SARS-CoV-2 triggers inflammatory responses and cell death through caspase-8 activation

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can lead to respiratory illness and multi-organ failure in critically ill patients. Although the virus-induced lung damage and inflammatory cytokine storm are believed to be directly associated with coronavirus disease 2019 (COVID-19) clinical manifestations, the underlying mechanisms of virus-triggered inflammatory responses are currently unknown. Here we report that SARS-CoV-2 infection activates caspase-8 to trigger cell apoptosis and inflammatory cytokine processing in the lung epithelial cells. The processed inflammatory cytokines are released through the virus-induced necroptosis pathway. Virus-induced apoptosis, necroptosis, and inflammation activation were also observed in the lung sections of SARS-CoV-2-infected HFH4-hACE2 transgenic mouse model, a valid model for studying SARS-CoV-2 pathogenesis. Furthermore, analysis of the postmortem lung sections of fatal COVID-19 patients revealed not only apoptosis and necroptosis but also massive inflammatory cell infiltration, necrotic cell debris, and pulmonary interstitial fibrosis, typical of immune pathogenesis in the lung. The SARS-CoV-2 infection triggered a dual mode of cell death pathways and caspase-8-dependent inflammatory responses may lead to the lung damage in the COVID-19 patients. These discoveries might assist the development of therapeutic strategies to treat COVID-19.


INTRODUCTION
Since its emergence in December 2019, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused a pandemic transmission in the world. Virus infection causes severe respiratory illness in the patients, called the coronavirus disease 2019 (COVID- 19). 1 As of 24 July 2020, SARS-CoV-2 had infected over 15 million people worldwide and had claimed the death of over 630,000 people. Severe disease manifested by fever and pneumonia, leading to acute respiratory distress syndrome (ARDS), has been described in up to 20% of COVID-19 cases. 2 Elevated levels of serum interleukin (IL)-6 and other inflammatory cytokines correlate with respiratory failure, ARDS, and adverse clinical outcomes. 3,4 Inflammatory responses from infected cells may further induce infiltration of immune cells into the lung causing overproduction of pro-inflammatory cytokines resulting in severe lung damage and multi-organ dysfunction. 5 Indeed, virusinduced cytokine storm and sepsis has claimed the death of 28% of fatal COVID-19 cases. 6 Understanding the underlying molecular mechanisms of virus-induced inflammatory responses would be highly valuable for developing effective therapeutic strategies for treating COVID-19 patients.
Caspase-8, which was previously viewed exclusively as an apoptotic caspase, has now emerged as a master regulator of the three major cell death pathways, including apoptosis, pyroptosis, and necroptosis. 7 Recently, it was reported that caspase-8 can induce the expression of pro-inflammatory cytokines and process pro-IL-1β and IL-18 in the same way as caspase-1, resulting in the release of bioactive cytokines, through either pyroptosis or necroptosis. 8 The role of caspase-8 in mediating inflammatory responses has been reported in the context of infection with fungal pathogens such as Candida albicans and Aspergillus fumigatus. 9 Whether caspase-8 activation can also mediate virusinduced inflammatory responses has not been reported before.
Necroptosis is an immunogenic cell death pathway that can eliminate virus-infected cells and mobilize both innate and adaptive immune responses to restrict virus replication. 10 The benefits of necroptosis to the host, however, may sometimes be outweighed by the potentially deleterious hyper-inflammatory consequences of activating this death pathway in pulmonary and other tissues. 11 For example, the influenza A virus (IAV) infectioninduced necroptosis in airway epithelial cells is associated with lung damage and hyper-inflammatory responses. Similar with IAV, SARS-CoV-2 can cause severe lung damage and disease-associated hyper-inflammatory responses. 12 Whether SARS-CoV-2 infection also induces necroptosis pathway to trigger cell death and inflammatory responses are currently unknown.
Here we report that SARS-CoV-2 infection of lung epithelial cells induces caspase-8 activation that triggers cell apoptosis and processing of inflammatory cytokines including IL-1β into the bioactive form. IL-1β was then secreted through the SARS-CoV-2induced necroptosis pathway resulting in inflammatory responses. The SARS-CoV-2 infection induced a dual mode of cell death pathways, apoptosis, and necroptosis, and inflammatory responses were also observed in the infected HFH4-hACE2 transgenic mouse model and in the postmortem lung sections of fatal COVID-19 patients. Thus, the cell death and inflammatory responses are intimately linked during SARS-CoV-2 infection. The inflammatory responses from the infected epithelial cells may further induce infiltration of inflammatory cells inducing strong immune pathogenesis as revealed in the postmortem lung sections of fatal COVID-19 patients. These discoveries have shed light on the underlying mechanisms of COVID-19 pathogenesis and will offer valuable insights for the rational development of effective therapeutic strategies to treat COVID-19.

