SARS-CoV-2 Nsp2 Contributes to Inflammation by Activating NF-κB

COVID-19 is associated with robust inflammation and partially impaired antiviral responses. The modulation of inflammatory gene expression by SARS-CoV-2 is not completely understood. In this study, we characterized the inflammatory and antiviral responses mounted during SARS-CoV-2 infection. K18-hACE2 mice were infected with a Wuhan-like strain of SARS-CoV-2, and the transcriptional and translational expression interferons (IFNs), cytokines, and chemokines were analyzed in mouse lung homogenates. Our results show that the infection of mice with SARS-CoV-2 induces the expression of several pro-inflammatory CC and CXC chemokines activated through NF-κB but weakly IL1β and IL18 whose expression are more characteristic of inflammasome formation. We also observed the downregulation of several inflammasome effectors. The modulation of innate response, following expressions of non-structural protein 2 (Nsp2) and SARS-CoV-2 infection, was assessed by measuring IFNβ expression and NF-κB modulation in human pulmonary cells. A robust activation of the NF-κB p65 subunit was induced following the infection of human cells with the corresponding NF-κB-driven inflammatory signature. We identified that Nsp2 expression induced the activation of the IFNβ promoter through its NF-κB regulatory domain as well as activation of p65 subunit phosphorylation. The present studies suggest that SARS-CoV-2 skews the antiviral response in favor of an NF-κB-driven inflammatory response, a hallmark of acute COVID-19 and for which Nsp2 should be considered an important contributor.

One of the first host defense mechanisms against pathogens such as viruses is the innate immune response that is initiated by the recognition of pathogen-associated molecular patterns (PAMPs) by cellular sensors [7]. One of the main effector systems triggered Wisent Inc., St-Bruno, QC, Canada) for at least one week then clones were isolated using a single-cell cloning approach.
Reporter assays. HEK293T cells were plated in 24-well plates (1.6 × 10 5 /well) and transfected with 50 ng to 100 ng of reporter vectors and 30 ng to 300 ng of Nsp1, Nsp2 or Nsp1-Nsp2 vectors brought to 0.5 µg/well with the empty expression vector. Twenty-four hours post-transfection, transfected cells were infected with 20 hemagglutinin units of SeV or stimulated with 500 units of IFNα (PBL Assay Science, Piscataway, NJ, USA). Sixteen hours later, cells were lysed, and the luciferase activity was determined as previously described [33].
IFNβ induction and ISG induction. HEK293T cells were plated (3 × 10 4 /well) in 12-well plates and transfected with 0.2 µg of Nsp1 vector, 0.6 µg of Nsp2 vector, or 0.8 µg of Nsp1-Nsp2 vector complete to 1 µg/well with empty pCDNA5. Twenty-four hours post-transfection, cells were infected with 40 hemagglutinin units of SeV for sixteen hours. Supernatants and cells were collected separately, and cells were lysed in 0.5 mL of QIAzol reagent (Qiagen, Toronto, ON, Canada). Samples were stored at −80 • C until future analysis.
Infection and poly(I:C) stimulation. A549-hACE2 cells were plated (7.5 × 10 4 /well) in 12-well plates and infected with Wuhan-like SARS-CoV-2 strain following the same procedure described above. Twenty-four hours post-infection, cells were transfected with 2 µg/mL of poly(I:C) for 16 h. Supernatants and cells were collected separately, and cells were lysed in 0.5 mL of QIAzol reagent (Qiagen). Supernatants were incubated with 1% triton for one hour at room temperature to inactivate SARS-CoV-2. Samples were stored at −80 • C until analyzed.
IFNß quantification. The IFNβ in the supernatant was quantified with the Human IFNbeta DuoSet enzyme-linked immunosorbent assay (ELISA) kit, according to the supplier recommendations (R&D Systems Inc., Toronto, ON, Canada).
Multiplex cytokines quantification. Cytokines in mouse lung homogenates were measured using a custom ProcartaPlexTM Mouse Mix & Match Panels kit (Invitrogen, Waltham, MA, USA) on the Bio-Plex 200 (Bio-Rad Laboratories Ltd.).
Quantitative real-time PCR analysis. Total RNA from cell cultures was extracted following QIAzol protocol and RNA from mouse lungs was extracted using the Bead Mill Tissue RNA Purification Kit and the Omni Bead Ruptor Bead Mill homogenizer (Kennesaw, GA). Following extraction, residual DNA was removed by treating the samples with DNAse I (Roche, Mississauga, ON, Canada). For the quantification of human gene expression and mouse Cxcl1, Ccl2, Isg56(Ifit1), Ifny, and Ifna, RNA was reverse transcribed to cDNA using SuperScript™ IV VILO™ mastermix (ThermoFisher Scientific). Quantitative real-time PCR (qPCR) was performed using the SsoAdvanced Universal Probes Supermix (Bio-Rad Laboratories Ltd.) for IFNB1 gene and GAPDH as the housekeeping gene. SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories Ltd.) was used for human ISG15 and ISG56 and mouse genes including Gapdh as the housekeeping gene on the Rotor-Gene Q 5plex (Qiagen). The RT-qPCR primers and probes are listed in Supplementary Table S2.
