The normal source import contains human CD19/P15391, 38 annotations, two alternative products and 22 references. All machine fields are preserved. Six existing publication caches and six Reactome caches are reused without replacement; their access flags describe the selected versions. The two family exports remain outside this review. No new GO annotation is proposed.
The normal Falcon deep-research command and its Perplexity fallback both failed to resolve the pinned deep-research-client dependency in the available offline environment. No provider report was authored or substituted. This document records manual research. All twelve selected publication headers and abstracts and six complete Reactome summaries were read. Selected original Methods and Results are identified below; this is not a claim to have inspected all figures, supplements or raw data.
CD19 provides a noncatalytic membrane platform connecting the B-cell receptor environment to intracellular signaling proteins. The core molecular function is signaling receptor complex adaptor activity (GO:0030159), supported by complex assembly and recruitment evidence. Regulation of B-cell receptor signaling (GO:0050855) is the core process. Plasma membrane is the core location. Proliferation, antibody responses and calcium mobilization are downstream outputs; CD19 is not assigned kinase, lipid-kinase or calcium-channel activity.
The antibody-coligation study establishes enhanced antigen sensitivity (PMID:1373518(https://pubmed.ncbi.nlm.nih.gov/1373518/)). Its cached abstract and PubMed record were read; the complete original methods were not obtained. The patient study connects biallelic CD19 mutations to defective antigenic responses despite preserved mature B-cell numbers (PMID:16672701(https://pubmed.ncbi.nlm.nih.gov/16672701/)). Publisher-indexed calcium text was available, but direct full-page and PDF requests failed. Its canonical cache remains abstract-only.
Human receptor-complex coprecipitation and surface patching support association with CD21/CR2, CD81/TAPA-1 and Leu-13 (PMID:1383329(https://pubmed.ncbi.nlm.nih.gov/1383329/), complete abstract). Raji and recombinant K562 experiments distinguish the ligand-recognition role of CR2 from the CD19-containing signaling assembly (PMID:1702139(https://pubmed.ncbi.nlm.nih.gov/1702139/), complete abstract; "CR2 associates directly with CD19"). These results do not assign complement-fragment binding to CD19 itself.
PMID:9382888(https://pubmed.ncbi.nlm.nih.gov/9382888/): selected complete cached Methods and Results concern human WT or Y484F/Y515F CD19 expressed in mouse J558Lμm3CD45+ cells, plus a separate mouse splenic-B-cell arm. PI3K, IP3 and calcium were measured; AKT is discussed as a possible downstream mechanism. Historical residue numbering is retained as reported. Figure pixels and results reported as “not shown” were not inspected.
PMID:29523808(https://pubmed.ncbi.nlm.nih.gov/29523808/): selected Results, Figure 8 caption and expression/antibody Methods describe GRB2-SH2 affinity purification from human DG75 lysates with GST and mutant-domain controls. CD19 was detected by immunoblot despite absence from the mass-spectrometry identification. This supports association, without proving purified binary binding or a native messenger-binding mechanism for CD19. The source coreceptor annotation remains UNDECIDED because GO:0015026 explicitly includes messenger binding; the established adaptor function is accepted separately.
PMID:23071339(https://pubmed.ncbi.nlm.nih.gov/23071339/): the existing 8 KB full-text-marked cache is abbreviated. Selected Results from the richer authenticated 29 KB extraction describe CD19 crosslinking and a CD19–IKAROS–SYK complex in human FL112 pro-B cells. Supplementary Figure S4 was not inspected. The SYK/Ikaros-focused title is not evidence of a wrong-gene annotation. Preserve the broad complex observation as non-core; do not infer direct binary interaction or CD19 kinase activity.
PMID:9317126(https://pubmed.ncbi.nlm.nih.gov/9317126/) is explicitly a transgenic-mouse study of a CD19 construct lacking its cytoplasmic region. Its abstract supports signaling dependence on that region. It is not described as a primary-human experiment. Antibody effects in the human study PMID:2463100(https://pubmed.ncbi.nlm.nih.gov/2463100/) depend on cell type and stimulus; its abstract does not support a universal activating direction.
The external-surface IDA annotations cite FcRL6 and microRNA papers (PMID:17213291(https://pubmed.ncbi.nlm.nih.gov/17213291/), PMID:20660734(https://pubmed.ncbi.nlm.nih.gov/20660734/)). Their selected abstracts do not expose the CD19 assays. The localizations agree with independent CD19 surface evidence, so the annotations are retained with explicit curator deference, without claiming those assays were inspected. The exosome observation (PMID:20458337(https://pubmed.ncbi.nlm.nih.gov/20458337/)) remains non-core; its target table was not inspected, and detection does not imply exosome-biogenesis activity.
All six Reactome summaries were read. Membrane complexes support location; lipid phosphorylation belongs to PI3K and complement recognition to CR2. The inhibitor-context reaction does not make CD19 an enzyme or a direct inhibitor target.
