INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The infectious virion of mammalian orthoreoviruses (reoviruses), prototype members of the Reoviridae family, has a genome composed of 10 double-stranded RNA (dsRNA) segments surrounded by two concentric protein capsids. The virion can be proteolytically cleaved in vitro to generate the intermediate subvirion particle (ISVP), which lacks outer-capsid protein 3 and contains fragments of outer-capsid protein µ1. Further proteolysis removes the µ1 fragments and releases outer-capsid protein 1, yielding the core particle (reviewed in reference 34). When provided with substrates, the core is transcriptionally active in vitro, using the dsRNA genome segments as templates for synthesis of the 10 full-length plus strand transcripts (3, 24, 39). In addition, these transcripts are modified to have a cap 1 structure (^m7NGpppG^m2'O) at their 5' ends by viral enzymes within the core (15, 37). The resulting capped transcripts, which are released from the core as they are synthesized (4), are competent for translation into the reovirus proteins (38). Three of these proteins, µNS, NS, and 1s, are synthesized in infected cells but are not found in purified virions (36, 49). The functions of these "nonstructural" proteins are not well understood, but all members of the Reoviridae family encode such proteins suggesting that they play important roles during infections by these viruses (13).

µNS (called µ0 in older papers), an 80,000-M_r (80K) nonstructural protein, is encoded by the reovirus M3 genome segment (27, 32). M3 encodes another protein, µNSC, whose M_r is approximately 5,000 smaller than that of µNS and that is recognized by µNS-specific monoclonal antibodies (23). µNSC is thought to be generated by translation initiation at a downstream start codon in the open reading frame that encodes µNS, such that µNSC lacks approximately 5,000 Da of sequences that are present at the amino (N) terminus of µNS (28, 44). Both µNS and µNSC are present in cells infected with the prototype isolates of all three reovirus serotypes (23), are expressed to moderate levels throughout infection (17), and are present in the infected cell at a µNS:µNSC ratio of 1:1 to 4:1 (44). Whether the two proteins have functional differences has not yet been addressed. Because the activities of µNSC have not been differentiated from those of µNS, the two proteins are generally referred to as µNS in this report.

Although the roles of µNS in the reovirus life cycle remain poorly understood, previous observations suggest several possibilities. In one study, antibodies to µNS coimmunoprecipitated the viral plus strand RNA transcripts soon after the transcripts were synthesized in infected cells (1). Through this RNA-protein interaction, µNS may be involved in translation of the viral transcripts, packaging of the RNA segments into new reovirus particles, synthesis of the minus strand RNA, or recognition and sorting of the 10 distinct RNA segments prior to packaging (1). In other studies, µNS was isolated from infected cells in association with different types of viral particles (30, 31, 50). These particles were believed to be assembly intermediates because they chased into virions later in infection (50). In one of these studies, newly assembled "transcriptase particles," capable of synthesizing the viral plus strand transcripts in vitro, were isolated from cells and shown to comprise core-like particles plus large amounts of µNS (30). The precise function of µNS in these particles was unclear because cores are transcriptionally active in vitro in the absence of µNS (3, 24, 39) and because transcriptionally active particles isolated from cells in other studies contained less or no µNS (31, 40). Nevertheless, the findings of Morgan and Zweerink (30) suggest that µNS may play a role in the regulation of reovirus transcription or particle assembly. Other observations concerning µNS include its association with the cytoskeletal fraction from infected cells (29), possession of predicted -helical coiled-coil motifs in the carboxyl (C)-terminal third of the µNS sequence (28), and coimmunoprecipitation with NS using NS-specific antibodies (23). M3 was additionally identified as a genetic determinant of genome segment deletion during passage of reovirus at high multiplicity in culture (6). These results suggest that µNS may be involved at several steps in the viral life cycle, but additional experiments should provide a better description of its roles.

To investigate the association of µNS with viral particles and its possible involvement in transcription or assembly, we obtained µNS protein from expression in insect cells using a recombinant baculovirus and studied its interactions with different types of reovirus particles. After demonstrating µNS binding to cores, but not to virions or ISVPs, we determined various characteristics of the µNS-bound cores (µNS cores), including their continued transcription and capping activities. We also showed that µNS can compete with outer-capsid proteins µ1 and 3 for binding to cores. The results of these studies suggest several possible roles for µNS cores in the reovirus life cycle, including a role in enhancing production of the viral plus strand transcripts by blocking or delaying outer-capsid assembly on these particles.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. All enzymes were from New England Biolabs, Inc. (Beverly, Mass.), unless otherwise stated. All chemicals were from Sigma (St. Louis, Mo.) unless otherwise stated.