Digital PCR analysis. SARS-CoV-2 viral RNA loads were determined using Droplet Digital PCR (ddPCR) supermix for probes without dUTP (Bio-Rad Laboratories Ltd.) and the QX200 Droplet Digital PCR System Workflow (Bio-Rad Laboratories Ltd.). The ddPCR primers and probes are listed in the Supplementary Table S2. RT 2 profiler PCR Arrays. The RNA extracted from mouse lungs as described above was cleaned up using On-Column DNAse using RNAse-Free DNase Set (Qiagen) and RNeasy Mini Kit (Qiagen). The RNA was reverse transcribed using RT2 First Strand Kit (Qiagen). qPCR and quality control were performed using RT 2 SYBR ® Green ROC FAST Mastermix (Qiagen) and RT 2 profiler PCR Arrays: Mouse Antiviral response (Qiagen). Data analyses were performed using the GeneGlobe (Qiagen) analyzing tool. Genes of the RT 2 profiler PCR Arrays are listed in the Supplementary Table S3. Immunofluorescence. A549-hACE2 cells were plated (1.6 × 10 4 /well) in 8-well chamber slides. Twenty-four hours later, cells were infected with SARS-CoV-2 as described above. Forty-eight hours post-infection, cells were fixed in 2% paraformaldehyde in PBS for one hour at room temperature. Cells were then incubated for thirty minutes in blocking solution (PBS with 0.1% bovine serum albumin (BSA), 3% FBS, 0.1% Triton X-100, and 1 mM EDTA) then with 17 µg/mL of rabbit anti-SARS-CoV-2-N (Rockland Immunochemicals Inc., Limerick, PA, USA) in the blocking solution for one hour at room temperature. After, cells were washed three times for 5 min with PBS and incubated with 4 µg/mL of goat anti-rabbit-Alexa-488 (Thermo Fisher Scientific) in the blocking solution for 30 min at room temperature. Finally, cells were washed for 5 min in PBS and incubated into PBS 1X with 1.67 µg/mL of DAPI (Invivogen). Cells were washed for 5 min in PBS, mounted with ProLong Diamond Antifade reagent (Thermo Fisher Scientific) and images were acquired using a Z2 confocal microscope with LSM 800 scanning system (Zeiss, Germany). Images were captured with a 20x objective (Zeiss, Apochromat). ZEN 2.3 software (Zeiss) was used to acquire and process images. Z-stack projections of 3 µm in total thickness are represented. Contrast adjustments and image processing were performed using Zen lite (Zeiss).

Results
SARS-CoV-2 induces robust CC and CXC chemokine production but limited type I IFN production in mice. To study the innate immune response developed during infection, K18-hACE2 mice were infected with the Wuhan-like strain SARS-CoV-2. Three days post-infection, the mice were euthanized and their lungs used for further analyses. No change in body weight or temperature was observed during this time [36]. As shown in Figure 1A,B, all the infected mice had significant pulmonary infectious viral loads and expressed abundant SARS-CoV-2 E gene copy numbers. When the antiviral-related gene expression was measured, Chemokine C-X-C Ligand 10 (Cxcl10 [Ip-10]), Chemokine C-C Ligand 2 (Ccl2 [Mcp-1]), Cxcl11 (Ip-9), Cxcl9 (Mig), Ifnb1, and Interleukin 6 (Il-6) were the main cytokine genes upregulated ( Figure 1C). Ifnγ, Tumor necrosis factor α (Tnfα, Cxcl1(Groa), Ifnα, Ccl3 (Mip-1α), Ccl4 (Mip-1β) genes were also upregulated, but to a lesser extent. In contrast, Il12α and Il18 genes were downregulated or poorly induced. When analyzed at the protein level, CC and CXC chemokines were efficiently produced in response to infection. In contrast, despite robust Ifnβ1, Ifnγ, Il-6, and Tnf gene expression, Viruses 2023, 15,334 6 of 17 few gene products were measured ( Figure 1C). In terms of relative abundance, chemokines were the mediators produced at the greatest levels, being several hundred to a thousand times more abundant than other pro-inflammatory cytokines and type I IFN ( Figure 1D). These results show that SARS-CoV-2 triggers an innate immune response in mice dominated by chemokines. expressed abundant SARS-CoV-2 E gene copy numbers. When the antiviral-related gene expression was measured, Chemokine C-X-C Ligand 10 (Cxcl10 [Ip-10]), Chemokine C-C Ligand 2 (Ccl2 [Mcp-1]), Cxcl11 (Ip-9), Cxcl9 (Mig), Ifnb1, and Interleukin 6 (Il-6) were the main cytokine genes upregulated ( Figure 1C). Ifnγ, Tumor necrosis factor α (Tnfα, Cxcl1(Groa), Ifnα, Ccl3 (Mip-1α), Ccl4 (Mip-1β) genes