FAS/CD95 is a cell-surface death receptor that recruits FADD and caspase8 into the CD95 death-inducing signaling complex. Receptor activation supports apoptosis; accessory interactions and conditional responses require their own assay context. The binding record table preserves exact imported partners and the basis of each decision. Supported generic interactions are retained as non-core under the ClinGen project policy.
Death-receptor activity is the appropriate molecular term. GO:0005031 specifically requires TNF cytokine binding, whereas FAS recognizes FASL. Antibody stimulation in PMID:12221075(https://pubmed.ncbi.nlm.nih.gov/12221075/) and the original APO-1 purification/transfection paper establishes receptor-dependent apoptosis without being a TNF-binding assay. The latter's Methods, BL60-P7 Results and Figures6/7 were read. Broad DISC annotations are refined to the CD95-specific complex where the actual experiments isolate it.
Interaction assays have distinct limits. Co-immunoprecipitation and affinity purification support association in a complex, not necessarily direct contact between two purified proteins. A cellular-component DISC term is not a replacement molecular function for binding. Partner kinase, helicase, phosphatase or ubiquitin-ligase activities are not transferred to FAS. Human receptor experiments in mouse host cells and mouse proteins used in structural or interaction assays are identified separately.
The original TNFR1 paper includes explicit FAS controls in Figure2C and Results: Flag-FASL isolates receptor complexes with FADD/caspase8 in human HT1080 cells. Its title does not invalidate those assays. The palmitoylation paper, original PDF pages2/4/9/10 (printed222/224/229/230), likewise contains positive human FAS–FADD/caspase8 complex assays in Figure1 and depletion/gradient analysis in Figure4. These pages were visually read; neither paper establishes purified binary FAS–caspase8 binding.
The HeLa proximity-ligation supplement provides actual FAS–DAXX, FAS–FASLG and FADD–FAS positive entries in TableS1 pages4/8 and exact assay antibodies in TableS7 pages76/89. FiguresS1/S2 describe thresholds and controls; both target antibody-table pages were visually inspected. The main Methods body was not recovered. PLA establishes cellular proximity, without proving a novel enzymatic function or the database-derived pathway cross-talk labels.
BioPlex293T AP-MS directly contains FAS–FADD and FAS–CASP8 pairs. The original Methods aggregate proteins to Entrez gene identifiers, so the displayed P25445-6 suffix is not proof that a natural soluble FAS isoform was functionally assayed. The leukocyte interaction screen independently contains reciprocal FAS/FASLG matrix entries with verified construct joins; a cited1.6nM affinity is prior literature, not a new target SPR assay in that screen.
Raft evidence describes a dynamic receptor pool. Human BEAS-2B fractionation/co-IP in PMID:21382479(https://pubmed.ncbi.nlm.nih.gov/21382479/) and human SW480 surface-flow, biotinylation and detergent-resistant fractions in PMID:25301068(https://pubmed.ncbi.nlm.nih.gov/25301068/) do not imply exclusive raft localization. FAS is a substrate in the palmitoylation experiments, not an acyltransferase. Engineered glycosylation/palmitoylation mutations are not natural isoform annotations.
Conditional outputs remain bounded. Glutamine deprivation changes FAS-triggered ASK1/JNK and apoptotic responses in human HeLa cells (PMID:11096076(https://pubmed.ncbi.nlm.nih.gov/11096076/)); this is not direct nutrient sensing by FAS. Caspase-inhibited FAS-triggered death and primary RIP-null/rescue evidence (PMID:16507998(https://pubmed.ncbi.nlm.nih.gov/16507998/), PMID:11101870(https://pubmed.ncbi.nlm.nih.gov/11101870/)) support a conditional necroptotic signaling context, without an assayed FAS–RIPK3–MLKL complex. FAS-peptide SPR and the 3EWT primary structure support calmodulin binding; inaccessible main Methods limit whole-receptor stoichiometry and regulation claims.
Specific unresolved contexts are retained honestly. The 18328427 repository document contains a summary, not full Methods/Results. SRC/P12931 is distinct from YES1/P07947; exact SRC/FADD/caspase8 assays remain unverified. The exact positive human FAS–FAS two-hybrid record EBI-494924 for PMID:7536190 supports ACCEPT for the gene-level self-association claim, together with independent FAS evidence in PMID:21382479 and PMID:7538908. The operative paper experiment and figure pixels remain unreadable. The inherited self-association quotation describes FADD and is preserved as an inherited review anchor, not as support for this FAS–FAS row; this does not establish an IntAct source defect. The B-cell-exosome paper's main Methods were read, but the FAS-specific supplemental proteomic entry was inaccessible; FASN is not FAS. The complete cached PMID:19593445 main article does not supply the imported FAS mechanical-stimulus assay, so the citation mismatch is flagged without inventing a replacement.