Construction of M3 recombinant baculoviruses. The type 3 Dearing (T3D) reovirus M3 genome segment was cloned by Cashdollar et al. (7) and subcloned into pUC8 by Wiener et al. (44). The clone was a generous gift from M. R. Roner (Florida Atlantic University). For our work, we cut the T3D M3 gene from pUC8 using the PshAI and AflII sites in the terminal nontranslated regions of T3D M3. The 5' overhang of the AflII site was filled in using the Klenow fragment of DNA polymerase I (Pharmacia, Piscataway, N.J.). The T3D M3 gene fragment was then blunt end ligated to SmaI-cut pGEM4Z (Promega, Madison, Wis.) to generate pGEM4Z-M3(T3D). The T3D M3 gene was excised from pGEM4Z-M3(T3D) at the BamHI and KpnI sites and ligated to pFastBacI (Gibco-BRL, Gaithersburg, Md.) that had been cut with the same enzymes. The pFastBacI vector containing T3D M3 was transformed into DH10Bac cells (Gibco-BRL) following the manufacturer's instructions to produce a recombinant bacmid. The isolated bacmid was transfected into Spodoptera frugiperda (Sf21) cells (Invitrogen, Carlsbad, Calif.) to yield a progeny recombinant baculovirus containing the reovirus T3D M3 gene [M3(T3D)-bac]. The stock was then subjected to two serial passages in Sf21 cells to increase viral titer.

Expression of recombinant µNS. To express µNS protein, 7.5 × 10⁶ Trichoplusia ni cells (High Five, Invitrogen) were plated on a 100-mm-diameter dish, infected with M3(T3D)-bac at a multiplicity of infection (MOI) of 1, and incubated at 27°C. The cells were harvested at 52 h postinfection by being pelleted at 500 × g for 10 min at 4°C, washed with phosphate-buffered saline (137 mM NaCl, 3 mM KCl, 8 mM Na₂HPO₄, 1 mM KH₂PO₄), and repelleted. The cell pellet was resuspended in 500 µl of lysis buffer (10 mM Tris [pH 7.5], 2.5 mM MgCl₂, 100 mM NaCl₂, 0.5% Triton X-100, 5 µg of leupeptin/ml, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol) and placed on ice for 30 min with mixing every 5 min. The insoluble fraction was removed by spinning the mixture at 500 × g for 10 min at 4°C. The amount of µNS expressed in the lysate varied from 0.3 to 0.8 µg of µNS per µl of lysate as estimated from Coomassie blue-stained gels with bovine serum albumin standards. The µNS protein used in the experiments described in this paper was T3D µNS unless otherwise specified.

Production of µNS polyclonal antibodies. To direct expression of µNS with an N-terminal histidine tag, the T3D M3 gene was removed from pGEM4Z-M3(T3D) at the BamHI and KpnI sites and ligated to pRSETB (Invitrogen) that had been cut with the same enzymes. The plasmid was transformed into BL21-DE3 cells (Novagen, Madison, Wis.), and the histidine-tagged µNS was expressed and purified following the protocol in the pET system manual (Novagen). In brief, expression was induced with isopropyl--D-thiogalactopyranoside (0.19 mg/ml), and the cells were grown at 37°C for 3 h. Intact cells were pelleted, resuspended, and lysed by sonication. The insoluble fraction containing µNS was spun down and solubilized in 8 M urea. The histidine-tagged µNS was purified with His-bind resin (Novagen) in column format. The eluent was dialyzed into phosphate-buffered saline and concentrated with polyethylene glycol. The antiserum was generated in a rabbit by the polyclonal antibody service in the animal care unit of the University of Wisconsin Medical School (Madison, Wis.).

Growth of reovirus and purification. Infections and purification of reovirus T1L and T3D virions were performed as described previously (14). ISVPs were prepared by digestion of particles with chymotrypsin as described previously (33). Reovirus cores were prepared by digestion as described for reoviruses T3D (25) and T1L (12). Cores were alternatively prepared using an expedited protocol as described previously (9). To obtain particles labeled with [³⁵S]methionine and [³⁵S]cysteine, 5 mCi of Easy Tag Express protein labeling mixture (Dupont, Wilmington, Del.) was added per 4 × 10⁸ cells in spinner culture. All particles were purified using equilibrium centrifugation in CsCl density gradients, followed by dialysis into virion buffer (VB) (10 mM Tris [pH 7.5], 10 mM MgCl₂, 150 mM NaCl) and storage at 4°C. The concentrations of virions and cores were determined using 1.0 A₂₆₀ = 2.1 × 10¹² virions/ml (42) and 1.0 A₂₆₀ = 4.2 × 10¹² cores/ml (10).

SDS-PAGE and immunoblot analysis. Samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in 10% polyacrylamide gels as described previously (18). Proteins were visualized by staining with Coomassie brilliant blue R-250 (Coomassie blue). Gels with radiolabeled proteins were dried onto filter paper and visualized by phosphorimaging (Molecular Dynamics, Sunnyvale, Calif.). To estimate relative amounts of reovirus proteins from Coomassie blue-stained gels, gels were scanned with a laser densitometer (Molecular Dynamics) and volume-based intensities of the protein bands were determined using the ImageQuant program (Molecular Dynamics). For immunoblots, protein samples were subjected to SDS-PAGE and transferred to nitrocellulose (Bio-Rad, Hercules, Calif.) at 4°C for 1 h at 100 V in transfer buffer (25 mM Tris, 192 mM glycine, pH 8.3). Binding of the primary antibody was detected with alkaline phosphatase-coupled goat anti-rabbit or goat anti-mouse immunoglobulin (Bio-Rad) and colorimetric reagents p-nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolylphosphate p-toluidine salt (Bio-Rad).