were also upregulated, but to a lesser extent. In contrast, Il12α and Il18 genes were downregulated or poorly induced. When analyzed at the protein level, CC and CXC chemokines were efficiently produced in response to infection. In contrast, despite robust Ifnβ1, Ifnγ, Il-6, and Tnf gene expression, few gene products were measured ( Figure 1C). In terms of relative abundance, chemokines were the mediators produced at the greatest levels, being several hundred to a thousand times more abundant than other pro-inflammatory cytokines and type I IFN ( Figure  1D). These results show that SARS-CoV-2 triggers an innate immune response in mice dominated by chemokines. The SARS-CoV-2 E gene copy number was evaluated by ddPCR using lung RNA and expressed as number per copy of Rpp30 mRNA (mean ± SD). (B) Infectious viral titers were determined in lung homogenates and expressed in TCID50/mL (mean ± SD). (C) Gene expression was evaluated by RT-qPCR and cytokine concentration in lung homogenates determined using a 13-plex Luminex panel. Cytokine gene expression and concentration levels are presented as heatmaps with results expressed as fold (log2) relative to mock-infected mice (mean). Statistical analyses were performed by comparing 2 (−ΔCt) values for each gene in the control group and infected groups with a nonparametric T-test. (D) Cytokine concentrations in lung homogenates. Results are expressed as mean ± SD. For protein quantification, statistical analyses were performed by comparing the normalized concentration for each cytokine in the control group and infected groups with a nonparametric t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. SARS-CoV-2 infection in mice does not induce an inflammatory reaction mediated by the inflammasome. Along with the cytokines studied above, the expression of several genes involved in the Toll-like receptor (TLR), NOD-like receptor (NLR) and RIG-like receptor (RLR) signaling pathways were measured. Despite the robust upregulation of the Mediterranean fever gene (Mefv) implicated in inflammasome formation [37] and The SARS-CoV-2 E gene copy number was evaluated by ddPCR using lung RNA and expressed as number per copy of Rpp30 mRNA (mean ± SD). (B) Infectious viral titers were determined in lung homogenates and expressed in TCID 50/mL (mean ± SD). (C) Gene expression was evaluated by RT-qPCR and cytokine concentration in lung homogenates determined using a 13-plex Luminex panel. Cytokine gene expression and concentration levels are presented as heatmaps with results expressed as fold (log 2 ) relative to mock-infected mice (mean). Statistical analyses were performed by comparing 2 (−∆Ct) values for each gene in the control group and infected groups with a nonparametric T-test. (D) Cytokine concentrations in lung homogenates. Results are expressed as mean ± SD. For protein quantification, statistical analyses were performed by comparing the normalized concentration for each cytokine in the control group and infected groups with a nonparametric t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. SARS-CoV-2 infection in mice does not induce an inflammatory reaction mediated by the inflammasome. Along with the cytokines studied above, the expression of several genes involved in the Toll-like receptor (TLR), NOD-like receptor (NLR) and RIG-like receptor (RLR) signaling pathways were measured. Despite the robust upregulation of the Mediterranean fever gene (Mefv) implicated in inflammasome formation [37] and proinflammatory cytokine release, inflammasome components such as the Apoptosis-associated speck-like protein containing a CARD (Pycard), Proline-serine-threonine phosphatase-interacting protein 1(Pspip1), Absent in melanoma 2 (Aim2), Caspase 1 (Casp1), and NLR family pyrin domain containing 3 (Nlrp3) were weakly modulated early in infection ( Figure 2A). Moreover, Caspase recruitment domain-containing protein 9 (Card9) and Mitogen-activated protein kinase 14 (Mapk14), implicated in the inflammatory process, were downregulated. This finding is consistent with the lack of inflammasome-characteristic cytokines such as IL1β and IL18 ( Figure 1C). sociated speck-like protein containing a CARD (Pycard), Proline-serine-threonine phosphatase-interacting