Input cores initially form peripheral core-only factories then transition into the perinuclear core-plus-OC factories while de-novo core-only peripheral factories are formed devoid of input cores

Recent findings reveal that reovirus replication does not occur within single, homogeneous factories. Instead, peripheral core-only factories initially form, containing core and non-structural proteins, but lacking outercapsid (OC) proteins, which are instead co-localize with LDs. These core-only factories permit amplification of cores away from OC proteins, which could otherwise cause premature full assembly and, consequently, halt further amplification. As replication proceeds, intermediate mid-cytoplasmic factories also form, where cores and LD–docked OC proteins mix. Finally, fully assembled and inert virions are deposited at large perinuclear compartments. Previous studies also found that the factory-forming non-structural µNS protein forms inclusion bodies in cells when transfected alone [34,51]. When cells with pre-existing µNS inclusion bodies were infected in the presence of cycloheximide, which thwarts viral replication, incoming cores co-localized with the µNS inclusion bodies [36]. These findings raised the possibility that incoming cores might reside in the early peripheral core-only factories. Alternatively, factories could be established at locations distant from input cores, using cytoplasmic viral proteins and RNAs produced by input cores. To distinguish between these possibilities, we aimed to develop a strategy to monitor input core localization throughout a bona fide reovirus replication cycle. To this end, core particles were generated in vitro from reovirus (strain T3D^PL) by chymotrypsin-mediated removal of OC proteins. The cores were then labelled with Alexa Fluor dyes. SDS-PAGE analysis confirmed that in vitro cores devoid of OC proteins were generated and efficiently fluorophore labelled (S1A Fig). Although core particles lack the OC machinery needed for cellular binding and entry events, they can still be effectively encapsidated in lipid-based transfection reagents and delivered into cells [52] (S1B Fig). To establish if Alexa Fluor labelled cores (AF-cores) had comparable replication kinetics to non-labelled cores, AF- and non-labelled cores were transfected into H1299 cells and de novo virus titers were quantified by plaque titration. While new virus production during whole virus infection is normally seen at 4–6hpi, transfected cores likely have accelerated onset of infection. Beyond the early timepoints when titers are close to the limit of detection, AF-cores demonstrated growth kinetics similar to those of non-labeled cores (S1C Fig), supporting their utility in tracking input core fate during infection.

To evaluate whether incoming cores reside in the early peripheral core-only factories, IF-CM was used to monitor whether AF-cores transfected into cells colocalized with the µNS-positive but OC-negative core-only factories. H1299 cells were transfected with AF647-cores, fixed at 6, 9, and 12 hours (S2A, C Fig) post-transfection (hpt), immunostained with antibodies directed against µNS and OC σ3 proteins, and subjected to IF-CM analysis (Fig 1A, B). It should be noted that σ3 immunostaining was used to identify OC-containing factories, and herein “OC” often refers to σ3. Three-dimensional images were then processed using the Volocity software for quantification as described in the Methods section. Aggregates of AF647-cores were sometimes visible (Fig 1, demarcated with asterisks), but such aggregates became in-focus at z-planes corresponding to the surface of cells, and were never associate within intracellular virus protein-positive factories; accordingly, such aggregates were excluded from analysis.” In the context of natural infection, ‘early’ infection refers to ~4–9 hpi, ‘mid’ infection to ~9–12 hpi, and ‘late’ infection to ~12–16 hpi, with titer saturation occurring around 15 hpi [49]. However, in experiments utilizing transfection of AF-core particles, the binding and uncoating steps of normal infection are bypassed, and thus, infection by AF-cores is accelerated, with saturation reached by 9–12 hpt. At an early time point of 6hpt, ~ 70% of all µNS-positive OC-negative factories were also positive for AF-cores; this suggested that the majority of initial peripheral core-only factories form around input cores (Fig 1C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Input cores initially form peripheral core-only factories then transition into the perinuclear core-plus-OC factories while de novo core-only peripheral factories are formed devoid of input cores.

H1299 cells were transfected with ~1000 AF 647-labelled reovirus core particles per cell (magenta). At 6, 9 and 12 hpt cells were fixed and subsequently immunostained with rabbit polyclonal sera generated against μNS (μNS, AF 488, green), mouse monoclonal anti-σ3 (10G10, Cy3, yellow) and nuclei were stained with DAPI (blue). (A) Representative images captured at 6 and 12 hpt. Aggregates formed by AF-cores are denoted with a white asterisk in image panels. (B) Select regions of OC(-) and OC(+) factories were artificially blown-up using Adobe Photoshop (white boxes). All images were acquired via immunofluorescence spinning disk confocal microscopy. (C) Of the total AF-core containing factories within a cell, the percentage of factories with or without outercapsid staining were graphed Statistical analysis is reported as ordinary two-way ANOVA with Tukey’s multiple comparisons test. (D) The edge-to-edge distance of each factory object with AF-cores, with or without OC staining, were quantified and graphed Statistical analysis is reported as ordinary one-way ANOVA with Sidak’s multiple comparisons test. (E) The volume of AF-core containing factories. Statistical analysis is reported as ordinary one-way ANOVA with Sidak’s multiple comparisons test. (F) Of the total factory volume containing AF-cores within a cell, the percentage of factory volume containing or not containing OC was graphed. Statistical analysis is reported as ordinary two-way ANOVA with Tukey’s multiple comparisons test. (G) Of the total number of OC(-) factories within a cell, the proportion of factories containing AF-cores or not were graphed. Statistical analysis is reported as ordinary two-way ANOVA with Tukey’s multiple comparisons test. (H) Of the total OC(-) factory volume within a cell, the percentage of factory volume containing or not containing AF-cores was graphed. Statistical analysis is reported as ordinary two-way ANOVA with Tukey’s multiple comparisons test. (I) Cartoon schematic describing conclusions of this work. Data represents five-ten images per condition and is representative of four independent experiments. Created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/o6xobjd. All data is plotted as the mean + /- 95% confidence intervals. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05.

https://doi.org/10.1371/journal.ppat.1013238.g001

Having determined that input cores reside in the early peripheral core-only factories during bona fide reovirus replication, it became possible to determine whether early core-only factories move towards the nucleus to become the larger mid-cytoplasmic factories, or whether input cores remain at the periphery while larger mid-cytoplasmic factories form independently. To distinguish between these possibilities, analysis of AF-core location over time was conducted. During early infection (6hpt), AF-cores were predominantly in OC(-) factories (Fig 1B top, and C), located farthest from the nucleus (Fig 1D) and were small in volume (Fig 1E). As time progressed, more AF-cores accumulated in OC-containing factories (Fig 1B bottom, and 1C) found closer to the nucleus (Fig 1D), which were larger than their OC(-) counterparts (Fig 1E) and constituted the majority of the viral factory volume by 12hpt (Fig 1F). These findings suggest that input cores initially start at peripheral core-only factories, but transition over time into mid-cytoplasmic OC-containing factories. Moreover, the population of OC(-) factories lacking input cores increased significantly in number (Fig 1G) and volume (Fig 1H) between 6–12hpt, suggesting that new de novo factories devoid of input cores are formed at the periphery of cells as infection process. Altogether, the data suggest that initial peripheral OC(-) factories are formed predominantly around input cores, which transition into the nuclear-proximal factories over time, while new de-novo OC(-) factories devoid of input cores are seeded in the periphery (Fig 1I).

