Human caspase-4 is an inflammatory cysteine protease of the caspase-1 subfamily, encoded in the CASP1/CASP4/CASP5 cluster, that acts as the cytosolic receptor and effector of the non-canonical inflammasome. Its N-terminal CARD binds the lipid A moiety of bacterial lipopolysaccharide that has reached the cytoplasm; ligand engagement converts the otherwise disordered CARD into an alpha-helical fold, drives CARD-mediated polymerisation, and dimerises and autoprocesses the p20/p10 protease domain at D270 and D289. The activated enzyme cleaves after aspartate, processing gasdermin D to liberate its pore-forming N-terminal fragment and thereby triggering pyroptosis, and directly maturing pro-interleukin-18 - an activity that human caspase-4 has but rodent caspase-11 lacks, so caspase-4 and mouse caspase-11 are not functionally interchangeable despite occupying the same position in their respective pathways. Activation on intracellular bacteria requires guanylate-binding proteins, chiefly GBP1, which coat and deform the LPS-rich outer membrane and expose lipid A. How caspase-4 engages its ligand is disputed: direct binding measurements show high-affinity association with free LPS and lipid A and identify a hydrophobic acyl-chain pocket in the CARD, whereas membrane work indicates that caspase-4 binds LPS surfaces of positive curvature rather than individual LPS molecules, and the resulting assembly is structurally heterogeneous rather than a complex of fixed stoichiometry. The substrate repertoire is also unsettled: alongside gasdermin D and pro-interleukin-18, caspase-4 has been reported to cleave pro-interleukin-1beta and to directly activate the executioner caspases CASP3 and CASP7, while in macrophages most gasdermin D cleavage during non-canonical signalling may instead be carried out by caspase-1. A separate, longer-standing arm of caspase-4 biology places the zymogen on the cytoplasmic face of the endoplasmic reticulum membrane via TMEM214, where it participates in apoptosis induced by ER stress and by amyloid-beta peptides.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005737 cytoplasm | IBA GO_REF:0000033 | ACCEPT | Summary: Pan-caspase phylogenetic assignment of cytoplasmic residence; CASP4 is a soluble cytoplasmic zymogen. Reason: Correct. CASP4 is synthesised as a cytosolic pro-enzyme and its LPS-sensing function is by definition a cytosolic one. The sibling cytosol annotation (GO:0005829) is the more informative of the two. |
| GO:0005829 cytosol | IBA GO_REF:0000033 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0070269 pyroptotic inflammatory response | IBA GO_REF:0000033 | ACCEPT | Summary: Pyroptosis is the defining outcome of CASP4 activation by cytosolic LPS. Reason: Core biological process, supported by reconstitution, by CRISPR screens identifying GSDMD as the executioner, and by pathogen-evasion genetics. The IBD node PTN002573059 is seeded by CASP4 itself plus mouse Casp4/11 and zebrafish caspa/caspb, so the phylogenetic assertion adds the claim that the role is inherited rather than human-specific. Supporting Evidence: PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:32109412 we show site-specific caspase-4/11 autoprocessing, generating a p10 product, is required and sufficient for cleaving GSDMD and inducing pyroptosis. PMID:34671164 S. flexneri evaded pyroptosis mediated by caspase-11 or caspase 4 (hereafter referred to as caspase-11/4) using a type III secretion system (T3SS) effector, OspC3. |
| GO:0006915 apoptotic process | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Generic apoptosis term inherited from the pan-caspase node and from the 1995 overexpression papers; CASP4 does have an apoptotic arm, but it is not what the gene is for. Reason: Retained but demoted. CASP4 overexpression induced apoptosis in the founding transfection studies, it participates in ER-stress apoptosis, and a 2026 study reports that CASP4/5 directly cleave and activate the executioner caspases CASP3 and CASP7 - which, if it holds, gives the apoptotic arm real mechanistic footing. That claim rests on a single study and is one of the points currently in dispute, so it is recorded here rather than promoted to a core function. Supporting Evidence: PMID:42044191 intracellular LPS and the gram-negative bacterial pathogen Salmonella activate CASP4/5 in macrophages to directly cleave and activate CASP3 and CASP7. PMID:42044191 CASP1, CASP4, CASP5 and to a lesser extent, CASP11, processed CASP3C163A and CASP7C186A but failed to process the D175A and D198A mutants PMID:7743998 Transfection experiments demonstrate that TX is a protease which is able to cleave itself and the p30 ICE precursor, but not to generate mature IL-1 beta from pro-IL-1 beta. |
| GO:0050729 positive regulation of inflammatory response | IBA GO_REF:0000033 | ACCEPT | Summary: CASP4 activation drives release of mature IL-18 and IL-1 family cytokines and downstream inflammation. Reason: Correct and central. CASP4 both matures pro-IL-18 directly and licenses canonical inflammasome output; the IBD node is seeded by CASP4's own experimental annotation together with mouse Casp4/11. Supporting Evidence: PMID:37993712 enables human cells to bypass the need for canonical inflammasomes and caspase-1 for IL-18 release. PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. |
| GO:0004197 cysteine-type endopeptidase activity | IBA GO_REF:0000033 | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0072558 NLRP1 inflammasome complex | IBA GO_REF:0000033 | MARK AS OVER ANNOTATED | Summary: Phylogenetic transfer of NLRP1 inflammasome membership from CASP5 and two zebrafish caspases; human CASP4 has not been placed in an NLRP1 inflammasome. Reason: The GO definition of this complex names only caspase-1 or caspase-5 as the protease component. Inspection of the PAINT file for PTHR47901 shows the IBD at PTN002573059 was seeded by UniProtKB:P51878 (human CASP5) plus zebrafish caspa and caspb, with no CASP4 evidence; the node spans the whole CASP1/4/5 clade, so the term reaches CASP4 mechanically. CASP4's experimentally defined complex is the non-canonical inflammasome (GO:0160074), which it is separately and correctly annotated to. Not removed: if CASP4 is shown to be recruited to NLRP1 inflammasomes the transfer would become sound. Propagation Review Root cause: PROPAGATION BAD Failure modes: WRONG ORTHOLOG OR PARALOG COMPARTMENT OR COMPLEX MISMATCH Sources checked: UniProtKB:P51878 Β· CASP5 SUPPORTS SOURCE BUT NOT TARGET Human caspase-5 is named in the GO term definition; its paralog CASP4 is not, and shares no NLRP1 evidence. ZFIN:ZDB-GENE-000616-3 Β· caspa SUPPORTS SOURCE BUT NOT TARGET Zebrafish inflammatory caspase; NLRP1 complex composition in teleosts does not establish human CASP4 membership. ZFIN:ZDB-GENE-020812-1 Β· caspb SUPPORTS SOURCE BUT NOT TARGET As for caspa. PANTHER:PTN002573059 Β· PTN002573059 SUPPORTS SOURCE BUT NOT TARGET Node spans the CASP1/CASP4/CASP5 clade, so a CASP5-specific complex assignment reaches CASP4 by descent alone. |
| GO:0043525 positive regulation of neuron apoptotic process | IBA GO_REF:0000033 | MARK AS OVER ANNOTATED | Summary: Neuronal apoptosis role propagated from the deep pan-caspase node, whose seeds are all apoptotic caspases, not inflammatory ones. Reason: The IBD sits at PTN000047947, the node covering the whole caspase family, and its seeds are rat Casp3 (RGD:2275), Casp8 (RGD:620945), Casp2 (RGD:69274), Casp6 (RGD:70967), mouse Casp9 (MGI:MGI:1277950) and C. elegans ced-3 - initiator and executioner apoptotic caspases assayed in neuronal death models. None is an inflammatory caspase, and CASP4's own biology is LPS sensing in myeloid and epithelial cells. Transferring a neuron-specific apoptotic regulatory role across that node to CASP4 over-reaches. Not removed, because CASP4 does have an apoptotic arm (ER stress, AΞ²) - but that arm is captured by GO:0097193 and GO:1904646, not by this term. Propagation Review Root cause: PROPAGATION BAD Failure modes: FUNCTIONAL DIVERGENCE CONTEXT OR TISSUE MISMATCH Sources checked: RGD:2275 Β· Casp3 SUPPORTS SOURCE BUT NOT TARGET Executioner caspase; neuronal apoptosis evidence does not transfer to an LPS-sensing inflammatory caspase. RGD:620945 Β· Casp8 SUPPORTS SOURCE BUT NOT TARGET Apoptotic initiator caspase with a death-receptor role CASP4 does not share. RGD:69274 Β· Casp2 SUPPORTS SOURCE BUT NOT TARGET RGD:70967 Β· Casp6 SUPPORTS SOURCE BUT NOT TARGET MGI:MGI:1277950 Β· Casp9 SUPPORTS SOURCE BUT NOT TARGET Apoptosome-activated initiator caspase. WB:WBGene00000417 Β· ced-3 SUPPORTS SOURCE BUT NOT TARGET Ancestral apoptotic caspase; seeds the deep node but says nothing about inflammatory-caspase subfunction. PANTHER:PTN000047947 Β· PTN000047947 SUPPORTS SOURCE BUT NOT TARGET Pan-caspase node; a term placed here reaches every caspase including the LPS receptors. |