protein 1(Pspip1), Absent in melanoma 2 (Aim2), Caspase 1 (Casp1), and NLR family pyrin domain containing 3 (Nlrp3) were weakly modulated early in infection ( Figure 2A). Moreover, Caspase recruitment domain-containing protein 9 (Card9) and Mitogen-activated protein kinase 14 (Mapk14), implicated in the inflammatory process, were downregulated. This finding is consistent with the lack of inflammasome-characteristic cytokines such as IL1β and IL18 ( Figure 1C). Modulation of the IFN activation pathways during SARS-CoV-2 infection. As shown in the Figure 2B, many downstream effector genes such as Interleukin-1 receptor-associated kinase 1 (Irak1), Transcription factor (Jun), Mavs [Ips-1], Mitogen-activated protein kinase kinase 7 (Map3k7[Tak1]) were downregulated during infection. Moreover, Canopy FGF signaling regulator 3 (Cnpy3), a TLR chaperon [38], was also downregulated, which could impair the recognition of viral PAMPs. The inhibitor of the nuclear factor kappa-B kinase subunit beta (Ikbkb[Ikkb]) was also downregulated; meanwhile, nuclear factor kappa-B1 (Nf-kb1) expression was not affected. As IKBKB inhibit the NF-κB complex formation [39], a reduction in IKBKB would result in higher overall NF-kB activity. On the other hand, cytoplasmic and endosomal ssRNA and dsRNA sensors, such as Tlr3, Tlr7, Tlr8, DExD/H-box helicase 58 (Ddx58[Rig-i)), 2′-5′-oligoadénylate synthetase 2 (Oas2), and Interferon-induced helicase C domain-containing protein 1 (Ifih1 [Mda-5]), were Results are expressed as fold (log 2 ) relative to mock-infected mice (mean, n = 4/group). For gene expression, statistical analyses were performed by comparing 2 (−∆Ct) values for each gene in the control group and infected groups with a nonparametric t-test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Robust ISG expression despite low-level type I IFN production during SARS-CoV-2 infection. The expression of genes associated with type I IFN signaling was monitored during infection ( Figure 2B). Interferon-stimulated genes such as Isg15, Isg56 (IFIT1), and Interferon-induced GTP-binding protein Mx1 (Mx1) were strongly induced following infection. In agreement with those results, Stat1 gene transcription was also upregulated. Conversely, Ifnar1 expression was downregulated. SARS-CoV-2 infection induces NF-κB phosphorylation. To verify the ability of SARS-CoV-2 to activate NF-κB signaling, NF-κB p65 subunit phosphorylation was quantified by Western Blot in infected A549-hACE2 cells. As shown in Figure 3A, at 24 h post-infection the virus did not affect the overall p65 expression while at 48 h post-infection, a threefold reduction in the p65 expression was noted relative to the mock treated cells. On the other hand, SARS-CoV-2 infection (48 h) induced p65 phosphorylation by more than 4-fold. The p65 phosphorylation increment correlated with the viral load, as determined by N protein expression ( Figure 3A). When p65 phosphorylation is normalized with the overall p65 expression levels, a 15-fold induction in p65 phosphorylation is measured 48 h post-infection relative to control cells ( Figure 3B).
SARS-CoV-2 induces IFNB1 transcription, inhibits IFNβ protein synthesis, and does not impair ISG transcription. A549-hACE2 cells were treated with the SARS-CoV-2 Wuhanlike strain and infection confirmed by immunofluorescence using anti-nucleocapsid (N) antibodies ( Figure 3C). The IFNB1, ISG15, ISG56 mRNA levels and IFNβ1 secretion were measured using mock-infected or SARS-CoV-2-infected cells treated or not with poly(I:C), a type I IFN inducer. We made use of poly(I:C) to determine whether SARS-CoV-2 would affect exogenous IFN inducing stimuli. The IFNβ1 mRNA quantification revealed that SARS-CoV-2 infection efficiently induced IFNB1 gene transcription. Stimulation with poly(I:C) amplified IFNB1 mRNA expression ( Figure 3D). Similar results were obtained for ISG15 and ISG56 gene expression ± poly(I:C) ( Figure 3E,F). At the protein level, no IFNβ1 was detected in the supernatant of infected cells with SARS-CoV-2 ( Figure 3G). The virus was also capable of partially inhibiting the poly(I:C)-induced IFNB1 secretion ( Figure 3G). This translational inhibition was recapitulated with Nsp1 transfection (Figure 3H,I).
