FXN

Summary: PMID:15641778 explicitly reports ferroxidase activity of assembled recombinant human frataxin, whereas the monomer does not prevent the tested iron-catalyzed radical reactions.

Reason: PMID:15641778 explicitly reports ferroxidase activity and iron sequestration by assembled recombinant human frataxin, whereas the monomer does not prevent the tested iron-catalyzed radical reactions. This positive biochemical observation is retained. PMID:15581888 also reports frataxin species from monomer to large polymers in human-heart mitochondria, supporting physiological assembly rather than an absence of oligomers in vivo. These assembly observations do not establish that iron oxidation is a physiological function of those species. The available evidence does not establish an evolved secondary ferroxidase role, so MARK_AS_OVER_ANNOTATED distinguishes a real measured activity from an established biological function. The accessible originals are abstract-only; no quantitative inefficiency, noise-level activity, promiscuity or unreported assay conditions are inferred.

Summary: The electronic ferroxidase assignment has independent human biochemical support from PMID:15641778, specifically for assembled frataxin.

Reason: The electronic ferroxidase assignment has independent positive human biochemical support from PMID:15641778, specifically for assembled frataxin. PMID:15581888 supports the occurrence of assembled species in human-heart mitochondria, but does not establish their physiological use of iron-oxidation chemistry. The electronic mapping does not resolve that functional gap. MARK_AS_OVER_ANNOTATED therefore retains the measured capability while declining to treat ferroxidase activity as an established evolved secondary function. This does not deny physiological assembly or claim that the activity is noise or promiscuous, and it does not imply that monomeric mitochondrial FXN normally acts as a ferritin-like iron-storage enzyme.

Summary: Human holofrataxin delivers iron to ferrochelatase in the reported reconstitution.

Reason: The authentic PMID:15123683 abstract identifies human holofrataxin binding to human ferrochelatase and delivering iron for heme formation. GO:0034986 iron chaperone activity, defined as directly binding and delivering iron ions to a target protein, describes this demonstrated capability more informatively than generic protein binding. Preserve the original P22830 partner, IPI evidence and reference. The replacement remains an assay-bounded, non-core property; full Methods, exact constructs and binding conditions were inaccessible, and this does not establish obligatory physiological iron delivery by FXN.

Proposed replacements: iron chaperone activity

Summary: The authentic abstract reports physical association of human frataxin with human succinate dehydrogenase subunits, supporting this SDHB context.

Reason: The original PMID:15961414 abstract explicitly reports physical interaction between human frataxin and human succinate dehydrogenase complex subunits. Retain the source-supplied SDHB partner as a supported non-core association. The current retrievable evidence does not independently establish residue ranges, antibody identity, expression host or purified binary affinity. Those details from a contemporaneous author-page reading are omitted from the current claim because its successful response body was not retained; the saved 403 response supplies no evidence.

Summary: The authentic abstract reports physical association of human frataxin with human succinate dehydrogenase subunits, supporting this SDHA context.

Reason: The original PMID:15961414 abstract explicitly reports physical interaction between human frataxin and human succinate dehydrogenase complex subunits. Retain the source-supplied SDHA partner as a supported non-core association. The current retrievable evidence does not independently establish residue ranges, antibody identity, expression host or purified binary affinity. Those details from a contemporaneous author-page reading are omitted from the current claim because its successful response body was not retained; the saved 403 response supplies no evidence.

Summary: The original abstract explicitly supports an HSPA9–frataxin association.

Reason: PMID:26702583 explicitly reports that HSPA9 interacts with and stabilizes frataxin and other ISC-biogenesis proteins, and separately says HSPA9 disease variants alter interactions. This is positive support for the P38646 association. Retain generic binding as non-core. Only the original abstract was accessible, so assay format, FXN construct, host and specific wild-type-versus-mutant comparisons remain unverified; no purified binary affinity is claimed.

Summary: FXN contacts NFS1 and ISCU in the assembled ISC complex.

Reason: PMID:31101807 resolves human FXN81–210 contacting NFS1 and ISCU in a recombinant assembly that includes bacterial ACP. Binding measurements use the assembled NFS1/ISD11/ACP/ISCU complex, not isolated FXN–ISCU affinity. Retain both original partner contexts as supported non-core generic binding; the separately annotated enzyme activation and specific assembly-complex membership express the core mechanism. Processed constructs do not establish alternative-product specificity.

Summary: The exact FXN–A6NEM1 HuRI assay contact is retained with its fragment scope.

Reason: Official IntAct records 0, 1 and 17 match FXN, A6NEM1 and PMID:32296183 exactly, with positive array, pooling and validated two-hybrid records. The human A6NEM1 participant is annotated as residues 159–432 and the host is yeast. Retain the supported assay interaction as non-core; the study-wide mammalian validation cohorts do not certify this exact pair, and no physiological mitochondrial co-localization or natural-isoform specificity is inferred.

Summary: HuRI supports the exact FXN–DAPP1 contact in yeast two-hybrid assays.

Reason: Official IntAct records 2, 18 and 19 match FXN, Q9UN19 and PMID:32296183, with positive two-hybrid assays in yeast. The pooling/array records specify tagged human constructs, including a C-terminal activation-domain tag on DAPP1. Retain this binding as non-core; a validated two-hybrid designation does not establish mammalian co-immunoprecipitation or a physiological FXN signaling role.

Summary: The original neurodegeneration-network dataset contains the FXN–PIAS1 pair.

Reason: PMID:32814053 Table S2, NN1.0 row 7826, reports FXN–PIAS1; exact O75925 IntAct records independently match the publication and positive yeast two-hybrid formats. Retain this assay-supported generic binding as non-core. No FXN pair appears in the retained Table S4 mammalian-validation sample, so do not infer endogenous interaction, sumoylation activity or transfer PIAS1 function to FXN.

Summary: SAMD3 binding is supported for the exact partner identifier used in the interaction record.

Reason: PMID:32814053 Table S2, NN1.0 row 9466, reports FXN–SAMD3, and the exact IntAct participant is Q8N6K7-2. Retain the positive yeast two-hybrid association as non-core while preserving that partner suffix. The partner isoform-comment concerns natural-product confirmation and is not a negative interaction flag. Neither physiological FXN–SAMD3 association nor specificity of an FXN natural product is established.

