FANCF is an integral organizing subunit of the Fanconi anemia core complex. Human assembly-mutant and complementation experiments support its structural contribution to interstrand cross-link repair; conserved hydroxyurea-response evidence preserves broader replication-stress scope. FANCL and UBE2T perform ubiquitin transfer. No FANCF nuclease, polymerase or ubiquitin-transfer activity is inferred from its presence in a reaction complex.
These are agent-authored evidence notes. The direct-original reviewer inspected the experimental sources and screen records identified below. The database/core reviewer independently read the human assembly sources and mouse donor chapter, database mappings and target-specific participant paths. The complete labeled reference assessments retain the exact reading boundaries of each reviewer; one reviewer's figure inspection is not attributed to the other.
PMID:11063725(https://doi.org/10.1093/hmg/9.18.2665), The Fanconi anemia protein FANCF forms a nuclear complex with FANCA, FANCC and FANCG, was recovered as a complete original from the OHSU institutional repository. The relevant Results, captions and Discussion, and patient/cell, construct, antibody, fractionation and reticulocyte-translation Methods were read at printed pages 2666–2673. Figures 4 and 5 on page 2669 were also inspected as pixels. Human endogenous FANCF co-immunoprecipitates with core subunits; complementation restores nuclear FANCA and FANCA–FANCG complex formation. The negative pairwise reticulocyte experiments remain explicit. They do not negate cellular assembly, nor do the cellular co-immunoprecipitations alone resolve every binary interface. FANCF-GFP and endogenous fractionation support predominant nuclear residence. The authors qualify the small cytoplasmic signal as potentially authentic or caused by diffusion/cross-contamination; it is not definitive native cytosolic quantification.
PMID:15262960(https://doi.org/10.1074/jbc.M407034200), The Fanconi anemia gene product FANCF is a flexible adaptor protein, was independently read from the complete author-hosted original. Experimental Procedures, the relevant Results and Figures 4–6/Table I were read; actual pages 39424–39425 were viewed. Human FANCF mutants distinguish C-terminal FANCG/core contacts from an N-terminal region needed for association with the FANCC/FANCE subcomplex. Some C-terminal deletions retain MMC complementation despite weak stable co-immunoprecipitation, so these readouts are not equated. The N-terminal 1–15 deletion remains nuclear but fails MMC rescue. In the three-hybrid experiment, FANCC mediates FANCE–FANCF association. Neither this result nor the paper's historical terminology assigns an enzymatic reaction to FANCF. Engineered deletions are not natural alternative products.
The complete abstract of PMID:12649160(https://pubmed.ncbi.nlm.nih.gov/12649160/) supports FANCG interaction with the FANCF C-terminal region and FANCG-mediated FANCA–FANCF association. Complete operative text was not recovered after the saved publisher and archival attempts. No unobserved Methods or additional constructs are inferred. The scaffold replacement uses the independently read human experiments above; it does not claim the inaccessible paper alone demonstrated every aspect of that activity or reassign the specific FANCG bridge to FANCF.
The database/core reviewer independently read the 2000 main text through Methods without viewing its figure pixels, and the 2004 Experimental Procedures, Results and captions through Figure 6/Table I plus selected Discussion/Figures 9–10 captions. Actual Figures 4–6/Table I were viewed. The 2004 author PDF has a checked journal imprint and bibliographic identity; the repository calls it another version, so version identity is not presumed.
PMID:17082180(https://pubmed.ncbi.nlm.nih.gov/17082180/) retains positive structural/assembly information in its abstract. Its historical findings and quotations remain intact. Complete Methods, Results and figure pixels were not independently recovered/read in this assessment. Current conclusions have independent support from the actual human assembly experiments above; inherited quotations do not imply a fresh full-text read.
For the two ICL-repair annotations (zero-based rows 3 and 11), the structured support now names both PMID:15262960 and PMID:17082180 explicitly. The former supplies independently read human FA-F cell complementation and assembly-mutant experiments (Experimental Procedures; Figures 4–6 and Table I, pages 39423–39425); the latter supplies FANCF-specific structural/mutant conclusions in its verified original abstract. The existing FANCI–FANCD2/pathway quotations remain historical support, not a substitute for these FANCF-specific sources. Adding the latter citation does not claim new access to its complete article.
The official GO:0030674 comment directs protein-complex scaffolds to GO:0140378, protein complex scaffold activity, an integral structural component holding a complex together. This fits FANCF's complementary assembly interfaces and rescue phenotypes. The narrower GO:0160072, ubiquitin ligase complex scaffold activity specifies ligase–substrate-adaptor organization; the inspected FANCF experiments do not establish that precise positioning. Neither substrate recognition by FANCE nor ubiquitin transfer by FANCL/UBE2T is assigned to FANCF.
