Interferon-inducible innate-immune protein built from two N-terminal Z-nucleic-acid-binding (Zalpha1, Zalpha2) domains followed by three RIP homotypic interaction motifs (RHIMs). It is expressed at low levels at rest and strongly upregulated by immunostimulation, and is found predominantly in the cytosol, with a functionally important nuclear pool. ZBP1 recognises double-stranded nucleic acids that are in, or predisposed to adopt, the left-handed Z conformation, and converts that recognition into cell death by using its RHIMs to nucleate a signalling complex with the kinases RIPK1 and RIPK3, driving necroptosis, apoptosis and inflammatory cell death. Z-form RNA arising during influenza A and herpes simplex virus infection, including host transcripts generated when viruses disrupt transcription termination, is a physiological activating ligand, and the antagonism of ZBP1 by vaccinia E3 and by the Zalpha domain of ADAR1 places it in a competitive balance that determines cell survival versus death. Two aspects of its biology are actively disputed. Whether the Zalpha domains can themselves drive the A-to-Z conversion of unmodified double-stranded RNA is contested: biophysical work indicates they convert unmodified DNA but not unmodified RNA, and bind Z-form RNA only when a chemical modification or other strain has already lowered the energetic barrier, which would restrict the protein to reading pre-formed Z-RNA rather than generating it. The human and mouse proteins also differ mechanistically, the human protein requiring all three RHIMs and signalling through RIPK1 largely independently of RIPK3, whereas the mouse protein signals mainly through RIPK3 using RHIM1 alone. The early proposal that the protein acts as a general cytosolic DNA sensor driving type I interferon induction has not held up in human cells.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0003692 left-handed Z-DNA binding | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic transfer of Z-DNA binding to the Zalpha-domain protein the gene is named for. The assertion is correct and is independently demonstrated for the human protein. Reason: Core molecular function. Worth noting how weakly GOA supports it: in the human record left-handed Z-DNA binding is carried only by this IBA and by an NAS, even though direct biophysical demonstration for human ZBP1 has existed since 2006, where circular dichroism and EMSA showed each Zalpha domain binds Z-DNA on its own and that linking them enhances binding. A 2026 preprint independently confirms that both human Zalpha domains convert and stabilise unmodified dsDNA in the Z form. The phylogenetic judgement is not in dispute and the human protein is the 1:1 ortholog of the mouse donor, so the node placement is sound; the annotation should be upgraded to IDA rather than argued with. Supporting Evidence: PMID:16990255 both Zalpha domains can bind Z-DNA independently and that substrate binding is greatly enhanced when both domains are linked PMID:42079158 While ZBP1's ZΞ± domains are able to convert and stabilize unmodified dsDNA in the Z-conformation |
| GO:0003723 RNA binding | IBA GO_REF:0000033 | ACCEPT | Summary: RNA binding transferred phylogenetically, with human ZBP1's own experimental annotation among the sources. Reason: Correct, though generic. Human ZBP1 appears in this IBA's WITH/FROM alongside mouse Zbp1 and the PANTHER node, which is expected and is a marker that experimental grounding exists on the target itself rather than a sign of circularity. The binding is real for both A-form and Z-form RNA: 2026 biophysical work finds the Zalpha domains engage unmodified duplex RNA in an A-form-like mode and stabilise the Z form once a Z-promoting modification lowers the barrier. The more informative term for the functional consequence, left-handed Z-RNA immune receptor activity, is separately annotated and is not an ontology ancestor of RNA binding, so the two coexist rather than one superseding the other. Supporting Evidence: PMID:42079158 We show that ZBP1's ZΞ± domains require dsRNAs with Z-promoting chemical modification in order for them to bind and stabilize the Z-conformation. |
| GO:0005634 nucleus | IBA GO_REF:0000033 | ACCEPT | Summary: ZBP1 has a nuclear pool in which it is genuinely active, most clearly during influenza A infection. Reason: The is_active_in qualifier is justified rather than merely tolerated here: influenza A-generated Z-RNA activates ZBP1 in the nucleus, and the resulting RIPK3/MLKL activation occurs there, disrupting the nuclear envelope. The steady-state distribution is predominantly cytoplasmic, so this is a secondary but functionally load-bearing site. Supporting Evidence: PMID:32200799 Here, we show that replicating IAV generates Z-RNAs, which activate ZBP1 in the nucleus of infected cells. |
| GO:0005737 cytoplasm | IBA GO_REF:0000033 | ACCEPT | Summary: The principal compartment in which ZBP1 senses ligand and assembles signalling complexes. Reason: Correct and core, with human ZBP1's own experimental annotations among the sources. The cytosol annotation carried by this gene (GO:0005829) is more informative about where the protein actually sits. |
