequencesinto infectious reovirus particles has been established, currenttechniques are limited to modifications of the S2 gene (encodingcore protein ${sigma}$ 2) (48-50). Rescue of engineered changes in nativereovirus RNAs has not been reported. Therefore, stable celllines expressing reovirus-specific siRNAs, such as those establishedin this study, offer an opportunity to study functions of viralproteins for which mutant strains are unavailable.

Reovirus replication and assembly are thought to occur withincytoplasmic viral inclusions where viral and cellular proteins(21, 54), viral RNAs (57), and immature and mature viral particles(21) are concentrated. The µ2, µNS, and ${sigma}$ NS proteinsplay key roles in forming viral inclusions (4, 5, 7, 9, 36,39, 45, 65); however, little is known about their individualand corporate functions in viral RNA replication and particleassembly. In this study, we investigated the functions of µ2,µNS, and ${sigma}$ NS proteins in viral replication by using RNAi,a process by which dsRNA directs sequence-specific degradationof mRNA. To overcome the barrier of low transfection efficiencyproblematic to the use of chemically synthesized siRNAs, weestablished cell lines stably expressing M1-, M3-, and S3-RNA-specificsiRNAs, which were highly effective in inhibiting the expressionof the T3D µ2, µNS, and ${sigma}$ NS proteins, respectively(Fig. 1). Viral yields of T3D and selected T1L x T3D reassortantviruses containing T3D genome segments with siRNA-targeted sequenceswere significantly decreased in siRNA-expressing cell lines(Fig. 2 and 3). In contrast, viral yields of T1L, which differsfrom T3D in the siRNA-targeted sequences (Table 2), were notaltered by T3D-based siRNAs (Fig. 2). Thus, inhibition of virusproduction by reovirus siRNAs is allele specific and, therefore,unlikely to result from induction of innate antiviral responsesin cell lines constitutively producing siRNAs.

Stable expression of siRNA specific for the µ2 proteinsubstantially diminished viral protein and dsRNA synthesis incells infected with T3D (Fig. 4 and 5), demonstrating a tightfunctional association of µ2 protein with the viral replicationmachinery. In a previous study, reovirus ts mutant tsH11.2,which contains a defect mapped to the M1 genome segment, producedneither detectable viral proteins nor dsRNA late in infectionat the nonpermissive temperature (13). These findings are consistentwith our results for stable cell lines expressing M1 siRNA.The µ2 protein is genetically associated with viral-strain-specificdifferences in the in vitro transcriptase and NTPase activitiesof viral particles (44, 64), and purified µ2 possessesin vitro NTPase activity stimulated in the presence of ${lambda}$ 3 (31).Thus, inhibition of reovirus replication by stable expressionof M1 siRNA may reflect suppression of transcriptase or NTPaseactivities of the viral core. It is also possible that the effectsof M1 siRNA are attributable to the inhibition of another µ2function. For example, µ2 determines strain-specific differencesin rate of viral inclusion formation in reovirus-infected cells(36). In cells infected with T3D, µNS and ${sigma}$ NS are detectedin very small inclusion-like structures in the absence of µ2(Fig. 7). These results suggest a critical function for µ2in the maturation of viral inclusions but not in their genesis.

Stable expression of S3 siRNA substantially diminished viraldsRNA and protein synthesis in cells infected with T3D (Fig.4 and 5). These results suggest that functional ${sigma}$ NS is essentialfor reovirus replication, consistent with a previous reportthat reovirus mutant tsE320, in which the ts phenotype mapsto the S3 genome segment, produced reduced levels of viral proteinsand dsRNA (16, 20) and exhibited diminished inclusion formationat the nonpermissive temperature (4). Expression of ${sigma}$ NS was notdetectable by immunofluorescence in T3D-infected cells stablyexpressing S3 siRNA (Fig. 6), although µNS was detectedin small inclusion-like structures (Fig. 6). These findingssuggest that ${sigma}$ NS functions in the maturation of functional viralinclusions but does not initiate inclusion formation in theabsence of other viral proteins. The ${sigma}$ NS protein associates withssRNA in a non-sequence-specific manner and forms higher-ordercomplexes stabilized by RNAs (24-26, 30, 47). Thus, possiblefunctions for ${sigma}$ NS consistent with the results of this and otherstudies include recruitment of viral ssRNA to inclusions; sequestrationof viral RNA in inclusions through a tripartite µNS- ${sigma}$ NS-RNAcomplex; organization of viral RNA for replication, assortment,and packaging; scaffolding of viral inclusions; and regulationof viral translation. These or other possible ${sigma}$ NS functions maybe mediated by interactions between ${sigma}$ NS and cellular proteinsyet to be identified.