Input cores establish independent peripheral compartments

When transfecting cores with one type of AF-fluorophore, it could not be distinguished whether input cores form distinct peripheral factories or whether different input cores share the same factories. Since incoming virions might bring distinct mutations and given that the peripheral factories are the first sites of core amplification, early coalescence of incoming cores could have implications for the timing of cooperativity between incoming virions. To address this question, in vitro-generated core particles were differentially labelled with AF-546 or AF-647 dyes. H1299 cells were first transfected with AF-546 labelled cores, and then one hour later, the cells were transfected with AF-647 labelled cores before fixation at seven hours after the initial transfection (Figs 2 and S3). AF-cores were transfected separately to prevent their co-encapsidation in lipoplexes and thus artificially delivered into the same location. Cells were immunostained and analyzed by IF-CM (Fig 2A, B). Only ~10% of AF-core puncta were shared between the two input cores (Fig 2B “Both”, C), with 90% being either AF-546–positive or AF-647–positive (Fig 2B “AF-546 Only” and “AF-647 Only”, C). It should be noted that there was a higher proportion of AF-647-independent puncta than AF-546, possibly due to increased transfection efficiency. These results suggest that input particles establish predominantly independent compartments at the beginning of infection (Fig 2D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Input cores establish independent peripheral compartments.

H1299 cells were first transfected with ~1000 AF-546 cores per cell. 1 hour later, the cells were transfected with ~1000 particles per cell of AF-647 labelled cores. 7 hpt, cells were fixed and processed for immunofluorescence confocal microscopy imaging. (A) Representative images of compressed Z-stacks. Cells were stained with DAPI to visualize nuclei. Aggregates formed by AF-cores are denoted with a white asterisk in image panels. (B) Compartments of every identity (AF-546 cores only, AF-647 cores only, and both cores) were selected and artificially blown-up using Adobe Photoshop (white boxes). (C) Using Volocity software (Quorum Technologies Inc.), the proportion of punctates containing AF546-cores alone, AF647-cores alone, or both cores co-localizing together was plotted. (D) Cartoon schematic summarizing findings, Created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/j1js5vhData is representative of four independent experiments with a minimum of five cells per condition analyzed. Data is plotted as mean + /- SD. Statistical analysis is reported as ordinary one-way ANOVA between the mean of each column. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05.

https://doi.org/10.1371/journal.ppat.1013238.g002

In early peripheral factories, most of the independent input core-containing OC(-) factories do not acquire newly synthesized core proteins from external factories

The finding that input cores initially produce independent peripheral compartments raised an interesting question: do de novo core proteins produced by an input core remain confined to their compartment of origin, or do de novo core proteins from independent input core factories intermix? Answering this question would provide insights into the potential for interaction and material exchange between distinct reoviral protein production centers. For example, if core proteins from different factories intermix, it could indicate a more interconnected network of viral replication sites, possibly allowing for resource sharing or adaptability in response to cellular defenses. However, addressing this question required us to create a new method for distinguishing core proteins expressed by distinct input cores, since there is a paucity of antibodies capable of discriminating between core proteins of distinct reoviruses, even among different strains or serotypes. Note that core proteins rather than OC proteins would need to be tracked because OC proteins are excluded from the core-only factories. To our knowledge, there are no examples of recombinant reoviruses containing exogenous sequences embedded into core proteins. To fill this gap, eight distinct flexible loops in the λ2 protein crystal structures were identified that may be amenable to insertion of FLAG and V5 epitope tags (Table 1 and S4 Fig, coloured sequences). The reovirus reverse genetics system was used to rescue recombinant viruses containing 2XFLAG or 1XV5 tags in each of the eight loops of λ2. Viruses were successfully recovered when 2XFLAG or 1XV5 tags were inserted at amino acids 1082 –1085, but not in the other seven loop regions (Table 1 and S4 Fig, orange sequences). We suspect this region is amenable to insertions because it does not hinder OC assembly and/or λ2 turret movement during transcription [46,48,53,54] (S4A–H Fig). The λ2 proteins of purified FLAG- and/or V5-tagged reoviruses (S4I Fig) could clearly be distinguished by Western blot analysis with corresponding tag-specific antibodies (Fig 3A). Similarly, cells infected by λ2-FLAG- and/or λ2-V5-tagged reoviruses could effectively be distinguished by immunofluorescence with FLAG- or V5- antibodies (Fig 3B). To determine whether the λ2-V5 and λ2-FLAG-tagged viruses had similar replication kinetics as the wild-type T3D^PL reovirus, we evaluated the attachment efficiency to cells by flow cytometric analysis (Fig 3C), the expression of viral proteins over time by Western blot analysis (Fig 3D, E), and progeny virus production by plaque titration (Fig 3F). Though slightly attenuated in all three respects, the tagged viruses displayed replication kinetics similar to wild-type and were deemed appropriate models for application to studies of de novo protein distribution dynamics in factories.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. λ2 regions tested for insertion of tag sequences.

https://doi.org/10.1371/journal.ppat.1013238.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Recombinant tagged reoviruses λ2-FLAG and λ2-V5 are comparable in their replication kinetics.