Incubation of µNS lysate with reovirus particles and gradient fractionation. The buffer conditions for binding were 10 mM Tris, 100 to 120 mM NaCl, 3 to 6 mM MgCl₂, and 0.3 to 0.5% Triton X-100. Fifty microliters of insect cell lysate containing µNS was mixed with 2 × 10¹² particles (cores, ISVPs, or virions) in VB. For the lysate-alone gradients, 50 µl of insect cell lysate containing µNS was mixed with VB. For the particle-alone gradients, 2 × 10¹² particles in VB were mixed with lysis buffer. Samples were incubated for 1 h at 37°C, except for ISVPs, which were incubated at room temperature due to instability at 37°C. The samples were layered onto preformed 3.5-ml CsCl density gradients (1.26 to 1.47 g/cm³ for cores and 1.2 to 1.45 g/cm³ for ISVPs and virions) in an SW60 tube and spun in a Beckman ultracentrifuge for 2 h at 50,000 rpm at 5°C. The gradients were fractionated with a peristaltic pump into 200-µl fractions. The optical density at 280 nm (OD₂₈₀) and refractive index of each fraction were determined. The density of each fraction was determined based on its refractive index.

Identification of proteins bound to reovirus particles. Cores, virions, or ISVPs (10¹²) were mixed with 140 µl of insect cell lysate containing µNS as described above except that all samples were incubated at room temperature. After centrifugation, the particle bands were visualized by light scattering and were harvested by puncturing the bottom of the tube, except for µNS cores, which were harvested with a Pasteur pipette from the top of the gradient. Samples were dialyzed into VB prior to analysis by SDS-PAGE and immunoblotting.

NS-expressing recombinant baculovirus and protein expression. The recombinant baculovirus expressing the T1L NS protein was previously described (18). T. ni cells were infected with recombinant baculovirus as described previously (18). Cell lysate was prepared as described above for µNS.

RNase A treatment of µNS cores and µNS lysate. µNS cores (approximately 2 × 10¹²) purified on a CsCl gradient and dialyzed into VB were treated with 20 µg of RNase A for 2 h at 37°C and then purified on another CsCl gradient. Untreated µNS cores were analyzed in parallel as a control. µNS lysate (50 µl) was treated with 4 µg of RNase A for 30 min at 25°C and then incubated with 2 × 10¹² cores for 1 h at 25°C. Untreated µNS lysate was also analyzed in parallel as a control. The samples were layered on CsCl gradients and spun to equilibrium. Particles were removed from the gradients with a Pasteur pipette and analyzed by SDS-PAGE.

Gradients with different amounts of µNS. Cores (10¹²) in 20 µl of VB were mixed with twofold-increasing amounts of insect cell lysate containing µNS (4.3, 8.5, 17, 34, 68, and 140 µl) and brought to a final volume of 170 µl with lysis buffer. Cores (10¹²) in 20 µl of VB were mixed with 150 µl of lysis buffer or 140 µl of lysate from insect cells infected with wild-type baculovirus and 10 µl of lysis buffer. For the lysate-alone sample, 140 µl of insect cell lysate containing µNS was mixed with 30 µl of lysis buffer. The samples were incubated at room temperature for 1 h and then layered onto preformed 3.5-ml CsCl gradients (1.31 to 1.50 g/cm³) in an SW60 tube. The samples were spun at 50,000 rpm for 2 h at 5°C and visualized with a high-intensity lamp.

Negative-stain electron microscopy (EM). Samples were negatively stained with 1% uranyl acetate and viewed with a Philips 120 transmission electron microscope at 100 kV as described previously (8).

Sucrose gradient velocity sedimentation of lysate containing µNS. Insect cell lysate containing µNS (150 µl) was layered on a preformed 11-ml sucrose gradient (5 to 20%) in an SW41 tube and spun in a Beckman ultracentrifuge at 40,000 rpm for 15 h at 5°C. An additional gradient contained 2 mg of each marker, gamma globulin (7S) and thyroglobulin (19S). The gradients were fractionated with a piston gradient fractionator (BioComp Instruments, Inc., New Brunswick, Canada) into 26 equal fractions. The positions of the markers were determined by OD₂₆₀. Samples from the gradient of µNS lysate were analyzed by immunoblotting.

Transcription and methylation assay. T3D cores were mixed with various amounts of µNS lysate, incubated for 1 h at room temperature, purified on CsCl gradients, and dialyzed into VB. Cores were also incubated with lysate from wild-type baculovirus-infected cells (the same amount of lysate used to make µNS cores with 980 molecules of µNS per core) and purified on CsCl gradients. µNS cores for transcription reactions were made with [³⁵S]methionine- and [³⁵S]cysteine-labeled T3D cores. The amount of µNS per core was determined from densitometry of Coomassie blue-stained gels. Transcription reactions and transcription/methylation reactions were performed as described previously (26). The aggregated nature of µNS cores led to variability of particle input, so transcription activity was expressed as a ratio of the ³²P incorporated into trichloroacetic acid (TCA)-precipitable counts to ³⁵S in the TCA-precipitable counts to standardize the amount of RNA synthesized to the number of ³⁵S-labeled cores in each reaction mixture. Methylation activity was expressed as a ratio of ³H incorporated into TCA-precipitable counts to ³²P incorporated into TCA-precipitable counts to standardize for differences in transcription activity.