Summary: The reported FXN–RNF138 interaction is a supported non-core screen contact.

Reason: PMID:32814053 Table S2, NN1.0 row 9155, contains FXN–RNF138, joined to exact Q8WVD3 IntAct records from the same paper. Positive yeast two-hybrid formats support retaining this binding as non-core. The retained mammalian-validation table does not establish this pair, and RNF138 ligase activity is not an activity of FXN.

Summary: The dataset supports an FXN–CORO2A yeast two-hybrid contact.

Reason: PMID:32814053 Table S2, NN1.0 row 2297, reports FXN–CORO2A, with matching Q92828 positive IntAct records. Retain the contact as non-core. The source does not establish a native mitochondrial or actin-regulatory role for FXN from this pair; the generic binding assertion is narrower than those functional interpretations.

Summary: FXN–RNF183 is present in the original interaction dataset.

Reason: PMID:32814053 Table S2, NN1.0 row 9199, reports FXN–RNF183 and matches the exact Q96D59 IntAct partner. The positive yeast two-hybrid assays support non-core binding. They do not establish endogenous interaction, a particular organelle contact or ubiquitin-ligase activity of FXN.

Summary: The Q9BVU5 fragment contact is retained without expanding its identity or construct scope.

Reason: PMID:32814053 Table S2, NN1.0 row 5988, names LOC440444, and the literal Q9BVU5 IntAct partner joins that gene identifier. IntAct retains a tested region of residues 29–110. Retain the positive yeast two-hybrid contact as non-core; do not silently substitute an ARL17 paralog, infer full-length binding or treat the tested fragment as an FXN natural product.

Summary: The original screen supports an FXN–WDR61 contact.

Reason: PMID:32814053 Table S2, NN1.0 row 12750, contains FXN–WDR61, matching exact Q9GZS3 positive IntAct records. Retain this yeast two-hybrid interaction as non-core. The pair does not establish that FXN participates in every WDR61-containing complex or transcriptional process, and no pair-specific mammalian validation is inferred.

Summary: FXN–CAPN10 binding is supported within the tested screen system.

Reason: PMID:32814053 Table S2, NN1.0 row 1724, reports FXN–CAPN10; exact Q9HC96 IntAct records retain positive yeast two-hybrid evidence. Keep the binding as non-core. This contact does not identify a cleavage substrate, show CAPN10-dependent FXN processing or give FXN protease activity.

Summary: The exact RASSF1 partner product is preserved in the supported assay contact.

Reason: PMID:32814053 Table S2, NN1.0 row 8735, reports FXN–RASSF1, while the corresponding IntAct records specifically identify Q9NS23-4. Retain the positive yeast two-hybrid binding as non-core and preserve the partner-product suffix. It does not establish an FXN isoform-specific or endogenous RASSF1 signaling function.

Summary: The Haenig dataset independently records the FXN–DAPP1 contact.

Reason: PMID:32814053 Table S2, NN1.0 row 2492, reports FXN–DAPP1 and matches Q9UN19 IntAct records from this publication. Retain the positive yeast two-hybrid association as non-core. Preserve this original publication context separately from the HuRI row; the two source contexts are not evidence of a defined physiological FXN signaling mechanism.

Summary: The authentic PMID:15581888 abstract explicitly reports frataxin assemblies in human heart mitochondrial fractions, in addition to recombinant-protein analyses.

Reason: The authentic PMID:15581888 abstract explicitly reports frataxin assemblies in human heart mitochondrial fractions, in addition to recombinant-protein analyses. The full original was not available through the accessed article page, so its detailed fraction-purity and oligomer measurements were not re-evaluated. The basic mitochondrial location is nevertheless directly described and independently established by other human studies. Accept that location without importing a fixed physiological oligomer stoichiometry from the abstract.

Summary: PMID:16091420 examined endogenous FXN in human Caco-2 cells.

Reason: PMID:16091420 examined endogenous FXN in human Caco-2 cells. Selected original Methods and Results/Figures 4 and 5 describe mitochondrial colocalization and subcellular fractionation at days 4 and 13, with VDAC-1 and citrate-synthase controls. Differentiated cells additionally contain a prominent soluble cytosolic pool. Thus the mitochondrial annotation is directly supported and explicitly nonexclusive; no current natural FXN splice-product identifier is established by these assays.

Summary: In PMID:16608849, selected original fractionation and localization experiments detect mature endogenous FXN in mitochondria as well as an extramitochondrial pool in several human cell settings.

Reason: In PMID:16608849, selected original fractionation and localization experiments detect mature endogenous FXN in mitochondria as well as an extramitochondrial pool in several human cell settings. Mitochondrial marker controls and full-length versus engineered targeting-deficient constructs were distinguished. Accept mitochondrial localization as a central site of FXN function, without converting the additional cytosolic pool or an engineered 56–210 construct into a contradiction or natural-product assignment.

Summary: The authentic PMID:18725397 abstract reports mitochondrial localization and processing of human FXN and functional rescue experiments involving the mature 81–210 form.

Reason: The authentic PMID:18725397 abstract reports mitochondrial localization and processing of human FXN and functional rescue experiments involving the mature 81–210 form. Full selected experimental text was not recovered, so the paper’s detailed fractionation conditions were not reassessed. Its basic mitochondrial conclusion agrees with independently read human-cell studies. Retain this location without generalizing an exclusivity statement from the paper to every cell state or all natural FXN products.

Summary: PMID:20053667 distinguishes mitochondrial and cytosolic human FXN by fractionation and N-terminal sequencing in HEK293 cells.

Reason: PMID:20053667 distinguishes mitochondrial and cytosolic human FXN by fractionation and N-terminal sequencing in HEK293 cells. Selected original Figure 1 and Methods identify mature 81–210 FXN in the mitochondrial fraction as well as a cytosolic fraction, using MnSOD and tubulin controls. The later patient-cell rescue experiments distinguish full-length 1–210 from engineered cytosolic 81–210. The mitochondrial location is supported; those results do not establish a measured export mechanism.

