Caspase-1 interdomain linker cleavage is required for pyroptosis

Abstract

Pathogen-related signals induce a number of cytosolic pattern-recognition receptors (PRRs) to form canonical inflammasomes, which activate pro-caspase-1 and trigger pyroptotic cell death. All well-studied inflammasome-forming PRRs oligomerize with the adapter protein ASC (apoptosis-associated speck-like protein containing a CARD) to generate a large structure in the cytosol, which induces the dimerization, autoproteolysis, and activation of the pro-caspase-1 zymogen. However, several PRRs can also directly interact with pro-caspase-1 without ASC, forming smaller "ASC-independent" inflammasomes. It is currently thought that little, if any, pro-caspase-1 autoproteolysis occurs during, and is not required for, ASC-independent inflammasome signaling. Here, we show that the related human PRRs NLRP1 and CARD8 exclusively form ASC-dependent and ASC-independent inflammasomes, respectively, identifying CARD8 as the first canonical inflammasome-forming PRR that does not form an ASC-containing signaling platform. Despite their different structures, we discovered that both the NLRP1 and CARD8 inflammasomes require pro-caspase-1 autoproteolysis between the small and large catalytic subunits to induce pyroptosis. Thus, pro-caspase-1 self-cleavage is a required regulatory step for pyroptosis induced by human canonical inflammasomes.

Figure 1.. NLRP1 is ASC-dependent and CARD8 is ASC-independent.

(A) Human NLRP1, CARD8, and ASC domain organization. The autoproteolysis sites are indicated. (B, C) HEK 293T cells stably expressing CASP1 and GSDMD (HEK 293T^{CASP1 + GSDMD}) were transiently transfected with constructs encoding the indicated proteins and treated with DMSO or VbP (10 μM, 6 h). (B, C) Supernatants were evaluated for LDH release (B) and lysates were analyzed by immunoblotting (C). Data are means ± SEM of three biological replicates. ***P < 0.001 by two-sided t test. (D, E) HEK 293T cells were transfected with constructs encoding GFP-tagged ASC and NLRP1 or CARD8, treated with DMSO or VbP (10 μM, 6 h), and evaluated for ASC speck formation by fluorescence microscopy. The cells were not fixed before analysis. (D, E) Shown are the mean ± SEM (D) and representative images (E) from 10 replicates from one of two independent experiments. ***P < 0.001 by two-sided t test. (F) HEK 293T cells transiently transfected with constructs encoding the indicated proteins and treated with DMSO or VbP (10 μM, 6 h). Lysates were harvested, subjected to disuccinimidyl suberate cross-linking, and evaluated by immunoblotting. All data, including immunoblots, are representative of three or more independent experiments. FL, full-length; WCL, whole cell lysate.

Figure S1.. The NLRP1 CARD forms an ASC-containing inflammasome.

(A) HEK 293T^{CASP1 + GSDMD} cells were transfected with plasmids encoding the UPA-CARD fragments of NLRP1 or CARD8 and ASC as indicated. After 24 h, lysates were evaluated by immunoblotting. (B, C) HEK 293T^{CASP1 + GSDMD} were transfected with plasmids encoding full-length NLRP1 or NLRP1 without a pyrin domain (NLRP1∆PYD). The next day, the cells were pretreated with DMSO or bortezomib (bort., 20 μM, 30 min), which inhibits inflammasome formation, before the DMSO or VbP (10 μM, 6 h). (B, C) Supernatants were evaluated for LDH release (B) and lysates were analyzed by immunoblotting (C). Data are means ± SEM of three biological replicates. **P < 0.01, ***P < 0.001 by two-sided t test. (D, E) HEK 293T cells were transfected with plasmids encoding GFP-tagged ASC and the UPA-CARD fragments of NLRP1 or CARD8, and then evaluated for ASC speck formation by fluorescence microscopy. The cells were not fixed before analysis. (D, E) Shown are the mean ± SEM (D) and representative images (E) from 10 replicates from one of two independent experiments. ***P < 0.001, by two-sided t test. (F) Cell lysates from HEK 293T cells transiently expressing SmBiT-tagged full-length ASC were mixed with lysates expressing the indicated LgBiT-tagged CARD and analyzed for the relative luminescence. All data, including immunoblots, are representative of three or more independent experiments. NS, not significant.

Figure 2.. Specific CARD–CARD interactions determine ASC-dependent or independent inflammasome assembly.

(A) Expression of the indicated LgBiT-tagged CARDs in HEK 293T cells was verified by immunoblotting. (B, C) Cell lysates from HEK 293T cells transiently expressing SmBiT-tagged ASC^CARD (B) or SmBiT-tagged CASP1^CARD (C) were mixed with lysates expressing LgBiT-tagged CARDs and analyzed for the relative luminescence. Data are means ± SEM of four independent replicates.

Figure 3.. Caspase-1 autoproteolysis is required for CARD8-mediated death.