| GO:0004197 cysteine-type endopeptidase activity | IEA GO_REF:0000120 | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0005576 extracellular region | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: CASP4 is released into the extracellular space by UVB-irradiated keratinocytes; a real but peripheral localisation. Reason: Grounded rather than spurious: UniProt records release into the extracellular milieu by keratinocytes after UVB, from the same study that supports the experimental annotation. It is a consequence of unconventional secretion / pyroptotic release, not a site of CASP4 activity, so it is kept out of the core set. Supporting Evidence: PMID:22246630 we show that caspase-4 expression is required for UVB-induced activation of proIL-1Ξ² and for unconventional protein secretion by skin-derived keratinocytes. |
| GO:0005739 mitochondrion | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Mitochondrial localisation reported alongside ER localisation in the ER-stress literature. Reason: Reported by two independent groups working on ER-stress apoptosis and carried by UniProt, so retained; but the mitochondrial pool has no assigned activity and is unrelated to the LPS-sensing function that defines the gene. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005789 endoplasmic reticulum membrane | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Procaspase-4 is recruited to the cytoplasmic face of the ER membrane by TMEM214. Reason: Well supported by two independent studies and by UniProt (peripheral membrane protein, cytoplasmic side). It belongs to the ER-stress arm of CASP4 biology rather than to the non-canonical inflammasome, hence non-core. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | IEA GO_REF:0000044 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0006508 proteolysis | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Generic proteolysis, true of any protease. Reason: Correct but uninformative next to the specific processes CASP4 drives (GO:0051604 protein maturation, GO:0016540 protein autoprocessing, GO:0160075 non-canonical inflammasome complex assembly). Retained as a true parent rather than modified, because no single child term subsumes the whole of CASP4's proteolytic activity. |
| GO:0008234 cysteine-type peptidase activity | IEA GO_REF:0000002 | MODIFY | Summary: InterPro-derived peptidase activity, one level too general. Reason: Correct but under-specific: CASP4 is an endopeptidase with strict P1-Asp specificity, and the specific term is already independently supported by EXP, IDA, IBA and TAS evidence on this gene. Proposed replacements: cysteine-type endopeptidase activity |
| GO:0042981 regulation of apoptotic process | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: Regulation of apoptosis inferred electronically from the CARD domain signature. Reason: Defensible but generic. CASP4 does modulate apoptosis - through ER stress and, per a 2026 report, through direct activation of CASP3/CASP7 - but the domain-based inference carries no direction and no context, so it stays out of the core set. |
| GO:0051604 protein maturation | IEA GO_REF:0000117 | ACCEPT | Summary: CASP4 proteolytically matures its substrates: GSDMD into its pore-forming N-terminal fragment, and pro-IL-18 into mature IL-18. Reason: Core biological process and the direct consequence of the catalytic activity. Two independent 2023 structural studies established direct pro-IL-18 maturation by human CASP4, and GSDMD processing is established by reconstitution and structure. Supporting Evidence: PMID:37993714 Here we show that activated human caspase-4, but not mouse caspase-11, directly and efficiently processes IL-18 in vitro and during bacterial infections. PMID:37993712 Here we demonstrate that the lipopolysaccharide receptor caspase-4 from humans and other mammalian species (except rodents) can cleave pro-IL-18 PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. |
| GO:0061702 canonical inflammasome complex | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: CASP4 is found in, and feeds into, canonical (NLRP3/NLRP6, caspase-1-activating) inflammasomes as well as its own non-canonical one. Reason: Not wrong - CASP4 physically interacts with caspase-1 and is required for NLRP3- and NLRP6-dependent IL-1Ξ² output - but the complex that defines CASP4 is the non-canonical inflammasome (GO:0160074), whose GO definition names CASP4 explicitly. Kept as a secondary, context-dependent complex membership. Supporting Evidence: PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. PMID:33377178 After LPS priming, cytosolic Streptococcus mutans LTA triggered NLRP6-caspase 4 inflammasome activation. |
| GO:0065003 protein-containing complex assembly | IEA GO_REF:0000117 | MODIFY | Summary: Generic complex-assembly term from an ARBA model. Reason: The assembly CASP4 actually carries out is that of the non-canonical inflammasome, a specific child of this term that the gene already carries under both TAS and IDA evidence. Proposed replacements: non-canonical inflammasome complex assembly |
| GO:0097193 intrinsic apoptotic signaling pathway | IEA GO_REF:0000117 | KEEP AS NON CORE | Summary: Intrinsic apoptotic signalling, driven by the ER-stress arm of CASP4 biology. Reason: Supported by knockdown experiments showing reduced ER-stress-induced apoptosis, and reproduced by an independent group via the TMEM214 anchor. Genuine, but a separate arm from the LPS-sensing function for which the gene is principally responsible. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | REMOVE | Summary: Bare protein-binding annotations from interaction screens and targeted co-IP. Reason: Per project guidance, bare protein binding conveys no functional information. The underlying interactions are real and in two cases mechanistically important - SERPINB1 blocks CARD oligomerisation, and cathepsin G binds and activates CASP4 - but each should be captured by an informative term (e.g. CARD domain binding) or as a regulatory relationship, not by GO:0005515. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. Supporting Evidence: PMID:30692621 Here we report that serpin family B member 1 (SERPINB1) limited the activity of those caspases by suppressing their caspase-recruitment domain (CARD) oligomerization and enzymatic activation. PMID:29077095 Td92 treatment resulted in the binding of cathepsin G to caspase-4 and the coaggregation of these two molecules. |
| GO:0004197 cysteine-type endopeptidase activity | EXP PMID:37558421 Caspase-4 dimerisation and D289 auto-processing elicit an in... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0004197 cysteine-type endopeptidase activity | TAS Reactome:R-HSA-9947861 | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0160075 non-canonical inflammasome complex assembly | TAS Reactome:R-HSA-9948001 | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9947861 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9947908 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9947987 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9948146 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9956624 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9958657 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9960526 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005576 extracellular region | EXP PMID:22246630 Caspase-4 is required for activation of inflammasomes. | KEEP AS NON CORE | Summary: CASP4 is released into the extracellular space by UVB-irradiated keratinocytes; a real but peripheral localisation. Reason: Grounded rather than spurious: UniProt records release into the extracellular milieu by keratinocytes after UVB, from the same study that supports the experimental annotation. It is a consequence of unconventional secretion / pyroptotic release, not a site of CASP4 activity, so it is kept out of the core set. Supporting Evidence: PMID:22246630 we show that caspase-4 expression is required for UVB-induced activation of proIL-1Ξ² and for unconventional protein secretion by skin-derived keratinocytes. |
