Ligands of orphan receptors
Which natural signalling molecules act on the orphan receptors, whose own ligands are still unknown?
Summary
Why it is open
A receptor is called an orphan when it resembles a signalling receptor but has no validated endogenous ligand, the natural molecule that activates it in the body. Abigail Jobe and Ranjit Vijayan write that orphans account for about 30% of the roughly 400 non-olfactory human G protein-coupled receptors (GPCRs), the seven-transmembrane receptors carrying many hormone and neurotransmitter signals [10]. The Concise Guide to Pharmacology 2025/26, current to mid-2025, lists 197 class A (rhodopsin-like) GPCRs with known ligands and 87 class A orphans, 54 of them with an endogenous ligand proposed in at least one publication; it lists 8 orphans in class C, 1 with a proposal, and 33 in the adhesion family, 7 with a proposal [13].
Nuclear receptors, transcription factors normally switched on by small lipophilic molecules, have their own orphans. In the same guide the NC-IUPHAR subcommittee states that the Rev-Erb, testicular, tailless-like, COUP-TF-like, oestrogen-related, nerve growth factor IB-like (NR4A), Fushi tarazu F1-like, germ cell nuclear factor and DAX-like receptors have yet to be officially paired with an endogenous ligand, and that the retinoic acid-related orphan (ROR) receptors have yet to be assigned a definitive one; Rev-Erbs are thought to be activated by heme, and RORα may be synthesised with a "captured" agonist such as cholesterol [14].
NC-IUPHAR, the nomenclature committee of the International Union of Basic and Clinical Pharmacology, sets the bar. Anthony Davenport and colleagues state that reproducibility is the minimum criterion: two or more refereed papers from independent groups should show activity at a potency consistent with a physiological function, and the pairing should be repeated by at least one independent laboratory, ideally by a different assay. They also ask that selective agonists mimic and antagonists block the ligand's action, that it be present in tissues at appropriate concentrations, and that deleting the gene abolish the receptor's characteristics [3]. Three questions are therefore open: whether a receptor has an endogenous ligand, which molecule it is, and whether its activity is intrinsic instead.
What is known
Screens and their yield (2013 to 2019). Chris Southern and colleagues assembled 10,500 candidate ligands and screened 82 GPCRs using β-arrestin recruitment, a readout independent of G proteins; they described one candidate endogenous ligand and several surrogate ligands, but for most orphan receptors studied the assay found no cognate ligand [2]. Simon Foster and colleagues combined comparative genomics across 313 species with analysis of human class A GPCR sequences and structures, then screened with three assays, pairing 17 proposed endogenous peptides with five orphan receptors: bombesin receptor 3, GPR1, GPR15, GPR55 and GPR68 [6].
Pairings proposed since 2023. Thibaut Laboute and colleagues reported in 2023 that GPR158 acts as a metabotropic glycine receptor: glycine and taurine bind directly to a Cache domain of the receptor, glycine acting through it reduces production of the second messenger cyclic AMP, and glycine, unlike taurine, changed the excitability of cortical neurons [9]. The evidence is one group's, from transfected cells and rodent neurons.
CXCL17, a mucosal protein, shows how a receptor and a molecule can be orphans of each other; a signal with no known target appears in this catalogue as endothelium-derived hyperpolarising factor. In 2015 José Maravillas-Montero and colleagues proposed GPR35 as the receptor of CXCL17 and suggested the name CXCR8 [4]. Nor Azreen Binti Mohd Amir and colleagues could not reproduce this in 2018: GPR35 failed to signal in response to CXCL17 in four assay types, and they concluded that it "is unlikely to be a bona fide receptor for CXCL17" and that the monocytic cell line THP-1 expresses an as yet unidentified receptor for it [5]. In 2024 Borja Ocón and colleagues identified GPR25 as a lymphocyte receptor for CXCL17 from omics datasets, with homing experiments in mice [11]. In 2025 Wen-Feng Hu and colleagues reported the same pairing, predicted with AlphaFold 3 and then tested: CXCL17 activated GPR25 with a half-maximal effect at about 100 nM in a β-arrestin assay and had no effect on 17 other receptors tested, though they write that the hypothesis that CXCL17 is its endogenous agonist "needs to be tested experimentally in future studies" [12].
