(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Copy-back viral genomes induce a cellular stress response that interferes with viral protein expression without affecting antiviral immunity [1]

['Lavinia J. González Aparicio', 'Department Of Molecular Microbiology', 'Center For Women Infectious Disease Research', 'Washington University School Of Medicine In St. Louis', 'Missouri', 'United States Of America', 'Yanling Yang', 'Matthew Hackbart', 'Carolina B. López']

Date: 2023-11

Antiviral responses are often accompanied by translation inhibition and formation of stress granules (SGs) in infected cells. However, the triggers for these processes and their role during infection remain subjects of active investigation. Copy-back viral genomes (cbVGs) are the primary inducers of the mitochondrial antiviral signaling (MAVS) pathway and antiviral immunity during Sendai virus (SeV) and respiratory syncytial virus (RSV) infections. The relationship between cbVGs and cellular stress during viral infections is unknown. Here, we show that SG form during infections containing high levels of cbVGs, and not during infections with low levels of cbVGs. Moreover, using RNA fluorescent in situ hybridization to differentiate accumulation of standard viral genomes from cbVGs at a single-cell level during infection, we show that SGs form exclusively in cells that accumulate high levels of cbVGs. PKR activation is increased during high cbVG infections and, as expected, is necessary for virus-induced SG. However, SGs form independent of MAVS signaling, demonstrating that cbVGs induce antiviral immunity and SG formation through 2 independent mechanisms. Furthermore, we show that translation inhibition and SG formation do not affect the overall expression of interferon and interferon stimulated genes during infection, making the stress response dispensable for global antiviral immunity. Using live-cell imaging, we show that SG formation is highly dynamic and correlates with a drastic reduction of viral protein expression even in cells infected for several days. Through analysis of active protein translation at a single-cell level, we show that infected cells that form SG show inhibition of protein translation. Together, our data reveal a new cbVG-driven mechanism of viral interference where cbVGs induce PKR-mediated translation inhibition and SG formation, leading to a reduction in viral protein expression without altering overall antiviral immunity.

Funding: Financial support during preparation of this work was provided by the US National Institutes of Health National Institute of Allergy and Infections Diseases AI137062 and AI134862 (to CBL), and the Principles of Pulmonary Research Training Grant T32-007317 (to LGA and MH). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2023 González Aparicio et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

The predicted overlapping antiviral roles of the PKR-driven cellular stress response and cbVGs led us to question if cbVGs are involved in SG formation during RSV and parainfluenza virus infections, and whether cbVG-mediated antiviral immunity depends on SG formation. Our data show that cbVGs are the primary inducers of canonical SG during Sendai virus (SeV) and RSV infections through PKR activation and that this induction is independent of the MAVS pathway. Contrary to previous reports, we found that MAVS does not localize to cbVG-induced SG and that translation inhibition and SG formation are not required for overall induction of antiviral immunity. Instead, we show that cbVGs induce protein translation inhibition in SG-positive cells, resulting in reduced levels of virus proteins at a single-cell level without affecting the expression of antiviral proteins at a population level. Overall, these data demonstrate that cbVGs orchestrate the induction of cellular stress and antiviral immunity independently, highlighting the importance of considering the presence of nsVGs when studying virus–host interactions. Importantly, our data uncover a new primary mechanism of interference by cbVGs via the induction of viral protein translational arrest.

In addition to the antiviral immune response, virus infections can induce cellular stress responses that lead to protein translation inhibition and the formation of stress granules (SGs) [ 10 ]. During most viral infections, the cellular stress response is initiated upon activation of the double-stranded RNA binding protein PKR, which phosphorylates the eukaryotic initiation factor 2 alpha (eIF2α), leading to cap-dependent translation arrest, disassembly of polysomes, and formation of SG. SGs are liquid phase-separated nonmembranous organelles composed mostly of untranslated mRNA and RNA binding proteins [ 11 , 12 ]. Activation of this cellular stress response during infection can lead to reduced viral protein expression [ 12 ] and has been proposed to mediate the antiviral immune response [ 13 – 17 ]. Accumulation of cbVGs during measles virus infection has been correlated with PKR activation, but whether cbVGs are the main triggers of PKR activation and SG formation is unknown [ 18 ].

