The purinergic receptor P2X7 and the NLRP3 inflammasome are druggable host factors required for SARS-CoV-2 infection

Purinergic receptors and NOD-like receptor protein 3 (NLRP3) inflammasome regulate inflammation and viral infection, but their effects on severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection remain poorly understood. Here, we report that the purinergic receptor P2X7 and NLRP3 inflammasome are cellular host factors required for SARS-CoV-2 infection. Lung autopsies from patients with severe coronavirus disease 2019 (COVID-19) reveal that NLRP3 expression is increased in host cellular targets of SARS-CoV-2 including alveolar macrophages, type II pneumocytes and syncytia arising from the fusion of infected macrophages, thus suggesting a potential role of NLRP3 and associated signaling pathways to both inflammation and viral replication. In vitro studies demonstrate that NLRP3-dependent inflammasome activation is detected upon macrophage abortive infection. More importantly, a weak activation of NLRP3 inflammasome is also detected during the early steps of SARS-CoV-2 infection of epithelial cells and promotes the viral replication in these cells. Interestingly, the purinergic receptor P2X7, which is known to control NLRP3 inflammasome activation, also favors the replication of D614G and alpha SARS-CoV-2 variants. Altogether, our results reveal an unexpected relationship between the purinergic receptor P2X7, the NLRP3 inflammasome and the permissiveness to SARS-CoV-2 infection that offers novel opportunities for COVID-19 treatment.


