Lysophosphatidic Acid Receptor
In 2003, a ligand screening assay for the purinergic GPCR, P2Y9, revealed that its agonist is LPA, not nucleotides, such as ATP (Noguchi et al. 2003). This result led investigators to pursue the identification of a ligand for the related receptor, GPR92, leading to the finding that its agonist is also LPA (Lee et al. 2006). Distinct human genetic studies and other in vitro studies further confirmed that P2Y5 is also the sixth LPA receptor (Pasternack et al. 2008; Yanagida et al. 2009). P2Y9, GPR92, and P2Y5 are structurally related to one another and form a large cluster in the phylogenetic tree, which is far distant from the cluster of LPAR1, LPAR2, and LPAR3 (Fukushima et al. 2015). According to these successive studies, P2Y9, GPR92, and P2Y5 were termed LPAR4, LPAR5, and LPAR6, respectively (Kihara et al. 2014).
The identification of six types of LPA receptors, together with discoveries of enzymes responsible for LPA production and degradation, has established LPA as an important lipid mediator that has essential roles in body development and also influences the pathophysiology of many diseases (Tsujiuchi et al. 2014; Yung et al. 2014). Thus, current studies are now expanding to translational and clinical trials aiming to develop treatments for several diseases.
The International Union of Basic and Clinical Pharmacology (IUPHAR) has updated the nomenclature for the six types of LPA receptors as LPA1–LPA6, which are encoded by the corresponding genes, Lpar1–Lpar6 in nonhumans and LPAR1–LPAR6 in humans (Kihara et al. 2014). Although some GPCRs are proposed to be LPA receptors, including GPR35 and GPR87, we will mention only Lpar1–Lpar6.
LPA1 couples to three types of heterotrimeric G proteins, including Gi, Gq, and G12/13 families, which link the signals to adenylyl cyclase inhibition and MAP kinase activation, phospholipase C activation, and Rho kinase stimulation, respectively (Fig. 1) (Ishii et al. 2004). LPA1-mediated cellular responses vary depending on cell types, and representative responses are cell proliferation, cell survival, intracellular Ca2+ mobilization, stress fiber formation, and neurite retraction. LPA1 also contains a unique sequence, called the PDZ (PSD95/Dlg1/ZO-1)-binding motif, consisting of SVV at the carboxyl terminal end. This motif of LPA1 interacts with PDZ proteins, including RhoGEF, to modulate LPA1-mediated intracellular signaling (Fukushima et al. 2015). This SVV sequence is highly conserved among vertebrates, suggesting the existence of common signaling pathways and biological roles in vertebrates (Fukushima et al. 2015).
LPA receptor-mediated functions revealed in mice lacking LPA receptor genes
Brain development, psychiatric function, pain transmission, fibrosis, vascular formation in concert with LPA2
Synaptic transmission in concert with PRG-1
Embryo implantation, spermatogenesis in concert with LPA1 and LPA2
Vasculogenesis, lymphogenesis, osteogenesis, lymphocyte circulation
No marked phenotype reported in the mice
Lpar2 was first found based on a homology search for Lpar1. Human LPAR2 encodes a 39 kDa protein consisting of 348 amino acids, which shares about 60% amino acid identity with human LPA1. The Lpar2 cluster is divided into two classes in phylogenetic analysis (Fukushima et al. 2015). One consists of a teleost group, which includes two subtypes of Lpar2, Lpar2a and Lpar2b, and the other consists of a non-teleost vertebrate group. Thus, Lpar2 might have acquired distinct roles specific to animal species during evolution. Computational and mutagenesis analyses of human LPA2 revealed that R3.28, Q3.29, R5.38, and K7.36 are critical for LPA2 activation (Valentine et al. 2008). These residues, except for Q3.29, are highly conserved in vertebrates (Fukushima et al. 2015).
