Identification and Characterization of a Novel Lysophosphatidic Acid Receptor, p2y5/LPA6*

p2y5 is an orphan G protein-coupled receptor that is closely related to the fourth lysophosphatidic acid (LPA) receptor, LPA4. Here we report that p2y5 is a novel LPA receptor coupling to the G13-Rho signaling pathway. “LPA receptor-null” RH7777 and B103 cells exogenously expressing p2y5 showed [3H]LPA binding, LPA-induced [35S]guanosine 5′-3-O-(thio)triphosphate binding, Rho-dependent alternation of cellular morphology, and Gs/13 chimeric protein-mediated cAMP accumulation. LPA-induced contraction of human umbilical vein endothelial cells was suppressed by small interfering RNA knockdown of endogenously expressed p2y5. We also found that 2-acyl-LPA had higher activity to p2y5 than 1-acyl-LPA. A recent study has suggested that p2y5 is an LPA receptor essential for human hair growth. We confirmed that p2y5 is a functional LPA receptor and propose to designate this receptor LPA6.

The fact that p2y5 shares the highest sequence homology with p2y9/LPA 4 among all GPCRs (12) strongly suggested that LPA is a ligand for p2y5. However, we could not detect LPAinduced Ca 2ϩ mobilization or cAMP level changes in p2y5overexpressing cells at the time of the identification of p2y9/ LPA 4 as the fourth LPA receptor in our laboratory (12). In the course of the further analysis of p2y5-overexpressing cells, we found that p2y5 actually responded to LPA with activation of the G 13 -Rho signaling pathway. Our results confirm the identification of p2y5 as an LPA receptor and extend the knowledge of the functional roles of LPA.
Cell Culture-RH7777 rat hepatoma cells and B103 rat neuroblastoma cells were kindly provided by Dr. J. Chun (Scripps Research Institute, La Jolla, CA). These cells were maintained on collagen-coated 100-mm dishes (Iwaki, Tokyo, Japan) in Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% fetal bovine serum (Invitrogen). B103 cells and RH7777 cells stably expressing p2y5, LPA 1 , or LPA 3 (see below) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 0.3 mg/ml G418 (Invitrogen). Human umbilical vein endothelial cells (HUVECs) were purchased from Cell Applications (San Diego, CA) and cultured in collagen-coated 100-mm dishes with endothelial cell growth medium (Cell Applications). The cells were used for assays from passages 4 -7. In some experiments, cells were pretreated with 5 M Y27632 (Calbiochem; from a 5 mM stock in water stored at Ϫ30°C) for 10 min.
Cloning of Human p2y5-A DNA fragment containing the entire open reading frame of human p2y5 (GenBank TM accession number NM_005767) was first amplified from human genomic DNA by PCR using Pfu turbo (Stratagene, La Jolla, CA) and oligonucleotides (sense primer, 5Ј-AAAGTGCTTC-CAAACTGAAAATTG-3Ј; antisense primer, 5Ј-CCCAGTG-AGTCCTAATGGTTTTAT-3Ј). The entire open reading frame of p2y5 with an additional sequence of hemagglutinin (HA) epitope at the 5Ј-end was subsequently amplified from the resultant PCR products using KOD-Plus (Toyobo, Osaka, Japan) and oligonucleotides (sense primer containing KpnI and HA tag sequences, 5Ј-GGGGTACCGCCATGTACCCCTAC-GACGTGCCCGACTACGCCGTAAGCGTTAACAGCTCC-3Ј; antisense primer containing XbaI sequence, 5Ј-GCTCT-AGATCAGGCAGCAGATTCATT-3Ј). The resultant DNA fragment was digested with KpnI and XbaI and subsequently cloned into the mammalian expression vector pCXN2.1, a slightly modified version of pCXN2 (17) with multiple cloning sites, between the KpnI and NheI sites.
