key: cord-336628-0evl3wnd
authors: Neufeldt, Christopher J.; Cerikan, Berati; Cortese, Mirko; Frankish, Jamie; Lee, Ji-Young; Plociennikowska, Agnieszka; Heigwer, Florian; Joecks, Sebastian; Burkart, Sandy S.; Zander, David Y.; Gendarme, Mathieu; El Debs, Bachir; Halama, Niels; Merle, Uta; Boutros, Michael; Binder, Marco; Bartenschlager, Ralf
title: SARS-CoV-2 infection induces a pro-inflammatory cytokine response through cGAS-STING and NF-κB
date: 2020-07-21
journal: bioRxiv
DOI: 10.1101/2020.07.21.212639
sha: 
doc_id: 336628
cord_uid: 0evl3wnd

SARS-CoV-2 is a novel virus that has rapidly spread, causing a global pandemic. In the majority of infected patients, SARS-CoV-2 leads to mild disease; however, in a significant proportion of infections, individuals develop severe symptoms that can lead to permanent lung damage or death. These severe cases are often associated with high levels of pro-inflammatory cytokines and low antiviral responses which can lead to systemic complications. We have evaluated transcriptional and cytokine secretion profiles from infected cell cultures and detected a distinct upregulation of inflammatory cytokines that parallels samples taken from infected patients. Building on these observations, we found a specific activation of NF-κB and a block of IRF3 nuclear translocation in SARS-CoV-2 infected cells. This NF-κB response is mediated by cGAS-STING activation and could be attenuated through STING targeting drugs. Our results show that SARS-CoV-2 curates a cGAS-STING mediated NF-κB driven inflammatory immune response in epithelial cells that likely contributes to inflammatory responses seen in patients and might be a target to suppress severe disease symptoms.

In late 2019 SARS-CoV-2 emerged as a highly infectious coronavirus that causes respiratory disease in humans, termed COVID-19. Since the initial identification, SARS-CoV-2 has spread around the world leading the World Health Organization to declare a pandemic. SARS-CoV-2 infection causes respiratory symptoms that range from mild to severe and can result in lasting lung damage or death in a significant number of cases 1 . One of the hallmarks of severe COVID-19 disease is low levels of type I interferons (IFNs) and overproduction of inflammatory cytokines such as IL-6 and TNF [2] [3] [4] [5] . This unbalanced immune response fails to limit virus spread and can cause severe systemic symptoms 3, 6 . Therapies aimed at modulating immune activation to attenuate the detrimental inflammatory response or promote an antiviral cytokine response represents an important avenue for treating patients with severe COVID-19.

SARS-CoV-2 is a plus-strand RNA virus that replicates its genome in the cytosolic compartment of the cells. Like all plus-strand RNA viruses, this replication process requires the production of a negative-strand RNA template in order to amplify the positive sense viral genome. This process and probably also the production of subgenomic RNAs of negative and positive polarity, produces double strand (ds)RNAs that can be sensed by cytosolic immune receptors (pattern recognition receptors, PRRs) that subsequently activate antiviral pathways 7 .

In addition to direct viral sensing, cells have also evolved ways to detect the indirect effects of virus infection, such as nuclear or mitochondrial damage caused by the heavy cellular burden of virus replication. Cytoplasmic DNA sensors including cGAS-STING, IFI16, or AIM2, recognize dsDNA from DNA viruses, but have also been shown to play an important role in RNA virus infection, either through directly recognising viral signatures or through sensing of cellular DNA released from mitochondria or nuclei due to cellular stress (reviewed in 8, 9 ) .

Substrate recognition by either RNA or DNA sensors leads to signalling cascades that activate two major branches of the innate immune response, the type I/III IFN response and the inflammatory cytokine response. The type I/III IFN pathways are directly involved in protecting neighboring cells from virus spread and are vital for the immediate cell-intrinsic antiviral response. The inflammatory cytokine response is involved in recruitment and activation of immune cells, which is required to initiate an adaptive immune response.

Due to the effective nature of innate immune sensing and responses plus-strand RNA viruses have evolved numerous ways to limit or block these cellular pathways. For many viruses, the initial line of defence is to hide viral replication intermediates within membrane compartments that block access to cytosolic PRRs, such as RIG-I or MDA5 10 . In the case of coronaviruses, this is achieved through the formation of replication organelles composed predominantly of double membrane vesicles, within which viral RNA replication occurs 11, 12 .