| GO:0005739 mitochondrion | EXP PMID:23661706 Transmembrane Protein 214 (TMEM214) mediates endoplasmic ret... | KEEP AS NON CORE | Summary: Mitochondrial localisation reported alongside ER localisation in the ER-stress literature. Reason: Reported by two independent groups working on ER-stress apoptosis and carried by UniProt, so retained; but the mitochondrial pool has no assigned activity and is unrelated to the LPS-sensing function that defines the gene. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005789 endoplasmic reticulum membrane | EXP PMID:15123740 Involvement of caspase-4 in endoplasmic reticulum stress-ind... | KEEP AS NON CORE | Summary: Procaspase-4 is recruited to the cytoplasmic face of the ER membrane by TMEM214. Reason: Well supported by two independent studies and by UniProt (peripheral membrane protein, cytoplasmic side). It belongs to the ER-stress arm of CASP4 biology rather than to the non-canonical inflammasome, hence non-core. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005789 endoplasmic reticulum membrane | EXP PMID:23661706 Transmembrane Protein 214 (TMEM214) mediates endoplasmic ret... | KEEP AS NON CORE | Summary: Procaspase-4 is recruited to the cytoplasmic face of the ER membrane by TMEM214. Reason: Well supported by two independent studies and by UniProt (peripheral membrane protein, cytoplasmic side). It belongs to the ER-stress arm of CASP4 biology rather than to the non-canonical inflammasome, hence non-core. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0061702 canonical inflammasome complex | EXP PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | KEEP AS NON CORE | Summary: CASP4 is found in, and feeds into, canonical (NLRP3/NLRP6, caspase-1-activating) inflammasomes as well as its own non-canonical one. Reason: Not wrong - CASP4 physically interacts with caspase-1 and is required for NLRP3- and NLRP6-dependent IL-1Ξ² output - but the complex that defines CASP4 is the non-canonical inflammasome (GO:0160074), whose GO definition names CASP4 explicitly. Kept as a secondary, context-dependent complex membership. Supporting Evidence: PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. PMID:33377178 After LPS priming, cytosolic Streptococcus mutans LTA triggered NLRP6-caspase 4 inflammasome activation. |
| GO:0061702 canonical inflammasome complex | EXP PMID:26508369 Human caspase-4 and caspase-5 regulate the one-step non-cano... | KEEP AS NON CORE | Summary: CASP4 is found in, and feeds into, canonical (NLRP3/NLRP6, caspase-1-activating) inflammasomes as well as its own non-canonical one. Reason: Not wrong - CASP4 physically interacts with caspase-1 and is required for NLRP3- and NLRP6-dependent IL-1Ξ² output - but the complex that defines CASP4 is the non-canonical inflammasome (GO:0160074), whose GO definition names CASP4 explicitly. Kept as a secondary, context-dependent complex membership. Supporting Evidence: PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. PMID:33377178 After LPS priming, cytosolic Streptococcus mutans LTA triggered NLRP6-caspase 4 inflammasome activation. |
| GO:0042742 defense response to bacterium | IDA PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | ACCEPT | Summary: CASP4 restricts intracellular Gram-negative bacteria by sensing their LPS and triggering pyroptosis. Reason: Core biological process, supported across Salmonella, Shigella and Burkholderia models and by the existence of dedicated bacterial evasion mechanisms (OspC3 ADP-riboxanation of the caspase), which is strong evidence that the pathway matters in vivo. Supporting Evidence: PMID:34671164 Mouse caspase-11 and human caspase-4 and caspase-5 recognize cytosolic lipopolysaccharide (LPS) to induce pyroptosis by cleaving the pore-forming protein GSDMD PMID:31268602 GBP1 facilitated caspase-4 recruitment to Salmonella leading to its enhanced activation and pyroptosis. |
| GO:0051604 protein maturation | IDA PMID:34671164 Shigella evades pyroptosis by arginine ADP-riboxanation of c... | ACCEPT | Summary: CASP4 proteolytically matures its substrates: GSDMD into its pore-forming N-terminal fragment, and pro-IL-18 into mature IL-18. Reason: Core biological process and the direct consequence of the catalytic activity. Two independent 2023 structural studies established direct pro-IL-18 maturation by human CASP4, and GSDMD processing is established by reconstitution and structure. Supporting Evidence: PMID:37993714 Here we show that activated human caspase-4, but not mouse caspase-11, directly and efficiently processes IL-18 in vitro and during bacterial infections. PMID:37993712 Here we demonstrate that the lipopolysaccharide receptor caspase-4 from humans and other mammalian species (except rodents) can cleave pro-IL-18 PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:34671164 Shigella evades pyroptosis by arginine ADP-riboxanation of c... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0004197 cysteine-type endopeptidase activity | IDA PMID:26375003 Cleavage of GSDMD by inflammatory caspases determines pyropt... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0005829 cytosol | IDA PMID:23661706 Transmembrane Protein 214 (TMEM214) mediates endoplasmic ret... | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0051604 protein maturation | IDA PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | ACCEPT | Summary: CASP4 proteolytically matures its substrates: GSDMD into its pore-forming N-terminal fragment, and pro-IL-18 into mature IL-18. Reason: Core biological process and the direct consequence of the catalytic activity. Two independent 2023 structural studies established direct pro-IL-18 maturation by human CASP4, and GSDMD processing is established by reconstitution and structure. Supporting Evidence: PMID:37993714 Here we show that activated human caspase-4, but not mouse caspase-11, directly and efficiently processes IL-18 in vitro and during bacterial infections. PMID:37993712 Here we demonstrate that the lipopolysaccharide receptor caspase-4 from humans and other mammalian species (except rodents) can cleave pro-IL-18 PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. |
| GO:0008289 lipid binding | IDA PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | MODIFY | Summary: Generic lipid binding; all three IDA rows derive from LPS / lipid A binding experiments. Reason: The parent term is true but uninformative, and the specific child GO:0001530 is already annotated from two of the same sources. Flagged caveat: a 2025 report identifies a non-LPS lipid ligand (sphingomyelin C12) for caspase-4; if that is confirmed the generic parent would become independently informative rather than redundant, and this MODIFY should be revisited. Proposed replacements: lipopolysaccharide binding Supporting Evidence: PMID:25119034 Human caspase-4 and the mouse homologue caspase-11 (hereafter referred to as caspase-4/11) and also human caspase-5, directly bound to LPS and lipid A with high specificity and affinity. PMID:32510692 Binding of polymerizing hGBP1 to the bacterial surface disrupts the O-antigen barrier, thereby unmasking lipid A, eliciting caspase-4 recruitment |
| GO:0008289 lipid binding | IDA PMID:32510692 Direct binding of polymeric GBP1 to LPS disrupts bacterial c... | MODIFY | Summary: Generic lipid binding; all three IDA rows derive from LPS / lipid A binding experiments. Reason: The parent term is true but uninformative, and the specific child GO:0001530 is already annotated from two of the same sources. Flagged caveat: a 2025 report identifies a non-LPS lipid ligand (sphingomyelin C12) for caspase-4; if that is confirmed the generic parent would become independently informative rather than redundant, and this MODIFY should be revisited. Proposed replacements: lipopolysaccharide binding Supporting Evidence: PMID:25119034 Human caspase-4 and the mouse homologue caspase-11 (hereafter referred to as caspase-4/11) and also human caspase-5, directly bound to LPS and lipid A with high specificity and affinity. PMID:32510692 Binding of polymerizing hGBP1 to the bacterial surface disrupts the O-antigen barrier, thereby unmasking lipid A, eliciting caspase-4 recruitment |
