Dipeptidyl peptidase 9 sets a threshold for CARD8 inflammasome formation by sequestering its active C-terminal fragment.

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and requests for resources and reagents should be directed and will be fulfilled by the Lead Contact, Hao Wu ( wu@crystal.harvard.edu ). Study plasmids will be made available on Addgene ( https://www.addgene.org/Hao_Wu/ and https://www.addgene.org/Daniel_Bachovchin/ ). Extended protocol are available on protocols.io ( https://www.protocols.io/groups/hao-wu-lab ). Pymol and chimera session files, in addition to other raw data, will be made available on our Open Science Framework ( https://osf.io/x7dv8/ ). Cryo-EM data were deposited to EMPIAR, the EMDB, and the PDB.

EXPERIMENTAL MODEL AND SUBJECT DETAILS HEK 293T Cell Lines

The human kidney epithelial cell line HEK 293T (ATCC) and related CASP1/GSDMD-expressing stable cell lines and DPP8 −/− , DPP9 −/− lines were maintained in DMEM (GIBCO, ThermoFisher) supplemented with 10% fetal bovine serum (GIBCO, ThermoFisher Scientific), at 37°C, and 5% CO 2 . HEK 293T cells were verified by the manufacturer. Cells were frequently checked for morphological features and tested for mycoplasma using MycoAlert Mycoplasma Detection kit (Lonza). Expi293F Cell Line Expi293F suspension cells were maintained in Expi293F Expression Medium (GIBCO, ThermoFisher) with constant shaking at 100 RPM, 37°C, 5% CO 2 . Expi293F cells were not authenticated nor tested for mycoplasma contamination. Sf9 Cell Line Sf9 insect cells were maintained in HyClone SFX-Insect Cell Media (Cytiva) supplemented with 1X antibiotic-antimycotic (ThermoFisher) at 27°C with constant shaking at 100 RPM. Sf9 cells were recently purchased from the manufacturer and were not authenticated.

THP-1 Cell Lines

THP-1 cells

(ATCC) and related THP-1 CARD8 −/− stable cell lines were grown in Roswell Park Memorial Institute (RPMI) medium 1640 with L-glutamine and 10% fetal bovine serum (GIBCO, ThermoFisher Scientific), at 37°C and 5% CO 2 . Cells were monitored for morphological features and regularly tested for mycoplasma using MycoAlert Mycoplasma Detection kit (Lonza). METHODS DETAILS Constructs and Cloning Full-length CARD8 (T60 isoform, Uniprot ID Q9Y2G2-5) was cloned into pcDNA3.1 LIC 6A (Addgene plasmid #30124) with a C-terminal FLAG tag and a modified pcDNA3.1 LIC 6D (Addgene plasmid #30127) construct (C-terminal TEV-GFP-FLAG tag). The short isoform of DPP9 (DPP9S, Uniprot ID Q86TI2-1) was also cloned into pcDNA3.1 LIC 6A (N-terminal FLAG-TEV tag or N-terminal His-TEV tag). CARD8 (with several synonymous mutations to avoid CRISPR/Cas9 editing) was also shuttled into pInducer20 vector (Addgene, #44012) using Gateway technology (Thermo Fischer Scientific) and pLEX_305-N-dTAG (Addgene, #91797). CARD8-ZUC containing the ZU5, UPA and CARD was cloned as previously described ( Chui et al., 2020 ) and shuttled to pLEX_305-N-dTAG (Addgene, #91797). CARD8-CT constructs were synthesized (GenScript) with an N-terminal ubiquitin sequence followed by CARD8 (S297-L537), cloned into the pcDNA3.1 vector (Ub-CARD8-CT) and shuttled into the pLEX307 vector using Gateway technology (Thermo Fischer Scientific). Point mutations were introduced with Q5 site-directed mutagenesis (NEB) or QuikChange site-directed mutagenesis (Agilent). Most constructs will be made available on Addgene. Cell Culture HEK 293T cells and THP-1 cells were purchased from ATCC. HEK 293T cells were grown in Dulbecco’s Modified Eagle’s Medium (DMEM) with L-glutamine and 10% fetal bovine serum (FBS). THP-1 cells were grown in Roswell Park Memorial Institute (RPMI) medium 1640 with L-glutamine and 10% FBS. All cells were grown at 37° C in a 5% CO 2 atmosphere incubator. Cell lines were regularly tested for mycoplasma using the MycoAlert Mycoplasma Detection Kit (Lonza). THP-1 CARD8 −/− cells and HEK 293T cells stably expressing caspase-1 and GSDMD were generated as previously described ( Johnson et al., 2020 ). Reconstituted THP-1 CARD8 −/− cells were generated via lentiviral infection of THP-1 CARD8 −/− cells with the indicated pInducer20 construct followed by selection with G418 (Geneticin) at 200 μg/mL until all control cells were dead (approximately 14 days). dTAG-ZUC THP-1 cells were generated via lentiviral infection followed by selection with puromycin 500 ng/mL until all control cells were dead (approximately 7 days).

Protein Expression and Purification

To express CARD8-DPP9 complexes, Expi293F cells (1 L, 2–3 × 10 6 cells/mL) were co-transfected with CARD8-TEV-GFP-FLAG (0.7 mg) and DPP9S (0.3 mg) following incubation with polyethylenimine (3 mL, 1 mg/mL) in Opti-MEM (100 mL) for 30 min. 24 h later, cells were supplemented with glucose (9 mL, 45%) and valproic acid (10 mL, 300 mM). Cells were harvested 5 d after transfection by centrifugation (2,000 RPM, 20 min), washed once with PBS, split into 3 pellets, flash-frozen in liquid nitrogen, and stored at −80 °C. Later, one thawed pellet was resuspended in lysis buffer (50–100 mL, 25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM TCEP), sonicated (2 s on 8 s off, 3.5 min total on, 40% power, Branson), and ultracentrifuged at 40,000 RPM for 1 h (45 Ti fixed-angle rotor, Beckman). The supernatant was incubated with pre-equilibrated (lysis buffer) anti-FLAG M2 affinity gel (Sigma, 1.0 mL) for 4 h at 4 °C, washed in batch once with lysis buffer (5 mL), and then washed by gravity flow with 25–50 column volumes (CV) lysis buffer. The CARD8-DPP9 complex was eluted by on-column cleavage at room temperature for 1 h using elution buffer (5 mL, 25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM TCEP, 0.2 mg TEV protease) and loaded onto a Mono Q 5/50 GL anion exchange column (Cytiva). Protein was eluted using a salt gradient from 150 mM to 1 M NaCl (25 mM Tris-HCl pH 8.0, 1 mM TCEP) over 15 CV. Mono Q eluent was concentrated using a 0.5 mL spin concentrator (Amicon Ultra, 50 kDa MW cutoff) to 0.5 mg/mL (assuming ε=1). Concentrated eluent was dialyzed overnight into EM buffer (25 mM HEPES pH 7.5, 150 mM NaCl, 1 mM TCEP) using a 0.5 mL Slide-A-Lyzer (ThermoFisher). Total protein yield varied between 2–3 mg per L of mammalian culture.

Expression and purification of CARD8

(S297A) followed an identical protocol. Expression and purification of the VbP-bound complex followed an identical protocol except for the addition of 10 μM VbP to all purification and dialysis buffers.