(A, B) Control and GSDMD^−/− THP-1 cells were treated with VbP (10 μM, 24 h) before supernatants were analyzed for LDH release (A) and lysates and supernatants were evaluated by immunoblotting (B). Data are means ± SEM of three biological replicates. ***P < 0.001 by two-sided t test. An asterisk indicates a background band. (C) Schematic of pro-caspase-1 depicting the CARD domain and large (p20, LS) and small (p10, SS) catalytic subunits. Predicted cleavage sites, sizes of potential cleavage products, and the catalytic cysteine are indicated. (D, E) HEK 293T cells stably expressing GSDMD and the indicated pro-caspase-1 constructs were transiently transfected with plasmids encoding RFP (mock), CARD8 WT, or autoproteolysis-defective CARD8 S297A (SA) for 24 h before addition of VbP (10 μM, 6 h). (D, E) Cell death was assessed by LDH release (D) and GSDMD and CASP1 cleavage by immunoblotting (E). Data are means ± SEM of three biological replicates. ***P < 0.001 by two-sided t test. All data, including immunoblots, are representative of three or more independent experiments. CL, cleaved; FL, full-length.

Figure 4.. Human pro-caspase-1 interdomain linker cleavage is required for pyroptosis.

(A) HEK 293T cells stably expressing GSDMD were transiently transfected with plasmids encoding RFP (mock), the indicated pro-caspase-1 constructs, and the CARD8^UPA-CARD for 24 h before lysates were analyzed by immunoblotting. Asterisks indicate background bands. (B) HEK 293T cells stably expressing the indicated pro-caspase-1 constructs were transiently transfected with plasmids encoding CARD8^UPA-CARD or NLRP1^UPA-CARD and ASC for 24 h before lysates were evaluated by immunoblotting. (C, D) Control and CASP1^−/− THP-1 cells ectopically expressing the indicated pro-caspase-1 proteins were treated with DMSO or VbP (10 μM, 24 h). (C, D) GSDMD cleavage was assessed by immunoblotting (C) and cell death by LDH release (D). Data are means ± SEM of three biological replicates. ***P < 0.001 by two-sided t test. An asterisk indicates background bands. All data, including immunoblots, are representative of three or more independent experiments. CL, cleaved; FL, full-length.

Figure S2.. Human caspase-1 autoproteolysis is critical for pyroptosis.

(A, B) HEK 293T cells stably expressing the indicated pro-caspase-1 constructs were transiently transfected with plasmids encoding CARD8-Flag (0.05 μg) or NLRP1-Flag (0.1 μg) and ASC (0.01 μg) without GSDMD (A) or with GSDMD (0.01 μg) (B). After 24 h, the cells were treated with VbP (10 μM, 6 h) before lysates were analyzed by immunoblotting. (C) HEK 293T cells stably expressing the indicated pro-caspase-1 were transiently transfected with residues 1–328 of NLRC4 for 24 h before lysates were evaluated by immunoblotting. (D) Schematic of the DmrB-caspase-1 constructs. Predicted cleavage sites, sizes of potential cleavage products, and the catalytic cysteine are indicated. (E) HEK 293T cells stably expressing GSDMD were transiently transfected with the indicated DmrB-caspase-1 constructs for 24 h before addition of AP-20187 (500 nM, 1 h). GSDMD and CASP1 cleavage were evaluated by immunoblotting. All data, including immunoblots, are representative of three or more independent experiments. CL, cleaved; FL, full-length.

Figure S3.. ASC activates apoptosis in the absence of caspase-1 activity.

(A) HEK 293T cells or HEK 293T cells ectopically expressing pro-caspase-1 C285A were transiently transfected with plasmids encoding RFP (mock, 2.0 μg), ASC (0.01 μg), CARD8^UPA-CARD (0.02 μg), or CARD8^UPA-CARD plus ASC. After 24 h, the cells were analyzed by immunoblotting. (B) HEK 293T cells were transiently transfected with plasmids encoding RFP (mock, 2.0 μg), ASC (0.01 μg), NLRP1^UPA-CARD (0.02 μg), or NLRP1^UPA-CARD plus ASC, or the indicated plasmids plus CASP1 C285A (0.01 μg), for 24 h and then the cell lysates were evaluated by immunoblotting. All data, including immunoblots, are representative of three or more independent experiments.

Figure 5.. Cleavage of the mouse pro-caspase-1 interdomain linker is critical for pyroptosis.

(A) Schematic of pro-caspase-1 domain organization and interdomain linker sequences of the indicated species. (B, C, D, E) HEK 293T cells stably expressing mouse GSDMD were transiently transfected with plasmids encoding the indicated mouse pro-caspase-1 proteins (0.01 μg) ± NLRP1B^UPA-CARD (0.02 μg) for 24 h before the collection of supernatants for quantification of LDH release (B, D) and harvesting cell lysates for immunoblots (C, E). (F, G) HEK 293T cells stably expressing various mouse pro-caspase-1 constructs were evaluated by immunoblotting after transient transfection of plasmids encoding NLRP1B^UPA-CARD alone (0.02 μg) (F) or together with mouse GSDMD (0.01 μg) (G) for 24 h. (F, G) Cropped images in (F, G) are from the same membrane. Data are mean values ± SEM of three biological replicates. *P < 0.01, **P < 0.01, ***P < 0.001 by two-sided t test. All data, including immunoblots, are representative of three or more independent experiments.