One scaffold unit carries interstrand cross-link repair, nuclear location and Fanconi anemia nuclear-complex membership. DNA damage response is an actual ancestor of interstrand cross-link repair and is not repeated as a redundant core process; the independently measured hydroxyurea response remains in the description and accepted broader annotation. The encompassing nuclear location retains both nucleoplasmic and chromatin-associated complex pools. A separate cytosolic core is not inferred.
A current exact-term comparator query for FANCA, FANCG and FANCB found no GO:0140378 rows. This is recorded as a negative comparator result, not evidence of curator omission. The scaffold core follows the actual definition and the supported MODIFY replacements; no NEW annotation is needed. The frozen repository GO-CAM index and cached model text contained no FANCF/Q9NPI8 match, a bounded repository result rather than a claim about all live models. Typed term ancestry, exact definitions and query responses are retained in the review evidence.
PMID:22343915(https://pmc.ncbi.nlm.nih.gov/articles/PMC3321854/) identifies FANCF in the human HeLaS3 nuclear core purification: Figure 1A/B explicitly labels its band. The main purification, construct and fractionation Methods and relevant Results were read, together with authentic supplemental Methods and Figure S1. Actual Figure 1, Figure 5 and supplemental S1 pixels were inspected. Figure 5H assays FANCA, FANCG, FAAP20 and H2A rather than a separate FANCF band. The retained chromatin assertion is a supported curator projection of the assembled complex's localization, using positive FANCF membership and complex loading. It is not an invented individual FANCF chromatin assay or evidence of isolated FANCF DNA binding; absence of an individual band is not contrary evidence.
PMID:20347428(https://pmc.ncbi.nlm.nih.gov/articles/PMC2847587/) explicitly identifies FANC-F in a core complex recovered by HeLa nuclear fractionation and FANCM immunoprecipitation, with silver staining, mass spectrometry and immunoblot analysis. Selected Results, purification Methods and Figure 1 caption were read. Separate measurements place most FANCM/MHF outside the FA core. They do not transfer FANCM/MHF enzymatic activity to FANCF or require all FANCM complexes to contain it. No new pixel or supplemental read is claimed for this source.
PMID:19965384(https://pmc.ncbi.nlm.nih.gov/articles/PMC2909596/) uses Xenopus egg extract and recombinant FANCI–FANCD2 depletion/add-back with a cisplatin-crosslinked plasmid. The complete main narrative and Figures 1–4 captions were read. This is pathway evidence, not an isolated human FANCF assay. Human FANCF assembly and complementation in the two studies above independently support its structural participation in ICL repair. FANCL/UBE2T perform ubiquitin transfer and downstream enzymes perform incision and synthesis.
The four FA nuclear-complex annotations (rows 1, 4, 12 and 24) also point to existing PMID:11063725 and PMID:17082180 support. The immutable UniProt record, CC -!- SUBUNIT block (lines 123–130), independently provides the curated complex-membership statement. Its six ECO:0000269 provenance IDs are 11063725, 12724401, 15502827, 16116422, 17082180 and 18550849. That block combines complex-membership and HES1-interaction statements; it is not treated as six independently reread FANCF experiments or evidence that FANCF performs the partners’ chemistry. The existing reference list is retained.
The two PAINT assertions derive from PANTHER family PTHR14449, ancestral node PTN001426273. The actual version 19 tree and alignment place human leaf PTN002494673, whose ungapped sequence exactly matches the displayed 374-residue Q9NPI8 protein, below that node. The target path contains two positive IBD assertions and no NOT. The DNA-damage-response assertion uses mouse MGI:3689889; the FA-complex assertion uses that mouse source and human Q9NPI8. Human self-WITH is legitimate experimental descendant evidence.
Public source locators are the PANTHER family tree, family alignment and GO PAINT IBD distribution. The exact retrieved URLs and saved version are retained in the source receipts; live services can change. The observed ancestry is PTN001426273 → PTN001025035 → PTN008352821 → PTN002683660 → PTN008555810 → PTN000374517 → PTN000374518 → PTN000374519 → PTN000374520 → PTN000374521 → PTN000374522 → PTN002494673.
Current mouse E9Q5Z5 records trace the two experimental donor assertions to PMID:21915857(https://pubmed.ncbi.nlm.nih.gov/21915857/). A genuine accessible author version is chapter 5 of Bakker's dissertation, FANCM, the mouse that roared, beginning on printed page 77. Methods on pages 79–81 and Results/Figures 5.1–5.2 on pages 82–85 establish the critical scope. Mouse and human FANCF complement human FA-F cells. Fancf-null mouse fibroblasts show MMC hypersensitivity, chromosome breaks and G2 accumulation, and fail FANCD2 monoubiquitination after either MMC or hydroxyurea. Figures 5.1 and 5.2 pixels were viewed. Selected Discussion was read; other tumor/reproductive figures, supplementary material, cited earlier assay protocols and the final journal PDF were not independently read.
