This guide records the evidence used for the 58 imported assertions and the integrated nucleotide-exchange mechanism. Publication-cache flags describe the machine cache; the access statements below describe what was actually inspected. External original sections can therefore be available even when the normal cache is abstract-only. Article bodies are not redistributed in the portable database archive.
The source-locator inventory supplies normal-cache hashes, public routes, retained snapshot hashes and precise text scopes. Independent source consultations are identified as such; they are not represented as duplicate original reads. No figure pixels were inspected for these assessments.
The reproducer checks every imported GOA context, the 721-residue official sequence, the PAINT lineage and four positive IBD records, exact ARBA predicates, complete saved donor queries, literal IntAct records and human Reactome paths. The HPA parser preserves the location-specific enhanced Cytosol score and all six antibody/cell-line records. Source absence from a current interaction query does not invalidate a historical curated association.
GO_REF references identify methods and have LOW relevance individually; they do not substitute for target evidence. Physical association and independently verified partner class support the specific binding refinements. No partner activity is attributed to epsilon.
Engineered epsilon534–547 peptide with substitutions/phosphoserine544 associates with GSK3B in kinase assays; partner class is independently verified.
Inspected: Normal abstract, official metadata, both source-specific human IntAct kinase-assay records and explicit target/partner feature fields.
Limit: Publisher PDF HTTP403; no intact epsilon/native-complex claim.
Source: https://pubmed.ncbi.nlm.nih.gov/10481074/; cache publications/PMID_10481074.md, SHA256 e01fe6e5e61b5e5f8bc7d33a8dddf8dc6fd88b1ffc35da5202aeb606399aceac.
Native human five-subunit complex depleted of eIF2 supports exchange and cytoplasmic complex membership.
Inspected: Author-uploaded original, selected factor-preparation, chromatography, subunit-analysis and GDP-dissociation Methods/Results (captured text lines 468–721).
Limit: Not a full-paper or isolated-epsilon preparation claim.
Source: https://pubmed.ncbi.nlm.nih.gov/11323413/; cache publications/PMID_11323413.md, SHA256 ec823995789c31c53a1f798bf81023d5d4e493a969308461b1b15e80f3bc7337.
Native human phosphorylation is separate from detailed recombinant rat epsilon mutants. Rat S712/713GA changes eIF2 association/complex exchange; free epsilon activity is very low relative to assembled complex.
Inspected: Original Introduction (including the insulin-regulation background), Results and Figure 8 on GSK3 docking/phosphorylation of full-length and C-terminally truncated rat epsilon, and Discussion on the in-cis docking region. Saved paragraph indices 4–6,10,16,39–41,43–49,54–58,61,63,65,67,69. Figure 8 uses GSK3 alpha/beta immunoprecipitation from HEK293 cells; the epsilon construct is rat. The insulin-specific narrative cites prior work and is not presented as a new insulin experiment in this paper.
Limit: Rat residue numbers not mapped to human without alignment; no complete supplemental vector inspection.
Source: https://pubmed.ncbi.nlm.nih.gov/11500362/; cache publications/PMID_11500362.md, SHA256 f2a620cd6154034a94b551809898986a63d0f731559eaa98b11baf9b10ef3532.
Rat epsilon S535A/S539A constructs in Rat1/PC12 preserve translation and resist GSK3/PI3K-inhibitor/serum-withdrawal apoptosis. Phosphorylation inhibits pro-translation activity; no active WT epsilon proapoptotic step established.
Inspected: Saved PMC indexed Methods10/17–20, Results23–45/47, Discussion49–54; exact causal reread included resistant-mutant, tau and cycloheximide controls.
Limit: No universal antiapoptotic function or opposite NEW. Figure captions/text read, not pixels. MAO is role scope, not a conclusion from WT overexpression alone.
Source: https://pubmed.ncbi.nlm.nih.gov/11756553/; cache publications/PMID_11756553.md, SHA256 aec5cb3341f3305a0fb5a7fcddd61377dd81eccdbf85b5e78b8d9751219fc529.