Cryo-EM Screening and Data Collection

Grids were screened at University of Massachusetts Worcester, Pacific Northwest Center for Cryo-EM (PNCC), and Harvard Medical school (HMS) using a Talos Arctica microscope (ThermoFisher) operating at an acceleration voltage of 200 keV equipped with a direct electron detector. Small initial dataset collection revealed severe particle orientation preference for the CARD8-DPP9 maps, causing anisotropic resolution, in addition to dissociation of CARD8-DPP9 complexes. To remedy these issues, we crosslinked samples to avoid dissociation and tilted the stage to collect missing views (below). The purified DPP9S-CARD8-WT or DPP9S-CARD8-S297A complex (0.40 mg/mL assuming ε=1; 25 mM HEPES pH 7.5, 150 mM NaCl, 1 mM TCEP, ±10 μM VbP) was crosslinked with 0.02% glutaraldehyde on ice for 5 min and immediately loaded onto a glow-discharged Quantifoil grid (R1.2/1.3 400-mesh gold-supported holey carbon, Electron Microscopy Sciences), blotted for 3–5 s under 100% humidity at 4 °C, and plunged into liquid ethane using a Mark IV Vitrobot (ThermoFisher). Grids were screened at Harvard Medical School for ice and particle quality prior to data collection. For data collection, movies were acquired at Harvard Medical School (DPP9-CARD8-WT) and National Center for CryoEM Access and Training (NCCAT) in New York (DPP9-CARD8-WT+VbP and DPP9-CARD8-S297A) using a Titan Krios microscope (ThermoFisher) at an acceleration voltage of 300 keV equipped with a BioQuantum K3 Imaging Filter (slit width 20 eV). Movies were recorded with a K3 Summit direct electron detector (Gatan) operating in counting mode at 105,000 × (0.825 Å/pix at HMS, 0.82 or 0.83 at Å/pix NCCAT). For CARD8-DPP9 at 0° and 37°: 3,306 and 2,488 movies at a stage tilt of either 0° or 37° were collected using SerialEM ( Mastronarde, 2005 ) at varying defocus values ranging between −0.8 to −2.2 μm and −1.5 to −3.0 μm, respectively. We used image shift to record two shots for each of the four holes per stage movement. Movies were exposed with a total dose of 58.5 e − /Å 2 for 2.22 s fractionated over 49 frames for 0° stage tilt movies and 64.99 e − /Å 2 for 2.25 s fractionated over 49 frames for 37° stage tilt movies. For CARD8-DPP9-VbP at 0° and 37°: 1,811 and 6,642 movies at a stage tilt of either 0° or 37°, respectively, were collected using Leginon ( Suloway et al., 2005 ) at varying defocus values ranging between −1.1 to −3.4 μm. One shot was recorded for each of the four holes per stage movement through image shift. All movies were exposed with a total dose of 67.06 e − /Å 2 for 1.5 s fractionated over 50 frames. For CARD8(S297A)-DPP9, 3,840 movies were collected at a stage tilt of 37° using Leginon ( Suloway et al., 2005 ) to vary the defocus range between −0.8 to −2.5 μm and to record one shot for each of the four holes per stage movement through image shift. All movies were exposed with a total dose of 63.67 e − /Å 2 for 1.5 s fractionated over 50 frames.

