Discrete Virus Factories Form in the Cytoplasm of Cells Coinfected with Two Replication-Competent Tagged Reporter Birnaviruses That Subsequently Coalesce over Time

Abstract

The Birnaviridae family, responsible for major economic losses to poultry and aquaculture, is composed of nonenveloped viruses with a segmented double-stranded RNA (dsRNA) genome that replicate in discrete cytoplasmic virus factories (VFs). Reassortment is common; however, the underlying mechanism remains unknown given that VFs may act as a barrier to genome mixing. In order to provide new information on VF trafficking during dsRNA virus coinfection, we rescued two recombinant infectious bursal disease viruses (IBDVs) of strain PBG98 containing either a split GFP11 or a tetracysteine (TC) tag fused to the VP1 polymerase (PBG98-VP1-GFP11 and PBG98-VP1-TC). DF-1 cells transfected with GFP1-10 prior to PBG98-VP1-GFP11 infection or stained with a biarsenical derivative of the red fluorophore resorufin (ReAsH) following PBG98-VP1-TC infection, had green or red foci in the cytoplasm, respectively, that colocalized with VP3 and dsRNA, consistent with VFs. The average number of VFs decreased from a mean of 60 to 5 per cell between 10 and 24 h postinfection (hpi) (P < 0.0001), while the average area increased from 1.24 to 45.01 μm² (P < 0.0001), and live cell imaging revealed that the VFs were highly dynamic structures that coalesced in the cytoplasm. Small VFs moved faster than large (average 0.57 μm/s at 16 hpi compared to 0.22 μm/s at 22 hpi), and VF coalescence was dependent on an intact microtubule network and actin cytoskeleton. During coinfection with PBG98-VP1-GFP11 and PBG98-VP1-TC viruses, discrete VFs initially formed from each input virus that subsequently coalesced 10 to 16 hpi, and we speculate that Birnaviridae reassortment requires VF coalescence.IMPORTANCE Reassortment is common in viruses with segmented double-stranded RNA (dsRNA) genomes. However, these viruses typically replicate within discrete cytoplasmic virus factories (VFs) that may represent a barrier to genome mixing. We generated the first replication competent tagged reporter birnaviruses, infectious bursal disease viruses (IBDVs) containing a split GFP11 or tetracysteine (TC) tag and used the viruses to track the location and movement of IBDV VFs, in order to better understand the intracellular dynamics of VFs during a coinfection. Discrete VFs initially formed from each virus that subsequently coalesced from 10 h postinfection. We hypothesize that VF coalescence is required for the reassortment of the Birnaviridae This study provides new information that adds to our understanding of dsRNA virus VF trafficking.

Keywords: IBDV; birnavirus; coinfection; double-stranded RNA virus; reassortment; viroplasm; virus factory.

FIG 1

Construction of tagged reporter IBDV strains. The PBG98 sequence was used as a backbone, and the nucleotide sequences encoding either the GFP11 tag (A) or TC tag (B) were added to the 3′ end of the coding region of segment B. (C) To obtain a positive GFP signal, DF-1 cells were transfected with a plasmid expressing GFP1-10 prior to transfection with a plasmid expressing GFP11 or infection with the PBG98-VP1-GFP11 virus. (D) To visualize the TC tag, DF-1 cells infected with the PBG98-VP1-TC virus were stained with ReAsH. The replication of both tagged viruses in DF-1 cells was compared to the recombinant wild-type (wt) PBG98 virus. The titer to which the GFP11-tagged virus (E) or TC-tagged virus (F) replicated was compared to the wt PBG98 virus at the indicated time points postinfection in triplicate. Virus titers were expressed as the log₁₀ TCID₅₀/ml, individual replicates were plotted as a scatterplot, and the means were plotted as a bar graph with error bars representing the standard deviations (SD) of the mean. The data are representative of three independent experiments. ns, not significant; *, P < 0.05; ****, P < 0.0001.

FIG 2

The PBG98-VP1-GFP signal is a marker for IBDV VFs. DF-1 cells were either infected with the recombinant wild-type (wt) PBG98 virus (A and C) or transfected with a plasmid expressing GFP1-10 and infected 24 h posttransfection with the PBG98-VP1-GFP11 virus (B and D) at an MOI of 1. Cells were fixed at 20 hpi and stained with DAPI and either anti-dsRNA (A and B) or anti-VP3 (C and D) mouse monoclonal antibodies, followed by a goat anti-mouse secondary antibody conjugated to Alexa Flour 568, and then imaged. Scale bars, 20 μm.

FIG 3

RNA from both segments A and B are located within the VFs. DF-1 cells were transfected with a plasmid expressing GFP1-10 and infected 24 h posttransfection with the PBG98-VP1-GFP11 virus at an MOI of 1. Cells were fixed at 24 hpi, and 48 Stellaris FISH probes conjugated to Quasar 570 dye were hybridized to RNA from segment A (A), 48 Stellaris FISH probes conjugated to Quasar 670 dye were hybridized to RNA from segment B (B), or a cocktail of the 96 probes were cohybridized (C). Scale bars, 20 μm.