One nsVG subpopulation, copy-back viral genomes (cbVGs), has critical roles in inducing the cellular antiviral immune response, controlling the rate of viral replication, and promoting the establishment of persistent infections [ 6 – 8 ]. Nonsegmented negative-sense RNA viruses generate cbVGs when the viral polymerase initiates replication at the promoter region, falls off the template, and then reattaches to the nascent strand [ 3 ]. The polymerase then uses the nascent strand as a template and continues replicating, copying back the already synthetized RNA ( S1A Fig ) [ 3 ]. The resulting RNA molecules contain highly structured immunostimulatory motifs and lack genes encoding viral proteins [ 7 , 9 ]. Although cbVGs can only replicate in the presence of a full-length standard genome that provides essential viral proteins, cbVGs are key interactors with the host and drive several cellular responses that determine the infection outcome. Notably, all the known effects of cbVGs on shaping the host response are dependent on the mitochondrial antiviral signaling (MAVS) pathway. cbVGs activate retinoic acid–inducible gene I (RIG-I)-like receptors (RLRs) leading to MAVS signaling, which then induces robust antiviral responses [ 9 ]. By activating the MAVS pathway, cbVGs stimulate the interferon (IFN) response that ultimately reduces virus spread and induces long-term protective immunity [ 6 ]. Additionally, cbVGs signal through MAVS to activate a cell survival mechanism that promotes the establishment of persistent infections in vitro [ 8 ]. Whether cbVGs can induce other cellular pathways that contribute to the outcome of the infection remains unknown.

Respiratory syncytial virus (RSV) and the parainfluenza viruses are endemic RNA viruses responsible for a large disease burden, especially involving children and older adults [ 1 , 2 ]. RNA viruses produce not only full-length standard viral genomes (stVGs) but also variants, hypermutated RNAs, and nonstandard viral genomes (nsVGs) that provide different functions and advantages to the virus [ 3 , 4 ]. nsVGs produced during RSV and parainfluenza virus infections are critical determinants of infection outcome in vitro and in vivo [ 5 – 7 ]. When produced early during infection, nsVGs significantly reduce virus spread and disease severity in mice and humans [ 5 , 7 ]. nsVGs impact the infection via stimulation of major signaling pathways that shape the cellular response to the infection. Identifying cellular pathways and molecular mechanisms by which nsVGs reduce virulence may lead to new strategies to prevent severe disease upon RNA virus infection.

Results

SGs form during RSV infection containing high levels of cbVGs To assess whether cbVGs induced SG formation, we infected lung epithelial A549 cells with RSV stocks containing high or low levels of cbVGs (RSV cbVG-high and RSV cbVG-low, respectively). To achieve high and low cbVG accumulation in these stocks, the virus was grown at different multiplicity of infection (MOI), as virus expansion at high MOI promotes the accumulation of cbVG, while virus expansion at low MOI reduces the accumulation of cbVGs [19]. cbVG contents in the stocks were confirmed by PCR (S1B Fig). Because cbVGs potently induce the IFN response, we expect cbVG-high stocks to induce higher expression of IL-29 than a cbVG-low stock [6]. As expected, IL-29 mRNA levels were increased in cells infected with cbVG-high stocks (S1C Fig). Additionally, presence of cbVGs during infection is expected to correlate with reduced levels of virus replication in infected cells as compared to cbVG-low stocks due to the activity of IFNs [6]. Using RSV G mRNA transcripts as a proxy for virus replication, we confirmed that infection with RSV cbVG-high stocks resulted in reduced levels of RSV G mRNA as compared to infection with an RSV cbVG-low stocks (S1B Fig). To visualize SG formation during RSV infections, cells were immunostained for the well-characterized SG associated protein Ras GTPase-activating protein-binding protein 1 (G3BP1), along with the RSV nucleoprotein (NP) to identify infected cells. Fluorescence imaging analysis showed SG in infected cells during RSV cbVG-high infections, while they were rarely detected in RSV cbVG-low infections. SGs were observed as early as 12 hours postinfection (hpi) and were still present at 24 hpi (Fig 1A). The percent of SG-positive cells during RSV cbVG-high infection increased over time, and approximately 10% of infected cells were SG positive at 24 hpi (Fig 1B). Of note, throughout our study, all SG-positive cells were positive for viral protein. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. SGs form during RSV infection containing high levels of cbVGs. (A) SG (G3BP1, magenta) and viral protein (RSV NP, yellow) detection 12 and 24 hpi with RSV cbVG-high or cbVG-low at MOI of 1.5 TCID 50 /cell. (B) Percent of SG-positive cells within the infected population 12 and 24 hpi with RSV cbVG-high and cbVG-low infections. Approximately 150 infected cells were counted per condition (average of 3 independent experiments shown). (C) SG (G3BP1, magenta) and viral protein (RSV NP, yellow) detection 24 hpi with RSV cbVG-high or cbVG-low at MOIs 0.1, 1.5, 5, and 10 TCID 50 /cell. (D) Percent of SG-positive cells within the infected population 24 hpi with RSV cbVG-high and cbVG-low infection at MOIs 0.1, 1.5, 5, and 10 TCID 50 /cell. Approximately 150 infected cells were counted per condition (average of 3 independent experiments shown). (E) SG detection (TIAR, white) in cells staining via FISH for stVG-high (orange) and cbVG (green) cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell. (F) Percent of SG-positive cells within the stVG-high and cbVG-high cell populations during RSV cbVG-high infection (average of 3 independent experiments shown). All widefield images were acquired with the Apotome 2.0 at 63× magnification and are representative of 3 independent experiments. Scale bar = 50 μm. Statistical analysis: one-way ANOVA (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.00001). Numerical data plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; FISH, fluorescence in situ hybridization; G3BP1, GTPase-activating protein-binding protein 1; hpi, hours postinfection; MOI, multiplicity of infection; NP, nucleoprotein; RSV, respiratory syncytial virus; SG, stress granule; stVG, standard viral genome; TIAR, TIA-1-related. https://doi.org/10.1371/journal.pbio.3002381.g001 Although cbVG-containing viral particles can infect cells, they are not considered fully infectious as they can only replicate in cells coinfected with standard virus particles. Thus, infections based on MOI only account for the number of fully infectious particles in the inoculum. We expect that RSV cbVG-high infections, which contain both infectious standard particles and noninfectious cbVG particles, will contain a higher amount of total viral particles. To determine if the observed differences in SG formation were due to differences in total viral particles added in the inoculum, we infected cells with RSV cbVG-high and RSV cbVG-low at increasing MOIs and compared percent of SG-positive cells. Increasing the MOI of RSV cbVG-low infection did not increase the percent of SG-positive cells even when using 10 times more RSV cbVG-low than RSV cbVG-high (Fig 1C and 1D). We observed an increase in the percent of SG-positive cells as we increased the MOI during RSV cbVG-high infection, which correlates with the increased number of cbVG-containing particles in the inoculum. However, no differences in percent of SG-positive cells were observed between MOI 5 and MOI 10 (Fig 1D), suggesting that there is a threshold on the amount of SG-positive cells we can obtain at a given time during the infection. Taken together, these data indicate that presence of cbVGs during RSV infection correlates with SG formation.