(A) Equivalent amounts of CsCl purified λ2-tagged viruses were subject to SDS-PAGE followed by either Coomassie blue staining or Western blot with polyclonal rabbit anti-λ2 (AF 657), monoclonal mouse anti-FLAG M2 (AF 488), or polyclonal rabbit anti-V5 (AF 647). (B) H1299 cells were infected with λ2-FLAG, λ2-V5, or both viruses at an MOI of 5, fixed at 12 hpi, and immunostained with monoclonal mouse anti-FLAG (AF 488, green) and polyclonal rabbit anti-V5 (AF 647, magenta) and imaged via immunofluorescence confocal microscopy. Created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/3n918ke(C) Equivalent particles of T3D^PL (WT), λ2-FLAG, and λ2-V5 were used to infect L929 cells at 4°C for 1 hour. After 1 hour, cells were collected using cell stripper and fixed. Cells were immunostained with mouse monoclonal anti-σ3 (10G10, AF 647) and the amount of virus bound was quantified by flow cytometry (MFI). Data is representative of three independent experiments. (D-F) H1299 or L929 cells were infected with WT, λ2-FLAG, or λ2-V5 at a MOI of 5. After one hour of infection, virus-containing media was replaced with complete media with or without 100 μg/mL of cycloheximide (CHX). Cell lysates were collected every three hours from 9-18hpi for either SDS-PAGE or plaque assay. (D) Quantifications of reovirus proteins by SDS-PAGE and Western blot analysis of cell lysates with polyclonal rabbit sera directed against reovirus, and monoclonal mouse anti-β-actin. Data are representative of three independent experiments. (E) Representative Western blots from which quantifications in (D) were drawn. (F) Viral titers as measured by plaque assay, and statistical analysis is reported as unpaired t-test between the mean of each column. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05. Data are representative of four independent experiments. All data is plotted as mean + /- 95% CI. Statistical analysis is reported as ordinary one-way ANOVA between the mean of each column (binding assay, C) or timepoint (D). ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05.

https://doi.org/10.1371/journal.ppat.1013238.g003

While FLAG and V5 antibodies effectively stained de novo λ2-FLAG and λ2-V5 produced from their respective viruses, they did not stain incoming cores. This was likely because there were insufficiently exposed epitopes on incoming cores to be detected by the antibodies, or because the epitope is masked within the cores (S4C–E for whole core, S4F–H for whole virus structures). Given that incoming AF-647- and AF-546-labelled cores appeared to predominantly form independent peripheral factories (Fig 2), we asked whether, nevertheless, these peripheral factories shared de novo produced core proteins. In other words, could a λ2-FLAG product associate with an AF-λ2-V5-containing early peripheral compartment? To explore this question, AF-labelled cores generated from λ2-FLAG or λ2-V5 viruses were transfected into H1299 cells, fixed at 7 hpt, and immunostained with α-FLAG antibodies to discriminate between input and de novo protein-containing compartments (Figs 4A, B and S5A–C). Among the de novo λ2-FLAG-containing factory population, just over 50% contained no input cores, suggesting that de-novo proteins establish new factories devoid of input cores altogether (Figs 4B “de novo FLAG alone”, C, and S5D, E). Approximately 21% of de-novo λ2-FLAG associated with factories produced by the parent input core (AF-λ2-FLAG) (Figs 4B “de novo FLAG and input FLAG”, C, and S5D, E), while only ~10% co-localized with their non-self input (AF-λ2-V5), and ~13% co-localized with factories positive for both input particles (Figs 4B “de novo FLAG and input FLAG and input V5”, C and S5D, E).Through a complementary analysis, a plurality of input AF-λ2-V5-positive compartments did not contain de novo or input λ2-FLAG (~47%), where the remaining puncta (with de novo and input λ2-FLAG, with only de novo λ2-FLAG, or with only input λ2-FLAG) each made up 14–19% of the population (Figs 4D, S5F, G). Considering these results together, it appears that the majority of de novo core proteins (~56%) become directed to new factories devoid of input cores, and the majority of input-core-containing factories (~47%) do not contain de novo proteins produced by non-self cores. Nevertheless, some (~15–20%) factories produced by input cores do receive de novo proteins produced by non-self cores.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. In early peripheral factories, most of the independent input core-containing OC(-) factories do not acquire newly synthesized core proteins from external factories.

H1299 cells were first transfected with ~1000 AF-546 labelled FLAG- or V5-tagged cores per cell. 1 hour later, the cells were transfected with ~1000 particles per cell of AF-647 labelled FLAG- or V5-tagged cores (opposite tag to the first transfection). 7 hpt, cells were fixed and processed for immunofluorescence confocal microscopy. (A-B) Representative images of compressed Z-stacks. Cells were immunostained with monoclonal mouse anti-FLAG (AF 488, green) and stained with DAPI to visualize nuclei. White asterisks indicate aggregates formed by AF-cores. (B) Compartments of different identities (de novo FLAG alone, de novo FLAG with input FLAG, de novo FLAG with both input FLAG and input V5) were selected (white boxes in (A) and artificially blown-up using Adobe Photoshop software. (C-D) The number of AF-FLAG cores, AF-V5 cores, and de novo λ2-FLAG punctates and the degree of colocalization between each were quantified. Circle data points represent data where AF-FLAG cores were transfected first, and triangles represent data where AF-V5 cores were added first. (C) The number of de novo λ2-FLAG (green) punctates were quantified and the proportion of punctates per cell lacking both AF-FLAG and AF-V5, containing only AF-FLAG, containing only AF-V5, or lacking both AF-cores were plotted. (D) The number of AF-V5 core punctates were quantified and the proportion of punctates co-localizing with both de novo and AF-FLAG, de novo FLAG only, AF-FLAG only, or neither de novo nor AF-FLAG were plotted. (E) The volume of punctates were quantified. Groups are categorized based on co-localization of input AF-FLAG cores, input AF-V5 cores, and/or de novo λ2-FLAG proteins. (F) Cartoon schematic, created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/nqc59nl, summarizing the findings from A-E. All data is representative of four independent experiments with a minimum of five cells per condition analyzed. All data is plotted as mean + /- SD. Statistical analysis is reported as ordinary one-way ANOVA between the mean of each column. ****p < 0.0001, ***p < 0.001, **p < 0. 05, ns > 0.05.

https://doi.org/10.1371/journal.ppat.1013238.g004

Although the majority of input-core-containing factories did not contain non-self cores or de novo proteins, an assessment of the relative volumes for each unique factory composition provided new insights into the expansion and mixing that occur between peripheral factories. The volumes of factories containing both input λ2-FLAG and input λ2-V5 cores were larger than those of factories with only one input core population (Fig 4E), suggesting mixing or merging of early single-input-core peripheral factories. In all cases, factory volumes were significantly larger when they also contained de novo core proteins produced by the reciprocal virus, with the largest volumes representing both input cores and non-self de novo core proteins. The analysis suggests that while input cores initially form a majority of independent peripheral factories with minimal de novo proteins from other input cores, larger, more-mature factories ultimately contain mixtures of input cores and de novo proteins (Fig 4F). Moreover, de novo proteins can also create new peripheral factories devoid of input cores, but such factories are smaller than those formed from input cores.

