Brain-enriched class A seven-transmembrane receptor, originally cloned as an endothelin-B-receptor-like gene and independently identified as the parkin substrate Pael-R. It has a long extracellular N-terminus that is constitutively cleaved by metalloproteinase and shed, so the receptor is present at the cell surface mainly in a truncated form, and it carries a C-terminal PDZ-binding motif through which MUPP1/MPDZ delivers it to dendrites and synapses in a complex with CASPR2. GPR37 signals through Gi, inhibiting adenylate cyclase and raising intracellular calcium and ERK phosphorylation, and is expressed in neurons, oligodendrocytes, astrocytes, sensory neurons and macrophages. The receptor is also strongly prone to misfolding: overexpressed or mutant GPR37 aggregates and is retained in the endoplasmic reticulum, where it is ubiquitinated by parkin together with CHIP/STUB1 and Hsp70, and by the ER ligase HRD1/SYVN1, and its accumulation triggers ER stress and dopaminergic cell death, linking it to autosomal-recessive juvenile parkinsonism. Its endogenous ligand is disputed. The secreted glycoprotein prosaposin and its peptide fragment prosaptide bind and activate GPR37 and are the best-replicated of the candidates; the docosahexaenoic-acid-derived lipid mediator protectin D1 (neuroprotectin D1) is reported to activate the same receptor in the same assays, which would make GPR37 a receptor for both a peptide and a lipid; the bone-derived hormone osteocalcin has been proposed as a third ligand acting through GPR37 in oligodendrocytes and in ventral tegmental area neurons; and an earlier pairing with the invertebrate peptide head activator was not reproduced and the peptide has no human gene. Loss of GPR37 in mice delays resolution of inflammatory and neuropathic pain, impairs macrophage phagocytosis and defence against infection, and alters oligodendrocyte differentiation and myelination.
Definition: Combining with protectin D1 (neuroprotectin D1), a docosahexaenoic-acid-derived specialized pro-resolving lipid mediator, and transmitting the signal across the membrane by activating an associated G-protein.
Justification: GPR37 is reported to be activated by protectin D1/neuroprotectin D1 in binding and calcium assays, and PD1's actions on macrophage phagocytosis, sepsis survival, chemotherapy-induced neuropathy, inflammatory pain resolution and efferocytosis are lost in Gpr37 knockout mice. No GO term exists for a protectin D1 receptor, so this activity currently has nowhere to go and is instead absorbed into generic GPCR terms. Proposed rather than asserted as a core function, because the claim carries a real chemical-class problem - the same receptor is asserted to be activated by a peptide (prosaptide) and by a lipid, and most of the supporting work shares a senior author. The decisive experiment, competition binding of protectin D1 against labelled prosaptide on the same receptor preparation, has not been reported.
Parent term: bioactive lipid receptor activity
Supporting Evidence:
Definition: Combining with osteocalcin, a bone-derived peptide hormone, and transmitting the signal across the membrane by activating an associated G-protein.
Justification: Osteocalcin has been proposed as an endogenous ligand of GPR37, first in oligodendrocytes, where GPR37 was identified as an OCN receptor controlling differentiation and myelination, and more recently in ventral tegmental area GABAergic neurons, where OCN suppresses THIK-1 potassium currents through GPR37-mediated cAMP reduction. No osteocalcin receptor term exists in GO, although two other receptors, GPRC6A and GPR158, have been proposed for the same hormone. Proposed rather than asserted as a core function because the two GPR37 studies share senior authors and therefore corroborate rather than independently replicate one another, and because the 2026 study carries a substantial published erratum.
Parent term: G protein-coupled peptide receptor activity
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005886 plasma membrane | IBA GO_REF:0000033 | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. The phylogenetic assertion matches the direct evidence. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0007193 adenylate cyclase-inhibiting G protein-coupled receptor signaling pathway | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic assertion of Gi-coupled, adenylate-cyclase-inhibiting signalling. Reason: Matches the direct evidence in human cells - prosaptide stimulation of GPR37-transfected cells inhibits forskolin-stimulated cAMP in a pertussis-toxin-sensitive manner - and the same Gi/cAMP direction is reported for the osteocalcin work. Supporting Evidence: PMID:23690594 Prosaptide stimulation of cells transfected with GPR37 or GPR37L1 induced the phosphorylation of ERK in a pertussis toxin-sensitive manner |
| GO:0008528 G protein-coupled peptide receptor activity | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic assertion that GPR37 is a peptide-activated GPCR. Reason: Correct at this level of generality and supported by the prosaptide/prosaposin work. It is also the safe term to retain if the more specific prosaposin pairing were ever overturned. |
| GO:0036505 prosaposin receptor activity | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic propagation of prosaposin receptor activity. Reason: Core molecular function, and the only proposed GPR37 ligand pairing with support from more than one laboratory. The original screen (Hall lab) found prosaptide to be the single hit among orphan neuropeptides, showed binding, receptor-dependent endocytosis, pertussis-toxin-sensitive ERK phosphorylation, GTPgammaS loading and cAMP inhibition, and reproduced the effects with purified full-length prosaposin. The Ji lab subsequently reproduced TX14 binding and GPR37-dependent calcium signalling in an independent system, and in 2026 showed that the in vivo actions of intrathecal TX14A require GPR37 specifically in TRPV1-lineage sensory neurons. Reservations remain and are recorded rather than suppressed - the agonist concentration needed is around 100 nM, higher than is usual for peptide/GPCR pairs; the pairing had not been ratified by the IUPHAR Nomenclature Committee as of the 2015 review of this receptor; and prosaposin may additionally act as a folding chaperone that increases GPR37 cell-surface expression, which would produce some of the same readouts without orthosteric agonism. The WITH/FROM