Coronaviruses can also evade recognition by immune receptors through modification of viral RNA to resemble host mRNA [13] [14] [15] [16] . In addition to these passive immune evasion strategies, coronaviruses utilize various mechanisms to actively target and block key immune sensors or signalling molecules (reviewed in 17, 18 ). For SARS-CoV-1, a closely related virus, several viral proteins have been shown to block RIG-I/MDA5 sensing, as well as the downstream activation of TBK1 and IRF3 [19] [20] [21] [22] [23] [24] . SARS-CoV-1 also efficiently blocks IFN receptor and JAK-STAT signalling to stop downstream immune activation 19, [25] [26] [27] . Additionally, the SARS-CoV-1 papain-like protease (PLP) has been shown to interfere with cGAS-STING activation also limiting activation of innate immune pathways 28 . The combination of these actions can lead to an imbalance between proinflammatory and antiviral immune responses.

Although numerous immune evasion mechanisms have been characterized for other pathogenic coronaviruses, it remains to be determined whether similar processes exist for SARS-CoV-2. Given the homology of SARS-CoV-1 to SARS-CoV-2, they may have many conserved antagonistic strategies, however, key differences in infection and disease could suggest divergent pathways. Early reports on SARS-CoV-2 demonstrated that infection is highly sensitive to type I/III IFN treatment [29] [30] [31] [32] . In combination with the low levels of IFN reported to be secreted in severe cases, this suggests that like SARS-CoV-1, SARS-CoV-2 infection actively blocks immune activation. Transcriptomic analyses of SARS-CoV-2 infected cells generated ambiguous results on the induction of type I/III IFNs and the subsequent expression of IFN stimulated genes (ISGs). On the one hand, it was shown that SARS-CoV-2 triggers only an attenuated immune response suggesting a block in PRR signaling pathways, which would parallel SARS-CoV-1 and MERS-CoV infections 29 . On the other hand, several studies argue for a strong induction of IFN responses in both lung and intestinal infection models 30, 33 . Additionally, proteomics approaches determining SARS-CoV-2 protein interactions with host factors in exogenous expression conditions revealed several interactions with key immune regulators including MAVS, TBK1 and several co-factors involved in IRF3 activation 34, 35 . However, many of these findings are still observational leaving the mechanisms of SARS-CoV-2 innate immune response modulation unresolved.

Here, we report the transcriptomic profiles derived from SARS-CoV-2 infected human lung cells showing a specific bias towards an NF-κB mediated inflammatory response and a restriction in the TBK1 specific IRF3/7 activation and subsequent IFN response. Consistently, secreted cytokine profiles from both severe COVID-19 patients and SARS-CoV-2 infected lung epithelial cells, were enriched for pro-inflammatory cytokines and lacked type I/III IFNs. We also demonstrate that SARS-CoV-2 infection leads specifically to NF-κB but not IRF3 nuclear localization and that poly(I:C)-induced pathway activation is attenuated in infected cells.

Finally, we show that the cGAS-STING pathway is activated by SARS-CoV-2 infection, leading to a specific NF-κB response and that inflammatory cytokine upregulation can be mitigated by STING inhibitory drugs. These results provide insight into how innate immune responses are modulated by SARS-CoV-2 in epithelial cells likely contributing to the strong inflammatory responses observed in severe COVID-19 cases.

SARS-CoV-2 predominantly infects airway and lung tissue in infected individuals. In order to determine the effects of SARS-CoV-2 on human lung epithelial cells, Calu-3 and A549 cells were infected with SARS-CoV-2 and virus growth, as well as host transcriptional architecture was determined over a time course of infection. In contrast to Calu-3, A549 cells lack endogenous expression of the major SARS-CoV-2 entry receptor ACE2 and, hence, are not naturally permissive to SARS-CoV-2 infection 36 . We therefore used an engineered A549 cell line stably expressing ACE2 (A549-ACE2), which is susceptible to SARS-CoV-2 infection 12, 37 . For both, A549-ACE2 and Calu-3 cells, we observed an increase in intracellular viral RNA starting at 4 h post infection, which continued to increase up to 24 h post infection ( Fig. 1a-b ).