| GO:0008289 lipid binding | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | MODIFY | Summary: Generic lipid binding; all three IDA rows derive from LPS / lipid A binding experiments. Reason: The parent term is true but uninformative, and the specific child GO:0001530 is already annotated from two of the same sources. Flagged caveat: a 2025 report identifies a non-LPS lipid ligand (sphingomyelin C12) for caspase-4; if that is confirmed the generic parent would become independently informative rather than redundant, and this MODIFY should be revisited. Proposed replacements: lipopolysaccharide binding Supporting Evidence: PMID:25119034 Human caspase-4 and the mouse homologue caspase-11 (hereafter referred to as caspase-4/11) and also human caspase-5, directly bound to LPS and lipid A with high specificity and affinity. PMID:32510692 Binding of polymerizing hGBP1 to the bacterial surface disrupts the O-antigen barrier, thereby unmasking lipid A, eliciting caspase-4 recruitment |
| GO:0042742 defense response to bacterium | IDA PMID:32510692 Direct binding of polymeric GBP1 to LPS disrupts bacterial c... | ACCEPT | Summary: CASP4 restricts intracellular Gram-negative bacteria by sensing their LPS and triggering pyroptosis. Reason: Core biological process, supported across Salmonella, Shigella and Burkholderia models and by the existence of dedicated bacterial evasion mechanisms (OspC3 ADP-riboxanation of the caspase), which is strong evidence that the pathway matters in vivo. Supporting Evidence: PMID:34671164 Mouse caspase-11 and human caspase-4 and caspase-5 recognize cytosolic lipopolysaccharide (LPS) to induce pyroptosis by cleaving the pore-forming protein GSDMD PMID:31268602 GBP1 facilitated caspase-4 recruitment to Salmonella leading to its enhanced activation and pyroptosis. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:32510692 Direct binding of polymeric GBP1 to LPS disrupts bacterial c... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0001530 lipopolysaccharide binding | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | ACCEPT | Summary: Direct, high-affinity binding of CASP4 to LPS and to its lipid A moiety - the receptor function of the non-canonical inflammasome. The abstraction is challenged by 2026 work showing the physiological ligand is a curved LPS membrane rather than a free LPS molecule. Reason: Retained as a core molecular function, with the dispute recorded. The IDAs rest on direct binding measurements (Shi et al. 2014) and on the 2023 structural work, and a 2026 HDX-MS study localises recognition to a hydrophobic pocket formed by helices Ξ±1, Ξ±2 and Ξ±5 of the CARD that engages the acyl chains of LPS - i.e. a genuine chemical interaction between the protein and the lipid. Against this, Began et al. 2026 report that caspase-4 binds LPS membranes with positive curvature and conclude it engages membranes with defined geometry rather than individual LPS molecules, with GBP1 required to generate that geometry on cytosolic bacteria. These are reconcilable at the level GO annotates: the same paper still calls lipid A the cognate ligand, and curvature governs presentation and avidity rather than the chemistry of recognition. GO has no term for binding a lipid surface of defined curvature; proposing one on a single study would be premature, so the molecular-binding term is kept and the question is raised in suggested_questions. Supporting Evidence: PMID:25119034 Human caspase-4 and the mouse homologue caspase-11 (hereafter referred to as caspase-4/11) and also human caspase-5, directly bound to LPS and lipid A with high specificity and affinity. PMID:41477831 Using HDX-MS, we identified a hydrophobic pocket formed by helices Ξ±1, Ξ±2, and Ξ±5 in caspase-4/11 CARD that is critical for recognizing the acyl chains of LPS. PMID:41702406 we investigated how caspase-4 accesses its cognate ligand, the hydrophobic lipid A moiety of LPS PMID:41702406 Fragmented LPS micelles presented additional micelle tips that served as binding and activation sites for caspase-4, indicating that caspase-4 engages LPS membranes with defined geometry rather than individual LPS molecules. |
| GO:0004197 cysteine-type endopeptidase activity | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0004197 cysteine-type endopeptidase activity | IDA PMID:37993714 Recognition and maturation of IL-18 by caspase-4 noncanonica... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0042742 defense response to bacterium | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | ACCEPT | Summary: CASP4 restricts intracellular Gram-negative bacteria by sensing their LPS and triggering pyroptosis. Reason: Core biological process, supported across Salmonella, Shigella and Burkholderia models and by the existence of dedicated bacterial evasion mechanisms (OspC3 ADP-riboxanation of the caspase), which is strong evidence that the pathway matters in vivo. Supporting Evidence: PMID:34671164 Mouse caspase-11 and human caspase-4 and caspase-5 recognize cytosolic lipopolysaccharide (LPS) to induce pyroptosis by cleaving the pore-forming protein GSDMD PMID:31268602 GBP1 facilitated caspase-4 recruitment to Salmonella leading to its enhanced activation and pyroptosis. |
| GO:0042742 defense response to bacterium | IDA PMID:37993714 Recognition and maturation of IL-18 by caspase-4 noncanonica... | ACCEPT | Summary: CASP4 restricts intracellular Gram-negative bacteria by sensing their LPS and triggering pyroptosis. Reason: Core biological process, supported across Salmonella, Shigella and Burkholderia models and by the existence of dedicated bacterial evasion mechanisms (OspC3 ADP-riboxanation of the caspase), which is strong evidence that the pathway matters in vivo. Supporting Evidence: PMID:34671164 Mouse caspase-11 and human caspase-4 and caspase-5 recognize cytosolic lipopolysaccharide (LPS) to induce pyroptosis by cleaving the pore-forming protein GSDMD PMID:31268602 GBP1 facilitated caspase-4 recruitment to Salmonella leading to its enhanced activation and pyroptosis. |
| GO:0050729 positive regulation of inflammatory response | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | ACCEPT | Summary: CASP4 activation drives release of mature IL-18 and IL-1 family cytokines and downstream inflammation. Reason: Correct and central. CASP4 both matures pro-IL-18 directly and licenses canonical inflammasome output; the IBD node is seeded by CASP4's own experimental annotation together with mouse Casp4/11. Supporting Evidence: PMID:37993712 enables human cells to bypass the need for canonical inflammasomes and caspase-1 for IL-18 release. PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. |
| GO:0050729 positive regulation of inflammatory response | IDA PMID:37993714 Recognition and maturation of IL-18 by caspase-4 noncanonica... | ACCEPT | Summary: CASP4 activation drives release of mature IL-18 and IL-1 family cytokines and downstream inflammation. Reason: Correct and central. CASP4 both matures pro-IL-18 directly and licenses canonical inflammasome output; the IBD node is seeded by CASP4's own experimental annotation together with mouse Casp4/11. Supporting Evidence: PMID:37993712 enables human cells to bypass the need for canonical inflammasomes and caspase-1 for IL-18 release. PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. |
| GO:0051604 protein maturation | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | ACCEPT | Summary: CASP4 proteolytically matures its substrates: GSDMD into its pore-forming N-terminal fragment, and pro-IL-18 into mature IL-18. Reason: Core biological process and the direct consequence of the catalytic activity. Two independent 2023 structural studies established direct pro-IL-18 maturation by human CASP4, and GSDMD processing is established by reconstitution and structure. Supporting Evidence: PMID:37993714 Here we show that activated human caspase-4, but not mouse caspase-11, directly and efficiently processes IL-18 in vitro and during bacterial infections. PMID:37993712 Here we demonstrate that the lipopolysaccharide receptor caspase-4 from humans and other mammalian species (except rodents) can cleave pro-IL-18 PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. |