field of this IBA contains UniProtKB:O15354, GPR37 itself, which is expected and correct - GPR37 carries its own IDA for this term and is therefore one of the descendant evidences the PAINT curator used to place the ancestral node. That is a marker of experimental grounding on the target, not circularity. Supporting Evidence: PMID:23690594 We screened these receptors for potential activation by various orphan neuropeptides, and these screens yielded a single positive hit: prosaptide, which promoted the endocytosis of GPR37 and GPR37L1, bound to both receptors and activated signaling in a GPR37- and GPR37L1-dependent manner. PMID:30010619 NPD1 and TX14 also bind to GPR37 and cause GPR37-dependent iCa2+ increases in peritoneal MΦs. PMID:42144155 Global GPR37 knockout or conditional knockout of GPR37 in TRPV1-lineage sensory neurons abolished the long-term inhibitory effect and unpriming effect of i.th. TX14A, confirming that GPR37 in this specific cellular population mediates these effects. |
| GO:0043235 signaling receptor complex | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Phylogenetic assertion that the receptor is part of a signalling receptor complex. Reason: Plausible and consistent with the documented GPR37 complexes (with MUPP1/CASPR2 and with the dopamine transporter), but the term is generic and says nothing about which complex or what it does. Retained, demoted from core. |
| GO:0043410 positive regulation of MAPK cascade | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic assertion of positive regulation of the MAPK cascade. Reason: Supported directly in human cells: prosaptide and prosaposin stimulation of GPR37 induces ERK phosphorylation in a pertussis-toxin-sensitive, receptor-dependent manner. Supporting Evidence: PMID:23690594 Prosaptide stimulation of cells transfected with GPR37 or GPR37L1 induced the phosphorylation of ERK in a pertussis toxin-sensitive manner |
| GO:0004930 G protein-coupled receptor activity | IEA GO_REF:0000002 | ACCEPT | Summary: Generic GPCR activity from the rhodopsin-like 7TM InterPro signature. Reason: Correct; GPR37 is a class A GPCR that couples to Gi. Same action as the TAS row for this term. |
| GO:0005789 endoplasmic reticulum membrane | IEA GO_REF:0000044 | ACCEPT | Summary: ER membrane localisation projected from the UniProt subcellular-location annotation. Reason: Correct and biologically important rather than incidental. GPR37 is strongly prone to misfolding and ER retention, is a substrate for parkin- and HRD1/SYVN1-mediated ER-associated degradation, and disease-associated variants such as R558Q accumulate in the ER. Supporting Evidence: PMID:17059562 This study shows that HRD1 was expressed in substantia nigra pars compacta (SNC) dopaminergic neurons and interacted with Pael-R through the HRD1 proline-rich region, promoting the ubiquitylation and degradation of Pael-R. |
| GO:0005886 plasma membrane | IEA GO_REF:0000044 | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0007186 G protein-coupled receptor signaling pathway | IEA GO_REF:0000002 | ACCEPT | Summary: Generic GPCR signalling pathway from the InterPro signatures. Reason: Correct; same action as the TAS row for this term. |
| GO:0007193 adenylate cyclase-inhibiting G protein-coupled receptor signaling pathway | IEA GO_REF:0000117 | ACCEPT | Summary: ARBA machine-learning assignment of Gi-coupled signalling. Reason: Concordant with the IBA and IDA rows for this term. |
| GO:0008528 G protein-coupled peptide receptor activity | IEA GO_REF:0000117 | ACCEPT | Summary: ARBA machine-learning assignment of peptide receptor activity. Reason: Concordant with the IBA and IDA rows for this term. |
| GO:0016020 membrane | IEA GO_REF:0000002 | ACCEPT | Summary: Generic membrane localisation from the 7TM signatures. Reason: Correct but uninformative next to the plasma membrane and ER membrane annotations this gene already carries. |
| GO:0030425 dendrite | IEA GO_REF:0000044 | ACCEPT | Summary: Dendritic localisation projected from the UniProt subcellular-location annotation. Reason: Supported experimentally: MUPP1 transports GPR37 to the dendrite and synapse in primary hippocampal neurons. Supporting Evidence: PMID:25977097 In primary hippocampal neurons, GPR37 co-localized with MUPP1 and CASPR2 at the synapse, but not GPR37(R558Q). |
| GO:0036505 prosaposin receptor activity | IEA GO_REF:0000117 | ACCEPT | Summary: ARBA machine-learning assignment of prosaposin receptor activity. Reason: Core molecular function, and the only proposed GPR37 ligand pairing with support from more than one laboratory. The original screen (Hall lab) found prosaptide to be the single hit among orphan neuropeptides, showed binding, receptor-dependent endocytosis, pertussis-toxin-sensitive ERK phosphorylation, GTPgammaS loading and cAMP inhibition, and reproduced the effects with purified full-length prosaposin. The Ji lab subsequently reproduced TX14 binding and GPR37-dependent calcium signalling in an independent system, and in 2026 showed that the in vivo actions of intrathecal TX14A require GPR37 specifically in TRPV1-lineage sensory neurons. Reservations remain and are recorded rather than suppressed - the agonist concentration needed is around 100 nM, higher than is usual for peptide/GPCR pairs; the pairing had not been ratified by the IUPHAR Nomenclature Committee as of the 2015 review of this receptor; and prosaposin may additionally act as a folding chaperone that increases GPR37 cell-surface expression, which would produce some of the same readouts without orthosteric agonism. Supporting Evidence: PMID:23690594 We screened these receptors for potential activation by various orphan neuropeptides, and these screens yielded a single positive hit: prosaptide, which promoted the endocytosis of GPR37 and GPR37L1, bound to both receptors and activated signaling in a GPR37- and GPR37L1-dependent manner. PMID:30010619 NPD1 and TX14 also bind to GPR37 and cause GPR37-dependent iCa2+ increases in peritoneal MΦs. PMID:42144155 Global GPR37 knockout or conditional knockout of GPR37 in TRPV1-lineage sensory neurons abolished the long-term inhibitory effect and unpriming effect of i.th. TX14A, confirming that GPR37 in this specific cellular population mediates these effects. |
| GO:0042277 peptide binding | IEA GO_REF:0000117 | KEEP AS NON CORE | Summary: Generic peptide binding, assigned by an ARBA model. Reason: True but uninformative; the specific GO:0036505 prosaposin receptor activity annotation carries the same information with the ligand identified. Same action as the IPI row for this term. |
