Reovirus Nonstructural Protein σNS Recruits Viral RNA to Replication Organelles

Figures

• 

  FIG 1. Mutations in σNS do not appear to alter protein folding. (A) Sequence alignment of the N-terminal 38 amino acids (green bar) of σNS encoded by mammalian reovirus strains T1L, T2J, and T3D. Positively charged residues are shown in blue. (B) 35S-labeled in vitro-expressed σNS was incubated with proteinase K for the times shown, resolved by SDS-PAGE, and visualized by autoradiography. HEK293T cells were transfected with σNS alone (C) or cotransfected with σNS and μNS (D) and incubated for 24 h. Total protein in cell lysates was immunoprecipitated using an IgG isotype antibody (−), conformation-specific σNS-specific monoclonal antibody 2A9 (C; +), or σNS-specific monoclonal antibody 3E10 (D; +), resolved by SDS-PAGE, and immunoblotted using antisera directed against σNS (C and D) or μNS (D). Percentages of total lysates in the immunoprecipitation reactions (4% [C] or 3% [D]) were used as loading controls.

• 

  FIG 2. Alanine substitution of positively charged residues in a predicted RNA-binding domain of σNS alters RNA-dependent oligomerization. 35S-labeled σNS was expressed in rabbit reticulocyte lysates (RRLs) and incubated with or without RNase A. Samples were resolved by native PAGE to preserve oligomeric species (A) or SDS-PAGE to monitor protein expression (B) and visualized by autoradiography. The scale bar to the left of the native polyacrylamide gel (A) marks the kilodalton (kDa) ranges used for densitometric analysis of each σNS construct. The green scale bar marks σNS bound to RNA, whereas the orange scale bar marks unbound σNS. (C) The efficiency with which each σNS construct forms RNA-dependent oligomers was calculated by dividing the density of σNS bound to RNA (panel A, green scale bar) by the total density of σNS present in the gel (panel A, green and orange scale bar).

• 

  FIG 3. Mutants of σNS incapable of RNA binding fail to complement reovirus replication in cells expressing σNS-specific siRNAs. (A) Cells that constitutively express siRNAs directed against GFP or σNS were transfected with expression plasmids encoding GFP or the σNS constructs shown and incubated for 24 h. Cells were adsorbed with reovirus strain T3D at an MOI of 5 PFU/cell and incubated for 48 h. (B) Cell culture supernatants were collected for infectious virus quantification by plaque assay. Titer values that differ significantly from those obtained from cells expressing siRNAs against GFP complemented with GFP by one-way ANOVA with Dunnett’s multiple-comparison test are shown. **, P < 0.0021; ***, P < 0.0002; ****, P < 0.0001. (C) Cells that constitutively express siRNAs against GFP or σNS were transfected with expression plasmids encoding the σNS constructs shown and incubated for 24 h. Cell lysates were resolved by SDS-PAGE and immunoblotted using antisera directed against σNS and monoclonal antibodies specific for alpha-tubulin (α-Tub). Panel A was prepared using BioRender.

• 

  FIG 4. Mutations in σNS that compromise RNA binding disrupt incorporation of the protein into reovirus replication organelles. (A) Cells that constitutively express siRNAs directed against σNS were transfected with expression plasmids encoding the σNS constructs shown and incubated for 24 h. Cells were adsorbed with reovirus strain T3D at an MOI of 5 PFU/cell, incubated for 48 h, fixed, stained using σNS-specific monoclonal antibody 3E10 (cyan), μNS-specific antiserum (magenta), and DAPI (blue), and imaged using confocal microscopy. Regions selected for magnification are indicated by dotted white boxes. Bar, 5 μm. (B) The percentage of σNS immunofluorescence signal in reovirus factories was quantified by dividing the sum of σNS signal in reovirus factories by the sum of cytoplasmic σNS signal. Individual data points represent single cells. Percentage values that differ significantly from those obtained from WT σNS-transfected reovirus-infected cells by one-way analysis of variance (ANOVA) with Dunnett’s multiple-comparison test are shown. **, P < 0.0021.