Cryo-EM Data Processing

Data processing leveraged SBgrid Consortium ( Morin et al., 2013 ) for support and computing resources. Movies collected at Harvard Medical School (CARD8-DPP9) were pre-processed on-the-fly by the facility’s pipeline script. Movies were corrected by gain reference and for beam-induced motion and summed into motion-corrected images using the Relion 3.08 implementation of the MotionCor2 algorithm ( Zheng et al., 2017 ). The CTFFIND4 program ( Rohou and Grigorieff, 2015 ) was used to determine the defocus of each micrograph. Relion 3.1 ( Scheres, 2012 ; Zivanov et al., 2018 ) was used for subsequent image processing. Movies collected at NCCAT (CARD8-DPP9-VbP) were similarly pre-processed using Relion 3.1. For the CARD8-DPP9 complex, template-free autopicking with crYOLO ( Wagner et al., 2019 ) (generalized training for on-the-fly picking at the HMS cryo-EM center) selected 487,952 particles from 3,306 micrographs, which were subjected to a single round of 2D classification. The heterogenous nature of the sample was evident in 2D class averages because complexes containing only DPP9, 2CARD8:2DPP9, and 4CARD8:2DPP9 were present. A randomized set of 100,000 particles was selected for the de novo reconstruction of an initial model, which was low-pass-filtered to 40 Å to use as the input reference for 3D classification. Multiple rounds of 3D classification including global and local fine angular search were performed. After visual inspection, one class with 2CARD8:2DPP9 (62,018 particles) was selected for 3D refinement. Subsequently, particles were CTF refined and Bayesian polished to reach an overall resolution of 3.3 Å. However, the map suffered from anisotropic resolution, and cryoEF ( Naydenova and Russo, 2017 ) analysis estimated that data collected at a tilt angle of 37° was ideal to fill the gaps in Fourier space. The overall processing scheme for tilt data was derived from the work flow of tilt dataset described in Zivanov et al. 2018 ( Zivanov et al., 2018 ). For CARD8-DPP9 data collected at a 37° tilt, micrographs were first motion-corrected with MotionCor2 ( Zheng et al., 2017 ) followed by Gctf ( Zhang, 2016 ) to calculate per micrograph defocus values in Relion 3.1. Template-free autopicking with crYOLO ( Wagner et al., 2019 ) then picked 313,425 particles from 2,488 micrographs. After multiple rounds of 2D classification, 243,459 visually homogenous particles remained for further processing. A random subset comprising 121,730 of these particles was used for de novo initial model construction, which was then low-passed filtered (30 Å) and used as the initial reference map for 3D classification. After one round of 3D classification, a 3D class with 58,980 particles was selected for high-resolution refinement. The first 3D refinement using these particles yielded a 5.2 Å resolution structure and the Fourier shell correlation (FSC) curve showed strong fluctuations indicating imprecisions in CTF estimation. We then utilized CTF refinement implemented in Relion 3.1, including higher order aberrations correction, anisotropic magnification corrections, and per-particle defocus estimation ( Zivanov et al., 2020 ). Iterative rounds of 3D refinement followed by CTF refinement and Bayesian polishing gradually improved the resolution and this iterative process was stopped when we observed a resolution plateau at 3.8 Å. Refined and polished particle sets from 0° and 37° stage-tilt data were merged and 3D refinement was performed. The merged data was then reconstructed to give an overall resolution of 3.3 Å as calculated by gold-standard FSC between half maps, with much improved CARD8 density ( Figure S1 ). For the CARD8-DPP9-VbP complex, template-free autopicking with crYOLO ( Wagner et al., 2019 ) (trained with 10 manually-picked high-contrast micrographs) yielded 1,404,573 particles from 1,811 micrographs that were collected without stage tilt. These particles were subjected to multiple rounds of 2D classification that yielded visually homogeneous 2D classes with 471,255 particles. A randomized set of 100,000 particles was selected for the de novo reconstruction of an initial model, which was low-pass-filtered to 30 Å to use as the input reference for 3D classification. Multiple rounds of 3D classification including global and local fine angular search were performed. After visual inspection, one 3D class (89,909 particles) was selected for 3D refinement. Subsequently, particles were CTF refined and Bayesian polished to reach an overall resolution of 3.5 Å. However, the map suffered from anisotropic resolution, and cryoEF ( Naydenova and Russo, 2017 ) analysis estimated that data collected at a tilt angle of 37° degrees was ideal to fill the gaps in Fourier space (below). The overall processing scheme for tilt data analysis of CARD8-DPP9-VbP was derived from CARD8-DPP9 data processing (described above). Briefly, micrographs were first motion-corrected with MotionCor2 ( Zheng et al., 2017 ) followed by Gctf ( Zhang, 2016 ) to calculate per micrograph defocus values in Relion 3.1. Template-free autopicking with crYOLO ( Wagner et al., 2019 ) then picked 1,889,993 particles from 6642 micrographs. After multiple rounds of 2D classification, 840,902 visually homogenous particles remained for further processing. A random subset comprising 100,000 of these particles was used for de novo initial model construction, which was then low-passed filtered (30 Å) and used as the initial reference map for 3D classification. After multiple rounds of 3D classification, a 3D class with 91,254 particles was selected for high-resolution refinement. The first 3D refinement using these particles yielded a 5.8 Å resolution structure, and the FSC curve showed strong fluctuations indicating imprecisions in CTF estimation. We then utilized CTF refinement implemented in Relion 3.1, including correction for higher order aberrations and anisotropic magnification, and per-particle defocus estimation ( Zivanov et al., 2020 ). Iterative rounds of 3D refinement followed by CTF refinement and Bayesian polishing gradually improved the resolution and this iterative process was stopped when we observed a resolution plateau at 3.4 Å. Refined particle sets from 0° and 37° stage-tilt data were merged and 3D refinement was performed. The merged data was then 3D classified with fine local search angles to give a final stack of 146,101 particles which was reconstructed to give an overall resolution of 3.3 Å as calculated by gold-standard FSC between half maps, with much improved CARD8 density ( Figure S4 ). The DPP9-CARD8-S297A complex was processed in cryoSPARC ( Punjani et al., 2017 ). 3,840 movies collected at 37° stage tilt were summed into motion-corrected micrographs with Patch-Motion in cryoSPARC ( Punjani et al., 2017 ). Next, defocus values were determined with cryoSPARC’s Patch-CTF function ( Punjani et al., 2017 ). 5,505,288 particles were picked by blob picking and the 2D classes following two rounds of classification were used as input to re-pick 1,634,615 particles. Following 3 rounds of 2D classification, 191,609 particles remained and were used to generate 3 ab-initio models. At this stage, heterogeneity between bound and unbound DPP9 particles was evident, and a round of heterogeneous refinement (2 classes) yielded a reconstruction from 92,404 particles at a nominal resolution of 4.88 Å. These particles were used for a final round of non-uniform refinement ( Punjani et al., 2019 ) which converged at an overall resolution of 3.86 Å as calculated by gold-standard FSC ( Figure S3 ). Atomic Model Building The cryo-EM maps were first fit with the crystal structure of DPP9 dimer (PDB ID: 6EOQ) ( Ross et al., 2018 ). A homology model of CARD8-FIIND was generated with Schrodinger Prime ( Jacobson et al., 2002 ) using the structure of NLRP1-FIIND ( Qin et al., 2020 ; Robert Hollingsworth et al., 2021 ) as template. Manual adjustment and de novo building of missing segments, rigid-body fitting, flexible fitting, and segment-based real-space refinement were performed in distinct parts of the initial model to fit in the density in Coot ( Emsley et al., 2010 ), with help of UCSF-Chimera ( Goddard et al., 2007 ) and real-space refinement in Phenix ( Klaholz, 2019 ). A few unstructured regions, including parts of the UPA, were omitted owing to poor density. The full model represents Asp18-Met1356 amino acids of DPP9 (short isoform, Uniprot ID Q86TI2-1), ZU5 A Gly166-Ser295, UPA A Arg304-Pro446 and UPA B Leu320-Pro446 of CARD8 (isoform 5, Uniprot ID Q9Y2G2-5). For the VbP-bound structure, the full model represents Asp18-Leu863 amino acids of DPP9 (short isoform, Uniprot ID Q86TI2-1), ZU5 A Gly166-Ser295, UPA A Arg304-Pro445 and UPA B Leu320-Pro446 of CARD8 (isoform 5, Uniprot ID Q9Y2G2-5). We modelled covalently linked DPP9 S730-VbP in the cryo-EM map density using NLRP1-DPP9-VbP (PDB ID: 6X6C) and DPP8-VbP (PDB ID: 6HP8) ( Díaz, 2018 ) as templates. DPP9-VbP interactions are similar to those observed in the DPP8-VbP complex in addition to DPPIV bound to substrate (PDB ID: 5YP3) ( Roppongi et al., 2018 ) ( Figure S4G ). For both structures, interaction analysis was conducted visually and using PISA ( Krissinel and Henrick, 2007 ). Structure representations were generated in ChimeraX ( Goddard et al., 2018 ), Pymol, and ResMap ( Kucukelbir et al., 2014 ). Ligand interaction analysis was conducted with Maestro ( Schrödinger Release 2020-1, 2020 ). Pymol and ChimeraX session files are available on our Open Science Framework repository ( https://osf.io/x7dv8/ ). Schematics were created with BioRender.