| GO:0045202 synapse | IEA GO_REF:0000120 | ACCEPT | Summary: Synaptic localisation, combined automated annotation. Reason: Concordant with the experimental localisation of GPR37 to synapses in hippocampal neurons. Supporting Evidence: PMID:25977097 In primary hippocampal neurons, GPR37 co-localized with MUPP1 and CASPR2 at the synapse, but not GPR37(R558Q). |
| GO:0005515 protein binding | IPI PMID:17519329 GPR37 associates with the dopamine transporter to modulate d... | REMOVE | Summary: Bare protein binding recording the GPR37-SLC6A3 (dopamine transporter) interaction. Reason: 'Project guidance: protein binding carries no functional information. The underlying observation, that GPR37 associates with the dopamine transporter and modulates dopamine uptake, would be worth a specific transporter-regulator or process annotation if it were replicated.'. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:23864651 The identification of novel proteins that interact with the ... | REMOVE | Summary: Bare protein binding from a screen for proteins that interact with the GLP-1 receptor. Reason: Uninformative as a molecular function, and a GPCR-GPCR association from a targeted interaction screen with no functional follow-up for this pair. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:28298427 Systematic protein-protein interaction mapping for clinicall... | REMOVE | Summary: Bare protein binding from a systematic interaction map of clinically relevant human GPCRs (eight partners). Reason: Uninformative as a molecular function; high-throughput affinity-capture data without pair-specific follow-up. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:32814053 Interactome Mapping Provides a Network of Neurodegenerative ... | REMOVE | Summary: Bare protein binding from a neurodegenerative-disease interactome study (thirteen partners). Reason: Uninformative as a molecular function. Worth noting that this study is about widespread protein aggregation in affected brains, which is consistent with GPR37 being aggregation-prone, but the annotation itself records nothing. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0009986 cell surface | IEA GO_REF:0000107 | ACCEPT | Summary: Cell surface localisation projected from the mouse ortholog. Reason: Supported directly in human cells - GPR37 matures, is exported from the ER and is present at the cell surface, predominantly as the N-terminally shed form. Supporting Evidence: PMID:26869225 GPR37 is expressed at the cell surface predominantly in the N-terminally truncated form. |
| GO:0016358 dendrite development | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Dendrite development, projected from the mouse ortholog. Reason: Plausible and consistent with the CASPR2-MUPP1-GPR37 dendritic complex, but a downstream organismal-level consequence rather than something the receptor protein does. Non-core. |
| GO:0030165 PDZ domain binding | IEA GO_REF:0000107 | ACCEPT | Summary: PDZ domain binding, projected from the mouse ortholog. Reason: Supported in human cells: GPR37 binds the PDZ11 domain of MUPP1 through its C-terminal PDZ-binding motif, and a GPR37 construct lacking that motif is not delivered to the cell surface. Supporting Evidence: PMID:25977097 GPR37, but not GPR37(R558Q) nor GPR37-deltaC which lacks its PDZ binding domain, was transported to the cell surface by MUPP1. |
| GO:0034614 cellular response to reactive oxygen species | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Cellular response to reactive oxygen species, projected from the mouse ortholog. Reason: Real but indirect. The evidence is that prosaposin and prosaptide protect astrocytes against oxidative stress in a GPR37/GPR37L1-dependent way; that is a downstream protective outcome of receptor signalling rather than an oxidative-stress-sensing role of the receptor itself. Same action as the ISS row for this term. Supporting Evidence: PMID:23690594 Moreover, both prosaptide and prosaposin were found to protect primary astrocytes against oxidative stress, with these protective effects being attenuated by siRNA-mediated knockdown of endogenous astrocytic GPR37 or GPR37L1. |
| GO:0043410 positive regulation of MAPK cascade | IEA GO_REF:0000120 | ACCEPT | Summary: Positive regulation of MAPK cascade, combined automated annotation. Reason: Concordant with the IBA and IDA rows for this term. |
| GO:0005886 plasma membrane | EXP PMID:25977097 CASPR2 forms a complex with GPR37 via MUPP1 but not with GPR... | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. MUPP1 delivers GPR37 to the cell surface. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0005886 plasma membrane | EXP PMID:26869225 The Parkinson's-disease-associated receptor GPR37 undergoes ... | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. The receptor is present at the surface predominantly as the N-terminally shed form. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0045202 synapse | EXP PMID:25977097 CASPR2 forms a complex with GPR37 via MUPP1 but not with GPR... | ACCEPT | Summary: Experimental colocalisation of GPR37 with MUPP1 and CASPR2 at synapses of primary hippocampal neurons. Reason: Well supported, and the ASD-associated R558Q variant fails to reach the synapse, which strengthens the localisation claim. Supporting Evidence: PMID:25977097 In primary hippocampal neurons, GPR37 co-localized with MUPP1 and CASPR2 at the synapse, but not GPR37(R558Q). |
| GO:0034614 cellular response to reactive oxygen species | ISS PMID:23690594 GPR37 and GPR37L1 are receptors for the neuroprotective and ... | KEEP AS NON CORE | Summary: Sequence-similarity-based assignment of a role in the cellular response to reactive oxygen species. Reason: Same reasoning as the IEA row for this term: an indirect, downstream protective outcome of prosaposin-GPR37 signalling in astrocytes rather than a direct activity of the receptor. Supporting Evidence: PMID:23690594 Moreover, both prosaptide and prosaposin were found to protect primary astrocytes against oxidative stress, with these protective effects being attenuated by siRNA-mediated knockdown of endogenous astrocytic GPR37 or GPR37L1. |