Introduction
The rapid worldwide spread of SARS-CoV-2 poses a global health emergency that has not been resolved.Despite the use of preventive COVID-19 vaccination, neutralizing monoclonal antibodies, steroid or IL-6-targeted treatment that drastically reduced the risk of severe symptoms and death due to SARS-CoV-2 infection (1-6), people who are vaccinated or treated, may still get infected with SARS-CoV-2.Thus, there is an urgent need to combine anti-inflammatory therapies with highly effective antivirals for treating COVID-19.Three antiviral compounds, remdesivir (7), molnupiravir (8,9), and nirmatrelvir (10) were authorized by the US Food and Drug administration for an emergency use for the treatment of COVID-19 (https://www.fda.gov).These drugs are proposed as therapeutic options, alone or in combination with modifiers of inflammatory responses (such as tocilizumab (11) or the Janus kinase inhibitor baricitinib (12)) for patients suffering from severe COVID-19 or for those with high risk of severe disease progression.Nevertheless, selection pressures exerted by these antivirals in patients with prolonged infection have been shown to increase viral diversity, to lead to drug resistance and to favor the emergence of mutated variants (13).To prevent further massive outbreaks with emerging variants, a better understanding of mechanisms whereby SARS-CoV-2 hijacks cellular host factors for viral replication and dysregulates anti-viral immune responses is still required and should pave the way for the identification of novel cellular or viral targets, thus offering alternative therapeutic opportunities for SARS-CoV-2-infected patients.
The SARS-CoV-2 entry into host cells begins with the binding of viral spike (S) glycoprotein to angiotensin converting enzyme 2 (ACE-2) and triggers through two distinct pathways, the proteolytic cleavage of S glycoprotein by the serine protease TMPRSS2 at cell surface or by cathepsin B/L in the host cell endosomes (14-16).This process ultimately allows the fusion of viral membranes with host cellular membranes (15,17) and leads to viral RNA release into host cytosol, to the viral replication using specialized enzymes (such as RNA-dependent RNA polymerase (RdRp) (18)), to viral structural protein expression (such as E and S proteins), and finally, to the assembly and release of the viral progeny (19).Host factors (such as RAB7A, p38MAPK, CK2, AXL and PIFFYVE kinases) are involved in the regulation of early and late steps of SARS-CoV-2 infection (20, 21), but the host cellular pathways used by SARS-CoV-2 to establish a viral infection are still poorly understood.Even though SARS-CoV-2-infected people are mainly asymptomatic or exhibit mild to moderate symptoms, approximately 15% of patients experience severe disease with atypical pneumonia and 5% develop an acute respiratory distress syndrome (ARDS) and/or multiple organ failure that is associated with a high mortality rate (around 20-30%) (22).Studies of COVID-19 patients with severe disease revealed a high level of plasma pro-inflammatory cytokines (including IL-1b, IL-6, IL-10, IL-18 and TNF) (23) and lactate dehydrogenase (LDH) (24), indicating overt hyper-inflammation during COVID-19, sometimes improperly termed "cytokine storm" or "cytokine release syndrome".Until now, few molecular mechanisms driving COVID-19 associated hyper-inflammation have been identified and proposed to explain COVID-19 pathogenesis.
SARS-CoV-2 infection activates several microbe-sensing receptors called pattern recognition receptors (PRRs) such as Toll-like receptors (TLRs), RIG-I-like receptors (RLRs) and C-Type lectin receptors (CLRs).Their activations lead to antiviral immune responses driven by the nuclear factor NF-kB, type I interferon secretion and interferon stimulated genes (ISGs) expression (25).Among those PRRs, nucleotide-binding domain, leucine-rich repeat-containing receptor (NLRs) proteins were widely found to be involved in SARS-CoV-2-mediated hyperinflammation.NLR proteins and purinergic (P2) receptors are the main germline-encoded pattern recognition receptors regulating the secretion of IL-1 family members in response to microbial infection, inflammation, and inflammatory diseases.Upon activation, NLR protein 3 (NLRP3), which is the most studied NLR protein (26), forms large complexes with the adaptor protein ASC (apoptosis-associated speck-like protein containing a CARD domain), called inflammasomes, which activate caspase-1, induce the release of mature cytokines IL-1b and 27) and can lead to the inflammatory cell death of stimulated, stressed or infected host cells, which is also known as pyroptosis (28).High levels of IL-1b, IL-18 and LDH positively correlate with disease severity in COVID-19 patients (29), suggesting that inflammasomes participate in SARS-CoV-2 pathogenesis.More recently, NLRP3 inflammasome was found activated in response to SARS-CoV-2 infection and identified as critical driver of COVID-19 (30-33).In addition, SARS-CoV-2 viral proteins such as viral spike (S) glycoprotein (34), SARS-CoV open reading frame-8b (ORF8b) (35), the transmembrane pore-forming viral Viroporin 3a (also known as SARS-COV 3a) (36) and the viral N protein (37) were shown to activate the NLRP3 inflammasome, thus revealing that NLRP3 inflammasome is potentially an interesting molecular target for the treatment of COVID-19.Purinergic receptors are membrane-bound innate receptors that bind extracellular nucleotides (such as adenosine triphosphate (ATP)), and control numerous cellular functions (such as cytokine secretion and migration) mainly in immune cells, but also on other cell types that are involved in SARS-CoV-2 pathogenesis such as type 1 and 2 pneumocytes, endothelial cells, platelets, cardiomyocytes and kidney cells (38,39).Purinergic P2 receptors are divided into two families, the ionotropic P2X receptors and the metabotropic P2Y receptors, which can regulate the NLRP3 inflammasome (39-41).P2X7 activation was extensively shown to control NLRP3 inflammasome activation and cytokine release in response to danger signals (42).We previously revealed that purinergic receptors (such as P2Y2 and P2X7), NLRP3 inflammasome and associated signaling pathways also control viral replication through the modulation of the fusogenic activity of HIV-1 envelope (41,43,44) or the pre-integrative steps of HIV-1 life cycle (45).The contribution of purinergic receptors to viral infection has been confirmed with other purinergic receptors and with several viruses (such as human cytomegalovirus and hepatitis B, C and D viruses) (46).In this context, we hypothesized that the purinergic receptor P2X7 and NLRP3 inflammasome activation could contribute to SARS-CoV-2 pathogenesis and thus, the therapeutic modulation of P2X7 and NLRP3 could regulate the permissiveness of host cells to SARS-CoV-2 infection.