As replication progresses, de-novo core proteins start to mix indiscriminately between independent compartments created by input cores

The focus on OC-negative peripheral factories (Fig 4) revealed that while the majority of core-only factories were composed of independent input cores and self-derived de novo proteins, some new core-only factories contained a mixture of different input cores and non-self de novo proteins. In addition, other factories were formed entirely by de novo core proteins, devoid of input cores. To understand how composition changes as factories transition from peripheral core-only to mid-cytoplasmic core-plus-OC compartments, we evaluated the distribution of de novo proteins from two independent sources (λ2-FLAG and λ2-V5 viruses) between core-only (OC-negative) peripheral factories and OC-containing intermediate compartments. Instead of using labelled input cores, a bona fide virus infection cycle was used, where H1299 cells were co-infected with λ2-FLAG and λ2-V5 viruses. Infected cells were fixed at 8 hpi, immunostained with α-FLAG and α-V5 antibodies to detect core λ2 proteins, as well as with OC σ3 protein-specific antibodies to discriminate between OC(-) and OC(+) factories (Figs 5A and S6). Following IF-CM analysis, there was an equivalent proportion of the OC(-) peripheral factories positive for de novo λ2-FLAG (FLAG+) only, de novo λ2-V5 (V5+) only, and both de novo λ2-FLAG and λ2-V5 (FLAG + /V5+) (Fig 5B). However, the factories that contained both de novo λ2-FLAG and λ2-V5 were significantly larger in volume relative to the de-novo λ2-FLAG or λ2-V5 only factories (Fig 5C), despite all three localizing at approximately the same peripheral region of the cell (Fig 5D, E). The larger size of OC(-) peripheral factories containing a mixture of λ2-FLAG and λ2-V5 de novo proteins may represent new factories formed by both de novo proteins and/or independent factories that merge into larger shared ones.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. De novo core proteins start to mix indiscriminately between independent compartments created by input cores.

H1299 cells were co-infected with λ2-FLAG and λ2-V5 at an MOI of 5 each. 8 hpi, cells were fixed and immunostained with monoclonal mouse anti-FLAG (AF 488, green), polyclonal rabbit anti-V5 (AF 647, magenta) and monoclonal mouse anti-σ3 directly conjugated to Alexa Fluor 594 (10C1, yellow) and stained with DAPI for nuclei visualization (blue). (A) Representative immunofluorescence confocal images. (B-D) Quantifications of OC-negative factory populations. (B) The number of FLAG(+), V5(+), or FLAG and V5(+) (Shared) OC(-) factory objects per cell graphed as a percentage of the total outercapsid-negative factory objects per cell. The (C) volume and (D) edge-to-edge distance from the nucleus of FLAG(+), V5(+), or shared OC(-) factory objects. (E) Cartoon model of the OC(-) factory populations observed at 8hpi. Created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/olzwmfp. (F-H) Quantifications of OC(+) factory populations. (F) The number of FLAG(+), V5(+), or shared OC(+) factory objects per cell graphed as a percentage of the total OC(+) factory objects per cell. The (G) volume and (H) edge-to-edge distance from the nucleus of FLAG(+), V5(+), or shared OC(+) factory objects. (I) Cartoon model of the OC(-) and (+) factory populations observed at 8 hpi. Created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/7blhjp7. All data is representative of three independent experiments with a minimum of 5 cells analyzed per experiment. Data is plotted as mean + /- 95% CI. Statistical analysis is reported as ordinary one-way ANOVA between the mean of each column. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05.

https://doi.org/10.1371/journal.ppat.1013238.g005

With respect to the OC(+) population, which represented factories further along in the replication cycle and containing de novo OC proteins (σ3), there were equivalent proportions of smaller, independent FLAG+ and V5 + factories, but a significant increase in the proportion of FLAG + /V5 + factories, which were larger in volume compared to only FLAG+ or only V5 + factories (Fig 5F, G). All OC(+) factories were located at similar distances from the nucleus, though they were closer to the nucleus than the OC(-) factories (Fig 5H, I). These observations suggest that as early as 8 hpi, when core amplification is exponential, de novo core proteins from different viral sources co-localize within the same factories, with approximately one-third of OC(-) core-only factories and one-half of OC(+) factories containing shared de novo proteins. This co-localization of proteins from different viral sources raises the possibility that core amplification factories could contribute proteins with distinct mutations or functions, although future studies would be needed to directly establish the extent of functional complementation.

Altogether, Figs 1–5 reveal the dynamic formation of peripheral core-only factories and transition to core-plus-OC mid-cytoplasmic factories as follows: Input core particles first seed independent initial OC(-) factories in the cell periphery, with the majority retaining only self-derived de novo proteins. As infection progresses, these input(+) factories move into intermediate regions where they acquire OC proteins. During this process, the originally independent factories are more likely to merge and/or acquire de novo-produced products from non-self origins, resulting in the “shared” factory phenotype. While the first wave of core-only factories move into intermediate regions, new de novo OC(-) factories that lack input particles are simultaneously seeded. During this second wave of replication, products mix indiscriminately, forming an equivalent amount of independent and shared de novo OC(-) factories. By late infection, input particles accumulate in OC(+) perinuclear factories of shared phenotype, where all components have merged, while new de novo OC(-) factories continue forming in the periphery during ongoing rounds of replication.

Microtubule polymerization does not affect OC(-) peripheral factories and their transition to OC(+) mid-cytoplasmic factories but is required for transition of whole assembled virions into perinuclear depots

The transition of independent input cores from peripheral core-only factories to mid-cytoplasmic OC+ factories, their later transitions into perinuclear condensates, and the sharing of de novo core proteins suggest a movement of factories and proteins that could be mediated by passive mechanisms such as diffusion along concentration gradients, or active mechanisms such as actin-based and/or microtubule-dependent transport [55]. Previous studies found that an intact microtubule (MT) network and motor protein dynein are necessary for factory accumulation at perinuclear areas [56,57]. However, these studies focused on late time points when most core amplification and progression into perinuclear factories would have already occurred and were conducted prior to discovery of distinct core-only versus core+OC factory types. Nocodazole treatment was also previously suggested to reduce genome packaging efficiency [58], however recent cryoET studies [59] suggest that electron microscopy structures previously thought to represent genome-devoid whole particles may sometimes represent shells formed by outercapsid proteins rather than empty whole particles. Accordingly, to explore the earlier stages of factory dynamics during reovirus replication, the role of MT was revisited at time points that capture exponential core amplification and all three compartment types: core-only peripheral factories; mid-cytoplasmic core+ OC+ factories, which are largely devoid of fully assembled viruses but adjacent to OC-decorated lipid droplets; and perinuclear compartments that house fully assembled virions.

