Human NLRP1 is a sensor of pathogenic coronavirus 3CL proteases in lung epithelial cells

Abstract

Inflammation observed in SARS-CoV-2-infected patients suggests that inflammasomes, proinflammatory intracellular complexes, regulate various steps of infection. Lung epithelial cells express inflammasome-forming sensors and constitute the primary entry door of SARS-CoV-2. Here, we describe that the NLRP1 inflammasome detects SARS-CoV-2 infection in human lung epithelial cells. Specifically, human NLRP1 is cleaved at the Q333 site by multiple coronavirus 3CL proteases, which triggers inflammasome assembly and cell death and limits the production of infectious viral particles. Analysis of NLRP1-associated pathways unveils that 3CL proteases also inactivate the pyroptosis executioner Gasdermin D (GSDMD). Subsequently, caspase-3 and GSDME promote alternative cell pyroptosis. Finally, analysis of pyroptosis markers in plasma from COVID-19 patients with characterized severe pneumonia due to autoantibodies against, or inborn errors of, type I interferons (IFNs) highlights GSDME/caspase-3 as potential markers of disease severity. Overall, our findings identify NLRP1 as a sensor of SARS-CoV-2 infection in lung epithelia.

Keywords: 3CL proteases; Gasdermins; NLRP1 inflammasome; SARS-CoV-2; epithelial cells; pyroptosis.

Graphical abstract

Figure 1

NLRP1 is an innate immune sensor of SARS-CoV-2 infection (A) Western blot examination of NLRP1, NLRP3, and ACTIN in various human epithelial cells and cell lines. (B) Cell death (LDH) and viability (ATP Glo) evaluation in A549^ACE2/NLRP1+ and A549^{ACE2/NLRP1−} airway epithelial cell lines infected with various multiplicity of infection (MOI) of SARS-CoV-2 for 24 h. (C) Florescence microscopy and associated quantifications of ASC-GFP specks in A549^{ACE2/NLRP1+/ASC-GFP} and A549^{ACE2/NLRP1−/ASC-GFP} airway epithelial cell lines infected with SARS-CoV-2 (MOI 0.05) for 24 h. Nucleus was stained with Hoechst (blue), and nucleocapsid (N) was stained in red after fixation. Brightfield/ASC-GFP pictures were taken in dish during cell infection. Images shown are from one experiment and are representative of n = 3 independent experiments. For quantifications, the percentage of cells with ASC complexes was determined by determining the ratios of cells positives for ASC speckles on the total nuclei. At least 10 fields from n = 3 independent experiments were analyzed. Values are expressed as mean ± SEM. (D) Cell death (LDH) evaluation in A549^NLRP1+ and A549^NLRP1− airway epithelial cell lines infected with SARS-CoV-2 (MOI 0.05) for 24 h in the presence/absence of the pan-caspase inhibitor Z-VAD (25 μM), NSP5 protease inhibitor PF-00835231 (10 μM), or the RNA-dependent RNA polymerase (RdRp) inhibitor remdesivir (5 μM). (E) Florescence microscopy and associated quantifications of ASC-GFP specks in A549^{ACE2/NLRP1+/ASC-GFP} and A549^{ACE2/NLRP1−/ASC-GFP} airway epithelial cell lines infected with SARS-CoV-2 (MOI 0.05) for 24 h. Nucleus was stained with Hoechst (blue). Brightfield/ASC-GFP pictures were taken in dish during cell infection. Images shown are from one experiment and are representative of n = 3 independent experiments; scale bars, 10 μm. For quantifications, the percentage of cells with ASC complexes was determined by determining the ratios of cells positives for ASC speckles on the total nuclei. At least 10 fields from n = 3 independent experiments were analyzed. Values are expressed as mean ± SEM. (F) Cell death (LDH) evaluation in NHBE airway epithelial cells infected with various multiplicity of infection (MOI) of SARS-CoV-2 for 36 h in the presence/absence of the pan-caspase inhibitor Z-VAD (25 μM), Necroptosis inhibitor Necrostatin-1s (Nec1s, 30 μM), or the Ferroptosis inhibitor Ferrostatin-1 (Fe1, 10 μM). (G) Western blot characterization of genetic invalidation of NLRP1 in NHBE population using CRISPR-Cas9 and measure of cell lysis (LDH release) in NHBE^WT and NHBE^NLRP1−/− airway epithelial cells infected with various multiplicity of infection (MOI) of Wuhan or Delta variant SARS-CoV-2 for 36 h. Efficiency of genetic invalidation by single-guide RNAs (sgRNAs were evaluated at the whole cell population. Data information: western blot (A and G) images are from one experiment performed 3 times. Graphs (B), (D), (F), and (G) show data presented as means ± SEM from n = 3 (F and G), n = 4 (B) and n = 6 (D) independent pooled experiments; ^∗∗∗p ≤ 0.001 for the indicated comparisons with t test. Images (C and E) are representative of one experiment performed 3 times.