Negative Stain Electron Microscopy

Copper grids coated with layers of plastic and thin carbon film (Electron Microscopy Sciences) were glow discharged before 4 μl of purified proteins were applied. Samples were left on the grids for 45 sec, blotted, and then stained with 1% uranyl formate for 40 s, blotted, and air dried. The grids were imaged on a JEOL 1200EX or Tecnai G 2 Spirit BioTWIN microscope at the HMS EM facility operating at 80 keV. Immunoblotting. Samples were run on either NuPAGE ™ 4 to 12%, Bis-Tris 1.0 mm, Mini Protein Gel (Invitrogen) for 30 min at 175 V or NuPAGE ™ 4 to 12%, Bis-Tris, 1.0 mm, Midi Protein Gel (Invitrogen) for 45–60 min at 175 V. Gels were transferred to nitrocellulose with the Trans-Blot Turbo Transfer System (BIO-RAD). Membranes were blocked with Intercept ™ (TBS) Blocking Buffer (LI-COR) for 30 min at ambient room temperature, prior to incubating with primary antibody (1:1,000 in Intercept ™ Blocking Buffer) overnight at 4 °C. Blots were washed 3 times with TBST buffer prior to incubating with secondary antibody (1:10,000 in Intercept ™ Blocking Buffer) for 60 min at ambient room temperature. Blots were washed 3 times with TBST buffer, rinsed with water and imaged via Odyssey CLx (LI-COR). Immunoprecipitation Assays For DPP9 mutants ( Figure 3B ), DPP8 −/− , DPP9 −/− HEK 293T cells ( Hollingsworth et al., 2020 ) were seeded at 1 × 10 6 cells/well in 6-well tissue culture dishes. The following day cells were transfected with plasmids encoding for CARD8 (2 μg) or the indicated FLAG-tagged DPP9 construct (2 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. DPP9 and CARD8 lysates were mixed in a 1:1 ratio prior to treating with DMSO or VbP (10 μM) for 1 h. They were incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS in microcentrifuge spin columns (Pierce), bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), CARD8 C terminus rabbit polyclonal Ab (Abcam, ab24186), and GAPDH rabbit polyclonal Ab (CST, 14C10). For CARD8 mutants ( Figure 3C ), HEK 293T cells were seeded at 0.5 × 10 6 cells/well in 6-well tissue culture dishes. The following day cells were transfected with plasmids encoding for the indicated FLAG-tagged CARD8 construct (2 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. The clarified lysates were normalized by DC protein assay (BIO-RAD) and incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS, bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), and GAPDH rabbit polyclonal Ab (CST, 14C10). For ternary complex capture experiments ( Figure 4A ), HEK 293T cells were seeded at 5 × 10 5 cells/well in 6-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for FLAG-tagged CARD8-CT (1 μg), the indicated MYC-tagged CARD8-FIIND-SA (1 μg), and RFP (to 2 μg) with FuGENE HD according to manufacturer’s instructions (Promega). After 48 h cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. Lysates were incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS in microcentrifuge spin columns (Pierce), bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), Myc-tag (71D10) rabbit monoclonal Ab (CST, #2278), and GAPDH rabbit polyclonal Ab (CST, 14C10). For on-bead displacement experiments ( Figure 5A – C ), HEK 293T cells were seeded at 3 × 10 6 cells in a 10 cm tissue culture dish. The following day the cells were transfected with plasmids encoding for FLAG-tagged DPP9 (2 μg), V5-tagged Ub-S297 CARD8 or V5-tagged Ub-S1213 NLRP1 (3 μg), and MYC-tagged CARD8 S297A or MYC-tagged NLRP1 S1213A (5 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. Lysates were incubated with 100 μL of anti-FLAG-M2 agarose resin (Sigma) for 2 h at 4 °C. The agarose was washed once with PBS, and subsequently split into 4 × 25 μL aliquots. 50 μL of PBS containing DMSO, VbP (10 μM), 8J (50 μM), Bestatin methyl ester (MeBS, 10 μM) at RT for 1 h in microcentrifuge spin columns (Pierce). Displaced proteins were collected via centrifugation. The resin was washed 3x with cold PBS, followed by elution with 3x-FLAG peptide as above. Protein content was evaluated by immunoblotting with the following antibodies: FLAG® M2 monoclonal Ab (Sigma, F3165), Myc-tag (71D10) rabbit monoclonal Ab (CST, #2278), and V5 rabbit polyclonal Ab (Abcam, ab9116). For FP-biotin displacement experiments ( Figure 5D ), HEK 293T cells were seeded at 3 × 10 6 cells in a 10 cm tissue culture dish. The following day the cells were transfected with plasmids encoding for MYC-tagged CARD8-S297A-FL (10 μg) or FLAG-tagged Ub-S297 CARD8 (10 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication, mixed and centrifuged at 20,000 × g for 10 min at 4 °C. Lysates were split into 3 aliquots and incubated with DMSO, VbP (10 μM) or FP PEG23 biotin (10 μM) at room ambient temperature for 1 h. Lysates were incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS in microcentrifuge spin columns (Pierce), bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), Myc-tag (71D10) rabbit monoclonal Ab (CST, #2278), IRDye 600 RD streptavidin (LI-COR, 926-68079) and GAPDH rabbit polyclonal Ab (CST, 14C10). For dTAG ternary complex experiments ( Figure 5H ), HEK 293T cells were seeded at 5 × 10 5 cells/well in 6-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for FLAG-tagged CARD8-FIIND-S297A (1 μg), and dTAG-CARD8 (1 μg). The next day cells were treated with dTAG-13 (500 nM) and VbP (10 μM) for 24 h prior to harvesting. Anti-FLAG immunoprecipitation was executed as described above. Protein content was evaluated by immunoblotting with the following antibodies: FLAG® M2 monoclonal Ab (Sigma, F3165), HA-Tag (C29F4) rabbit monoclonal (CST, #3724), DPP9 rabbit polyclonal Ab (Abcam, Ab42080), and GAPDH rabbit polyclonal Ab (CST, 14C10). LDH Cytotoxicity Assay For THP-1 CARD8 −/− cells stably expressing the indicated pInducer20 CARD8 construct ( Figure 3D ), cells were seeded at 2.5 × 10 5 cells/mL in 12-well tissue culture dishes. The cells were treated with the indicated combination of doxycycline (1 μg/mL) and VbP (10 μg/mL) for 24 h. For HEK 293T cells stably expressing CASP-1 and GSDMD-V5 ( Figure 3E ), cells were seeded at 1.5 × 10 5 cells/mL in 12-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for the indicated CARD8-CT construct (0.1 μg) and RFP as a filler vector (to 2 μg) per 125 μL of Opti-MEM with FuGENE HD according to manufacturer’s instructions (Promega). In all cases, supernatants were analyzed for LDH activity using the CyQUANT ™ LDH Cytotoxicity Assay Kit (ThermoFisher) and lysates protein content was evaluated by immunoblotting with a combination of the following antibodies: CARD8 CT rabbit polyclonal Ab (Abcam, Ab24186), CASP1 rabbit polyclonal Ab (CST, #2225), GSDMDC1 rabbit polyclonal Ab (Novus, NBP2-33422), HA-Tag (C29F4) rabbit monoclonal (CST, #3724) and GAPDH rabbit polyclonal Ab (CST, 14C10). dTAG Assays HEK 293T cells stably expressing CASP-1 and GSDMD-V5 ( Figure 4C – D , F – H and Figure 5D – E ) were seeded at 1.5 × 10 5 cells/mL in 12-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for the indicated amount of plasmids encoding for dTAG-CARD8 or dTAG-CARD8 ZUC and CARD8 FIIND-S297A or NLRP1 FIIND-S1213A constructs with RFP as a filler vector (to 2 μg) per 125 μL of Opti-MEM with FuGENE HD according to manufacturer’s instructions (Promega). After 24 h (dTAG-CARD8) or 48 h (dTAG-CARD8 ZUC) cells were treated with DMSO, dTAG-13 (500 nM, unless otherwise indicated), and/or VbP (10 μM) for 3h prior to harvesting for analysis of supernatants for LDH release and lysates for protein content as detailed above (see Immunoblotting ). THP-1 CARD8 −/− cells stably expressing dTAG-CARD8 ZUC ( Figure 5F ) were seeded at 2 × 10 5 cells/mL in 12-well tissue culture dishes. Cells were treated with doxycycline (1 μg/mL) for 48h followed by dTAG-13 (5 nM) and VbP (10 μM) for 3h prior to harvesting for analysis of supernatants for LDH release as detailed above (see LDH Cytotoxicity Assay ).

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical significance was calculated by using GraphPad Prism version 7, GraphPad Software, San Diego, California USA, www.graphpad.com . The number of independent experiments, the statistical significance, and the statistical test used to determine the significance are indicated in each figure or figure legend or method section where quantification is reported.

