TRAF2 (TNF receptor-associated factor 2) is a cytoplasmic signaling adaptor of the TRAF family, built from an N-terminal RING domain, a series of TRAF-type zinc fingers, a coiled-coil (TRAF-N) region and a C-terminal MATH (TRAF-C) domain. The trimeric MATH domains bind the cytoplasmic tails of ligand-engaged TNF receptor superfamily members (such as TNFR2 and CD40) and adaptor proteins in the TNFR1 complex, while the coiled-coil region recruits the cIAP1/cIAP2 ubiquitin ligases. Through this recruitment TRAF2 couples receptor engagement to canonical NF-kappaB and JNK/MAP kinase activation and protects cells from TNF-induced apoptosis. In resting cells a TRAF2-TRAF3-cIAP assembly keeps the kinase NIK low and so restrains non-canonical NF-kappaB signaling. TRAF2 also binds the ER-stress sensor IRE1 and links it to ASK1-JNK activation. Whether the atypical TRAF2 RING has intrinsic ubiquitin ligase activity is disputed. Structural work finds no productive E2 interface, a sphingosine-1-phosphate cofactor model is not supported by Sphk1-deficient cells, and much TRAF2-associated ubiquitination is carried out by the recruited cIAPs. The horse protein F7BIV4 is a full-length member of the TRAF2 orthologous subfamily, in which all domains and key binding residues are conserved; its functions are inferred from human and mouse TRAF2.
Summary: Intrinsic TRAF2 ubiquitin-transfer activity is unresolved. A RING/E2 structural and biochemical study finds no productive E2 interface. A purified-protein study reports S1P-dependent RIP1 ubiquitination, but Sphk1 deficiency does not reproduce TRAF2 loss in TNF signaling in vivo. Much TRAF2-dependent ubiquitination is carried out by the cIAPs that TRAF2 recruits.
Reason: The PAINT curators placed GO:0004842 at the vertebrate TRAF2-clade node PTN002562154 (PTHR10131), seeded by human Q12933 and mouse Traf2 experimental annotations. F7BIV4 sits in the same PANTHER subfamily (PTHR10131:SF21) and keeps the RING Cys/His ligands, so placement is not the problem. The problem is the donor evidence itself. Yin et al. (PMID:19810754) report a RING that cannot engage Ubc13. Alvarez et al. (PMID:20577214) rescue activity only with an S1P cofactor. Etemadi et al. (PMID:26701909) show that TRAF2-dependent TNF signaling does not require Sphk1, and they summarise the field as holding that TRAF2 mainly acts by recruiting cIAPs. The human review keeps the same term UNDECIDED, and this row follows it.
Summary: The TRAF2 MATH domain binds the cytoplasmic tails of TNF receptor superfamily members. Recombinant human TRAF2 binds the CD40 cytoplasmic domain directly. In horse F7BIV4 the MATH domain is 142/146 identical to the human one.
Reason: Receptor binding is the defining activity of the TRAF-C/MATH domain. The PAINT curators also placed GO:0005164 at the eumetazoan node PTN000805720, which is ancestral to the TRAF2 subfamily (PTHR10131:SF21) that contains F7BIV4. The term means binding the intracellular part of the receptor, not binding the TNF cytokine.
Summary: TRAF2 is a soluble cytoplasmic adaptor that is recruited to the cytoplasmic face of ligand-engaged receptors and to IRE1 at the ER.
Reason: Cytoplasm is the right compartment. The PAINT curators placed GO:0005737 at the deep node PTN008938480 (PTHR10131), which includes the TRAF2 subfamily. Experimental human TRAF2 localisation (UniProt Q12933, PubMed:15383523 and PubMed:19150425) agrees. Cytosol, the more specific term, is also annotated and accepted. Cytoplasm is kept as well, because TRAF2 also acts at the cytoplasmic face of membranes.
Summary: TRAF2 has no transmembrane or signal sequence. It is a soluble cytosolic adaptor that is recruited from the cytosol to receptor complexes.
Reason: This InterPro-based term is consistent with human TRAF2, which has experimental and Reactome cytosol annotations accepted in the human review. Horse F7BIV4 keeps the same domain architecture.