Increased extracellular virus RNA was observed starting at 6 h post infection which was paralleled by the release of infectious virus (Fig. 1a-c) . The levels of viral RNA, production of infectious virus and virus spread were significantly higher in Calu-3 cells compared to A549- 1a ). We did not observe an overall decrease in total mRNA quality or large differences in probe intensity, thus showing no indication that SARS-CoV-2 infection causes a general transcriptional shutdown. Importantly, we observed a high degree of overlap between top significantly upregulated or downregulated genes from both cell lines (Extended Data Fig. 1b- c). However, in A549-ACE2 cells, there was less overall change in transcript levels following infection (Fig. 2b , and Extended Data Fig. 1a and 1d) , which is likely due to the lower levels of infection (~40% vs 80% at 24 h of A549-ACE2 vs Calu-3, respectively) (Fig. 1e) . Gene-set enrichment analysis of the transcriptional changes using curated "Hallmark" pathways showed were also observed at early time points after infection of Calu-3 cells as well as high levels of the chemokine CXCL10/IP-10 at 24 h post infection (Fig. 3e) . In contrast, no consistently detectable upregulation was found for IFNα, IFNβ, IFNγ and IFNλ2/3, whereas a rather moderate increase of IFNλ1 expression was found, but only at the late time point of Calu-3 cell infection. Together, these results corroborate published data showing that, in severe cases, SARS-CoV-2 infection preferentially induces a pro-inflammatory cytokine production with little activation of the antiviral responses. Additionally, these data indicate that infected epithelial cells secrete cytokines that can contribute to induction of tissue-level inflammation.

To confirm that the lack of IFN response in Calu-3 or A549-ACE2 cells infected with SARS-CoV-2 was not due to defects in the activation of innate immune pathways, we To test if IFNs could limit virus replication even after establishment of infection, A549-ACE2 cells were treated with high levels of various IFNs at the time point of infection or 6 h thereafter. Pre-treatment with type I IFNs, serving as control, blocked virus infection, whereas co-or post-treatment had significantly less effects ( Fig. 4e-f ). Of note, the 2-fold decrease in virus replication following post-treatment with type I IFNs likely represents a block in virus spread following the first round of infection, as only ~40-50% of cells were observed to be infected at the 6 h time point (Fig. 1e) . Together, these observations suggest that SARS-CoV-2 likely supresses the production of IFNs and antiviral ISGs, and that it furthermore rapidly and potently blocks IFN signalling in infected cells.

The high levels of inflammatory gene activation and the lack of IFNs and ISGs in response to SARS-CoV-2 infection lead us to investigate which transcription factors of the cell- 

To determine if SARS-CoV-2 can actively block immune stimulation through cytosolic 

To determine the source of the SARS-CoV-2 induced inflammatory response or downstream immune activation, we evaluated the effects of innate immune receptor knockout or overexpression. We first looked at RNA receptors that have previously been described to Together, these data indicate that recognition of viral RNA via cellular RNA sensors is not involved in NF-κB activation in SARS-CoV-2 infected cells.

Although cGAS is a sensor of cytosolic DNA, induction of the cGAS-STING-signaling axis leading to activation of NF-κB and IRF3 has been reported for several RNA virus infections, most likely through cellular stress responses 9 . To determine whether the cGAS-STING pathway is triggered in SARS-CoV-2 infection, we first evaluated changes in localization of cGAS or STING in infected cells. Indeed, both cGAS and STING were observed to re-localize to perinuclear clusters in infected cells, indicative of activation ( Fig. 6a-b) . Costaining for cGAS and dsDNA in infected cells also showed that dsDNA colocalized with cGAS in infected cells (Fig. 6c) . Additionally, we observed that, unlike in poly(I:C)-mediated activation of RLRs, SARS-CoV-2 infection did not interfere with activation of the cGAS-STING pathway by dsDNA transfection (Extended Data Fig. 3c-d) .

To confirm that cGAS-STING activation is involved in the observed induction of proinflammatory cytokines, we examined the effects of pharmacologically blocking STING in SARS-CoV-2 infected cells. One hour post infection, cells were treated with the STING specific inhibitor H-151, the TBK1 inhibitor amlexanox (Amx), or DMSO. At 24 h post infection, we observed a significant decrease in the levels of TNF mRNA in infected cells treated with H-151 compared to DMSO treated cells (Fig. 6d) , both, in A549-ACE2 and Calu-3. This decrease was not observed for Amx treated cells. Infection levels and cell viability were not significantly affected at the effective concentration (Extended Data Fig. 3e -g). Together

these results indicate that SARS-CoV2-infection triggers the cGAS-STING pathway, leading to NF-κB-mediated induction of pro-inflammatory cytokines, and that this response can be controlled with STING inhibitors.