| GO:0043065 positive regulation of apoptotic process | IBA GO_REF:0000033 | ACCEPT | Summary: ZBP1 activation drives apoptosis as one of its cell-death outputs. Reason: Supported for the human protein by 2026 work showing hZBP1 is a potent inducer of apoptosis, requiring the scaffold function of RIPK1 but not its kinase activity. The human/mouse divergence reported in that paper concerns which kinase transduces the signal, not whether apoptosis is an output, so the phylogenetic transfer of the process term is unaffected. Supporting Evidence: PMID:42436309 In contrast to mZBP1, which signals primarily through RIPK3, hZBP1-induced cell death depends on RIPK1 in a RIPK3-independent manner. |
| GO:0060340 positive regulation of type I interferon-mediated signaling pathway | IBA GO_REF:0000033 | MARK AS OVER ANNOTATED | Summary: Inherited from the original DAI-as-cytosolic-DNA-sensor model. The term describes regulation of signalling downstream of the interferon receptor, which is not what was reported, and the underlying interferon role specifically failed to replicate in human cells. Reason: Two independent problems. First, term scoping. GO:0060340 is regulation of the type I interferon-mediated signaling pathway, i.e. the response to interferon; what Takaoka et al. reported is DAI-dependent induction of interferon genes, which is interferon production (GO:0032481). The mouse PAINT seed for this term is an IMP from that same paper, so the mismatch was inherited rather than introduced by the transfer. Second, species specificity. Lippmann et al. tested this directly in human cells and found that multiple hZBP1 siRNAs failed to suppress IFN-beta responses to cytosolic poly(dA-dT) or to intracellular bacteria, while mouse Zbp1 siRNA did impair the response in L929 cells; cGAS was subsequently established as the dominant cytosolic DNA sensor. This is the kind of human/mouse divergence that the 2026 EMBO Reports study warns against transferring. Marked over-annotated rather than modified to GO:0032481, because asserting that human ZBP1 positively regulates type I interferon production would assert the very thing that was looked for in human cells and not found. Not removed: ZBP1 is itself interferon-inducible and can amplify IRF3-dependent reporter activity when overexpressed, so the term is not baseless. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: ROLE CONFLATION FUNCTIONAL DIVERGENCE Sources checked: MGI:MGI:1927449 Β· Zbp1 (mouse) SUPPORTS SOURCE BUT NOT TARGET Mouse Zbp1 carries this term as an IMP from PMID:17618271, which measured induction of type I interferon genes (production), not potentiation of interferon receptor signalling. Human ZBP1 was subsequently shown not to be required for IFN-beta responses in human cells (PMID:18771559). PANTHER:PTN008557399 Β· PAINT ancestral node for the ZBP1 family SUPPORTS TRANSFER The node placement itself is not challenged; human ZBP1 is the 1:1 ortholog of the mouse donor and sits inside the clade. The objection is to the GO term chosen and to this particular pathway role in human cells, not to the phylogeny. Supporting Evidence: PMID:18771559 intracellular bacteria and cytosolic poly(dA-dT) activate IFNbeta responses in different human cells without requiring human ZBP1 PMID:17618271 DAI (DLM-1/ZBP1) binds to double-stranded DNA and, by doing so, enhances its association with the IRF3 transcription factor and the TBK1 serine/threonine kinase. |
| GO:0060545 positive regulation of necroptotic process | IBA GO_REF:0000033 | ACCEPT | Summary: The best-established biological output of ZBP1 activation. Reason: Core. Supported for human ZBP1 by its own experimental annotations (which appear in this IBA's WITH/FROM, as expected for a gene that contributed to the ancestral inference), by the influenza Z-RNA work, and by the vaccinia E3 antagonism study. The 2026 human/mouse comparison reports that human ZBP1 is if anything a more potent inducer of cell death than the mouse protein, so the divergence it documents does not undercut this term. Supporting Evidence: PMID:34192517 In the absence of this E3 domain, Z-form RNA accumulates during the early phase of VACV infection, triggering ZBP1 to recruit receptor interacting protein kinase (RIPK)3 and execute necroptosis. |
| GO:0140374 antiviral innate immune response | IBA GO_REF:0000033 | ACCEPT | Summary: ZBP1-triggered cell death is an antiviral effector mechanism. Reason: Core. Independently supported by direct human-record experimental evidence from the vaccinia E3 study and by the influenza work; the fact that two DNA and RNA viruses encode dedicated ZBP1 antagonists is itself strong evidence that the antiviral role is real. Supporting Evidence: PMID:34192517 Necroptosis mediated by Z-nucleic-acid-binding protein (ZBP)1 (also called DAI or DLM1) contributes to innate host defense against viruses by triggering cell death to eliminate infected cells. |
| GO:0002376 immune system process | IEA GO_REF:0000117 | MODIFY | Summary: True but at the root of the immune branch, carrying essentially no information. Reason: Correct in kind but far too general for a protein whose immune role is specifically characterised. The gene already carries the informative child term; this ARBA-generated parent should be replaced by it. Proposed replacements: antiviral innate immune response |