µNS and µ1-3 competition for core binding. The baculovirus to express T1L µ1 and 3 proteins was previously described (9). µ1- and 3-containing lysate (µ1-3 lysate) was prepared as described for µNS. T3D cores (5 × 10¹⁰) were mixed with 25 µl of lysis buffer, with 20 µl of µ1-3 lysate and 5 µl of lysis buffer, or with 5 µl of µNS lysate and 20 µl of lysis buffer as controls for the positions of the resulting particles, cores, recoated cores and µNS cores, respectively. T3D cores (5 × 10¹⁰) were mixed with 5 µl of µNS lysate and incubated for 1 h at 37°C, and then 20 µl of µ1-3 lysate was added, followed by incubation for 2 h at 37°C. T3D cores (5 × 10¹⁰) were mixed with 20 µl of µ1-3 lysate and incubated at 37°C for 2 h, and then 5 µl of µNS lysate was added, followed by incubation for 1 h at 37°C. The amount of µ1-3 lysate added was sufficient to recoat cores completely as demonstrated previously (9). The amount of µNS lysate was chosen to add approximately the same amount of µ1 and µNS to each sample as estimated from Coomassie blue-stained gels of lysate. After incubation at 37°C, the samples were layered on performed 5.4-ml CsCl density gradients (1.30 to 1.40 g/cm³) in an SW50 tube and spun in a Beckman ultracentrifuge at 40,000 rpm for 2 h 45 min at 5°C. Fractions were TCA precipitated by adding 30 µg of bovine serum albumin as a carrier and 900 µl of cold TCA. Samples were incubated on ice for 20 min and spun for 15 min at 16,000 × g in a microcentrifuge. The supernatant was removed, and the pellet was washed with 500 µl of 70% ethanol, allowed to air dry, resuspended in 20 µl of 1× Laemmli sample buffer (125 mM Tris [pH 8.0], 2% -mercaptoethanol, 1% SDS, 0.1% bromophenol blue), and incubated at 100°C for 5 min. Samples were analyzed by immunoblotting. µNS, µ1, and the core  proteins were detected using polyclonal antiserum specific to µNS, polyclonal antiserum specific to reovirus core proteins (S. Noble and M. L. Nibert, unpublished data), and a monoclonal antibody specific to µ1, 10H2 (43).

Computer software. Images for the figures were scaled uniformly and adjusted for optimal brightness and contrast in Photoshop 4.0 (Adobe System, San Jose, Calif.). All figures were produced in Illustrator 7.0 (Adobe).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Reovirus µNS protein is expressed to high levels in insect cells. To obtain large amounts of µNS for study, we generated recombinant baculovirus M3(T3D)-bac containing the entire coding region of the T3D reovirus M3 genome segment under transcriptional control of the baculovirus polyhedrin promoter. Insect cells infected with this virus produced a prominent protein doublet with a size of approximately 80K in SDS-polyacrylamide gels, similar to those of the µNS and µNSC proteins from reovirus-infected L cells (Fig. 1A). The doublet is found mostly in the cytoplasmic fraction of lysed insect cells (data not shown). Both the 80K doublet from M3(T3D)-bac-infected cells and µNS and µNSC from reovirus-infected cells were recognized in immunoblots by a polyclonal antiserum that we raised against µNS expressed in Escherichia coli (Fig. 1B). These proteins were also recognized in immunoblots by a monoclonal antibody to µNS (23) (data not shown). The ratio of the two proteins in the doublet varied with each preparation (data not shown). Experiments are in progress to determine whether the lower protein in this doublet has the same origin as µNSC from reovirus-infected cells (44) or may instead represent a breakdown product of µNS. In the subsequent text, we refer to both proteins in the 80K doublet from M3(T3D)-bac-infected cells as µNS. A recombinant baculovirus [M3(T1L)-bac] containing the T1L M3 gene was also constructed and directed expression of an 80K doublet, recognized by the polyclonal µNS antiserum, to levels as high as those for the T3D protein (data not shown).

View larger version (61K): [in this window] [in a new window]   FIG. 1.   SDS-PAGE and immunoblot analysis of µNS expression. Murine L cells were infected with T3D reovirus at an MOI of 40 and harvested at 24 h postinfection. Insect cells were infected with M3(T3D)-bac at an MOI of 1 and harvested at 52 h postinfection. (A) An SDS-polyacrylamide gel was loaded with 5 × 1010 reovirus virions (V), 5 × 1010 reovirus cores (C), cytoplasmic lysate of 8.8 × 104 L cells infected with reovirus (I), cytoplasmic lysate of 8.8 × 104 uninfected L cells (U), cytoplasmic lysate of 2.8 × 104 insect cells infected with M3(T3D)-bac (R), cytoplasmic lysate of 3.9 × 104 insect cells infected with wild-type baculovirus (WT), and cytoplasmic lysate of 2.3 × 104 uninfected insect cells (U). (B) An identically loaded gel was analyzed by immunoblotting with a polyclonal antiserum specific to µNS. The positions of µNS and µNSC from infected L cells are indicated to the left, and those of molecular mass markers (kilodaltons) are indicated to the right of each panel.