Although STING activation is usually associated with both, NF-κB as well as IRF3 activation, several reports have suggested that interfering with proper translocation of STING from the ER to Golgi compartments can selectively stimulate the NF-κB pathway 39, 40 . To test whether this is the case in SARS-CoV-2 infected cells, we determined the localization of STING relative to Golgi markers by microscopy. Consistent with previous reports, in cells transfected with dsDNA, we observed STING translocation to the Golgi compartment (Extended Data Fig. 3h) . No significant colocalization of Golgi markers and STING were observed in either mock or SARS-CoV-2 infected cells, suggesting that STING translocation may be impaired (Fig. 6e-f ). Moreover, we found that clusters of STING in SARS-CoV-2 infected cells colocalized with viral nucleocapsid (N) protein ( Fig. 6g-h; Extended Data Fig.   3i ). Together, these results suggest that STING is activated in SARS-CoV-2 infection, but inhibited from translocating to the Golgi, leading to a specific NF-κB inflammatory response in infected cells.

In this study, we combine transcriptional profiling and cytokine secretion analyses to characterize the pro-inflammatory response induced by SARS-CoV-2 infection, and evaluate the virus-induced signalling pathways mediating this response. We report that both virus- Further evaluation of the SARS-CoV-2 induced pro-inflammatory response showed a specific induction of NF-κB, but not of IRF3 or the subsequent IFN signalling. NF-κB can be activated through numerous immune or stress stimuli including the ER stress responses or increase in cytosolic reactive oxygen species, as well as through detection of cytosolic DNA released from the nucleus or mitochondria (reviewed in 8, 45, 46 ). Our results indicate that cGAS-STING activation is a major contributor to NF-κB activation in SARS-CoV-2 infected cells.

Since cGAS is a dsDNA sensor that would not be expected to directly recognize SARS-CoV-2 RNA, it is likely that cellular stress or cytokine responses induced by the infection leads to nuclear or mitochondrial DNA release which is sensed by cGAS [47] [48] [49] [50] [51] . Similar activation of cGAS-STING has been observed for other positive strand RNA viruses including flaviviruses and both SARS-CoV-1 and NL63 coronaviruses (reviewed in 9 ). For the coronaviruses, STING activation is perturbed through the action of the viral PLP leading to an inhibition of STING oligomerization and downstream activation of TBK1 and IRF3 28, 52, 53 . Intriguingly, mechanisms for cGAS-STING modulation seem to be different in SARS-CoV-2 infected cells, highlighting a major immunological difference between these related viruses. Of note, the activation of NF-κB through cGAS-STING does not exclude other sources of NF-κB activation.

Indeed, we observed increases in FOS/JUN and ATF3 mRNA levels in infected cells suggesting activation of multiple cell stress pathways 54 . Moreover, pharmacological inhibition of STING did not completely block TNF upregulation, further indicating a role for other sources of NF-κB-activation. We speculate that therapeutic inhibition of multiple NF-κB activation pathways could serve to further reduce pro-inflammatory responses in SARS-CoV-2 infected cells.

The selective activation of NF-κB, rather than a general block in all immune activation pathways, indicates a pro-viral role for NF-κB signalling. In addition to functions in inflammation, NF-κB is also important for cell survival and proliferation 55 . These NF-κB cell survival signals could be beneficial for the virus by promoting vitality in cells in order to facilitate efficient and sustained virus replication and spread. Mechanisms for NF-κB pathway interference have been reported for numerous DNA and RNA viruses [56] [57] [58] . Selective modulation of the cGAS-STING pathways may allow SARS-CoV-2 to promote an NF-κB mediated cell survival signal while limiting ISG induction.

Classical cGAS-STING induction activates not only NF-κB, but also TBK1 and IRF3 pathways. We envisage several mechanisms that could contribute to the selective NF-κB activation. First, the virus could actively block TBK1 activation in infected cells. Indeed, protein interaction studies indicate that viral NSP13 and NSP15 proteins interact with TBK1 or its adaptor proteins 34 . Additionally, a block in TBK1 activation has been reported for both SARS-CoV-1 and MERS-CoV infections and early reports demonstrate a lack of TBK1 phosphorylation in SARS-CoV-2 infected cells 29, 59 . Moreover, our results support a model where SARS-CoV-2 infection prevents activated STING from translocating from the ER to the Golgi. Activation of STING at the ER has been shown to be sufficient for NF-κB activation but not for TBK1 activation and the subsequent IRF3 phosphorylation 40, 60 . It may be that fragmentation of the Golgi by SARS-CoV-2 infection leads to an impairment of STING translocation to the ERGIC. Consistently, our transcriptomic analysis shows impairment in protein secretion pathways, specifically including downregulation of several COP coatomer proteins involved in ER to Golgi transport. Alternatively, SARS-CoV-2 proteins could actively block cGAS-STING translocation. Colocalization between STING and N protein in infected cells suggests a direct role for N protein in limiting STING translocation. A similar mechanism has been suggested for murine cytomegalovirus, where viral m152 protein associates with STING and limits exit from the ER, thereby promoting an NF-κB specific response.