| GO:0003677 DNA binding | IEA GO_REF:0000120 | MODIFY | Summary: Electronic assignment of generic DNA binding from the ZBP1 family InterPro signature. The measured activity is duplex-specific. Reason: DNA binding is correct but under-specific. Every experimental characterisation of this activity, whether B-form or Z-form, is of double-stranded DNA, and the gene separately carries left-handed Z-DNA binding for the Z-form case. GO:0003690 double-stranded DNA binding is the accurate generalisation of the experiment. Proposed replacements: double-stranded DNA binding |
| GO:0003723 RNA binding | IEA GO_REF:0000120 | ACCEPT | Summary: RNA binding assigned electronically from the Z-binding domain signature. Reason: Correct. The Zalpha domains do bind duplex RNA; this half of the InterPro:IPR042371 mapping is sound, in contrast to the deaminase term the same signature also carries. Supporting Evidence: PMID:42079158 We show that ZBP1's ZΞ± domains require dsRNAs with Z-promoting chemical modification in order for them to bind and stabilize the Z-conformation. |
| GO:0003726 double-stranded RNA adenosine deaminase activity | IEA GO_REF:0000002 | REMOVE | Summary: A catalytic activity ZBP1 cannot perform, imported wholesale from ADAR1 because the two proteins share a Zalpha domain. Reason: Demonstrably wrong and traceable to its cause. The annotation derives from InterPro:IPR042371, the generic Z-binding domain entry, which carries both GO:0003723 and GO:0003726 because the prototypical Zalpha-domain protein, ADAR1, is an adenosine deaminase. ZBP1 is not. Its architecture is two Zalpha domains followed by three RHIMs, with no deaminase domain and no catalytic residues anywhere in the Q9H171 record, and no A-to-I editing activity has ever been reported for it. This is a domain-driven electronic mis-mapping in which a shared accessory module drags along the catalytic activity of a different protein, and it is exactly the class of IEA that should be argued down rather than retained. The InterPro2GO mapping on IPR042371 should be corrected at source. Supporting Evidence: PMID:42079158 The full-length gene product includes two N-terminal ZBDs (ZΞ±1 and ZΞ±2) responsible for Z-NA binding and three RIPK Homotypic Interaction Motifs (RHIMs) |
| GO:0005634 nucleus | IEA GO_REF:0000120 | ACCEPT | Summary: Nuclear localisation assigned from the UniProt subcellular location vocabulary. Reason: Agrees with direct microscopy and with the functionally important nuclear activity seen during influenza infection. |
| GO:0005737 cytoplasm | IEA GO_REF:0000120 | ACCEPT | Summary: Cytoplasmic localisation assigned from the UniProt subcellular location vocabulary. Reason: Agrees with direct microscopy; the principal steady-state compartment. |
| GO:0051607 defense response to virus | IEA GO_REF:0000117 | ACCEPT | Summary: Antiviral defence assigned by ARBA, consistent with the experimental record. Reason: Core. The gene independently carries this term with IMP evidence from the influenza Z-RNA study, so the electronic assignment adds nothing but is not wrong. |
| GO:0060340 positive regulation of type I interferon-mediated signaling pathway | IEA GO_REF:0000002 | MARK AS OVER ANNOTATED | Summary: The same interferon assertion arriving by a second route, from the InterPro ZBP1 family entry whose description still characterises the protein as a cytosolic DNA sensor that activates the innate immune system. Reason: Same adjudication as the IBA row for this term. The mapping sits on InterPro:IPR042361, whose free-text description reproduces the 2007 DAI-as-DNA-sensor model. That model was not reproduced in human cells: multiple hZBP1 siRNAs failed to suppress IFN-beta responses to cytosolic DNA or intracellular bacteria, whereas mouse Zbp1 knockdown did impair the mouse response. Independently, the term itself describes regulation of signalling downstream of the interferon receptor rather than interferon production, so it would not be the right term even if the biology held. Retained but demoted rather than removed, since ZBP1 is interferon-inducible and can amplify IRF3-dependent reporters on overexpression. Supporting Evidence: PMID:18771559 multiple hZBP1 siRNAs did not suppress IFNbeta or IL-8 expression induced by poly(dA-dT) or bacterial infection in human cells |
| GO:0060545 positive regulation of necroptotic process | IEA GO_REF:0000117 | ACCEPT | Summary: Necroptosis promotion assigned by ARBA, agreeing with the experimental record. Reason: Core; redundant with better-evidenced annotations of the same term but not wrong. |