DATA AVAILABILITY

The cryo-EM maps have been deposited in the Electron Microscopy Data Bank under the accession numbers EMD-22367 (CARD8-DPP9), EMD-22402 (CARD8-DPP9-VbP), and EMD-22974 (CARD8-S297A-DPP9). The atomic coordinates have been deposited in the Protein Data Bank under the accession numbers PDB ID 7JKQ (CARD8-DPP9) and PDB ID 7JN7 (CARD8-DPP9-VbP). Raw cryo-EM data have been deposited into EMPIAR under the accession numbers EMPIAR-10597 (CARD8-DPP9), EMPIAR-10596 (CARD8-DPP9-VbP), and EMPIAR-10600 (CARD8-S297A-DPP9). Extended protein purification protocols will be made available on Protocols.io ( https://www.protocols.io/groups/hao-wu-lab ). Pymol session files and full-sized figures will be made available on OSF ( https://osf.io/x7dv8/ ). Most constructs will be made available on Addgene. All other data and materials can be obtained from the corresponding authors upon reasonable request.

EXPERIMENTAL MODEL AND SUBJECT DETAILS HEK 293T Cell Lines

The human kidney epithelial cell line HEK 293T (ATCC) and related CASP1/GSDMD-expressing stable cell lines and DPP8 −/− , DPP9 −/− lines were maintained in DMEM (GIBCO, ThermoFisher) supplemented with 10% fetal bovine serum (GIBCO, ThermoFisher Scientific), at 37°C, and 5% CO 2 . HEK 293T cells were verified by the manufacturer. Cells were frequently checked for morphological features and tested for mycoplasma using MycoAlert Mycoplasma Detection kit (Lonza). Expi293F Cell Line Expi293F suspension cells were maintained in Expi293F Expression Medium (GIBCO, ThermoFisher) with constant shaking at 100 RPM, 37°C, 5% CO 2 . Expi293F cells were not authenticated nor tested for mycoplasma contamination. Sf9 Cell Line Sf9 insect cells were maintained in HyClone SFX-Insect Cell Media (Cytiva) supplemented with 1X antibiotic-antimycotic (ThermoFisher) at 27°C with constant shaking at 100 RPM. Sf9 cells were recently purchased from the manufacturer and were not authenticated.

THP-1 Cell Lines

THP-1 cells

(ATCC) and related THP-1 CARD8 −/− stable cell lines were grown in Roswell Park Memorial Institute (RPMI) medium 1640 with L-glutamine and 10% fetal bovine serum (GIBCO, ThermoFisher Scientific), at 37°C and 5% CO 2 . Cells were monitored for morphological features and regularly tested for mycoplasma using MycoAlert Mycoplasma Detection kit (Lonza).

METHODS DETAILS Constructs and Cloning Full-length CARD8 (T60 isoform, Uniprot ID Q9Y2G2-5) was cloned into pcDNA3.1 LIC 6A (Addgene plasmid #30124) with a C-terminal FLAG tag and a modified pcDNA3.1 LIC 6D (Addgene plasmid #30127) construct (C-terminal TEV-GFP-FLAG tag). The short isoform of DPP9 (DPP9S, Uniprot ID Q86TI2-1) was also cloned into pcDNA3.1 LIC 6A (N-terminal FLAG-TEV tag or N-terminal His-TEV tag). CARD8 (with several synonymous mutations to avoid CRISPR/Cas9 editing) was also shuttled into pInducer20 vector (Addgene, #44012) using Gateway technology (Thermo Fischer Scientific) and pLEX_305-N-dTAG (Addgene, #91797). CARD8-ZUC containing the ZU5, UPA and CARD was cloned as previously described ( Chui et al., 2020 ) and shuttled to pLEX_305-N-dTAG (Addgene, #91797). CARD8-CT constructs were synthesized (GenScript) with an N-terminal ubiquitin sequence followed by CARD8 (S297-L537), cloned into the pcDNA3.1 vector (Ub-CARD8-CT) and shuttled into the pLEX307 vector using Gateway technology (Thermo Fischer Scientific). Point mutations were introduced with Q5 site-directed mutagenesis (NEB) or QuikChange site-directed mutagenesis (Agilent). Most constructs will be made available on Addgene. Cell Culture HEK 293T cells and THP-1 cells were purchased from ATCC. HEK 293T cells were grown in Dulbecco’s Modified Eagle’s Medium (DMEM) with L-glutamine and 10% fetal bovine serum (FBS). THP-1 cells were grown in Roswell Park Memorial Institute (RPMI) medium 1640 with L-glutamine and 10% FBS. All cells were grown at 37° C in a 5% CO 2 atmosphere incubator. Cell lines were regularly tested for mycoplasma using the MycoAlert Mycoplasma Detection Kit (Lonza). THP-1 CARD8 −/− cells and HEK 293T cells stably expressing caspase-1 and GSDMD were generated as previously described ( Johnson et al., 2020 ). Reconstituted THP-1 CARD8 −/− cells were generated via lentiviral infection of THP-1 CARD8 −/− cells with the indicated pInducer20 construct followed by selection with G418 (Geneticin) at 200 μg/mL until all control cells were dead (approximately 14 days). dTAG-ZUC THP-1 cells were generated via lentiviral infection followed by selection with puromycin 500 ng/mL until all control cells were dead (approximately 7 days).

Protein Expression and Purification

To express CARD8-DPP9 complexes, Expi293F cells (1 L, 2–3 × 10 6 cells/mL) were co-transfected with CARD8-TEV-GFP-FLAG (0.7 mg) and DPP9S (0.3 mg) following incubation with polyethylenimine (3 mL, 1 mg/mL) in Opti-MEM (100 mL) for 30 min. 24 h later, cells were supplemented with glucose (9 mL, 45%) and valproic acid (10 mL, 300 mM). Cells were harvested 5 d after transfection by centrifugation (2,000 RPM, 20 min), washed once with PBS, split into 3 pellets, flash-frozen in liquid nitrogen, and stored at −80 °C. Later, one thawed pellet was resuspended in lysis buffer (50–100 mL, 25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM TCEP), sonicated (2 s on 8 s off, 3.5 min total on, 40% power, Branson), and ultracentrifuged at 40,000 RPM for 1 h (45 Ti fixed-angle rotor, Beckman). The supernatant was incubated with pre-equilibrated (lysis buffer) anti-FLAG M2 affinity gel (Sigma, 1.0 mL) for 4 h at 4 °C, washed in batch once with lysis buffer (5 mL), and then washed by gravity flow with 25–50 column volumes (CV) lysis buffer. The CARD8-DPP9 complex was eluted by on-column cleavage at room temperature for 1 h using elution buffer (5 mL, 25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM TCEP, 0.2 mg TEV protease) and loaded onto a Mono Q 5/50 GL anion exchange column (Cytiva). Protein was eluted using a salt gradient from 150 mM to 1 M NaCl (25 mM Tris-HCl pH 8.0, 1 mM TCEP) over 15 CV. Mono Q eluent was concentrated using a 0.5 mL spin concentrator (Amicon Ultra, 50 kDa MW cutoff) to 0.5 mg/mL (assuming ε=1). Concentrated eluent was dialyzed overnight into EM buffer (25 mM HEPES pH 7.5, 150 mM NaCl, 1 mM TCEP) using a 0.5 mL Slide-A-Lyzer (ThermoFisher). Total protein yield varied between 2–3 mg per L of mammalian culture.

Expression and purification of CARD8

(S297A) followed an identical protocol. Expression and purification of the VbP-bound complex followed an identical protocol except for the addition of 10 μM VbP to all purification and dialysis buffers.

