EIF2B2

Summary: The cited ovarian-failure screen does not demonstrate follicle-development activity.

Reason: PMID:12707859 reports ovarian failure and white-matter disease in patients 944 and 992 with EIF2B2 variants; this is positive human disease evidence. The original source PMID:15507143 then screened 93 women with isolated premature ovarian failure for selected known EIF2B2/EIF2B4/EIF2B5 variants and found none. Neither result establishes which step of follicle formation or maturation depends on EIF2B2. The earlier paper explicitly leaves decreased follicle formation, increased loss, altered recruitment and interrupted maturation as alternatives. The GO:0001541 assertion therefore overstates the source-specific developmental resolution; this does not deny EIF2B2-associated ovarian disease or generalize the negative screen to all variants.

Summary: EIF2B2 contributes to cytoplasmic translation initiation.

Reason: The shared independently inspected PMID:27023709 Methods 12 and Results 23/Figure 2h include a human eIF2/eIF2B comparator in reconstituted HCV-IRES translation. The detailed recombinant and structural work is chiefly S. pombe, but the human comparator directly supports complex participation in cytoplasmic initiation. Beta is a regulatory/structural component, not an independently assayed exchange enzyme. Human subunit reconstitution in PMID:29599213 corroborates this complex role.

Summary: Beta contributes to the initiation-factor activity of eIF2B.

Reason: PMID:16289705 reports that eIF2/eIF2B supplementation increases translation in HeLa extracts. The inspected abstract does not resolve the species or preparation of the added factors, and full operative Methods remain unavailable. The retained contributes_to assertion is independently supported by the direct human-complex translation/reconstitution evidence in PMID:27023709 and PMID:29599213; no isolated-beta activity or factor species is inferred from the HeLa host.

Summary: The ancestral beta lineage contributes to eIF2B nucleotide exchange.

Reason: The literal GO:0005085 IBD at PTN000027996, dated 20190208, carries contributes_to and lies on the exact human P49770 leaf path in PTHR45859. The aligned human sequence matches all 351 official residues. The stated human, rat and budding-yeast descendants include current experimental grounding; the target appearing among the sources is legitimate self-grounding, not circularity. The annotation describes contribution within eIF2B, not autonomous beta catalysis.

Summary: Beta is a component of the active human GDP-release factor.

Reason: The inspected PMID:11323413 Methods purify native eIF2B from HeLa-cell cytoplasm and measure GDP release by the whole complex. Beta belongs to this active preparation. Its separate nucleotide-binding assay does not establish beta-only exchange or GTP hydrolysis. The imported contributes_to qualifier accurately states the subunit-level contribution; patient EIF2B2 variants independently reduce complex exchange in PMID:15054402.

Summary: The exchange activity is established for the beta-containing complex.

Reason: The experimentally attributed UniProt P49770 SUBUNIT annotation places alpha/beta/delta in the regulatory subcomplex and gamma/epsilon in the catalytic subcomplex, citing PMID:25858979, PMID:27023709 and PMID:31048492. The independent PMID:25858979 advisory identifies whole-complex GDP-release experiments and CHO Eif2b4 delta-mutant experiments, not isolated-beta catalysis. Supplementary Methods and the Table S1 exception to unchanged coding sequences were unavailable. Preserve the curator-supported term and imported enables qualifier without silently editing the source; the functional synthesis must explicitly use contribution to complex GEF activity. Direct human beta-variant and reconstitution evidence supplies independent corroboration. The preserved source relation is not thereby endorsed: a probable WRONG_QUALIFIER query asks UniProt to check contributes_to. This is a relation-level question based on the complex contribution, not a requirement for isolated protein activity or a rejection of the GEF function.

Summary: Beta participates in the exchange-factor machinery of eIF2B.

Reason: The experimentally attributed UniProt P49770 SUBUNIT annotation places alpha/beta/delta in the regulatory subcomplex and gamma/epsilon in the catalytic subcomplex, citing PMID:25858979, PMID:27023709 and PMID:31048492. PMID:27023709 chiefly investigates S. pombe subunits and includes a human whole-complex translation comparator. This comparator is not a beta-only GDP-release assay. Retain the established complex-level GEF interpretation, corroborated by the direct human experiments in PMID:15054402 and PMID:29599213, while preserving the imported enables qualifier and avoiding an autonomous-beta enzymatic claim. The preserved source relation is not thereby endorsed: a probable WRONG_QUALIFIER query asks UniProt to check contributes_to. This is a relation-level question based on the complex contribution, not a requirement for isolated protein activity or a rejection of the GEF function.