| GO:0042923 neuropeptide binding | IPI PMID:16443751 The neuropeptide head activator is a high-affinity ligand fo... | REMOVE | Summary: Neuropeptide binding asserted from the 2006 report that the Hydra peptide head activator binds GPR37 with nanomolar affinity. The WITH/FROM entity is UniProtKB:P69251, the Hydra vulgaris head-activator peptide. Reason: The sole evidence is the 2006 head-activator (HA) pairing, which has since been contradicted from three directions. (i) Dunham et al. 2009 attempted to replicate it and found no HA-mediated internalisation, ERK1/2 phosphorylation or cAMP stimulation. (ii) HA was included in a screen across all remaining orphan GPCRs and was not a hit for any receptor (Southern et al. 2013). (iii) Head activator is a Hydra undecapeptide and no corresponding gene has been found in the human genome (Davenport et al. 2013) - note that the WITH/FROM field of the GOA row is literally UniProtKB:P69251, MORN_HYDVU, the Hydra vulgaris morphogenetic neuropeptide. Even the prosaposin paper, which is sympathetic to the HA result, concedes that HA is not a true ortholog of prosaptide and proposes only that it may act as an agonist by similarity. This is a published, direct failure to replicate rather than a disagreement with a curator over a paper I have not read, so REMOVE is appropriate. No content is lost: the peptide-receptor function of GPR37 is retained by the GO:0008528 and GO:0036505 annotations, which rest on the prosaposin work. REMOVE rather than MARK_AS_OVER_ANNOTATED: over-annotation would imply the claim is partly true of the human protein, and it is not. The ligand is a Hydra undecapeptide with no human gene, the WITH/FROM on the binding row names UniProtKB:P69251 (MORN_HYDVU) directly, and an independent group failed to reproduce the internalisation, ERK1/2 and cAMP responses. That is refutation, not over-reach. CLAUDE.md lists genuinely contradicted functions as valid grounds for REMOVE and bars only second-guessing an experimental annotation from an unread abstract; the contradiction here is published and independent. Nothing is lost: the peptide-receptor claim survives on GO:0008528 and GO:0036505 via prosaposin. Supporting Evidence: PMID:26635605 These studies contrast with that of Dunham et al. (2009), who attempted to replicate the finding that HA was a ligand for GPR37 but found no evidence of HA-mediated internalization, ERK1/2 phosphorylation or cAMP stimulation. PMID:26635605 perhaps the most damning evidence against HA as the endogenous ligand for GPR37 is the fact that it has not been found in the human genome (Davenport et al., 2013). Thus, it appears that HA is unlikely to be an agonist at GPR37 and is certainly not its endogenous ligand. PMID:23690594 HA does not seem to be a true ortholog of prosaptide, but nonetheless we propose that this invertebrate peptide may possess the ability to act as a GPR37 agonist owing to its similarity to prosaptide and prosaposin. |
| GO:0005737 cytoplasm | IDA PMID:16443751 The neuropeptide head activator is a high-affinity ligand fo... | MODIFY | Summary: Cytoplasmic localisation observed when GPR37 is overexpressed, where the receptor aggregates and is retained intracellularly. Reason: The observation is real but the term misdescribes it. GPR37 is a multi-pass membrane protein; what the authors saw was intracellular retention of an aggregation-prone receptor, which other work localises specifically to the endoplasmic reticulum, where GPR37 is a substrate for parkin- and HRD1-mediated degradation and where disease-associated variants accumulate. Endoplasmic reticulum membrane is the accurate compartment. Proposed replacements: endoplasmic reticulum membrane Supporting Evidence: PMID:16443751 Overexpression of GPR37 led to aggregate formation, retention of the receptor in the cytoplasm and low survival rates of transfected cells PMID:17059562 This study shows that HRD1 was expressed in substantia nigra pars compacta (SNC) dopaminergic neurons and interacted with Pael-R through the HRD1 proline-rich region, promoting the ubiquitylation and degradation of Pael-R. |
| GO:0007218 neuropeptide signaling pathway | IDA PMID:16443751 The neuropeptide head activator is a high-affinity ligand fo... | REMOVE | Summary: Neuropeptide signalling pathway asserted from head-activator-evoked calcium and Gi responses in GPR37-expressing cells. Reason: The sole evidence is the 2006 head-activator (HA) pairing, which has since been contradicted from three directions. (i) Dunham et al. 2009 attempted to replicate it and found no HA-mediated internalisation, ERK1/2 phosphorylation or cAMP stimulation. (ii) HA was included in a screen across all remaining orphan GPCRs and was not a hit for any receptor (Southern et al. 2013). (iii) Head activator is a Hydra undecapeptide and no corresponding gene has been found in the human genome (Davenport et al. 2013) - note that the WITH/FROM field of the GOA row is literally UniProtKB:P69251, MORN_HYDVU, the Hydra vulgaris morphogenetic neuropeptide. Even the prosaposin paper, which is sympathetic to the HA result, concedes that HA is not a true ortholog of prosaptide and proposes only that it may act as an agonist by similarity. This is a published, direct failure to replicate rather than a disagreement with a curator over a paper I have not read, so REMOVE is appropriate. No content is lost: the peptide-receptor function of GPR37 is retained by the GO:0008528 and GO:0036505 annotations, which rest on the prosaposin work. REMOVE rather than MARK_AS_OVER_ANNOTATED: over-annotation would imply the claim is partly true of the human protein, and it is not. The ligand is a Hydra undecapeptide with no human gene, the WITH/FROM on the binding row names UniProtKB:P69251 (MORN_HYDVU) directly, and an independent group failed to reproduce the internalisation, ERK1/2 and cAMP responses. That is refutation, not over-reach. CLAUDE.md lists genuinely contradicted functions as valid grounds for REMOVE and bars only second-guessing an experimental annotation from an unread abstract; the contradiction here is published and independent. Nothing is lost: the peptide-receptor claim survives on GO:0008528 and GO:0036505 via prosaposin. Supporting Evidence: PMID:26635605 These studies contrast with that of Dunham et al. (2009), who attempted to replicate the finding that HA was a ligand for GPR37 but found no evidence of HA-mediated internalization, ERK1/2 phosphorylation or cAMP stimulation. PMID:26635605 perhaps the most damning evidence against HA as the endogenous ligand for GPR37 is the fact that it has not been found in the human genome (Davenport et al., 2013). Thus, it appears that HA is unlikely to be an agonist at GPR37 and is certainly not its endogenous ligand. PMID:23690594 HA does not seem to be a true ortholog of prosaptide, but nonetheless we propose that this invertebrate peptide may possess the ability to act as a GPR37 agonist owing to its similarity to prosaptide and prosaposin. |