Primary cells
Minor salivary gland biopsy specimens were obtained from patients referred for suspected primary Sjögren's syndrome diagnosis to the rheumatology department of Bicetre Hospital (AP-HP, UniversitéParis-Saclay).The subjects selected for this study were patients who presented with sicca symptoms but without any autoantibodies and with normal or subnormal minor salivary gland findings (i.e., focus score <1), and thus were considered as controls.Primary cultures of salivary gland epithelial cells were established from minor salivary glands based on previously described protocol (48).Briefly, the minor salivary gland biopsy was cut into small fragments, which were placed in 75-cm² flasks (BD Falcon) (prehydrated with 3 ml of s-BEM medium during 30 minutes at 37°C) with 1 ml of s-BEM medium.The supplemented-Basal Epithelial Medium (s-BEM) consisted of 3:1 mix of Ham's F12 (Gibco, Life Technologies, UK) and DMEM (Gibco, Life Technologies, UK), supplemented with 2.5% FBS (PAN BIOTECH, Aidenbach, Germany), 10 ng/ml epidermal growth factor (EGF) (Gibco), 0.4 μg/ml hydrocortisone (Upjohn 100mg SERB), 0.5 μg/ml insulin (Novo Nordisk, Novorapid Flexpen) and 1X penicillin/streptomycin (LONZA, Verviers, Belgium).The flasks were incubated at 37°C with 5% CO 2 in a humidified incubator.The day after, 3ml of s-BEM were added slowly in the flasks.At day 6, 5 ml of s-BEM were added.Then, cultures were fed with fresh medium twice a week.After biopsy, 2-3 small pieces were seeded into T75 flask and cultured in a DMEM-F12 medium supplemented with 2.5% FBS, 100 UI/ml penicillin/streptomycin, 0.4 μg/ml hydrocortisone, 0.5 μg/ml insulin and 10 ng/ml EGF.Once tissues had expended, cells were harvested and 1x10 5 cells per well were plated in 48-well plates that were pretreated with collagen for 1 hour at 37°C.Twenty-four hours later, cells were treated with OxATP and infected with SARS-CoV-2 at a multiplicity of infection (MOI) of 0.5 for 48 hours.Cellular RNA was collected and extracted with NucleoSpin RNA Plus XS, Micro kit for RNA purification with DNA removal column (#740990.250,Macherey-Nagel), according to manufacturer instructions.Samples were then subjected to reverse transcription and PCR using probes against RdRp and b-actin.Data were normalized to b-actin and to control condition following the 2 -DDCt method.

Human autopsies and patients' serum
Human autopsies were performed after ethics committee approval (n°9/2020) at the National Institute for Infectious Diseases Lazzaro Spallanzani-IRCCS Hospital (Rome, Italy), according to guidance for post-mortem collection and submission of specimens and biosafety practices (CDC March 2020, Interim Guidance and (49)) to reduce the risk of transmission of infectious pathogens during and after the post-mortem examination.Autopsies were performed in accordance with the law owing to the unknown cause of death, and to both scientific and public interest in a pandemic novel disease.Post-mortem lung sections were obtained from 3 non-COVID-19 patients and 7 COVID-19 patients with severe disease.Controls (n=3) were patients deceased after hemorragia, cardiorespiratory failure or interstitial pneumonia associated with pulmonary capillaritis.All COVID-19 patients (n=7) deceased after cardiorespiratory failure and exhibited in their vast majority diffuse alveolar damage (n=6) or fibrosis (n=1).Detection of SARS-CoV-2 was performed by RT-PCR on all patients using nasopharyngeal swabs.Convalescent COVID-19 serum was obtained from patients that received three doses of COVID-19 mRNA vaccine (Pfizer).Serum was collected one month after the last injection.All patients had a titer > 40 000 units/ml.The serum was obtained from Dr. Samuel Lebourgeois and Dr. Nadhira Houhou-Fidouh (Hopital Bichat Claude Bernard, France).