Interestingly, pathway analysis of our microarray data indicate that cytokine transcriptional responses from SARS-CoV-2 infected cells resemble signatures from human cytomegalovirus infected cells. Further experimentation is required to define the precise mechanisms of STING activation in SARS-CoV-2 infected cells and to determine whether viral proteins are directly associating with components of the cGAS-STING pathway.

The majority of documented SARS-CoV-2 infections lead to mild or no symptoms, indicating that even the observed low level of antiviral pathway activation induced by infected cells can be sufficient to limit and resolve the infection. On the other hand, in patients with underlying conditions or attenuated immune responses, these antiviral responses do not limit virus replication and a sustained virus load eventually leads to a long term inflammatory response. In these latter cases, one important avenue of treatment is to modulate the immune response in order to alleviate hyper-inflammation. In addition to other immune modulators that are currently being used or clinically evaluated (eg. IL-6 inhibitors or corticosteroids) 61-65 , our results indicate that disease severity might be suppressed at the epithelial cell level through the use of cGAS-STING inhibitors or through blocking NF-κB mediated inflammatory responses.

In this respect, NF-κB inhibitors analogous to CAPE or parthenolide, prolonging survival of SARS-CoV-1 infected mice 44 , might help to reduce the disease burden imposed by COVID-

supplemented with Glutamax (Gibco),10% fetal bovine serum, 100 U penicillin/ml, 100 µg streptomycin/ml, 2 mM L-glutamine and nonessential amino acids. 

SARS-CoV-2 stocks were produced using VeroE6 cell line. Passage 2 BavPat1/2020 (MOI: 0.01) strain was used to generate the seed virus (passage 3). After 48 h the supernatant was harvested, cell debris was removed by centrifugation at 1,000xg for 5 min and supernatant filtered with a 0.45 mm pore-size filter. Passage 4 virus stocks were produced by using 500 µl of the seed virus (passage 3) to infect 9E+06 VeroE6 cells. The resulting supernatant was harvested, filtered 48 h later as described above and stored in aliquots at -80°C. Stock virus titers were determined by plaque assay.

Total RNA was isolated from cells or supernatants using the NucleoSpin RNA extraction kit (Macherey-Nagel) according to the manufacturer's specification. cDNA was synthesized from the total RNA using the high capacity cDNA reverse transcription (RT) kit (ThermoScientific) according to the manufacturer's specifications. Each cDNA sample was diluted 1:15 in nuclease free H2O prior to qPCR analysis using specific primers and the iTaq Universal SYBR green mastermix (Bio-Rad). Primers for qPCR were designed using Primer3 software and include: Gene set enrichment analysis was performed according to Subramanian et al. 67 . We use the practical R implementation "fgsea" 68 and the hallmark pathway gene set published by Liberzon et al. 69 . The barcode plot implementation was inspired by Zhan et al. 70 .

Primary antibodies and specific dilutions used for western blot or immunofluorescence 

After infection with SARS-CoV-2 cells were fixed with 6% formaldehyde solution, washed twice with phosphate buffered saline (PBS) and permeabilized with 0.2% Triton X-100 in PBS.

Next, the Triton X-100 solution was replaced with 2.5% (w/v) milk solution (in PBS) and cells were blocked for 1 h at room temperature. Primary antibodies were diluted in 2.5% milk solution and samples were incubated with primary antibodies for 1 h. After washing three times with PBS, samples were incubated with Fluorophore-conjugated secondary antibodies, diluted in milk solution, for 30 min. After washing three times with PBS samples were mounted in Fluoromount G solution containing DAPI (Southern biotech) for DNA staining. Microscopic analyses were conducted with a Nikon Eclipse Ti microscope (Nikon, Tokio, Japan) or a Leica SP8 confocal microscope (Leica) for the subcellular localization analyses.

For quantification of the nuclear translocation of NF-κB p65/RELA or IRF3, nuclei were segmented using DAPI signal first. Secondly, the segmented nucleus was dilated and finally dilated nucleus was subtracted by original nucleus mask to detect perinuclear fluorescent signal. To determine SARS-CoV-2 infected cells, dsRNA intensity was measured within the perinuclear area. The status of NF-κB or IRF3 nuclear signals was determined based on the ration between perinuclear intensity divided by nuclear intensity.

