SUMMARY Heterotrimetic G proteins consist of four subfamilies (GS, Gi/O, Gq/11, and G12/13) that mediate signaling via G-protein-coupled receptors (GPCRs), principally by receptors binding Gα C termini. G-protein-coupling profiles govern GPCR-induced cellular responses, yet receptor sequence selectivity determinants remain elusive. Here, we systematically quantified ligand-induced interactions between 148 GPCRs and all 11 unique Gα subunit C termini. For each receptor, we probed chimeric Gα subunit activation via a transforming growth factor-α (TGF-α) shedding response in HEK293 cells lacking endogenous Gq/11 and G12/13 proteins, and complemented G-protein-coupling profiles through a NanoBiT-G-protein dissociation assay. Interrogation of the dataset identified sequence-based coupling specificity features, inside and outside the transmembrane domain, which we used to develop a coupling predictor that outperforms previous methods. We used the predictor to engineer designer GPCRs selectively coupled to G12. This dataset of fine-tuned signaling mechanisms for diverse GPCRs is a valuable resource for research in GPCR signaling.

This article is available online at http://www.jlr.org with plasma concentrations of 10-30 nM and several hundred nanomoles, respectively ( 3, 4 ). Both LPA and S1P have critical roles in multiple cellular events through G protein-coupled receptors (GPCRs). Six GPCRs have been identifi ed for LPA (LPA 1-6 ) and fi ve GPCRs have been identifi ed for S1P (S1P 1-5 ) ( 5 ), and nomenclature of these LysoGPs receptors has recently been proposed by Kihara et al. ( 6 ). These receptors are grouped into two classes, the Edg and P2Y families, respectively. LPA 1-3 and all fi ve of the S1P receptors, S1P [1][2][3][4][5] , are members of the Edg family, while LPA 4-6 are members of the P2Y family. In addition, LPA is an endogenous ligand for PPAR ␥ ( 7 ), and was shown to activate transient receptor potential cation channel subfamily V member 1 (TRPV1) channels leading to an infl ux of Ca 2+ ions through TRPV1 ( 8 ). Studies on gene-targeted mice and human genetic diseases have clearly shown that each receptor has specifi c roles in both physiological and pathological conditions. For example, LPA has a pivotal role in neurogenesis ( 9 ) and has also been implicated in the development of lung fi brosis ( 10 ) via LPA 1 . LPA exhibits unique roles in implantation of fertilized eggs via LPA 3 ( 11 ) and hair follicle formation via LPA 6 ( 12 ). LPA is produced by at least two pathways where multiple phospholipase activities are involved ( 3 ). Lysophospholipase D/autotaxin/NPP2 produces LPA from LysoGPs such as LPC, while phosphatidic acid (PA)-selective PLA 1 ␣ (PA-PLA 1 ␣ ) and PA-PLA 1 ␤ produce LPA from PA by their PLA 1 activities. In contrast, S1P is produced intracellularly by phosphorylation Abstract It is now accepted that lysophospholipids (LysoGPs) have a wide variety of functions as lipid mediators that are exerted through G protein-coupled receptors (GPCRs) specifi c to each lysophospholipid. While the roles of some LysoGPs, such as lysophosphatidic acid and sphingosine 1-phosphate, have been thoroughly examined, little is known about the roles of several other LysoGPs, such as lysophosphatidylserine (LysoPS), lysophosphatidylthreonine, lysophosphatidylethanolamine, lysophosphatidylinositol (LPI), and lysophosphatidylglycerol. Recently, a GPCR was found for LPI (GPR55) and three GPCRs (GPR34/LPS 1 , P2Y10/ LPS 2 , and GPR174/LPS 3 ) were found for LysoPS. In this review, we focus on these newly identifi ed GPCRs and summarize the actions of LysoPS and LPI as lipid mediators.