Endogenous mouse NIH3T3 epsilon shows nuclear and cytoplasmic staining and stress-granule redistribution; current Q8CHW4 IDA donor is genuine.
Inspected: Publisher-indexed original Results, Figures 1–2 captions and Discussion, including the fixation comparison. Direct publisher requests failed; complete Methods and figure pixels were not inspected.
Limit: Positive mouse localization is retained; fixation affects relative staining, and this does not establish exclusive human nuclear residence.
Source: https://pubmed.ncbi.nlm.nih.gov/12388085/; cache publications/PMID_12388085.md, SHA256 a2a6881d2c74be404fc22bcb8def4ce8e1b862370b00f9ba58af4fab9269d872.
Human epsilon-eGFP in BHK cells shows basal cytoplasm/nuclear exclusion; engineered NLS plus leptomycin experiments have different scope.
Inspected: Selected original Results, Methods and captions concerning eIF2B-epsilon–eGFP, nuclear-localization-signal constructs and leptomycin B; indexed paragraphs 9,18,49,50,52,53,55,57–60,69–71,90,91.
Limit: No complete insert-sequence verification or universal nuclear-exclusion claim; text/captions read, not pixels.
Source: https://pubmed.ncbi.nlm.nih.gov/12426392/; cache publications/PMID_12426392.md, SHA256 58b29f9210cc36331900b9a1caa960ef75dac4896714d6b7578ddca1cb9a89e2.
Homozygous EIF2B5 severe disease with low uric acid/elevated GDP sugars; heat-specific experiment not exposed.
Inspected: Verified abstract; full original unavailable.
Limit: Publisher failure; no title/abstract-only wrong-gene claim.
Source: https://pubmed.ncbi.nlm.nih.gov/12499492/; cache publications/PMID_12499492.md, SHA256 efcb1d43c86350b0b95d0fb684be641ebb1ca10b63b2d6af582256727d89b2fe.
Seven mutated patients include EIF2B5 R113H homozygotes and R113H/R195C; ovarian failure and white-matter disease are positive clinical evidence.
Inspected: Normal indexed blocks13–35, human cases/genotyping/clinical findings and Discussion.
Limit: No follicle-development assay; complete table cells not reproduced in the normal cache.
Source: https://pubmed.ncbi.nlm.nih.gov/12707859/; cache publications/PMID_12707859.md, SHA256 44f03632f92045e00942a2d3055e3050718af3c394fdec01b4f5623b0db69f71.
Rat hippocampal eIF2B-subunit expression changes developmentally; no functional hippocampus-development assay reported in inspected scope.
Inspected: Official Europe PMC core metadata/abstract, checked against the selected normal cache metadata; no full body inspected.
Limit: Donor species and the limited abstract scope retained; normal shared bytes reused where available.
Source: https://pubmed.ncbi.nlm.nih.gov/12850562/; cache publications/PMID_12850562.md, SHA256 8ee4b4d22ba09cefb30535d8971b10903b1a9e4878a0cbef0d10f2ca7a214efe.
EIF2B5 patients7(V430A/S447L) and8(Y495C homozygous) have severe abnormal myelination/rarefaction.
Inspected: Original Methods, clinical/MRI Results and Table 1; indexed paragraphs 14–28, including MRI Results 18, caption 20 and Table 1 at 26.
Limit: Clinical/pathological observations do not themselves identify a direct epsilon myelin-sheath formation step.
Source: https://pubmed.ncbi.nlm.nih.gov/14566705/; cache publications/PMID_14566705.md, SHA256 6bd9ac2ef7155c6eb4f57e7490a187758325666c17f77da848671e626e0b536a.
Patient EBV lymphoblast extracts, rat-liver eIF2 substrate, triplicate GDP dissociation; multiple literal EIF2B5 genotypes reduce complex activity, heterozygotes normal.