Soluble-product studies and old quotations require contextual reading. Accessible originals support apoptosis antagonism for specified human soluble FAS products, but do not establish cytosolic localization or identical activity of every alternative product. The complete original PMID:7575433 scans distinguish observed transcripts from undetectable minor proteins and predicted ligand-binding consequences. PMID:8648105 reports inhibition of antibody- and ligand-induced death in its abstract, without proving sequestration. PMID:8598453 concerns a deletion-insertion mutation in resistant HUT78 clones. Current UniProt sequence features show that the isoform6 deletion largely removes the membrane segment while retaining the death domain; isoform7 retains the membrane segment while losing the death domain. These topology facts are not measured functions. Product descriptions and the core summary use the reviewed successor objects. Inherited quotations retain their original positions and are not silently rewritten into current conclusions.
The whole-reference assessments specify actual Methods/Results, captions, abstract-only or supplement-only access for every cited direct source. Missing figure pixels and unavailable target supplements are stated explicitly. Genuine original recovery supplements the normal publication cache without altering its bytes.
Database provenance covers 20 reference groups across 33 current contexts and four historical source contexts. Whole-reference assessments retain both primary-evidence and database-source scopes.
The actual PTHR46874 v19 tree and alignment place human P25445 at PTN002508085 below PTN002930185. The ten-node target path, nine exact IBD records and absence of family NOT records are preserved in the accepted intake proof. Human self-WITH is expected experimental grounding. Neither donor count nor self-WITH was used as a negative criterion.
The inherited death-receptor mechanism is retained, with the overly specific literal TNF-cytokine activity refined to death receptor activity (GO:0005035). The corresponding electronic BP premise is refined to Fas signaling pathway (GO:0036337). Actual human anti-CD95 experiments support these decisions; TNF receptor family membership alone does not establish TNF cytokine binding PMID:12221075.
Human FAS/FADD assembly and signaling support DISC membership; the crystallized complex in PMID:20935634 is mouse FAS-DD/human FADD-DD, while human complexes and human cellular assays were also examined. No one oligomeric stoichiometry is imposed on native FAS.
Raft association is supported by human fractionation, with dynamic redistribution and a substantial non-raft pool [PMID:21382479; PMID:25301068]. External cell-surface exposure is supported by intact-cell labeling and cold surface biotinylation PMID:25301068. HPA supports plasma-membrane staining with HPA027444 in A-431 and U-251 MG; its additional cytosol/nuclear-body records and lack of natural-isoform resolution are retained as assay limits.
Conditional outputs remain non-core: ASK1/JNK activation after FAS ligation in glutamine-deprived HeLa cells PMID:11096076; soluble-product inhibition [PMID:7595210; PMID:7510905]; injury-associated mouse motoneuron apoptosis PMID:13679421; and necroptotic signaling under permissive conditions [PMID:21052097; PMID:16507998]. PMID:21052097 specifically uses human FAS expressed in mouse L929sA cells. The inherited FADD quotation does not prove a motor-neuron role, and the inherited RIP-binding quotation does not alone prove necroptosis.
The original murine FasB experiment was recovered at Methods/Results/Figure 2 caption level: FasB-expressing P815 cells and their conditioned medium resist FasL-dependent killing while surface FasA remains PMID:7595210. The mouse splice product differs in architecture from human soluble products. No direct ligand-sequestration assay, equivalent function of every human isoform, or universal endogenous mechanism is inferred from that experiment.
The actual InterPro target matches, current InterPro2GO mappings and all four ARBA JSON rules were inspected. The current DISC and extrinsic-apoptosis rule clauses match the target FunFam and Eukaryota. The historical cell-surface and apoptosis-regulation rule clauses still match the target, so current GOA absence is not proof of rule withdrawal.
All 12 FAS Reactome source events reach P25445 entity 66248 at the plasma membrane. Every selected route was replayed against the saved raw nested input/output/component/member entities. The TP53-expression event reaches FAS as its output; the location annotation does not attribute transcription-factor activity to FAS. Processing events containing FAS within a larger DISC complex are likewise valid CC sources even when their titles foreground another protein.
The historical keyword calmodulin-binding source has positive peptide-level evidence: the original FAS-peptide SPR caption and RCSB 3EWT support binding, but the article body was inaccessible and full-receptor stoichiometry or universal regulation is not established PMID:24914971. Its source pipeline is now documented as virus-only. The historical combined apoptosis source has a current InterPro counterpart. These are source-lifecycle questions, distinct from the positive biological judgments. Historical source records and their evidence remain retained. Six historical contexts are marked retired from the active set only for the pinned current GOA snapshot; that marker does not assert biological invalidity or official withdrawal.
The official record enumerates seven product sequences; the FasExo8Del functional evidence is specifically from mutation-bearing lymphoma clones. Isoform 6 deletes residues 169–189, overlapping most of the annotated 174–190 membrane segment, while retaining the death-domain sequence; isoform 7 retains the transmembrane sequence while losing the death domain. Sequence architecture is not a measured cellular-localization or functional assay. The reviewed product and core successor objects retain all seven product identities and sequence notes. They do not claim that only isoform 1 is membrane-containing or that all soluble products lose the death domain.