Virus, pseudoviral constructs and viral production
The SARS-CoV-2 BetaCoV/France/IDF0372/2020 strain (D614G) was provided by Dr. Benoit Visseaux from the group of Prof. Diane Descamps (UMR S 1135, Hopital Bichat Claude Bernard, Paris) and by the National Reference Center For Respiratory Viruses (Institut Pasteur, Paris, France).The SARS-CoV-2 Alpha variant (lineage B1.1.7)was obtained from Pr. Mauro Pistello.Viral stocks were prepared by propagation in African green monkey kidney epithelial (Vero E6) cells in a biosafety level-3 (BLS-3) laboratory and titrated using lysis plaque assay as previously described (50).SARS-CoV-2 stock titer was 2x10 6 PFU/ml.After 72 hours of infection, the supernatant was centrifuged at 1500 rpm, 12°C , for 5 minutes to remove dead cells and then centrifuged at 3000 rpm for 20 minutes at 12°C.Supernatant was then aliquoted and stored at -80°C.For the production of green fluorescent protein (GFP)-tagged HIV-1NL4-3 Denv variant (defective in viral envelope) pseudotyped with the SARS-CoV-2 spike (S) envelope (S-GFP-LV), the plasmid NLENG1-ES-IRES coding for defective virus carrying two stop codons in the reading frame of the envelope and expressing the fluorescent reporter gene EGFP (GFP) was derived from the HIV-1 virus NL4-3 as previously published (51).Envelope defective virus was pseudotyped with HU-1 SARS-CoV-2 Spike protein (pLV-Spike, Invivogen) and used for single round infection assays.To produce viral stocks, 293T cells were transfected with virus-encoding plasmids by the calcium phosphate method.Briefly, 293T cells were transfected using 20 μg of NLENG1-ES-IRES and 4 μg of the corresponding plasmid encoding spike (S) envelop.Two days post-transfection, supernatants containing the viruses were submitted to a low-speed centrifugation step, filtered through a 0.45-μm low-protein-binding Durapore filter (Millipore), and stored at -80°C until use.Viral stocks were titrated using HIV-1 P24 ELISA assay (#NEK05001KT, Perkin Elmer) and the concentration was 200 ng/ml.

Flow cytometry
Control, treated-ACE2-A549 and treated-Vero E6 cells were harvested in DMEM medium, washed three times in PBS and saturated with 2% PBS-BSA at 4°C for 20 minutes.Primary antihuman ACE-2 antibody (#AF933, R&D systems) or isotypematched antibody (#AB-108-C, R&D systems) was incubated at 0.2 μg/10 6 cells for ACE2-A549 and at 0.4 μg/10 6 cells for Vero E6 cells during 30 minutes at room temperature in 2% PBS-BSA.Cells were washed three times with 2% PBS-BSA and incubated with secondary anti-goat IgG (H+L) APC-conjugated antibody (F0108, R&D systems) in 2% PBS-BSA for 20 minutes at room temperature at 0.05 μg/10 6 cells for ACE2-A549 cells or at 0.1 μg/10 6 cells for Vero E6 cells.Cells were then washed three times with 2% PBS-BSA, resuspended in 200 μl 2% PBS-BSA and analyzed for mean fluorescence intensity (MFI) of ACE2 expression with CytoFLEX flow cytometer (Beckman).To determine the biological effects of the purinergic receptor P2X7 and NLRP3 inflammasome on viral entry, Vero E6 cells, ACE2-A549 cells and ACE2-TMPRSS2-A549 cells were pretreated for 4 hours with indicated drugs and infected for 24 hours with Spike-GFP-LVs.As negative control, Spike-GFP-LVs were preincubated with convalescent COVID19 patient's serum during 40 minutes at room temperature.After extensive washes, cells were incubated in complete medium for 24 hours, fixed with 4% PFA and GFP-positive cells were analyzed using BD FACSCelesta flow cytometer (BD Biosciences).

Detection of IL-1b cytokine release
Supernatants harvested from PMA-stimulated THP1 macrophages, Caco-2 cells or ACE2-A549 cells that were treated with MCC950, depleted for NLRP3 or incubated with COVID-19 patient's serum, and infected as indicated with SARS-CoV-2, or from control uninfected cells were analyzed using ELISA assay for IL-1b, according to the manufacturers' instructions.IL-1b secretion from PMA-THP1 macrophages was detected using IL-1 beta Human ELISA kit (#KHC0011, ThermoFisher Scientific) and IL-1 beta Human ELISA kit, High Sensitivity (#BMS224HS, ThermoFisher Scientific) was used for the detection of IL-1b secretion from epithelial cells.