Cryo-EM Screening and Data Collection

Grids were screened at University of Massachusetts Worcester, Pacific Northwest Center for Cryo-EM (PNCC), and Harvard Medical school (HMS) using a Talos Arctica microscope (ThermoFisher) operating at an acceleration voltage of 200 keV equipped with a direct electron detector. Small initial dataset collection revealed severe particle orientation preference for the CARD8-DPP9 maps, causing anisotropic resolution, in addition to dissociation of CARD8-DPP9 complexes. To remedy these issues, we crosslinked samples to avoid dissociation and tilted the stage to collect missing views (below). The purified DPP9S-CARD8-WT or DPP9S-CARD8-S297A complex (0.40 mg/mL assuming ε=1; 25 mM HEPES pH 7.5, 150 mM NaCl, 1 mM TCEP, ±10 μM VbP) was crosslinked with 0.02% glutaraldehyde on ice for 5 min and immediately loaded onto a glow-discharged Quantifoil grid (R1.2/1.3 400-mesh gold-supported holey carbon, Electron Microscopy Sciences), blotted for 3–5 s under 100% humidity at 4 °C, and plunged into liquid ethane using a Mark IV Vitrobot (ThermoFisher). Grids were screened at Harvard Medical School for ice and particle quality prior to data collection. For data collection, movies were acquired at Harvard Medical School (DPP9-CARD8-WT) and National Center for CryoEM Access and Training (NCCAT) in New York (DPP9-CARD8-WT+VbP and DPP9-CARD8-S297A) using a Titan Krios microscope (ThermoFisher) at an acceleration voltage of 300 keV equipped with a BioQuantum K3 Imaging Filter (slit width 20 eV). Movies were recorded with a K3 Summit direct electron detector (Gatan) operating in counting mode at 105,000 × (0.825 Å/pix at HMS, 0.82 or 0.83 at Å/pix NCCAT). For CARD8-DPP9 at 0° and 37°: 3,306 and 2,488 movies at a stage tilt of either 0° or 37° were collected using SerialEM ( Mastronarde, 2005 ) at varying defocus values ranging between −0.8 to −2.2 μm and −1.5 to −3.0 μm, respectively. We used image shift to record two shots for each of the four holes per stage movement. Movies were exposed with a total dose of 58.5 e − /Å 2 for 2.22 s fractionated over 49 frames for 0° stage tilt movies and 64.99 e − /Å 2 for 2.25 s fractionated over 49 frames for 37° stage tilt movies. For CARD8-DPP9-VbP at 0° and 37°: 1,811 and 6,642 movies at a stage tilt of either 0° or 37°, respectively, were collected using Leginon ( Suloway et al., 2005 ) at varying defocus values ranging between −1.1 to −3.4 μm. One shot was recorded for each of the four holes per stage movement through image shift. All movies were exposed with a total dose of 67.06 e − /Å 2 for 1.5 s fractionated over 50 frames. For CARD8(S297A)-DPP9, 3,840 movies were collected at a stage tilt of 37° using Leginon ( Suloway et al., 2005 ) to vary the defocus range between −0.8 to −2.5 μm and to record one shot for each of the four holes per stage movement through image shift. All movies were exposed with a total dose of 63.67 e − /Å 2 for 1.5 s fractionated over 50 frames.