Summary: Human EIF2B2 variants reduce complex-level nucleotide exchange.

Reason: The genuine author-deposited PMID:15054402 original Methods and Table 1 assay extracts of human EBV-immortalized lymphoblasts using rat-liver eIF2 as the GDP-bound substrate. EIF2B2 E213G homozygotes and E213G/K273R compound heterozygotes show reduced exchange relative to controls. This supports the preserved contributes_to annotation; neither substrate species nor whole-cell extracts establish isolated-beta GEF activity. No figure-pixel or sequenced experimental-insert claim is made.

Summary: Resolve the binding assertion to a translation-initiation-factor partner.

Reason: The original source identifies human EIF2B4/delta within eIF2B assemblies. The inspected PMID:15060152 Results and Figure 1 explicitly report no stable isolated beta association with an individual subunit under that assay, but recover beta and delta together in four/five-subunit complexes. GO:0031369 fits physical association with the independently established initiation-factor partner without asserting purified binary affinity, a particular interface or autonomous delta catalysis. Later source-specific beta–delta positives corroborate the partner identity, but do not change the negative isolated-pair result in this paper.

Proposed replacements: translation initiation factor binding

Summary: Retain the curated C9orf72 association without a speculative functional class.

Reason: PMID:21516116 uses human ORFeome constructs, removes autoactivators and sequence-confirms pairwise-retested Y2H positives. The original target-specific Table S2 record was not independently recovered and the current IntAct target query has no record for this paper; neither absence demonstrates a false imported IPI. The same P49770–Q96LT7 pair has explicit later validated Y2H records in PMID:32296183. C9orf72 has GEF/GAP-complex functions, which do not by themselves establish a catalytic-enzyme-binding class or transfer autophagy function to EIF2B2. Retain supported generic association as non-core under the project instruction.

Summary: The tested EIF2B4 partner is a translation initiation factor.

Reason: The actual PMID:25910212 human ORFeome Methods and current IntAct IM-25511-1761/1133/505 records identify P49770 and Q9UI10. The first two records list delta-side substitutions at 228, 306 and 391; these are not beta mutations. The study compares mutant and wild-type interaction profiles and removes autoactivators. A positive partner-class refinement to GO:0031369 is justified without assigning catalytic activity to either subunit or claiming every pair was orthogonally validated.

Proposed replacements: translation initiation factor binding

Summary: Human subunit reconstitution supports initiation-factor association.

Reason: The independently inspected PMID:29599213 Methods express five codon-optimized human subunits in E. coli and assemble alpha dimers with beta/gamma/delta/epsilon material for exchange assays. This supports beta association with the EIF2B4 initiation-factor partner within the reconstituted complex. The current IntAct query does not contain this paper, and the exact synthesized sequences in the supplement were not inspected; no isolated binary affinity, natural-product equivalence or autonomous beta GEF activity is inferred.

Proposed replacements: translation initiation factor binding

Summary: Refine the tested beta–delta interaction to initiation-factor binding.

Reason: PMID:31515488 Methods describe sequence-verified human wild-type/mutant clones, autoactivation controls and repeated pairwise Y2H assays. The literal IM-27438-1304 record joins P49770 to Q9UI10, with beta position 127 recorded as a mutation with no effect. GO:0031369 describes the verified partner class; the reference-sequence checksum is not proof of the precise assayed insert and no target-specific PCA result is assumed.

Proposed replacements: translation initiation factor binding

Summary: The C9orf72 pair is supported, but no narrower binding class is established.

Reason: PMID:32296183 reports quadruplicate pairwise retesting and sequence confirmation. Five current P49770–Q96LT7 records include validated two-hybrid IM-25472-127836. Their terminal AD/DB fusion annotations describe tags, not protein truncations. The curated participant checksums match current 351-aa EIF2B2 and 481-aa C9orf72 reference sequences, without proving insert identity. C9orf72 GEF/GAP and trafficking functions do not warrant an enzyme-binding or scaffolding term for this association. Retain generic binding as non-core under the project instruction.

Summary: The validated EIF2B4 association has an informative partner class.

Reason: Three current PMID:32296183 human beta–delta records include validated two-hybrid IM-25472-123370. The source uses sequence-confirmed, quadruplicate-retested human pairs; current partner Q9UI10 is independently classified as a regulatory subunit of translation initiation factor eIF2B. GO:0031369 captures that association without claiming a specific domain, direct catalytic contact or a new physiological pathway.