| GO:0008188 neuropeptide receptor activity | IDA PMID:16443751 The neuropeptide head activator is a high-affinity ligand fo... | REMOVE | Summary: Neuropeptide receptor activity asserted from the head-activator pairing. Reason: The sole evidence is the 2006 head-activator (HA) pairing, which has since been contradicted from three directions. (i) Dunham et al. 2009 attempted to replicate it and found no HA-mediated internalisation, ERK1/2 phosphorylation or cAMP stimulation. (ii) HA was included in a screen across all remaining orphan GPCRs and was not a hit for any receptor (Southern et al. 2013). (iii) Head activator is a Hydra undecapeptide and no corresponding gene has been found in the human genome (Davenport et al. 2013) - note that the WITH/FROM field of the GOA row is literally UniProtKB:P69251, MORN_HYDVU, the Hydra vulgaris morphogenetic neuropeptide. Even the prosaposin paper, which is sympathetic to the HA result, concedes that HA is not a true ortholog of prosaptide and proposes only that it may act as an agonist by similarity. This is a published, direct failure to replicate rather than a disagreement with a curator over a paper I have not read, so REMOVE is appropriate. No content is lost: the peptide-receptor function of GPR37 is retained by the GO:0008528 and GO:0036505 annotations, which rest on the prosaposin work. REMOVE rather than MARK_AS_OVER_ANNOTATED: over-annotation would imply the claim is partly true of the human protein, and it is not. The ligand is a Hydra undecapeptide with no human gene, the WITH/FROM on the binding row names UniProtKB:P69251 (MORN_HYDVU) directly, and an independent group failed to reproduce the internalisation, ERK1/2 and cAMP responses. That is refutation, not over-reach. CLAUDE.md lists genuinely contradicted functions as valid grounds for REMOVE and bars only second-guessing an experimental annotation from an unread abstract; the contradiction here is published and independent. Nothing is lost: the peptide-receptor claim survives on GO:0008528 and GO:0036505 via prosaposin. Supporting Evidence: PMID:26635605 These studies contrast with that of Dunham et al. (2009), who attempted to replicate the finding that HA was a ligand for GPR37 but found no evidence of HA-mediated internalization, ERK1/2 phosphorylation or cAMP stimulation. PMID:26635605 perhaps the most damning evidence against HA as the endogenous ligand for GPR37 is the fact that it has not been found in the human genome (Davenport et al., 2013). Thus, it appears that HA is unlikely to be an agonist at GPR37 and is certainly not its endogenous ligand. PMID:23690594 HA does not seem to be a true ortholog of prosaptide, but nonetheless we propose that this invertebrate peptide may possess the ability to act as a GPR37 agonist owing to its similarity to prosaptide and prosaposin. |
| GO:0009986 cell surface | IDA PMID:16443751 The neuropeptide head activator is a high-affinity ligand fo... | ACCEPT | Summary: Cell surface localisation of GPR37 observed in the head-activator study. Reason: Kept despite the ligand claim in the same paper being refuted: the localisation itself is independently confirmed by later work showing that GPR37 matures, is exported from the ER and is present at the cell surface. A refuted pharmacology does not invalidate a separate localisation observation. Supporting Evidence: PMID:26869225 GPR37 is expressed at the cell surface predominantly in the N-terminally truncated form. |
| GO:0007193 adenylate cyclase-inhibiting G protein-coupled receptor signaling pathway | IDA PMID:23690594 GPR37 and GPR37L1 are receptors for the neuroprotective and ... | ACCEPT | Summary: Prosaptide and prosaposin stimulation of GPR37 inhibits forskolin-stimulated cAMP production in a pertussis-toxin-sensitive manner. Reason: Direct experimental evidence for Gi coupling, the best-defined transducer arm of this receptor. Supporting Evidence: PMID:23690594 Prosaptide stimulation of cells transfected with GPR37 or GPR37L1 induced the phosphorylation of ERK in a pertussis toxin-sensitive manner, stimulated 35 S-GTPΞ³S binding, and promoted the inhibition of forskolin-stimulated cAMP production. |
| GO:0043410 positive regulation of MAPK cascade | IDA PMID:23690594 GPR37 and GPR37L1 are receptors for the neuroprotective and ... | ACCEPT | Summary: Ligand-stimulated, GPR37-dependent ERK phosphorylation. Reason: Direct experimental evidence; core downstream process. Supporting Evidence: PMID:23690594 Prosaptide stimulation of cells transfected with GPR37 or GPR37L1 induced the phosphorylation of ERK in a pertussis toxin-sensitive manner |
| GO:0008528 G protein-coupled peptide receptor activity | IDA PMID:23690594 GPR37 and GPR37L1 are receptors for the neuroprotective and ... | ACCEPT | Summary: GPR37 is activated by a peptide ligand and transduces the signal through a heterotrimeric G protein. Reason: Direct experimental evidence, and the parent term of the more specific prosaposin receptor activity annotation. Supporting Evidence: PMID:23690594 Prosaptide stimulation of cells transfected with GPR37 or GPR37L1 induced the phosphorylation of ERK in a pertussis toxin-sensitive manner, stimulated 35 S-GTPΞ³S binding, and promoted the inhibition of forskolin-stimulated cAMP production. |
| GO:0042277 peptide binding | IPI PMID:23690594 GPR37 and GPR37L1 are receptors for the neuroprotective and ... | KEEP AS NON CORE | Summary: Peptide binding, recorded with mouse prosaposin (Q61207) as the partner. Reason: The binding is real, but the generic term adds nothing over GO:0036505 prosaposin receptor activity from the same reference. Same action as the IEA row for this term. |