To test the role of MTs during factory transitions, wild-type T3D^PL reovirus-infected cells were treated with the MT depolymerizing drug nocodazole at 1 hpi for the duration of infection. Nocodazole was added post-virus adsorption to avoid potential effects of MT disruption on reovirus endocytosis [50]. Cells were fixed at 8 and 12 hpi, followed by IF-CM to monitor protein and factory distribution and composition (Figs 6 and S7). Importantly, β-Tubulin staining appeared filamentous under control DMSO treatment but diffuse and granular following nocodazole treatment, indicating that MT integrity was successfully disrupted by nocodazole treatment (Figs 6A, 100x β-Tubulin panels in magenta, S7A,B Fig). When IF-CM with polyclonal core protein–specific antibodies and OC σ3 was used to distinguish core-only from core+OC factories, there was no difference between DMSO and nocodazole treatment in the percentage of OC(-) and OC(+) factories (Fig 6B), nor in the volumes of either OC(−) or OC(+) factories (Fig 6C). These results suggest that MT networks are not required for the establishment or growth of core-only factories, nor for their transition into mid-cytoplasmic core-plus-OC factories.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Microtubule polymerization does not affect OC(-) peripheral factories and their transition to OC(+) mid-cytoplasmic factories but is required for transition of whole assembled virions into perinuclear depots.

H1299 cells were infected with T3D^PL MOI 3. 1 hpi, complete media containing 10μM nocodazole, or an equivalent volume of DMSO (control), was added to the cells. 8 or 12 hpi, cells were fixed and immunostained for immunofluorescence confocal microscopy imaging. (A) Representative images of cells stained with: monoclonal mouse anti-tubulin (12G10, AF 647, magenta), BODIPY 493/503 for lipid droplet visualization (LDs, yellow), DAPI for nuclei visualization (blue), monoclonal mouse anti-σ3 (10C1, directly conjugated to AF 594 in the middle panel (yellow), or 10G10 followed by AF 647 conjugated secondary antibodies in the right panel (magenta)), and/or polyclonal rabbit antibodies raised against reovirus cores (AF 488, green). Images were captured at 100x and 60x magnification. Images were created from Z-stacks and are representative of a minimum of five images captured for each condition from three or four independent experiments. (B) The number of factories positive for OC staining were quantified and graphed as a percentage of total factories per cell. (C) The factory object volumes and (D) edge-to-edge distances from the nucleus were quantified. (F) The total number of lipid droplets per cell were quantified. (F) The number of lipid droplets associated with OC proteins per cell were graphed as a percentage of the total number of lipid droplets. (G) The edge-to-edge distance to the nucleus of each lipid droplet was graphed. (H) The mean edge-to-edge distance between each lipid droplet per cell was graphed. All data is representative of 3 independent experiments and plotted as mean + /- 95% CI. Statistical analysis is reported one-way ANOVA between the mean of each column. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05.

https://doi.org/10.1371/journal.ppat.1013238.g006

As was previously observed by others at late timepoints of infection [56], there was a lack of large interconnected perinuclear factories adjacent to the nucleus following nocodazole treatment compared with DMSO (Figs 6A [σ3 and core staining in bottom panels of both 100x and 60x] and S7C). Quantitatively, OC(-) factories remained significantly further from the nucleus at both 8 and 12 hpi, and OC(+) factories at 12 hpi following nocodazole treatment (Fig 6D). Moreover, because previous findings have implicated an association between OC proteins and LDs that may impact assembly [49,60], the number of LDs and the proportion associated with OC proteins following nocodazole treatment were also assessed. Quantitative IF-CM revealed that neither metric was altered following nocodazole treatment at either 8 or 12 hpi (Fig 6E, F). In addition to associating with OC proteins, a previous study observed that LDs can congregate closer to the nucleus and each other following reovirus infection [49]. Both of these effects were reduced following nocodazole treatment; LDs were found farther from the nucleus (Fig 6G) and farther away from each other than in the DMSO control (Fig 6H). Altogether, the early events of reovirus amplification, including the formation of core-only factories, core-plus-OC factories, and LD-associated OC proteins, can occur and transition to mid-cytoplasmic compartments in an MT-independent manner. However, in the absence of an intact MT network, factory progression is stalled at the mid-cytoplasmic OC(+) intermediate phase and fails to produce the large perinuclear depots of fully assembled virus particles.

Results presented previously demonstrated that input core particles transition from predominantly OC(-) factories in the cell periphery to OC(+) factories in intermediate and perinuclear areas as infection progresses (Fig 1). When MTs were disrupted, OC(+) factory progression became stalled at the mid-cytoplasmic location with circular-factory morphology, rather than progress to the amorphic nuclear-proximal location (Fig 6). The OC(-) factories were still present in the periphery. To determine if the dynamics of input core particles could be altered following nocodazole treatment, or in other words, if the localization of input particles might be MT-dependent, H1299 cells were transfected with AF546-cores, fixed at 12 hpt, and immunostained with antibodies directed against core proteins and OC σ3 (Fig 7). 1 hpt, transfected cells were treated with the MT depolymerizing drug nocodazole at 1 hpi for the duration of infection, or with DMSO as a control (Figs 7A, B, and S8A, B). Nocodazole-mediated MT disruption did not affect the proportion of factories associated with input AF-cores (Fig 7C) nor the proportion of AF-cores associated with OC(+) factories (Fig 7D) at 12 hpt. Moreover, single pinpoint foci of AF-core fluorescence could be easily visualized and counted within factories, with average diameter of 464nm and high circularity. The pinpoints were unlike the aggregates demarked by asterisks, which were much larger, were poorly circular, and were never associated with factories. Although the measured diameter of each pinpoint exceeds the size of a single virus particle (60–80 nm), objects smaller than the ~ 200 nm diffraction limit of confocal microscopy often appear blurred, causing their fluorescence signals to be larger than their true size. Consequently, it was not possible to definitively determine whether each pinpoint focus represented a single core or multiple cores clustered together. Nonetheless, the number of pinpoints per factory ranged from 0 to over 50. Counting these pinpoints revealed that factories with only a few input cores behaved similarly to those with many input cores, suggesting that the number of input cores does not alter the phenotype of the factory. Importantly, nocodazole treatment did not significantly alter the distribution of input core number per factory. These data suggest that the distribution of AF-cores across factories, including those containing OC proteins, is not mediated by MTs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. The localization of input AF-core particles is not microtubule dependent.