Rat donor PMIDs 16154149, 17195944 and 16493077 and mouse donor PMIDs 7602121, 12702495 and 8640868 were identifier/abstract-verified, not fully read. The human AICD corroboration PMID:7530335 was read at original publisher-abstract level. No experimental annotation was removed because of these limits. For these database-source read scopes, HPA image pixels were inspected and publication figure captions were read where documented; publication figure pixels and supplements were not independently inspected. The primary-evidence scopes above identify the additional visual and supplemental reads.
Exact selected normal-source bytes remain unchanged. Whole-reference assessments preserve complete primary-evidence and database-source scopes, with the provenance and access limits recorded in the associated notes.
All seven product identities and sequence notes are retained. The canonical precursor contains 335 residues; splice products 2–7 contain 103, 86, 149, 132, 314 and 220 residues. The official sequence edits agree between the preserved UniProt cache and the fresh sequence records. Transcript structures and assay limitations come from PMID:7575433(https://pubmed.ncbi.nlm.nih.gov/7575433/); inhibition by soluble products is bounded to the assays reported in PMID:8648105(https://pubmed.ncbi.nlm.nih.gov/8648105/); the membrane-containing, death-domain-lacking FasExo8Del variant was studied in lymphoma cells in PMID:8598453(https://pubmed.ncbi.nlm.nih.gov/8598453/). The first paper’s scanned pages were read, while the latter two assessments use their original abstracts. Gene-level annotation and engineered constructs do not establish natural isoform-specific function.
The database decision table complements the binding table above. Both tables preserve their source-index namespaces; fresh and historical indices are distinct.
The current QuickGO death-receptor record places death receptor activity below transmembrane signaling receptor activity. The broad InterPro receptor annotation is therefore TOO_GENERAL while the supported death-receptor replacements remain. Both broad signal-transduction records are likewise TOO_GENERAL because the retained Fas signaling pathway supplies the specific positive process. These actions do not invalidate the original broad assertions.
The soluble-isoform source explicitly introduces Fas/Apo-1 as an apoptosis-signaling cell-surface antigen. The imported NAS row has no product restriction, so its gene-level receptor assertion is refined to death receptor activity. The abstract's soluble-transcript, COS-expression and inhibition observations remain distinct: they do not establish signaling by every soluble FAS product. Full operative Methods/Results remain unavailable, and no natural isoform or product-specific experiment is guessed.
Live QuickGO searches for Fas, CD95 and FASLG and the death-receptor child list were checked on 2026-10-26. The list includes TNF and TRAIL receptor activities but no equivalent FASLG-specific child. The proposed ontology term is a request to express that ligand specificity; it has no invented identifier and adds no NEW annotation. Plasma membrane is retained in the concise core location summary. Positive raft annotations and the dynamic, nonexclusive receptor-pool evidence above remain unchanged.
Six historical source contexts retain their source objects, quotations and scientific judgments. Their retired marker means absence from the active GOA snapshot with SHA-256 24cd034e13a76b4e1d6e39bed1f87d29313a51cd288a28087dc10bc713b2a1dd; it is not a biological negation or an assertion of official withdrawal. The two authored tables now include each complete source object, so term, evidence, reference, qualifier, partner and optional source fields can be joined directly without relying on a private index.
Inherited quotation positions are retained under the project preservation instruction. A title or citation-metadata anchor does not establish an assay, and the inherited glioblastoma invasion sentence does not prove the unresolved CASP8 interaction. Such limitations are described in the scientific reasons; no replacement quotation is manufactured. The current notes distinguish sequence architecture from measured function and bound the isoform questions to the original expression and tumor-cell contexts.
The two revised actions use the actual assertion scope: the NAS receptor row is gene-level, and the self-binding row has an exact positive FAS–FAS interaction record. These decisions do not turn unreadable primary experiments into independently verified assays. The two decision tables cover generic protein-binding and database contexts; neither contains annotation index 43 or 95, so their existing source joins remain unchanged.
The proposed FASLG-specific receptor term additionally cites the already assessed reciprocal recombinant-ectodomain screen PMID:35922511(https://pubmed.ncbi.nlm.nih.gov/35922511/). The primary matrix and construct metadata identify human FAS/P25445 and FASLG/P48023. Secondary-matrix validation and a newly measured target affinity are not claimed.
The description now uses the experimentally supported ASK1/JNK output in glutamine-deprived human HeLa cells PMID:11096076(https://pubmed.ncbi.nlm.nih.gov/11096076/). This is an apoptotic assay context, not a claim of non-apoptotic signaling or direct nutrient sensing. It avoids making the less-resolved glioblastoma partner assays the sole additional exemplar. Inherited quotation and journal content remain intact.