Increased NLRP3 expression level in alveolar macrophages, pneumocytes and syncytia detected in lung autopsies of patients with SARS-CoV-2 pneumonia
To study the potential contribution of NLRP3 to SARS-CoV-2 pathogenesis, we first analyzed the expression of NLRP3 in lung tissue samples obtained from three uninfected individuals and seven SARS-CoV-2-infected patients who died from COVID-19 (Tables 1, 2).Autopsies were stained with hematoxylin and eosin, or incubated with antibody against NLRP3 and analyzed.As previously described (55), parenchymal multifocal damages with intra-alveolar inflammation, fibrin and hyaline membrane formation that are consistent with a diagnosis of diffuse alveolar damage were observed in all SARS-CoV-2-infected patients.The pattern of developed pneumonia (with fibrotic organization and type 2 pneumocyte hyperplasia) and fibroblastic foci formed by the loss of organized connective tissue consistent with alveolar duct fibrosis were also detected (Figure 1A).Besides, the presence of inflammatory cells (composed mainly of macrophages and lymphocytes) was the main characteristic of COVID-19 patient autopsies (Figure 1A).NLRP3 expression was then examined on these samples, and immuno-reactive NLRP3 was mainly detected in Type II pneumocytes (Figures 1B, C) and alveolar macrophages (Figures 1B, D) from both SARS-CoV-2 patients and uninfected specimens, which have been previously shown to be cellular targets of SARS-CoV-2 (56,57).Interestingly, we also detected in SARS-CoV-2-infected lung tissues, some syncytia that expressed macrophage marker, CD68 (Figure 1E), NLRP3 (Figure 1F) and viral RNA (Figure 1G), revealing a potential link between NLRP3 expression and viral infection.Given that macrophages are key cellular players for COVID-19 pathogenesis (58) and inflammasome activation was detected in SARS-CoV-2-infected macrophages (30), we first analyzed the expression of NLRP3 during the infection of macrophages with SARS-CoV-2.Interestingly, we found that NLRP3 expression level increased 24 hours after SARS-CoV-2 infection in phorbol-12myristate-13-acetate (PMA) -differentiated human THP1 macrophages with different multiplicities of infection (MOI) (Figure 2A, Supplementary Figure 1A) or at different time points after infection with a MOI of 0.2 (Figure 2B, Supplementary Figure 1B).Increased NLRP3 expression level was observed in the absence of detectable intracellular SARS-CoV-2 spike (S) protein expression, indicating a lack of viral replication (Figures 2A, B).These results thus confirm, as previously published (30), that NLRP3 expression level is increased in absence of productive infection or during abortive SARS-CoV-2 infection in macrophages.Then, we evaluated the effect of SARS-CoV-2 infection on two hallmarks of inflammasome activation, the oligomerization of the adaptor protein ASC and the production of mature IL-1b (59).The oligomerization of ASC, revealed by the quantification of ASC speck formation (Figures 2C, D), and the release of mature IL-1b (Figures 2E, F) were increased in PMA-THP1 macrophages that were infected with SARS-CoV-2, as compared to uninfected cells (Figures 2C-F).Inactivation of NLRP3 inflammasome with a pharmacological inhibitor MCC950 (Figure 2G, Supplementary Figure 1C) or using specific NLRP3 CRISPR guide RNAs (gRNAs) and the CAS9 gene (CrNLRP3)  (Figure 2H, Supplementary Figure 1D) significantly reduced the release of mature IL-1b by PMA-THP1 macrophages that were infected with SARS-CoV-2 during 24 hours (Figures 2I, J).These results reveal that NLRP3 inflammasome is activated during abortive SARS-CoV-2 infection in macrophages.Given that the purinergic receptor P2X7 has been extensively involved in the activation of NLRP3 inflammasome (59), we also investigated the impact of P2X7 antagonist oxidized ATP (OxATP) or agonist 2'(3')-O-(4-Benzoylbenzoyl) adenosine 5'-triphosphate (BzATP) on NLRP3 inflammasome activation elicited by SARS-CoV-2.Pharmacological inhibition or activation of P2X7 with respectively, OxATP or BzATP (Figures 2K, L), did not modulate increased expression level of NLRP3 (Figure 2K, Supplementary Figure 1E) nor IL-1b release (Figure 2L) from PMA-treated THP1 macrophages that were infected with SARS-CoV-2.Results reveal that the activation of NLRP3 inflammasome detected during the infection of PMA-treated THP1 macrophages with SARS-CoV-2 does not require the purinergic receptor P2X7.Taken together, these results indicate that NLRP3 inflammasome activation is detected during the early steps of human macrophage infection with SARS-CoV-2.