Like LPA1, LPA2 also couples with Gi, Gq, and G12/13 families (Fig. 2) (Ishii et al. 2004). Therefore, LPA2-mediated cellular responses are thought to be the same as those of LPA1 and to be redundant with signaling through LPA1 under certain conditions (Yung et al. 2014). Moreover, like LPA1, human and mouse LPA2 contain the PDZ-binding motif consisting of STL at the carboxyl terminal end. This motif of LPA2 interacts with Na+-H+ exchanger regulatory factor 2 or RhoGEF, resulting in a change in intracellular signaling (Fukushima et al. 2015). However, unlike the SVV sequence in LPA1, the STL sequence in LPA2 is not highly conserved among vertebrates, and STL-associated signaling might be restricted in some animal species (Fukushima et al. 2015).
Lpar2 expression is mainly observed in the lung, kidney, testis, spleen, and thymus of adult mice as well as in the brain of embryonic and postnatal mice. Lpar2-deficient mice show no obvious phenotypic alterations. However, when these mice are crossed with other mice lacking Lpar1 or Prg-1, which encodes LPA phosphatase, their offspring display some defects in blood vessel formation or synaptic transmission, respectively (Table 1) (Trimbuch et al. 2009; Yung et al. 2014).
Like Lpar2, Lpar3 was also found based on a homology search for Lpar1. Human LPAR3 encodes a 40 kDa protein consisting of 353 amino acids, which shares about 50% amino acid identity with human LPA1 and LPA2. Mutagenesis studies of human LPA3 have identified that R3.28, Q3.29, W4.64, R5.38, and K7.35 are critical for LPA3 activation (Valentine et al. 2008). Moreover, R7.36 and K7.35 in the seventh transmembrane domain are important for LPA recognition by LPA3 and are conserved in vertebrates (Fukushima et al. 2015).
LPA3 interacts with Gi and Gq, but not G12/13 (Fig. 2) (Ishii et al. 2004). Thus, LPA3-dependent cellular responses are likely different from those of LPA1 and LPA2. Nonetheless, LPA3 is also able to induce stress fiber formation involving the pathway via Gq, but not G12/13 as observed for LPA1 or LPA2. Furthermore, the LPA3-Gi pathway is also involved in a distinct type of actin rearrangement, which we called actin patch-like structures (Fukushima et al. 2015). These findings suggest that LPA3 also exerts diverse cellular responses.
A unique feature of LPA3 is its ligand preference, which is different from that of LPA1 and LPA2. LPA3 prefers 2-acyl-LPA to 1-acyl-LPA, while, in the case of both LPA1 and LPA2, cellular responses (i.e., EC50 values) to 1-acyl-LPA and 2-acyl-LPA are similar (Fig. 2) (Yung et al. 2014). The biological and functional linkage of LPA3 and 2-acyl-LPA-producing enzyme remains unknown.
Lpar3 expression is enriched in the lung, kidney, testis, and small intestine in adult mice as well as in the brain of postnatal mice. Lpar3-deficient female mice show abnormal embryo implantation in the uterus (Table 1) (Yung et al. 2014). Moreover, triple deletion of Lpar1–Lpar3, but not single deletion of Lpar3, results in defects in germ cell survival in aged male mice (Yung et al. 2014). This result is consistent with the finding that Lpar1, Lpar2, and Lpar3 are highly expressed in the testis (Ishii et al. 2004).
The Lpar1–Lpar3 genes consist of multiple exons and introns, and exon-intron boundaries within the genes encode the sixth transmembrane domains (Fukushima et al. 2015). Such gene structures are highly conserved in vertebrates except for teleost Lpar2b, which additionally contains exon-intron boundaries within the regions encoding the second and third transmembrane domains. This suggests that after a duplication event in the ancestral teleost Lpar2, Lpar2b independently evolved, as suggested by the diverged phylogenetic tree for Lpar2 (Fukushima et al. 2015).
Lpar4 was originally isolated as an orphan purinergic GPCR, p2y9, belonging to the p2y family. A ligand screening assay led to the identification of P2Y9 as the fourth LPA receptor (Noguchi et al. 2003). Human LPAR4 encodes a 42 kDa protein consisting of 370 amino acids, which shares less than 25% amino acid identity with human LPA1, LPA2, and LPA3. There are no amino acid residues corresponding to R3.28, Q3.29, and K7.36 in LPA1 and LPA2 (Li et al. 2009). Instead, LPA4 activation involves three distinct residues, S3.28, a threonine in the second extracellular loop, and Y6.51.