Density of cores, but not virions and ISVPs, is decreased after incubation with insect cell lysate containing µNS. Based on the observation from Morgan and Zweerink (30) that transcriptase particles from reovirus-infected cells represent core-like particles plus µNS protein, we tested the capacity of µNS from M3(T3D)-bac-infected insect cells to bind reovirus cores as well as other particle types. In initial experiments, lysate from insect cells expressing µNS (µNS lysate) was mixed with purified T3D cores, T1L ISVPs, or T3D virions. T3D ISVPs were not tested because of their instability at high concentrations. As controls, cores, ISVPs, and virions without insect cell lysate and µNS lysate without cores were analyzed separately. After a period of incubation, a floccular precipitate became visible in the cores-plus-µNS sample, but not in the other samples (data not shown). To separate the reovirus particles and particle-bound proteins from nonbound proteins, the samples were subjected to equilibrium centrifugation in CsCl density gradients. Following centrifugation, virus particles and/or other abundant proteins formed visible bands in the gradients (data not shown). The virions-plus-µNS and ISVPs-plus-µNS gradients contained well-defined bands near the positions of virions and ISVPs observed in the gradients with each of those particles alone (  1.36 and 1.38 g/cm³, respectively). In the cores-plus-µNS gradient, however, the core band was absent from its expected position, and instead a flocculent white band was seen at a higher position (lower density) than that of cores in the gradient with cores alone (  1.43 g/cm³). A second prominent band was observed at the top of the cores-plus-µNS, ISVPs-plus-µNS, and virions-plus-µNS gradients, at the same position as the band in the gradient with µNS lysate alone (  1.30 g/cm³). The loss of the core band and the appearance of a new band at lower density suggested that the RNA-to-protein ratio of the particles had been lowered by the binding of a lysate protein(s) to the cores.

View larger version (30K): [in this window] [in a new window]   FIG. 2.   Analysis of reovirus particle density shifts in the presence of lysate containing µNS. Reovirus cores (A), virions (B), and ISVPs (C) (2 × 1012) were incubated with 50 µl of buffer or insect cell lysate containing µNS. Lysate was also incubated alone. Samples were subjected to equilibrium centrifugation in CsCl density gradients and fractionated. The buoyant density and OD280 were determined for each fraction. The strong absorbance seen at the top of the gradient (lowest density) is due to detergent in the lysis buffer.

µNS binds cores but not virions or ISVPs. To determine if a specific lysate protein was interacting with cores, experiments similar to those described above were performed with an increased amount of µNS lysate to allow for detection of particle-bound protein by Coomassie blue staining. The prominent bands in the CsCl gradients were collected and analyzed by SDS-PAGE. The sample collected from the gradient of cores plus µNS contained both core proteins and an additional protein of approximately 80K (Fig. 3A, lane 2). This protein was not present in the samples collected from gradients of cores alone or of cores plus wild-type baculovirus lysate (Fig. 3A, lanes 1 and 3). Samples isolated from gradients containing virions and ISVPs, whether previously incubated with µNS lysate or not, contained only proteins present in the respective virus particles (Fig. 3A, lanes 4 through 7). Immunoblot analysis of the gradient-isolated material confirmed that µNS was present as an 80K protein doublet comigrating with cores (Fig. 3B, lane 2). µNS and cores comigrated in gradients after incubation at temperatures from 4 to 37°C and with core concentrations between 1.8 × 10¹² and 2.4 × 10¹³ cores/ml (data not shown). At all ratios of cores to µNS lysate tested to date, both proteins in the µNS doublet bound to cores (data not shown). No detectable µNS comigrated with virions or ISVPs (Fig. 3B, lanes 5 and 6). Similar results were obtained for T3D µNS binding to reovirus T1L particles: the µNS doublet bound to cores but not to virions (data not shown). Thus, the capacity to interact with cores was not specific to the strain or serotype of the virus particles. Moreover, the T1L µNS protein doublet was shown to bind to T1L and T3D cores but not to T1L and T3D virions or T1L ISVPs (data not shown), demonstrating that the selectivity of µNS binding to cores was not specific to the strain or serotype of the virus from which µNS was derived.

View larger version (59K): [in this window] [in a new window]   FIG. 3.   SDS-PAGE and immunoblot analysis of µNS binding to reovirus particles. Reovirus cores, virions, or ISVPs (2 × 1012) were incubated in the presence and absence of µNS lysate. Cores were also incubated with wild-type baculovirus lysate. Samples were subjected to equilibrium centrifugation in CsCl density gradients. (A) Particle bands isolated from CsCl density gradients were analyzed by SDS-PAGE with 5 × 1010 particles loaded per lane. Samples (lanes): no, no lysate added; µNS, µNS lysate added; WT, wild-type baculovirus lysate added. (B) An identically loaded gel was subjected to immunoblot analysis with a polyclonal antiserum generated against µNS. For both panels, molecular mass markers (kilodaltons) are on the left and protein mobilities are on the right.