NLRP3 acts as a proviral host factor for SARS-CoV-2 infection
Considering that epithelial cells, which have been identified as primary targets of viral replication in the lung and the nasal epithelium of SARS-CoV-2-infected patients (56,60,61), also express NLRP3 inflammasome (62), we then analyzed the activation of NLRP3 inflammasome during the infection of permissive host Caco-2 cells and ACE2-overexpressing A549 (ACE2-A549) cells.Surprisingly, a weak, but significant SARS-CoV-2-MOI-dependent release of IL-1b was detected in response

Purinergic receptor P2X7 dictates NLRP3 inflammasome activation for SARS-CoV-2 replication
As previously published (41,43,70), we initially revealed the contribution of purinergic receptors to the early steps of human immunodeficiency virus-1 (HIV-1) infection.To evaluate the possibility that P2X7, an important partner of NLRP3 inflammasome activation, might also affect SARS-CoV-2 viral replication, we next examined the impact of a non-selective antagonist of purinergic receptors P2X, the pyridoxal-phosphate-6-azophenyl-2',4'-disulfonate (PPADS) and the P2X7-specific antagonist OxATP, on SARS-CoV-2 viral replication and related cellular damage.The inhibition of P2X and P2X7 activities by PPADS and/or OxATP reduced the frequency of spike (S)-positive cells (Figures 5A-D) and the expression level of viral E RNA (Figure 5E) elicited by SARS-CoV-2 in Vero E6 cells (Figures 5A,  C) and ACE2-A549 (Figures 5B, D, E).PPADS and OxATP partially affected viability of uninfected Vero E6 cells (Supplementary Figures 4A, B) at the concentration of 100 μM, but not at working concentration (10 μM).Moreover, OxATP did not show any effect on the viability of ACE2-A549 cells (Supplementary Figure 4C).Accordingly, P2X7 depletion using small interfering RNA significantly reduced intracellular spike (S) expression level detected after 48-hour infection of ACE2-A549 cells with SARS-CoV-2 (Figure 5F, Supplementary Figures 4D, E).Conversely, the activation of P2X7 with BzATP increased replication of SARS-CoV-2, as revealed by the increase of intracellular spike (S) expression levels (Figures 5B, D) and viral E expression level (Figure 5G), but did not show a significant impact on cellular viability (Supplementary Figure 4F).We then analyzed impact of OxATP on the replication of SARS-CoV-2 into primary salivary gland epithelial cells, which have been identified as primary targets for SARS-CoV-2 in SARS-CoV-2-infected patients (71), and confirmed that OxATP significantly reduced SARS-CoV-2 replication, as revealed by the reduction of RdRp mRNA expression level in primary salivary gland epithelial cells that were treated with OxATP and infected during 48 hours with SARS-CoV-2 (Figure 5H).This process occurs without affecting the viability of primary salivary gland epithelial cells (Supplementary Figure 4G).Finally, we analyzed the effect of P2X7 inhibition on the replication of SARS-CoV-2 Alpha variant and determined the amount of viral RdRp RNA released in the supernatant of liver Huh7 cells that were infected with SARS-CoV-2 Alpha variant in presence or in absence of OxATP.We observed that OxATP significantly reduced the release of viral particles in the supernatant of SARS-CoV-2-infected Huh7 cells (Figure 5I), thus revealing that purinergic receptor P2X7 also regulates the replication of SARS-CoV-2 Alpha variant.To determine whether the purinergic receptor P2X7 may act upstream of NLRP3 inflammation activation, ACE2-A549 cells were infected during 48 hours with SARS-CoV-2 in presence of indicated concentrations of BzATP and/or Tranilast.As previously shown (Figures 3F-K, 5B, D, G), the intracellular expression of spike (S) protein elicited by the infection of ACE2-A549 cells with SARS-CoV-2 was increased or decreased in presence of BzATP or Tranilast, respectively (Figure 5J).Interestingly, we observed that the treatment of ACE2-A549 cells with both BzATP and Tranilast significantly reduced the intracellular spike (S) expression level detected after 48 hours of infection, as compared to ACE2-A549 cells that were infected in presence of BzATP (Figure 5J), thus demonstrating that the purinergic receptor P2X7 acts upstream to NLRP3 inflammasome activation and dictates viral replication.