For pretreatment experiments, cells were treated for 6 h with serial dilution of IFNα2 

For Calu-3 cell stimulation (Fig. 4a,b ) cells were transfected with the indicated amount of poly(I:C) using lipofectamine 2000 reagent as per the manufacturer's protocol. 16 h after transfection, total RNA was isolated and RT-qPCR was used to determine transcript levels as describe above.

For transfection in SARS-CoV-2 infected cells ( Fig. 5 and extended data Fig. 3) , cells seeded in 24-well plates were infected with SARS-CoV-2 at MOI=5 for 6 h. Cells were then transfected with poly(I:C) or herring DNA (500ng/well) using lipofectamine 2000 reagent as per the manufacturer's protocol. 6 h after transfection, cells were either fixed with 4% paraformaldehyde and processed for immunofluorescence, or total RNA was isolated for RT-qPCR analysis as described above.

Infected and mock cells were washed with PBS and lysed with 100 µl of sample buffer (120 For cytopathic effect (CPE) assays, Calu-3 or A549-ACE2 cells were plated into 96 well plates. Cells were infected with SARS-CoV-2 for 48 h followed by fixation in 5%

formaldehyde for 1 h. Cells were then stained with 1% crystal violet solution and scanned.

Patient sera were collected and stored at -80°C until cytokine measurement. 

A549-ACE2 cells were infected with SARS-CoV-2 for 16 h. a, Cells were fixed and stained with antibodies specific for IRF3 (green), p65/RELA (red) and dsRNA (grey). Turquoise arrows point to cells showing p65/RELA nuclear accumulation. Scale bars, 10 µm. b, Graph shows the mean nuclear accumulation of IRF3 and p65/RELA for images from cells treated as in panel (a). c, Infected cells we either untreated or transfected with Poly(I:C) 6 h post infection. Cells were fixed and stained with antibodies against dsRNA followed by analysis using widefield microscopy. Graph shows the average percent of infected cells for 7 fields of view collected from 3 independent experiments determined by dsRNA fluorescence signal (N>200 cells per field of view). d-e, Cells were infected with SARS-CoV-2 for the indicated times. Cells were lysed and levels of given proteins were determined by western blot using monospecific primary antibodies. e, Western blot signals for phosopho-p65/RELA (pRELA) were quantified and compared to the corresponding total p65/RELA proteins levels. Graph shows the mean and SEM for pRELA vs. total p65/RELA protein levels for 3 independent experiments. f-i, A549-ACE2 cells were infected with SARS-CoV-2 for 6 h followed by transfection with Poly(I:C) and incubated for 4 h. f. Total RNA was isolated and the mRNA levels of IFIT1 and IFIT3 were determined by RT-qPCR. Graphs show the mean and SEM from 3 independent experiments. g, Cells were fixed and stained with antibodies specific for IRF3 (green), p65/RELA (red) and dsRNA (grey). Magenta arrows point to cells with nuclear accumulation of both IRF3 and p65/RELA and turquoise arrows point to cells with only p65/RELA nuclear signal. Scale bars, 10 µm. h-i, Quantification of nuclear translocation of fluorescence signals from p65/RELA or IRF3 from 7 fields of view collected from 2 independent experiments conducted as in panel (g). Graphs show the mean number of cells with nuclear signal in uninfected or infected cells in either the Mock or SARS-CoV-2 treatment conditions. Infection was determined by the dsRNA signal. Quantification was done using an in house Fiji macro.

cells. a-c, A549-ACE2 cells were infected with SARS-CoV-2 for 16 h followed by fixation and staining with the indicated antibodies. Cells were analyzed by confocal microscopy. Scale bars 10 µm (a-b), 5 µm (c). d, Cells were infected with SARS-CoV-2. One hour after infection cells were treated with the indicated drugs at the given concentrations. Total RNA was isolated and the TNF mRNA transcript levels were determined by RT-qPCR. Graph shows the average fold change and SEM for TNF transcript levels compared to DMSO treated cells for 4 independent experiments. e, A549-ACE2 cells were infected with SARS-CoV-2 for 16 h followed by fixing and staining with the indicated antibodies. Cells were analyzed by confocal microscopy. Scale bars, 5 µm. f, Pearson's correlation coefficient for fluorescence signals pertaining to STING and TGN46 were calculated for 7 fields of view over 2 independent experiments (N>20 cells). g, A549-ACE2 cells were infected SARS-CoV-2 for 16 h followed by fixing and staining with antibodies specific to STING (green) or N protein (red). Cells were analyzed by confocal microscopy. Scale bars, 10 µm upper panels, 1 µm for inset that is indicated with a rectangle in middle right panel. h, Pearson's correlation coefficient for fluorescence signal pertaining to STING and N protein were calculated for 6 fields of view over 2 independent experiments (N>20 cells).