• 

  FIG 5. Mutations in σNS that compromise RNA binding disrupt incorporation of the protein into factory-like structures. (A) Cells were transfected with expression plasmids encoding the σNS constructs shown along with μNS and incubated for 24 h. Cells were fixed, stained using σNS-specific monoclonal antibody 3E10 (cyan), μNS-specific antiserum (magenta), and DAPI (blue), and imaged using confocal microscopy. Regions selected for magnification are indicated by dotted white boxes. Bar, 5 μm. (B) The percentage of σNS immunofluorescence signal in reovirus factory-like structures was quantified by dividing the sum of σNS signal in reovirus factory-like structures by the sum of cytoplasmic σNS signal. Individual data points represent single cells. Percentage values that differ significantly from those obtained from WT σNS and μNS cotransfected cells by one-way analysis of variance (ANOVA) with Dunnett’s multiple-comparison test are shown. ****, P < 0.0001.

• 

  FIG 6. Reovirus transcripts are present in viral factories only when σNS is capable of binding RNA. Cells that constitutively express siRNAs directed against σNS were transfected with expression plasmids encoding the σNS constructs shown and incubated for 24 h. Cells were adsorbed with reovirus strain T3D at an MOI of 5 PFU/cell, incubated for 48 h, fixed, stained using RNA FISH probes specific for σNS mRNA (white) or σ3 mRNA (yellow), μNS-specific antiserum (magenta), and DAPI (blue), and imaged using confocal microscopy. Bar, 4.8 μm. The percentage of cytoplasmic σNS-mRNA (B, white bars) and σ3-mRNA (B, yellow bars) FISH signals in reovirus factories was quantified by dividing the sum of the FISH signal in reovirus factories by the sum of the cytoplasmic FISH signal. Individual data points represent single cells. Percentage values that differ significantly from those obtained from WT σNS-transfected reovirus-infected cells by one-way analysis of variance (ANOVA) with Dunnett’s multiple-comparison test are shown. *, P < 0.0332; ****, P < 0.0001.

• 

  FIG 7. Reovirus mRNAs are recruited to factory-like structures by σNS. Cells were transfected with expression plasmids encoding μNS, σ3, and the σNS constructs shown and incubated for 24 h. Cells were fixed, stained using RNA FISH probes specific for σNS mRNA (white) or σ3 mRNA (yellow), μNS-specific antiserum (magenta), and DAPI (blue), and imaged using confocal microscopy. Bar, 4.8 μm. The percentage of cytoplasmic σNS-mRNA (B, white bars) and σ3-mRNA (B, yellow bars) FISH signals in factory-like structures was quantified by dividing the sum of the FISH signal in factory-like structures by the sum of the cytoplasmic FISH signal. Individual data points represent single cells. Percentage values that differ significantly from those obtained from cells cotransfected with WT σNS, σ3, and μNS by one-way analysis of variance (ANOVA) with Dunnett’s multiple-comparison test are shown. ****, P < 0.0001.

• 

  FIG 8. Model of σNS mRNA recruitment for reovirus factory formation. (A) Following viral attachment and internalization, reovirus cores enter the cytoplasm and transcribe viral mRNAs that are translated to yield structural and nonstructural (NS) proteins. Factory scaffolds are formed by μNS (large magenta sphere), which recruits reovirus structural proteins. Viral mRNAs in the cytoplasm are bound by σNS and delivered to factory scaffolds. Viral components that form nascent viral cores are concentrated in factories, promoting viral packaging and core maturation for secondary rounds of transcription and translation. (B) Mutations in σNS that compromise RNA-binding capacity also impede recruitment of mRNAs to factories and stall factory enlargement, as nascent cores do not form and are thus incapable of amplifying viral replication. This figure was prepared using BioRender.