SARS-CoV-2 infection induces inflammatory responses in lung epithelial cells
To investigate the mechanisms of SARS-CoV-2-induced inflammatory responses, Calu-3 cells, a lung epithelial cell model, were infected and the cell lysates were collected at different time points post infection. Quantitative reverse-transcriptase PCR (qRT-PCR) analysis showed a time-dependent upregulation of inflammatory cytokines and chemokines including IL-7, IL-8, tumor necrosis factor-α (TNF-α), CXCL10, and CCL5 in the virus-infected cells (Fig. 1a), of which upregulation were also reported in the COVID-19 patients. 13 Furthermore, SARS-CoV-2 infection of Calu-3 cells induced IL-1β processing and secretion in an multiplicity of infection (MOI)-dependent manner (Fig. 1b). The expression level of pro-IL-1β was also upregulated in the infected cells (Fig. 1c). To analyze whether SARS-CoV-2-induced inflammatory responses depend on viral replication, SARS-CoV-2 was inactivated with ultraviolet (UV) treatment. UV inactivation abolished virus replication (Fig. 1d), virus-induced upregulation of TNF-α and IL-1β (Fig. 1e, f), and also prevented pro-IL-1β processing and secretion (Fig. 1g). On the other hand, treatment with MLN120B, an inhibitor of nuclear factor-κB (NFκB), inhibited virus-induced upregulation of pro-IL-1β and TNF-α without affecting viral replication ( Fig. 1h-j). Taken together, these results suggested that SARS-CoV-2-induced inflammatory responses depend on viral replication and activation of the NFκB pathway.
SARS-CoV-2-induced IL-1β maturation and secretion depends on caspase-8 activation We next analyzed the mechanism that mediate IL-1β processing and secretion. Activation of caspase-1 through the NLRP3 inflammasome is often involved in virus-induced IL-1β processing and secretion. 14 However, treatment with the caspase-1 inhibitor, VX-765, did not affect SARS-CoV-2-induced pro-IL-1β production or IL-1β processing and secretion (Fig. 2a, b). Similar results were observed with treatment of the NLRP3 inhibitor MCC950 (Fig. 2c,  d). Furthermore, both western blotting and quantitative PCR (qPCR) assays showed that NLRP3 is undetectable in the Calu-3 cells (Fig. 2e, f). Together, these results indicated that SARS-CoV-2induced IL-1β processing and secretion is independent of caspase-1 and NLRP3 inflammasome in the Calu-3 cells.
It was recently reported that caspase-8 can mediate IL-1β maturation through direct cleavage of the pro-IL-1β at the same site with caspase-1. 15 We found that SARS-CoV-2 infection of Calu-3 cells induced pronounced caspase-8 cleavage (Fig. 2g). The cleavage of BH3-interacting domain death agonist (BID), a substrate of caspase-8, into the truncated form further confirmed activation of caspase-8 in the SARS-CoV-2-infected cells (Fig. 2g). To analyze whether activated caspase-8 mediated SARS-CoV-2-induced inflammatory responses, Calu-3 cells were infected by SARS-CoV-2 in the presence of inhibitors against different caspases. The inhibition effects of these inhibitors were verified ( Supplementary Fig. S1). The treatment of caspase-8-specific inhibitor Z-IETD-FMK strongly reduced IL-1β processing and secretion in the SARS-CoV-2infected cells without affecting viral replication (Fig. 2h). Similar results were observed for Z-DEVD-FMK treatment, which has a broader inhibition effects against caspase-3, caspase-6, caspase-7, caspase-8, and caspase-10 ( Fig. 2h). Consistent with the caspase-8 inhibitor treatment, depletion of caspase-8 with small interfering RNA (siRNA) also reduced the processing and secretion of IL-1β without affecting virus replication (Fig. 2i).
We further analyzed processing of pro-IL-1β through cotransfection experiments. As shown in Fig. 2j, co-transfection of pro-IL-1β with caspase-1 but not caspase-3 resulted in IL-1β processing and secretion. Co-transfection of caspase-8 with pro-IL-1β resulted in IL-1β processing and secretion similarly with caspase-1, which was inhibited by treatment with caspase-8 inhibitor Z-IETD-FMK as previously reported 15 (Fig. 2j). These results suggested that SARS-CoV-2 infection induces caspase-8 activation to mediate pro-IL-1β processing.