Figure 2

SARS-CoV-2 NSP5 protease-cleaved NLRP1 at the Q333 site nucleates NLRP1 inflammasome (A) Florescence microscopy and associated quantifications of ASC-GFP specks in A549^{NLRP1+/ASC-GFP} and A549^{NLRP1−/ASC-GFP} airway epithelial cell lines infected with SARS-CoV-2 (MOI 0.05) for 24 h in the presence or absence of proteasome inhibitor bortezomib (0.1 μM) or inhibitor of the glycine N-degron pathway MLN4924 (1 μM). Images shown are from one experiment and are representative of n = 3 independent experiments; scale barss 10 μm. For quantifications, the percentage of cells with ASC complexes was determined by determining the ratios of cells positives for ASC speckles on the total cells presents in the wells. At least 10 fields from n = 3 independent experiments were analyzed. Values are expressed as mean ± SEM. (B) Western blot examination of NLRP1 cleavage using an anti-NLRP1 N-terminal antibody (aa 1–323) upon coincubation of SARS-CoV-2, SARS-CoV-1, or MERS-CoV 3CL (NSP5) proteases (5 μM) with A549^NLRP1+ airway epithelial cell lysates in presence or absence of the 3CL inhibitors GC-376 (10 μM) or PF-00835231 (10 μM). NLRP1 N-terminal, NLRP1 C-terminal, NSP5, and ACTIN were immunoblotted. (C) Florescence microscopy and associated quantifications of ASC-GFP specks in A549^{NLRP1+/ASC-GFP} airway epithelial cell lines transduced with a doxycycline (dox)-inducible plasmid encoding NSP5 or its catalytically inactive mutant NSP5^C145A. Images shown are from one experiment and are representative of n = 3 independent experiments; scale bars, 10 μm. For quantifications, the percentage of cells with ASC complexes was determined by determining the ratios of cells positives for ASC speckles on the total cells presents in the wells. At least 10 fields from n = 3 independent experiments were analyzed. Values are expressed as mean ± SEM. (D) Schematic representation of the approximate NLRP1 N-terminal fragment generated by NSP5 protease cut. (E) Western blot examinations of the ability of NSP5 to cleave various NLRP1 constructs mutated in glutamine (Q) at various sites. Immunoblots show anti-N-terminal NLRP1, ACTIN, and NSP5. (F) Florescence microscopy and associated quantifications of ASC-GFP specks in A549^{NLRP1+/ASC-GFP} or A549^{NLRP1Q333A/ASC-GFP} airway epithelial cell lines transduced with a doxycycline (dox)-inducible plasmid encoding NSP5 or its catalytically inactive mutant NSP5^C145A. Images shown are from one experiment and are representative of n = 3 independent experiments; scale bars, 10 μm. For quantifications, the percentage of cells with ASC complexes was determined by determining the ratios of cells positives for ASC speckles on the total nuclei (Blue). At least 10 fields from n = 3 independent experiments were analyzed. Values are expressed as mean ± SEM. (G) Cell death (LDH) evaluation in A549^NLRP1+ and A549^NLRP1− airway epithelial cell lines expressing a doxycycline (dox)-inducible plasmid encoding NSP5 in the presence or absence of the proteasome inhibitor bortezomib (0.1 μM) or inhibitor of the glycine N-degron pathway MLN4924 (1 μM). (H) Western blot examination of NLRP1 cleavage using an anti-NLRP1 N-terminal antibody (aa 1–323) after infection of A549^NLRP1−, A549^NLRP1+ A549^NLRP1Q333A, or A549^NLRP1Q130A airway epithelial cells with SARS-CoV-2 (MOI 0.05) for 24 h in the presence/absence of the proteasome inhibitor bortezomib (0.1 μM, Bort.). NLRP1 N-terminal, NSP5, and ACTIN were immunoblotted. ^∗NS: prominent nonspecific bands, not specific. (I) Cell death (LDH) evaluation in A549^NLRP1+ and A549^NLRP1Q333A airway epithelial cell lines infected with SARS-CoV-2 (MOI 0.05) for 24 h. (J) Western blot examination of NLRP1 cleavage using an anti-NLRP1 N-terminal antibody (aa 1–323) after infection of NHBE^WT airway epithelial cells with SARS-CoV-2 (MOI 1) for 36 h in the presence/absence of the proteasome inhibitor bortezomib (0.1 μM) or inhibitor of the glycine N-degron pathway MLN4924 (1 μM). NLRP1 N-terminal, nucleocapsid, and ACTIN were immunoblotted. (K) Measure of cell lysis (LDH release) in NHBE^WT and NHBE^NLRP1−/− airway epithelial cells infected with SARS-CoV-2 (MOI 1) for 36 h in the presence/absence of the proteasome inhibitor bortezomib (0.1 μM) or inhibitor of the glycine N-degron pathway MLN4924 (1 μM). Data information: images (A, C, and F) show one experiment performed 3 times. Western blot (B, E, H, and J) images are from one experiment performed 3 times. Graphs (C, G, I, and K) show data presented as means ± SEM from n = 3 independent pooled experiments; ^∗∗∗p ≤ 0.001 for the indicated comparisons with t test.