Summary: The authentic PMID:9241270 abstract explicitly reports targeting of epitope-tagged human frataxin to mitochondria.

Reason: The authentic PMID:9241270 abstract explicitly reports targeting of epitope-tagged human frataxin to mitochondria. Yeast loss-of-function and iron phenotypes in the same paper are a separate experimental component and are not substituted for the human localization experiment. Full original Methods were not recovered, so the tag and cell details are not invented. The basic human mitochondrial location is also corroborated by independently read later studies.

Summary: The authentic PMID:9302253 abstract explicitly reports localization of human and mouse frataxin using immunofluorescence and immunoelectron microscopy, with mitochondrial/crista-associated observations and an N-terminal targeting region.

Reason: The authentic PMID:9302253 abstract explicitly reports localization of human and mouse frataxin using immunofluorescence and immunoelectron microscopy, with mitochondrial/crista-associated observations and an N-terminal targeting region. Full original Methods were not recovered. Accept the broad mitochondrial annotation, which is independently corroborated, without inferring a fixed membrane topology, exclusive compartment, or current splice-product specificity from the abstract.

Summary: The retained full main text of PMID:34800366 describes MitoCoP mapping using human HEK293T, HeLa, Huh7 and U2OS fractions, combining subtractive and spatial proteomics with import and literature evidence.

Reason: The retained full main text of PMID:34800366 describes MitoCoP mapping using human HEK293T, HeLa, Huh7 and U2OS fractions, combining subtractive and spatial proteomics with import and literature evidence. The selected main-text strategy and Figure 1 were read; the exact FXN supplementary-table row was not independently extracted. Retain the incoming HTP mitochondrial assertion because the location is securely established by independent FXN-specific human experiments, without claiming that the target’s individual MitoCoP measurements were re-audited.

Summary: The mitochondrial IBA represents the valid inherited organelle-level location of FXN.

Reason: The GO:0005739 IBD at AN12/PTN000423706 is genuinely ancestral to human FXN, with the complete literal donor set and target self-membership preserved. The established mitochondrial assembly role supports this organelle-level is_active_in assertion. The more specific matrix location is part_of mitochondrion, but does not invalidate or uniquely demote the ancestral organelle assertion relative to the other accepted mitochondrial rows. This is a nonexclusive localization claim: it neither denies experimentally supported extramitochondrial pools nor assigns every natural product to one compartment. The original qualifier, donor order, evidence and source fields remain unchanged.

Summary: The current official Human Protein Atlas FXN subcellular record identifies mitochondria as a supported main location and cytosol as a supported additional location.

Reason: The current official Human Protein Atlas FXN subcellular record identifies mitochondria as a supported main location and cytosol as a supported additional location. Antibody CAB022164 is documented, including mitochondrial and cytosolic staining in U2OS cells; images were not independently scored. This incoming HPA assertion agrees with direct fractionation and localization studies of human FXN. Retain the mitochondrial location without requiring exclusivity or assigning every natural product to it.

Summary: PMID:17468497 directly examines human FXN maturation in cells, using endogenous material, transfected constructs, fractionation and N-terminal sequencing.

Reason: PMID:17468497 directly examines human FXN maturation in cells, using endogenous material, transfected constructs, fractionation and N-terminal sequencing. Figure 1 includes mitochondrial fractions and Figure 2 identifies the mature 81–210 species. Selected original Results and Methods support mitochondrial localization while also reporting a cytosolic mature pool in particular cells. Processing products are not silently equated with the currently declared alternative splice isoforms.

Summary: PMID:21298097 analyzes human FXN in mitochondrial fractions and affinity-purified complexes in HeLa cells.

Reason: PMID:21298097 analyzes human FXN in mitochondrial fractions and affinity-purified complexes in HeLa cells. It also distinguishes mitochondrial-targeted human 81–210 from untargeted cytosolic 81–210 in a mouse fibroblast rescue setting. These are human protein constructs in a specified host, not evidence of a wrong-gene annotation. The original Results and Methods support the mitochondrial site of FXN-dependent assembly and viability, without making the recombinant complex a universal native stoichiometry.

Summary: This combined electronic assertion is consistent with the experimentally established mitochondrial pool of human FXN.

Reason: This combined electronic assertion is consistent with the experimentally established mitochondrial pool of human FXN. The original ARBA, UniProtKB, Ensembl, InterPro and subcellular-location source identifiers are preserved, without claiming that every individual propagation chain was re-audited. Human-cell fractionation and processing studies independently support the location. Additional cytosolic FXN in selected conditions does not contradict a mitochondrial annotation.

Summary: The accessible PMID:27519411 abstract concerns the human mitochondrial Fe-S assembly machinery and recombinant human FXN42–210 with NFS1, ISD11 and ISCU.

Reason: The accessible PMID:27519411 abstract concerns the human mitochondrial Fe-S assembly machinery and recombinant human FXN42–210 with NFS1, ISD11 and ISCU. Its engineered, bacterially produced oligomeric assembly is not itself a measurement of endogenous human-cell localization or native stoichiometry. The NAS mitochondrial assignment is nevertheless consistent with the directly established human mitochondrial FXN pool. Retain the location, while keeping the abstract-only source access and construct limits explicit.

Summary: The curator infers matrix localization from the mitochondrial [2Fe-2S] assembly complex.

Reason: The curator infers matrix localization from the mitochondrial [2Fe-2S] assembly complex. This is consistent with the independently established mitochondrial core ISC machinery. Preserve the IC inference and its original supporting term; the mixed human/fly reconstitution in PMID:29491838 is not itself a human cell-localization assay.

Summary: The literal ISS donor Q07540 is official Saccharomyces cerevisiae S288c Yfh1.

Reason: The literal ISS donor Q07540 is official Saccharomyces cerevisiae S288c Yfh1. Its complete current QuickGO response contains a mitochondrial-matrix IDA annotation, in addition to electronic and Reactome assertions. The exact donor record and its publication identifier are retained in source provenance. This verifies donor identity and experimental grounding under GO_REF:0000024, not a fresh reading of that donor paper. Human FXN’s role in the mitochondrial matrix assembly system independently supports the transferred location. Preserve the literal donor and distinguish this ISS from a PAINT/IBA inference.