Cloning of Human LPA Receptors-HA-tagged human LPA 1 and LPA 4 DNA were constructed and cloned into pCXN2.1, as described previously (18). The entire open reading frame of human LPA 2 (GenBank TM accession number NM_004720) with an additional sequence of HA epitope at the 5Ј-end was amplified from LPA 2 cDNA in pcDNA3.1 (kindly provided by Dr. Izumi, Gunma University, Japan) using KOD-Plus and oligonucleotides (sense primer containing KpnI and HA tag sequences, 5Ј-GGGGTACCGCCATGTACCCCTACGA-CGTGCCCGACTACGCCGTCATCATGGGCCAGTGC-3Ј; antisense primer containing XbaI sequence, 5Ј-GCTCTA-GACTAAAGGGTGGAGTCCATCAG-3Ј). A DNA fragment containing the entire open reading frame of human LPA 3 (GenBank TM accession number NM_012152) was first amplified from a cDNA prepared from human kidney poly(A) ϩ RNA (Clontech, Palo Alto, CA) by PCR using Pfu turbo and oligonucleotides (sense primer, 5Ј-TCTAGGATGTTCACTTCTTC-TCCAC-3Ј; antisense primer, 5Ј-GAGAGGCATCCAGAGT-TTAGGAAGT-3Ј). The entire open reading frame of LPA 3 with an additional sequence of HA epitope at the 5Ј-end was subsequently amplified from the resultant PCR products using KOD-Plus and oligonucleotides (sense primer containing KpnI and HA tag sequences, 5Ј-GGGGTACCGCCATGTACCC-CTACGACGTGCCCGACTACGCCAATGAGTGTCACTA-TGAC-3Ј; antisense primer containing XbaI sequence, 5Ј-GCTCTAGATTAGGAAGTGCTTTTATTGCA-3Ј). A DNA fragment containing the entire open reading frame of human LPA 5 (GenBank TM accession number NM_020400) was first  amplified from human genomic DNA by PCR using Pfu turbo  and oligonucleotides (sense primer, 5Ј-ACTTGAGGCTGAG-AAAAGATTCAG-3Ј; antisense primer, 5Ј-ACTTTGTACTC-TTCTGCGTTGCTAA-3Ј). The entire open reading frame of LPA 5 with an additional sequence of HA epitope at the 5Ј-end was subsequently amplified from the resultant PCR products using KOD-Plus and oligonucleotides (sense primer containing KpnI and HA tag sequences, 5Ј-GGGGTACCGCCATGTAC-CCCTACGACGTGCCCGACTACGCCTTAGCCAACAGC-TCCTCA-3Ј; antisense primer containing XbaI sequence, 5Ј-GCTCTAGATCAGAGGGCGGAATCCTG-3Ј). All resultant DNA fragments were digested with KpnI and XbaI and subsequently cloned into pCXN2.1.
Transient Expression and Membrane Preparation-RH7777 cells (2 ϫ 10 6 ) were seeded in collagen-coated 100-mm dishes and transfected with each receptor expression plasmid using the Lipofectamine 2000 reagent (Invitrogen). After 24-h serum starvation with Dulbecco's modified Eagle's medium containing 0.1% bovine serum albumin (BSA; fatty acid-free; Sigma), the cells were washed with phosphate-buffered saline twice and scraped off. After further washing with binding buffer (25 mM HEPES-NaOH (pH 7.4), 10 mM MgCl 2 , and 0.25 M sucrose), the cells were suspended in the buffer containing an additional protease inhibitor mixture (Complete; Roche Applied Science), sonicated three times at 15 watts for 30 s, and centrifuged at 8,000 ϫ g for 10 min at 4°C. The supernatant was further centrifuged at 100,000 ϫ g for 60 min at 4°C, and the resultant pellet was homogenized in ice-cold binding buffer. The protein concentration of the homogenate was determined with a Bradford assay (Bio-Rad) using BSA as a standard.
Western Blotting-The membrane fraction (4 g of protein) from RH7777 cells was digested with 2 units of N-glycosidase F (Roche Applied Science) to remove asparagine-bound N-glycans according to the manufacturer's instructions. The protein sample (with 5% 2-mercaptoethanol) was analyzed by 10% SDS-PAGE without heat denaturation, followed by transfer to a polyvinylidene difluoride membrane (Millipore Corp., Bedford, MA). The membrane was blocked with 5% skim milk (Difco) and probed with a 3F10 rat monoclonal anti-HA antibody (Roche Applied Science). The bands were visualized with an ECL chemiluminescence detection system (GE Healthcare) using horseradish peroxidase-conjugated anti-rat IgG (Santa Cruz Biotechnology, Inc., Santa Cruz, CA).