Figure 3

NSP5 protease cleaves Gasdermin D in its pore-forming domain (A) Cell death (LDH) evaluation in A549^NLRP1+, A549^NLRP1−, or A549^{NLRP1+/GSDMD−} airway epithelial cell lines infected for 24 h with SARS-CoV-2 (MOIs 0.1, 0.01, and 0.001) or stimulated with doxycycline (dox)-induced NSP5 expression. (B) Measure of cell lysis (LDH release) and cell viability (Cell titer Glo) in NHBE^WT and NHBE^GSDMD−/− airway epithelial cells infected with various SARS-CoV-2 viral strains (MOI 1) for 36 h. (C) Western blot examination of Gasdermin D (GSDMD) processing in A549^NLRP1+ cells infected with SARS-CoV-2 at MOI of 0.1 for 24 h. GSDMD was immunoblotted using an anti-C-terminal antibody (recognizes full-length and C-terminal cleaved forms of GSDMD) or with an anti-GSDMD active N-terminal fragment (30 kDa) specific antibody. NLRP1, ACTIN, and SARS-CoV-2 nucleocapsid were also evaluated. (D) Western blot examination of GSDMD and NLRP1 cleavages upon coincubation of SARS-CoV-2 NSP3 protease or SARS-CoV-2/SARS-CoV1 3CL (NSP5) proteases (5 μM) with A549^NLRP1+ or A549^NLRP1− cell lysates in the presence or absence of the 3CL inhibitor PF-00835231 (10 μM). GSDMD (anti C-terminal), NLRP1 N-terminal, NSP5, and ACTIN were immunoblotted. (E) Coomassie observation of recombinant GSDMD cleavage by various amounts of SARS-CoV-2 NSP5 protease and top-down mass-spectrometry identification of GSDMD-cleaved fragments. In blue are represented the various GSDMD fragments identified upon NSP5 coincubation. In red is the NSP5 protease detected by mass spectrometry. (F) Western blot examination and schematic representation of GSDMD cleavage by SARS-CoV-2 3CL (NSP5) or recombinant human caspase-1 (CASP1) proteases in cell lysates from A549 expressing WT GSDMD or GSDMD^193A constructs. GSDMD (anti-C-terminal), NSP5, and ACTIN were immunoblotted. (G) Cell death (LDH) evaluation in A549 cells expressing doxycycline-inducible GSDMD fragments, including GSDMD full-length (FL), caspase-1-generated active GSDMD (1–275), or NSP5-generated 1–193 and 194–484 GSDMD fragments. Cell lysis was determined 18 h after doxycycline (dox) addition in the culture medium. (H and I) Cell death (LDH) evaluation in A549^{NLRP1+/GSDMD−} cells complemented or not with constructs coding for WT GSDMD or GSDMD^193A. Cells were transduced with dox-inducible NSP5 plasmids, treated with the NLRP1 activator Val-boro (10 μM) or infected (I) with SARS-CoV-2 (MOI 0.01). Cell lysis was determined 18 h after doxycycline (dox) addition, 10 h after Val-boro addition in the culture medium, or 24 h after infection. Data information: graphs (A, B, G, H, and I) show data presented as means ± SEM from n = 3 independent pooled experiments; ^∗∗∗p ≤ 0.001 for the indicated comparisons with t test. Western blot (C, D, and F) images are from one experiment performed 3 times. Images (E) show one experiment performed 3 times.