Proposed replacements: translation initiation factor binding

Summary: The affinity-purified EIF2B4 partner is an initiation factor.

Reason: The two current PMID:33961781 records IM-29278-4587/25409 identify human P49770–Q9UI10 association by anti-tag coimmunoprecipitation, with a beta C-terminal tandem tag. The inspected original Methods use sequence-verified human ORFs, replicate LC-MS and background-aware CompPASS-Plus filtering. Physical complex association plus the verified EIF2B4 class supports GO:0031369; AP-MS does not establish a purified binary interface. The two curated records are not treated as two independently characterized physiological mechanisms.

Proposed replacements: translation initiation factor binding

Summary: ATP binding is observed, with an unresolved physiological role.

Reason: The inspected PMID:11323413 native human eIF2B experiments cross-link radiolabeled ATP to the beta-position band, with unlabeled nucleotide competition and no-UV controls. These experiments support ATP binding, but do not establish a physiological ATP-dependent reaction or an obligatory nucleotide-bound exchange intermediate. The physiological role remains unresolved, so retain this positive binding observation as non-core. The separate recombinant GTP-binding experiment is assessed under GO:0005525.

Summary: GTP binding is observed, with an unresolved physiological role.

Reason: The inspected PMID:11323413 native human eIF2B experiments cross-link radiolabeled GTP to the beta-position band, with unlabeled nucleotide competition and no-UV controls. A separate recombinant His-beta experiment confirms direct GTP binding; the historical recombinant ORF species and exact sequence were not specified in the inspected cloning passage. These observations support nucleotide binding but not beta GTPase activity or an obligatory nucleotide-bound exchange intermediate. The paper itself separates this binding from the catalytic exchange mechanism, so retain it as non-core.

Summary: The broad cytoplasm assertion can be refined to independently supported cytosol.

Reason: PMID:11323413 Methods recover active native beta-containing eIF2B from HeLa-cell cytoplasm by fractionation; that experiment itself resolves cytoplasm, not a narrower compartment or a microscopy localization. Independent curated Reactome paths from R-HSA-72670 and R-HSA-72722 reach human EIF2B2 entity 72519, UniProt P49770 residues 1–351, explicitly in GO:0005829 cytosol. The preserved UniProt subcellular annotation also specifies cytosol, with its homology attribution kept distinct from the human pathway model. Cytosol is part of cytoplasm, so MODIFY refines a correct but broad location using this combined evidence. The imported GO:0005737 term, IDA source and located_in relation remain unchanged; no claim is made that the original fractionation alone established cytosol or excluded other pools.

Proposed replacements: cytosol

Summary: The location mapping correctly identifies cytosol.

Reason: The source maps UniProt SL-0091 to GO:0005829. The beta-containing exchange-factor machinery acts in cytoplasmic translation, and the two inspected Reactome event paths reach exact human P49770 residues 1–351 at the cytosol entity. This is consistent with the native HeLa cytoplasmic preparation; it does not imply exclusion from every other cellular compartment.

Summary: The curated eIF2B event places beta in cytosol.

Reason: Reactome R-HSA-72670 reaches the eIF2B complex 72526 and EIF2B2 entity 72519 through literal component/catalyst paths. The target entity maps to UniProt P49770, residues 1–351, and GO:0005829. This is curated pathway localization, not a new direct imaging experiment or an isolated-beta catalytic assay.

Summary: The curated eIF2B event places beta in cytosol.

Reason: Reactome R-HSA-72722 reaches the eIF2B complex 72526 and EIF2B2 entity 72519 through literal component/catalyst paths. The target entity maps to UniProt P49770, residues 1–351, and GO:0005829. This is curated pathway localization, not a new direct imaging experiment or an isolated-beta catalytic assay.

Summary: Conserved beta membership in the eIF2B complex is well supported.

Reason: The GO:0005851 IBD at PTN000027996 is an ancestor of the exact P49770 leaf. The recovered tree/MSA and literal rat, budding-yeast, fission-yeast and human source identities support inheritance of beta-complex membership. Current donor records include direct experimental annotations; target self-grounding is expected and does not create circular evidence. Direct native-human and reconstituted-human complex observations corroborate the assertion.

Summary: Beta is present in native human eIF2B.

Reason: PMID:11323413 Methods and Results identify beta among the subunits of native HeLa eIF2B purified for GDP-release measurements. The subunit-specific nucleotide-binding result is additional evidence, not a substitute for the demonstrated complex membership.