| GO:0036505 prosaposin receptor activity | IDA PMID:23690594 GPR37 and GPR37L1 are receptors for the neuroprotective and ... | ACCEPT | Summary: Direct experimental pairing of GPR37 with prosaptide and with purified full-length prosaposin - binding, receptor-dependent endocytosis, GTPgammaS loading, ERK phosphorylation and cAMP inhibition. Reason: Core molecular function, and the only proposed GPR37 ligand pairing with support from more than one laboratory. The original screen (Hall lab) found prosaptide to be the single hit among orphan neuropeptides, showed binding, receptor-dependent endocytosis, pertussis-toxin-sensitive ERK phosphorylation, GTPgammaS loading and cAMP inhibition, and reproduced the effects with purified full-length prosaposin. The Ji lab subsequently reproduced TX14 binding and GPR37-dependent calcium signalling in an independent system, and in 2026 showed that the in vivo actions of intrathecal TX14A require GPR37 specifically in TRPV1-lineage sensory neurons. Reservations remain and are recorded rather than suppressed - the agonist concentration needed is around 100 nM, higher than is usual for peptide/GPCR pairs; the pairing had not been ratified by the IUPHAR Nomenclature Committee as of the 2015 review of this receptor; and prosaposin may additionally act as a folding chaperone that increases GPR37 cell-surface expression, which would produce some of the same readouts without orthosteric agonism. Supporting Evidence: PMID:23690594 We screened these receptors for potential activation by various orphan neuropeptides, and these screens yielded a single positive hit: prosaptide, which promoted the endocytosis of GPR37 and GPR37L1, bound to both receptors and activated signaling in a GPR37- and GPR37L1-dependent manner. PMID:30010619 NPD1 and TX14 also bind to GPR37 and cause GPR37-dependent iCa2+ increases in peritoneal MΦs. PMID:42144155 Global GPR37 knockout or conditional knockout of GPR37 in TRPV1-lineage sensory neurons abolished the long-term inhibitory effect and unpriming effect of i.th. TX14A, confirming that GPR37 in this specific cellular population mediates these effects. |
| GO:0005783 endoplasmic reticulum | IDA PMID:17059562 A ubiquitin ligase HRD1 promotes the degradation of Pael rec... | ACCEPT | Summary: Experimental localisation of GPR37/Pael-R to the endoplasmic reticulum, where HRD1 ubiquitylates it. Reason: Well supported and reproduced across laboratories; the ER pool is a genuine and functionally consequential compartment for this misfolding-prone receptor. Supporting Evidence: PMID:17059562 This study shows that HRD1 was expressed in substantia nigra pars compacta (SNC) dopaminergic neurons and interacted with Pael-R through the HRD1 proline-rich region, promoting the ubiquitylation and degradation of Pael-R. |
| GO:0031625 ubiquitin protein ligase binding | IPI PMID:17059562 A ubiquitin ligase HRD1 promotes the degradation of Pael rec... | KEEP AS NON CORE | Summary: GPR37/Pael-R binds the E3 ubiquitin ligase HRD1/SYVN1 through its proline-rich region. Reason: Robustly supported and, in reproducibility terms, the strongest molecular claim about GPR37: the parkin interaction was found by unbiased screening, the parkin/CHIP/Hsp70/Pael-R complex was confirmed in vitro and in vivo, and a second E3, HRD1/SYVN1, was independently shown to ubiquitylate the same substrate in dopaminergic neurons. It is nevertheless marked non-core, because what it describes is ER quality control acting on an aggregation-prone receptor - a consequence of how badly GPR37 folds - rather than a signalling function the protein evolved to perform. The receptor role captured by GO:0036505 and GO:0007193 remains the core function. Supporting Evidence: PMID:17059562 This study shows that HRD1 was expressed in substantia nigra pars compacta (SNC) dopaminergic neurons and interacted with Pael-R through the HRD1 proline-rich region, promoting the ubiquitylation and degradation of Pael-R. |
| GO:0005886 plasma membrane | IDA PMID:11439185 An unfolded putative transmembrane polypeptide, which can le... | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0031625 ubiquitin protein ligase binding | IPI PMID:11439185 An unfolded putative transmembrane polypeptide, which can le... | KEEP AS NON CORE | Summary: GPR37/Pael-R binds parkin (PRKN) and is ubiquitinated by it. Reason: Robustly supported and, in reproducibility terms, the strongest molecular claim about GPR37: the parkin interaction was found by unbiased screening, the parkin/CHIP/Hsp70/Pael-R complex was confirmed in vitro and in vivo, and a second E3, HRD1/SYVN1, was independently shown to ubiquitylate the same substrate in dopaminergic neurons. It is nevertheless marked non-core, because what it describes is ER quality control acting on an aggregation-prone receptor - a consequence of how badly GPR37 folds - rather than a signalling function the protein evolved to perform. The receptor role captured by GO:0036505 and GO:0007193 remains the core function. Supporting Evidence: PMID:11439185 A putative G protein-coupled transmembrane polypeptide, named Pael receptor, was identified as an interacting protein with Parkin, a gene product responsible for autosomal recessive juvenile Parkinsonism (AR-JP). PMID:11439185 Parkin specifically ubiquitinates this receptor in the presence of ubiquitin-conjugating enzymes resident in the endoplasmic reticulum and promotes the degradation of insoluble Pael receptor |
| GO:0005515 protein binding | IPI PMID:14532270 A product of the human gene adjacent to parkin is a componen... | REMOVE | Summary: Bare protein binding recording the GPR37-PACRG interaction. Reason: Uninformative as a molecular function. The underlying report, that the parkin- coregulated gene product is a Lewy body component and suppresses Pael-receptor-induced cell death, belongs to the ER-stress side of GPR37 biology and is captured by the ubiquitin-ligase- binding annotations. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0000151 ubiquitin ligase complex | IDA PMID:12150907 CHIP is associated with Parkin, a gene responsible for famil... | MARK AS OVER ANNOTATED | Summary: Part-of assignment to a ubiquitin ligase complex, from the demonstration that CHIP, Hsp70, parkin and Pael-R form a complex. Reason: The complex is real, but GPR37 is the substrate held by that complex, not a component of the ligase. Standard GO practice does not make substrates part_of the E3 complex that modifies them, and doing so here implies GPR37 contributes to ubiquitin-ligase activity, which it does not. The relationship is better carried by the ubiquitin protein ligase binding annotations this gene already has. Supporting Evidence: PMID:12150907 Here, we show that CHIP, Hsp70, Parkin, and Pael-R formed a complex in vitro and in vivo. |