The three PAINT assertions retain the curator's evolutionary judgment. The IBD placement and alignment were not independently inspected, and no structured ancestral-gain claim is invented. CD19 appearing in its own descendant-evidence list is legitimate, not circular support. The two alternative products remain exactly as seeded; no product-specific functional distinction is inferred from their names.
GO definitions were checked in official GO/MOD displays: signaling receptor complex adaptor activity and coreceptor activity. The single pathway refinement narrows antigen receptor signaling to B-cell receptor signaling. It creates no new assertion and preserves the original source fields.
Normal schema, term and best-practice validation passed without diagnostics. The status check computed COMPLETE; the initial manually selected DRAFT was corrected to that computed value. The one unresolved coreceptor annotation remains explicit. Independent whole-review assessment is pending.
The source-specific coreceptor decision is now MARK_AS_OVER_ANNOTATED. In the inspected human DG75 experiment, GRB2-SH2 affinity purification with WT, R86L and GST controls and a CD19 immunoblot establishes signaling-complex association (PMID:29523808(https://pubmed.ncbi.nlm.nih.gov/29523808/)). The official GO:0015026 definition requires combining with a messenger and cooperating with another receptor to initiate signaling. The inspected CD19 experiment does not establish that complete activity. The decision is limited to this source-to-term mismatch; it does not assert that CD19 has no possible physiological messenger or that every receptor-complex subunit is excluded. The independently supported signaling receptor complex adaptor activity remains the core. This paragraph supersedes the earlier unresolved-action account without altering its historical text.
The core description now states phosphotyrosine-dependent p85-containing PI3K recruitment and activation explicitly. The selected original Methods and Results in PMID:9382888(https://pubmed.ncbi.nlm.nih.gov/9382888/) distinguish human WT or Y484F/Y515F CD19 in mouse J558L cells from the separate mouse splenic-B-cell arm. Antigen-stimulated anti-p85-associated kinase measurements and calcium controls support this proximal signaling mechanism. The mutant PI3K result described as not shown was not independently inspected, and AKT was not experimentally measured in the inspected assay. The core anchor now uses the paper's own result statement rather than its cited background sentence. The existing broader PI3K/protein-kinase-B pathway row remains non-core as an arm of the B-cell receptor signaling unit; absence of an AKT assay is not treated as evidence against the PI3K result. No separate catalytic function or redundant new process annotation is added.
Reference assessment is source-specific. Official identifier/title records and the complete available abstracts support VERIFIED for PMID:1383329, PMID:1702139, PMID:2463100 and PMID:9317126; this does not imply complete-body access. PMID:9382888 is VERIFIED with the selected Methods/Results limits above. Four official Reactome event identities and complete cached summaries explicitly identify CD19 or its receptor complex (R-HSA-199518, R-HSA-2076220, R-HSA-8853252 and R-HSA-9606887), supporting bounded VERIFIED assessments. The generic PI3K reaction R-HSA-2316434 and inhibitor overview R-HSA-2400009 do not expose the specific CD19 target-containing complex in their summaries and remain UNVERIFIED. Three inaccessible exact CD19 assay/table mappings and the four method-reference target mappings also remain explicitly unverified. Access flags now describe all twelve chosen publication caches: three have full text available and nine are abstract-only. PMID:23071339's full-text flag still describes the abbreviated canonical extraction; the richer authenticated selected Results and unread supplement remain separately documented above.
All 38 machine annotation objects, two products, original reference identifiers/titles and core ontology terms are unchanged. Only the source-specific coreceptor action changes; the PI3K row receives a scope clarification. Existing broader B-cell activation assertions remain supported and are not new redundancy. Immutable GOA, UniProt, publication and Reactome files are preserved. The original CREATE history is append-only: its suggested_experiments section label did not correspond to an authored YAML field, and no experiment is invented or old record rewritten to match it. Any subsequent EDIT record should identify only the sections actually changed.
Normal schema, term and best-practice validation, computed-status checking and rendering all passed for this focused follow-up. COMPLETE is unchanged, with zero annotation advisories and no status mismatch; the former single UNDECIDED action is now the bounded over-annotation decision above. The renderer includes the appended notes and unchanged machine assertions.
Re-read the complete 38-row existing review, including human CD19 constructs expressed in mouse J558L cells (9382888), patient B-cell responses (16672701), receptor coligation (1373518), and full DG75 GRB2-SH2/CD19 experiments (29523808). Existing actions are retained: the adaptor function is supported; the latter experiment does not establish the messenger-binding requirement of GO0015026 coreceptor activity. PI3K activation in9382888 is measured, whereas AKT is discussed rather than directly assayed. This source distinction does not erase supported participation in the broader PI3K/AKT pathway. Exosome proteomics remains contextual localization without an exosome-biogenesis claim. No new annotation or mechanistic claim was added.
All original annotation payloads, evidence codes, supporting entities, relationship fields and flags are preserved. Review decisions judge biological claims; ordinary relationship qualifiers are not used to change their meaning. No new annotation rows were proposed. Actual Falcon and perplexity-lite attempts failed authentication/quota; manual source inspection is documented here instead of a fabricated provider report. The existing PIK3R1 provider report was reused as discovery, with primary evidence assessed separately.