LPA4 couples to four distinct G proteins, namely, Gs, Gi, Gq, and G12/13 (Fig. 2) (Yung et al. 2014). LPA4-dependent Gs, Gq, and G12/13 activation leads to cAMP accumulation, intracellular Ca2+ mobilization, and cytoskeletal changes, respectively. However, LPA4-Gi coupling seems to link to intracellular Ca2+ mobilization, but not to adenylyl cyclase inhibition. Although LPA4 activates diverse G protein signaling similar to LPA1 and LPA2, LPA4-dependent cellular responses regarding cell motility are opposite to LPA1- or LPA2-mediated responses. Thus, LPA4 may have an inhibitory or antagonistic role in cellular responses induced by other LPA receptors. This idea is supported by the finding that Lpar4-deficient cells are hypersensitive to LPA exposure in terms of cell motility (Yung et al. 2014).
Lpar4 expression is relatively restricted and observed in the heart, skin, and thymus of adult mice, as well as the brain of embryonic mice. Lpar4-deficient mice show no obvious phenotype but have an increased bone volume (Table 1) (Yung et al. 2014). Lpar4-deficient mice of different genetic backgrounds show prenatal lethality due to abnormal blood and lymphatic vessel formation (Yung et al. 2014). In addition, surviving Lpar4-deficient mice exhibit decreased transmigration of lymphocytes in lymph nodes (Hata et al. 2016). Together, studies of these mice indicate that Lpar4 is partly involved in osteogenesis, lymphogenesis, vasculogenesis, and lymphocyte circulation.
Human LPAR5 encodes a 41 kDa protein consisting of 372 amino acids, which shares about 34% amino acid identity with human LPA4. The Lpar5 cluster in the phylogenetic tree is more characteristic than that of Lpar4 or Lpar6 and is divided into two subclusters, a teleost group and a non-teleost group, as seen for the Lpar2 clusters (Fukushima et al. 2015). The teleost Lpar5 cluster includes Lpar5a and Lpar5b, suggesting duplication of teleost Lpar5 during evolution. Computational analysis of human LPA5 identified four amino acid residues critical for LPA recognition and LPA5 activation, R2.60, H4.64, R6.62, and R7.32 in LPA5 (Williams et al. 2009). However, these residues vary among animal species from teleosts to mammals, indicating that the ability of LPA5 to recognize LPA for activation is distinct among animal species (Fukushima et al. 2015). Indeed, teleost LPA5a and LPA5b, both of which lack H4.64 and R6.62, show no LPA response (Fukushima et al. 2015).
LPA5 couples to Gq and G12/13, resulting in intracellular Ca2+ mobilization and cytoskeletal changes, respectively (Fig. 2) (Yung et al. 2014). LPA5-dependent cAMP accumulation is mediated through non-Gs-dependent pathways. LPA5 is activated by 1-acyl-LPA but shows a tenfold preference for 1-alkyl-LPA compared with 1-acyl-LPA (Fig. 2) (Williams et al. 2009). Furthermore, farnesyl pyrophosphate and N-arachidonylglycine, which are structurally distinct from LPA, bind to and activate LPA5 (Fig. 1) (Oh et al. 2008).
Lpar5 is expressed in many mouse tissues, including the spleen, heart, platelets, dorsal root ganglion, and embryonic brains. LPAR5 is also abundantly expressed in human platelets and is primarily involved in LPA-induced platelet activation (Williams et al. 2009). Lpar5-deficient mice show loss of neuropathic pain, similar to that observed in Lpar1-deficient mice (Table 1). However, LPA1-mediated pathways involve protein kinase C and Ca channels, leading to demyelination, whereas LPA5-dependent signals include cAMP-dependent transcription, not leading to demyelination (Yung et al. 2014). Thus, Lpar1 and Lpar5 might be independently involved in the initiation and/or maintenance of neuropathic pain.