Lysophosphatidic acid (LPA) is a bioactive lipid composed of a phosphate group, a glycerol backbone, and a single acyl chain that varies in length and saturation. LPA activates six class A G-protein-coupled receptors to provoke various cellular reactions. Because LPA signalling has been implicated in cancer and fibrosis, the LPA receptors are regarded as promising drug targets. The six LPA receptors are subdivided into the endothelial differentiation gene (EDG) family (LPA-LPA) and the phylogenetically distant non-EDG family (LPA-LPA). The structure of LPA has enhanced our understanding of the EDG family of LPA receptors. By contrast, the functional and pharmacological characteristics of the non-EDG family of LPA receptors have remained unknown, owing to the lack of structural information. Although the non-EDG LPA receptors share sequence similarity with the P2Y family of nucleotide receptors, the LPA recognition mechanism cannot be deduced from the P2Y and P2Y structures because of the large differences in the chemical structures of their ligands. Here we determine the 3.2 Å crystal structure of LPA, the gene deletion of which is responsible for congenital hair loss, to clarify the ligand recognition mechanism of the non-EDG family of LPA receptors. Notably, the ligand-binding pocket of LPA is laterally open towards the membrane, and the acyl chain of the lipid used for the crystallization is bound within this pocket, indicating the binding mode of the LPA acyl chain. Docking and mutagenesis analyses also indicated that the conserved positively charged residues within the central cavity recognize the phosphate head group of LPA by inducing an inward shift of transmembrane helices 6 and 7, suggesting that the receptor activation is triggered by this conformational rearrangement.

Lysophosphatidylserine (LysoPS) is an endogenous lipid mediator generated by hydrolysis of membrane phospholipid phosphatidylserine. Recent ligand screening of orphan G-protein-coupled receptors (GPCRs) identified two LysoPS-specific human GPCRs, namely, P2Y10 (LPS2) and GPR174 (LPS3), which, together with previously reported GPR34 (LPS1), comprise a LysoPS receptor family. Herein, we examined the structure-activity relationships of a series of synthetic LysoPS analogues toward these recently deorphanized LysoPS receptors, based on the idea that LysoPS can be regarded as consisting of distinct modules (fatty acid, glycerol, and l-serine) connected by phosphodiester and ester linkages. Starting from the endogenous ligand (1-oleoyl-LysoPS, 1), we optimized the structure of each module and the ester linkage. Accordingly, we identified some structural requirements of each module for potency and for receptor subtype selectivity. Further assembly of individually structure-optimized modules yielded a series of potent and LysoPS receptor subtype-selective agonists, particularly for P2Y10 and GPR174.

Lysophosphatidylserine (1-oleoyl-2 R-lysophosphatidylserine, LysoPS) has been shown to have lipid mediator-like actions such as stimulation of mast cell degranulation and suppression of T lymphocyte proliferation, although the mechanisms of LysoPS actions have been elusive. Recently, three G protein-coupled receptors (LPS1/GPR34, LPS2/P2Y10 and LPS3/GPR174) were found to react specifically with LysoPS, raising the possibility that LysoPS serves as a lipid mediator that exerts its role through these receptors. Previously, we chemically synthesized a number of LysoPS analogues and evaluated them as agonists for mast-cell degranulation. Here, we used a transforming growth factor-α (TGFα) shedding assay to see if these LysoPS analogues activated the three LysoPS receptors. Modification of the serine moiety significantly reduced the ability of the analogues to activate the three LysoPS receptors, whereas modification of other parts resulted in loss of activity in receptor-specific manner. We found that introduction of methyl group to serine moiety (1-oleoyl-lysophosphatidylallothreonine) and removal of sn-2 hydroxyl group (1-oleoyl-2-deoxy-LysoPS) resulted in reduction of reactivity with LPS1 and LPS3, respectively. Accordingly, we synthesized a LysoPS analogue with the two modifications (1-oleoyl-2-deoxy-lysophosphatidylallothreonine) and found it to be an LPS2-selective agonist. These pharmacological tools will definitely help to identify the biological roles of these LysoPS receptors.