Radioligand Binding Assay-A total of 20 g each of the membrane fractions was incubated in 200 l of binding buffer containing 0.25% BSA with the indicated concentra- Stable Expression-B103 cells and RH7777 cells stably expressing p2y5 (referred to hereafter as B103-p2y5 cells and RH7777-p2y5 cells, respectively) were established as follows. B103 cells and RH7777 cells were transfected with the p2y5expression plasmid (see above). Expression of the HA epitope on the cell surface was confirmed by flow cytometric analysis (EPICS XL, Beckman Coulter, Fullerton, CA) using a 3F10 anti-HA antibody and phycoerythrin-labeled anti-rat IgG (Beckman Coulter) as the secondary antibody. Stable transfectants were selected with 1 mg/ml G418 for 20 days. After labeling the drug-resistant cells with phycoerythrin, as described above, a group of HA-positive cells was sorted using an automated magnetic cell sorter (autoMACS; Miltenyi Biotec, Auburn, CA) with the use of anti-phycoerythrin MicroBeads (Miltenyi Biotec). B103 cells and RH7777 cells stably expressing LPA 1 or LPA 3 were established essentially as described previously (18). was separated from free by rapid filtration through a Unifilter-96-GF/C, which was rinsed 10 times with ice-cold TMN buffer (10 mM Tris-HCl (pH 7.5), 25 mM MgCl 2 , and 100 mM NaCl). Radioactivity was measured as described above.
[ 35 S]GTP␥S Binding to the G␣ 13 Protein-The membrane fraction (40 g of protein) from RH7777 cells transiently expressing p2y5 was incubated in 100 l of GTP␥S binding buffer containing 0.5 nM [ 35 S]GTP␥S with or without 10 M 1-oleoyl-LPA for 30 min at 30°C. Then membrane protein was resuspended in 900 l of solubilization buffer (20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 5 mM MgCl 2 , 1 mM EDTA, 1.25% Nonidet P-40, 0.2% SDS, 100 M GDP, 100 M GTP, and Complete). 50 l of protein A/G PLUS-agarose beads (Santa Cruz Biotechnology) were added to the extracts and mixed by rotation for 3 h at 4°C to preclear nonspecific binding. Agarose beads were then pelletted, and the supernatant was incubated with 1:100 anti-G␣ 13 antibody (A-20; Santa Cruz Biotechnology) for 60 min at 4°C, and then 50 l of protein A/G PLUS-agarose was added. Following rotation for 12-16 h at 4°C, the immune complexes were collected by centrifugation and washed three times with wash buffer (20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 5 mM MgCl 2 , 1 mM EDTA, 100 M GDP, and 100 M GTP). [ 35 S]GTP␥S binding in the immunoprecipitates was quantified by liquid scintillation counting. Nonspecific binding was determined by the immunoprecipitation without antibody.
Construction of G s/13 Chimeric Protein Expression Plasmid-Human G s -encoding cDNA in pcDNA3.1 (G␣ s long subunit; cDNA Resource Center, Rolla, MO) was used as a template to replace its C-terminal five amino acids (QYELL) with those of G 13 (QLMLQ) (19) using PCR-based mutagenesis with two oli-gonucleotides (sense primer, 5Ј-ATTAATACGACTCACTA-TAGG-3Ј; antisense primer, 5Ј-GGTCTAGATTACTGT-AGCATAAGCTGACGAAGGTGCATGCGCTGAA-3Ј). The resultant DNA fragment was digested with KpnI and XbaI and subsequently cloned into the pcDNA3.1 vector. The NheI-XbaI DNA fragment (containing the G s/13 sequence) in the resultant plasmid replaced the Renilla luciferase sequence of the pRL-CMV vector (Promega, Madison, WI). Intact pRL-CMV vector served as the control for the G s/13 expression vector.