SARS-CoV-2-induced IL-1β secretion depends on the necroptosis pathway We further analyzed the pathway that mediates IL-1β secretion during SARS-CoV-2 infection. The independence of caspase-1 activity for SARS-CoV-2-induced IL-1β processing and secretion suggests that pyroptosis might not play a major role during IL-1β secretion. Necroptosis is another form of cell death that can mediate secretion of inflammatory cytokines including IL-1β. 16 We therefore analyzed whether SARS-CoV-2 can trigger necroptosis in the Calu-3 cells. The mixed lineage kinase domain-like (MLKL) is the effector of necroptosis and is phosphorylated by the receptorinteracting protein kinase-3 (RIPK3) during necroptosis. SARS-CoV-2 infection of Calu-3 cells induced phosphorylation of MLKL (pMLKL) similar with the positive control (Fig. 3a). Staining with the antibody against pMLKL confirmed the upregulation of pMLKL in the infected cells and showed a staining pattern on the plasma membrane, consistent with the localization of activated pMLKL (Fig. 3b). UV inactivation of SARS-CoV-2 prevented virus replication and virus-induced MLKL phosphorylation, indicating of viral replication-dependent activation of the necroptosis pathway ( Fig.  3c). Inhibition with RIPK3 inhibitor GSK-872 inhibited pMLKL in the infected cells, suggesting that SARS-CoV-2 induces RIPK3 phosphorylation to trigger the necroptosis pathway (Fig. 3d). Moreover, treatment with RIPK3 inhibitor GSK-872 reduced IL-1β maturation and secretion without affecting viral replication ( Fig. 3d and Supplementary Fig. S2). Similarly, treatment with necrosulfonamide (NSA), a pMLKL membrane translocation inhibitor, reduced IL-1β maturation and secretion without affecting viral replication, confirming the role of necroptosis in SARS-CoV-2-induced inflammatory responses (Fig. 3d).

SARS-CoV-2 infection triggers apoptosis through caspase-8 activation
Caspase-8 is known to trigger apoptosis through both the extrinsic and intrinsic pathways. 17 To assess whether SARS-CoV-2 infection induced cell apoptosis in lung epithelial cells, the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay was performed on SARS-CoV-2-infected Calu-3 cells. TUNEL staining revealed an increased number of apoptotic cells in infected cells, compared with the mock control cells, indicating activation of cellular apoptosis (Fig. 4a). Western blotting analysis showed that caspase-8, caspase-9, and caspase-3 were cleaved into the activated forms in SARS-CoV-2-infected cells, similarly as in the control cells treated with staurosporine ( Fig. 4b). In addition, cleavage of the poly (ADP-ribose) polymerase 1 (PARP-1), a hallmark of apoptosis activation, was observed in the infected cells (Fig. 4b). In cells that were infected by the UV-inactivated SARS-CoV-2 virus, the cleavage of caspase-8, caspase-9, caspase-3, and PARP-1 was prevented, indicating that the apoptosis was dependent on viral replication (Fig. 4c). Treatment with caspase-8-specific inhibitor Z-IETD-FMK inhibited virus-induced BID cleavage, caspase-3 activation, and PARP-1 cleavage (Fig. 4d). In contrast, treatment with caspase-1 inhibitor, VX-765, did not show such an inhibition (Fig. 4e). These results suggested that SARS-CoV-2-triggered apoptosis depends on viral replication and caspase-8 activation.   18 The HFH4-hACE2 mice were intranasally infected with SARS-CoV-2 or mock-treated and the lung tissue was collected at 3 days post infection. TUNEL staining revealed a large number of apoptotic cells in the lung section of the infected mice but not in that of the mock-infected mice, indicating that SARS-CoV-2 infection induces pronounced apoptosis in the lung of infected mice (Fig. 5a). Immunohistochemical assays showed strong staining of virus NP protein in the lung section, confirming the productive viral infection of the lung tissues (Fig. 5b). IL-1β was also detected in the lung section of the infected mice but not in that of the mock control, confirming virus-induced inflammatory responses (Fig. 5b). Importantly, staining with the antibody against pMLKL revealed strong signals in the lung section of infected mouse but not in that of the mock control (Fig. 5b). These results suggested that SARS-CoV-2 infection of the HFH4-hACE2 mouse model leads to pronounced apoptosis and necroptosis concomitantly with inflammatory responses in the lung.