The decisive original is now directly inspectable as printed page 84, Figure 5.2, PDF page 85, with source and rendering provenance. Panel D compares Fancf-null and wild-type mouse MEFs after HU (1 mM, 24 h), MMC (3 micromolar, 24 h) or no treatment. The upper FANCD2-L species is detected in the wild-type response and absent in the null cells; this is the complex-dependent monoubiquitination readout, not a direct catalytic assay of FANCF. The full page was rendered without cropping or alteration from the linked author dissertation. Its SHA-256 is a36f5493226d97f588e33c8c1a1bc3c4444bd1942082beead53b9aba957cc98b; the PDF SHA-256 is d599d3afc0ee3ec5dd3fcff5930404b6242e054d1289b852ff1fa2d40ace14aa. The selected normal cache remains abstract-only; the public artifact documents the separately inspected external original.
The measured hydroxyurea response supports retaining broader DNA damage response alongside interstrand cross-link repair. This is evidence of additional stress-response breadth, not retention of a parent merely for completeness. The mouse knockout is functional evidence for the core machinery; it is not purified FANCF ubiquitin-transfer chemistry or direct complex purification. Tumor and gonadal outcomes do not create additional human process assertions here.
The official UniProt SL-0191 mapping supports nucleus. Human imaging/fractionation establishes that enclosing compartment without restricting all FANCF to chromosome-excluding nucleoplasm. The Human Protein Atlas record separately supports nucleoplasm in SH-SY5Y cells. HPA008899's 130-residue antigen matches Q9NPI8 residues 231–360 exactly. Both linked merged images were viewed. They support nuclear-interior signal without quantitative colocalization, a damage-focus claim or product-specific interpretation.
InterPro IPR035428 matches human FANCF and maps to ICL repair and FA nuclear-complex membership. ARBA00086467 requires Eukaryota and exact CATH FunFam 1.25.40.490:FF:000001. The independently captured CATH hit establishes the encompassing superfamily only; exact FunFam membership was not independently re-executed. This limitation does not negate the separate verified InterPro map or the human experimental complex evidence.
All 11 assigned events were traced through actual participant links to FANCF itself. The ten nuclear events reach human Q9NPI8 entity 419537, residues 1–374, with its own nucleoplasm compartment. The cytosolic event reaches entity 420794, also residues 1–374, with its own cytosol compartment. Twenty-six target routes are complete; eight unexpanded objects elsewhere in the graph are not treated as complete off-target coverage.
| Event / row | FANCF-specific interpretation and inspectable participant route |
|---|---|
| R-HSA-6785126 / 14 | Input FANCF and assembled core output; DNA loading is distinguished from assembly. Representative source route: 6785126 → 419537 (input). |
| R-HSA-6785342 / 15 | Core membership during ID2/UBE2T recruitment; no autonomous ubiquitin transfer. Representative source route: 6785342 → 6785124 → 6785123 → 419537 (input → hasComponent → hasComponent). |
| R-HSA-6785361 / 16 | Component of the modeled catalyst assembly; FANCL/UBE2T perform the chemistry. Representative source route: 6785361 → 6788398 → 6788397 → 6788399 → 419537 (input → hasComponent → hasComponent → hasComponent). |
| R-HSA-6785732 / 17 | Core-containing complexes during nuclease recruitment; no FANCF nuclease activity. Representative source route: 6785732 → 6785367 → 6788397 → 6788399 → 419537 (input → hasComponent → hasComponent → hasComponent). |
| R-HSA-6785986 / 18 | Component location during unhooking; named nucleases make the incisions. Representative source route: 6785986 → 6786151 → 419526 → 419537 (output → hasComponent → hasComponent). |
| R-HSA-6786155 / 19 | Core location during POLN recruitment. Representative source route: 6786155 → 6786151 → 419526 → 419537 (input → hasComponent → hasComponent). |
| R-HSA-6786166 / 20 | Core component during modeled POLN synthesis; no FANCF polymerase activity. Representative source route: 6786166 → 6788399 → 419537 (output → hasComponent). |
| R-HSA-6786171 / 21 | Substrate/output complex location; USP1:WDR48 performs deubiquitination. Representative source route: 6786171 → 6788397 → 6788399 → 419537 (output → hasComponent → hasComponent). |
| R-HSA-6788385 / 22 | Core location during ATR–ATRIP recruitment; no inferred binary FANCF–ATRIP interaction. Representative source route: 6788385 → 6785340 → 6785124 → 6785123 → 419537 (input → hasComponent → hasComponent → hasComponent). |
| R-HSA-6788392 / 23 | Core component in the composite assembly; ATR performs phosphorylation. Representative source route: 6788392 → 6788398 → 6788397 → 6788399 → 419537 (output → hasComponent → hasComponent → hasComponent). |
| R-HSA-9835411 / 13 | Cytosolic FA-core:HSP70 participant, retained as a contextual location. The summation emphasizes FANCC/PKR and A/C/G phenotypes, not an independently verified FANCF-specific PKR mechanism. Representative source route: 9835411 → 9835414 → 9835415 → 420794 (input → hasComponent → hasComponent). |
Rows 13–23 each cite their own existing Reactome reference in supported_by. The ten nucleoplasm routes terminate at R-HSA-419537.1, FANCF, whose own record identifies human Q9NPI8 residues 1–374 and nucleoplasm. The cytosol route terminates at R-HSA-420794.1, FANCF, with the same sequence bounds and its own cytosol compartment. The table gives one complete linkable route per event; all 26 source routes were checked against the saved official object responses. For the cytosolic row, the event summation chiefly describes FANCC/PKR and A/C/G phenotypes; the target-containing participant model and the qualified small cytoplasmic pool in PMID:11063725 support contextual retention, without a FANCF-specific PKR mechanism. The inherited nuclear-localization quotation is preserved but does not itself establish cytosol.