Purinergic receptor P2X7 and NLRP3 inflammasome regulate SARS-CoV-2 replication at a post-entry level
Then, we determined whether purinergic receptor P2X7 and NLRP3 might control SARS-CoV-2 replication through the modulation of viral entry.We analyzed viral entry by using SARS-CoV-2 spike (S) pseudotyped HIV-1.Thus, we incubated ACE2 and transmembrane serine protease 2 (TMPRSS2)expressing A549 (ACE2-TMPRSS2-A549) highly permissive cells with OxATP, BzATP or serum of convalescent COVID-19 patients during 4 hours before their infection with green fluorescent protein (GFP)-tagged HIV-1NL4-3 DEnv variant (defective in viral envelope) pseudotyped with the SARS-CoV-2 spike (S) envelope (S-GFP-LV) (Figures 6A-E).Then, we analyzed the viral entry by detecting after 48-hour infection, the frequency of GFP + cells and observed that OxATP and BzATP did not reduce the frequency of GFP + cells detected in S-GFP-LV-infected ACE2-TMPRSS2-A549 cells (Figures 6A-D, F).Of note, serum of convalescent COVID-19 patients efficiently reduced the frequency of GFP + cells detected after 48 hours of infection (Figures 6E, F).These results were confirmed by detecting the intracellular HIV-1 Cap24 capsid in ACE2-TMPRSS2-A549 cells (Figure 6G, Supplementary Figure 5A), ACE2-A549 cells (Supplementary Figure 5B) and Vero E6 cells (Figure 6H, Supplementary Figure 5C) that were infected by S-GFP-LV during 4 hours in presence (or not) of OxATP, BzATP or serum of convalescent COVID-19 patients (Figures 6G, H, Supplementary Figures 5A-C) and revealed that purinergic receptor P2X7 does not regulate SARS-CoV-2 entry into host cells.Similarly, we analyzed the frequency of GFP + cells detected after 48-hour infection of ACE2-TMPRSS2-A549 cells with S-GFP-LV in presence (or not) of Tranilast, YVAD or serum of convalescent patients.We observed that Tranilast and YVAD did not reduce the frequency of GFP + cells, as compared to cells treated with serum of convalescent patients (Figures 6I-N).These results were confirmed by detecting the intracellular HIV-1 Cap24 capsid in ACE2-TMPRSS2-A549 cells (Figure 6O, Supplementary Figure 5D) and ACE2-A549 cells (Figure 6P, Supplementary Figure 5E) that were infected with S-GFP-LV during 4 hours in presence (or not) of Tranilast.Since ACE2 expression was shown to control viral entry (72), Vero E6 cells (Supplementary Figures 6A, B) and ACE2-A549 cells (Supplementary Figures 6C-F) were treated with OxATP, BzATP, Tranilast and/or YVAD, for 6 hours (Supplementary Figures 6A, B), 4 hours (Supplementary Figures 6C, D), or 24 hours (Supplementary Figures 6E, F), and ACE2 membrane expression was analyzed by flow cytometry.The expression of ACE2 did not change in the presence of the agonist of P2X7 (BzATP), the selective antagonist of P2X7 (OxATP), the selective inhibitor of NLRP3 (Tranilast) and the selective inhibitor of caspase-1 (YVAD), implying that the purinergic receptor P2X7, the protein NLRP3 and the caspase-1 do not affect the membrane expression of ACE2.Altogether, these results demonstrate that the activation of the purinergic receptor P2X7 and the NLRP3 inflammasome supports SARS-CoV-2 replication into target cells without modulating viral entry.