SGs form exclusively in cbVG-high cells during RSV cbVG-high infection Using a previously described RNA fluorescence in situ hybridization (FISH)-based assay that allows differentiation of full-length genomes from cbVGs at a single-cell level [8], our lab reported that cells infected with RSV or SeV cbVG-high stocks have heterogenous accumulation of viral genomes; some cells accumulate high levels of standard genomes (stVG-high), and others accumulate high levels of cbVGs (cbVG-high) [8,20,21]. To determine if SG formed differentially within these 2 populations of cells, we combined RNA FISH with immunofluorescence to detect SG during RSV cbVG-high infection. At 24 hpi, SG formed almost exclusively in cbVG-high cells (green) and not stVG-high cells (orange) (Fig 1E). Interestingly, only around 30% of the cbVG-high cells had SG (Fig 1F). This could suggest that a threshold of cbVG accumulation in the cells is needed for SG formation or that SG formation occurs asynchronously during infection, which is observed during HCV infection [22]. Nevertheless, these data demonstrate that cbVGs trigger SG formation.

cbVGs induce SG during SeV infection To determine whether cbVG induction of SG also occurs during infection with parainfluenza viruses, we infected cells with cbVG-high or cbVG-low SeV, a member of the paramyxovirus family and close relative to the human parainfluenza virus 1. Like infection with RSV, SG formed predominantly during SeV cbVG-high infections (Fig 2A) where approximately 20% of the infected cells were positive for SG at 24 hpi (Fig 2B). Compared to cells with undetected SGs or NP (Fig 2A, right panel inset 1), some SG-positive cells had notably low NP signal (Fig 2A, right panel inset 2), while other SG-positive cells showed high NP signal (Fig 2A, right panel inset 3). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. SeV cbVGs induce SG formation. (A) SG (G3BP1, white) and viral protein (SeV NP) detection 24 hpi with SeV cbVG-low and cbVG-high (NP, yellow) at MOI 1.5 TCID 50 /cell. Digital zoomed images for each of the marked cells are shown in the panel on the right. (B) Percent of infected SG-positive cells 24 hpi with SeV cbVG-low and cbVG-high at MOI 1.5 TCID 50 /cell. Approximately 150 infected cells were counted per condition (average of 3 independent experiments shown). (C) SG (G3BP1, white) and viral protein (SeV NP) detection 24 hpi at MOI 1.5 TCID 50 /cell supplemented with either purified cbVG particles or UV-inactivated cbVG particles at increasing HAUs. (D) Percent of SG-positive cells at increasing HAU doses of active/UV-inactive cbVG particles. Approximately 200 infected cells were counted per condition (average of 3 independent experiments shown). All widefield images were acquired with the Apotome 2.0 at 63× magnification and are representative of 3 independent experiments. Scale bar = 50 μm. Statistical analysis: one-way ANOVA (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.00001). Numerical data plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; G3BP1, GTPase-activating protein-binding protein 1; HAU, hemagglutination unit; hpi, hours postinfection; MOI, multiplicity of infection; NP, nucleoprotein; SeV, Sendai virus; SG, stress granule. https://doi.org/10.1371/journal.pbio.3002381.g002 To further establish the role of cbVGs in inducing SG, we performed a dose-dependent experiment using purified cbVG-containing viral particles. We infected cells with SeV cbVG-low and supplemented the infection with increasing doses of purified cbVG-containing particles. The percent of SG-positive cells increased in proportion to the amount of purified cbVG-containing particles added (Fig 2C, upper panel, and 2D). SG were not observed, however, when we added the same amounts of UV-inactivated purified cbVG particles (Fig 2C, lower panel, and 2D). These data demonstrate that only replication-competent cbVGs induce SG formation during RNA virus infection.