| GO:0051604 protein maturation | IDA PMID:37993714 Recognition and maturation of IL-18 by caspase-4 noncanonica... | ACCEPT | Summary: CASP4 proteolytically matures its substrates: GSDMD into its pore-forming N-terminal fragment, and pro-IL-18 into mature IL-18. Reason: Core biological process and the direct consequence of the catalytic activity. Two independent 2023 structural studies established direct pro-IL-18 maturation by human CASP4, and GSDMD processing is established by reconstitution and structure. Supporting Evidence: PMID:37993714 Here we show that activated human caspase-4, but not mouse caspase-11, directly and efficiently processes IL-18 in vitro and during bacterial infections. PMID:37993712 Here we demonstrate that the lipopolysaccharide receptor caspase-4 from humans and other mammalian species (except rodents) can cleave pro-IL-18 PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:26375003 Cleavage of GSDMD by inflammatory caspases determines pyropt... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:37993712 Structural insights into cytokine cleavage by inflammatory c... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:37993714 Recognition and maturation of IL-18 by caspase-4 noncanonica... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0160074 non-canonical inflammasome complex | IDA PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | ACCEPT | Summary: CASP4 is the defining protease component of the non-canonical inflammasome. Reason: Core cellular component; the GO definition names CASP4 by name. Two 2026 structural studies refine what this 'complex' is - it is heterogeneous rather than stoichiometric, with three major LPS:caspase species, and assembly proceeds by CARD folding-on-LPS and polymerisation - but they do not challenge membership. Supporting Evidence: PMID:42546204 we determine the stoichiometry of the non-canonical inflammasome showing that it is heterogeneous, comprised of three major complexes with different numbers of LPS and caspase molecules. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. |
| GO:0160074 non-canonical inflammasome complex | IDA PMID:31268602 Human GBP1 is a microbe-specific gatekeeper of macrophage ap... | ACCEPT | Summary: CASP4 is the defining protease component of the non-canonical inflammasome. Reason: Core cellular component; the GO definition names CASP4 by name. Two 2026 structural studies refine what this 'complex' is - it is heterogeneous rather than stoichiometric, with three major LPS:caspase species, and assembly proceeds by CARD folding-on-LPS and polymerisation - but they do not challenge membership. Supporting Evidence: PMID:42546204 we determine the stoichiometry of the non-canonical inflammasome showing that it is heterogeneous, comprised of three major complexes with different numbers of LPS and caspase molecules. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0160075 non-canonical inflammasome complex assembly | IDA PMID:31268602 Human GBP1 is a microbe-specific gatekeeper of macrophage ap... | ACCEPT | Summary: LPS engagement drives CARD-mediated oligomerisation of CASP4 into the non-canonical inflammasome. Reason: Core biological process. Oligomerisation on LPS is the activation step itself, and 2026 structural work shows how it happens - the CARD is disordered until LPS engagement, then folds and polymerises into assemblies of eight or more protomers, with heterogeneous LPS:caspase stoichiometry. Supporting Evidence: PMID:25119034 Insect-cell purified caspase-4/11 underwent oligomerization upon LPS binding, resulting in activation of the caspases. PMID:41477831 caspase-4/11 CARDs are intrinsically unstructured in their resting state and adopt an Ξ±-helical conformation upon LPS engagement. PMID:41702406 In vitro, GBP1 fragmented LPS micelles and promoted caspase-4/LPS complex formation, thereby enhancing LPS-induced caspase-4 activation. |
| GO:0070269 pyroptotic inflammatory response | IDA PMID:31268602 Human GBP1 is a microbe-specific gatekeeper of macrophage ap... | ACCEPT | Summary: Pyroptosis is the defining outcome of CASP4 activation by cytosolic LPS. Reason: Core biological process, supported by reconstitution, by CRISPR screens identifying GSDMD as the executioner, and by pathogen-evasion genetics. The IBD node PTN002573059 is seeded by CASP4 itself plus mouse Casp4/11 and zebrafish caspa/caspb, so the phylogenetic assertion adds the claim that the role is inherited rather than human-specific. Supporting Evidence: PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:32109412 we show site-specific caspase-4/11 autoprocessing, generating a p10 product, is required and sufficient for cleaving GSDMD and inducing pyroptosis. PMID:34671164 S. flexneri evaded pyroptosis mediated by caspase-11 or caspase 4 (hereafter referred to as caspase-11/4) using a type III secretion system (T3SS) effector, OspC3. |
| GO:0001530 lipopolysaccharide binding | IDA PMID:25119034 Inflammatory caspases are innate immune receptors for intrac... | ACCEPT | Summary: Direct, high-affinity binding of CASP4 to LPS and to its lipid A moiety - the receptor function of the non-canonical inflammasome. The abstraction is challenged by 2026 work showing the physiological ligand is a curved LPS membrane rather than a free LPS molecule. Reason: Retained as a core molecular function, with the dispute recorded. The IDAs rest on direct binding measurements (Shi et al. 2014) and on the 2023 structural work, and a 2026 HDX-MS study localises recognition to a hydrophobic pocket formed by helices Ξ±1, Ξ±2 and Ξ±5 of the CARD that engages the acyl chains of LPS - i.e. a genuine chemical interaction between the protein and the lipid. Against this, Began et al. 2026 report that caspase-4 binds LPS membranes with positive curvature and conclude it engages membranes with defined geometry rather than individual LPS molecules, with GBP1 required to generate that geometry on cytosolic bacteria. These are reconcilable at the level GO annotates: the same paper still calls lipid A the cognate ligand, and curvature governs presentation and avidity rather than the chemistry of recognition. GO has no term for binding a lipid surface of defined curvature; proposing one on a single study would be premature, so the molecular-binding term is kept and the question is raised in suggested_questions. Supporting Evidence: PMID:25119034 Human caspase-4 and the mouse homologue caspase-11 (hereafter referred to as caspase-4/11) and also human caspase-5, directly bound to LPS and lipid A with high specificity and affinity. PMID:41477831 Using HDX-MS, we identified a hydrophobic pocket formed by helices Ξ±1, Ξ±2, and Ξ±5 in caspase-4/11 CARD that is critical for recognizing the acyl chains of LPS. PMID:41702406 we investigated how caspase-4 accesses its cognate ligand, the hydrophobic lipid A moiety of LPS PMID:41702406 Fragmented LPS micelles presented additional micelle tips that served as binding and activation sites for caspase-4, indicating that caspase-4 engages LPS membranes with defined geometry rather than individual LPS molecules. |
| GO:0016540 protein autoprocessing | IDA PMID:34671164 Shigella evades pyroptosis by arginine ADP-riboxanation of c... | ACCEPT | Summary: CASP4 self-cleaves in the interdomain linker, and this autoprocessing is required for substrate cleavage. Reason: Core: autoprocessing is not a by-product but the licensing step. Structural work shows the p10-form autoprocessed caspase is what engages GSDMD, and mutational work maps the two self-cleavage sites (D270, D289) and links D289 processing to acquisition of full protease activity. Supporting Evidence: PMID:32109412 we show site-specific caspase-4/11 autoprocessing, generating a p10 product, is required and sufficient for cleaving GSDMD and inducing pyroptosis. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0051604 protein maturation | IDA PMID:26375003 Cleavage of GSDMD by inflammatory caspases determines pyropt... | ACCEPT | Summary: CASP4 proteolytically matures its substrates: GSDMD into its pore-forming N-terminal fragment, and pro-IL-18 into mature IL-18. Reason: Core biological process and the direct consequence of the catalytic activity. Two independent 2023 structural studies established direct pro-IL-18 maturation by human CASP4, and GSDMD processing is established by reconstitution and structure. Supporting Evidence: PMID:37993714 Here we show that activated human caspase-4, but not mouse caspase-11, directly and efficiently processes IL-18 in vitro and during bacterial infections. PMID:37993712 Here we demonstrate that the lipopolysaccharide receptor caspase-4 from humans and other mammalian species (except rodents) can cleave pro-IL-18 PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. |