Summary: PMID:17285345 is an authentic human-frataxin biochemical paper whose accessible abstract concerns recombinant 56–210 material and iron-related self-cleavage.

Reason: PMID:17285345 is an authentic human-frataxin biochemical paper whose accessible abstract concerns recombinant 56–210 material and iron-related self-cleavage. It was not treated as a newly inspected matrix-localization experiment. The NAS matrix assertion is retained because human FXN acts in the mitochondrial matrix Fe-S assembly system, supported independently by the read cellular assembly study and the curated Reactome complex. No natural isoform or protease mechanism is assigned from the background localization statement.

Summary: Reactome R-HSA-1362408 locates the FXN-containing NFS1/ISD11/ISCU assembly reaction in the mitochondrial matrix.

Reason: Reactome R-HSA-1362408 locates the FXN-containing NFS1/ISD11/ISCU assembly reaction in the mitochondrial matrix. This agrees with the core biochemical machinery. The localization can be retained independently of the source's explicitly alternative iron-donor and sulfur-transfer-activation models.

Summary: Reactome R-HSA-1362416 models frataxin iron binding in the mitochondrial matrix, consistent with the established site of core FXN function.

Reason: Reactome R-HSA-1362416 models frataxin iron binding in the mitochondrial matrix, consistent with the established site of core FXN function. Retaining location does not require endorsing a fixed iron-binding stoichiometry or an obligatory donor mechanism.

Summary: The Reactome reaction describing FXN and mitochondrial aconitase places FXN in the matrix, independently supported by the core ISC machinery.

Reason: The Reactome reaction describing FXN and mitochondrial aconitase places FXN in the matrix, independently supported by the core ISC machinery. Retain this location without treating the source's aconitase-repair model as the sole or defining FXN activity.

Summary: The matrix location is consistent with FXN's established mitochondrial assembly role.

Reason: The matrix location is consistent with FXN's established mitochondrial assembly role. The retained R-HSA-9854984 summary mostly discusses respiratory complex II maturation and does not isolate FXN's molecular step; therefore retain localization while avoiding an unsupported claim that FXN directly inserts all SDHB clusters.

Summary: Matrix localization is independently well supported for the core FXN assembly function.

Reason: Matrix localization is independently well supported for the core FXN assembly function. Retain the Reactome-linked location while keeping the more specific molecular attribution to downstream UQCRFS1 cluster insertion unresolved in the separate iron-chaperone row.

Summary: The official HPA FXN record reports supported cytosolic localization, including U-251MG and U2OS staining with antibody CAB022164, alongside a supported mitochondrial main location.

Reason: The official HPA FXN record reports supported cytosolic localization, including U-251MG and U2OS staining with antibody CAB022164, alongside a supported mitochondrial main location. The metadata and sample annotations were inspected; the images were not independently scored. Retain the cytosolic pool as contextual non-core localization, supported also by human fractionation studies, without converting a cell-line signal into universal localization or natural-product-specific function.

Summary: PMID:17468497 Figure 1 and the selected original Results/Methods describe mature human FXN in cytosolic as well as mitochondrial fractions, with endogenous and transfected material distinguished.

Reason: PMID:17468497 Figure 1 and the selected original Results/Methods describe mature human FXN in cytosolic as well as mitochondrial fractions, with endogenous and transfected material distinguished. The processing experiments identify mature 81–210 FXN; they do not establish a current alternative-splicing isoform assignment for every pool. Retain cytosolic localization as a supported contextual property. The source’s existence of a cytosolic mature species does not by itself prove a mitochondrial export mechanism.

Summary: The UniProt subcellular-location mapping to cytosol is supported by endogenous human-cell fractionation and localization studies, including differentiation-dependent Caco-2 results and mature cytosolic FXN in human cell lines.

Reason: The UniProt subcellular-location mapping to cytosol is supported by endogenous human-cell fractionation and localization studies, including differentiation-dependent Caco-2 results and mature cytosolic FXN in human cell lines. Retain this nonexclusive, contextual pool as non-core rather than denying it because mitochondrial assembly is the principal function. The mapping does not identify which currently declared natural splice product supplies every cytosolic signal.

Summary: PMID:15123683 directly reports iron delivery to human ferrochelatase and heme formation in a reconstituted system.

Reason: PMID:15123683 directly reports iron delivery to human ferrochelatase and heme formation in a reconstituted system. This is a reported contributory biochemical role, not merely a disease consequence or necessity argument. Retain it as non-core without claiming that FXN is the physiological heme iron donor in every setting.

Summary: The InterPro mapping is compatible with FXN-dependent iron handling.

Reason: The InterPro mapping is compatible with FXN-dependent iron handling. Human-cell depletion and rescue studies link loss of frataxin to altered iron distribution and iron-regulatory readouts, but do not isolate a dedicated homeostatic mechanism separate from its established role in Fe-S assembly. Retain intracellular iron homeostasis as a physiological consequence, rather than a second defining molecular function. The original IPR017789 source remains intact.

Summary: The authentic PMID:18160053 abstract describes FXN-silenced human HeLa cells, mitochondrial ferritin rescue, reduced cytosolic ferritin, and iron-handling abnormalities.

Reason: The authentic PMID:18160053 abstract describes FXN-silenced human HeLa cells, mitochondrial ferritin rescue, reduced cytosolic ferritin, and iron-handling abnormalities. Its full experimental text was not recovered; the attempted HAL route returned an access challenge. These observations support this broad homeostasis annotation in its perturbation context, with independent human-cell evidence from PMID:15509595. They do not establish a dedicated FXN homeostatic activity independent of Fe-S client maturation. Retain the annotation as non-core without treating the access limitation as a curator error.

Summary: The original PMID:15123683 abstract directly reports human holofrataxin interaction with human ferrochelatase, iron delivery and promotion of heme formation.

Reason: The original PMID:15123683 abstract directly reports human holofrataxin interaction with human ferrochelatase, iron delivery and promotion of heme formation. Preserve the conjunction specifying ferrochelatase input, mitochondrial location and heme synthesis. This is evidence for the reported reconstituted system, not for a second obligatory physiological mechanism or the NFS1-directed core activity.