Claims that did not hold. Jobe and Vijayan record a decline in deorphanisation over the past decade, with pairings retracted for failed reproducibility [10]. In 2021 Tony Ngo and colleagues retracted their 2017 Nature Chemical Biology paper reporting the first surrogate ligands for the orphan GPR37L1, after finding a cloning error in the construct used for compound screening: the corrected construct "no longer displays constitutive activity" and the predicted compounds "do not evoke a GPR37L1-mediated response", while the computational method is unaffected [8]. Davenport and colleagues note that a pairing reported by two independent groups has sometimes not been reproduced by others, as with obestatin and GPR39, where the original findings "may have been caused by contamination of the sample of obestatin by impurities" [3].
Receptors that may not need a ligand. Zhulun Wang and colleagues solved the ligand-binding domain of the orphan nuclear receptor Nurr1 at 2.2 Å: its fold resembles that of agonist-bound, active domains, but bulky hydrophobic side chains fill the region ligands normally occupy, leaving no cavity. The domain can still be regulated in mammalian cells, which they present as a model for ligand-independent function [1]. In the orphan GPR52, Xi Lin and colleagues found that extracellular loop 2 occupies the orthosteric binding pocket and operates as a built-in agonist, giving the receptor an intrinsically high level of basal activity [7].
Recent claims using AI methods. Besides the AlphaFold 3 prediction for GPR25 [12], Larissa Ferguson and colleagues published DeorphaNN in Molecular Cell in August 2026: a graph neural network over AlphaFold-multimer predictions of active-state receptor-peptide complexes, trained on measured peptide activity in the nematode Caenorhabditis elegans. Applied to two orphan nematode receptors, it ranked 364 candidate peptides, and of the five highest-ranked for each receptor one activated NPR-44 and one activated NPR-33. The authors report that performance varied among receptor families and did not track similarity to the training receptors, so reliability for a new receptor is hard to estimate in advance [15].
Main approaches
Reverse pharmacology. Start from the receptor: express it in cells and look for a candidate that activates it [10]. Strongest evidence: the approach produced the first deorphanisations in 1988, of the serotonin 5-HT1A and dopamine D2 receptors [10], and in 2013 NC-IUPHAR recommended new receptor names on the basis of 11 class A pairings [3]. Main difficulty: Jobe and Vijayan call such techniques expensive, laborious and inconsistent [10], and receptor densities above native levels can produce pairings that do not reproduce [3].
High-throughput and β-arrestin screens. Screen large candidate libraries across many receptors in parallel. Strongest evidence: β-arrestin recruitment detected the established pairings used as controls [2] and found pairings no other platform in the Foster study did [6]. Main difficulty: for most of the 82 receptors screened against 10,500 candidates it found no cognate ligand [2]; the GPR55 and GPR68 peptides were all inactive in β-arrestin recruitment, while the GPR1 pairings appeared only there, so a pairing can depend on the assay [6].
Computational and structure-based approaches. Rank candidates by sequence, structure and machine learning before any assay. Strongest evidence: prediction followed by functional testing paired peptides with five orphan receptors [6]; an AlphaFold 3 model preceded the GPR25 experiments [12]; DeorphaNN placed an active peptide in the top five of 364 for each of two orphan receptors [15]. Main difficulty: predictions must still be confirmed in cells, and the DeorphaNN authors list the absence of a membrane context and of post-translational modifications in AlphaFold 2 predictions as limits [15].