Characterization of IFNβ1 promoter activation by Nsp2. Th IFNB gene is regulated by a complex promoter containing positive regulatory domains I and III (PRD-I-III) (IRF3/7 responsive elements) and a PRD-II (NF-κB-responsive element) ( Figure 4A) [40]. As Nsp2 s effects on IFNB1 are somewhat controversial in the literature, we opted to determine whether Nsp2 is a Type I IFN antagonist or not. An expression vector coding for Nsp2 was co-transfected into HEK293T cells together with an IFNβ1 promoter luciferase reporter. As shown in Figure 4B, Nsp2 expression activated the IFNβ1 promoter and amplified the response to SeV infection. [40]. The results with PRD-I-III and PRD-II demonstrated that Nsp2 activated the NF-κB binding domain of the IFNβ1 promoter ( Figure 4C) while having no effects on the PRD-I-III elements ( Figure 4D). To determine whether Nsp2 could induce IFNβ1 production and secretion, we generated Nsp2-or GFP-inducible A549(hACE2) and measured IFNβ1 release in the supernatant following mock or poly (I:C) stimulation. The results showed that Nsp2 expression did not impair nor induce IFNB1 production in a detectable way ( Figure 4E). fold reduction in the p65 expression was noted relative to the mock treated cells. other hand, SARS-CoV-2 infection (48 h) induced p65 phosphorylation by more fold. The p65 phosphorylation increment correlated with the viral load, as determ N protein expression ( Figure 3A). When p65 phosphorylation is normalized with th all p65 expression levels, a 15-fold induction in p65 phosphorylation is measur post-infection relative to control cells ( Figure 3B).   Figure S1) and expressed as normalized levels (Norm. vol. ×10 8 ) below the blot. Then, NF-κB p65 phosphorylation is normalized with p65 subunit expression and expressed as fold of induction relative to the mock-infected control (mean ± SD, n = 2 replicates/condition) (B). SARS-CoV-2 Nucleocapsid staining on A549-hACE2 infected cells. Forty-eight hours post-infection cells were fixed and stained as described in the materials and methods section. This staining was performed a least three times and a representative picture was shown. (C). Effect of SARS-CoV-2 infection on poly(I:C) IFNB1 mRNA expression (D), ISG15 mRNA (E) or ISG56 mRNA expression (F), induced IFNβ1 production (G). Twenty-four hours post-seeding, A549-hACE2 cells were infected with SARS-CoV-2 as described in the materials and methods section. Thirty-two hours post-infection, SARS-CoV-2 and mock-infected cells were stimulated with poly(I:C) and RNA extracted and analyzed with RT-qPCR while IFNβ1 was measured in the supernatant by ELISA. Effect of Nsp1 on IFNb1 mRNA (H) and protein (I) expression. Cells were transfected with expression vectors Nsp1 or an empty vector and infected or not with SeV as IFNβ inducer. RNA was isolated and analyzed for IFNB1 mRNA (H), Supernatants were collected and assayed for IFNβ production (I). All experiments were performed twice in triplicate and the compilation of the data is shown. Results are expressed as activation percentage relative to the poly I:C transfected or SeV-infected negative control (mean ± SD (n = 6 replicates/condition)). Statistical analyses were performed by comparing mock-infected or pcDNA-transfected control with corresponding condition with a nonparametric T-test. * p < 0.05, ** p < 0.01, *** p < 0.001.  Reporter essays were performed as described above with 300 ng, 200 ng, and 100 ng of vector encoding for Nsp2. All experiments were performed twice in triplicate, and one representative experiment is shown. Results are expressed as fold activation relative to the pcDNA mock-infected control (mean ± SD, n = 3 replicates/condition). Effect of Nsp2 expression on the IFNβ1 production (E). Twenty-four hours post-seeding, A549-hACE2 SB-Nsp2 or SB-GFP cells were induced with 0,5μg/mL of Doxycycline. Thirty-two hours post-induction, cells were stimulated with poly(I:C) for sixteen hours and IFNβ1 was measured in the supernatant by ELISA. All experiments were performed twice in triplicate and the compilation of the data is shown. Results are expressed as activation percentage relative to the poly I:C mock-infected control (mean ± SD (n = 6 replicates/condition)). Statistical analyses were performed by comparing corresponding condition with mock-transfected control with a nonparametric one-way ANOVA with Dunn's correction. ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001.
Nsp2 co-transfection fails to reduce the inhibitory effect of Nsp1. Since Nsp1 was identified as the main type I IFN antagonist by several groups, we aimed at evaluating whether both proteins might antagonize each other. First, Nsp1 the expression vector was co-transfected into HEK293T cells together with the IFNβ1 reporter to validate its inhibi- Reporter essays were performed as described above with 300 ng, 200 ng, and 100 ng of vector encoding for Nsp2. All experiments were performed twice in triplicate, and one representative experiment is shown. Results are expressed as fold activation relative to the pcDNA mock-infected control (mean ± SD, n = 3 replicates/condition). Effect of Nsp2 expression on the IFNβ1 production (E). Twenty-four hours post-seeding, A549-hACE2 SB-Nsp2 or SB-GFP cells were induced with 0.5 µg/mL of Doxycycline. Thirty-two hours post-induction, cells were stimulated with poly(I:C) for sixteen hours and IFNβ1 was measured in the supernatant by ELISA. All experiments were performed twice in triplicate and the compilation of the data is shown. Results are expressed as activation percentage relative to the poly I:C mock-infected control (mean ± SD (n = 6 replicates/condition)). Statistical analyses were performed by comparing corresponding condition with mock-transfected control with a nonparametric one-way ANOVA with Dunn's correction. ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001. Nsp2 co-transfection fails to reduce the inhibitory effect of Nsp1. Since Nsp1 was identified as the main type I IFN antagonist by several groups, we aimed at evaluating whether both proteins might antagonize each other. First, Nsp1 the expression vector was co-transfected into HEK293T cells together with the IFNβ1 reporter to validate its inhibitory activity ( Figure 5A). Second, the Nsp1 and Nsp2 vectors were cotransfected with IFNβ1 reporter activation. No significant differences between cells co-transfected with Nsp1 and Nsp2 vectors and cells singly transfected with an Nsp1 vector alone were detected ( Figure 5B). When transfected cells were analyzed for Nsp1 and Nsp2 expression, Nsp2 could be efficiently detected only in the absence of Nsp1, suggesting that Nsp1 inhibited Nsp2 translation ( Figure 5C). HEK293T cells were seeded, transfected, and simulated according to the procedure described above and the luciferase activity was measured then standardized as described for other reporter assays. Doses of 300 ng, 100 ng, 30 ng, and 3 ng per well of Nsp1 vectors for Nsp1 alone experimentation (A). HEK293T cells were seeded, transfected, and simulated according to the procedure described above and the luciferase activity was measured then standardized as described for other reporter assays. Doses of 300 ng of Nsp2 aand 100 ng of Nsp1 per well were used (B). All experiments were performed twice in triplicate and the compilation of the data is shown. Results are expressed as fold activation relative to the pcDNA mock-infected control (mean ± SD, n = 6 replicates/condition). The Nsp1-Nsp2 cotransfection and Nsp1 transfection conditions were compared to the mock-transfected control using nonparametric one-way ANOVA with Dunn's correction and Nsp2 conditions were compared using nonparametric one-way ANOVA. The Nsp1-Nsp2 cotransfection and Nsp1 transfection conditions were compared using nonparametric t-test. * p < 0.05, ** p < 0.01, **** p < 0.0001, ns: not significant. Protein expression of Nsp1 and Nsp2 in individual transfection and cotransfection (C). HEK293T were seeded in 6-well plate. Twenty-four hours after, cells were transfected with control vector or Nsp1 expression vector or Nsp2 expression vector. Doses of 800 ng, 500 ng, and 100 ng of Nsp2 vector were used. Forty-eight hours post-transfection, cells were lysed in SDS PAGE 2X buffer and detected by Western Blot with an anti-FLAG (viral protein) and rabbit anti-tubulinβ (loading control) on radiological films. The Western Blot was performed two times and one representative experiment is shown.
Effects of Nsp1-Nsp2 polyprotein on IFNβ pathways. To circumvent the fact that Nsp1 prevented the expression of Nsp2, we designed a vector expressing a Nsp1-P2A-Nsp2 polyprotein (schematized in Figure 6A). The polyprotein vector was transfected into HEK293T cells and the polyprotein along with Nsp1 and Nsp2 individual proteins were detected by Western Blot (Figure 6B). The p65 and phosphorylated p65 expression levels were also quantified on the same cell lysate showing that Nsp2 increases the p65 phos- Figure 5. Effect of SARS-CoV-2-Nsp1 and Nsp1-Nsp2 cotransfection on SeV-induced IFNβ1 promoter activation. HEK293T cells were seeded, transfected, and simulated according to the procedure described above and the luciferase activity was measured then standardized as described for other reporter assays. Doses of 300 ng, 100 ng, 30 ng, and 3 ng per well of Nsp1 vectors for Nsp1 alone experimentation (A). HEK293T cells were seeded, transfected, and simulated according to the procedure described above and the luciferase activity was measured then standardized as described for other reporter assays. Doses of 300 ng of Nsp2 aand 100 ng of Nsp1 per well were used (B). All experiments were performed twice in triplicate and the compilation of the data is shown. Results are expressed as fold activation relative to the pcDNA mock-infected control (mean ± SD, n = 6 replicates/condition). The Nsp1-Nsp2 cotransfection and Nsp1 transfection conditions were compared to the mock-transfected control using nonparametric one-way ANOVA with Dunn's correction and Nsp2 conditions were compared using nonparametric one-way ANOVA. The Nsp1-Nsp2 cotransfection and Nsp1 transfection conditions were compared using nonparametric t-test. * p < 0.05, ** p < 0.01, **** p < 0.0001, ns: not significant. Protein expression of Nsp1 and Nsp2 in individual transfection and cotransfection (C). HEK293T were seeded in 6-well plate. Twenty-four hours after, cells were transfected with control vector or Nsp1 expression vector or Nsp2 expression vector. Doses of 800 ng, 500 ng, and 100 ng of Nsp2 vector were used. Forty-eight hours posttransfection, cells were lysed in SDS PAGE 2X buffer and detected by Western Blot with an anti-FLAG (viral protein) and rabbit anti-tubulinβ (loading control) on radiological films. The Western Blot was performed two times and one representative experiment is shown.