Inspected: Author-uploaded original, selected Methods, Table 1 and GDP-dissociation Results (captured text lines 336–786).
Limit: Patient lysates are not isolated epsilon; no figure pixels.
Source: https://pubmed.ncbi.nlm.nih.gov/15054402/; cache publications/PMID_15054402.md, SHA256 25c3ade454409e484565cfd159428846636bc653117aa2a6a2e02c63d46be939.
Full-length sequence-verified human clones in HEK293; gamma binds epsilon in a two-subunit expression context, delta is recovered with assembled complex. Human epsilon variants affect exchange/substrate association.
Inspected: Partial normal original text plus externally recovered full-length cloning, complex-formation Results and Figure 1 caption. The normal cache does not contain the complete Methods.
Limit: Cache flag true does not imply complete Methods; selected recovered sections only. No synthesized insert-to-current-reference equivalence assumed.
Source: https://pubmed.ncbi.nlm.nih.gov/15060152/; cache publications/PMID_15060152.md, SHA256 156d464d437c17d85a21181b9cb56c5a01f86eef50a76d9fa1a267f6de81b1dc.
Four patients have EIF2B5 and one EIF2B2; oligodendrocyte density, apoptosis-associated staining and myelin pathology measured. Increased density has unresolved proliferation/survival explanations.
Inspected: Author-uploaded original, selected autopsy Methods, genotypes, quantification and Discussion; captured text lines 250–661, 477–1200, 1360–1545 and 1864–1882 in the saved views.
Limit: Selected original windows, not complete-paper/pixel inspection; no lineage-progress assay inferred.
Source: https://pubmed.ncbi.nlm.nih.gov/15217090/; cache publications/PMID_15217090.md, SHA256 2d08c635e04d508330ae04bbac0a2ff52a48f3b6b47a6b8204a82d8957da940d.
93 isolated POF women were negative for screened known variants; prior positive ovarioleukodystrophy is discussed, not experimentally recapitulated.
Inspected: Original Methods, Results and Discussion; indexed paragraphs 14–34.
Limit: No assertion that all ovarian EIF2B5 variants were excluded.
Source: https://pubmed.ncbi.nlm.nih.gov/15507143/; cache publications/PMID_15507143.md, SHA256 fbe1a6164c493f2f8592828e21367e6dae6b5a4ccd50ff92117a0860c5fc0eab.
Affected human cultures produce few abnormal GFAP-positive astrocytes; normal human glial-progenitor EIF2B5 RNAi impairs GFAP induction. Normal-appearing oligodendrocytes reported.
Inspected: Normal abstract, Nature publisher preview, figure titles3/4 and available supplement link metadata.
Limit: Full operative main Methods/supplements inaccessible; direct figure-page attempts failed. No heat or sheath-formation experiment reconstructed.
Source: https://pubmed.ncbi.nlm.nih.gov/15723074/; cache publications/PMID_15723074.md, SHA256 88b91229a68d240723900a3d8680e5e9daaa3ee99a22f0a78f5a1cbe257c92bf.
Patient eIF2B-mutant fibroblasts show heightened ATF4 response with similar broad translational inhibition/upstream kinase activation.
Inspected: Normal and official Europe PMC abstracts; publisher payload inspected and identified as HTML challenge.
Limit: Specific EIF2B5 genotype and operative experiments unresolved; HTTP200 HTML is not a PDF.
Source: https://pubmed.ncbi.nlm.nih.gov/16041584/; cache publications/PMID_16041584.md, SHA256 b124358737d83cbbcb20f9342a4f2f6da0f4ff33e05dcefe7962b1831e288e31.
eIF2/eIF2B supplementation improves HeLa-extract translation; the added factor species/preparation is unresolved in this access scope.
Inspected: Verified abstract and independent access assessment; complete operative Methods unavailable.
Limit: Full operative Methods unavailable; broad complex function has independent support.