| GO:0004197 cysteine-type endopeptidase activity | IDA PMID:34671164 Shigella evades pyroptosis by arginine ADP-riboxanation of c... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0050729 positive regulation of inflammatory response | IDA PMID:33377178 NLRP6-caspase 4 inflammasome activation in response to cario... | ACCEPT | Summary: CASP4 activation drives release of mature IL-18 and IL-1 family cytokines and downstream inflammation. Reason: Correct and central. CASP4 both matures pro-IL-18 directly and licenses canonical inflammasome output; the IBD node is seeded by CASP4's own experimental annotation together with mouse Casp4/11. Supporting Evidence: PMID:37993712 enables human cells to bypass the need for canonical inflammasomes and caspase-1 for IL-18 release. PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. |
| GO:0050729 positive regulation of inflammatory response | IDA PMID:34671164 Shigella evades pyroptosis by arginine ADP-riboxanation of c... | ACCEPT | Summary: CASP4 activation drives release of mature IL-18 and IL-1 family cytokines and downstream inflammation. Reason: Correct and central. CASP4 both matures pro-IL-18 directly and licenses canonical inflammasome output; the IBD node is seeded by CASP4's own experimental annotation together with mouse Casp4/11. Supporting Evidence: PMID:37993712 enables human cells to bypass the need for canonical inflammasomes and caspase-1 for IL-18 release. PMID:26508369 Here we report that caspase-4 and caspase-5 mediate IL-1Ξ± and IL-1Ξ² release from human monocytes after LPS stimulation. |
| GO:0050830 defense response to Gram-positive bacterium | IDA PMID:33377178 NLRP6-caspase 4 inflammasome activation in response to cario... | KEEP AS NON CORE | Summary: CASP4 responds to lipoteichoic acid from Gram-positive Streptococcus mutans via the NLRP6 inflammasome. Reason: A real but indirect and context-restricted role: here CASP4 acts downstream of a distinct sensor (NLRP6) in dental pulp cells, rather than as the direct pattern receptor. It should not be read as CASP4 binding a Gram-positive ligand. Supporting Evidence: PMID:33377178 After LPS priming, cytosolic Streptococcus mutans LTA triggered NLRP6-caspase 4 inflammasome activation. |
| GO:0070269 pyroptotic inflammatory response | IDA PMID:34671164 Shigella evades pyroptosis by arginine ADP-riboxanation of c... | ACCEPT | Summary: Pyroptosis is the defining outcome of CASP4 activation by cytosolic LPS. Reason: Core biological process, supported by reconstitution, by CRISPR screens identifying GSDMD as the executioner, and by pathogen-evasion genetics. The IBD node PTN002573059 is seeded by CASP4 itself plus mouse Casp4/11 and zebrafish caspa/caspb, so the phylogenetic assertion adds the claim that the role is inherited rather than human-specific. Supporting Evidence: PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:32109412 we show site-specific caspase-4/11 autoprocessing, generating a p10 product, is required and sufficient for cleaving GSDMD and inducing pyroptosis. PMID:34671164 S. flexneri evaded pyroptosis mediated by caspase-11 or caspase 4 (hereafter referred to as caspase-11/4) using a type III secretion system (T3SS) effector, OspC3. |
| GO:2000494 positive regulation of interleukin-18-mediated signaling pathway | IDA PMID:33377178 NLRP6-caspase 4 inflammasome activation in response to cario... | MODIFY | Summary: Annotated as positive regulation of IL-18 receptor signalling, whereas CASP4's actual role is to generate mature IL-18. Reason: Wrong level of the pathway. CASP4 cleaves pro-IL-18 at the caspase-1 tetrapeptide site and enables its release through the GSDMD pore; it does not act within the IL-18-receptor signalling cascade. Positive regulation of interleukin-18 production is the term that matches what was measured and what the 2023 structural studies established. Proposed replacements: positive regulation of interleukin-18 production Supporting Evidence: PMID:37993712 enables human cells to bypass the need for canonical inflammasomes and caspase-1 for IL-18 release. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. |
| GO:0005515 protein binding | IPI PMID:29077095 Caspase-4 activation by a bacterial surface protein is media... | REMOVE | Summary: Bare protein-binding annotations from interaction screens and targeted co-IP. Reason: Per project guidance, bare protein binding conveys no functional information. The underlying interactions are real and in two cases mechanistically important - SERPINB1 blocks CARD oligomerisation, and cathepsin G binds and activates CASP4 - but each should be captured by an informative term (e.g. CARD domain binding) or as a regulatory relationship, not by GO:0005515. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. Supporting Evidence: PMID:30692621 Here we report that serpin family B member 1 (SERPINB1) limited the activity of those caspases by suppressing their caspase-recruitment domain (CARD) oligomerization and enzymatic activation. PMID:29077095 Td92 treatment resulted in the binding of cathepsin G to caspase-4 and the coaggregation of these two molecules. |
| GO:0004197 cysteine-type endopeptidase activity | IDA PMID:32109412 Structural Mechanism for GSDMD Targeting by Autoprocessed Ca... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0016540 protein autoprocessing | IDA PMID:32109412 Structural Mechanism for GSDMD Targeting by Autoprocessed Ca... | ACCEPT | Summary: CASP4 self-cleaves in the interdomain linker, and this autoprocessing is required for substrate cleavage. Reason: Core: autoprocessing is not a by-product but the licensing step. Structural work shows the p10-form autoprocessed caspase is what engages GSDMD, and mutational work maps the two self-cleavage sites (D270, D289) and links D289 processing to acquisition of full protease activity. Supporting Evidence: PMID:32109412 we show site-specific caspase-4/11 autoprocessing, generating a p10 product, is required and sufficient for cleaving GSDMD and inducing pyroptosis. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0050727 regulation of inflammatory response | IDA PMID:32109412 Structural Mechanism for GSDMD Targeting by Autoprocessed Ca... | MODIFY | Summary: Unsigned regulation-of-inflammation term where the direction is known. Reason: CASP4's effect on inflammation is unambiguously positive - it drives GSDMD-dependent pyroptosis and cytokine release - and the positive child term is already carried by this gene under both IBA and IDA evidence. Proposed replacements: positive regulation of inflammatory response |
| GO:0070269 pyroptotic inflammatory response | IDA PMID:32109412 Structural Mechanism for GSDMD Targeting by Autoprocessed Ca... | ACCEPT | Summary: Pyroptosis is the defining outcome of CASP4 activation by cytosolic LPS. Reason: Core biological process, supported by reconstitution, by CRISPR screens identifying GSDMD as the executioner, and by pathogen-evasion genetics. The IBD node PTN002573059 is seeded by CASP4 itself plus mouse Casp4/11 and zebrafish caspa/caspb, so the phylogenetic assertion adds the claim that the role is inherited rather than human-specific. Supporting Evidence: PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:32109412 we show site-specific caspase-4/11 autoprocessing, generating a p10 product, is required and sufficient for cleaving GSDMD and inducing pyroptosis. PMID:34671164 S. flexneri evaded pyroptosis mediated by caspase-11 or caspase 4 (hereafter referred to as caspase-11/4) using a type III secretion system (T3SS) effector, OspC3. |