Summary: NFS1-directed enzyme activation is a core FXN activity.

Reason: NFS1-directed enzyme activation is a core FXN activity. PMID:28634302 describes the human cysteine-desulfurase assembly and the structural rationale for FXN activation; its full assay details were not recovered here. Independent direct human sulfur-transfer and assembly measurements in PMID:24971490 and the mechanistic experiments in PMID:38627381 substantiate the function. Preserve the original NFS1 has_input extension and do not reinterpret FXN itself as the cysteine desulfurase.

Summary: PMID:38627381 directly compares core ISC complexes with and without FXN.

Reason: PMID:38627381 directly compares core ISC complexes with and without FXN. FXN binding rearranges the ISCU2 assembly site and accelerates specific transfer of the NFS1 persulfide to ISCU2 Cys138; the cysteine-desulfurase and scaffold carry the chemical intermediates. This provides a concrete enzyme-activation mechanism. Retain the NFS1 input extension; neither FXN-bound iron nor an absolute requirement for FXN in every reconstituted reaction is implied.

Summary: The ferrous-binding IBA has a verified ancestral placement and compatible human biochemical evidence.

Reason: The literal GO:0008198 IBD is at AN12/PTN000423706 on the actual PTHR16821 lineage to human FXN AN33/PTN002497669. Its fly, yeast and human descendant sources match the incoming WITH/FROM set; human self-membership is valid experimental grounding. Human iron-binding measurements support the property, retained as non-core because assay-dependent binding does not establish an obligatory physiological iron-donor role. No lineage loss is asserted.

Summary: The original PMID:12785837 abstract directly quantifies ferrous binding to human frataxin using iron-dependent fluorescence.

Reason: The original PMID:12785837 abstract directly quantifies ferrous binding to human frataxin using iron-dependent fluorescence. Retain the observed binding as non-core. Its assay-dependent capacity and affinity do not establish the physiological iron donor to mitochondrial ISCU.

Summary: PMID:15123683 directly describes iron-loaded human frataxin delivering ferrous iron to human ferrochelatase.

Reason: PMID:15123683 directly describes iron-loaded human frataxin delivering ferrous iron to human ferrochelatase. Retain the metal-binding property in that biochemical context without making iron delivery the universal mechanism of FXN function.

Summary: PMID:17285345 reports an iron-binding site in human frataxin residues 56-210 and changes following N-terminal self-cleavage.

Reason: PMID:17285345 reports an iron-binding site in human frataxin residues 56-210 and changes following N-terminal self-cleavage. This supports ferrous binding for the tested processed construct, with the physiological significance explicitly discussed rather than established. The processed fragment is not a natural alternative-product identifier.

Summary: PMID:29576242 directly measures ferrous binding to human FXN81-210 by NMR and iron quantification after desalting.

Reason: PMID:29576242 directly measures ferrous binding to human FXN81-210 by NMR and iron quantification after desalting. Its Results support binding and iron release under defined cysteine/reductant conditions. The Discussion explicitly does not prove that released iron enters ISCU or exclude other in vivo sources; retain the biochemical property without promoting that unresolved delivery model to the core.

Summary: Reactome R-HSA-1362416 models frataxin binding ferrous iron and also discusses its role in activating sulfur production.

Reason: Reactome R-HSA-1362416 models frataxin binding ferrous iron and also discusses its role in activating sulfur production. The retained source supports the binding assertion, while distinct experimental constructs and conditions give different stoichiometries. Keep this non-core and avoid presenting a fixed two-iron capacity as universal.

Summary: Ferric binding remains unresolved because positive reports conflict with controlled negative human NMR evidence.

Reason: The actual AN12/PTN000423706 placement and human descendant path remain valid, and the target appearing as an experimental donor is legitimate. The supporting human ferric-binding finding is disputed by the controlled human NMR result in PMID:29576242. Retain the IBA as UNDECIDED pending resolution of that experimental conflict, without asserting wrong ancestry, circular evidence or an evolutionary loss. Human FXN81-210 gave no significant ferric-induced NMR shifts with five-fold FeCl3 in PMID:29576242. The investigators observed no iron precipitation and oxidation of previously bound Fe2+ returned the spectrum toward iron-free FXN. These controls directly challenge a ferric-binding interpretation; an assay-condition explanation is not established. Older reported positive measurements and this negative result remain in conflict because the older full experimental conditions were not recovered.

Summary: Ferric binding remains unresolved because positive reports conflict with controlled negative human NMR evidence.

Reason: The authentic original abstract reports ferric-binding fluorescence and calorimetric measurements. Human FXN81-210 gave no significant ferric-induced NMR shifts with five-fold FeCl3 in PMID:29576242. The investigators observed no iron precipitation and oxidation of previously bound Fe2+ returned the spectrum toward iron-free FXN. These controls directly challenge a ferric-binding interpretation; an assay-condition explanation is not established. Older reported positive measurements and this negative result remain in conflict because the older full experimental conditions were not recovered. Keep the IDA assertion UNDECIDED; neither dismiss the original observation nor treat ferric binding as an established conditional property.

Summary: Ferric binding remains unresolved because positive reports conflict with controlled negative human NMR evidence.

Reason: The authentic original abstract reports mutant-dependent iron binding and precipitation, but it does not resolve the ferric-specific assay conditions; the full original was inaccessible. Human FXN81-210 gave no significant ferric-induced NMR shifts with five-fold FeCl3 in PMID:29576242. The investigators observed no iron precipitation and oxidation of previously bound Fe2+ returned the spectrum toward iron-free FXN. These controls directly challenge a ferric-binding interpretation; an assay-condition explanation is not established. Older reported positive measurements and this negative result remain in conflict because the older full experimental conditions were not recovered. Keep the IDA assertion UNDECIDED rather than using the other disputed positive report as confirmation.

Summary: Ferric binding remains unresolved because positive reports conflict with controlled negative human NMR evidence.

