SARS-CoV-2 variant-specific differences in inhibiting the effects of the PKR-activated integrated stress response

Abstract

The integrated stress response (ISR) is a eukaryotic cell pathway that triggers translational arrest and the formation of stress granules (SGs) in response to various stress signals, including those caused by viral infections. The SARS-CoV-2 nucleocapsid protein has been shown to disrupt SGs, but SARS-CoV-2 interactions with other components of the pathway remains poorly characterized. Here, we show that SARS-CoV-2 infection triggers the ISR through activation of the eIF2α-kinase PKR while inhibiting a variety of downstream effects. In line with previous studies, SG formation was efficiently inhibited and the induced eIF2α phosphorylation only minimally contributed to the translational arrest observed in infected cells. Despite ISR activation and translational arrest, expression of the stress-responsive transcription factors ATF4 and CHOP was not induced in SARS-CoV-2 infected cells. Finally, we found variant-specific differences in the activation of the ISR between ancestral SARS-CoV-2 and the Delta and Omicron BA.1 variants in that Delta infection induced weaker PKR activation while Omicron infection induced higher levels of p-eIF2α, and greatly increased SG formation compared to the other variants. Our results suggest that different SARS-CoV-2 variants can affect normal cell functions differently, which can have an impact on pathogenesis and treatment strategies.

Keywords: Integrated stress response; Omicron; SARS-COV-2; SARS-COV-2 variants; Stress granules; Translational arrest.

Fig. 1

ISR activation and SG formation in ancestral SARS-CoV-2 infected cells. (A) eIF2α-phosphorylation in SARS-CoV-2 infected Vero E6 and A549-hACE2 cells. Uninfected and SARS-CoV-2 infected cells were treated with 1 mM sodium arsenite (SA), an ISR-inducing chemical, for 1 h, as positive control. Cell lysates were collected at 24 hpi and analyzed for levels of p-eIF2α. Calnexin was used as a loading control. (B) Protein levels and phosphorylation of PKR in SARS-CoV-2 infected A549-hACE2 cells. Cells were infected SARS-CoV-2. Cell lysates were collected at 24 hpi and analyzed for expression and phosphorylation levels of PKR. b-actin was used as a loading control. (C) Quantification of (B). Protein levels were normalized to cellular levels of b-actin. The data are presented as the fold change in relation to the protein levels of uninfected cells. Data are represented as mean ± SEM. n = 3. (D) G3BP1 staining in SARS-CoV-2 infected A549-hACE2 cells. Cells were infected with SARS-CoV-2 and fixed after 24 h. Uninfected cells that were either untreated or treated with 1 mM SA were used as a control. The cells were stained with antibodies against SARS-CoV-2 N protein (magenta) and G3BP1 (green) and with DAPI (blue). Imaging was performed with confocal microscopy (20X). Scale bars 20 mm. (E) Localization of the viral N protein within SGs. A549-hACE2 cells were fixed 24 h hpi and stained with antibodies against SARS-CoV-2 N protein (magenta) and G3BP1 (green) and with DAPI (blue). Arrows show colocalization of viral proteins with G3BP1. Imaging was performed with confocal microscopy (20X). Scale bars 20 mm. *, p < 0.05; **, p < 0.005.

Fig. 2

The effect of different stress stimuli on SG formation during ancestral SARS-CoV-2 infection and on viral replication in Vero E6 cells. (A) Experimental setup for Figs. 2B - G. Uninfected and ancestral SARS-CoV-2 infected cells were (B) treated with 1 mM sodium arsenite for 1 h, (C) transfected with 2.5 mg Poly(I:C) overnight, (D) treated with 5 mM thapsigargin for 5 h or (E) starved for 5 h and fixed at 24 hpi. They were stained for SARS-CoV-2 N protein and G3BP. For the analysis using fluorescence microscopy, ≥5 images of a total of >100 cells were taken at random positions and the number of infected cells showing SGs was determined. Data are represented as mean ± SEM. n ≥ 3. (F) Effect of different stress inducers on infectivity. SARS-CoV-2 infected cells were treated with sodium arsenite, Poly(I:C), thapsigargin or starved (see materials and methods) and fixed at 24 hpi They were stained for SARS-CoV-2 N protein. For the analysis using fluorescence microscopy, ≥5 images of a total of >100 cells were taken at random positions and the number of infected cells was determined and normalized against the infection rate of untreated cells. Data are represented as mean ± SEM. n = 3. (G) Effect of different stress inducers on viral replication. SARS-CoV-2 infected cells were treated with sodium arsenite, Poly(I:C), thapsigargin or starved (see materials and methods) and 6 h after treatment the viral titers were determined. Data are represented as mean ± SEM. n ≥ 3. *, p < 0.05; **, p < 0.005; ****, p < 0.0001.