| GO:0005515 protein binding | IPI PMID:30692621 SERPINB1-mediated checkpoint of inflammatory caspase activat... | REMOVE | Summary: Bare protein-binding annotations from interaction screens and targeted co-IP. Reason: Per project guidance, bare protein binding conveys no functional information. The underlying interactions are real and in two cases mechanistically important - SERPINB1 blocks CARD oligomerisation, and cathepsin G binds and activates CASP4 - but each should be captured by an informative term (e.g. CARD domain binding) or as a regulatory relationship, not by GO:0005515. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. Supporting Evidence: PMID:30692621 Here we report that serpin family B member 1 (SERPINB1) limited the activity of those caspases by suppressing their caspase-recruitment domain (CARD) oligomerization and enzymatic activation. PMID:29077095 Td92 treatment resulted in the binding of cathepsin G to caspase-4 and the coaggregation of these two molecules. |
| GO:0004197 cysteine-type endopeptidase activity | IDA PMID:16920334 Protective role of Cop in Rip2/caspase-1/caspase-4-mediated ... | ACCEPT | Summary: CASP4 is a cysteine protease that cleaves after Asp; this is its uncontested core molecular function. Reason: Core molecular function, established from the 1995 cloning papers through to 2023 cryo-EM and crystal structures of substrate complexes, and not in dispute in any of the 2026 literature. The substrate repertoire is contested (GSDMD vs pro-IL-18 vs CASP3/CASP7 vs pro-IL-1Ξ²) but the catalytic activity itself is not. Supporting Evidence: PMID:7797510 Purified ICH-2 is functional as a protease in vitro. PMID:26375003 Caspase-1 and caspase-4/5/11 specifically cleaved the linker between the amino-terminal gasdermin-N and carboxy-terminal gasdermin-C domains in GSDMD, which was required and sufficient for pyroptosis. PMID:37993714 Caspase-4 cleaves the same tetrapeptide site in pro-IL-18 as caspase-1. PMID:37558421 Here, we show that caspase-4 first dimerises then self-cleaves at two sites-D270 and D289-in the interdomain linker to acquire full proteolytic activity, cleave GSDMD, and induce cell death. |
| GO:0032991 protein-containing complex | IDA PMID:16920334 Protective role of Cop in Rip2/caspase-1/caspase-4-mediated ... | MARK AS OVER ANNOTATED | Summary: Root-level complex membership from a HeLa overexpression study. Reason: Carries no functional information: every protein that touches another protein satisfies it. The underlying observation is a CARD-CARD interaction with the CARD-only protein Cop/CARD16, which is already captured more informatively by the GO:0050700 CARD domain binding annotation from the same paper. Supporting Evidence: PMID:16920334 in addition to the known interaction of Cop and caspase-1, we demonstrated a novel interaction of Cop with caspase-4. |
| GO:0050700 CARD domain binding | IPI PMID:16920334 Protective role of Cop in Rip2/caspase-1/caspase-4-mediated ... | KEEP AS NON CORE | Summary: CARD-CARD interaction between CASP4 and the CARD-only protein Cop/CARD16. Reason: An informative binding term and a real interaction, unlike the bare protein-binding rows. It is kept non-core because the annotated interaction is with an inhibitory decoy CARD rather than with the CARD-CARD self-association that drives non-canonical inflammasome assembly; the latter is captured by GO:0160075. Supporting Evidence: PMID:16920334 in addition to the known interaction of Cop and caspase-1, we demonstrated a novel interaction of Cop with caspase-4. |
| GO:1903265 positive regulation of tumor necrosis factor-mediated signaling pathway | IDA PMID:16920334 Protective role of Cop in Rip2/caspase-1/caspase-4-mediated ... | MARK AS OVER ANNOTATED | Summary: TNF-pathway regulation inferred from a HeLa overexpression/inhibitor study of the CARD-only protein Cop. Reason: The experiment establishes that Cop inhibits TNF-alpha-induced death and that Cop binds CASP4; positive regulation of TNF-mediated signalling by CASP4 is an inference from over-expressed components in an epithelial cell line, several steps removed from a demonstrated CASP4 activity. The informative result of that paper - the CARD-CARD interaction - is retained under GO:0050700. Supporting Evidence: PMID:16920334 in addition to the known interaction of Cop and caspase-1, we demonstrated a novel interaction of Cop with caspase-4. |
| GO:0097193 intrinsic apoptotic signaling pathway | IMP PMID:15123740 Involvement of caspase-4 in endoplasmic reticulum stress-ind... | KEEP AS NON CORE | Summary: Intrinsic apoptotic signalling, driven by the ER-stress arm of CASP4 biology. Reason: Supported by knockdown experiments showing reduced ER-stress-induced apoptosis, and reproduced by an independent group via the TMEM214 anchor. Genuine, but a separate arm from the LPS-sensing function for which the gene is principally responsible. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. |
| GO:1904646 cellular response to amyloid-beta | IMP PMID:15123740 Involvement of caspase-4 in endoplasmic reticulum stress-ind... | KEEP AS NON CORE | Summary: CASP4 is cleaved on amyloid-beta exposure and knockdown reduces A-beta-induced death. Reason: Supported by the original ER-stress study's siRNA experiments and biologically coherent with the ER-stress arm, but it is a disease-context response in a specialised setting rather than part of the gene's core evolved function. Supporting Evidence: PMID:15123740 Caspase-4 is also cleaved by administration of Abeta, and Abeta-induced apoptosis is reduced by small interfering RNAs to caspase-4. |
| GO:0005739 mitochondrion | IDA PMID:15123740 Involvement of caspase-4 in endoplasmic reticulum stress-ind... | KEEP AS NON CORE | Summary: Mitochondrial localisation reported alongside ER localisation in the ER-stress literature. Reason: Reported by two independent groups working on ER-stress apoptosis and carried by UniProt, so retained; but the mitochondrial pool has no assigned activity and is unrelated to the LPS-sensing function that defines the gene. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005783 endoplasmic reticulum | IDA PMID:15123740 Involvement of caspase-4 in endoplasmic reticulum stress-ind... | KEEP AS NON CORE | Summary: ER localisation, the less specific counterpart of the ER-membrane annotation. Reason: Same evidence and same standing as GO:0005789, which is the more informative term of the pair. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. |
| GO:0070059 intrinsic apoptotic signaling pathway in response to endoplasmic reticulum stress | IMP PMID:15123740 Involvement of caspase-4 in endoplasmic reticulum stress-ind... | KEEP AS NON CORE | Summary: The ER-stress-specific form of the intrinsic apoptotic annotation. Reason: The most informative statement of the ER-stress arm: CASP4 cleavage is triggered specifically by ER-stress inducers and not by other apoptotic stimuli, and knockdown reduces ER-stress-induced death. Non-core for the same reason as GO:0097193. Supporting Evidence: PMID:15123740 We found that human caspase-4, a member of caspase-1 subfamily that includes caspase-12, is localized to the ER membrane, and is cleaved when cells are treated with ER stress-inducing reagents, but not with other apoptotic reagents. PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9686096 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9686271 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9947934 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9947940 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9948160 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9956613 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9956667 | ACCEPT | Summary: Cytosolic localisation, asserted by phylogeny, by UniProt subcellular-location mapping, by Reactome and by direct observation. Reason: Core location. The non-canonical inflammasome is a cytosolic sensor of LPS that has reached the host cytoplasm; the PAINT node PTN008306143 is seeded in part by CASP4's own experimental annotation, which is the expected marker of experimental grounding on the target rather than circularity. Supporting Evidence: PMID:23661706 TMEM214 was localized on the outer membrane of the ER and constitutively associated with procaspase 4, which was also critical for ER stress-induced apoptosis. |