| GO:0005515 protein binding | IPI PMID:19590578 DAI/ZBP1 recruits RIP1 and RIP3 through RIP homotypic intera... | REMOVE | Summary: Bare protein binding standing in for what is arguably ZBP1's best-evidenced human molecular function, RHIM-mediated recruitment of the RIP kinases. Reason: Per project curation guidance, protein binding carries no functional information. It is especially unhelpful here, because the underlying paper is the one that identified the RHIMs in DAI and showed they recruit RIPK1 and RIPK3; that is a signalling adaptor activity and is proposed as a new annotation in this review. Note also that IntAct records the partners for this reference as UniProtKB:Q13601 and UniProtKB:Q9Y572. Q9Y572 is RIPK3 and matches the paper; Q13601 resolves to KRR1, a small-subunit processome component that the abstract does not discuss (RIPK1 is Q13546). Flagged for a curator to check rather than acted on here, since the full text has not been read. 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:19590578 we report the identification of two receptor-interacting protein (RIP) homotypic interaction motifs (RHIMs) in the DAI protein sequence, and show that these domains relay DAI-induced NF-kappaB signals through the recruitment of the RHIM-containing kinases RIP1 and RIP3 |
| GO:0005515 protein binding | IPI PMID:32814053 Interactome Mapping Provides a Network of Neurodegenerative ... | REMOVE | Summary: A hit from a large-scale neurodegeneration interactome screen. Reason: Bare protein binding, from a high-throughput interactome map rather than a focused characterisation. The recorded partner is UniProtKB:P42858 (huntingtin), with no described mechanism linking it to ZBP1 function. Uninformative as a molecular function annotation. 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. |
| GO:0140639 positive regulation of pyroptotic inflammatory response | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Pyroptotic arm of ZBP1-driven cell death, transferred by sequence similarity from mouse Zbp1. Reason: Biologically plausible and well documented for mouse Zbp1 in the PANoptosis literature, but this human annotation rests entirely on ISS from UniProtKB:Q9QY24 with no human-specific experimental support. The 2026 characterisation of human ZBP1 examined apoptosis and necroptosis and found signalling routed through RIPK1 rather than RIPK3, and explicitly cautions against transferring mouse mechanism to human. Retained because nothing contradicts it, but demoted from core until human evidence exists. Supporting Evidence: PMID:42436309 highlight the limitations of translating preclinical findings in animal models to human therapeutic strategies targeting this pathway |
| GO:0097527 necroptotic signaling pathway | IMP PMID:32200799 Influenza Virus Z-RNAs Induce ZBP1-Mediated Necroptosis. | ACCEPT | Summary: ZBP1 initiates the RIPK-dependent signalling cascade that executes necroptosis, shown by loss-of-function in the influenza Z-RNA study. Reason: Core, and the strongest process-level evidence code on the human record. The GO definition of this term centres on activation of RIPK1 and/or RIPK3, which is exactly what ZBP1's RHIMs accomplish; the 2026 human work refines which of the two kinases carries the signal in human cells without disturbing the term. Supporting Evidence: PMID:32200799 ZBP1 then initiates RIPK3-mediated MLKL activation in the nucleus, resulting in nuclear envelope disruption |
| GO:7770073 left-handed Z-RNA immune receptor activity | ISS GO_REF:0000024 | ACCEPT | Summary: The flagship molecular function, and the contested one. Retained: the disputed claim is that the Zalpha domains convert A-form RNA to Z-form, which is not what this term asserts. Reason: This annotation has to be read against the term's actual definition, which is combining with a left-handed Z-RNA and transmitting the signal to initiate an innate immune response, i.e. binding plus signal transduction, with no requirement that the receptor generate the Z form itself. The 2026 bioRxiv preprint by Krall et al. (a preprint, not peer reviewed, and weighed as one) makes a narrower claim than it first appears: that the long-assumed A-to-Z conversion activity was never tested and does not occur for unmodified RNA, while the same experiments show the Zalpha domains do bind and stabilise Z-form RNA once a modification such as 8-methyl- or 8-oxoguanine lowers the energetic barrier. The authors' own conclusion is that ZBP1 sensing is restricted to pre-Z-forming nucleic acids, which narrows the ligand set rather than abolishing the receptor function. On the other side, Yin et al. (Nature 2025) identify host-encoded Z-RNAs generated by virus-driven disruption of transcription termination as bona fide, sufficient ZBP1-activating ligands, and these fall squarely in the pre-Z-forming category; a Science Bulletin commentary (PMID:41580360) frames host Z-RNAs as the endogenous trigger. The two positions are therefore more compatible than the framing suggests, and UNDECIDED would overstate the disagreement. What is genuinely weak is the evidence code: for the function this gene is best known for, human GOA carries only ISS from mouse Zbp1 (UniProtKB:Q9QY24). Direct human evidence exists in the cited literature and this row should be upgraded to IDA/IMP. Supporting Evidence: PMID:41082924 host cell-encoded Z-RNAs are major and sufficient ZBP1-activating ligands after infection by these two human pathogens PMID:42079158 This functional difference likely restricts ZBP1 sensing to pre-Z-forming nucleic acids, preventing aberrant immune activation PMID:32200799 Here, we show that replicating IAV generates Z-RNAs, which activate ZBP1 in the nucleus of infected cells. |