Density of µNS-core complexes changes with the amount of µNS. To determine if µNS forms a defined layer on the outside of the core similar to the outer-capsid proteins µ1 and 3 in virions (9), we investigated whether µNS binding to cores is saturable. Equal numbers of cores were incubated with increasing volumes of µNS lysate. The mixtures were then analyzed in CsCl density gradients to determine their buoyant densities. As controls, cores alone, µNS lysate alone, and cores mixed with wild-type baculovirus lysate were also analyzed. Following centrifugation to equilibrium, the abundant proteins were visualized by direct observation. µNS cores generated with different volumes of µNS lysate were found at different positions in the separate gradients and, therefore, exhibited different buoyant densities. As the volume of added µNS lysate was increased in gradients 5 through 7 (Fig. 4), the buoyant densities of µNS cores continuously decreased. The complexes formed with smaller amounts of µNS lysate in gradients 3 and 4 (Fig. 4), appeared to migrate at a slightly higher density than cores alone in this experiment. This may have been due either to a small increase in the density of the complexes or to slight inconsistencies in the way the different gradients were formed. The complexes formed using the largest lysate volume (Fig. 4, gradient 7) migrated close to the position of the protein alone in the gradient of µNS lysate alone (Fig. 4, gradient 8). After isolation of the µNS-core complexes from the gradients, SDS-PAGE analysis showed that increasing amounts of µNS were bound to the cores, corresponding to the increase in added lysate (data not shown). As in previous experiments, migration of the core band was unchanged after incubation with wild-type baculovirus lysate (Fig. 4, gradient 9). The overall trend of decreasing buoyant density of complexes with increasing µNS, with the density ultimately approaching that of protein alone, suggested that the binding of µNS to cores is not saturable. The progressively larger amounts of µNS bound to a fixed number of cores may reflect the capacity of µNS molecules to self-associate.

View larger version (77K): [in this window] [in a new window]   FIG. 4.   CsCl density gradients of cores incubated with increasing amounts of lysate containing µNS. Cores (1012) were incubated alone (gradient 1); with 4.3 (gradient 2), 8.5 (gradient 3), 17 (gradient 4), 34 (gradient 5), 68 (gradient 6), or 140 µl (gradient 7) of µNS lysate; or with 140 µl of wild-type baculovirus lysate (gradient 9). µNS lysate (140 µl) was incubated alone (gradient 8). The samples were subjected to equilibrium centrifugation in CsCl density gradients and visualized with a high-intensity light. Solid arrowheads, bands of cores; open arrowheads, complexes of µNS cores; line, position of bands from the lysate alone.

Negative-stain EM of µNS cores reveals cores linked together in large complexes. To investigate the morphology of µNS cores, the complexes from the gradients in Fig. 4 were examined by negative-stain EM. Micrographs of µNS cores from gradients 5 and 6 revealed intact cores embedded within large complexes (Fig. 5A and B). Cores, the appearance of which is well defined (8, 42), are shown for comparison (Fig. 5B, inset). The cores in the µNS core samples were linked together by an amorphous density which we attribute to µNS. The large complexes and amorphous density surrounding the cores were not present in samples containing either cores alone or cores that had been mixed with wild-type baculovirus-infected cell lysates before gradient isolation (Fig. 5B, inset, and data not shown). µNS cores from other gradients exhibited similar morphology (data not shown). Notably, the cores in samples with smaller amounts of µNS were also linked together in complexes (data not shown). Since nonphysiological cysteine bond formation between µNS molecules was one possible explanation for the observations, in a subsequent experiment 1 mM dithiothreitol was added to the lysis buffer, the core-binding reaction mixture, and the CsCl gradient; however, the presence of this reducing agent did not affect the aggregated nature of µNS cores (data not shown). In sum, these data provide evidence that (i) µNS bound to cores such that a regularly structured outer capsid of protein was not formed and (ii) the intact particles were linked together.

View larger version (64K): [in this window] [in a new window]   FIG. 5.   Negative-stain EM of µNS cores and velocity sedimentation analysis of µNS lysate. µNS cores were purified on a CsCl gradient, dialyzed into VB, stained with uranyl acetate, and viewed by EM. (A) µNS cores viewed at low resolution. Bar, 200 nm. (B) µNS cores viewed at higher resolution. The inset provides an image of cores for comparison. Bars, 100 nm. (C) µNS lysate was subjected to velocity sedimentation on a sucrose gradient. The gradient was fractionated and subjected to immunoblot analysis with polyclonal antiserum specific to µNS. Markers (7S and 19S) were analyzed on a parallel gradient, and the positions were determined by OD260. The positions of the 7S and 19S markers and the top and bottom of the gradient are indicated. The 19S marker was pelleted at the bottom of the gradient with the centrifugation conditions used in this experiment. The position of µNS is indicated to the right.