| GO:0006508 proteolysis | TAS PMID:7797510 Identification and characterization of ICH-2, a novel member... | KEEP AS NON CORE | Summary: Generic proteolysis, true of any protease. Reason: Correct but uninformative next to the specific processes CASP4 drives (GO:0051604 protein maturation, GO:0016540 protein autoprocessing, GO:0160075 non-canonical inflammasome complex assembly). Retained as a true parent rather than modified, because no single child term subsumes the whole of CASP4's proteolytic activity. |
| GO:0006915 apoptotic process | TAS PMID:7743998 A novel human protease similar to the interleukin-1 beta con... | KEEP AS NON CORE | Summary: Generic apoptosis term inherited from the pan-caspase node and from the 1995 overexpression papers; CASP4 does have an apoptotic arm, but it is not what the gene is for. Reason: Retained but demoted. CASP4 overexpression induced apoptosis in the founding transfection studies, it participates in ER-stress apoptosis, and a 2026 study reports that CASP4/5 directly cleave and activate the executioner caspases CASP3 and CASP7 - which, if it holds, gives the apoptotic arm real mechanistic footing. That claim rests on a single study and is one of the points currently in dispute, so it is recorded here rather than promoted to a core function. Supporting Evidence: PMID:42044191 intracellular LPS and the gram-negative bacterial pathogen Salmonella activate CASP4/5 in macrophages to directly cleave and activate CASP3 and CASP7. PMID:42044191 CASP1, CASP4, CASP5 and to a lesser extent, CASP11, processed CASP3C163A and CASP7C186A but failed to process the D175A and D198A mutants PMID:7743998 Transfection experiments demonstrate that TX is a protease which is able to cleave itself and the p30 ICE precursor, but not to generate mature IL-1 beta from pro-IL-1 beta. |
| GO:0006915 apoptotic process | TAS PMID:7797510 Identification and characterization of ICH-2, a novel member... | KEEP AS NON CORE | Summary: Generic apoptosis term inherited from the pan-caspase node and from the 1995 overexpression papers; CASP4 does have an apoptotic arm, but it is not what the gene is for. Reason: Retained but demoted. CASP4 overexpression induced apoptosis in the founding transfection studies, it participates in ER-stress apoptosis, and a 2026 study reports that CASP4/5 directly cleave and activate the executioner caspases CASP3 and CASP7 - which, if it holds, gives the apoptotic arm real mechanistic footing. That claim rests on a single study and is one of the points currently in dispute, so it is recorded here rather than promoted to a core function. Supporting Evidence: PMID:42044191 intracellular LPS and the gram-negative bacterial pathogen Salmonella activate CASP4/5 in macrophages to directly cleave and activate CASP3 and CASP7. PMID:42044191 CASP1, CASP4, CASP5 and to a lesser extent, CASP11, processed CASP3C163A and CASP7C186A but failed to process the D175A and D198A mutants PMID:7743998 Transfection experiments demonstrate that TX is a protease which is able to cleave itself and the p30 ICE precursor, but not to generate mature IL-1 beta from pro-IL-1 beta. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Is the annotatable ligand of CASP4 a lipopolysaccharide molecule or an LPS membrane surface of defined curvature? Direct binding assays and HDX-MS support a lipid A acyl-chain pocket in the CARD, while micelle-fragmentation experiments argue that caspase-4 engages membranes with geometry rather than individual LPS molecules. If the latter is the operative description, GO would need a curvature-sensing molecular-function term rather than GO:0001530.
Suggested experts: Petr Broz, Jan Began, Jianbin Ruan, Feng Shao
Q: Are CASP3 and CASP7 physiological substrates of CASP4, or is their processing during non-canonical signalling dominated by other routes in most cell types? A single 2026 study reports direct cleavage at D175 and D198 by recombinant CASP4/5; independent replication would be needed before curating them as CASP4 substrates.
Suggested experts: Cornelius Y. Taabazuing, Mandar Kulkarni
Q: What fraction of gasdermin D cleavage during non-canonical inflammasome signalling is carried out by CASP4 itself rather than by CASP1 acting downstream? The answer does not change the molecular-function annotation, which rests on reconstitution and structure, but it determines how a GO-CAM should wire the pathway.
Suggested experts: Cornelius Y. Taabazuing, Feng Shao, Judy Lieberman
Q: Does CASP4 directly process pro-interleukin-1beta? UniProt asserts that it does not, citing the 1995 characterisations, whereas a 2023 study reports that the D289-autoprocessed p34/p9 species cleaves pro-IL-1beta independently of the NLRP3 inflammasome.
Suggested experts: Kate Schroder, Dave Boucher
Q: Given that human CASP4 matures pro-IL-18 while mouse caspase-11 does not, which mouse Casp4/Casp11 phenotypes can legitimately be used to annotate human CASP4, and which should be restricted to the shared LPS-sensing and gasdermin D axis?
Suggested experts: Feng Shao, Jonathan C. Kagan, Pascal Devant
Q: Is there a non-LPS lipid ligand for CASP4? A 2025 report identifies sphingomyelin C12 as a caspase-4 ligand; if confirmed, the generic lipid binding annotation should be retained rather than collapsed into lipopolysaccharide binding.
Experiment: Titrate purified CASP4 (and its isolated CARD) against defined lipid A/LPS presentations of matched chemical composition but controlled geometry - monomeric lipid A below the critical micelle concentration, planar supported bilayers, and nanodiscs or vesicles of graded diameter giving a curvature series - and measure affinity and oligomerisation in parallel by isothermal titration calorimetry, mass photometry and native mass spectrometry. A geometry requirement predicts binding that tracks curvature at constant lipid A concentration; a molecular-ligand model predicts binding to sub-CMC monomeric lipid A.
Hypothesis: CASP4 binding requires lipid A presented on a curved membrane surface, not free LPS molecules.
Type: Quantitative binding and oligomerisation assays across a defined membrane-curvature series
Experiment: In human macrophages (THP-1 and primary monocyte-derived) carrying CASP1, CASP8, CASP9, GSDMD and GSDME knockouts, reconstitute CASP4 knockouts with wild-type, catalytically dead (C258A) and autoprocessing-defective (D270A/D289A) CASP4, then quantify CASP3 D175 and CASP7 D198 neo-epitope cleavage by targeted mass spectrometry with a stable-isotope-labelled standard. Direct, CASP4-dependent cleavage should require CASP4 catalytic activity and persist in the CASP1/CASP8/CASP9 triple knockout.
Hypothesis: CASP3 and CASP7 are direct, physiologically relevant CASP4 substrates during cytosolic LPS sensing.
Type: Genetic reconstitution with neo-epitope-resolved quantitative proteomics
Experiment: Compare CASP4 activation kinetics in GBP1-null cells challenged with (i) transfected free LPS, (ii) cytosolic Gram-negative bacteria, and (iii) synthetic LPS-bearing vesicles of defined curvature delivered to the cytosol. If GBP1's only role is to generate curvature, pre-curved synthetic vesicles should rescue activation in GBP1-null cells; residual GBP1 dependence would indicate an additional scaffolding or recruitment function.
Hypothesis: The CASP4-GBP1 requirement can be separated from the CASP4-LPS binding step.
Type: Genetic complementation with defined synthetic ligand delivery
Experiment: Side-by-side, activity-normalised in vitro cleavage of a common substrate panel (GSDMD, pro-IL-18, pro-IL-1beta, pro-CASP3, pro-CASP7, CGAS) by recombinant human CASP4, human CASP5 and mouse caspase-11, quantified by densitometry and by kinetic assays on the corresponding tetrapeptide substrates, to produce an explicit substrate-specificity matrix that curators can use to decide which mouse results may be transferred.
Hypothesis: Human CASP4 and mouse caspase-11 differ in substrate repertoire in a way that limits cross-species annotation transfer.
Type: Comparative enzymology across orthologues and paralogues
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)