| GO:0003723 RNA binding | IDA PMID:34192517 Vaccinia virus E3 prevents sensing of Z-RNA to block ZBP1-de... | ACCEPT | Summary: Direct demonstration of RNA binding in the vaccinia E3 competition study. Reason: Correct, and the only direct experimental code supporting any nucleic-acid binding by ZBP1 on the human record. Generic, but the competition with the viral Zalpha-containing E3 protein makes clear that the relevant ligand is Z-form RNA. Supporting Evidence: PMID:34192517 In the absence of this E3 domain, Z-form RNA accumulates during the early phase of VACV infection, triggering ZBP1 to recruit receptor interacting protein kinase (RIPK)3 and execute necroptosis. |
| GO:0005737 cytoplasm | IC PMID:34192517 Vaccinia virus E3 prevents sensing of Z-RNA to block ZBP1-de... | ACCEPT | Summary: Curator inference that ZBP1 acts in the cytoplasm, where it meets vaccinia Z-RNA. Reason: Sound inference: poxvirus replication and the E3/ZBP1 competition are cytoplasmic, so the site of action follows. |
| GO:0140374 antiviral innate immune response | IDA PMID:34192517 Vaccinia virus E3 prevents sensing of Z-RNA to block ZBP1-de... | ACCEPT | Summary: Direct evidence that ZBP1 mounts an antiviral response that vaccinia must actively suppress. Reason: Core, and the best-evidenced biological-process annotation on the human record alongside the influenza IMP. Supporting Evidence: PMID:34192517 Necroptosis mediated by Z-nucleic-acid-binding protein (ZBP)1 (also called DAI or DLM1) contributes to innate host defense against viruses by triggering cell death to eliminate infected cells. |
| GO:0050832 defense response to fungus | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Antifungal role transferred by sequence similarity from mouse Zbp1. Reason: Documented for mouse Zbp1 in fungal PANoptosis studies, but with no human experimental support; the human record carries this only as ISS from UniProtKB:Q9QY24. It is a context-specific deployment of the same cell-death machinery rather than a distinct core function, and given the reported human/mouse mechanistic divergence it should not be treated as established for the human protein. |
| GO:0051607 defense response to virus | ISS GO_REF:0000024 | ACCEPT | Summary: Antiviral defence transferred from mouse Zbp1. Reason: Core, and here the mouse transfer is corroborated rather than merely assumed: the same term is independently carried with IMP evidence from the influenza study and with direct evidence from the vaccinia study, both bearing on the human protein. |
| GO:0005515 protein binding | IPI PMID:33348174 Herpes simplex virus encoded ICP6 protein forms functional a... | REMOVE | Summary: RHIM-mediated heteromeric amyloid assembly with the herpes simplex virus protein ICP6, recorded as bare protein binding. Reason: 'The interaction itself is mechanistically meaningful: ICP6 (UniProtKB:P08543) is a viral RHIM protein that subverts host RHIM signalling, and the paper shows the RHIM core tetrad drives both amyloid formation and engagement of host RHIM proteins. But protein binding conveys none of that. The informative representation is ZBP1''s RHIM-dependent signalling adaptor activity, proposed as a new annotation in this review.'. 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:33348174 The core tetrad sequence of the ICP6 RHIM is important for both amyloid formation and interaction with host RHIM-containing proteins. |
| GO:0002218 activation of innate immune response | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Generic innate-immune activation, transferred from mouse Zbp1. Reason: True in substance for human ZBP1, whose RHIMs relay signals to NF-kappaB in human cells, but the term is high-level and the gene carries far more specific and better-evidenced annotations of the same biology. Kept as background rather than core. |
| GO:0005634 nucleus | IDA PMID:16876127 Intracellular localization of human ZBP1: Differential regul... | ACCEPT | Summary: Direct microscopy: ZBP1 forms nuclear foci overlapping PML bodies in interferon-treated and leptomycin B-treated cells. Reason: Correct. The nuclear pool is modest at steady state and is accentuated by blocking nuclear export, but it is real and, per the influenza work, functionally significant. Supporting Evidence: PMID:16876127 ZBP1 was distributed primarily in the cytoplasm and occasionally as nuclear foci in interferon (IFN)-treated primary hepatocellular carcinoma cells |
| GO:0005634 nucleus | IDA PMID:16990255 ZBP1 subcellular localization and association with stress gr... | ACCEPT | Summary: Localisation study of full-length ZBP1 and the Zalpha-deleted splice variant. Reason: Correct; consistent with the independent microscopy in PMID:16876127 and with the functional nuclear pool described during influenza infection. |
| GO:0005737 cytoplasm | IDA PMID:16876127 Intracellular localization of human ZBP1: Differential regul... | ACCEPT | Summary: Direct microscopy showing predominantly cytoplasmic ZBP1. Reason: Correct and core; the principal steady-state compartment. Supporting Evidence: PMID:16876127 ZBP1 was distributed primarily in the cytoplasm and occasionally as nuclear foci in interferon (IFN)-treated primary hepatocellular carcinoma cells |