µNS cores retain transcriptional activity. If the µNS-core complexes formed in vitro are similar to transcriptase particles isolated from reovirus-infected cells (30), the complexes should be transcriptionally active. The transcriptional activities of µNS cores containing different amounts of bound µNS were therefore tested and compared to that of cores. Cores mixed with wild-type baculovirus-infected lysate and then gradient purified were included as a control to address whether cellular factors from the lysate could affect the transcriptional activity of cores. Because the aggregated nature of µNS cores made them difficult to aliquot consistently, the transcriptional activity of each sample was standardized to the number of input particles by use of cores labeled with [³⁵S]methionine and [³⁵S]cysteine. Transcriptional activity was expressed as the ratio of ³²P incorporated into acid-precipitable counts to the ³⁵S counts in each reaction mixture. Results of a representative experiment are shown in Fig. 6A. µNS cores containing approximately 50 or 1,300 µNS molecules per core were slightly less active at transcription than either the original cores or cores mixed with wild-type baculovirus-infected lysate prior to purification. In four different experiments, the transcriptional activities of µNS cores ranged from 46 to 84% (mean, 66% ± 13%) of that of cores. The transcripts from cores and µNS cores were indistinguishable when separated on denaturing gels (data not shown). To verify that µNS remained bound to cores during transcription, transcription reaction mixtures containing µNS cores were layered on CsCl density gradients, spun to equilibrium, and fractionated. µNS continued to comigrate with cores in this experiment (data not shown), suggesting that it remained bound to cores. In sum, the results indicate that the binding of µNS to cores did not inhibit the core transcriptional activity as assembly of outer-capsid proteins is believed to do (2, 11, 21, 46; D. L. Farsetta, K. Chandran, and M. L. Nibert, unpublished data).

View larger version (61K): [in this window] [in a new window]   FIG. 6.   Analysis of µNS cores for transcription and methylation activities. (A) Standard transcription reactions were performed in triplicate with purified 35S-labeled cores (core), purified 35S-labeled cores previously mixed with wild-type baculovirus lysate (WT), and purified 35S-labeled cores with two different amounts of µNS bound. The approximate amount of µNS per core was determined by densitometry. The mean transcriptional activity ± standard deviation is reported as TCA-precipitable 32P counts relative to 35S counts to normalize for particle input. (B) Methylation activity was monitored using standard transcription reaction mixtures containing the methyl donor [3H]SAM. Activity is expressed as TCA-precipitable 3H counts normalized to the amount of transcription with TCA-precipitable 32P counts. These results are from four separate experiments with three separately prepared µNS core samples. The average activity of cores in each experiment was normalized to one, and the other samples were scaled appropriately. Cores mixed with wild-type baculovirus lysate and purified (WT), µNS cores with 350 molecules of µNS per core, and µNS cores with 980 molecules of µNS per core were assayed once in triplicate. The other samples of µNS cores were assayed twice in triplicate.

µNS cores may have elevated transcript 5'-end capping activities. To evaluate the effects of µNS on the capping of the viral transcripts, µNS cores and cores were quantitatively compared for 5' cap methylation. RNA methylation activity was assayed in a standard transcription assay mixture containing [-³²P]GTP and the methyl donor S-adenosyl-L-[methyl-³H]-methionine ([³H]SAM). In addition to assaying cap methylation, this approach indirectly assesses the RNA triphosphatase and guanylyltransferase capping activities of particles since these reactions must precede methylation (15). Methylation activity was expressed as a ratio of acid-precipitable ³H counts to acid-precipitable ³²P counts to adjust for any differences in transcription activity between samples. The activities of µNS cores with approximately 60 or 350 molecules of µNS bound per core and of cores mixed with wild-type baculovirus-infected lysate and then purified were similar to that of cores (Fig. 6B). However, when more µNS molecules were bound per core (530 to 3,800 molecules of µNS), the methylation activity was increased to approximately twofold that of cores (Fig. 6B). To test if the increased methylation activity was due to a contaminating methylase or nonspecific trapping of [³H]SAM when such large amounts of µNS were bound to cores, we repeated the assay using µNS cores containing either 60 or 980 molecules of µNS under conditions that are not permissive for full-length transcript production: in the presence of GTP only at 45°C, ATP only at 45°C, or all four nucleotides at 4°C. None of these conditions supported detectable methylation activity with either preparation of µNS cores or cores alone (data not shown). Thus, we conclude that the methylation increase observed with higher levels of µNS per core was dependent on transcript production, suggesting that it was due to an increase in transcript capping.

