ZAP70 (human, P43403) curation notes

Project: ADAPTIVE_IMMUNITY (T cell receptor trunk).

Biology summary

• Syk-family cytoplasmic PTK; tandem N-SH2/C-SH2, interdomain B (Y292, Y315, Y319), C-terminal kinase domain (active fragment 327-606) PMID:15292186.
• Recruitment: Lck phosphorylates both ITAM tyrosines; ZAP70 binds via both SH2 domains PMID:7509083. Binding needs both ITAM tyrosines phosphorylated PMID:7528772. Kd ~25 nM for CD3 epsilon ITAM PMID:7761456.
• Activation: SH2 engagement relieves autoinhibition [PMID:10704231; PMID:8901551]; Lck phosphorylates Y493 and pY319 binds the Lck SH2 domain PMID:10318843.
• Substrates: LAT [PMID:9489702; PMID:11368773 "Zap-70 efficiently phosphorylates LAT on tyrosine residues at positions 226, 191, 171, 132 and 127."], SLP-76/LCP2 [Reactome:R-HSA-202216], VAV1 PMID:23620790, DUSP3/VHR Y138 PMID:12447358.
• Negative regulation: PTPN22 dephosphorylates Y493 PMID:16461343; UBASH3B/Sts-1 PMID:24256567; CBL TKBD binds pY292 PMID:22266821.
• Disease: selective T cell deficiency/SCID with absent CD8 SP T cells PMID:8613493; combined hypomorphic + activating alleles cause autoimmunity PMID:26783323.

Annotation review decisions (98 rows)

• MF kinase terms (GO:0004672, GO:0004713, GO:0004715; all evidence types incl. Reactome TAS): ACCEPT. Four Reactome TAS rows are SYK-centred reactions (DAP12 signaling) where ZAP70 is presumably in a candidate set; the MF itself is correct.
• GO:0001784 phosphotyrosine residue binding (IEA): ACCEPT; core.
• protein binding (28 IPI rows): 12 MODIFY -> GO:0001784 (11 rows) (SH2-pITAM binding to CD247/CD3E; FCRL3 phospho-ITAM-like motif) or GO:0019901 protein kinase binding (1 row, LCK via pY319); 16 REMOVE (substrate-trap/phosphatase, E3 ligase, kinase-substrate, high-throughput SH2 screens with EGFR/MET/KIT, and abstracts that do not describe a ZAP70 function).
• Locations: cytoplasm/cytosol/plasma membrane/membrane/immunological synapse ACCEPT; cell-cell junction, membrane raft (colocalizes_with) and TCR complex (recruited, not a constitutive subunit) KEEP_AS_NON_CORE.
• Processes: TCR signaling pathway ACCEPT (core; ZAP70 performs the phosphorylation steps). Protein phosphorylation / peptidyl-tyrosine phosphorylation / intracellular signal transduction ACCEPT (general but correct). Adaptive immune response, immune response, T cell activation, T cell differentiation, positive thymic selection, positive regulation of T cell differentiation, leukocyte adhesion/migration, T cell aggregation/migration: KEEP_AS_NON_CORE (necessity or downstream outcomes per project rule 3). B cell activation: MARK_AS_OVER_ANNOTATED.

Issues noted

• PMID:22732588 (peptidyl-tyrosine phosphorylation IDA): full text says "Src kinases (Lck or Fyn) but not the ZAP-70 kinase can phosphorylate both Themis1 and Themis2", and Themis1 was phosphorylated in ZAP70-deficient P116 cells; the reference weakly supports this row. Kept (ACCEPT) because the term is correct for ZAP70, flagged reference as MISCITED.
• PMID:12150984 and PMID:8648092 are abstract-only and the abstracts do not mention ZAP70; deferred to curators on location, removed the generic protein binding row.
• THEMIS: PMID:23460737 reports ZAP70 phosphorylates THEMIS in vitro and with Lck in HEK293, while PMID:22732588 reports it does not; conflicting.