Canonical SGs are formed during cbVG-high infection Some viruses can induce formation of SG-like granules that differ compositionally from canonical SG and can relocalize SG components to viral replication centers [23–25]. Other viruses induce formation of RNAseL-dependent bodies (RLBs), which contain common proteins also found in SG but are structurally and functionally distinct from SG [26]. To better characterize the granules observed during RSV cbVG-high infection, we began by testing if cbVG-dependent granules require polysome disassembly, a crucial step for proteins to bind ribosome-free mRNA and form canonical SG. For this, we treated RSV cbVG-high infected cells with cycloheximide (CHX), which inhibits canonical SG by preventing polysome disassembly [27]. Sodium arsenite, a chemical known to induce canonical SG, was used as a positive control [28]. Treatment with CHX during RSV cbVG-high infection led to a decrease in SG-positive cells compared to treatment with the drug’s vehicle alone (DMSO) (Fig 3A and 3B). To rule out any effect the drugs could have on G3BP1 localization, we costained with another SG marker, TIA-1-related (TIAR) protein. Costaining with TIAR showed colocalization with G3BP1 in SG in the DMSO-treated cells and disassembly from granules in the drug-treated conditions (Fig 3A), demonstrating that cbVG-dependent SGs are canonical SGs. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. RNA granules formed during RSV cbVG-high infection are canonical SGs. (A) G3BP1 (red) and TIAR (green) staining for SG in cells treated with SA (0.5 mM) for 1 h or infected with RSV cbVG-high (RSV NP, white) at MOI 1.5 TCID 50 /cell 23 hpi and treated with DMSO or CHX (10 μg/mL) for 1 h. (B) Quantification of SG-positive cells after drug treatment in SA or RSV cbVG-high infected cells. Approximately 150 cells were counted for each condition. Fold change relative to DMSO-treated cells is shown. (C) RNA granule detection (G3BP1, red; TIAR, green) in A549 control and RNAseL KO cells transfected with poly I:C 10 μg/mL or infected with RSV cbVG-high (RSV NP, white) 24 hpi at MOI 1.5 TCID 50 /cell. (D) RNA granule detection (G3BP1, red; and TIAR, green) A549 cells transfected with poly I:C or RSV and SeV cbVG derived oligonucleotides RSV 238 and SeV 268. All widefield images were acquired with the Apotome 2.0 at 63× or 40× magnification. Scale bar = 50 μm. Numerical values plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; CHX, cycloheximide; G3BP1, GTPase-activating protein-binding protein 1; hpi, hours postinfection; KO, knockout; MOI, multiplicity of infection; NP, nucleoprotein; RSV, respiratory syncytial virus; SA, sodium arsenite; SeV, Sendai virus; SG, stress granule; TIAR, TIA-1-related. https://doi.org/10.1371/journal.pbio.3002381.g003 We next tested whether RSV-induced granules were RLBs [29]. To do this, we infected RNAseL knockout (KO) cells with RSV cbVG-high virus and looked at differences in SG formation comparing to poly I:C transfection, which is known to induce RLB formation [29]. Structurally, RLBs are smaller, more punctate, and contain less TIAR than canonical SG (Fig 3C, left panel). RNAseL activation prevents canonical SG from forming by degrading free mRNA necessary for SG to form and only when knocking out RNAseL can canonical SGs form upon stimulation [29,30]. SGs are structurally bigger and less uniform than RLBs. SGs formed during RSV cbVG-high infection even in RNAseL KO cells, and the structure of these granules was unchanged between cell lines, demonstrating that RSV-dependent SGs are not RLBs (Fig 3C). We then investigated if, out of the context of an infection, cbVG RNA would still induce formation of canonical SG or would induce RLBs similar to poly I:C. We transfected in vitro transcribed RSV and SeV cbVG-derived oligonucleotides that maintain the key stimulatory domains of cbVGs (RSV 238 and SeV 268 [9]) into A549 cells and compared to poly I:C-induced RLBs. We saw no differences in RNA granule formation and G3BP1 and TIAR contents between poly I:C RLBs and the granules observed with transfected cbVG-derived oligonucleotides (Fig 3D), indicating that cbVGs induce canonical SG only in the context of SeV or RSV infection while RLBs are produced in response to naked cbVG RNA.