| GO:0005886 plasma membrane | IDA PMID:12150907 CHIP is associated with Parkin, a gene responsible for famil... | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0030544 Hsp70 protein binding | IPI PMID:12150907 CHIP is associated with Parkin, a gene responsible for famil... | KEEP AS NON CORE | Summary: GPR37/Pael-R associates with Hsp70 (HSPA8), which is displaced by CHIP during ER stress. Reason: Correct and specific, but it describes chaperone handling of a misfolding-prone receptor rather than a core signalling function. Non-core. Supporting Evidence: PMID:12150907 CHIP promoted the dissociation of Hsp70 from Parkin and Pael-R, thus facilitating Parkin-mediated Pael-R ubiquitination. |
| GO:0031072 heat shock protein binding | IPI PMID:12150907 CHIP is associated with Parkin, a gene responsible for famil... | KEEP AS NON CORE | Summary: Heat shock protein binding, the parent claim of the Hsp70 interaction above. Reason: Correct but less specific than the Hsp70 protein binding annotation from the same reference, and non-core for the same reason. |
| GO:0031625 ubiquitin protein ligase binding | IPI PMID:12150907 CHIP is associated with Parkin, a gene responsible for famil... | KEEP AS NON CORE | Summary: GPR37/Pael-R binds parkin and CHIP/STUB1 within a quality-control complex. Reason: Robustly supported and, in reproducibility terms, the strongest molecular claim about GPR37: the parkin interaction was found by unbiased screening, the parkin/CHIP/Hsp70/Pael-R complex was confirmed in vitro and in vivo, and a second E3, HRD1/SYVN1, was independently shown to ubiquitylate the same substrate in dopaminergic neurons. It is nevertheless marked non-core, because what it describes is ER quality control acting on an aggregation-prone receptor - a consequence of how badly GPR37 folds - rather than a signalling function the protein evolved to perform. The receptor role captured by GO:0036505 and GO:0007193 remains the core function. Supporting Evidence: PMID:12150907 Here, we show that CHIP, Hsp70, Parkin, and Pael-R formed a complex in vitro and in vivo. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-380073 | ACCEPT | Summary: Plasma membrane localisation asserted by a Reactome reaction in which GPR37 acts as a liganded Gi-activating GPCR. Reason: Correct location. Note that these Reactome reactions encode the prosaptide pairing (R-HSA-5336182 is "GPR37 binds prosaptide"), so Reactome has already taken a position on the ligand question that matches the GO:0036505 annotation retained here. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-5336182 | ACCEPT | Summary: Plasma membrane localisation asserted by a Reactome reaction in which GPR37 acts as a liganded Gi-activating GPCR. Reason: Correct location. Note that these Reactome reactions encode the prosaptide pairing (R-HSA-5336182 is "GPR37 binds prosaptide"), so Reactome has already taken a position on the ligand question that matches the GO:0036505 annotation retained here. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-749454 | ACCEPT | Summary: Plasma membrane localisation asserted by a Reactome reaction in which GPR37 acts as a liganded Gi-activating GPCR. Reason: Correct location. Note that these Reactome reactions encode the prosaptide pairing (R-HSA-5336182 is "GPR37 binds prosaptide"), so Reactome has already taken a position on the ligand question that matches the GO:0036505 annotation retained here. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-749456 | ACCEPT | Summary: Plasma membrane localisation asserted by a Reactome reaction in which GPR37 acts as a liganded Gi-activating GPCR. Reason: Correct location. Note that these Reactome reactions encode the prosaptide pairing (R-HSA-5336182 is "GPR37 binds prosaptide"), so Reactome has already taken a position on the ligand question that matches the GO:0036505 annotation retained here. |
| GO:0043235 signaling receptor complex | IDA PMID:23382219 Structural basis for endosomal trafficking of diverse transm... | KEEP AS NON CORE | Summary: Part-of a signalling receptor complex, curated from a structural study of endosomal sorting by PX-FERM proteins (SNX17/SNX27/SNX31). Reason: Retained, not removed. GPR37 does not appear anywhere in the cached text of this paper, so I cannot verify what the curator saw - most likely a supplementary cargo-interaction table behind the statement that the PX-FERM proteins bind a wide array of putative cargo molecules. Under project policy an experimental annotation is not removed because the cached text does not mention the gene. The term is generic and the evidence unverifiable from here, so it is kept as non-core. Same action as the IBA row for this term. |
| GO:0004930 G protein-coupled receptor activity | TAS PMID:9144577 A novel endothelin receptor type-B-like gene enriched in the... | ACCEPT | Summary: Author statement from the paper that cloned GPR37 as an endothelin-B-receptor-like gene enriched in brain. Reason: Correct by sequence and confirmed functionally. Note that neither GPR37 nor GPR37L1 was ever found to bind endothelins or related peptides, so the receptor name records ancestry, not pharmacology. Supporting Evidence: PMID:26635605 Originally identified through searches for homologs of endothelin and bombesin receptors, neither GPR37 nor GPR37L1 were found to bind endothelins or related peptides. |
| GO:0005886 plasma membrane | TAS PMID:9144577 A novel endothelin receptor type-B-like gene enriched in the... | ACCEPT | Summary: GPR37 is a multi-pass plasma membrane receptor. Reason: Correct and independently established by several studies; this is where the receptor engages its G protein. |
| GO:0007186 G protein-coupled receptor signaling pathway | TAS PMID:9144577 A novel endothelin receptor type-B-like gene enriched in the... | ACCEPT | Summary: Author statement that the cloned receptor signals as a GPCR. Reason: Correct; same action as the IEA row for this term. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Can a single receptor really be both a peptide receptor and a lipid receptor? Prosaptide/TX14A and protectin D1 are asserted to activate GPR37 in the same assays and often in the same papers, yet they belong to entirely different chemical classes. Nobody has reported competition binding of protectin D1 against labelled prosaptide, or mutagenesis separating the two responses, so it is unresolved whether GPR37 has two distinct sites, whether one of the two acts indirectly, or whether the shared calcium/phagocytosis readout is simply not a clean measure of orthosteric occupancy.