Source: https://pubmed.ncbi.nlm.nih.gov/16289705/; cache publications/PMID_16289705.md, SHA256 c62fe42768288a45221c04cf0fec987399e51cf494933d632429092a65121792.
Rat epsilon-FLAG muscle electroporation increases GEF; restores sepsis-depressed synthesis, while healthy-muscle synthesis need not rise.
Inspected: PMC indexed9,10,13–16,20,24,25,28,30,32,36,37.
Limit: Rat muscle/homogenate context, not isolated human epsilon activity.
Source: https://pubmed.ncbi.nlm.nih.gov/20484009/; cache publications/PMID_20484009.md, SHA256 bf571a26cf79d435559fd2170986b9de92695434df644606d89122e71b4b6de9.
Whole-complex GDP release, CHO delta-mutant and mouse-fibroblast contexts are distinct. Purified complex species/construct and the sequence exception cannot be resolved without the missing Methods/table.
Inspected: Independent original-source consultation: main text and captions read, with supplementary Table S1 and complete purification details unavailable.
Limit: Whole-complex exchange and CHO delta-mutant experiments do not establish an isolated-epsilon assay; unread supplementary sequence details remain unresolved.
Source: https://pubmed.ncbi.nlm.nih.gov/25858979/; cache publications/PMID_25858979.md, SHA256 48e7de1b46a1e59dea4d02c75879e4402e9d957cc202ddf87b844dbe91775dc1.
Human eIF2/eIF2B reconstituted-translation comparator exists; detailed purification/crystal work is primarily S. pombe.
Inspected: Independent original-source consultation: operative Methods/Results, including the human comparator in Methods paragraph 12, Results paragraph 23 and Figure 2h.
Limit: Most detailed purification and structural work concerns S. pombe; the human comparator is retained separately.
Source: https://pubmed.ncbi.nlm.nih.gov/27023709/; cache publications/PMID_27023709.md, SHA256 24fbdc63dc56c647ba96c8d3cae410754631c826d7584d48c8456ded906aa52b.
Five codon-optimized human genes expressed in E. coli; alpha2 plus beta/gamma/delta/epsilon reconstitution and exchange support complex function.
Inspected: Independent original-source consultation: printed pages 1–3 and selected Discussion plus cloning, purification and GDP-assay Methods on pages 7–9 of the author-hosted article.
Limit: Supplementary exact insert sequence and figure pixels were not inspected; no isolated human epsilon assay claimed.
Source: https://pubmed.ncbi.nlm.nih.gov/29599213/; cache publications/PMID_29599213.md, SHA256 8e41c97522808119bef1e4a699e5da88ae84997783a655d903fcb30f7f84fb22.
Human epsilon entity5 chains I/J expressed in E. coli has exact official 721-residue sequence; whole-complex identity does not establish contact geometry or isolated activity.
Inspected: Verified abstract and official PDB 6K72 entity 5 sequence/organism metadata; full original Methods/Results unavailable.
Limit: Full original Methods/Results unavailable; metadata is not operative Methods.
Source: https://pubmed.ncbi.nlm.nih.gov/31048492/; cache publications/PMID_31048492.md, SHA256 afbee8df7ae0688f4a14df19dde73521629405a230f0997e5c36f6380728d1d3.
Human ORFeome9.1 binary screening with pair verification; exact target pair positive in three Y2H configurations.
Inspected: Normal indexed blocks16–19,44–45,50,52; all literal target STX11 IntAct records and features.
Limit: Dataset-level orthogonal validation is not assigned to this pair; exact natural isoform insert not proven.
Source: https://pubmed.ncbi.nlm.nih.gov/32296183/; cache publications/PMID_32296183.md, SHA256 de8eeaee8fd10c8537ad6b2f69602d414b6f1896221b208edb8ffbae9eaa263b.
Positive Y2H screen records support identified dynein-heavy-chain and protease partners.
Inspected: All11 normal-cache blocks; human Q13144–Q14204 and Q13144–Q92876 source-specific IntAct records.