cbVG-dependent SGs are PKR dependent and MAVS independent To better understand the molecular mechanisms leading to SG formation in response to cbVGs during infection, we investigated the role of major dsRNA sensors in SG induction. SG formation during infection with many viruses, including RSV, depends on PKR activation [11]. To confirm that cbVGs induce PKR activation, we probed for PKR activation during RSV cbVG-high infection. As expected, PKR phosphorylation is increased during RSV cbVG-high infections compared to RSV cbVG-low or mock infection (Fig 4A). Because PKR is an IFN-stimulated gene (ISG) and cbVGs strongly induce the IFN response, higher levels of unphosphorylated PKR are expected during cbVG-high infection (Fig 4A, middle blot). To determine if cbVG-induced SGs are PKR dependent, we infected A549 PKR KO cells (Fig 4B, middle lane) and visualized SG formation. Consistent with the literature, PKR KO cells infected with RSV cbVG-high virus did not show SG-positive cells (Fig 4D and 4E, middle panel and bar). RSV G mRNA levels were similar between cell types, confirming that inhibition of SG in PKR KO cells was not due to lower replication of the virus (Fig 4C, middle bar). Together, these data suggest that the SG observed during RSV cbVG-high infection are PKR dependent and that cbVG induction of SG is mediated through PKR activation. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. cbVG-dependent SGs are PKR dependent and MAVS independent. (A) Phosphorylation of PKR 24 hpi with RSV cbVG-low and cbVG-high infection at MOI 1.5 TCID 50 /cell. p-PKR inverted mean intensity relative to α-tubulin is shown. Western blot images shown are representative of 3 independent experiments. Statistical analysis: one way ANOVA (*p < 0.05). (B) Western blot analysis showing efficient KO of PKR and MAVS in A549 cells. (C) Expression of RSV G gene mRNA relative to the HKI 24 hpi with RSV cbVG-high infection at MOI 1.5 TCID50/cell in control, MAVS, or PKR KO A549 cells (average of 3 independent experiments are shown). (D) SG (G3BP1, magenta) and viral protein (RSV NP) detection in PKR KO and MAVS KO A549 cells 24 hpi with RSV cbVG-high virus at MOI 1.5 TCID 50 /cell. (E) Quantification of SG-positive cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell in PKR or MAVS KO A549 cells. Approximately 300 cells were counted per condition. All widefield images were acquired with the Apotome 2.0 at 63× magnification and are representative of 3 independent experiments. Scale bar = 50 μm. Statistical analysis: one-way ANOVA (*p < 0.05, **p < 0.01). Numerical values plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; G3BP1, GTPase-activating protein-binding protein 1; HKI, housekeeping index; hpi, hours postinfection; KO, knockout; MAVS, mitochondrial antiviral signaling; MOI, multiplicity of infection; NP, nucleoprotein; RSV, respiratory syncytial virus; SG, stress granule. https://doi.org/10.1371/journal.pbio.3002381.g004 Because cbVGs exert most of their functions through RLR stimulation, which leads to MAVS activation and enhanced production of IFN, we sought to investigate whether cbVGs also induced SG through MAVS signaling. To our surprise, MAVS KO cells (Fig 4B, right lane) infected with RSV cbVG-high virus showed SG-positive cells (Fig 4D and 4E, right panel and bar). The percent of SG-positive cells trended slightly lower than control but was not statistically significant (Fig 4E). This is most likely due to a reduced expression of PKR, a well-known ISG. Contrary to reports in the literature, we did not observe localization of MAVS in SG (S2A Fig), nor recruitment of RIG-I to SG during SeV cbVG-high infection (S2B Fig). These data indicate that cbVGs induce SG independent of cbVGs immunostimulatory activity. To our knowledge, this is the first time cbVGs have shown to modulate cellular processes that are independent of MAVS signaling.

