Gene: human GPR25 (UniProt O00155), class A GPCR, 360 aa, chr 1q32.1.
Reviewed as part of a contested-function batch (see also GPR158, GPR75). Within that
batch GPR25 is the positive control: the deorphanization here is far stronger than for
either of the other two.
GPR25 was cloned in 1997 by PCR with somatostatin-receptor primers:
PMID:9020062, and had essentially no function attached for
27 years.
1. Ocón et al., Nature 2024 (Butcher lab, Stanford).
PMID:39293486, with in vivo homing:
PMID:39293486
The decisive adoptive-transfer experiment:
PMID:39293486, with
the result
PMID:39293486 — and this was assayed in Cxcl17-null as well as wild-type
recipients, tying the homing to the ligand.
2. Hu et al., FEBS J 2025 (Guo lab, Tongji).
PMID:40279398, with structure-guided mutagenesis:
PMID:40279398 and a matched-pair specificity test
PMID:40279398
3. Yang, Giblin & Pease, Basic Clin Pharmacol Toxicol 2026 (Imperial College).
A third, fully independent laboratory:
PMID:42207165 concluding
PMID:42207165. Note that this group independently recovered the
same W95 and R178 residues (plus R264) as essential.
4. Independent in vivo genetics.
PMID:41270189 with
PMID:41270189
This is a fourth lab (La Jolla Institute) and, importantly, a non-chemotaxis phenotype
that converges on the same tissue-residency biology.
PMID:41167449 — the same Tongji group reports that CXCL17 also
activates MRGPRX1, MRGPRX2 and MAS1. But the mechanistic dissection points the other way
from what that sentence suggests:
PMID:41167449
In other words, the MRGPR responses require micromolar CXCL17 and a different
structural determinant, whereas the GPR25 response is nanomolar-to-~100 nM, requires the
conserved C-terminal fragment, and maps to a defined orthosteric pocket. A promiscuous
ligand does not make one of its receptors wrong. The direction of residual doubt is
"which receptor carries CXCL17's physiological signal in which tissue", not "is GPR25 a
CXCL17 receptor".
There is also a live taxonomic argument about CXCL17 itself — the Imperial group
"advocate the reclassification of CXCL17 as a chemoattractant distinct from the chemokine
family" (PMID:42207165). That is a question about the ligand's nomenclature; it does not
disturb GO:0016494, whose scope is C-X-C-named chemokine ligands, and CXCL17 is one.
GO:0097021 lymphocyte migration into lymphoid organs is annotated twice with IDA
(PMID:39293486, PMID:40279398). Its definition is "The movement of a lymphocyte within
the lymphatic system into lymphoid organs such as lymph nodes, spleen or Peyer's
patches...". The full text of PMID:39293486 states the opposite for GPR25: homing was
enhanced to mucosal barrier tissues "but not to the intestines or control peripheral
lymph nodes (PLNs) or spleen". This is a term-scoping error, not a gene error — the
underlying biology (lymphocyte migration to peripheral tissue) is right. Action:
MODIFY both rows to GO:0072676 lymphocyte migration, which is true and does not
assert lymphoid-organ entry. A mucosal-homing child term would be better still and is
raised in suggested_questions.
GO:0016494 C-X-C chemokine receptor activity (IDA ×2) → ACCEPT, core. This isGO:0070098 chemokine-mediated signaling pathway (IDA ×2) → ACCEPT, core.GO:0048247 lymphocyte chemotaxis (IDA ×2) → ACCEPT, core.GO:0030595 leukocyte chemotaxis (IDA) → ACCEPT (correct parent of the above).GO:0097021 lymphocyte migration into lymphoid organs (IDA ×2) → MODIFY toGO:0072676, see above.GO:0004930 GPCR activity (IBA, IEA, IDA ×2, TAS) → ACCEPT. Correct, now subsumedGO:0007186, GO:0005886, GO:0016020 → ACCEPT.GO:0005515 protein binding (IPI ×20, from three interactome screens and one