| GO:0005737 cytoplasm | IDA PMID:16990255 ZBP1 subcellular localization and association with stress gr... | ACCEPT | Summary: Cytoplasmic distribution, punctate for full-length ZBP1 and granular for the Zalpha-deleted variant. Reason: Correct and core. This paper also establishes that the Zalpha domains control the cytoplasmic distribution pattern, which is why the isoform lacking exon 2 behaves differently. Supporting Evidence: PMID:16990255 While the full length protein showed a finely punctate cytoplasmatic distribution, ZBP1DeltaZalpha accumulated in large cytoplasmic granules. |
| GO:0043065 positive regulation of apoptotic process | ISS GO_REF:0000024 | ACCEPT | Summary: Apoptosis promotion transferred from mouse Zbp1. Reason: Unlike most of the mouse ISS transfers on this record, this one has been directly corroborated in human cells: the 2026 EMBO Reports study shows human ZBP1 drives apoptosis, requiring the RIPK1 scaffold. The mechanism differs from mouse; the process does not. Supporting Evidence: PMID:42436309 In contrast to mZBP1-activated signaling, RIPK1 plays a critical and universal role in transducing these downstream signaling cascades in response to hZBP1 activation across various human cell lines independently of RIPK3. |
| GO:0050727 regulation of inflammatory response | ISS GO_REF:0000024 | MODIFY | Summary: Direction-neutral parent of an annotation the same record already carries in its positive form. Reason: Uninformative as it stands. The direction of ZBP1's effect on inflammation is known and the gene is separately annotated to the positive child term, so the unsigned parent should be replaced by it. Proposed replacements: positive regulation of inflammatory response |
| GO:0050729 positive regulation of inflammatory response | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Pro-inflammatory role transferred from mouse Zbp1. Reason: Real but downstream. Inflammation follows from lytic ZBP1-driven cell death and from RHIM-dependent NF-kappaB activation rather than being a separate activity of the protein; the 2026 human study does report that hZBP1 induction upregulates NF-kappaB pathway genes. Kept as a secondary process. |
| GO:0051607 defense response to virus | IMP PMID:32200799 Influenza Virus Z-RNAs Induce ZBP1-Mediated Necroptosis. | ACCEPT | Summary: Loss-of-function evidence for ZBP1 in the antiviral response to influenza A. Reason: Core, and the strongest evidence code for this term on the human record. Supporting Evidence: PMID:32200799 Here, we show that replicating IAV generates Z-RNAs, which activate ZBP1 in the nucleus of infected cells. |
| GO:2000659 regulation of interleukin-1-mediated signaling pathway | ISS GO_REF:0000024 | MARK AS OVER ANNOTATED | Summary: Describes the wrong arm of interleukin-1 biology, and rests on mouse sequence similarity alone. Reason: GO:2000659 is regulation of the interleukin-1-mediated signaling pathway, i.e. of signal transduction downstream of the IL-1 receptor. What the mouse literature supports is that Zbp1 promotes inflammasome activation and thereby IL-1-beta maturation and release, which is production, not receptor signalling. There is no evidence that ZBP1 acts on cells receiving an IL-1 signal. Combined with the absence of any human experimental support, this is an over-annotation. Not removed outright, because the connection to IL-1 biology is genuine in mouse and a curator with the source paper may be able to re-point it to an IL-1-beta production term. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-1591234 | ACCEPT | Summary: Cytosolic localisation asserted by the Reactome ZBP1 cytosolic-DNA pathway. Reason: The pathway model these Reactome events belong to, ZBP1 as a cytosolic DNA sensor driving interferon induction, is the part of ZBP1 biology that did not replicate in human cells. The localisation claim is separable from that model and is independently supported by direct evidence, so the component annotation stands. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-1606327 | ACCEPT | Summary: Cytosolic localisation asserted by a Reactome IRF3 phosphorylation event. Reason: Correct as a location. As with the other Reactome rows, the surrounding interferon- induction pathway model is the contested element, not the compartment. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-1810457 | ACCEPT | Summary: Cytosolic localisation asserted by the Reactome ZBP1/RIP1/RIP3 recruitment event. Reason: Correct, and this particular Reactome event corresponds to the well-supported RHIM adaptor function rather than to the contested DNA-sensing arm. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-3249386 | ACCEPT | Summary: Cytosolic localisation asserted by a Reactome NLRP4/DTX4/TBK1 event. Reason: Correct as a location; redundant with the direct evidence for the same compartment. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-8948703 | ACCEPT | Summary: Cytosolic localisation asserted by a Reactome NLRP4/DTX4/TBK1 association event. Reason: Correct as a location; redundant with the direct evidence for the same compartment. |
| GO:0005829 cytosol | TAS Reactome:R-HSA-9679819 | ACCEPT | Summary: Cytosolic localisation asserted by a Reactome TBK1 inhibitor-binding event. Reason: Correct as a location, though this event is about amlexanox binding TBK1 and involves ZBP1 only as a pathway member. |