Summary: Human beta participates in assembled eIF2B complexes.

Reason: The recovered PMID:15060152 cloning, Results and Figure 1 describe tagged human subunits and stable beta/gamma/delta/epsilon and five-subunit assemblies in HEK293 cells. Mutant and wild-type assembly comparisons directly support beta membership. Failure of isolated beta to bind individual subunits in that assay is distinct from its positive membership in the assembled complex.

Summary: The structural and human-comparator evidence supports eIF2B membership.

Reason: PMID:27023709 principally resolves S. pombe complex architecture, but its inspected Methods/Results also assay a human eIF2B comparator. Independently verified native human membership and human subunit reconstitution support the existing annotation; do not relabel the yeast mutant structures as human experiments.

Summary: The exact ARBA predicate supports the beta complex assignment.

Reason: ARBA00088579 condition set 6 requires FunFam 3.40.50.10470:FF:000004 and Craniata. Both are present in the saved official human record/lineage. The rule conclusion is eIF2B-complex membership, independently supported by native and reconstituted human evidence. This assessment uses the literal predicate rather than inferring rule conditions from its name.

Summary: Human beta is part of the reconstituted eIF2B complex.

Reason: The independently inspected PMID:29599213 Methods and Results assemble human beta with gamma, delta and epsilon, with alpha-dependent oligomerization and exchange measurements. This directly supports beta membership in eIF2B. Bacterial expression host and the uninspected synthetic-sequence supplement do not establish a different human product or autonomous beta catalysis.

Summary: Resolve the conserved initiation role to cytoplasmic translation.

Reason: The literal GO:0006413 IBD at PTN000027996 descends to exact human P49770 and cites the human and fission-yeast beta sources. Current Q9UT76 carries direct cytoplasmic-initiation evidence from PMID:27023709, and the same paper includes a human reconstituted translation comparator. GO:0002183 is a verified descendant of GO:0006413 and specifies the supported cytoplasmic machinery without introducing a new process. The ancestral placement is not rejected for donor count or target self appearance.

Proposed replacements: cytoplasmic translational initiation

Summary: The supported initiation occurs in the cytoplasm.

Reason: PMID:16289705 reports factor-supplemented translation in a HeLa-extract system. Exact added-factor preparation details remain unavailable in the abstract-only operative cache, but the cytoplasmic eIF2B role is independently established by human reconstituted translation in PMID:27023709. GO:0002183 is the verified more specific term; this does not assign species to an unreported preparation.

Proposed replacements: cytoplasmic translational initiation

Summary: The beta-containing exchange factor regulates initiation.

Reason: PMID:15060152 directly tests human beta variants in complex assembly and exchange assays and reports altered ATF4-UTR reporter translation. These data support regulation of initiation as part of the exchange-factor mechanism. The response of an ATF4 translational reporter is not a direct transcription assay or evidence for an independent beta signaling enzyme.

Summary: The family predicate does not establish a CNS-development mechanism.

Reason: ARBA00026561 condition set 56 matches the target FunFam 3.40.50.10470:FF:000004. This authentic beta-family match establishes neither the tissue context nor the developmental progression defined by GO:0007417. Human EIF2B2 disease variants and white-matter abnormalities establish neurological importance, but general translation machinery and clinical pathology alone are insufficient to mechanistically propagate a CNS-development role. Mark this electronic developmental assertion as over-annotated without denying the human disease association.

Summary: The source-specific CNS-development assertion remains unresolved.

Reason: The authentic PMID:15776425 abstract identifies an EIF2B2 Pro196Ser variant among patients with eIF2B-related white-matter disorders. Publisher and PDF routes were blocked and full operative genotype/phenotype analyses were not recovered. This supports clinical relevance but does not resolve the source-specific developmental-process assertion. Preserve the experimental source rather than infer a wrong gene or replace unknown observations with a mechanism.

Summary: Heat-responsive translation is retained as a conserved complex context.

Reason: The literal ISS donor is rat Eif2b2 Q62818, independently joined to RGD:620820. Its current experimental GO:0009408 record cites PMID:9341116; the authentic abstract measures heat-dependent eIF2B activity and translation in rat hepatoma cells. Retain the conserved response as contextual complex function, not a beta-specific temperature sensor. Full donor Methods were unavailable, and no direct human heat assay or unique beta modification is claimed.

Summary: The clinical citation does not yet resolve the human heat-response assertion.