Discussion
With the goal of studying the potential contribution of the purinergic receptor P2X7 and the NLRP3 inflammasome to the pathogenesis of SARS-CoV-2 infection, our study revealed that these two major sensors of danger signals participate to the early steps of SARS-CoV-2 infection.The activation of inflammasomes was detected during COVID-19 and positively correlates with disease severity in humans (29).Despite the fact that several inflammasomes were proposed to be activated and to contribute to organ dysfunction and disease severity during COVID-19 (29), the activation of NLRP3 inflammasome was recently associated with chronic stages of COVID-19 in humans (29)  remain to be defined.In addition to its well-established role in triggering inflammation during microbial infection (68), our results highlighted an unexpected contribution of NLRP3 inflammasome activation to SARS-CoV-2 pathogenesis through the promotion of viral replication.Our results are in agreement with recent report revealing that pharmacological inhibition of NLRP3 inflammasome with MCC950 or knockout of NLRP3 abrogate both SARS-CoV-2 replication and associated inflammation during infection of human ACE2 mice (84).Considering the ability of caspase-1 to regulate the activation of sterol regulatory element binding proteins (SREBPs), whose expressions have been detected increased in patients with severe COVID-19 (85) and genetic manipulations revealed their contributions to SARS-CoV-2 life cycle (83,86,87), further investigations to define the potential interplay between NLRP3 inflammasome activation, fatty acid and cholesterol metabolism and SARS-CoV-2 replication would be of interest and should help to identify key mechanisms regulating the replication of SARS-CoV-2 that could be druggable targeted for COVID-19 treatment.
Our study demonstrated that the purinergic receptor P2X7 is also a key host factor in the SARS-CoV-2 replication cycle.The activation of purinergic receptor P2X7 supports SARS-CoV-2 replication without interfering with ACE2 membrane expression and viral entry.By deciphering the relationship between purinergic receptor P2X7 and NLRP3 inflammasome activation during the early steps of SARS-CoV-2 infection, our study demonstrated that purinergic receptor P2X7 acts upstream the activation of NLRP3 inflammasome.Since we previously observed that the purinergic receptors control HIV-1 envelope-elicited fusion (41,43), it is conceivable that upon the interaction of spike (S) protein with host cell receptors and/or in response to the fusogenic activity of spike (S) protein, ATP is released from SARS-CoV-2 target cells and acts in an autocrine fashion on purinergic receptor P2X7 present at the membrane of SARS-CoV-2 target cells to stimulate NLRP3 inflammasome activation and to regulate viral replication.
The therapeutic inhibition of the key host factors of the s ig n a li n g p a t hw a y t h at w e i d e n t i fi e d i n t h i s s t u d y (P2X7➔NLRP3➔Caspase-1) could block the SARS-CoV-2 life cycle at the level of viral replication.Ivermectin, which is known to potentiate the activity of the purinergic receptor P2X4 (88), which can form heterotrimeric receptors with P2X7 (89), was recently found to inhibit the replication of SARS-CoV-2 in vitro (90), but did not show benefits for COVID-19 patients (91).Nevertheless, further molecular investigations are required to fully understand the role of purinergic receptor family members and associated signaling pathways during the early steps of SARS-CoV-2 infection and COVID-19-associated hyper-inflammation.
The clinical randomized clinical trials (RCT) having evaluating IL-1 inhibitors in severe COVID-19 yielded to controversial results.One RCT (NCT04362813) with canakinumab (a monoclonal antibody against anti-IL-1b) was negative (92).With anakinra, an IL-1 receptor antagonist, one study (NCT04680949) in patients with increased soluble urokinase plasminogen receptor (suPAR) plasma level was positive (93), whereas 2 other trials (NCT04341584, NCT04364009) were negative (94).Our work suggests that targeting inflammasome-dependent IL-1b/IL-18 and/or purinergic signaling pathways may offer a novel opportunity for the treatment of viral infection and hyperinflammation associated with COVID-19.Our study provides the first rationale for testing specific P2X7 antagonists such as CE-224,535 or JNJ-54175446, which were assessed without success for the treatment of rheumatoid arthritis (NCT00628095) and are under evaluation for the treatment of major depressive disorders (NCT04116606), respectively, alone or in combination with specific antagonists of NLRP3 such as Tranilast or Dapansutrile, which were assessed for the treatment of Cryopyrin-Associated Periodic Syndrome (CAPS) (NCT03923140) or systolic heart failures (NCT03534297), respectively.Repurposing of these existing drugs for blocking both viral replication and COVID-19-associated hyper-inflammation should rapidly improve the health of COVID-19 patients.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author(s) declared that they were an editorial board member of Frontiers, at the time of submission.This had no impact on the peer review process and the final decision.

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers.Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

TABLE 1
Demographics, autopsy cause of death and lung histopathology findings of COVID-19 patients.

TABLE 2
Demographics, autopsy cause of death and lung histopathology findings of non-COVID-19 patients.