H1299 were transfected with ~1000 AF-546 reovirus core particles per cell (magenta). 1 hpt, cells were treated with (A) DMSO or (B) 10μM nocodazole. 12 hpt, cells were fixed and immunostained with monoclonal mouse antibodies directed against σ3 (10G10, α-σ3, AF 647, cyan) polyclonal rabbit antibodies raised against reovirus cores (α-core, AF 488, green) and DAPI was used to stain nuclei. All images are of compressed Z-stacks captured by IF-CM. White asterisks indicate aggregates formed by AF-cores. Factory objects within the cell were artificially blown-up using Adobe Photoshop software to showcase regions containing de novo cores only – no input particles, de novo cores with input, core and OC with no input, and core and OC with input. (C) The number of factories per cell, and the number of input AF-core particles within the factories, were quantified from 3D unedited images using Volocity software. Statistical analysis is reported one-way ANOVA between the mean of each column. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05. (D) The percentage of OC factories that contained input AF-cores out of the total number of factories per cell were quantified. Statistical analysis is reported as unpaired t-test between the means of the columns. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05. All data is plotted as mean + /- 95% CI. All data is representative of a minimum of five cells per three independent experiments.

https://doi.org/10.1371/journal.ppat.1013238.g007

Numerous studies have suggested that while some laboratory strains of reovirus such as T3D^PL and T1L produce similar factory morphologies, these differ from the factory morphology of the T3D^N laboratory strain. The differences in factory morphologies were previously associated with the amino acid at position 208 of the MT-associating μ2 protein [57]. It was important, therefore, to establish if the factory characteristics discovered in our studies using T3D^PL are representative of reoviruses more generally. Environmentally derived reoviruses that have not undergone lab culture adaptation provide a strong reflection of the reovirus quasispecies as a whole. The phenotypes observed for T3D^PL following nocodazole treatment (Fig 6) were consistent among environmentally derived reoviruses (T1E1, T2E1, T2E2), as well as laboratory strain T1L (S9A Fig). Moreover, with the exception of T3D^N, the remaining viruses all share the same amino acid at position 208 of μ2 (S9B Fig). Accordingly, the role MT networks play in factory transitions is not limited to only T3D^PL or T1L reoviruses but applies to naturally derived reoviruses as well.

Microtubule polymerization in part mediates the core to whole virus transition

Although nocodazole treatment inhibited the accumulation of large perinuclear factories, OC(+) factories still formed in intermediate areas. Previous studies using nocodazole focused on late time points of infection and reported similar results [57]. Studies assessing the effects of nocodazole on reovirus titers, however, produced conflicting results. While one study found no significant differences in titers at 24 hpi following nocodazole treatment beginning at 6 hpi with either T1L or T3D^N [61], another study using the same experimental approach found an approximately 50% reduction in titers at 24 hpi using the T1L strain. No significant reduction was observed with the T3D strain, although it was not explicitly stated which T3D strain was used [58]. Hence, to investigate whether nocodazole treatment impacts T3D^PL progeny titers over the course of replication, T3D^PL-infected H1299 cells, with or without nocodazole treatment, were subjected to single-step growth kinetic analysis. There was a ~ 10-fold decrease in viral progeny titers under nocodazole treatment beginning at 12 hpi (Fig 8A). To determine if the diminished progeny production could have been due to a decrease in de novo protein production following MT disruption, duplicate wells were subjected to SDS-PAGE and quantitative Western blot for assessment of reovirus protein levels (Fig 8B). An ~ 2-fold decrease in core protein production at 12 hpi following nocodazole treatment was observed (Fig 8C), with a similar decrease in OC protein expression between 12–18 hpi (Fig 8D). However, a 2-fold decrease in protein production mid-to-late infection was unlikely to fully account for a 10-fold decrease in progeny virus generation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Nocodazole-mediated microtubule disruption delays progeny virus full assembly.

(A-D) H1299 cells were infected with T3D^PL at an MOI of 3. 1 hpi, media containing either 10μM nocodazole or an equivalent volume of DMSO (control) were added to the cells. (A-B) Beginning at 6 hpi, cell lysates were collected every 6 hours for 24 hours. (A) The total virus titer from media plus cell lysates were determined. Titer data is plotted as mean + /- SD. Statistical analysis is reported one-way ANOVA between the mean of each timepoint. ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05. (B-D) Beginning at 12 hpi, cell lysates were collected every 6 hours for 30 hours. Lysates were subject to SDS-PAGE and Western blot analysis using polyclonal rabbit sera directed against reovirus and monoclonal mouse anti-β-actin. (B) Representative blots of dilution series of each sample for accurate quantification. Western blot was performed with antibodies directed against purified core particles (α-core, top), whole virus particles (α-reo, middle) and β-actin (bottom) Quantification of (C) core proteins(λ1/λ2) normalized to β-actin and (D) outercapsid proteins (μ1 + μ1C+σ3) normalized to β-actin. Western blot quantifications are plotted as mean + /- SD. Data from (A) is representative of four independent experiments, data from (B-D) are representative of three independent experiments and statistical analysis is reported as multiple unpaired t-tests between the mean of each timepoint and/or column ****p < 0.0001, ***p < 0.001, **p < 0.05, ns > 0.05. (E-F) H1299 cells were first transfected with siRNAs against the M2 gene, or a non-targeting control (in DMSO alone and Noco columns). Cells were infected with T3D^PL at an MOI of 3. 1 hpi, cells were treated with 10μM nocodazole or DMSO in the presence of absence of 100μg/mL cycloheximide. Every three hours for 24 hours, both RNA and cell lysates were collected to measure de novo core and virus generation, respectively. (E) The single-step growth kinetics of whole viruses (black) and cores (blue). (F) The percentage of core particles (blue) and whole virus particles (grey) out of the total particles grouped over the timeframes. (G) Cartoon schematics summarizing the findings shown in (E) and (F)created in BioRender. Shmulevitz, M. (2025) https://BioRender.com/ocj82yz for DMSO summary, for the M2KD summary: Shmulevitz, M. (2025) https://BioRender.com/br68zkb, for the NT summary: Shmulevitz, M. (2025) https://BioRender.com/xrumjpx, and for the bottom icons: Shmulevitz, M. (2025) https://BioRender.com/ydy4k1q. Data for (E) and (F) is plotted as mean + /- SD and represents three independent experiments.

https://doi.org/10.1371/journal.ppat.1013238.g008

The effects of nocodazole treatment were explored further by electron microscopy as a direct but qualitative assessment (S10 Fig). Recent cryo-electron tomography studies have identified multiple structures within infected cells, including virus cores, intact virions containing dsRNA genomes, empty virions lacking genomes, and newly discovered outer capsid (OC) protein shells [59]. Both dsRNA-containing virions and OC shells were observed within peripheral and mid-cytoplasmic factories, in both nocodazole-treated and untreated samples. The number of outer capsid shells varied significantly among different cells and even among factories within the same cell, making precise quantification from single-plane EM images challenging. Nevertheless, qualitatively, there was no striking difference in the overall abundance of outer capsid shells between treated and untreated cells. Minimal numbers of empty, genome-devoid virions were observed in either condition, suggesting that aberrant genome packaging is unlikely to account for the approximately 10-fold difference in viral titers. The most notable difference observed was that in untreated cells, microtubules appeared to infiltrate mid-cytoplasmic factories. Fully assembled virions accumulated with increasingly greater density within microtubule-rich domains, giving the impression that microtubules were releasing, guiding and/or extracting mature particles from the factories and redistributing them into increasingly tighter fully assembled virion storage depots.