Effects of Nsp1-Nsp2 polyprotein on IFNβ pathways. To circumvent the fact that Nsp1 prevented the expression of Nsp2, we designed a vector expressing a Nsp1-P2A-Nsp2 polyprotein (schematized in Figure 6A). The polyprotein vector was transfected into HEK293T cells and the polyprotein along with Nsp1 and Nsp2 individual proteins were detected by Western Blot (Figure 6B). The p65 and phosphorylated p65 expression levels were also quantified on the same cell lysate showing that Nsp2 increases the p65 phosphorylation ( Figure 6B,C). Next, the polyprotein-encoding vector was co-transfected with PRD-II and IFNβ1 luciferase reporters.  Figure S2) expressed as normalized levels (Norm. vol. ×10 8 ) below the blot. Then, NF-κB p65 phosphorylation was normalized with p65 subunit expression and expressed as fold of induction relative to the mock-infected control. This experiment was performed twice, and one representative experiment is shown. Effects of Nsp1-Nsp2 on PRDII reporter activation, (D) IFNβ1-luc, (E). IFNβ1 essay was performed as described above with 20 or 80 fmol of vector encoding for Nsp1,   Figure S2) expressed as normalized levels (Norm. vol. ×10 8 ) below the blot. Then, NF-κB p65 phosphorylation was normalized with p65 subunit expression and expressed as fold of induction relative to the mock-infected control. This experiment was performed twice, and one representative experiment is shown. Effects of Nsp1-Nsp2 on PRDII reporter activation, (D) IFNβ1luc, (E). IFNβ1 essay was performed as described above with 20 or 80 fmol of vector encoding for Nsp1, Nsp2 or Nsp1-Nsp2 polyprotein. PRDII reporter essays were performed using 40 or 80 fmol of following vectors. All experiments were performed twice in triplicate and the compilation of the data is shown. Results are expressed as fold activation relative to the pcDNA mock-infected control (mean ± SD, n = 6 replicates/condition). The Nsp1-Nsp2 and Nsp1 conditions were compared to the mock control using nonparametric one-way ANOVA with Dunn's correction. The Nsp1-Nsp2 and Nsp1 conditions were compared together using nonparametric T-test. The Nsp2 transfection conditions were compared to the control using nonparametric one-way ANOVA with Dunn's correction. * p < 0.05, ** p < 0.01, *** p < 0.001.
As for Nsp2 alone, Nsp1 and Nsp2 coexpression activated the PRD-II ( Figure 6D). Under basal conditions, a significant increase in IFNβ promoter activity was observed in the presence of Nsp1 and Nsp2 ( Figure 6E). These results indicate that in the presence of Nsp1, the Nsp2 protein remains capable of activating the IFNβ1 promoter and NF-κB responsive element, suggesting that Nsp2 partially antagonizes Nsp1 effects while favoring the expression of inflammatory genes.

Discussion
In this study, we demonstrated that SARS-CoV-2 induced robust Ifnb1 gene transcription during the infection of K18-hACE2 mouse and human cells, but low to even undetectable IFNβ1 protein release. Ifnb1 gene expression is regulated by the coordinated actions of IRF3 and NF-κB, which are constitutively expressed in most cells [41,42]. Infected cells can therefore respond rapidly to incoming viruses by inducing Ifnb1 gene expression and IFNβ production even before viruses can deploy their anti-viral defense mechanisms. In the case of SARS-CoV-2, several viral proteins are reported to possess activities that antagonize the innate immune response such as type I IFN production. The most potent SARS-CoV-2 protein antagonizing the IFN response is the Nsp1 protein that induces a global shutdown of cellular mRNA translation [26,27]. The fact that several ISGs, such as Isg15, Igs56, and Mx1 are highly upregulated ( Figure 2B) during infection suggests that infected cells release sufficient IFNβ to induce the expression of genes associated with antiviral defense mechanisms. However, the establishment of an antiviral state is contingent on efficient ISG mRNA translation. To find out whether the anti-viral state is efficiently put in place, we examined Irf7 gene expression and IFNα production. IRF7, constitutively expressed in plasmacytoid dendritic cells and B cells and induced in many other cell types by viral infections, is the main transcription factor responsible for the activation of IFNα promoters [12,43]. Plasmacytoid dendritic cells and B cell types do not appear to express ACE2 nor transmembrane serine protease 2 (TMPRSS2) [6], suggesting that they cannot be directly infected by SARS-CoV-2. In response to infection by SARS-CoV-2, Irf7 is among the genes most highly induced in infected mice ( Figure 2B), suggesting that the recognition of the infection by cellular sensors and downstream signaling molecules is functional.
