The normal imported record is reviewed human CD27/TNFRSF7, UniProt P26842 (secondary B2RDZ0), 260 residues, sequence version 2 and CRC64 43C38B895B2AD4E2; HGNC:11922, NCBI Gene:939 and ENSG00000139193.5 agree. The frozen association is autosomal recessive lymphoproliferative syndrome 2, MONDO:0014054, with Definitive validity. This disease classification is context rather than a substitute for functional assays. Normal UniProt supplies no alternative-products block, so no product records are manufactured from the differing RefSeq isoform a/b numbering.
All 43 source assertions are retained with original term labels, evidence, references, qualifiers and WITH/FROM entities. Forty-four raw GOA rows collapse to 43 unique source objects: the repeated PMID9177220/SIVA1 protein-binding row differs only in date/assigned-by metadata. The two PMID16683188 SIVA1 versus SIVA1-1 partner edges remain distinct. Source caches, UniProt, GOA, family-export quarantine and frozen data are unchanged. ROOT's authenticated source import created 12 files and preserved 146 preimages; authorship occurs separately.
The documented normal deep-research wrapper attempted falcon and then the configured perplexity-lite fallback. Both terminated because uv could not obtain the deep-research-client dependency through the unavailable DNS/network route. No provider result was produced or synthesized under a provider filename. This review therefore rests on manual source reading, not a purported deep-research report. The actual initial and terminal tool envelopes are retained in the proposal evidence directory.
All 17 original cached publication headers and complete abstracts and the complete human Reactome event summary were read. Full-text availability is recorded independently of what was actually inspected. For PMID28011863, selected complete Results, Methods and captions were read from the immutable XML-derived cache, including human CD27-deficient patient controls and ligand rescue/CRISPR experiments. For PMID12324477, the canonical cache is abstract-only; selected original Methods, Results and captions were read in the author-uploaded JBC reproduction, DOI:10.1074/jbc.M209050200, public original reproduction. For PMID9177220, official indexed original captions and selected Methods context were inspected, separately from the abstract-only cache. No figure-pixel measurement, supplementary target-table reconstruction or exhaustive whole-paper claim is made for these accesses.
Existing normal cached PMID38354704 was read completely at the header/abstract level and supplies bounded corroboration of CD27 internalization, TRAF2/SHP1 recruitment and T-cell memory programming. Its full Methods were not inspected. Normal publication-only retrieval supplied PMID34419446 (human CD27–CD70 structure; full XML), PMID9582383 (TRAF2/5 signaling; abstract-only) and PMID15485626 (receptor-specific NIK/p100 signaling; abstract-only). The authenticated transport and exclusive source import preserved all existing inputs. Complete headers/abstracts and selected original structural-paper sections were read; full Methods for the two older signaling papers remain inaccessible. All original source objects are preserved.
CD27's central activity is recognition of CD70 and transduction of a costimulatory signal through intracellular partners. Human CD27-Fc binding and the ligand-expression controls in PMID:28011863(https://pubmed.ncbi.nlm.nih.gov/28011863/) complement the original CD27 extracellular-domain–IgG expression-cloning assay in PMID:8387892(https://pubmed.ncbi.nlm.nih.gov/8387892/). Although the 2017 title emphasizes CD70 deficiency, it directly tests T cells from a CD27 p.W110X patient. CD27 surface expression is absent, CD70-dependent proliferation with anti-CD3 is impaired, and anti-CD3/CD28 stimulation remains effective. Anti-CD27 blockade and CD70 rescue/CRISPR controls support a receptor mechanism rather than necessity inferred from a clinical phenotype alone.
The full cached GO-CAM model 666b894f00000352 and exact P26842 index entry were read. CD27 appears as transmembrane signaling receptor activity at the plasma membrane, in CD27 signaling and T-cell activation, with CD70 ligand activity upstream and TRAF2 adaptor activity downstream. A single receptor core captures this role. No new process annotation is proposed, and the broad receptor-signaling process is not duplicated alongside its existing specific CD27-signaling child in the core. The cytokine-family relationship does not license GO:0005031, whose actual definition specifies TNF binding: CD70 is not TNF itself. The receptor core retains the current curated GO:0004888 representation, with ligand specificity stated biologically.