Summary: The term is true but very broad. TRAF2 acts in TNF-family receptor signaling, which is captured by the accepted tumor necrosis factor-mediated signaling pathway (GO:0033209).
Reason: GO:0033209 tumor necrosis factor-mediated signaling pathway is accepted on this gene and is a descendant of GO:0007165 (through cell surface receptor signaling pathway; checked in OLS), so this term adds nothing. The claim itself is correct.
TRAF2 is required for TNF-alpha-mediated activation of c-Jun N-terminal kinase (JNK), contributes to activation of NF-kappaB, and mediates anti-apoptotic signals
Summary: The TRAF2 RING and TRAF-type zinc fingers coordinate zinc structurally. The crystal structure of the human RING and first zinc finger resolves these sites, and all 8 RING Cys/His positions are conserved in F7BIV4.
Reason: Zinc binding is real and is the precise metal-binding term. It is structural and does not by itself establish catalysis, so it is non-core to the adaptor function.
Summary: TRAF2 is a core component of TNFR1 and TNFR2 signaling complexes. It recruits cIAPs and is required in mice for full TNF-induced NF-kappaB and MAP kinase activation in fibroblasts and keratinocytes.
Reason: This is a core process. The PAINT curators placed GO:0033209 at node PTN000805720, which includes the TRAF2 subfamily (PTHR10131:SF21) that contains F7BIV4. TRAF2 takes part directly, as the scaffold that recruits the E3 ligases to the receptor complex.
TRAF2 is required for TNF-alpha-mediated activation of c-Jun N-terminal kinase (JNK), contributes to activation of NF-kappaB, and mediates anti-apoptotic signals
Summary: TRAF2 limits TNF-induced apoptosis. Traf2-deficient cells are more sensitive to TNF killing, and TRAF2-cIAP recruitment supports survival signaling. TRAF2 can also promote death in other contexts, such as ER-stress-driven ASK1-JNK signaling, so regulation in both directions applies.
Reason: Because TRAF2 regulates apoptosis in both directions depending on the signal, the general regulation term is the right level of precision. The anti-apoptotic role in TNF signaling is a major, conserved output of the receptor complex that TRAF2 organises.
TRAF2 is required for TNF-alpha-mediated activation of c-Jun N-terminal kinase (JNK), contributes to activation of NF-kappaB, and mediates anti-apoptotic signals
Summary: The evidence gives a direction. TRAF2 is required for TNF-induced JNK activation, and dominant-negative TRAF2 blocks IRE1-induced JNK activation.
Reason: The general regulation term is correct, but the evidence specifically supports positive regulation of the JNK cascade (GO:0046330), from both TNF receptors and IRE1. No evidence was found that TRAF2 negatively regulates JNK.
Summary: True but uninformative, because the more specific zinc ion binding (GO:0008270) is kept on this gene.
Reason: GO:0008270 zinc ion binding, kept as non-core, is a descendant of GO:0046872 (checked in OLS). The metals coordinated by the TRAF2 RING and zinc fingers are zinc.
Summary: Recombinant TRAF2 autoubiquitination has been reported only with an S1P cofactor. TRAF2 is also a known cIAP1 substrate, so ubiquitinated TRAF2 does not prove self-catalysis.
Reason: This term depends on intrinsic E3 activity, which is disputed (see GO:0004842). The S1P-cofactor model is not supported by Sphk1 genetics. The human review keeps the same term UNDECIDED.
Summary: This term comes from the EC 2.3.2.27 number that the PIRNR family rule assigns to the horse record. Whether TRAF2 is itself a ligase or only recruits the cIAPs is disputed.
Reason: The same considerations as for GO:0004842 apply. The EC number on the horse record is a family-level assignment and adds no independent evidence. The human review keeps the same term UNDECIDED.
Summary: TRAF2 links activated receptors to downstream effectors by recruiting cIAP1/2 through its coiled-coil region. This adaptor activity is the core molecular function of TRAF2.
Reason: The PAINT curators placed GO:0035591 (IBD) at the TRAF-family node PTN008301857. Human TRAF2 in the same subfamily (PTHR10131:SF21) inherits it as IBA, and the curators negated it (IRD) only on separate subclades. F7BIV4 carries no IBA rows in GOA. All three cIAP-binding residues are conserved in the horse protein.
cIAP-binding residues: human I285, V288 and E292 align to identical horse I300, V303 and E307.
