PTPRC (CD45) review notes

Review created 2026-10-03 for the ADAPTIVE_IMMUNITY project (T cell receptor area).

Biology summary

• CD45 is a type I transmembrane receptor-type PTP on all nucleated hematopoietic cells; intrinsic PTPase activity shown biochemically PMID:2853967.
• Positive role: dephosphorylates the inhibitory C-terminal tyrosine of SFKs in T and B cells PMID:9197241.
• Negative role: dephosphorylates activating tyrosines (HCK/LYN in macrophages; LCK Y394) PMID:9197241; PMID:17719247.
• Substrate recognition: D1 binds LCK kinase, SH2 and unique domains; D2 binds kinase subdomain X PMID:14625311.
• Beyond T cells: CD45 with CD148 required for BCR and macrophage Fc receptor signaling PMID:18249142; JAK phosphatase dampening cytokine signaling PMID:11201744; human anti-CD45 suppresses IL-4/JAK/STAT6 PMID:12574355.
• Human CD45 deficiency causes SCID PMID:10700239.

Curation decisions

• Core MF kept general: GO:0004725 protein tyrosine phosphatase activity. GO:0005001 (transmembrane receptor PTP) rows accepted as the standard RPTP classification, but not used as core because its definition requires signal binding and no physiological ligand regulating CD45 activity is established (question raised).
• 17 protein binding rows: 3 LCK rows (PMID:7507203, PMID:8576115 x2) changed to protein kinase binding; ITGAL row changed to integrin binding; the rest removed (enzyme-substrate peptide assays from PMID:19167335, CSK phosphorylating CD45, galectin-1 lectin binding, CD2 crosslinking, SKAP1 substrate, EGFR screen, beta-catenin).
• Beta-catenin IPI (PMID:12370829): the abstract says CD45's catalytic domain did NOT detectably bind catenins. Flagged as a suggested question for curators.
• synapse (GO:0045202, EXP PMID:35767951 and the UniProt SubCell mapping) changed to immunological synapse (GO:0001772): the UniProt "Synapse" location refers to the T cell contact, not a neuronal synapse.
• protein tyrosine kinase inhibitor activity (IDA PMID:12100025) removed: the paper measured LCK protein expression after antibody crosslinking; CD45 is not an enzyme inhibitor of kinases.
• Knockout, antibody-blockade and cytokine-output process rows kept as non-core; cell cycle, bone marrow development, stem cell development/proliferation, calcium release, defense response to virus, DN2 thymocyte differentiation marked over-annotated; response to gamma radiation and response to aldosterone removed; heparin and HSPG binding left undecided (no accessible source).
• No NEW rows: immunological synapse arrives via MODIFY; peptidyl-tyrosine dephosphorylation is implied by the MF.

Additional publications cached

• PMID:17719247 (McNeill 2007, LCK rheostat), PMID:11201744 (Irie-Sasaki 2001, JAK phosphatase), PMID:10700239 (Kung 2000, human SCID).

Deep research integration (falcon)

Report: PTPRC-deep-research-falcon.md (Edison/falcon, completed in ~20 min). Used as leads only; every adopted claim was traced to a cached primary paper.

Adopted (verified and cached):
- CD45 as a "gatekeeper": maintains a regulatable pool of active LCK while suppressing zeta-chain phosphorylation PMID:31641081. Added to core function 1 and to the regulation of TCR signaling rows.
- Size-based ectodomain segregation (kinetic segregation) PMID:23580664. Added to core function 1 and the description.
- No known ligand, but extracellular dimerizing ligands inhibit CD45 PMID:39454026. Used in the GO:0005001 rows, description, suggested question and a suggested experiment.
- Second human SCID case PMID:11145714.

Rejected / not used:
- JAK phosphatase role described by the report as "less securely established". Kept as a separate core function because the primary mouse paper shows direct in vitro dephosphorylation and binding of JAKs (PMID:11201744) and human anti-CD45 data agree (PMID:12574355); recorded as moderate confidence.
- UBR4/IKZF1 IgG4-related disease and circulating-tumor-cell CD45 (2024): context effects, not PTPRC functions; not used.
- Therapeutic sections (base editing, CAR T, 131I-BC8 radioimmunotherapy): outside GO scope.
- Roberts 2012 uniparental-disomy SCID case: not needed beyond the two cached SCID reports.

Report errors / caveats:
- No factual errors found in the claims checked. Citations are by internal keys (e.g. "hermiston2009cd45cd148and pages 4-5") rather than PMIDs, and several statements rely on reviews (Hermiston 2009, Al Barashdi 2021) rather than primary data.
- It describes the D2 domain as supporting "folding, substrate recruitment, and regulation"; only the substrate-binding role is supported in the papers cached here (PMID:14625311).