Deep research

• just deep-research-falcon human ZAP70 --fallback perplexity-lite launched in background at the start of the review. The wrapper reported a falcon timeout after 600s and the perplexity-lite fallback failed (provider not available), but the falcon job itself completed (~1218 s, 29 citations) and wrote ZAP70-deep-research-falcon.md. Its content (tandem SH2 binding of doubly phosphorylated ITAMs, LCK-dependent activation via Y315/Y319/Y493, LAT and SLP-76 as principal substrates, ZAP70 not phosphorylating ITAMs) is consistent with this review and it is cited in core_functions.

Deep research integration (falcon)

Report: ZAP70-deep-research-falcon.md (Edison/falcon, 11 distinct sources, no PMIDs given; DOIs resolved to PMIDs via PubMed title search). Treated as secondary literature.

Claim tally (approx. 30 substantive claims): ~20 confirm the review; 4 add something new; 2 conflict; ~5 not relevant or unsupported for GO purposes.

Claims adopted

• Electrostatic substrate selection (acidic residues around the target Tyr, exclusion of Lys/Arg) -> verified in primary paper and added to description and core_functions[0] supported_by PMID:27700984.
• ZAP70 does not phosphorylate ITAMs (depends on LCK) -> description and core_functions[0] PMID:34923645. Removed "CD3 zeta itself" from the substrate suggested question accordingly.
• C-SH2 patient variants (R170C, R192W) impair phospho-zeta ITAM binding -> added PMID:37313400 as corroborating support on the GO:0005515 -> GO:0001784 MODIFY row (PMID:26783323) and core_functions[1]. No action change.
• ZAP70 augments tonic BCR/CCR7 signaling in IGHV-unmutated CLL (PMID:38113463) -> added as context on the B cell activation (GO:0042113, TAS) row; action MARK_AS_OVER_ANNOTATED unchanged because the evidence is from malignant cells, not normal B cell activation. Folded into the B cell suggested question.
• Cyclophilin A binds interdomain B and downregulates ZAP70 activity (PMID:36495335, primary, full text) -> raised as a suggested question only (single-lab, mouse/Jurkat). Not a NEW annotation: no existing GO row concerns this and ZAP70 is the regulated entity, not the regulator.

Claims not acted on

• "Thermodynamic brake", "allosteric hot spot" in the tandem SH2 (Gangopadhyay 2020/2021) and "kinetic avidity"/CD45-regulated unbinding (Goyette 2020): bioRxiv preprints as cited; mechanistic detail below GO resolution. Not adopted.
• Schultz 2022 (Cys residue in kinase domain regulating Lck): single citation for Y493 claim; not needed.
• Aktas 2024 CLL cytokine book chapter: low-quality source for VAV/structure claims; not used.
• Disease/prognosis statements (SCID, autoimmunity, CLL marker): already in review description; no GO impact.

Conflicts / report errors

• Autophosphorylation: the report repeatedly says ZAP70 autophosphorylates its regulatory tyrosines as part of activation. PMID:27700984 states "the inability of ZAP-70 to undergo robust autophosphorylation is due to steric and electrostatic repulsion". The review already attributes Y493/Y319 phosphorylation to LCK; raised as a suggested question.
• Report describes the Anto et al. CypA paper as "Dec 2023"; the DOI (s00018-022-...) and PubMed record (PMID:36495335) indicate a 2022/2023 Cell Mol Life Sci paper - minor date confusion. Goyette 2020 kinetic avidity is cited only as the bioRxiv preprint.
• No hallucinated PMIDs (the report gives none); all five DOIs checked resolved to real papers with matching titles.

THEMIS (PMID:22732588 vs PMID:23460737)

• The falcon report does not mention THEMIS at all, so it does not resolve the conflict. Indirectly relevant: the electrostatic filter (PMID:27700984) predicts that genuine ZAP70 sites are acidic-flanked; whether THEMIS phosphosites fit this was added to the substrate suggested question, and a direct in vitro kinase/phosphosite-mapping experiment was added to suggested_experiments. Existing decisions on both rows unchanged (PMID:22732588 IDA ACCEPT with MISCITED reference flag; PMID:23460737 protein binding REMOVE as uninformative).