Human LPAR6, originally known as P2Y5, was discovered to be a gene responsible for hypotrichosis simplex with characteristic wooly hair or hair loss frequently found in the Middle East (Pasternack et al. 2008). Human LPAR6 encodes a 39 kDa protein consisting of 344 amino acids, which shares about 57% amino acid identity with human LPA4 and 36% amino acid identity with human LPA5. Sequence analysis of LPAR6 in families with hypotrichosis simplex identified several missense mutations that could affect LPA6 activation (Raza et al. 2014). These included S3T, D63V, G146R, I188F, E189K, D248Y, and L277P. Computational analysis revealed that these mutated LPA6 proteins showed a shift in orientation of 1-acyl-LPA at the binding site, which might result in alterations of LPA6 activation. However, the influences of these mutations on the binding of 2-acyl-LPA, a more preferred agonist of LPA6, remain unknown.
LPA6 transduces signaling through Gi and G12/13 (Fig. 2) (Yung et al. 2014). LPA6-Gi signaling includes intracellular Ca2+ mobilization and MAP kinase activation, and LPA6-G12/13 signaling leads to activation of the serum-responsive element and cytoskeletal changes. As mentioned above, LPA6 prefers 2-acyl-LPA to 1-acyl-LPA, which is reminiscent of the ligand preference of LPA3 (Fig. 2) (Yanagida et al. 2009). However, the biological correlation of 2-acyl-LPA production and LPA6 activation is better characterized in mice and humans (Yung et al. 2014). LPA6 co-localizes with lipase member H (LIPH), an enzyme producing 2-acyl-LPA from phosphatidic acid, in hair follicles for their cooperation in hair development in mice. In human, LIPH mutations as well as LPAR6 mutations are observed in hypotrichosis patients.
Lpar6 seems to be ubiquitously expressed and might have fundamental functions in the body. However, the only reported LPAR6-associated phenotype is abnormal hair development in human as mentioned above, and these patients show no other characteristic phenotypes (Pasternack et al. 2008). Moreover, recently generated Lpar6-deficient mice exhibit no marked phenotypic changes (Table 1) (Hata et al. 2016). There might be some redundancy in the utilization of LPA receptors.
Unlike Lpar1–Lpar3, the Lpar4–Lpar6 genes in mammals consist of a single exon (Fukushima et al. 2015). Such gene structures are conserved across vertebrates except for some teleosts, in which Lpar4 includes exon-intron boundaries within the regions encoding the fourth transmembrane domains.
As we discussed above, there is growing evidence that LPA signaling plays important and diverse roles not only in body development but also in the onset or progression of many diseases. Thus, LPA receptors could be a suitable target for the treatment of diseases. Indeed, many pharmaceutical companies are now developing various synthetic agonists and antagonists specific to LPA receptor subtypes, some of which are entering clinical trials for the treatment of diseases. The recent crystal structure study of human LPA1 and future structural analyses of other LPA receptors will accelerate structure-based drug design.
In addition to therapeutic-based studies, another interesting issue in LPA research is how LPA signaling has evolved and acquired biological significance. Genes for LPA receptors, S1P receptors, LPA-producing enzymes, and S1P-producing enzymes (sphingosine kinase) are found in the genomes of vertebrates, from human to lamprey (Petromyzon marinus). In genomes of chordatas, ancestral animals of vertebrates, only two GPCR genes, deposited as s1pr3 orthologs, and sphingosine kinase genes seem to be present. In fly and worm genomes, no lysophospholipid receptor genes are reported. Therefore, unveiling LPA or S1P signaling in lower vertebrates and chordatas may lead to better understanding of the fundamental roles of lysophospholipid signaling in vertebrate evolution.
This work was supported by KINDAI research funding KD02 (N.F. and T.T.) and KINDAI School of Science and Engineering funding RK-052 and RK-062 (N.F.). We apologize to all authors whose primary research papers could not be cited directly due to space limitations.
- Valentine WJ, Fells JI, Perygin DH, Mujahid S, Yokoyama K, Fujiwara Y, et al. Subtype-specific residues involved in ligand activation of the endothelial differentiation gene family lysophosphatidic acid receptors. J Biol Chem. 2008;283(18):12175–87. Epub 2008/03/05PubMedPubMedCentralCrossRefGoogle Scholar