Cryo-EM Data Processing

Data processing leveraged SBgrid Consortium ( Morin et al., 2013 ) for support and computing resources. Movies collected at Harvard Medical School (CARD8-DPP9) were pre-processed on-the-fly by the facility’s pipeline script. Movies were corrected by gain reference and for beam-induced motion and summed into motion-corrected images using the Relion 3.08 implementation of the MotionCor2 algorithm ( Zheng et al., 2017 ). The CTFFIND4 program ( Rohou and Grigorieff, 2015 ) was used to determine the defocus of each micrograph. Relion 3.1 ( Scheres, 2012 ; Zivanov et al., 2018 ) was used for subsequent image processing. Movies collected at NCCAT (CARD8-DPP9-VbP) were similarly pre-processed using Relion 3.1. For the CARD8-DPP9 complex, template-free autopicking with crYOLO ( Wagner et al., 2019 ) (generalized training for on-the-fly picking at the HMS cryo-EM center) selected 487,952 particles from 3,306 micrographs, which were subjected to a single round of 2D classification. The heterogenous nature of the sample was evident in 2D class averages because complexes containing only DPP9, 2CARD8:2DPP9, and 4CARD8:2DPP9 were present. A randomized set of 100,000 particles was selected for the de novo reconstruction of an initial model, which was low-pass-filtered to 40 Å to use as the input reference for 3D classification. Multiple rounds of 3D classification including global and local fine angular search were performed. After visual inspection, one class with 2CARD8:2DPP9 (62,018 particles) was selected for 3D refinement. Subsequently, particles were CTF refined and Bayesian polished to reach an overall resolution of 3.3 Å. However, the map suffered from anisotropic resolution, and cryoEF ( Naydenova and Russo, 2017 ) analysis estimated that data collected at a tilt angle of 37° was ideal to fill the gaps in Fourier space. The overall processing scheme for tilt data was derived from the work flow of tilt dataset described in Zivanov et al. 2018 ( Zivanov et al., 2018 ). For CARD8-DPP9 data collected at a 37° tilt, micrographs were first motion-corrected with MotionCor2 ( Zheng et al., 2017 ) followed by Gctf ( Zhang, 2016 ) to calculate per micrograph defocus values in Relion 3.1. Template-free autopicking with crYOLO ( Wagner et al., 2019 ) then picked 313,425 particles from 2,488 micrographs. After multiple rounds of 2D classification, 243,459 visually homogenous particles remained for further processing. A random subset comprising 121,730 of these particles was used for de novo initial model construction, which was then low-passed filtered (30 Å) and used as the initial reference map for 3D classification. After one round of 3D classification, a 3D class with 58,980 particles was selected for high-resolution refinement. The first 3D refinement using these particles yielded a 5.2 Å resolution structure and the Fourier shell correlation (FSC) curve showed strong fluctuations indicating imprecisions in CTF estimation. We then utilized CTF refinement implemented in Relion 3.1, including higher order aberrations correction, anisotropic magnification corrections, and per-particle defocus estimation ( Zivanov et al., 2020 ). Iterative rounds of 3D refinement followed by CTF refinement and Bayesian polishing gradually improved the resolution and this iterative process was stopped when we observed a resolution plateau at 3.8 Å. Refined and polished particle sets from 0° and 37° stage-tilt data were merged and 3D refinement was performed. The merged data was then reconstructed to give an overall resolution of 3.3 Å as calculated by gold-standard FSC between half maps, with much improved CARD8 density ( Figure S1 ). For the CARD8-DPP9-VbP complex, template-free autopicking with crYOLO ( Wagner et al., 2019 ) (trained with 10 manually-picked high-contrast micrographs) yielded 1,404,573 particles from 1,811 micrographs that were collected without stage tilt. These particles were subjected to multiple rounds of 2D classification that yielded visually homogeneous 2D classes with 471,255 particles. A randomized set of 100,000 particles was selected for the de novo reconstruction of an initial model, which was low-pass-filtered to 30 Å to use as the input reference for 3D classification. Multiple rounds of 3D classification including global and local fine angular search were performed. After visual inspection, one 3D class (89,909 particles) was selected for 3D refinement. Subsequently, particles were CTF refined and Bayesian polished to reach an overall resolution of 3.5 Å. However, the map suffered from anisotropic resolution, and cryoEF ( Naydenova and Russo, 2017 ) analysis estimated that data collected at a tilt angle of 37° degrees was ideal to fill the gaps in Fourier space (below). The overall processing scheme for tilt data analysis of CARD8-DPP9-VbP was derived from CARD8-DPP9 data processing (described above). Briefly, micrographs were first motion-corrected with MotionCor2 ( Zheng et al., 2017 ) followed by Gctf ( Zhang, 2016 ) to calculate per micrograph defocus values in Relion 3.1. Template-free autopicking with crYOLO ( Wagner et al., 2019 ) then picked 1,889,993 particles from 6642 micrographs. After multiple rounds of 2D classification, 840,902 visually homogenous particles remained for further processing. A random subset comprising 100,000 of these particles was used for de novo initial model construction, which was then low-passed filtered (30 Å) and used as the initial reference map for 3D classification. After multiple rounds of 3D classification, a 3D class with 91,254 particles was selected for high-resolution refinement. The first 3D refinement using these particles yielded a 5.8 Å resolution structure, and the FSC curve showed strong fluctuations indicating imprecisions in CTF estimation. We then utilized CTF refinement implemented in Relion 3.1, including correction for higher order aberrations and anisotropic magnification, and per-particle defocus estimation ( Zivanov et al., 2020 ). Iterative rounds of 3D refinement followed by CTF refinement and Bayesian polishing gradually improved the resolution and this iterative process was stopped when we observed a resolution plateau at 3.4 Å. Refined particle sets from 0° and 37° stage-tilt data were merged and 3D refinement was performed. The merged data was then 3D classified with fine local search angles to give a final stack of 146,101 particles which was reconstructed to give an overall resolution of 3.3 Å as calculated by gold-standard FSC between half maps, with much improved CARD8 density ( Figure S4 ). The DPP9-CARD8-S297A complex was processed in cryoSPARC ( Punjani et al., 2017 ). 3,840 movies collected at 37° stage tilt were summed into motion-corrected micrographs with Patch-Motion in cryoSPARC ( Punjani et al., 2017 ). Next, defocus values were determined with cryoSPARC’s Patch-CTF function ( Punjani et al., 2017 ). 5,505,288 particles were picked by blob picking and the 2D classes following two rounds of classification were used as input to re-pick 1,634,615 particles. Following 3 rounds of 2D classification, 191,609 particles remained and were used to generate 3 ab-initio models. At this stage, heterogeneity between bound and unbound DPP9 particles was evident, and a round of heterogeneous refinement (2 classes) yielded a reconstruction from 92,404 particles at a nominal resolution of 4.88 Å. These particles were used for a final round of non-uniform refinement ( Punjani et al., 2019 ) which converged at an overall resolution of 3.86 Å as calculated by gold-standard FSC ( Figure S3 ). Atomic Model Building The cryo-EM maps were first fit with the crystal structure of DPP9 dimer (PDB ID: 6EOQ) ( Ross et al., 2018 ). A homology model of CARD8-FIIND was generated with Schrodinger Prime ( Jacobson et al., 2002 ) using the structure of NLRP1-FIIND ( Qin et al., 2020 ; Robert Hollingsworth et al., 2021 ) as template. Manual adjustment and de novo building of missing segments, rigid-body fitting, flexible fitting, and segment-based real-space refinement were performed in distinct parts of the initial model to fit in the density in Coot ( Emsley et al., 2010 ), with help of UCSF-Chimera ( Goddard et al., 2007 ) and real-space refinement in Phenix ( Klaholz, 2019 ). A few unstructured regions, including parts of the UPA, were omitted owing to poor density. The full model represents Asp18-Met1356 amino acids of DPP9 (short isoform, Uniprot ID Q86TI2-1), ZU5 A Gly166-Ser295, UPA A Arg304-Pro446 and UPA B Leu320-Pro446 of CARD8 (isoform 5, Uniprot ID Q9Y2G2-5). For the VbP-bound structure, the full model represents Asp18-Leu863 amino acids of DPP9 (short isoform, Uniprot ID Q86TI2-1), ZU5 A Gly166-Ser295, UPA A Arg304-Pro445 and UPA B Leu320-Pro446 of CARD8 (isoform 5, Uniprot ID Q9Y2G2-5). We modelled covalently linked DPP9 S730-VbP in the cryo-EM map density using NLRP1-DPP9-VbP (PDB ID: 6X6C) and DPP8-VbP (PDB ID: 6HP8) ( Díaz, 2018 ) as templates. DPP9-VbP interactions are similar to those observed in the DPP8-VbP complex in addition to DPPIV bound to substrate (PDB ID: 5YP3) ( Roppongi et al., 2018 ) ( Figure S4G ). For both structures, interaction analysis was conducted visually and using PISA ( Krissinel and Henrick, 2007 ). Structure representations were generated in ChimeraX ( Goddard et al., 2018 ), Pymol, and ResMap ( Kucukelbir et al., 2014 ). Ligand interaction analysis was conducted with Maestro ( Schrödinger Release 2020-1, 2020 ). Pymol and ChimeraX session files are available on our Open Science Framework repository ( https://osf.io/x7dv8/ ). Schematics were created with BioRender.