Cell-surface, external-facing membrane and plasma-membrane source assertions agree with direct human flow cytometry and topology. Soluble serum CD27 in PMID:12624711(https://pubmed.ncbi.nlm.nih.gov/12624711/) supports a separate non-core extracellular localization; that observational study's renal-apoptosis model is not treated as proven mechanism or a core function.
The original 2002 experiments use EBV-negative human Ramos cells and human CD70 or CD154 expressed on murine 300-19 feeder cells, with vector feeders, BCR ligation, antibody blockade and p53-mutant controls. They examine Bcl-2/Bcl-XL/Bik, mitochondrial potential, cytochrome-c redistribution and caspase/PARP cleavage. These support context-dependent suppression of BCR-induced mitochondrial apoptosis. The peptide caspase inhibitors are separate reagents; CD27 was not thereby shown to be a caspase-binding inhibitor.
The official GO:0043027 definition requires binding to and reducing the activity of the relevant endopeptidase. The distinction between this direct molecular activity and upstream receptor signaling supports bounded MARK_AS_OVER_ANNOTATED for row28. The supported negative-regulation-of-apoptosis rows remain non-core. This does not imply that a noncatalytic regulator must itself catalyze a reaction; the missing criterion is the demonstrated direct inhibitory activity, not enzyme chemistry.
The 1997 CD27/Siva study contains ligand-triggered apoptotic responses in human B-cell lines, human CD27-expressing murine pre-B cells and activated human lymphocytes, as well as human CD27/GFP-Siva coimmunoprecipitation in transfected 293 cells. These are distinct from Siva-overexpression effects. The extrinsic-apoptotic-signaling assertion is retained non-core with context limits. The 2002 authors discuss the difference from the earlier result; their proposed explanations are not treated as a resolved universal mechanism. Neither a CD27 death domain nor a universal prosurvival or prodeath behavior is asserted.
The selected 2002 original sections do not identify an original p100/RelB-processing experiment. Broad introductory NF-kappaB citations and a related 1998 paper appended after the reproduction must not be mistaken for the 2002 Results. The original noncanonical-pathway NAS row remains UNDECIDED after independent normal-source integration. PMID15485626's complete canonical abstract describes CD70-dependent p100 processing and CD27-associated NIK/IKK recruitment, positively supporting this biological branch. Its full construct and organism Methods remain unresolved, and it does not establish that the original 2002 citation contains that assay. The uncertainty concerns the exact source mapping and assay scope, not a claimed absence of noncanonical signaling by CD27.
PMID9692890 explicitly measures JNK in primary murine lymph-node T cells and reports receptor-tail/TRAF2 interaction and dominant-negative experiments. The two human ortholog-transfer JNK assertions are retained non-core without relabeling those measurements as direct human assays. The exact donor IDs and original ISS/IEA evidence remain intact.
The combined T- and B-cell immune-memory IMP assertion from PMID28011863 remains UNDECIDED at the exact source-mapping level. The reduced memory-B-cell counts belong to the CD70-deficient index patient, while the CD27 patient experiments primarily test T-cell expansion. The authors explicitly separate expansion from intrinsic memory differentiation and indicate that memory T-cell immunity can remain present. This does not contradict the broader CD27 memory-programming role described in PMID38354704, nor does it justify removing a curator assertion merely because the title names CD70. The unresolved issue is the exact combined-memory annotation from that specific study.
Human B-cell differentiation and cytokine-dependent antibody production are supported by the complete review abstract PMID10809378 and primary abstract PMID14556986. These receptor outputs remain non-core. The IBA antiapoptosis assertion is retained as inherited biology; P26842's presence among descendants is legitimate experimental grounding, not circularity. No original PAINT IBD placement is claimed to have been reconstructed.