cbVG-dependent SG inhibition is both G3BP1 and G3BP2 dependent To form SG, nucleating factors initiate RNA protein aggregation and liquid phase separation [10]. Studies suggest that one of these nucleating factors, G3BP1, is necessary and sufficient for SG to form during viral infections [31–33]. To determine if G3BP1 is sufficient for cbVG-dependent SG, we infected G3BP1 KO cells (Fig 5A, second lane) with RSV cbVG-high virus and looked at SG using TIAR staining as proxy for SG formation. RSV G mRNA levels confirmed that there were not significant differences in viral replication between cell types (Fig 5B). Unexpectedly, we observed TIAR-containing SG in G3BP1 KO cells (Fig 5C, upper panel). To confirm that these were canonical SGs and not aggregation of TIAR as an artifact of knocking out G3BP1, we treated the cells with CHX. Indeed, TIAR-containing SGs in G3BP1 KO cells are sensitive to CHX, suggesting that these were canonical SGs (Fig 5C, lower panel). These data indicate that knocking out G3BP1 is not sufficient to inhibit RSV-dependent SG, contradicting what has previously been suggested in the literature [31]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. cbVG-dependent SG inhibition is both G3BP1 and G3BP2 dependent. (A) Western blot analysis validating A549 G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO. (B) Expression of RSV G gene mRNA relative to the HKI 24 hpi with RSV cbVG-high infection at MOI 1.5 TCID 50 /cell in A549 control, G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO. (C) G3BP1 (red) and TIAR (green) staining for SG and viral protein (RSV NP) detection in control and G3BP1 KO cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell and treated with DMSO (upper panel) or CHX (10 μg/mL) (lower panel). (D) G3BP1 (red) and TIAR (green) staining for SG and viral protein (RSV NP) detection in A549 control, G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO cells infected 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell. No SG-positive cells were detected in any field of G3BP1/2 dKO A549 cells. All widefield images were acquired with the Apotome 2.0 at 63× magnification and are representative of 3 independent experiments. Scale bar = 50 μm. Statistical analysis: one-way ANOVA (**p < 0.01). Numerical values plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; CHX, cycloheximide; dKO, double KO; G3BP1, GTPase-activating protein-binding protein 1; HKI, housekeeping index; hpi, hours postinfection; KO, knockout; MOI, multiplicity of infection; NP, nucleoprotein; RSV, respiratory syncytial virus; SG, stress granule; TIAR, TIA-1-related. https://doi.org/10.1371/journal.pbio.3002381.g005 In the context of some nonvirus-induced stresses, knocking out both G3BP1 and G3BP2 have shown to be necessary for SG inhibition [34]. To test if cbVG-dependent SG inhibition requires KO of both G3BP1 and G3BP2, we next generated a G3BP2 KO cell line as well as a G3BP1/2 double KO (dKO) cell line (Fig 5A). When we infected G3BP1/2 dKO cells with RSV cbVG-high virus stocks, we no longer observed SG upon staining for TIAR, but SG were still formed in G3BP1 and G3BP2 single KO cells (Fig 5D). These data demonstrate that cbVG-dependent SG inhibition requires KO of both G3BP1 and G3BP2.