FIG 4

VFs decrease in number and increase in size over time. DF-1 cells were either transfected with GFP1-10 and infected with the PBG98-VP1-GFP11 virus (A) or infected with the wt PBG98 virus (B) an MOI of 1 and fixed at 10, 18, and 24 hpi, stained with DAPI, and imaged. The wt PBG98-infected cells were also stained with mouse anti-VP3 and goat anti-mouse-568 for visualization of VFs. Scale bars, 20 μm. The number of VFs per cell was determined for 30 infected cells at each time point and plotted for cells infected with the PBG98-VP1-GFP11 virus (C) and the wt PBG98 virus (E). The average area of the VFs in an infected cell was determined using the surface tool in Imaris 9 software (bitplane) and plotted for 30 infected cells at each time point for cells infected with the PBG98-VP1-GFP11 virus (D) and the wtPBG98 virus (F). The line represents the mean, and the error bars represent the SD of the mean. The numbers and sizes of foci were compared using Kruskal-Wallis one-way ANOVA. ***, P < 0.001; ****, P < 0.0001.

FIG 5

VFs coalesce in the cytoplasm of infected cells throughout infection. DF-1 cells were transfected with a plasmid expressing GFP1-10 and infected 24 h posttransfection with the PBG98-VP1-GFP11 virus at an MOI of 1. (A) One live infected cell was imaged every 4 min from 16 to 18 hpi. Images of the boxed region are shown at the indicated time points postinfection, where three VF coalescence events were witnessed (white, blue, and red arrows). (B) Another live infected cell was imaged from 22 to 25 hpi. Images of the boxed region are shown at the indicated time points postinfection where one VF fusion event was witnessed (white arrows) and one VF fission event was witnessed (orange arrows).

FIG 6

The distribution of VFs in the cytoplasm is dependent on an intact actin cytoskeleton and microtubule network, but alterations in VF distribution do not alter viral replication. DF-1 cells were infected with the PBG98-VP1-TC virus at an MOI of 1. Cells were fixed at 20 hpi and stained with DAPI and either an anti-tubulin mouse monoclonal antibody, followed by a goat anti-mouse secondary antibody conjugated to Alexa Flour 488 to visualize the microtubule network (A) or phalloidin conjugated to Alexa Flour 488 to visualize the actin cytoskeleton (B). Scale bars, 20 μm. Images of the boxed regions are shown enlarged, and the locations where the cytoskeleton colocalized with the VFs are indicated by white arrows. (C) The number of VFs per cell was determined for 30 infected cells and plotted. (D) The average area of the VFs in an infected cell was determined using the surface tool in Imaris 9 software (bitplane) and plotted for 30 infected cells. The line represents the mean, and the error bars represent the SD of the mean. (E) Cell supernatants obtained 24 hpi were titrated, and the TCID₅₀ values were determined. Virus titers were compared by one-way ANOVA and a Tukey’s multiple-comparison test following a Shapiro-Wilk normality test to confirm whether the data followed a normal distribution for parametric or nonparametric testing. The mean area and number of VFs were compared using Kruskal-Wallis one-way ANOVA. ns, not significant; ***, P < 0.001; ****, P < 0.0001.

FIG 7

VF coalescence is dependent on an intact microtubule network and actin cytoskeleton. DF-1 cells were transfected with a plasmid expressing GFP1-10 and infected 24 h posttransfection with the PBG98-VP1-GFP11 virus at an MOI of 1. Cells were treated with either nocodazole (A) or cytochalasin-D (B) from 2 hpi. One live infected cell was imaged every 4 min from 20 to 22 hpi. Images of the boxed region are shown at the indicated time points postinfection. (A) In an infected cell treated with nocodazole, VFs were witnessed transiently interacting (white arrows). (B) In an infected cell treated with cytochalasin-D, four VFs were observed (white, blue, red, and purple arrows) that did not coalescence or interact with each other over the 2-h period. Scale bars, 20 μm.

FIG 8

Discrete VFs from different strains of IBDV form in the cytoplasm of coinfected cells and coalesce throughout infection. DF-1 cells were transfected with a plasmid expressing GFP1-10 and coinfected 24 h posttransfection with the PBG98-VP1-GFP11 virus and the PBG98-VP1-TC virus at an MOI of 1. (A) Cells were fixed at 10, 14, 18, and 24 hpi and stained with ReAsH and DAPI. Scale bars, 20 μm. (B) The Mander’s correlation coefficient between red and green foci was plotted for 10 cells at 10, 18, and 24 hpi. The line represents the mean, and error bars represent the SD of the mean. Significance was determined by one-way ANOVA. ****, P < 0.0001. (C) The percentages of red, green, and colocalized (yellow) foci were quantified for 30 coinfected cells per time point, and the average was plotted at 10, 12, 14, and 16 hpi in the presence of nocodazole (noc), cytochalasin-D (cyto-D), or DMSO alone (mock).