Limit: No complete Methods or exact insert-sequence proof; reference sequence checksums do not establish assayed inserts.
Source: https://pubmed.ncbi.nlm.nih.gov/32814053/; cache publications/PMID_32814053.md, SHA256 df9eea608bd1079b9e3320715db0c16712ded397fd7cfde26b291d9ad54e6f3b.
C-terminally tagged STX11 associates with epsilon; syntaxin class is independently verified.
Inspected: Selected normal original Introduction/Discussion and exact STX11 anti-tag co-IP record.
Limit: Cache flag true does not prove complete Methods; physical association is not necessarily binary binding.
Source: https://pubmed.ncbi.nlm.nih.gov/33961781/; cache publications/PMID_33961781.md, SHA256 4763a1c8018e1b047fc1c2bf3d0e72c837e4b4c3dffbed428befa479e1c92286.
Rabbit epsilon-directed antibodies inhibit exchange and translation initiation; other-subunit antibodies are distinct.
Inspected: Official Europe PMC core metadata/abstract, checked against the selected normal cache metadata; no full body inspected.
Limit: Donor species and the limited abstract scope retained; normal shared bytes reused where available.
Source: https://pubmed.ncbi.nlm.nih.gov/8168527/; cache publications/PMID_8168527.md, SHA256 038e6230e9d739b183cae1f98e868c90bbfeac17cb235aebbacbbd3d2f522d65.
Human-insulin-receptor-transfected CHO cells show insulin-dependent GSK3 inactivation; that kinase phosphorylates the largest eIF2B subunit.
Inspected: Complete selected normal abstract.
Limit: Host is hamster, receptor human; no direct epsilon hormone-response step or factor preparation species inferred.
Source: https://pubmed.ncbi.nlm.nih.gov/8397507/; cache publications/PMID_8397507.md, SHA256 b2c7296f98bacffcf5fe02c3d8c563d7eb0a863e9aec3566d60325c40570f46d.
Diabetic rat muscle exchange activity differs by tissue; NADPH/CKII assays reported, explicit peptide-hormone stimulus unresolved.
Inspected: Official Europe PMC core metadata/abstract, checked against the selected normal cache metadata; no full body inspected.
Limit: Donor species and the limited abstract scope retained; normal shared bytes reused where available.
Source: https://pubmed.ncbi.nlm.nih.gov/8430778/; cache publications/PMID_8430778.md, SHA256 db70d61bfaf724f85a3c21ac95a45e4889d4038509d8f7e5e0d629eb6d5225db.
Rat pancreatic-islet glucose increases exchange activity rapidly; mannoheptulose did not activate eIF2B. The accessible abstract does not demonstrate blockade of the glucose response.
Inspected: Official Europe PMC core metadata/abstract, checked against the selected normal cache metadata; no full body inspected.
Limit: Donor species and the limited abstract scope retained; normal shared bytes reused where available.
Source: https://pubmed.ncbi.nlm.nih.gov/8567668/; cache publications/PMID_8567668.md, SHA256 f1e63f45add5ce6dfd63eadec562b0e855d8a04db3542ab47e5fd03a6b626bbd.
Human T-cell activation increases whole-complex eIF2B activity and changes GSK3/beta abundance.
Inspected: Verified abstract and independent source consultation covering publisher-indexed selected Results/Discussion; complete Methods unavailable.
Limit: Complete Methods and epsilon-specific TCR mechanism unresolved; beta focus is not negative epsilon evidence.
Source: https://pubmed.ncbi.nlm.nih.gov/8626696/; cache publications/PMID_8626696.md, SHA256 396404cd79b61da5068d4b6e5c9e022a60d8d5d38c47a2dfa22214c4c6e1f5e7.
Rabbit epsilon cloning and human U937 fragment establish identity/conservation; an initially incomplete clone is not a natural product or isolated-human catalytic assay.
Inspected: Verified abstract; full publisher text returned HTTP 403.