Finally, we analyzed the SARS-CoV-2 infection in the postmortem lung sections from fatal COVID-19 patients. Staining with the antibody against virus NP protein confirmed viral infection in the lung section (Fig. 5c). TUNEL assay performed on the lung section revealed distinct signals, indicating pronounced cell apoptosis similar with the situation in the infected HFH4-hACE2 mouse model (Fig. 5d). Staining with antibodies against IL-1β and pMLKL revealed that strong inflammation and necroptosis were induced in the lung tissue (Fig. 5e, f). Consistent with the virusinduced immunopathogenesis, severe interstitial inflammatory cell aggregation and infiltration (Fig. 6a, b), necrotic cell debris (Fig. 6c), early fibrosis formation (Fig. 6d), extravasated blood (Fig. 6e), and hemorrhage (Fig. 6f) were observed in the postmortem lung sections from fatal COVID-19 patients through hematoxylin-eosin (HE) staining assay. Taken together, these results suggested that SARS-CoV-2 infection induces cell death concomitantly with inflammatory responses and necroptosis activation in the lung tissue of the COVID-19 patients.

DISCUSSION
Here we report that SARS-CoV-2 infection of lung epithelial cells triggers cell death and inflammatory responses through the activation of caspase-8. Caspase-8 is a master regulator of several cell death pathways, including apoptosis, necroptosis, and pyroptosis. 19 Its role in regulation of inflammatory responses has been recently reported in the context of fungal infection. We found that SARS-CoV-2 induces caspase-8 activation to trigger cell apoptosis and to directly process inflammatory factors such as pro-IL-1β. The processed IL-1β is then secreted through the SARS-CoV-2-triggered necroptosis pathway (Fig. 6g). Thus, these discoveries have expanded the role of caspase-8 in the regulation of inflammatory responses to the context of virus infection.
The caspase-8-mediated apoptosis activation and inflammatory responses in infected lung epithelial cells may induce downstream immune pathogenesis in the lung tissue. In line with this notion, massive infiltration of inflammatory cells, necrotic cell debris, and pulmonary interstitial fibrosis were observed in the postmortem lung sections of fatal COVID-19 patients. TUNEL and immunohistochemistry (IHC) staining revealed pronounced apoptosis and necroptosis in the lung sections. Together, these mechanisms might lead to severe lung damage and immune pathogenesis during SARS-CoV-2 infection. This suggests that caspase-8 activation plays a central role in SARS-CoV-2-induced apoptosis and inflammatory responses. As the other two highly pathogenic coronaviruses, the SARS-CoV and MERS-CoV, also infect the lower respiratory track and cause severe pneumonia in patients; whether similar mechanisms were conserved among these viruses may be interesting questions for further investigation.
Necroptosis is a form of immunogenic cell death that can trigger inflammatory responses through releasing inflammatory cytokines and cellular damage associated molecular patterns (DAMPs). 20 Although SARS-CoV-2 activates caspase-8 to process pro-IL-1β, high level of caspase-8 activation is known to inhibit necroptosis pathway mediated by the RIPK3 and MLKL. 21 Thus, although caspase-8 activation is important for processing of pro-IL-1β in the context of SARS-CoV-2 infection, its activation did not reach a level that fully inhibited necroptosis and prevented IL-1β secretion. This suggests that SARS-CoV-2 induced a regulated level of caspase-8 activation to process pro-IL-1β and, meanwhile, to allow sufficient necroptosis activation for IL-1β release. Further efforts are needed to understand the mechanisms of how does SARS-CoV-2 trigger the necroptosis pathway and achieve the regulated caspase-8 activation to induce the inflammatory responses. The SARS-CoV-2-induced inflammatory responses were reported to be a major determinant for the disease progression in COVID-19 patients. 22 As blocking necroptosis strongly inhibited SARS-CoV-2-induced inflammatory responses, whether antinecroptosis treatment can be integrated into the therapeutic strategies to treat COVID-19 warrants further evaluation. It should also be noted that the necroptosis inhibitor did not fully block IL-1β secretion during SARS-CoV-2 infection, indicating that other pathways, such as pyroptosis, may also be involved in the inflammatory responses. Future efforts can be devoted to investigate whether pyroptosis and activation of gasdermins also play a role during SARS-CoV-2 inflammation.
Depending on the status of activation, the virus-triggered cell death pathways might also be beneficial to the host. It has been reported that influenza virus infection triggers both apoptosis and necroptosis in the lung as redundant cell death processes resulting in immune responses to restrict viral replication. 11 However, when these two pathways were triggered to a hyperactivation status, IAV infection leads to severe lung damage and fatal infection outcome. 11 Similar with IAV, the SARS-CoV-2 induced a dual mode of cell death pathways in the lung epithelial cells, which may pose anti-viral responses or immune pathogenesis depending on the activation status. Many factors may contribute to the level of cell death activation, including initial dose of viral infection, the underlying co-morbidity in the patient, age factor, etc. It would be important to study the regulated cell death activation and other concurrent factors in the clinical settings that determine the infection outcome of COVID-19 patients.