Q: Is the osteocalcin-GPR37 pairing independently replicable? The 2021 oligodendrocyte study and the 2026 ventral tegmental area study share senior authors and the same institute, so they corroborate rather than replicate. Given that osteocalcin already has two other claimed receptors (GPRC6A and GPR158), a demonstration of direct OCN binding to purified GPR37 from a laboratory with no stake in the original claim is needed before a molecular function term is justified.
Q: Does prosaposin agonise GPR37 or chaperone it? The alternative reading of the prosaposin data, raised in the 2015 review of this receptor and never formally excluded, is that prosaposin improves GPR37 folding and cell-surface delivery, and that downstream readouts follow from increased surface receptor rather than from orthosteric activation. Agonist-independent and agonist-dependent effects have not been separated.
Q: Has the IUPHAR Nomenclature Committee ratified any GPR37 deorphanization? As of the most recent authoritative review of this receptor the prosaposin pairing had not been ratified. Whether that position has changed matters directly for how confidently GO:0036505 should be treated as core.
Q: Is being a parkin and HRD1 substrate a function or a liability? The Pael-R work is the most reproducible molecular observation about GPR37, but it describes ER quality control of a protein that folds badly under overexpression. Whether endogenous GPR37 is a physiological ERAD client at normal expression levels, or whether the phenomenon is an artefact of overexpression that happens to be disease-relevant, has not been settled.
Q: What does constitutive N-terminal shedding do? GPR37 is cleaved by metalloproteinase between E167 and Q168 and the ectodomain is released, so most surface receptor is N-terminally truncated. Whether the shed ectodomain has activity of its own, and whether the truncated receptor differs pharmacologically from the full-length form, is unknown and could explain discrepant ligand results between laboratories.
Experiment: Perform saturation and competition radioligand or fluorescence-polarisation binding on purified GPR37 in nanodiscs, using labelled TX14A and unlabelled protectin D1 and vice versa, in parallel with GPR37L1 and an unrelated class A GPCR as controls. Mutually competitive displacement would establish a shared site; non-competitive behaviour would point to two sites and would justify two separate molecular function terms.
Hypothesis: Prosaptide and protectin D1 occupy the same site on GPR37.
Type: binding
Experiment: Reconstitute purified GPR37 with heterotrimeric Gi in nanodiscs and measure GTP turnover or conformational change (BRET/FRET sensor, or HDX-MS) in response to protectin D1 in a system containing no other cellular protein. A cell-free response would settle the direct-agonism question that cell-based calcium assays cannot.
Hypothesis: Protectin D1 activates GPR37 directly rather than through an intermediate.
Type: biochemistry
Experiment: In an independent laboratory, test binding of labelled osteocalcin to purified GPR37 and to GPRC6A and GPR158 side by side, and test whether osteocalcin-evoked cAMP reduction survives in cells expressing GPR37 but lacking GPR158 and GPRC6A. Include the carboxylated and uncarboxylated forms, since these differ in the peripheral osteocalcin literature.
Hypothesis: Osteocalcin binds GPR37 directly.
Type: binding
Experiment: Compare surface GPR37 by biotinylation and by a non-permeabilised antibody feed after short (minutes) and long (hours) prosaposin exposure, and test whether prosaposin still produces ERK and cAMP responses when surface receptor number is clamped, for example by using a receptor construct with a trafficking-independent surface anchor or by pre-saturating the surface pool.
Hypothesis: Prosaposin acts as a chaperone for GPR37 as well as, or instead of, as an agonist.
Type: cell biology
Experiment: A definitive negative would let the GO:0008188, GO:0007218 and GO:0042923 annotations be retired with a citable modern reference rather than a chain of secondary assertions. Test synthetic head activator across a modern transducer panel (TRUPATH, arrestin recruitment, calcium, cAMP) on human GPR37, with prosaptide as the positive control, in a GPR37-null background.
Hypothesis: Head activator has no action at human GPR37.
Type: pharmacology
Experiment: Compare ligand responses of full-length GPR37 with a receptor whose shedding site is mutated, and with a construct beginning at Q168, for prosaptide, protectin D1 and osteocalcin. If the truncated and full-length forms differ, the proportion of each form in a given expression system would explain much of the inter-laboratory disagreement about this receptor.
Hypothesis: N-terminal shedding changes GPR37 pharmacology.
Type: pharmacology
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)