cbVG-dependent SGs are not required to induce the antiviral response As SG formation is often associated with induction of the intrinsic antiviral immunity [12,14,16,17], we then determined if SGs are necessary for the expression of antiviral genes in response to cbVGs. To do this, we infected A549 control, G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO cells with RSV cbVG-high and looked for differences in expression of genes involved in antiviral immunity, including IFNs and ISGs, at 24 hpi by qPCR. Expression of IL-29, ISG56, and IRF7 mRNAs was not impaired when comparing control and G3BP1/2 dKO cells, and only statistically significant differences were observed in IL-29 expression between G3BP1 KO and dKO (Fig 6A–6C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. cbVG-dependent SGs are not required for the antiviral response during RSV cbVG-high infection. mRNA copy numbers of (A) IL29, (B) ISG56, and (C) IRF7 relative to the HKI in A549 control, G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell. Statistical analysis: one-way ANOVA (*p < 0.05). (D) Log 2-fold change analysis of genes related to the antiviral response in A549 control, G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell relative to mock-infected cells. Genes that had less than a 2-fold decrease difference between control and G3BP1/2 dKO are represented in grey color. Genes that had more than a 2-fold decrease difference are highlighted in color. Genes with 2-fold decrease or more difference between control and G3BP1/2 dKO are shown in the right panel and compared to the log 2-fold change of G3BP1 and G3BP2 single KOs. (E) Western blot analysis of RIG-I, IFIT1, and IRF7 in A549 control, G3BP1 KO, G3BP2 KO, and G3BP1/2 dKO cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell. (F) Inverted mean intensity quantification of IRF7, IFIT1, and RIG-I western blot bands relative to α-tubulin loading control. Statistical analysis: one-way ANOVA. No statistical significance was found. (G) IL29 mRNA levels relative to the HKI in control and PKR KO cells cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell. Statistical analysis: one-way ANOVA. No statistical significance was found. (H) Western blot analysis of IFIT1 and IRF7 in control and PKR KO cells 24 hpi with RSV cbVG-high at MOI 1.5 TCID 50 /cell. Statistical analysis: one-way ANOVA. No statistical significance was found. All western blot images shown are representative of 3 independent experiments. Numerical values plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; dKO, double KO; G3BP1, GTPase-activating protein-binding protein 1; HKI, housekeeping index; hpi, hours postinfection; KO, knockout; MOI, multiplicity of infection; RSV, respiratory syncytial virus; SG, stress granule. https://doi.org/10.1371/journal.pbio.3002381.g006 To assess the impact of SG on the host antiviral response more broadly, we looked at the whole transcriptome in A549 control and KO cells at 24 hpi. Most ISGs were expressed at similar levels in control and dKO cells (difference in expression were less than 2-fold; Fig 6D). In the few cases when there were differences of 2-fold decrease or more in expression, the difference was also observed in the G3BP1 or G3BP2 single KO conditions, suggesting that the difference is driven by processes independent of SG formation (Fig 6D, right panel). Additionally, we tested whether absence of SG leads to reduced protein expression of ISGs. Expression of IFIT1, IRF7, and RIG-I was not different between the cell lines, demonstrating that the antiviral immune response is not dependent on SG formation (Fig 6E and 6F). Because the role G3BPs have in the stress response is directly in SG formation and not the translation inhibition that occurs upstream of the pathway, we looked at the direct role of PKR signaling in antiviral immunity. For this, we infected PKR KO cells with RSV cbVG-high virus and compared IL-29 transcript levels and IFIT1 protein levels to control infected cells and saw no significant differences (Fig 6G and 6H). Similarly, cells infected with SeV cbVG-high virus had no differences in phosphorylation of IRF-3, the primary transcription factor leading to type I IFN expression, nor differences in protein expression of the antiviral gene IFIT1 (S3A and S3B Fig). Altogether, these data suggest that PKR activation and SG formation are dispensable for global induction of antiviral immunity.