Given that the major effect of disrupting MT networks was cessation of perinuclear factories that normally represent areas of fully assembled virus deposition, we wondered if the reduced titers in nocodazole-treated cells might reflect a change in the transition of core particles to fully assembled particles. To determine the impact of nocodazole on particle assembly, the kinetics of core and whole virion amplification over the first round of infection were assessed. H1299 cells were infected with reovirus in the presence or absence of nocodazole, and every 3 hours for 24 hours, RNA and whole cell lysates were collected to measure core and virus production, respectively. Since cores synthesize RNAs, RT-qPCR of reoviral RNAs was used as a proxy for core amplification. Whole virus levels were simultaneously measured by plaque titration. To differentiate RNA and whole virions synthesized linearly from input particles as opposed to logarithmic amplification from de-novo synthesized particles, cycloheximide was used to thwart secondary rounds of viral amplification by inhibiting translation. As a positive control for delayed OC assembly onto cores, cells were treated with siRNAs to knockdown the M2 gene, that encodes the OC μ1 protein, prior to infection in the absence of nocodazole. The single-step kinetics of core-derived RNAs and whole infectious virions were logarithmic during normal infection (Fig 8E, left (DMSO)). Core RNA amplification was not disrupted by nocodazole treatment. However, compared to DMSO-treated cells, there was a delay in whole virus production relative to core amplification in both M2-siRNA– and nocodazole-treated cells (Fig 8E, middle (M2-KD) and right (Noco), respectively). When plotted as the ratio of core-to-whole-virus production at early (6–9 hpi), mid (9–15 hpi), and late (15–24 hpi) stages of reovirus replication (Fig 8F), it was evident that whole virus assembly was not only delayed relative to core amplification under MT disruption, but that fewer whole viruses were ultimately produced. These results suggest that MTs are necessary not only for depositing whole viruses from mid-cytoplasmic core-plus-OC factories into perinuclear compartments (Fig 8G), but also for ensuring the timely and efficient transition of cores into whole viruses.

Cell lines and viruses

BHK-T7 (a generous gift from Ursula Buchholz, NIH), H1299 and L929 cells (purchased from the American Type Culture Collection (ATCC)) were maintained at 37˚C with 5% CO2. BHK-T7 cells were cultured in high-glucose Dulbecco’s modified Eagle medium (DMEM) (D5796, Millipore Sigma) supplemented with 10% fetal bovine serum (FBS) (F1051, Millipore Sigma), 2 mM l-glutamine (G7513, Millipore Sigma), 100 U/ml penicillin, 100 μg/ml streptomycin and 25 ng/ml amphotericin B (A5955, Millipore Sigma). Adherent L929s were cultured in Minimal essential medium (MEM) (M4655, Millipore Sigma) supplemented with 5% FBS, 1 × nonessential amino acids (M7145, Millipore Sigma) and 1mM sodium pyruvate (S8636, Millipore Sigma). H1299 cells were cultured in Roswell Park Memorial Institute medium (RPMI) (R8758, Millipore Sigma) supplemented with 10% FBS. L929s cultured in suspension were grown in Joklik’s modified MEM (JMEM) (pH 7.4) (M0518, Millipore Sigma) supplemented with 2g/L sodium bicarbonate (BP328, Fisher Scientific), 1.2g/L 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (BP310, Fisher Scientific), 1 × non-essential amino acids and 1mM sodium pyruvate. Reovirus serotype type 3 Dearing-PL (T3D^PL; Dr. Patrick Lee, Dalhousie University), and λ2-FLAG- and λ2-V5-tagged viruses were propagated in suspension L929 cells from a seed stock to preserve genetic identity, extracted with Vertrel XF (Dymar Chemicals) three times, and purified by ultracentrifugation on cesium chloride gradients, as previously described [82].

Plaque assays

For reovirus titers, plaque assays were done in L929 cells. Reovirus dilutions in serum-free media were added to confluent monolayers of L929 cells for 1hr with gentle rocking every 10 minutes. After 1 hour, an agar overlay was added (1:2 ratio of 2X JMEM media, 1:4 ratio of complete MEM media, and 1:4 ratio of 2% agar). Overlays were allowed to solidify at room temperature for 30 minutes before incubation at 37˚C/5% CO2 for 4–5 days. Then, 4% paraformaldehyde (PFA; in PBS) was added to the overlay for 30mins at room temperature. PFA was then discarded, and agar overlays were gently removed. Cells were then fixed with methanol for 15 minutes at 4˚C. After discarding the methanol, plaques were visualized by staining with a 1% (wt/vol) crystal violet solution (2.5g crystal violet, 50mL methanol, 197.5mL Milli-Q water) and plaques were manually counted.

Core particle generation, AlexaFluor labelling, and transfections

Immunofluorescent confocal microscopy

Transmission electron microscopy sample preparation

Cells were cultured on Aclar sheets and chemically fixed using a solution of 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M sodium cacodylate buffer containing 2 mM CaCl₂ (pH 7.4) for 20 minutes at 37 °C, followed by an additional 40 minutes at room temperature. After removal of the primary fixative and washing with buffer, samples were post-fixed for 1 hour in 1% osmium tetroxide supplemented with 10% potassium hexacyanoferrate(II). Following thorough washing, cells were treated with sodium acetate (pH 5.2) and stained with 2% aqueous uranyl acetate for 1 hour.

Samples were then dehydrated through a graded ethanol series and embedded in Spurr’s resin. After complete infiltration, the resin was polymerized at 65 °C for 72 hours. The Aclar sheets were subsequently removed, and 1 × 1 mm resin blocks were excised using a fret saw. These blocks were mounted onto epoxy stubs using cyanoacrylate adhesive and trimmed (Trim90, Diatome). Ultrathin sections (~100 nm) were prepared using a diamond knife (Ultra 45, Diatome).

Sections were post-stained with lead citrate and a second round of uranyl acetate staining, then carbon coated with a Leica EM ACE600 prior to imaging. Micrographs were acquired using a Hitachi H-7650 transmission electron microscope operated at 60 kV, equipped with an AMT digital camera and AMT Image Capture Engine software, as well as on a JEOL JEM-2100 operated at 200 kV, equipped with a Gatan Orius camera and DigitalMicrograph software (Gatan).

Cloning and recombinant virus generation

Cloning tagged L2 sequences.