Stable expression of siRNAs specific for the µNS proteindiminished viral protein and dsRNA synthesis in cells infectedwith T3D (Fig. 4 and 5). This is the first report that µNSplays an essential role in reovirus replication in infectedcells. In previous studies, µNS protein was shown to bindcore particles (8) and, when expressed from plasmid vectors,formed structures with morphologies similar to those of viralinclusions in infected cells and specifically recruited viralproteins ${lambda}$ 1, ${lambda}$ 2, µ2, ${sigma}$ NS, and ${sigma}$ 2 into inclusion-like structures(5, 7, 39). In T3D-infected cells stably expressing M3 siRNA,we found that ${sigma}$ NS protein was diffusely distributed in the cytoplasmand µNS was not detectable (Fig. 6). In contrast, ${sigma}$ NS andµNS were colocalized in small viral inclusions in infectedcells stably expressing M1 siRNA (Fig. 7). Therefore, our resultsand previous findings suggest that µNS plays a centralrole in viral inclusion formation and recruits other viral proteinsand viral RNA into the inclusions where replication and particleassembly occur.

We developed a complementation system for functional studiesof µNS in cells that stably express M3-specific siRNAs.The complementing construct pcM3_1861m contains three nucleotidesubstitutions in the siRNA-targeted T3D µNS-encoding sequence,resulting in resistance to siRNA-mediated degradation. The efficiencywith which this construct restored viral replication in infectedcells was less than anticipated, even when correcting for potentiallylow transfection frequencies. It is possible that the low complementationefficiency is related to the process of inclusion formationor maturation during reovirus infection. In our complementationsystem, exogenous µNS was expressed in M3 siRNA-expressingcells prior to viral infection and, therefore, viral inclusion-likestructures were likely nucleated by µNS prior to the expressionof other inclusion-associated proteins, such as µ2 and ${sigma}$ NS. Perhaps concurrent production of µNS and other replicationproteins, such as µ2 and ${sigma}$ NS, is required for maturationof fully functional inclusions. In support of this idea, coexpressionof µ2 or ${sigma}$ NS with exogenous µNS enhances viral yieldsin cells expressing M3 siRNA (T. Kobayashi and T. S. Dermody,unpublished data).

Our complementation strategy afforded the opportunity to definethe importance of µNS isoform µNSC in viral replication.We found that transient expression of µNSC was incapableof restoring viral replication in reovirus-infected cells madedeficient in µNS expression by RNAi (Fig. 8). Conversely,expression of µNSC was not requisite to the complementationof viral replication by µNS (Fig. 8). µNSC lacksthe 40-residue amino-terminal segment of µNS, which containsinteracting domains for µ2 and ${sigma}$ NS (9, 39). Therefore,these results indicate that the association of µNS withµ2 or ${sigma}$ NS (or both proteins) is required for viral replication.Although µNSC did not restore viral replication in thecomplementation system, this protein may nevertheless exercisefunctions beneficial to viral growth, consistent with preservationof the µNSC open reading frame. Additionally, it is possiblethat µNSC contributes to viral growth or spread in vivo.

Our study demonstrates the remarkable utility of stable celllines expressing reovirus-specific siRNAs as tools to investigatethe replication machinery of reovirus in infected cells. Targetingof other reovirus genes should lead to additional insights aboutthe contribution of individual viral proteins to viral replicationand assembly. The RNAi approach, combined with individual viralreplication-complementation assays, opens new opportunitiesto understand biological activities of reovirus replicationproteins.

ACKNOWLEDGMENTS

We thank Earl Brown for providing µ2-specific antiserum,members of our laboratory for many useful discussions, and DeniseWetzel, Kristen Guglielmi, and Tim Peters for review of themanuscript.

This work was supported by a fellowship from the Naito Foundation(T.K.), Public Health Service awards K08 AI062862 (J.D.C.) andR01 AI32539 from the National Institute of Allergy and InfectiousDiseases, and the Elizabeth B. Lamb Center for Pediatric Research.Additional support was provided by Public Health Service awardsCA68485 for the Vanderbilt-Ingram Cancer Center and DK20593for the Vanderbilt Diabetes Research and Training Center.

FOOTNOTES

* Corresponding author. Mailing address: Lamb Center for Pediatric Research, D7235 MCN, Vanderbilt University School of Medicine, Nashville, TN 37232. Phone: (615) 343-9943. Fax: (615) 343-9723. E-mail: terry.dermody{at}vanderbilt.edu.

REFERENCES

Top