A systematic review of pathological findings in COVID-19: a pathophysiological timeline and possible mechanisms of disease progression

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

COVID-19: consider cytokine storm syndromes and immunosuppression

Immune cell profiling of COVID-19 patients in the recovery stage by single-cell sequencing

COVID-19 severity correlates with airway epithelium-immune cell interactions identified by single-cell analysis

Impaired type I interferon activity and exacerbated inflammatory responses in severe Covid-19 patients. medRxiv

RIG-I-like receptors: their regulation and roles in RNA sensing

Molecular mechanisms and cellular functions of cGAS-STING signalling

The cGAS-STING Defense Pathway and Its Counteraction by Viruses

Architecture and biogenesis of plus-strand RNA virus replication factories

SARS-coronavirus replication is supported by a reticulovesicular network of modified endoplasmic reticulum

SARS-CoV-2 structure and replication characterized by <em>in situ</em> cryo-electron tomography. bioRxiv

In vitro reconstitution of SARS-coronavirus mRNA cap methylation

Functional screen reveals SARS coronavirus nonstructural protein nsp14 as a novel cap N7 methyltransferase

Coronavirus endoribonuclease targets viral polyuridine sequences to evade activating host sensors

Multiple enzymatic activities associated with severe acute respiratory syndrome coronavirus helicase

Immunology of COVID-19: Current State of the Science

SARS and MERS: recent insights into emerging coronaviruses

Severe acute respiratory syndrome coronavirus open reading frame (ORF) 3b, ORF 6, and nucleocapsid proteins function as interferon antagonists

Severe acute respiratory syndrome coronavirus M protein inhibits type I interferon production by impeding the formation of TRAF3.TANK.TBK1/IKKepsilon complex

Severe acute respiratory syndrome coronavirus papain-like protease ubiquitin-like domain and catalytic domain regulate antagonism of IRF3 and NF-kappaB signaling

Regulation of IRF-3-dependent innate immunity by the papainlike protease domain of the severe acute respiratory syndrome coronavirus

SARS-coronavirus open reading frame-9b suppresses innate immunity by targeting mitochondria and the MAVS/TRAF3/TRAF6 signalosome

Coronavirus nonstructural protein 15 mediates evasion of dsRNA sensors and limits apoptosis in macrophages

The SARS Coronavirus 3a protein causes endoplasmic reticulum stress and induces ligand-independent downregulation of the type 1 interferon receptor

Severe acute respiratory syndrome coronavirus ORF6 antagonizes STAT1 function by sequestering nuclear import factors on the rough endoplasmic reticulum/Golgi membrane

Severe acute respiratory syndrome coronavirus evades antiviral signaling: role of nsp1 and rational design of an attenuated strain

Coronavirus papain-like proteases negatively regulate antiviral innate immune response through disruption of STING-mediated signaling

Imbalanced Host Response to SARS-CoV-2 Drives Development of COVID-19

Critical Role of Type III Interferon in Controlling SARS-CoV-2 Infection in Human Intestinal Epithelial Cells

SARS-CoV-2 is sensitive to type I interferon pretreatment. bioRxiv

Antiviral activities of type I interferons to SARS-CoV-2 infection

Bulk and single-cell gene expression profiling of SARS-CoV-2 infected human cell lines identifies molecular targets for therapeutic intervention. bioRxiv

A SARS-CoV-2 protein interaction map reveals targets for drug repurposing

Multi-level proteomics reveals host-perturbation strategies of SARS-CoV-2 and SARS-CoV. bioRxiv

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Microscopy-based assay for semi-quantitative detection of SARS-CoV-2 specific antibodies in human sera. bioRxiv

Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles

The herpesviral antagonist m152 reveals differential activation of STING-dependent IRF and NF-kappaB signaling and STING's dual role during MCMV infection

Modular Architecture of the STING C-Terminal Tail Allows Interferon and NF-kappaB Signaling Adaptation

NF-kappaB, inflammation, immunity and cancer: coming of age

Severe acute respiratory syndrome coronavirus ORF3a protein activates the NLRP3 inflammasome by promoting TRAF3-dependent ubiquitination of ASC

Inhibition of NF-kappaB-mediated inflammation in severe acute respiratory syndrome coronavirus-infected mice increases survival