The durable CLINGEN_MENDELIAN project override retains correct but uninformative protein-binding assertions as KEEP_AS_NON_CORE. CD70 and SIVA1 pairs have direct independent primary support. ACTN2, CIB1, SPRED1, TRAF2 and the exact viral replication-protein edge are independently listed in the reviewed UniProt record. This corroborates pair retention without claiming inspection of original screen tables or transferring a partner's enzyme, actin, calcium or viral-process activity to CD27. The original source references, partners and SIVA1-1 specificity are preserved. Screen reference assessments distinguish verified identity from unresolved target-assay provenance.
The official 2024 correction to PMID35922511 was independently read. It corrects a CLC4M figure label, a concentration label and missing single-cell-method references. It does not announce a retraction or identify a CD27-edge correction. This bounded notice finding does not substitute for inspecting the original target-specific dataset.
All 43 source rows have manual decisions: 13 ACCEPT, 27 KEEP_AS_NON_CORE, one MARK_AS_OVER_ANNOTATED and two UNDECIDED. One receptor core is retained and no NEW annotation is proposed. The three normal references are integrated with their actual access flags. Four finite cached anchors support the review and core; quotation counts are checked across all occurrences. No verbatim source passage is added in these notes. Final term/schema/status/render results and exact source-preservation checks are recorded with the final review packet.
The first preliminary validation stopped on question/experiment object shape and an absent neighboring TMP GOA file. The original attempt is preserved. The second preliminary snapshot corrects only those schema shapes and supplies exact baseline GOA/UniProt copies; no biological decisions changed.
The 2021 structural paper (PMID34419446; https://doi.org/10.1016/j.jbc.2021.101102) directly examines human CD27 and human CD70. Its crystallized constructs are human CD27 A23–S127 and CD70 Q45–P193 expressed in Expi293 cells, with tag and glycosylation manipulations described in Methods. The receptor construct is truncated before stalk C185; the 3:3 crystal assembly must not be conflated with direct observation of the full-length disulfide-linked receptor dimer. Selected complete Results, Methods and embedded captions were read, including binding mutagenesis and plate-bound anti-CD3/CD70 stimulation of primary human peripheral-blood T cells. IFN-gamma and Jurkat AP-1 luciferase are actual measured outputs. The concentration- and presentation-dependent soluble-ligand results do not justify an absolute prohibition on soluble signaling. No figure pixels, supplementary figures or raw source data were reanalyzed.
PMID9582383 (https://doi.org/10.1074/jbc.273.21.13353) independently supports receptor-tail signaling through TRAF2/TRAF5 and NIK, but its original construct/species Methods remain unavailable despite bounded access attempts. PMID15485626 (https://doi.org/10.1016/j.immuni.2004.08.009) supplies independent receptor-specific p100/NIK/IKK evidence at the complete-abstract level. Neither paper is presented as a verified human-construct experiment on the basis of an abstract that does not state that detail. Three added reference records, a source-provenance amendment to the existing noncanonical-signaling review, and two additional core evidence pointers leave every existing action and core term unchanged.
Final normal validation, status checking and rendering passed on 2026-10-05. Validation reports fourteen generic-protein-binding advisories, all covered by the durable project override and the source-specific reasons above; there are no blocking errors or status mismatches. The computed status remains DRAFT. The final scope is 43 preserved source assertions, 27 assessed references, no alternative-product records, one receptor core and no NEW annotations; both source-specific UNDECIDED decisions remain explicit.
The final exact-substring audit detected source line breaks missing from two newly added quotations; the normal validator had accepted their whitespace-equivalent text. Both line breaks were restored to match immutable cache substrings exactly. The earlier candidate and successful normal checks are retained, and the corrected snapshot was validated again; no word, evidence identity, action or biological interpretation changed.