Negative Stain Electron Microscopy

Copper grids coated with layers of plastic and thin carbon film (Electron Microscopy Sciences) were glow discharged before 4 μl of purified proteins were applied. Samples were left on the grids for 45 sec, blotted, and then stained with 1% uranyl formate for 40 s, blotted, and air dried. The grids were imaged on a JEOL 1200EX or Tecnai G 2 Spirit BioTWIN microscope at the HMS EM facility operating at 80 keV. Immunoblotting. Samples were run on either NuPAGE ™ 4 to 12%, Bis-Tris 1.0 mm, Mini Protein Gel (Invitrogen) for 30 min at 175 V or NuPAGE ™ 4 to 12%, Bis-Tris, 1.0 mm, Midi Protein Gel (Invitrogen) for 45–60 min at 175 V. Gels were transferred to nitrocellulose with the Trans-Blot Turbo Transfer System (BIO-RAD). Membranes were blocked with Intercept ™ (TBS) Blocking Buffer (LI-COR) for 30 min at ambient room temperature, prior to incubating with primary antibody (1:1,000 in Intercept ™ Blocking Buffer) overnight at 4 °C. Blots were washed 3 times with TBST buffer prior to incubating with secondary antibody (1:10,000 in Intercept ™ Blocking Buffer) for 60 min at ambient room temperature. Blots were washed 3 times with TBST buffer, rinsed with water and imaged via Odyssey CLx (LI-COR). Immunoprecipitation Assays For DPP9 mutants ( Figure 3B ), DPP8 −/− , DPP9 −/− HEK 293T cells ( Hollingsworth et al., 2020 ) were seeded at 1 × 10 6 cells/well in 6-well tissue culture dishes. The following day cells were transfected with plasmids encoding for CARD8 (2 μg) or the indicated FLAG-tagged DPP9 construct (2 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. DPP9 and CARD8 lysates were mixed in a 1:1 ratio prior to treating with DMSO or VbP (10 μM) for 1 h. They were incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS in microcentrifuge spin columns (Pierce), bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), CARD8 C terminus rabbit polyclonal Ab (Abcam, ab24186), and GAPDH rabbit polyclonal Ab (CST, 14C10). For CARD8 mutants ( Figure 3C ), HEK 293T cells were seeded at 0.5 × 10 6 cells/well in 6-well tissue culture dishes. The following day cells were transfected with plasmids encoding for the indicated FLAG-tagged CARD8 construct (2 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. The clarified lysates were normalized by DC protein assay (BIO-RAD) and incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS, bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), and GAPDH rabbit polyclonal Ab (CST, 14C10). For ternary complex capture experiments ( Figure 4A ), HEK 293T cells were seeded at 5 × 10 5 cells/well in 6-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for FLAG-tagged CARD8-CT (1 μg), the indicated MYC-tagged CARD8-FIIND-SA (1 μg), and RFP (to 2 μg) with FuGENE HD according to manufacturer’s instructions (Promega). After 48 h cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. Lysates were incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS in microcentrifuge spin columns (Pierce), bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), Myc-tag (71D10) rabbit monoclonal Ab (CST, #2278), and GAPDH rabbit polyclonal Ab (CST, 14C10). For on-bead displacement experiments ( Figure 5A – C ), HEK 293T cells were seeded at 3 × 10 6 cells in a 10 cm tissue culture dish. The following day the cells were transfected with plasmids encoding for FLAG-tagged DPP9 (2 μg), V5-tagged Ub-S297 CARD8 or V5-tagged Ub-S1213 NLRP1 (3 μg), and MYC-tagged CARD8 S297A or MYC-tagged NLRP1 S1213A (5 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication and centrifuged at 20,000 × g for 10 min at 4 °C. Lysates were incubated with 100 μL of anti-FLAG-M2 agarose resin (Sigma) for 2 h at 4 °C. The agarose was washed once with PBS, and subsequently split into 4 × 25 μL aliquots. 50 μL of PBS containing DMSO, VbP (10 μM), 8J (50 μM), Bestatin methyl ester (MeBS, 10 μM) at RT for 1 h in microcentrifuge spin columns (Pierce). Displaced proteins were collected via centrifugation. The resin was washed 3x with cold PBS, followed by elution with 3x-FLAG peptide as above. Protein content was evaluated by immunoblotting with the following antibodies: FLAG® M2 monoclonal Ab (Sigma, F3165), Myc-tag (71D10) rabbit monoclonal Ab (CST, #2278), and V5 rabbit polyclonal Ab (Abcam, ab9116). For FP-biotin displacement experiments ( Figure 5D ), HEK 293T cells were seeded at 3 × 10 6 cells in a 10 cm tissue culture dish. The following day the cells were transfected with plasmids encoding for MYC-tagged CARD8-S297A-FL (10 μg) or FLAG-tagged Ub-S297 CARD8 (10 μg) with FuGENE HD according to manufacturer’s instructions (Promega). Cells were harvested and washed 3x with PBS. Pellets were lysed in Tris-Buffered Saline (TBS) with 0.5% NP-40 using pulse sonication, mixed and centrifuged at 20,000 × g for 10 min at 4 °C. Lysates were split into 3 aliquots and incubated with DMSO, VbP (10 μM) or FP PEG23 biotin (10 μM) at room ambient temperature for 1 h. Lysates were incubated with 20 μL of anti-FLAG-M2 agarose resin (Sigma) overnight at 4 °C. After washing 3 × 500 μL with cold PBS in microcentrifuge spin columns (Pierce), bound proteins were eluted by incubating resin with 40 μL of PBS with 150 ng/μL 3x-FLAG peptide for 1 h at 4 °C. An equal volume of 2x sample loading was added to the eluate and boiled. Protein content was evaluated by immunoblotting with the following antibodies: DPP9 rabbit polyclonal Ab (Abcam, Ab42080), FLAG® M2 monoclonal Ab (Sigma, F3165), Myc-tag (71D10) rabbit monoclonal Ab (CST, #2278), IRDye 600 RD streptavidin (LI-COR, 926-68079) and GAPDH rabbit polyclonal Ab (CST, 14C10). For dTAG ternary complex experiments ( Figure 5H ), HEK 293T cells were seeded at 5 × 10 5 cells/well in 6-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for FLAG-tagged CARD8-FIIND-S297A (1 μg), and dTAG-CARD8 (1 μg). The next day cells were treated with dTAG-13 (500 nM) and VbP (10 μM) for 24 h prior to harvesting. Anti-FLAG immunoprecipitation was executed as described above. Protein content was evaluated by immunoblotting with the following antibodies: FLAG® M2 monoclonal Ab (Sigma, F3165), HA-Tag (C29F4) rabbit monoclonal (CST, #3724), DPP9 rabbit polyclonal Ab (Abcam, Ab42080), and GAPDH rabbit polyclonal Ab (CST, 14C10). LDH Cytotoxicity Assay For THP-1 CARD8 −/− cells stably expressing the indicated pInducer20 CARD8 construct ( Figure 3D ), cells were seeded at 2.5 × 10 5 cells/mL in 12-well tissue culture dishes. The cells were treated with the indicated combination of doxycycline (1 μg/mL) and VbP (10 μg/mL) for 24 h. For HEK 293T cells stably expressing CASP-1 and GSDMD-V5 ( Figure 3E ), cells were seeded at 1.5 × 10 5 cells/mL in 12-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for the indicated CARD8-CT construct (0.1 μg) and RFP as a filler vector (to 2 μg) per 125 μL of Opti-MEM with FuGENE HD according to manufacturer’s instructions (Promega). In all cases, supernatants were analyzed for LDH activity using the CyQUANT ™ LDH Cytotoxicity Assay Kit (ThermoFisher) and lysates protein content was evaluated by immunoblotting with a combination of the following antibodies: CARD8 CT rabbit polyclonal Ab (Abcam, Ab24186), CASP1 rabbit polyclonal Ab (CST, #2225), GSDMDC1 rabbit polyclonal Ab (Novus, NBP2-33422), HA-Tag (C29F4) rabbit monoclonal (CST, #3724) and GAPDH rabbit polyclonal Ab (CST, 14C10). dTAG Assays HEK 293T cells stably expressing CASP-1 and GSDMD-V5 ( Figure 4C – D , F – H and Figure 5D – E ) were seeded at 1.5 × 10 5 cells/mL in 12-well tissue culture dishes. The following day the cells were transfected with plasmids encoding for the indicated amount of plasmids encoding for dTAG-CARD8 or dTAG-CARD8 ZUC and CARD8 FIIND-S297A or NLRP1 FIIND-S1213A constructs with RFP as a filler vector (to 2 μg) per 125 μL of Opti-MEM with FuGENE HD according to manufacturer’s instructions (Promega). After 24 h (dTAG-CARD8) or 48 h (dTAG-CARD8 ZUC) cells were treated with DMSO, dTAG-13 (500 nM, unless otherwise indicated), and/or VbP (10 μM) for 3h prior to harvesting for analysis of supernatants for LDH release and lysates for protein content as detailed above (see Immunoblotting ). THP-1 CARD8 −/− cells stably expressing dTAG-CARD8 ZUC ( Figure 5F ) were seeded at 2 × 10 5 cells/mL in 12-well tissue culture dishes. Cells were treated with doxycycline (1 μg/mL) for 48h followed by dTAG-13 (5 nM) and VbP (10 μM) for 3h prior to harvesting for analysis of supernatants for LDH release as detailed above (see LDH Cytotoxicity Assay ).

Supplementary Material 1