Crosstalk of reactive oxygen species and NF-kappaB signaling

The Crosstalk of Endoplasmic Reticulum (ER) Stress Pathways with NF-kappaB: Complex Mechanisms Relevant for Cancer

Mitochondrial DNA stress primes the antiviral innate immune response

Dengue virus activates cGAS through the release of mitochondrial DNA

Collateral Damage during Dengue Virus Infection: Making Sense of DNA by cGAS

Interleukin-1beta Induces mtDNA Release to Activate Innate Immune Signaling via cGAS-STING

cGAS surveillance of micronuclei links genome instability to innate immunity

Deubiquitinating and interferon antagonism activities of coronavirus papain-like proteases

SARS coronavirus papain-like protease inhibits the type I interferon signaling pathway through interaction with the STING-TRAF3-TBK1 complex

Cell Signaling and Stress Responses

More to Life than NF-kappaB in TNFR1 Signaling

NF-kappaB and virus infection: who controls whom

Who's Driving? Human Cytomegalovirus, Interferon, and NFkappaB Signaling

Manipulation of Non-canonical NF-kappaB Signaling by Non-oncogenic Viruses

The Global Phosphorylation Landscape of SARS-CoV-2 Infection

Cyclic dinucleotides trigger ULK1 (ATG1) phosphorylation of STING to prevent sustained innate immune signaling

Can we use interleukin-6 (IL-6) blockade for coronavirus disease 2019 (COVID-19)-induced cytokine release syndrome (CRS)?

IL-6 Inhibitors in the Treatment of Serious COVID-19: A Promising Therapy

Dexamethasone for COVID-19? Not so fast

Can steroids reverse the severe COVID-19 induced "cytokine storm

Early Short Course Corticosteroids in Hospitalized Patients with COVID-19

limma powers differential expression analyses for RNAsequencing and microarray studies

GSEA-P: a desktop application for Gene Set Enrichment Analysis

Fast gene set enrichment analysis. bioRxiv

The Molecular Signatures Database (MSigDB) hallmark gene set collection

MEK inhibitors activate Wnt signalling and induce stem cell plasticity in colorectal cancer

EBImage--an R package for image processing with applications to cellular phenotypes

Natural killer cells are scarce in colorectal carcinoma tissue despite high levels of chemokines and cytokines

Calu-3 cells show separation between infection and mock but the A549-ACE2 cells have less separation due to lower levels of infection. b, Plot comparing enriched genes between the two cell lines. c, Venn diagram of significantly (FDR < 10 %) upregulated (top) or downregulated (bottom) genes comparing Calu-3 and A549-ACE2 cells

Calu-3 or A549-ACE2 cells were treated with increasing concentrations of given IFNs; highest concentration corresponds to 100X the EC50 value (determined for ISG activation by qPCR) for each IFN. Six hours post treatment cells were infected with SARS-CoV-2 (MOI=5) and 48 h thereafter

The graphs show the means and SEMs compared to HPRT mRNA levels for 3 independent experiments. c-d, Cells were infected with SARS-CoV-2 or mock-infected for 6 h, then transfected with Poly(I:C) or herring DNA and 4 h thereafter, total RNA was isolated and the mRNA levels of given immune genes were determined by RT-qPCR. c, The graphs show the mean mRNA levels of IFIT1 or TNF, corrected for HPRT, for 4 independent experiments for mock-infected cells. d, Analogous to panel c, but for SARS-CoV-2 infected cells. Values were corrected for mock of the same treatment for 3 independent experiments. e-g, Cells were infected with SARS-CoV-2 and, 1 h later, cells were treated with the given drugs or DMSO only. e, total RNA was isolated and the viral RNA levels were determined by RT-qPCR. The graph shows the mean and SEM for 3 independent experiments corrected for HPRT. f, Cells were fixed and stained with antibodies specific for dsRNA. Graph shows the mean percent of infected cells from 3 different experiments for each condition. g, Graph shows the average fold change and SEM for the number of cells for each drug treatment compared to the DMSO treated cells. h, Cells either un-transfected or transfected with herring DNA for 6 h were fixed, stained with the indicated antibodies and examined by confocal microscopy. Scale bar, 5 µm. i, A549-ACE2 cells were infected with SARS-CoV-2 for 16 h followed by fixing and staining with the indicated antibodies

We thank all members of the Molecular Virology department at Heidelberg University for helpful discussions and support during different stages of the COVID-19 related lockdown. We also thank Dr. Monica Boxberger for taking patient samples, Sandra Wüst for excellent 

The authors declare no competing interests