GO:0043123 positive regulation of canonical NF-kappaB signal transduction
IBA GO_REF:0000033
NEW
Summary: TRAF2 promotes TNF-induced canonical NF-kappaB activation by recruiting cIAPs to the TNFR1 signaling complex. Traf2 loss delays NF-kappaB activation in fibroblasts and keratinocytes.
Reason: The PAINT curators placed GO:0043123 (IBD) at the vertebrate TRAF2-clade node PTN002562154, seeded by human and mouse TRAF2. F7BIV4 falls within this clade. TRAF2 takes part directly, as the receptor-complex scaffold for the E3 ligases. Note that TRAF2 also suppresses non-canonical NF-kappaB, which is a different pathway.
Summary: TRAF2 binds the cytoplasmic domain of the ER-stress sensor IRE1 and forms the IRE1-TRAF2-ASK1 complex that couples ER stress to JNK activation.
Reason: The PAINT curators placed GO:1990604 (IBD) at the vertebrate TRAF2-clade node PTN002562154, which contains F7BIV4. TRAF2 is a defining subunit of the complex. This is a non-core, stress-specific deployment of the adaptor.
Acts as a signaling adaptor in TNF-family receptor complexes. TRAF2 binds receptor cytoplasmic tails through its MATH domain and recruits the cIAP1/2 ubiquitin ligases through its coiled-coil region, coupling receptor engagement to canonical NF-kappaB and JNK activation and to protection from TNF-induced apoptosis.
Q: Does the TRAF2 RING have intrinsic E3 ligase activity in any physiological context (for example K48 ubiquitination of caspase-8, or NIK turnover), or is all TRAF2-dependent ubiquitination carried out by the recruited cIAPs?
Q: Is the 15-residue linker insertion in horse F7BIV4 (after human positions 260 and 265) supported by horse transcript data, or is it an artefact of the Ensembl gene model?
Suggested Experiments
Experiment: Reconstitute Traf2-null (or Traf2/Traf5 double-null) cells with RING-mutant TRAF2 that still binds cIAPs, and measure TNF-induced RIPK1 ubiquitination, JNK activation and NIK levels, to separate RING-dependent from recruitment-dependent outputs.
Type: genetic complementation
External Prediction Reviews
These computational predictions are reviewed separately from the GOA annotation set used for this review. The assessments below are from this project and do not constitute official GO annotations or endorsement by GO/UniProt. They are not included in the existing annotation review above.
Horse F7BIV4 is a full-length member of the TRAF2 orthologous subfamily, a cytoplasmic adaptor that binds TNF receptor superfamily tails (including CD40) through its MATH domain and recruits cIAP1/2 ubiquitin ligases. All five ProtNLM2 terms are supported by experimental evidence on human or mouse TRAF2 and are absent from the horse GOA set, which carries only TNF receptor binding (not an ancestor of CD40 receptor binding), signaling, apoptosis, JNK and disputed E3-activity terms.
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: TRAF2 knockdown suppresses TCR-dependent interleukin-2 production in the primary mammalian study. The selected horse protein retains the full human adaptor architecture at 92.0% paired identity, supporting conserved positive control of T-cell cytokine production. This inference does not require TRAF2 itself to catalyze ubiquitin transfer. The supported annotation is absent from the frozen horse GOA; model training membership is unknown.
Supporting Evidence:
PMID:15125833: "RNAi-mediated silencing of MALT1, TAK1, TRAF6, and TRAF2 suppressed TCR-dependent IKK activation and interleukin-2 production in T cells."
file:HORSE/TRAF2/TRAF2-bioinformatics/RESULTS.md: "The downloaded human Q12933 sequence (501 residues) and selected horse F7BIV4 sequence (516 residues) share 92.0% identity among 501 paired residues. Paired coverage is 100.0% of human and 97.1% of horse."
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: Recombinant human TRAF2 directly binds the CD40 cytoplasmic domain. Full coverage of the characterized human sequence, including the C-terminal receptor-binding TRAF region, supports the same molecular interaction for horse F7BIV4. The supported annotation is absent from the frozen horse GOA; model training membership is unknown.