Fig. 3

ISR activation and SG formation during infection with different SARS-CoV-2 variants (A) Time kinetics of SG formation in SARS-CoV-2 infected Vero E6 cells. Cells infected with different SARS-CoV-2 variants were fixed at 6, 24 or 48 hpi. They were stained for SARS-CoV-2 N protein and G3BP. For the analysis using fluorescence microscopy, ≥5 images of a total of >100 cells were taken at random positions and the number of infected cells showing SGs was determined. Data are represented as mean ± SEM. n ≥ 3. (B) SG formation in SARS-CoV-2 infected A549-hACE2 cells. Cells infected with different SARS-CoV-2 variants were fixed at 24 hpi. They were stained for SARS-CoV-2 N protein and G3BP. For the analysis using fluorescence microscopy, ≥5 images of a total of >100 cells were taken at random positions and the number of infected cells showing SGs was determined. Data are represented as mean ± SEM. n ≥ 3. (C) SG formation after sodium arsenite challenge in SARS-CoV-2 infected Vero E6 and A549-hACE2 cells. Cells were infected with different SARS-CoV-2 variants and treated with 1 mM sodium arsenite for 1 h. Cells were fixed at 24 hpi. They were stained for SARS-CoV-2 N protein and G3BP. For the analysis using fluorescence microscopy, ≥5 images of a total of >100 cells were taken at random positions and the number of infected cells showing SGs was determined. Data are represented as mean ± SEM. n = 3. (D) Replication of different SARS-CoV-2 variants in A549-hACE2 cells. Cells were infected with different SARS-CoV-2 variants and at 24 hpi the viral titers in the supernatant were determined. Data are represented as mean ± SEM. n = 4. (E) eIF2α- and PKR-phosphorylation in SARS-CoV-2 infected A549-hACE2 cells. Cells were infected with ancestral (A), Delta (D), or Omicron (O) SARS-CoV-2. Lysates were collected at 24 hpi and analyzed for levels of p-eIF2α and p-PKR. Calnexin was used as a loading control. (F) Quantification of (D). Protein levels were normalized to cellular levels of b-actin. The data are presented as the fold change in relation to the protein levels of cells infected with ancestral SARS-CoV-2. Data are represented as mean ± SEM. n = 3. *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p < 0.0001.

Fig. 4

The effect of different stress stimuli on infectivity and replication of ancestral SARS-CoV-2, Delta, and Omicron. A) Effect of different stress inducers on infectivity. SARS-CoV-2 infected A549-hACE2 cells were treated with sodium arsenite, Poly(I:C), thapsigargin or starved as described in Fig. 2 and fixed at 24 hpi. They were stained for SARS-CoV-2 N protein. For the analysis using fluorescence microscopy, ≥5 images of a total of >100 cells were taken at random positions and the number of infected cells was determined and normalized against the infection rate of untreated cells. Data are represented as mean ± SEM. n = 3. (B) Effect of different stress inducers on viral replication. SARS-CoV-2 infected A549-hACE2 cells were treated with sodium arsenite, Poly(I:C), thapsigargin or starved as described in Fig. 2. 6h after treatment the viral titers were determined and normalized against the titers in untreated cells. Data are represented as mean ± SEM. n = 3. *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p < 0.0001.

Fig. 5

The effect of ISR activation during SARS-CoV-2 infection on global translational levels and viral replication in A549-hACE2 cells. (A) Puromycin incorporation in A549-hACE2 cells infected with SARS-CoV-2 variants. Uninfected and infected cells were treated with sodium arsenite and/or ISRIB. Before sample collection at 24 hpi, the cells were treated with puromycin to visualize translational levels. (B) Quantification of (A). Puromycin levels were normalized to cellular levels of b-actin. The data are presented as the fold change in relation to the puromycin levels of untreated, uninfected cells and are represented as mean ± SEM. n ≥ 3. (C) ISRIB treatment decreases viral replication. SARS-CoV-2 infected cells were treated with ISRIB, and the viral titers of the supernatant collected at 24 hpi were determined. The data are represented as mean ± SEM. n = 3. *, p < 0.05; **, p < 0.005.

Fig. 6

Suppression of ATF4 and CHOP production in SARS-CoV-2 infected A549-hACE2 cells. (A) Uninfected and ancestral strain-infected cells were treated with sodium arsenite (SA), thapsigargin (T) or remained untreated. At 24 hpi, cell lysates were collected and analyzed for expression of ATF4, CHOP and p-eIF2α via western blot. Calnexin was used as a loading control. (B) Quantification of (A). Levels of ATF4 and CHOP were normalized against cellular levels of calnexin. The data are represented as the fold change in relation to the protein levels of untreated, uninfected cells and are represented as mean ± SEM. n = 3. (C) Uninfected and ancestral strain-infected cells were treated with T and ISRIB (I) or remained untreated. At 24 hpi, cell lysates were collected and analyzed for expression of ATF4 via western blot. Calnexin was used as a loading control. (D) Quantification of (C). Levels of ATF4 and CHOP were normalized against cellular levels of calnexin. The data are represented as the fold change in relation to the protein levels of untreated, uninfected cells and are represented as mean ± SEM. n = 3. (E) Uninfected cells and cells infected with different SARS-CoV-2 variants were treated with T or remained untreated. At 24 hpi, cell lysates were collected and analyzed for expression of ATF4 via western blot. Calnexin was used as a loading control. (F) Quantification of (E). Levels of ATF4 and CHOP were normalized against cellular levels of calnexin. The data are represented as the fold change in relation to the protein levels of untreated, uninfected cells and are represented as mean ± SEM. n = 3. *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p < 0.0001.