Reason: The family mapping is traceable, but family membership does not resolve the conflicting human ferric-binding experiments. Human FXN81-210 gave no significant ferric-induced NMR shifts with five-fold FeCl3 in PMID:29576242. The investigators observed no iron precipitation and oxidation of previously bound Fe2+ returned the spectrum toward iron-free FXN. These controls directly challenge a ferric-binding interpretation; an assay-condition explanation is not established. Older reported positive measurements and this negative result remain in conflict because the older full experimental conditions were not recovered. Keep the electronic assertion UNDECIDED; no family-wide loss or universal refutation is inferred.

Summary: PMID:18424449 directly tested iron depletion and supplementation in human HEK293 cells, together with patient-derived cells.

Reason: PMID:18424449 directly tested iron depletion and supplementation in human HEK293 cells, together with patient-derived cells. Its Results and Figures 1–5 describe iron-dependent FXN transcript/protein regulation, including DFO-sensitive promoter reporters and prolonged protein reduction. The official response-to-iron definition includes changes in expression or activity, so absence of an autonomous iron-sensor assay does not invalidate this term. Retain the experimentally supported response as contextual rather than a defining FXN activity.

Summary: The ancestral assembly annotation is correct but broader than the established initial-cluster process.

Reason: The GO:0016226 IBD at the actual family root AN0/PTN004495696 is on the human target lineage, with all literal descendant sources retained, including human FXN. The function is supported; the existing GO:0044571 [2Fe-2S] cluster assembly term is its verified direct is_a child and describes the human core step more specifically. TOO_GENERAL concerns present annotation specificity, not rejection of this deep ancestral placement or propagation.

Summary: PMID:12785837 explicitly describes human frataxin-supported [2Fe-2S] assembly on ISU.

Reason: PMID:12785837 explicitly describes human frataxin-supported [2Fe-2S] assembly on ISU. Thus the broad iron-sulfur cluster assembly annotation is supported but less informative than the existing [2Fe-2S] assembly term. This judgment concerns specificity, not rejection of the experimental evidence.

Summary: The InterPro Fe-S assembly inference is biologically sound, but the existing [2Fe-2S] cluster assembly annotations describe FXN's experimentally demonstrated role in the core mitochondrial assembly machinery more specifically.

Reason: The InterPro Fe-S assembly inference is biologically sound, but the existing [2Fe-2S] cluster assembly annotations describe FXN's experimentally demonstrated role in the core mitochondrial assembly machinery more specifically. The broader term is not a separate core function.

Summary: PMID:29097656 supports the core ISC machinery through structural and functional experiments, including explicitly distinguished human and Chaetomium thermophilum systems.

Reason: PMID:29097656 supports the core ISC machinery through structural and functional experiments, including explicitly distinguished human and Chaetomium thermophilum systems. The broad process is correct, while existing [2Fe-2S] assembly annotations better capture the core step. Do not describe the fungal SAXS FXN subcomplex as an all-human structure.

Summary: Reactome R-HSA-1362409 places FXN in mitochondrial Fe-S biogenesis and distinguishes initial ISCU [2Fe-2S] formation from subsequent larger-cluster maturation.

Reason: Reactome R-HSA-1362409 places FXN in mitochondrial Fe-S biogenesis and distinguishes initial ISCU [2Fe-2S] formation from subsequent larger-cluster maturation. The existing [2Fe-2S] assembly term is the more informative core assignment; the broad pathway term does not establish direct participation in every downstream transfer.

Summary: The accessible PMID:12785837 abstract concerns recombinant human frataxin, iron donation, and ISU-type cluster assembly; it does not expose the experiment underlying this autoprocessing annotation.

Reason: The accessible PMID:12785837 abstract concerns recombinant human frataxin, iron donation, and ISU-type cluster assembly; it does not expose the experiment underlying this autoprocessing annotation. The full original could not be recovered through the attempted publisher routes. Autoprocessing means processing performed by the protein itself: mitochondrial processing peptidase cleavage and the presence of mature FXN fragments do not independently demonstrate it. Preserve this experimental annotation unresolved, without declaring the paper miscited or the activity absent. The separate self-cleavage claim in the PMID:17285345 abstract is not a substitute for the missing assay context.

Summary: The narrower retained assertion GO:0006879, intracellular iron ion homeostasis, is an official direct is_a child of this monoatomic-cation homeostasis term.

Reason: The narrower retained assertion GO:0006879, intracellular iron ion homeostasis, is an official direct is_a child of this monoatomic-cation homeostasis term. FXN-specific iron-handling evidence supplies the relevant scope; this broader electronic annotation adds no distinct cation process. The narrower assertion is retained as non-core because the physiological homeostasis effects do not establish an additional dedicated molecular mechanism.

Summary: The official Reactome pathway R-HSA-9865881 includes the UQCRFS1 Fe-S insertion event R-HSA-9866272.

Reason: The official Reactome pathway R-HSA-9865881 includes the UQCRFS1 Fe-S insertion event R-HSA-9866272. Its modeled complex R-HSA-1362398 explicitly includes human FXN(81–210), so FXN is genuinely represented in the pathway. Retain the pathway-specific assembly consequence as non-core. The broader primary evidence establishes FXN participation in Fe-S cluster production, not a separately demonstrated FXN-catalyzed transfer directly into UQCRFS1. The component-detail endpoint was unavailable, and no current natural splice-product identifier or direct downstream transfer mechanism is inferred from the complex model.

Summary: The inherited iron-chaperone annotation is retained with a conditional biochemical interpretation.

Reason: The exact GO:0034986 IBD at AN12/PTN000423706 is on the human target path and preserves fly, yeast and human experimental descendant sources. Human holofrataxin-to-ferrochelatase iron transfer in PMID:15123683 supports the reported capability. Keep this non-core: neither the valid ancestral placement nor that experiment establishes FXN as the obligatory physiological iron donor to ISCU. No self-donor or donor-count objection applies.

Summary: PMID:15123683 directly reports delivery of iron by human holofrataxin to human ferrochelatase.

Reason: PMID:15123683 directly reports delivery of iron by human holofrataxin to human ferrochelatase. An iron-chaperone description is appropriate for that experiment. Retain it as non-core rather than generalizing it to an obligatory iron-donor mechanism in mitochondrial Fe-S assembly, where activation of sulfur transfer has stronger mechanistic support.