Supporting Evidence:
PMID:9718306: "Recombinant human TRAF proteins overexpressed in insect cells were biochemically characterized and used to finely map TRAF binding regions in the human CD40 cytoplasmic domain. TRAF1, TRAF2, TRAF3, and TRAF6, but not TRAF4 or TRAF5, bound directly to the CD40 cytoplasmic domain."
file:HORSE/TRAF2/TRAF2-bioinformatics/RESULTS.md: "The downloaded human Q12933 sequence (501 residues) and selected horse F7BIV4 sequence (516 residues) share 92.0% identity among 501 paired residues. Paired coverage is 100.0% of human and 97.1% of horse."
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: A reconstituted TRAF2βcIAP1 complex establishes recruitment of a genuine E3 ligase by TRAF2. The conserved horse scaffold supports membership in a ubiquitin ligase complex; this annotation does not assert autonomous catalytic activity of TRAF2. The supported annotation is absent from the frozen horse GOA; model training membership is unknown.
Supporting Evidence:
PMID:20447407: "To fulfill this role, cIAPs must be recruited to the receptor complex by TNF-receptor-associated factor (TRAF) 2. In this study, we reconstituted the complex between baculoviral IAP repeat (BIR) 1 of cIAP1 and the coiled-coil region of TRAF2, solved the structure of BIR1 from cIAP1, and mapped key binding residues on each molecule using mutagenesis."
file:HORSE/TRAF2/TRAF2-bioinformatics/RESULTS.md: "The downloaded human Q12933 sequence (501 residues) and selected horse F7BIV4 sequence (516 residues) share 92.0% identity among 501 paired residues. Paired coverage is 100.0% of human and 97.1% of horse."
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: TRAF2 is the receptor-associated adaptor that recruits cIAP1/2 into TNF receptor superfamily signaling assemblies. The conserved horse architecture supports complex membership independently of the disputed intrinsic E3 mechanism. The supported annotation is absent from the frozen horse GOA; model training membership is unknown.
Supporting Evidence:
PMID:20447407: "To fulfill this role, cIAPs must be recruited to the receptor complex by TNF-receptor-associated factor (TRAF) 2. In this study, we reconstituted the complex between baculoviral IAP repeat (BIR) 1 of cIAP1 and the coiled-coil region of TRAF2, solved the structure of BIR1 from cIAP1, and mapped key binding residues on each molecule using mutagenesis."
file:HORSE/TRAF2/TRAF2-bioinformatics/RESULTS.md: "The downloaded human Q12933 sequence (501 residues) and selected horse F7BIV4 sequence (516 residues) share 92.0% identity among 501 paired residues. Paired coverage is 100.0% of human and 97.1% of horse."
GO:0002637 regulation of immunoglobulin productionGO_BP
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: TRAF2-deficient mouse B cell lines are defective in CD40-mediated, TNF-dependent IgM production, and TRAF2-deficient mice have defective T helper-dependent antibody responses; mouse Traf2 carries this term by IMP. Horse F7BIV4 is a full-length member of the same TRAF2 orthologous subfamily with the receptor-binding MATH domain conserved, which justifies transfer. The horse gene has no immunoglobulin-production annotation in GOA. The role is a downstream, non-core consequence of TRAF2's CD40 adaptor function.
Supporting Evidence:
PMID:12958312: "We demonstrate that TRAF2 participates in synergy between CD40 and B cell antigen receptor signals, and in CD40-mediated, TNF-dependent IgM production."
PMID:10514016: "in the absence of TRAF2, the T helper-dependent antibody response, CD40-mediated proliferation, and NF-kappaB activation are defective."
PMID:9718306: "Recombinant human TRAF proteins overexpressed in insect cells were biochemically characterized and used to finely map TRAF binding regions in the human CD40 cytoplasmic domain. TRAF1, TRAF2, TRAF3, and TRAF6, but not TRAF4 or TRAF5, bound directly to the CD40 cytoplasmic domain."
file:HORSE/TRAF2/TRAF2-bioinformatics/RESULTS.md: "The downloaded human Q12933 sequence (501 residues) and selected horse F7BIV4 sequence (516 residues) share 92.0% identity among 501 paired residues. Paired coverage is 100.0% of human and 97.1% of horse."