| GO:0003677 DNA binding | IDA PMID:17618271 DAI (DLM-1/ZBP1) is a cytosolic DNA sensor and an activator ... | MODIFY | Summary: Direct demonstration that DAI/ZBP1 binds DNA, in the paper that proposed it as a cytosolic DNA sensor. Reason: The binding observation is sound and is not being second-guessed here; what was measured was binding to double-stranded DNA, and the follow-up study showed that full activation requires three DNA-binding regions. GO:0003690 double-stranded DNA binding states that accurately. Note that the broader conclusion this paper drew, that ZBP1 is the cytosolic DNA sensor driving type I interferon induction, did not hold up in human cells; that is addressed on the GO:0060340 rows and does not bear on the binding activity itself. Proposed replacements: double-stranded DNA binding Supporting Evidence: PMID:17618271 DAI (DLM-1/ZBP1) binds to double-stranded DNA and, by doing so, enhances its association with the IRF3 transcription factor and the TBK1 serine/threonine kinase. PMID:18375758 We first show that DAI directly interacts with DNA in vitro and that it requires three DNA-binding domains for full activation in vivo |
| GO:0005829 cytosol | IDA PMID:17618271 DAI (DLM-1/ZBP1) is a cytosolic DNA sensor and an activator ... | ACCEPT | Summary: Direct evidence for cytosolic ZBP1. Reason: Correct and core; the most informative compartment term on the record and the site where ZBP1 assembles its signalling complexes. |
| GO:0005634 nucleus | IDA PMID:19158679 An orthogonal proteomic-genomic screen identifies AIM2 as a ... | ACCEPT | Summary: Nuclear localisation observed in the screen that identified AIM2 as the cytoplasmic DNA sensor for the inflammasome. Reason: ZBP1 localisation is incidental to that paper's main finding, but the annotation is experimental, is concordant with two dedicated localisation studies, and the full text has not been read here, so it is accepted rather than second-guessed. |
| GO:0005737 cytoplasm | IDA PMID:19158679 An orthogonal proteomic-genomic screen identifies AIM2 as a ... | ACCEPT | Summary: Cytoplasmic localisation observed in the same proteomic-genomic screen. Reason: Concordant with the dedicated localisation studies and with the UniProt record. |
| GO:0003692 left-handed Z-DNA binding | NAS PMID:11842111 Complex regulation of the human gene for the Z-DNA binding p... | ACCEPT | Summary: The function the gene is named for, supported in GOA only by a non-traceable author statement from the 2002 cloning paper. Reason: The assertion is correct, but NAS is close to the weakest code in the ontology and it is striking that it is one of only two rows supporting Z-DNA binding for human ZBP1. The 2002 paper identifies the domain by homology to the ADAR1 Zalpha domain rather than measuring binding. Direct measurement for the human protein followed in 2006 by circular dichroism and EMSA, and again in 2026 by NMR, CD and ITC. Accepted on the strength of that later work, with the recommendation that the evidence code be upgraded. Supporting Evidence: PMID:11842111 DLM-1 contains an N-terminal Z-DNA binding domain homologous to the Zalpha domain in the RNA editing enzyme ADAR1. PMID:16990255 both Zalpha domains can bind Z-DNA independently and that substrate binding is greatly enhanced when both domains are linked |
| GO:0035591 signaling adaptor activity | IPI PMID:19590578 DAI/ZBP1 recruits RIP1 and RIP3 through RIP homotypic intera... | NEW | Summary: Proposed new annotation capturing what ZBP1's three RHIMs actually do: bring RIPK1 and RIPK3 together into a death-signalling complex without any catalytic contribution of ZBP1's own. Reason: This is arguably the best-evidenced molecular function of the human protein, and GOA represents it only as bare protein binding. Rebsamen et al. identified the RHIMs in human DAI and showed they recruit RIPK1 and RIPK3 to relay signals to NF-kappaB; the 2026 EMBO Reports study shows that human ZBP1 requires all three RHIMs to trigger cell death and that the C-terminal RHIM2/RHIM3 region determines assembly and RIPK1 engagement. ZBP1 has no catalytic activity, which matches the GO definition of signaling adaptor activity: a molecule that brings together two or more molecules in a signaling pathway, itself without catalytic activity. Proposed with the caveat that the human and mouse proteins route the signal differently, so the adaptor's preferred partner is RIPK1 in human cells. Supporting Evidence: PMID:19590578 we report the identification of two receptor-interacting protein (RIP) homotypic interaction motifs (RHIMs) in the DAI protein sequence, and show that these domains relay DAI-induced NF-kappaB signals through the recruitment of the RHIM-containing kinases RIP1 and RIP3 PMID:42436309 Unlike mZBP1, which primarily uses only its RHIM1 domain, hZBP1 requires all three RHIM domains (RHIM1, RHIM2, and RHIM3) to trigger cell death. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Can human ZBP1 be activated by unmodified duplex RNA under physiological conditions, or does activation strictly require RNA that is already predisposed to the Z form by chemical modification, base damage such as 8-oxoguanine, or negative supercoiling?