Summary: Reactome R-HSA-9866272 attributes insertion of a preformed [2Fe-2S] cluster into UQCRFS1 to an FXN-containing transfer complex.

Reason: Reactome R-HSA-9866272 attributes insertion of a preformed [2Fe-2S] cluster into UQCRFS1 to an FXN-containing transfer complex. The retained summary alone does not establish that FXN itself chaperones iron in this downstream step, and its cited detailed experiments were not inspected here. Keep this specific source-to-function claim unresolved; neither general Fe-S dependence nor the source's catalyst wording establishes the molecular role of each subunit.

Summary: Selected original PMID:16608849 Methods, Results and Figures 1–4 and 7 report increased ceramide-, GD3- and staurosporine-associated apoptosis in patient lymphoblasts and protection by transfected human FXN.

Reason: Selected original PMID:16608849 Methods, Results and Figures 1–4 and 7 report increased ceramide-, GD3- and staurosporine-associated apoptosis in patient lymphoblasts and protection by transfected human FXN. Full-length 1–210 and engineered 56–210 constructs were distinguished; protection did not extend to every tested stimulus, including UVB. Retain the supported negative regulation of apoptosis as a cellular-context effect. It is not evidence for a universal anti-apoptotic activity or for assigning the engineered fragment to a current natural isoform.

Summary: PMID:24971490 directly demonstrates that human FXN promotes sulfur incorporation into NFS1 and ISCU2 and supports enzymatic [2Fe-2S] assembly.

Reason: PMID:24971490 directly demonstrates that human FXN promotes sulfur incorporation into NFS1 and ISCU2 and supports enzymatic [2Fe-2S] assembly. Radiolabel tracking and CD assays distinguish protein-bound intermediates from nonspecific iron/sulfide chemistry. This is direct participation by an activating assembly component, not merely a knockout phenotype.

Summary: Despite its Drosophila-centered title, PMID:29491838 explicitly tests human FXN with purified fly Nfs1Δ1-53/Isd11/IscU.

Reason: Despite its Drosophila-centered title, PMID:29491838 explicitly tests human FXN with purified fly Nfs1Δ1-53/Isd11/IscU. Section The Drosophila Nfs1/Isd11/IscU/frataxin complex and Figure 3 report larger complexes on addition of human FXN and increased desulfurase output. Retain the curator's assembly annotation with this mixed-species assay scope; it is corroborated by direct human reconstitution in PMID:24971490.

Summary: PMID:29576242 directly investigates human FXN in a reconstituted cysteine-desulfurase/ISCU system and shows support for Fe-S cluster formation using iron-loaded FXN under defined conditions.

Reason: PMID:29576242 directly investigates human FXN in a reconstituted cysteine-desulfurase/ISCU system and shows support for Fe-S cluster formation using iron-loaded FXN under defined conditions. Retain the [2Fe-2S] assembly process while separating the demonstrated assembly role from the unproven conclusion that FXN is the exclusive physiological iron donor.

Summary: PMID:38627381 combines structural snapshots with rapid persulfidation assays to show how FXN promotes the sulfur-transfer step of de novo [2Fe-2S] assembly on ISCU2.

Reason: PMID:38627381 combines structural snapshots with rapid persulfidation assays to show how FXN promotes the sulfur-transfer step of de novo [2Fe-2S] assembly on ISCU2. FXN contributes the conformational activation that accelerates this chemical step; it is not the cluster scaffold or the electron donor. This is a core process.

Summary: The human acts_upstream_of assertion remains unresolved at the specific [4Fe-4S] linkage.

Reason: The original qualifier is acts_upstream_of, not involved_in. The retained PMID:29576242 Results support human FXN-assisted initial [2Fe-2S] assembly, consistent with an upstream role in later cluster pathways. They do not identify the specific downstream [4Fe-4S] linkage represented by this annotation sufficiently to resolve its source-specific attribution here. Keep UNDECIDED without denying upstream biology or requiring FXN itself to synthesize the final cluster. The separate mouse involved_in experiments supporting the ortholog transfers are assessed independently and do not replace this human annotation's original reference or qualifier.

Summary: The mouse experimental chain supports a bounded [4Fe-4S] assembly role transferred to human FXN.

Reason: The literal mouse O35943 donor record includes an independent IDA involved_in annotation to PMID:25597503, separately from an ISO acts_upstream_of row transferred from human Q16595; the chain is not merely circular propagation. The authentic cached abstract of PMID:23265191 reports formation of an [Fe4S4] cluster within the quaternary ISCU/NFS1/ISD11/FXN complex in the presence of iron and L-cysteine, followed by transfer to mammalian aconitase to generate active enzyme. It also reports that the ternary complex can assemble an Fe-S cluster without FXN, but transfers that cluster inefficiently to aconitase. These positive complex-assembly and transfer observations, together with the independently annotated mouse donor, support keeping the ortholog-transferred involvement as non-core. This does not establish cluster binding to isolated FXN, a direct physiological human [4Fe-4S] synthesis mechanism, or the distinct source attribution of the human acts_upstream_of annotation.

Summary: The mouse experimental chain supports a bounded [4Fe-4S] assembly role transferred to human FXN.

Reason: The literal mouse O35943 donor record includes an independent IDA involved_in annotation to PMID:25597503, separately from an ISO acts_upstream_of row transferred from human Q16595; the chain is not merely circular propagation. The authentic cached abstract of PMID:23265191 reports formation of an [Fe4S4] cluster within the quaternary ISCU/NFS1/ISD11/FXN complex in the presence of iron and L-cysteine, followed by transfer to mammalian aconitase to generate active enzyme. It also reports that the ternary complex can assemble an Fe-S cluster without FXN, but transfers that cluster inefficiently to aconitase. These positive complex-assembly and transfer observations, together with the independently annotated mouse donor, support keeping the ortholog-transferred involvement as non-core. This does not establish cluster binding to isolated FXN, a direct physiological human [4Fe-4S] synthesis mechanism, or the distinct source attribution of the human acts_upstream_of annotation.