Despite the relatively low type I IFN production and the Ifnar downregulation, ISGs were highly expressed in mice following SARS-CoV-2 infection suggesting that ISG activation could also rely on an IFN-independent mechanism. Indeed, it had been shown that the human cytomegalovirus and the vesicular stomatitis virus could induce ISG expression in Ifnar KO mice by an IRF1-and an IRF3-dependent manner [44][45][46]. A study from Zhou, Z. et al. [30] also detected a robust ISG expression in the bronchioalveolar lavage fluids (BALFs) of patient with severe COVID-19 but failed to detect type I IFN expression Banerjee, A. et al. [29] recently reported that SARS-CoV-2 efficiently induced a type I IFN transcriptional response upon the infection of pulmonary epithelial cells. Our work supports similar findings. In contrast, work by others [26,27] clearly shows that this virus can also strongly inhibit the IFNβ1 protein expression. This apparent contradiction can be explained by the fact that certain studies measure RNA expression, while others evaluate protein synthesis. In fact, knowing that SARS-CoV-2 Nsp1 suppresses mRNA translation, the study of both mRNA and protein synthesis is necessary to reach proper conclusions [26,27]. In that regard, our work confirms that the infection of pulmonary epithelial cells by SARS-CoV-2 induced IFNβ gene expression and even potentiated the response to IFN-inducing agents such as poly(I:C) ( Figure 3D). When IFNβ production in the supernatant was assessed however, partial inhibition in IFNβ production was measured only when infection was combined with poly(I:C) stimulation ( Figure 3G).
Relative to type I IFNs and interleukins, CC and CXC chemokines were produced at high levels during infection by all SARS-CoV-2 strains in agreement with observations made in the lungs of humans infected with SARS-CoV-2 and suffering from severe COVID-19 [22,30]. In the lungs of patients with severe COVID-19, CXCL8 (IL-8) and CXCL1 (GROα) were the predominant CXC chemokines. Mice do not encode the Cxcl8 gene, but do have CXCL1, which was produced at high levels ( Figure 1C,D). As in humans, CCL2 was the most prominent CC chemokine produced during infection ( Figure 1C,D). While CXCL1 and CXCL8 (human) are mainly involved in neutrophil recruitment and activation, CCL2 is the main cytokine implicated in monocyte recruitment as well as TH1 polarization [47][48][49]. Put together, the concerted actions of these chemokines likely lead to a massive recruitment of leukocytes responsible of the ARDS observed in severe COVID-19 case [22,23].
The weak IL-1β production combined with the lack of modulation or downregulation of inflammasome effector and IL-18 suggest that the pathological inflammation following the K18-hACE2 mice did not rely on the inflammasome. This inflammation fueled by the inflammasome has been describe in the works of Zeng, J. et al. [50] and Sefik et al. [51] in AAV-hACE2-transduced mice. This discrepancy may be explained by the difference between the two mouse models. K18-hACE2 mice express the hACE2 receptor and support SARS-CoV-2 infection in a broad range of tissues, while in AAV-hACE2 mice, the expression of the hACE2 receptor is limited to the respiratory tract [52]. This major difference could impact the disease course and may change the inflammatory pattern. Also, in our study we euthanized mice on day 3 before the apparition of clinical symptoms, while other studies have made analyses at later time points.
While our data in infected mice suggest that inflammasome was the inflammation trigger, we show that early infection induces robust NF-κB activation that could drive the expression of chemokines such as CXCL1,9,10,11 or CCL2,3,4,5. Combined with the lack of type I IFN production, in in vitro conditions, this finding suggests that the virus skews the immune response toward an exaggerated inflammatory response rather than an antiviral response, as previously hypothesized [53]. This overwhelming inflammatory response represents a major determinant of the pathogenesis and morbidity observed during COVID-19. In support, the use of dexamethasone, a non-specific anti-inflammatory drug, has proven effective in reducing mortality and the length of hospital stay for patients with COVID-19 requiring oxygen supply [54].
Original to this work, we provided evidence that Nsp2 activates the NF-κB pathway and the IFNβ1 promoter. This effect of Nsp2 could be demonstrated when Nsp1-Nsp2 were generated from a common polyprotein as with the situation observed during infection ( Figure 6D,E). When expressed individually, Nsp1 prevented the efficient expression of Nsp2 ( Figure 5C).
In summary, the current study reveals that SARS-CoV-2 infection triggers the vigorous expression of antiviral and inflammatory genes. However, in both mouse and cell lines, IFN synthesis is sub-optimal, a consequence of the translational shutdown mediated by Nsp1. The IFN shutdown in infected cells is however incomplete, in part due to the action of other viral proteins such as Nsp2 that partially antagonize the actions of Nsp1. As such, our work highlights the importance of studying viral protein functions in the context of infection. The use of recombinant mutant viruses will be helpful in delineating the synergistic/antagonizing functions of non-structural and accessory proteins during infection. In contrast to IFN, elevated inflammatory gene expression did translate into the production of high levels of several inflammatory chemokines, many of which are regulated by NF-κB. Considering our results demonstrating that Nsp2 activates the NF-κB pathway, Nsp2 should be considered as a potential contributor to the pathogenesis observed during SARS-CoV-2 infection.