Suggested experts: Jeffrey B Krall, Beat VΓΆgeli, Siddharth Balachandran
Q: Are the host Z-RNAs generated by disruption of transcription termination chemically modified or torsionally strained in a way that makes them competent ZBP1 ligands without requiring ZBP1-driven A-to-Z conversion?
Suggested experts: Chaoran Yin, Alan Herbert, Siddharth Balachandran
Q: Given that human ZBP1 signals through RIPK1 largely independently of RIPK3 and needs all three RHIMs, how many of the mouse-derived ISS annotations on the human record, in particular pyroptosis and antifungal defence, actually hold for the human protein?
Suggested experts: Fen Lu, Huifang Jiao
Q: Is there any human-cell context in which ZBP1 contributes non-redundantly to type I interferon induction, or is the interferon arm of the original DAI model best regarded as mouse-specific?
Suggested experts: Juliane Lippmann, Bastian Opitz
Q: Is Z-DNA, as distinct from Z-RNA, ever a physiological activating ligand for human ZBP1, or is Z-DNA binding a conserved biochemical property of the Zalpha fold with no separate signalling role?
Suggested experts: Nikolas Deigendesch, Stefan Rothenburg
Experiment: Reconstitute ZBP1 activation in human cells with defined synthetic duplex RNAs delivered at matched concentrations: unmodified (CpG)n repeats, the same sequence bearing 8-oxoguanine or 8-methylguanine at the C8 position, and a scrambled non-Z-forming control. Read out ZBP1 puncta formation, RIPK1 recruitment by co-immunoprecipitation, MLKL phosphorylation and cell death. A requirement for the modified ligand would confirm in cells what the biophysics predicts, and would justify tightening the GO:7770073 annotation to pre-formed Z-RNA.
Hypothesis: Human ZBP1 activation requires Z-RNA ligands that are already predisposed to the Z conformation, rather than ligands converted by ZBP1 itself.
Type: Defined-ligand reconstitution in human cells
Experiment: In human macrophage lines with ZBP1 knocked out and reconstituted with wild-type, Zalpha mutant, or individual RHIM mutant hZBP1, challenge with Candida albicans and with influenza A and measure gasdermin D cleavage, IL-1-beta release, caspase-1 activation and lytic death side by side with the same panel in mouse macrophages. This would test directly whether GO:0140639 and GO:0050832 are supportable for the human protein or should be retired from the human record.
Hypothesis: The mouse-derived ISS annotations for pyroptosis and antifungal defence do not transfer to human ZBP1.
Type: Cross-species knockout-reconstitution comparison
Experiment: Using CRISPR knockouts rather than the siRNA approach available in 2008, compare IFN-beta induction in human cells lacking ZBP1, cGAS, or both, after stimulation with cytosolic dsDNA, influenza infection and herpes simplex infection. Include a ZBP1 RHIM-mutant reconstitution to separate any interferon contribution from the cell-death output. A null result would settle the GO:0060340 annotations; a positive result would identify the context in which the original DAI model applies to human cells.
Hypothesis: Human ZBP1 does not contribute non-redundantly to type I interferon induction.
Type: CRISPR epistasis in human cells
Experiment: Determine the structure of the human ZBP1 RHIM assembly with RIPK1, by cryo-EM of the amyloid-like filaments formed on activation, and pair it with separation-of-function mutants in each of the three RHIM core tetrads. Quantifying which RHIM contacts which partner would provide the direct structural basis for a signaling adaptor activity annotation and explain why the human protein needs all three motifs while the mouse protein needs only one.
Hypothesis: ZBP1 acts as a non-catalytic signalling adaptor whose RHIM-dependent assembly is the rate-limiting step of activation.
Type: Cryo-EM plus structure-guided mutagenesis
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)