Structure determination and constitutive activity. Solve the structure, including in an active state with no added ligand. Strongest evidence: GPR52 reaches a fully active state coupled to Gs with no external agonist, because its own extracellular loop acts as the agonist [7], and the Nurr1 domain has no ligand cavity [1]. Main difficulty: an active structure shows that a receptor can signal without a ligand, not that none exists, and the GPR52 authors expect their structures to help identify endogenous ligands [7]; constitutive activity also depends on the construct, as the GPR37L1 retraction showed [8].
Open sub-questions
- CXCL17 and GPR25 against the committee's bar. Do the two reports [11, 12] meet the NC-IUPHAR criterion of reproduction by an independent laboratory using a different assay, and is CXCL17 present at the expected concentrations where GPR25 is expressed [3]?
- Glycine and GPR158. Can an independent laboratory reproduce glycine-dependent signalling at GPR158 in a different assay, and does deleting the gene abolish it in the reported neurons [3, 9]?
- Assay-dependent pairings. Do the peptide pairings reported for GPR55 and GPR68 reproduce in further G protein readouts, and does the GPR1 result survive outside the β-arrestin assay [6]?
- Orphans with nothing proposed. For class A orphans that the 2025/26 guide lists without any proposed endogenous ligand [13], is constitutive activity measurable, and does it arise from a built-in agonist like that of GPR52 [7]?
- Cavity-less ligand-binding domains. Which other orphan nuclear receptor ligand-binding domains lack a cavity as Nurr1's does [1], and which families the guide lists as unpaired [14] could be regulated without a ligand?
- Transfer of a predictor between species. Does a model trained on nematode peptide-receptor activity rank true agonists highly for human orphan receptors prospectively, given that performance varied by receptor family [15]?
Work on this in Papex
If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.
Key references
- Wang, Z., Benoit, G., Liu, J., Prasad, S., Aarnisalo, P., Liu, X., Xu, H., Walker, N. P. & Perlmann, T. (2003). Structure and function of Nurr1 identifies a class of ligand-independent nuclear receptors. Nature 423(6939), 555–560. 10.1038/nature01645 · Checked 16 Oct 2026
- Southern, C., Cook, J. M., Neetoo-Isseljee, Z., Taylor, D. L., Kettleborough, C. A., Merritt, A. et al. (2013). Screening β-arrestin recruitment for the identification of natural ligands for orphan G-protein-coupled receptors. Journal of Biomolecular Screening 18(5), 599–609 (Crossref lists the journal under its present name, SLAS Discovery). 10.1177/1087057113475480 · Checked 16 Oct 2026
- Davenport, A. P., Alexander, S. P. H., Sharman, J. L., Pawson, A. J., Benson, H. E., Monaghan, A. E. et al. (2013). International Union of Basic and Clinical Pharmacology. LXXXVIII. G protein-coupled receptor list: recommendations for new pairings with cognate ligands. Pharmacological Reviews 65(3), 967–986. 10.1124/pr.112.007179 · Checked 16 Oct 2026
- Maravillas-Montero, J. L., Burkhardt, A. M., Hevezi, P. A., Carnevale, C. D., Smit, M. J. & Zlotnik, A. (2015). Cutting edge: GPR35/CXCR8 is the receptor of the mucosal chemokine CXCL17. The Journal of Immunology 194(1), 29–33 (published online 19 November 2014). 10.4049/jimmunol.1401704 · Checked 16 Oct 2026