Specimen collection and ethics
Post mortem lung sections were prepared from a patient who died of COVID-19. The specimen collection and research was approved by the ethics committee of Wuhan Institute of Virology (WIV) and the designated hospital for COVID-19. The mice infection study was approved by WIV animal welfare committee.

Viruses
The SARS-CoV-2 (IVCAS 6.7512) was isolated from the bronchoalveolar lavage fluid sample of patient 23 and was propagated in Vero E6 cells. Viral titer (TCID 50 ) was determined on Vero E6 cells. Briefly, confluent monolayers were incubated with tenfold serial dilutions of virus for 3 days. Then the cytopathic effect was observed and the TCID 50 was calculated according to the Reed-Muench formula. Virus was inactivated by exposing to UV light for 1 h and the inactivation was confirmed by titer determination on Vero E6 cells.   RNA isolation and qRT-PCR analysis Total RNAs in collected cells were extracted with RNAiso Plus (TAKARA, Japan) following the manufacturer's instructions. qRT-PCR was performed with a two-step procedure using the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, China) and ChamQ Universal SYBR qPCR Master Mix (Vazyme, China). Primer sequences are available upon request.
Western blotting analysis Cells treated and collected as indicated were lysed with cell lysis buffer (Beyotime, China). Cell lysates were subjected to 10-15% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and further transferred to polyvinylidene difluoride membranes (Millipore). After blocking, proteins were blotted with the indicated primary antibodies and then corresponding horseradish peroxidase-conjugated secondary antibodies. Protein bands were detected by an enhanced chemiluminescence kit (Millipore) using a Chemiluminescence Analyzer (Chemiscope600pro).
Mature IL-1β detection Proteins in 500 μl of the collected supernatant were precipitate with equal volume of methanol and a quarter volume of chloroform. The precipitant was dissolved in 2 × SDS loading buffer for western blotting analyses as described previously using antibodies against IL-1β (ABclonal, China).
TUNEL assay Apoptosis in SARS-CoV-2-infected Calu-3 cells was detected by a TUNEL Assay kit purchased from Thermo Fisher (C10617) according to the manufacturer's protocols. Briefly, fixed cells were permeabilized with 0.25% Triton™X-100 for 20 min at room temperature. Then the TdT reaction mixture containing TdT and EdUTP was added to label the fragmented DNA in cells. Cells were washed and incubated with Click-iT™ Plus TUNEL reaction cocktail, followed by incubation with DAPI (Beyotime, China) and analyzed using confocal microscope (Andor Dragonfly 202).
Immunofluorescence assay Calu-3 cells were mock-treated or infected with SARS-CoV-2 (MOI = 0.1) for 36 h. Cells were fixed and permeabilized, and then blocked with phosphate-buffered saline containing 3% bovine serum albumin. The cells were incubated with rabbit anti-pMLKL (S358, Abcam) and mouse-anti-NP (SR24) for 1 h at room temperature, followed by incubation with Fluor 488 goat antimouse IgG and Fluor 561 goat anti-rabbit IgG for 1 h. Cells were washed and incubated with DAPI (Beyotime, China), and then analyzed using confocal microscope (Andor Dragonfly 202).
Histology and immunohistochemistry Lung samples obtained from infected and mock-infected mice were fixed with 4% paraformaldehyde, embedded in paraffin and cut into sections of 3.5 μm, and further used for TUNEL assay or processed to IHC. TUNEL assay was performed using an In Situ Cell Death Detection Kit (Roche) according to the manufacturer's protocols. Briefly, the sections were deparaffinized, rehydrated, and permeabilized with proteinase K. The sections were then incubated with a mixture of TdT and dUTP (2 : 29), followed by incubation with DAPI (Beyotime, China) solution for 10 min. For IHC analysis, the indicated antibodies were used as primary antibodies, then sections were incubated with secondary antibody (Rabbit/Mouse Envision, Dako, Denmark), followed by visualizing with a detection kit (DAB, Dako, Denmark).
The biopsy lung tissue was fixed with 4% paraformaldehyde, paraffin-embedded and cut into sections of 5 μm. For routine histology, sections were stained with HE. Immunostaining was performed using PANO 7-plex IHC kit (Panovue, China) according to the manufacturer's instructions. TUNEL assay was conducted as described above.
Statistical analyses All statistical analyses were performed in GraphPad Prism version 6, as defined in the text and figure legends.

DATA AVAILABILITY
The data sets used for the current study are available from the corresponding author upon reasonable request.