br Materials and methods br
Materials and methods
Discussion In this study, the orphan receptor GPR25 was cloned from several representative vertebrate species including zebrafish. Although the overall amino KPT-9274 sequence identity among these species is not high (38–52%), our synteny analysis clearly indicates that the cloned zebrafish, spotted gar and pigeon GPR25 are orthologous to human GPR25 (Supplementary Fig. 4). In addition, we noted that the seven transmembrane domains of vertebrate GPR25 share higher sequence identities with those of human GPR25, ranging from 48% to 53%. Like other members of class A GPCR, the conserved D/ERY motif for G protein-coupling and NPxxY motif are also observed in these cloned GPR25. All these conserved structural features imply that GPR25 may be a functional receptor in vertebrates. Moreover, we noted that vertebrate GPR25 shows a relatively high degree of amino acid sequence identity with that of zebrafish APLNR (34%) and human APLNR (29%), particularly in their transmembrane domains. This finding implies that GPR25 may have a close evolutionary relationship with APLNR. This speculation was supported by the phylogenetic analysis shown in Supplementary Fig. 5, in which vertebrate GPR25 shows a much closer evolutionary relationship to vertebrate APLNR, than to other members of class A GPCR family, such as angiotensin II receptor type 1/2 (AGTR1/2). The structural similarity, together with the close phylogenetic relationship, between GPR25 and APLNR led us to hypothesize that GPR25 and APLNR may share functional similarity in vertebrates, and GPR25 can also be activated by the two known endogenous ligands of APLNR: Apelin and Apela. In agreement with this hypothesis, we proved that like zebrafish/human APLNR, zebrafish GPR25 could be activated by both Apelin and Apela, and receptor activation could inhibit forskolin-stimulated cAMP production, monitored by a pGL3-CRE-luciferase reporter system [17,18]. Our finding, for the first time, indicates that zebrafish GPR25 possesses two potential ligands ‘Apelin and Apela’ and is functionally coupled to Gi protein(s), whose activation inhibits adenylate cyclase (AC) activity, similar to APLNR. Like zebrafish GPR25, pigeon and spotted gar GPR25 could also be activated by Apelin and Apela (Fig. 3A-D). Strikingly, we found that chicken Apelin-36 and Apela-32 seem unable to activate human GPR25 (Fig. 3E), though both peptides could activate human APLNR potently (Fig. 3F). Since confocal microscopy indicates human GPR25 could localize to the cell membrane (Fig. 4D), the inability of Apelin/Apela in activating human GPR25 is likely due to their extremely low (or no) binding affinity to GPR25 in vitro. This is partially congruent with the finding in a recent study, in which the recombinant alkaline phosphatase-conjugated Apela cannot bind to rat GPR25 in vitro . Taken together, our data suggest that like APLNR, GPR25 could be activated by Apelin and Apela in vertebrates, or, at the very least, in non-mammalian vertebrates. Ligand-induced receptor endocytosis is a common feature of many GPCRs in response to ligand stimulation . In this study, we found that zebrafish GPR25 is localized at the cell surface. Nevertheless, human GPR25 and APLNR could also show some degree of intracellular retention (Fig. 4), as suggested in previous studies reporting intracellular retention of human APLNR expressed in HEK293 cells [6,20]. Like human APLNR (Fig. 4) [6,20], zebrafish GPR25 is rapidly internalized and forms the large, hollow vesicles upon Apelin and Apela stimulation (Fig. 4), whereas human GPR25 seems unable to undergo internalization. This finding further supports the notion that both Apelin and Apela could activate GPR25 in non-mammalian vertebrates. The activation of GPR25 by Apelin and Apela in non-mammalian vertebrates, and not in mammals, suggests that mammalian GPR25 may have undergone a dramatic structural change after the divergence of birds and mammals, thus leading to its functional difference between non-mammalian and mammalian species (Fig. 2, Fig. 3, Fig. 4). Similar to the findings in the present study, another orphan receptor Bombesin receptor subtype-3 (BRS3) was demonstrated to be potently activated by gastrin-releasing peptide (GRP) and neuromedin B (NMB) in non-mammalian vertebrates (chickens and spotted gars), but not in mammals (mice) in our recent study . These studies from our laboratory may provide an alternative key approach to de-orphanize other orphan GPCRs from an evolutionary perspective in future. In this study, we also noted that activation of GPR25 in zebrafish, spotted gars and pigeons requires high concentrations of Apelin or Apela (≥100 nM), therefore, we still cannot exclude the possibility that other potential ‘high-affinity’ ligand(s) for GPR25 may exist in vertebrates.