These are related modeled location contexts, not 11 independent localization experiments. The cytosolic pool is retained without claiming inactivity, contamination or a FANCF-specific PKR assay. Event-cited originals were not all independently read.
PMID:32296183(https://doi.org/10.1038/s41586-020-2188-x) and the actual HuRI dataset contain TCP11 isoform Q8WWU5-7 with FANCF Q9NPI8. The literal records use human ORFeome clones 100080789 and 55485, historical Ensembl identifiers and N-terminal Gal4 fusions in yeast. Array, pooling and validated records describe stages of one screening pipeline. The relevant main Methods and all three target records were read. Study-wide MAPPIT/GPCA benchmarks are not asserted as individual validation of this pair. A previously saved file named as a supplementary DOCX was actually HTML and was excluded.
PMID:32814053(https://doi.org/10.1016/j.celrep.2020.108050) Table S2 records FANCF/HTT at NN1.0 row 4403 and NN1.0ext row 9678. Six authentic workbooks were mechanically searched; target records and headers were inspected. The author-main STAR Methods describe yeast autoactivation controls and growth/beta-galactosidase selection. Current IntAct records IM-28217-54611, IM-28217-38845 and IM-28217-69841 specify HTT bait residues 1–546 with mutation annotations at 18–38 and FANCF prey. Table A/B order alone was not used to infer orientation. These are engineered yeast-assay contexts, not evidence for every native full-length HTT state or a FANCF role in neurodegeneration. No target-specific mammalian DULIP/LuTHy result was established by the inspected material.
Current ontology searches and partner-class assessment did not establish a defensible more-specific FANCF activity for either screen pair. Broad tokenized search results were not treated as an exhaustive absence proof. The supported generic associations remain non-core under the project's explicit binding policy.
The displayed human Q9NPI8 sequence is 374 residues. UniProt lists no alternative-products field. Engineered deletions, fusion proteins, the mouse ortholog and external transcript records do not create natural human isoforms or establish identical assay behavior.
The current source version contains 27 imported contexts: 20 ACCEPT, four MODIFY and three KEEP_AS_NON_CORE. Genuine GOA seeding added supporting-entity lists to the first nine inherited records and split each of two older FANCA/FANCG binding groups into partner-specific contexts. The FANCG records retain their original positions; the two FANCA contexts are appended at 25 and 26. At the source-adoption checkpoint, all 25 inherited reviews were preserved. The present evidence-link update changes only the documented support links, source-specific prose and two reference assessments; all 42 inherited quotation paths, texts and reference associations remain exact, including the five original core supporting objects. The appended contexts use the accepted partner-specific decisions with bare supporting reference IDs only. This is an explicit source-version adoption, not two newly invented annotations. At the published checkpoint, canonical validation computed DRAFT with three advisories: two supported generic-binding screen associations retained as non-core under the standing campaign instruction, and the unused inherited research file. All six canonical checks passed at that checkpoint; no annotation remains UNDECIDED. The evidence-link successor also passed all six canonical checks, with the same three advisories and DRAFT status. All 60 review narratives render correctly.
All 26 reference identities and findings remain. For the two shared references, the historical assessment appears once, followed by both complete labeled current assessments. Availability flags are explicitly reconciled to immutable selected cache metadata: only PMID:32296183 has full text in that selected cache; the other nine selected PMID caches are abstract-only. External complete originals and supplements can nevertheless have been available and read as described above. No normal cache or quotation is rewritten, and no new quoted field is added. The 31-record normal union contains 24 selected dependencies and seven unselected private records, with the genuine refreshed GOA adopted explicitly and the inherited UniProt file kept exact.