µNS incubation with cores greatly decreases the efficiency of recoated core formation. To test if µNS can inhibit the binding of the outer-capsid proteins to cores, we mixed µNS lysate and lysate containing µ1 and 3 (µ1-3 lysate) with cores in vitro. Relative amounts of the two lysates were chosen to provide approximately the same amount of µNS and µ1 for binding to the core surface. Cores were incubated alone, with µNS lysate, or with µ1-3 lysate to provide controls for the positions of the resulting particles, cores, µNS cores, and recoated cores, respectively, in CsCl gradients. Cores were incubated with both µNS lysate and µ1-3 lysate together, with µNS lysate first then with µ1-3 lysate, or with µ1-3 lysate first then with µNS lysate to see if the incubation of µNS with cores during or before the addition of µ1 and 3 would affect the formation of recoated cores. The samples were layered on CsCl density gradients, spun to equilibrium, fractionated, and subjected to immunoblot analysis. Cores alone migrated to the bottom of the gradient (Fig. 7A), whereas recoated cores migrated into the lower half of the gradient (Fig. 7B). Note that some core proteins and µ1 remained trapped at the top of the gradient, migrating with protein alone (Fig. 7B). µNS cores migrated only into the upper half of the gradient (Fig. 7C), near protein alone at the top of the gradient. When cores were incubated first with µ1-3 lysate, allowing the formation of recoated cores before µNS lysate was added, both the core proteins and µ1 were detected in the lower half of the gradient at the position of recoated cores (Fig. 7D). In addition, no µNS was detected in the fractions containing core proteins and µ1 at the position of recoated cores, confirming that recoated cores had been formed. This agrees with the result that µNS does not bind to virions (Fig. 3). When cores were incubated first with µNS lysate, allowing the formation of µNS cores before µ1-3 lysate was added, little or no core proteins or µ1 was detected in the lower half of the gradient (Fig. 7E), suggesting that the formation of recoated cores was greatly reduced by prior formation of µNS cores. When µ1-3 lysate and µNS lysate were mixed prior to addition of cores, little or no core proteins or µ1 was detected in the lower half of the gradient (Fig. 7F), suggesting that formation of recoated cores was greatly reduced by simultaneous incubation with µNS. From these results, we conclude that the incubation of µNS with cores greatly reduces the capacity of µ1 and 3 to bind to cores in a manner conducive to recoated core formation.

View larger version (64K): [in this window] [in a new window]   FIG. 7.   Incubation of cores with µNS lysate and/or µ1 and 3 lysate. Cores were incubated alone or with combinations of lysate containing µNS or µ1 and 3. Samples were layered on CsCl density gradients and spun to equilibrium. Gradients were fractionated, and the fractions were subjected to immunoblot analysis. (A) Cores incubated with lysis buffer. (B) Cores incubated with µ1 and 3 lysate. (C) Cores incubated with µNS lysate. (D) Cores incubated with µ1 and 3 lysate followed by addition of µNS lysate and further incubation. (E) Cores incubated with µNS lysate followed by addition of µ1 and 3 lysate and further incubation. (F) µ1 and 3 lysate and µNS lysate mixed, followed by addition of cores and incubation. Antibodies used in the immunoblots are listed to the left, and proteins are labeled on the right. The top and bottom of the gradients are labeled below. The positions of recoated cores and cores are labeled at the top.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

µNS cores are similar to transcriptase particles. The µNS-coated cores that we formed in vitro from purified cores and recombinant µNS protein share a number of characteristics with the transcriptase particles previously isolated from reovirus-infected cells (30). (i) They have similar protein compositions, including µNS in place of outer-capsid proteins. (ii) They have a complete dsRNA genome (30, 31), in contrast to "replicase" particles isolated from infected cells, which contained little or no µNS and were in the process of converting single-stranded RNA to dsRNA (30, 31). (iii) They are capable of synthesizing the viral plus strand transcripts, although the transcription activity of µNS cores was slightly lower than the activity of cores in this study (Fig. 6A). The activity of transcriptase particles was not quantitatively compared to that of cores in the previous study (30). The µNS cores that we generated in vitro also exhibit mRNA capping activities (Fig. 6B), whereas the capping activities of transcriptase particles were not tested in the previous study (30).

µNS core large-complex formation. The interaction of recombinant µNS with cores links them together within large complexes, as viewed by negative-stain EM. µNS binding to cores could produce such aggregates if µNS were present in insect cell lysate as large oligomers formed solely from µNS-µNS interactions; these large oligomers could bind to many cores, yielding large complexes. However, this seems unlikely since µNS sediments as a monomer or small oligomer in velocity gradients (Fig. 5C), suggesting that formation of large complexes is specific to the interaction with cores. If µNS is a monomer or small oligomer, it could link cores together by the binding of one µNS molecule to multiple core particles or by the oligomeric association of µNS molecules bound to different core particles. Once all of the µNS binding sites on the cores are occupied, additional µNS might be added to the complexes by µNS-µNS interactions so that the binding of µNS to cores is not saturable. Further studies on the oligomeric status of µNS and the localization of the µNS binding site(s) for cores may provide insight into the formation of the µNS-core complexes.

Possible role(s) of µNS-core interaction. In the infected cell, the interaction of the nonstructural protein µNS with the reovirus core may function in steps as diverse as regulation of outer-capsid assembly, virus particle assembly, and RNA translation or sorting. The binding of µNS to newly formed cores within infected cells could prevent the assembly of outer-capsid proteins, as suggested by in vitro findings in this study (Fig. 7), allowing those particles to continue synthesizing viral plus strand transcripts. Prevention of outer-capsid assembly onto certain cores may be required in the cell because assembly of the outer-capsid proteins is believed to shut off transcription by the enclosed core-like particles (2, 11, 21, 46; D. L. Farsetta, K. Chandran, and M. L. Nibert, unpublished data) and because the outer-capsid proteins are present throughout infection in high levels (17). It has been calculated that secondary transcriptase particles (newly formed particles within infected cells) produce 95% of the reovirus mRNA in infected cells (19, 20, 22). µNS could bind a subset of newly assembled cores, sequestering them away from outer-capsid proteins and dedicating them to produce mRNA. Alternatively, µNS cores could be formed only transiently on the assembly pathway from core to virion, but long enough to allow more transcript production than would occur in the absence of µNS.