Limit: Full original unavailable; no isolated-human catalytic assay or natural alternative product inferred.
Source: https://pubmed.ncbi.nlm.nih.gov/8688466/; cache publications/PMID_8688466.md, SHA256 bda26f939c869f938da959968fd7e943570a8d5fb170f75324557c19cde12de7.
Five rat genes expressed in Sf9; epsilon alone has exchange activity in extracts, assembled subunits strongly enhance it; alpha controls phosphorylated-eIF2 inhibition.
Inspected: Official Europe PMC core metadata/abstract, checked against the selected normal cache metadata; no full body inspected.
Limit: Donor species and the limited abstract scope retained; normal shared bytes reused where available.
Source: https://pubmed.ncbi.nlm.nih.gov/9139680/; cache publications/PMID_9139680.md, SHA256 7817491cd36fd35d28068ff897ebc5ba4cd50a0ed288cfa70c6668497ddbeca9.
Rat H35 heat exposure changes whole-complex exchange activity and protein synthesis; no epsilon thermosensor assay.
Inspected: Official Europe PMC core metadata/abstract, checked against the selected normal cache metadata; no full body inspected.
Limit: Donor species and the limited abstract scope retained; normal shared bytes reused where available.
Source: https://pubmed.ncbi.nlm.nih.gov/9341116/; cache publications/PMID_9341116.md, SHA256 4bedb999860ab6aff71a95a5f89b3c7bd45f1c611a64d6f34432084ac707594c.
Rows 48, 49 and 57 remain UNDECIDED. These concern the shared myelination claim and an epsilon-specific T-cell-receptor pathway step. The inaccessible original patient ER-stress and rat hormone-donor experiments remain source-specific limitations, while independent mammalian epsilon evidence supports the broad response claims as non-core. PMID:11500362 provides a direct GSK3 docking experiment and an explicitly secondary account of insulin-dependent epsilon regulation; PMID:11756553 is not an insulin-challenge study.
The STX11, DYNC1H1 and KLK6 generic positive interactions remain non-core; their partner identities do not supply independently established epsilon-specific functional roles. Missing screen Methods limit the proposed refinements, not the existence of the curated positive pairs. STX11 is corroborated across binary-screen and cellular-association datasets. No native-interface requirement is added to GO binding definitions. Heat and glucose responses retain their existing non-core scope. Clinical and perturbation observations are not promoted into autonomous developmental or apoptotic mechanisms. The positive mouse nuclear observation remains contextual: stronger fixation increases cytoplasmic relative staining without demonstrating loss of nuclear signal.
The existing UniProtKB Q13144 record is retained byte-for-byte and now has an explicit file reference in the review. Its statements have different evidence scopes:
| Record section and lines | Coverage and limit |
|---|---|
| FUNCTION, 303–312 | ECO:0000269 citations 25858979, 27023709 and 31048492 support eIF2 GDP/GTP exchange and regulation by phosphorylated eIF2. This corroborates the independently assessed catalytic core; it is not evidence of isolated human epsilon activity. |
| SUBUNIT, 317–326 | The same cited studies support the heterodecamer and epsilon/gamma catalytic subcomplex. Complex architecture does not turn every co-recovered partner into a demonstrated binary interface. |
| ACTIVITY REGULATION, 313–316 | PMID:25858979 is the curated basis for ISRIB activation of eIF2B. No new annotation is proposed here. |
| INTERACTION, 327–331 | Exact DYNC1H1, GSK3B, KLK6 and STX11 records corroborate the partner identifiers. The existing source-specific binding judgments remain unchanged. |
| PTM, 334–338 | Ser-544 priming by DYRK2 is explicitly by similarity (ECO:0000250). The separately stated GSK3B phosphorylation/inhibition evidence cites PMID:8397507. Conserved-site context does not remove the engineered human-peptide or rat-construct limits of the kinase-binding evidence. |
These are reads of the curated record, not additional claims of underlying primary-paper access. Existing primary read scopes above remain controlling.