Linear DNA sequences encoding a linker region, restriction enzyme site NcoI, the T3D^PL L2 gene sequence with a fragment removed and either 2xFLAG or 1xV5 sequences added in place, restriction enzyme site PpuM1, and another linker region were synthesized by Invitrogen (Genestrings). The DNA sequences are listed in Table 2. The fragments were amplified by PCR, products were purified, and then digested with NcoI (FD0575, Thermo) and PpuM1 (FD0764, Thermo) restriction enzymes. The plasmid pBacT7-L2 was similarly digested and gel purified. The fragments were then ligated into the pBacT7-L2 plasmid using T4 ligase (15224–017, Invitrogen). Purified plasmids were sequenced via Sangar Sequencing at The Molecular Biology Service Unit (MBSU) at the University of Alberta.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. FLAG and V5 L2 DNA sequences.

https://doi.org/10.1371/journal.ppat.1013238.t002

SDS-PAGE and Western blot

For standard Western blot analysis, cells were first washed with 1xPBS before lysing with RIPA buffer (50mM Tris pH 7.4, 150mM NaCl, 1% NP-40, 0.5% sodium deoxycholate) supplemented with protease inhibitor cocktail (P8340, Sigma). After collection, 5x protein sample buffer (250mM Tris pH 6.8, 5% SDS, 45% glycerol, 9% β-mercaptoethanol, and 0.01% bromophenol blue) was added to a final concentration of 1x. Samples were then boiled at 100˚C for 10 minutes. Samples were electrophoresed on 10% SDS-acrylamide gels, and proteins were transferred from the gel onto nitrocellulose membranes using a Trans-Blot Turbo Transfer System (Bio-Rad). Membranes were incubated with block buffer (3% newborn calf serum (NBCS) in Tris-buffered saline (.05 M Tris and 0.15 M sodium chloride, pH 7.4 with 0.1% TritonX-100 (TBST)) for 1 hour at room temperature, followed by incubation with a primary antibody solution (block buffer containing: rabbit polyclonal antibodies against whole virus (α-Reo) at 1:10,000 (Biologics International Corp), rabbit polyclonal antibodies against core (α-core) at 1:1000 (Biologics International Corp), rabbit polyclonal antibodies against λ2 (α-λ2), monoclonal mouse anti-FLAG M2 (Sigma F1804), and/or polyclonal rabbit anti-V5 (Novus NB600–381)) for either 1 hour at room temperature or overnight at 4˚C followed by a secondary antibody solution (block buffer containing: goat α-rabbit AF 488 or 647 at 1:1000, goat α-rabbit HRP 1:10,000, goat α-mouse AF 488 or 647 at 1:1000, and/or goat α-mouse HRP 1:10,000) for 1 hour at room temperature. Membranes were washed 3x for 5 minutes each in TBST between primary and secondary antibody incubations. Prior to imaging, membranes with HRP-conjugated secondaries were incubated with ECL Plus Western Blotting Substrate (32132, Thermo Fisher Scientific) for 5 minutes at room temperature. Membrane imaging was carried out using a Biorad’s Chemidoc Imaging System. Densitometric analysis was performed with ImageJ software, and images were processed for display in Adobe Photoshop. All incubation steps for membranes were done with gentle rocking.

Binding assays and flow cytometry

To assess reovirus binding, 100% confluent L929s were prechilled for 1hr at 4°C and then infected with serial dilutions of virus at 4°C for 1hr with gentle agitation every 10mins. Three washes with cold PBS were done to remove any unbound virus, and then cells were detached using Cellstripper (25–056-CI Corning). Cells were then fixed with 4% PFA for 30 min at 4°C. Cells were incubated with monoclonal mouse anti-σ3 (1:100, 10C1 or 10G10, DSHB) antibody for 1hr at 4°C, then washed three times with flow buffer (2%FBS, 1mM EDTA in PBS). After each wash, cells were pelleted at 500xg for 5mins at 4°C. Then, cells were incubated with AF-conjugated goat anti-mouse 647 secondary antibody (1:1000) for 30 mins in the dark at room temperature. Samples were analyzed using BD LSR Fortessa (BD Biosciences), and a minimum of 10,000 cells were collected for each sample. Data was analyzed using FlowJo v10 Software (BD Biosciences)

Nocodazole and cycloheximide treatments

80-90% confluent H1299 cells were infected as indicated in specific experiments. In short, cell media was removed, cells were washed once with PBS, and virus diluted in serum-free media was added to the cells for 1 hour with gentle rocking every 10 minutes. One hour later, virus media was removed, and complete media containing 10μM nocodazole, or an equivalent volume of DMSO, was added to the cells. Cells were then incubated until desired time points were reached. Where indicated, cycloheximide (100 μg/mL) was added to complete media containing nocodazole or DMSO as well to inhibit translation and thus synthesis of de-novo viral particles.

DsiRNA-mediated knockdowns

For DsiRNA knock-down studies, cells were transfected with DsiRNAs (Integrated DNA Technology) directed against either a non-targeting control (G7L, vaccinia virus gene, Antisense Sequence: rUrUrUrArUrUrUrGrArUrGrArArUrCrUrArGrUrUrGrGrUrUrCrUrC Sense Sequence: rGrArArCrCrArArCrUrArGrArUrUrCrArUrCrArArArUrAAA), or M2 (reovirus μ1 protein, Antisense Sequence: rGrArCrCrCrUrGrArGrArUrGrArArUrUrArUrUrArUrCrUTT Sense Sequence: rArArArGrArUrArArUrArArUrUrCrArUrCrUrCrArGrGrGrUrCrArG). Cells were plated and transfected simultaneously by collecting cells via trypsinization and diluting to 2x10⁵ cells/mL and mixing with a pre-incubated cocktail of 1:2 volume Lipofectamine 2000 (11668–019, Invitrogen) to 2mM DsiRNA in Opti-MEM (31985–070, Gibco) prior to plating 1mL/well per 4 cm2. Cells were allowed to adhere for at least 6 hours prior to infection.

RT-qPCR

Cells were lysed in TRI Reagent (T9424, Millipore Sigma) and the aqueous phase was separated following chloroform extraction as per the TRI Reagent protocol. Isopropanol was mixed with the aqueous phase and RNA was isolated as per the Monarch Total RNA Miniprep kit (T2010S, New England Biolabs) protocol. RNA was eluted using RNAse free water and total RNA was quantified using a NanoDrop spectrophotometer. Using 200ng RNA per 12μl reaction, cDNA synthesis was performed with random primers (48190011, ThermoFisher Scientific) and M-MLV reverse transcriptase (28025013, ThermoFisher Scientific) as per the manufacturers protocol. Following a ¼ cDNA dilution, RT-PCR reactions were executed following the SYBR Select (4472920, Invitrogen) protocol using reovirus gene-specific primers (listed in Table 3) and the CFX96 system (Bio-Rad). All RT-qPCR reaction plates included a no template and no reverse transcription control.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Gene specific primers used in RT-qPCR reactions.

https://doi.org/10.1371/journal.ppat.1013238.t003