SeV cbVG-dependent SGs form dynamically during infection and correlate with reduced viral protein expression To study the dynamics of SG assembly and disassembly as well as assess the impact of SG during infection, we generated G3BP1-GFP expressing A549 cells to visualize SG formation in real time. Using live-cell imaging of cells infected with a recombinant SeV expressing miRFP670 (rSeV-CmiRF670) and supplemented with purified cbVG particles, we show dynamic formation and disassembly of SG throughout the course of the infection (S1 Movie). During the period of 6 to 72 hpi, we identified several subpopulations of cells (Fig 7A). Some cells formed SGs after infection and eventually disassembled them (Fig 7A, series 1). These cells showed faint levels of miRFP670 signal early in infection. Once SG disassembled, the miRFP670 signal increased. Other cells formed SG and eventually died (Fig 7A, series 2). A few cells assembled and disassembled SG and remained very low in miRFP670 signal throughout the infection (Fig 7A, series 3). Moreover, formation of SG persisted in the population even 13 dpi (Fig 7B). These data demonstrate that SeV cbVG-dependent SGs form asynchronously and that formation of SG continues throughout the infection. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. SeV cbVG-dependent SGs form asynchronously and are maintained at the population level throughout the infection. (A) G3BP1-GFP (green) expressing A549 cells infected with rSeV-CmiRF670 (magenta) reporter virus at MOI 3 TCID 50 /cell with 20 HAU of supplemented cbVG purified particles, time-lapse microscopy 6–72 hpi, images every 6 h at a 20× magnification. Series show focus of different cells in the population. (B) Time-lapse microscopy images of G3BP1-GFP (green) expressing A549 cells infected with rSeV-CmiRF670 (magenta) reporter virus at MOI 3 TCID 50 /cell with 20 HAU of supplemented cbVG purified particles from day 1 to day 13. (C) G3BP1-GFP (green) expressing A549 cells infected with rSeV-CmiRF670 (magenta) reporter virus at MOI 3 TCID 50 /cell with 20 HAU of supplemented cbVG purified particles, time-lapse microscopy 8–18 hpi, images taken every 1 h. All time-lapse images were acquired with a widefield microscope at 20× magnification. cbVG, copy-back viral genome; G3BP1, GTPase-activating protein-binding protein 1; HAU, hemagglutination unit; hpi, hours postinfection; MOI, multiplicity of infection; SeV, Sendai virus; SG, stress granule. https://doi.org/10.1371/journal.pbio.3002381.g007 In these experiments, we observed that the signal for the viral reporter gene miRFP670 was low in SG-positive cells, to the point where some cells appeared uninfected. This is similar, but more extreme, than our observation via immunofluorescence that SeV NP-positive SG-positive cells often showed lower signal for SeV NP compared to those that were SG-negative cells (Fig 2A). We observed similar findings in RSV cbVG-high infection when staining for the RSV F protein (S4 Fig). We hypothesized that a single cell could gain and lose miRFP670 signal within a 6-h window, resulting in SG-positive cells that appeared uninfected at the time of imaging. To confirm that SG-positive cells during live imaging were infected, we performed time-lapse imaging starting at 6 hpi before we begin to see SG-positive cells during the infection and tracked SG-positive cells every 30 min from 6 to 24 hpi to assess changes in the miRFP670 signal with a higher temporal resolution. SG-positive cells showed miRFP670 before forming SG and lost the signal as time went by, demonstrating that SG formation is correlated with a reduction in viral protein expression (Fig 7C and S2 Movie).

cbVG-mediated interference with viral protein expression is independent on MAVS signaling The reduction on virus protein levels in SG-positive cells led us to hypothesize that the well-established virus interference function of cbVGs is at least in part mediated by the induction of the cellular stress response. Because cbVGs are known to interfere with virus replication through the induction of MAVS signaling and IFN production, which consequently leads to a reduction of viral protein levels, we determined if this viral protein reduction observed in SG-positive cells was due to the IFN response and independent on SG formation. To test this, we infected MAVS KO cells with SeV cbVG-high and compared viral protein SeV NP expression to control infected cells. SG-positive cells showed similar SeV NP fluorescence in control and MAVS KO cells (Fig 8A). These data suggest that the interference in viral protein expression observed in cbVG and SG-positive cells is not due to the IFN response and, instead, suggest a direct role for the cellular stress response in reducing viral protein expression. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 8. cbVGs induce translation inhibition in SG-positive cells. (A) SG (G3BP1 white) and viral protein (SeV NP) detection in control and MAVS KO A549 cells 24 hpi with SeV cbVG-high virus at MOI 1.5 TCID 50 /cell 24 hpi. CTCF quantification of SeV viral protein NP in control and MAVS KO A549 SG-positive cells. (B) G3BP1 (green) for SG detection and PMY (magenta) for translation in cells infected with SeV cbVG-high (SeV NP, red) at MOI 3 TCID 50 /cell 24 hpi or treated with sodium arsenite, with and without treatment with PMY for 5 min. (C) Quantification of PMY intensity (CTCF) in cells after drug treatment with sodium arsenite or SG-positive and SG-negative SeV cbVG-high infected cells. Each dot represents the CTCF average of approximately 100 cells counted for each condition. Widefield images were acquired with the Apotome 2.0 at 63× magnification, scale bar = 50 μm. Statistical analysis: one-way ANOVA (*p < 0.05). (D) Diagram summarizing the role cbVGs have in inducing virus interference through activation of MAVS signaling and induction of translation inhibition. Numerical values plotted can be found in the Supporting information: S1 Data. cbVG, copy-back viral genome; CTCF, corrected total cell fluorescence; G3BP1, GTPase-activating protein-binding protein 1; hpi, hours postinfection; KO, knockout; MAVS, mitochondrial antiviral signaling; MOI, multiplicity of infection; NP, nucleoprotein; PMY, puromycin; SeV, Sendai virus; SG, stress granule. https://doi.org/10.1371/journal.pbio.3002381.g008

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002381

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.