- Binti Mohd Amir, N. A. S., Mackenzie, A. E., Jenkins, L., Boustani, K., Hillier, M. C., Tsuchiya, T., Milligan, G. & Pease, J. E. (2018). Evidence for the existence of a CXCL17 receptor distinct from GPR35. The Journal of Immunology 201(2), 714–724. 10.4049/jimmunol.1700884 · Checked 16 Oct 2026
- Foster, S. R., Hauser, A. S., Vedel, L., Strachan, R. T., Huang, X.-P., Gavin, A. C. et al. (2019). Discovery of human signaling systems: pairing peptides to G protein-coupled receptors. Cell 179(4), 895–908.e21. 10.1016/j.cell.2019.10.010 · Checked 16 Oct 2026
- Lin, X., Li, M., Wang, N., Wu, Y., Luo, Z., Guo, S. et al. (2020). Structural basis of ligand recognition and self-activation of orphan GPR52. Nature 579(7797), 152–157. 10.1038/s41586-020-2019-0 · Checked 16 Oct 2026
- Ngo, T., Ilatovskiy, A. V., Stewart, A. G., Coleman, J. L. J., McRobb, F. M., Riek, R. P., Graham, R. M., Abagyan, R., Kufareva, I. & Smith, N. J. (2021). Retraction note: orphan receptor ligand discovery by pickpocketing pharmacological neighbors. Nature Chemical Biology 17(4), 501, retracting Nature Chemical Biology 13(2), 235–242 (2017; published online 19 December 2016). 10.1038/s41589-021-00746-1 · Checked 16 Oct 2026
- Laboute, T., Zucca, S., Holcomb, M., Patil, D. N., Garza, C., Wheatley, B. A., Roy, R. N., Forli, S. & Martemyanov, K. A. (2023). Orphan receptor GPR158 serves as a metabotropic glycine receptor: mGlyR. Science 379(6639), 1352–1358. 10.1126/science.add7150 · Checked 16 Oct 2026
- Jobe, A. & Vijayan, R. (2024). Orphan G protein-coupled receptors: the ongoing search for a home. Frontiers in Pharmacology 15, 1349097. 10.3389/fphar.2024.1349097 · Checked 16 Oct 2026
- Ocón, B., Xiang, M., Bi, Y., Tan, S., Brulois, K., Ayesha, A. et al. (2024). A lymphocyte chemoaffinity axis for lung, non-intestinal mucosae and CNS. Nature 635(8039), 736–745 (published online 18 September 2024). 10.1038/s41586-024-08043-2 · Checked 16 Oct 2026
- Hu, W.-F., Yu, J., Wang, J.-J., Sun, R.-J., Zheng, Y.-S., Zhang, T., Liu, Y.-L., Xu, Z.-G. & Guo, Z.-Y. (2025). Identification of orphan GPR25 as a receptor for the chemokine CXCL17. The FEBS Journal 292(22), 5998–6015 (published online 25 April 2025). 10.1111/febs.70117 · Checked 16 Oct 2026
- Alexander, S. P. H., Davenport, A. P., Kelly, E., Gibb, A. J., Mathie, A. A., Peach, C. J. et al. (2025). The Concise Guide to Pharmacology 2025/26: G protein-coupled receptors. British Journal of Pharmacology 182(S1), S24–S151. 10.1111/bph.70230 · Checked 16 Oct 2026
- Alexander, S. P. H., Cidlowski, J. A., Gibb, A. J., Kelly, E., Mathie, A. A., Peach, C. J. et al. (2025). The Concise Guide to Pharmacology 2025/26: nuclear hormone receptors. British Journal of Pharmacology 182(S1), S242–S258. 10.1111/bph.70232 · Checked 16 Oct 2026
- Ferguson, L., Ouellet, S., Vandewyer, E., Wang, C., Wunna, Z., Lim, T. K. Y., Schafer, W. R. & Beets, I. (2026). DeorphaNN: virtual screening of GPCR peptide agonists using AlphaFold-predicted active-state complexes and deep learning embeddings. Molecular Cell 86(15), 3102–3117.e5 (first published 23 July 2026; preprint bioRxiv 10.1101/2025.03.19.644234, 20 March 2025). 10.1016/j.molcel.2026.07.006 · Checked 16 Oct 2026
Source
Open, as listed by Wikipedia · Unsolved problems in biology. Checked 08 Oct 2026.
The status is the source’s. Papex does not decide whether a problem is solved.
Records on this problem
No public record addresses this problem yet.