Summary: PMID:20053667 reports interaction of extramitochondrial human FXN with aconitase/IRP1 and restoration of cytosolic aconitase activity in FRDA cells.

Reason: PMID:20053667 reports interaction of extramitochondrial human FXN with aconitase/IRP1 and restoration of cytosolic aconitase activity in FRDA cells. Retain this positive regulation of lyase activity as a non-core contextual effect. It is distinct from FXN's activation of mitochondrial NFS1 and does not establish that FXN catalyzes the aconitase reaction.

Summary: The inherited node is valid, but the specific FXN-bound cluster evidence remains unresolved.

Reason: The literal GO:0051537 IBD at AN12/PTN000423706 lies on the verified human FXN lineage, and its human experimental descendant is legitimate rather than circular. However, the accessible PMID:12785837 abstract describes cluster formation on ISU after iron delivery, not an isolated cluster bound by FXN; its full text was inaccessible. The molecular interpretation therefore remains unresolved. This is not evidence against the node placement, a target-specific loss or grounds to remove the original IBA.

Summary: The accessible original PMID:12785837 abstract describes a [2Fe-2S] cluster formed on ISU after iron delivery by FXN.

Reason: The accessible original PMID:12785837 abstract describes a [2Fe-2S] cluster formed on ISU after iron delivery by FXN. It does not establish that FXN itself binds the assembled cluster, and the original full text was inaccessible. Preserve the experimental GOA assertion as unresolved rather than inferring a wrong-gene annotation from this incomplete view.

Summary: Protein maturation here means attainment of functional Fe-S client proteins, not FXN serving as a substrate for cleavage.

Reason: Protein maturation here means attainment of functional Fe-S client proteins, not FXN serving as a substrate for cleavage. In PMID:15509595, repeated FXN RNA interference in human HeLa cells reduced aconitase and succinate-dehydrogenase activities and altered the IRP1/IRE readout, while non-Fe-S enzyme controls were spared and recovery followed restoration of frataxin. Together with the established assembly-complex role, this supports FXN doing work needed for Fe-S client maturation. The selected original Results and Methods were read; no specific proteolytic or self-processing mechanism is inferred.

Summary: The authentic PMID:15641778 abstract describes purified human frataxin expressed in bacteria, iron-induced assembly, and attenuation of iron-dependent radical chemistry.

Reason: The authentic PMID:15641778 abstract describes purified human frataxin expressed in bacteria, iron-induced assembly, and attenuation of iron-dependent radical chemistry. The available evidence does not expose the exact experiment supporting a cellular response to hydrogen peroxide, and full original text was not recovered. A purified biochemical protective effect alone does not establish the cellular process, but incomplete access is not grounds to remove an experimental curator assertion or declare that no cellular assay exists. Keep this context unresolved.

Summary: PMID:16608849 directly measured cytochrome-c release by flow cytometry after selective permeabilization, with mitochondrial controls.

Reason: PMID:16608849 directly measured cytochrome-c release by flow cytometry after selective permeabilization, with mitochondrial controls. Figure 2 reports greater staurosporine-induced release in FXN-deficient patient lymphoblasts; the paper also tests protection and rescue with human FXN constructs. Retain this mitochondrial apoptotic endpoint as a supported context-dependent regulatory effect, distinct from the core Fe-S assembly activity. Do not infer a direct FXN–cytochrome-c binding mechanism or universal protection against all death stimuli.

Summary: The retained GO:0006879 intracellular iron ion homeostasis assertion is an official direct is_a child of inorganic ion homeostasis.

Reason: The retained GO:0006879 intracellular iron ion homeostasis assertion is an official direct is_a child of inorganic ion homeostasis. It captures the specific iron-handling consequence supported for FXN, while this ARBA-derived ancestor adds no distinct ion process. This granularity judgment does not promote the physiological homeostasis effect to a separate defining molecular function.

Summary: PMID:21298097 reports mammalian frataxin interaction with the preformed ISCU/NFS1/ISD11 assembly complex.

Reason: PMID:21298097 reports mammalian frataxin interaction with the preformed ISCU/NFS1/ISD11 assembly complex. This is consistent with later human FXN-bound structures. Retain membership in the mitochondrial [2Fe-2S] assembly complex without conflating complex membership with direct contact to every constituent.

Summary: PMID:29491838 directly adds HUMAN frataxin to the purified Drosophila Nfs1/Isd11/IscU complex and observes its association by size-exclusion chromatography; the original Figure 3 caption identifies the human protein.

Reason: PMID:29491838 directly adds HUMAN frataxin to the purified Drosophila Nfs1/Isd11/IscU complex and observes its association by size-exclusion chromatography; the original Figure 3 caption identifies the human protein. Keep the experimentally grounded complex association with its mixed-species context, corroborated by human structural studies.

Summary: PMID:31101807 resolves FXN bound to the active human NFS1/ISD11/ISCU assembly, with bacterial acyl-carrier protein retained from recombinant expression.

Reason: PMID:31101807 resolves FXN bound to the active human NFS1/ISD11/ISCU assembly, with bacterial acyl-carrier protein retained from recombinant expression. FXN contacts NFS1 and ISCU in the observed structure. This directly supports the specific complex assignment; lack of one contact in this structure does not refute interactions measured under other conditions.

Summary: The orthology-derived mitochondrial [2Fe-2S] assembly-complex assignment agrees with direct human interaction and structural evidence in PMID:21298097 and PMID:31101807.

Reason: The orthology-derived mitochondrial [2Fe-2S] assembly-complex assignment agrees with direct human interaction and structural evidence in PMID:21298097 and PMID:31101807. FXN is an activating component with dynamic occupancy, not necessarily a constitutive stoichiometric subunit in every intermediate.

Summary: The broad iron-sulfur cluster assembly-complex annotation is supported by PMID:29097656 and related primary work.

Reason: The broad iron-sulfur cluster assembly-complex annotation is supported by PMID:29097656 and related primary work. The existing mitochondrial [2Fe-2S] assembly-complex term specifies the core entity more precisely. Human crystallographic preparations and fungal FXN-containing SAXS complexes are distinguished in the source guide.