Reason: PMID:12499492 is an authentic severe infantile VWM report whose abstract foregrounds an EIF2B5 variant and metabolic abnormalities. The full operative text was not recovered. This does not establish beta misattribution, but the source-specific heat-response evidence remains inaccessible; preserve it as undecided. The separate rat-ortholog heat-response row is evaluated on its own evidence.

Summary: Glucose-responsive translation is a contextual complex response.

Reason: Rat beta Q62818 has the current experimental GO:0009749 assertion from PMID:8567668. Its authentic abstract reports glucose-dependent activation of eIF2B in isolated rat islets, with a nonmetabolizable glucose-analogue control and no change in eIF2-alpha phosphorylation. Retain the curated conserved complex response as non-core; this does not establish a beta glucose-binding sensor, a beta-specific modification or a directly tested human islet response. Full operative donor Methods remain unavailable.

Summary: Human beta-associated oligodendrocyte pathology is real; the developmental step is unresolved.

Reason: The recovered PMID:15217090 Methods, Results and Discussion examine five VWM autopsies; patient 5 is explicitly EIF2B2-mutant, whereas patients 1–4 have EIF2B5 mutations. The beta-associated infant has prominent oligodendrocyte apoptosis, and the study measures cell density, survival and proliferation. These are positive observations of beta-associated glial pathology. GO:0014003 concerns progression from commitment to a fully differentiated oligodendrocyte. The selected original evidence leaves unresolved whether EIF2B2 contributes to that developmental progression, altered survival, or both; the paper discusses the unresolved mechanism. UNDECIDED records this scientific uncertainty and preserves the curator assertion. It is not based solely on the abstract-only normal cache, and it does not require an isolated biochemical differentiation activity.

Summary: Myelination is retained as a non-core human disease context.

Reason: The full PMID:14566705 original identifies patients 1–3 with compound-heterozygous EIF2B2 G200V/P291S variants in Table 1. MRI text and Figure 1 caption document insufficient myelination, including the internal capsule in patient 3, alongside progressive white-matter abnormalities. Retain this curated IMP context of beta-dependent translation machinery without assigning beta a direct myelin-synthesis reaction or claiming that imaging alone establishes the cellular mechanism. Text/captions were inspected, not image pixels.

Summary: Peptide-hormone response is a context of regulated translation.

Reason: The exact rat Q62818 donor has a current experimental GO:0043434 record from PMID:8430778. The authentic abstract measures tissue-dependent eIF2B activity in diabetic rats and biochemical regulation of partially purified factor. Retain the curated ortholog response as a contextual complex function, without inventing a beta peptide-hormone receptor, a beta-specific insulin mechanism or a human assay. Full donor Methods and hormone-treatment details were not independently recovered.

Summary: Assembly of active human eIF2B promotes initiation.

Reason: The independently inspected PMID:29599213 human subunit reconstitution and GDP-exchange assays show assembly-dependent activation of the eIF2B machinery. Beta forms part of the regulatory assembly needed for productive factor function, which promotes regeneration of initiation-competent eIF2. Retain this integrated positive-initiation role, without treating beta alone as a GEF or claiming that the alpha-addition experiment selectively perturbed beta.

Summary: A beta-family match does not independently establish animal-organ development.

Reason: ARBA00029247 condition set 241 matches the genuine target FunFam 3.40.50.10470:FF:000004. The conclusion GO:0048513 concerns formation and maturation of animal organs, whereas the directly established role is translation initiation and its regulation. Severe multi-organ disease in EIF2B2 variant carriers does not by itself identify beta participation in an organ-building step. Mark this electronic assertion as over-annotated while preserving the authentic family match and the clinical observations.

Summary: Mitogen-responsive eIF2B does not yet resolve receptor-initiated signaling participation.

Reason: The recovered PMID:8626696 Results and Discussion demonstrate rapid whole-eIF2B exchange activation in primary human T cells after ionomycin/PMA, with PHA responsiveness also reported. Later increases in beta protein and mRNA are distinct from the early activation of existing factor. These are positive translational responses during T-cell activation. GO:0050852 concerns the molecular signals initiated by antigen-receptor cross-linking. The available observations do not resolve whether beta performs part of that signaling process or is regulated downstream as translation increases. UNDECIDED preserves this source-specific scientific uncertainty without asserting absence of a TCR role. Selected Results/Discussion were read; complete operative Methods remain unavailable, but that access limit is not the sole basis for the decision.