Figure 4

NLRP1 engages a caspase-3/Gasdermin E-dependent pyroptosis pathway upon SARS-CoV-2 infection (A) Western blot examination of Gasdermin E, caspases-3, and caspases-8 processing in A549^NLRP1+ and A549^NLRP1− cells after 24 h of infection with SARS-CoV-2 (MOI 0.05) in the presence or absence of the pan-caspase inhibitor Z-VAD (25 μM). Immunoblots were performed against full-length and processed forms of Gasdermin E (p55 and p30), caspase-8 (p54 and p15), caspase-3 (p35 and p17/19), SARS-CoV-2 nucleocapsid (p40), NLRP1 N-terminal (p130/110), and ACTIN (p40). (B) Western blot examination of Gasdermin E and caspases-3 processing in NHBE^WT and NHBE^NLRP1−/− cells after 36 h of infection with SARS-CoV-2 (MOI 1). Immunoblots were performed against full-length and processed forms of Gasdermin E (p55 and p30), caspase-3 (p35 and p17/19), SARS-CoV-2 nucleocapsid (p40), NLRP1 N-terminal (p130/110), and ACTIN (p40). (C and D) Measure of caspase-1 (C) and caspase-3/-7 (D) activities in SARS-CoV-2-infected (MOI 0.5) NHBE^WT or A549^NLRP1+ cells for 36 h in the presence or absence of inhibitors of caspase-1 (Z-YVAD, 40 μM) or caspase-3/-7 (Z-DEVD, 30μM). Val-boro (5 μM) was used a positive control of NLRP1-driven caspase activity for 10 H. (E) Measure of cell lysis (LDH release) in A549^NLRP1+ or NHBE-infected cells with SARS-CoV-2 (MOI 0.05 and 1, respectively) for 24 h in the presence/absence of the pan-caspase inhibitor Z-VAD (25 μM), the caspase-1 inhibitor Z-YVAD (40 μM), the caspase-8 inhibitor Z-IETD (40 μM), or the caspase-3 inhibitor Z-DEVD (30 μM). (F) Western blot characterization of genetic invalidation of CASP3 in A549 population cells using CRISPR-Cas9 approaches and measure of cell lysis (LDH release) in A549^NLRP1+ or A549^{NLRP1+/CASP3−}-infected cells with SARS-CoV-2 (MOI 0.05) for 24 h in the presence/absence of the pan-caspase inhibitor Z-VAD (25 μM), the caspase-1 inhibitor Z-YVAD (40 μM), or the caspase-3 inhibitor Z-DEVD (30 μM). Efficiency of genetic invalidation by single-guide RNAs (sgRNAs) targeting GFP or caspase-3 was evaluated at the whole cell population. (G) Western blot characterization of genetic invalidation of GSDME in A549 population cells using CRISPR-Cas9 approaches and measure of cell lysis (LDH release) in A549^NLRP1+, A549^{NLRP1+/GSDMD−}, or A549^{NLRP1+/GSDME−}-infected cells with SARS-CoV-2 (MOIs 0.001, 0.01, and 0.1) for 24 h. Efficiency of genetic invalidation by single-guide RNAs (sgRNAs) targeting GFP or GSDME was evaluated at the whole cell population. (H) Western blot characterization of genetic invalidation of GSDME in NHBE population cells using CRISPR-Cas9 approaches and measure of cell lysis (LDH release) in NHBE^WT or NHBE^GSDME−/−-infected cells with SARS-CoV-2 (MOIs 0.1, 0.5, and 1) for 36 h. Efficiency of genetic invalidation by single-guide RNAs (sgRNAs) targeting GFP or GSDME was evaluated at the whole cell population. Data information: western blot (A, B, and E–G) images are from one experiment performed 3 times. Graphs (C–G) show data presented as means ± SEM from n = 3 independent pooled experiments; ^∗∗∗p ≤ 0.001 for the indicated comparisons with t test.

Figure 5

Caspase-3/IL-16 and GSDME as potential markers of COVID-19 severity (A) Measure of the presence of various inflammatory mediators in plasmas from hospitalized patients presenting COVID-19 disease and analyzed according to their disease severity degree (n = 60 patients, including 15 with moderate COVID-19, 15 with moderate COVID-19 on admission, 15 with severe COVID-19 on admission, and 15 IFN alterations with severe COVID-19). Samples were prepared at day 0 posthospitalization. Data information: data shown as means from n = 12 different donors per category (moderate/moderate → severe/severe/IFN alterations); each category is represented with a colored circle; ^∗ p ≤ 0.05, ^∗∗p ≤ 0.01, ^∗∗∗p ≤ 0.001 for the indicated comparisons using t test with Bonferroni correction.