cAMP Measurement-To determine whether LPA receptors affect the activity of adenylyl cyclases, an AlphaScreen cAMP assay kit (PerkinElmer Life Sciences) was used as recommended by the manufacturer. Cells (4 ϫ 10 4 ) were seeded in collagencoated 96-well plates (Iwaki), followed by a 12-h serum starvation in the presence or absence of 100 ng/ml pertussis toxin (PTX; List Biological Laboratories, Campbell, CA; from a 400 g/ml stock in 10 mM Tris-HCl (pH 7.4) and 2 M urea stored at 4°C). The cells were washed with buffer A (Hanks' balanced salt solution containing 25 mM HEPES-NaOH (pH 7.4) and 0.1% BSA) and incubated in 100 l of buffer A containing 0.5 mM 3-isobutyl-1-methylxanthine (Sigma; from a 50 mM stock in dimethyl sulfoxide stored at Ϫ30°C) for 20 min at room temperature. The reaction was initiated by adding 50 l of various concentrations of LPA in buffer A with 10 M forskolin (Sigma; from a 10 mM stock in dimethyl sulfoxide stored at Ϫ30°C). After 30 min of incubation at room temperature, the reaction was terminated by adding 15 l of 10% Tween 20, followed by overnight storage at 4°C. The cAMP concentration in the cell lysate was measured in triplicate with the Fusion system (PerkinElmer Life Sciences). In some experiments, the cells were pretreated with 10 M Ki16425 (Sigma; from a 10 mM stock in dimethyl sulfoxide stored at Ϫ30°C) for 20 min at room temperature.
mRNA Expression Profile of LPA Receptors in HUVECs-Total RNA was extracted from HUVECs using an RNeasy minikit (Qiagen, Tokyo, Japan) and reverse-transcribed using Superscript II reverse transcriptase (Invitrogen). An equal aliquot of the cDNA solution was used to amplify the different receptor transcripts using ExTaq DNA polymerase (Takara, Tokyo, Japan). The PCR program was as follows: denaturation at 94°C for 2.5 min and 40 cycles of amplification consisting of denaturation at 96°C for 10 s, annealing at 57°C for 10 s, and extension at 72°C for 20 s. The sequences of primers used for LPA 1 , LPA 2 , LPA 3 , LPA 4 , and LPA 5 are the same as those used by Kotarsky et al. (13). The primers for p2y5 were designed to amplify a 358-bp fragment (sense primer, 5Ј-AGAATTGTGA-GAAAGCGACCTC-3Ј; antisense primer, 5Ј-TCTGTGACC-AGAATGAAACCAC-3Ј). The PCR products were separated by electrophoresis on a 2% agarose gel.
LPA-induced Morphological Changes of HUVECs-To evaluate LPA-induced cell morphological changes of HUVECs, we utilized the xCELLigence system (Roche Applied Science), which measures cell-substrate impedance (20). Briefly, HUVECs were seeded at 8,000 cells/well in an E-plate (Roche Applied Science) coated with poly-D-lysine, followed by a 12-h culture. The cells were serum-starved with 100 l of endothelial cell basal medium (Cell Applications) containing 0.1% BSA for 4 h and stimulated with 5 M of 1-oleoyl-LPA (by adding 5 l of 100 M stock in phosphate-buffered saline containing 3% BSA). Phosphate-buffered saline containing 3% BSA served as the control for LPA stimulation. The cell index was continuously monitored every 15 s for the duration of the experiments. For the experiments with siRNAs, the cells (60,000 cells/well) cultured in collagen-coated 12-well plates (Iwaki) were transfected with 5 nM human p2y5 siRNAs (Silencer Select siRNA, ID numbers s19796 and s19798; Ambion, Austin, TX) and control siRNA (Silencer Select negative control 1; Ambion) using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. After a 24-h incubation, the cells were reseeded at 8,000 cells/well in a poly-D-lysine-coated E-plate and used for the assay as described above.
Quantitative Real Time PCR-In the siRNA experiments, the mRNA levels of LPA 1 , p2y5, and ␤-actin were quantified using the LightCycler system (Roche Applied Science). The PCRs were set up in microcapillary tubes in a volume of 20 l, consisting of 2 l of cDNA solution, 1ϫ FastStart DNA Master SYBR Green I (Roche Applied Science), and 0.5 M each sense and antisense primers. The PCR program was as follows: denaturation at 95°C for 3 min and 45 cycles of amplification consisting of denaturation at 95°C for 15 s, annealing at 65°C for 5 s, and extension at 72°C for 7 s. The primers for LPA 1 (sense primer, 5Ј-GGCTATGTTCGCCAGAGGACTAT-3Ј; antisense primer, 5Ј-TCCAGGAGTCCAGCAGATGATAA-3Ј) were designed to amplify a 135-bp fragment. The primers for p2y5 (sense primer, 5Ј-GGTAAGCGTTAACAGCTCCCACT-3Ј; antisense primer, 5Ј-TTTGAGGACGCAGATGAAAATGT-3Ј) were designed to amplify a 139-bp fragment. The primers for ␤-actin (sense primer, 5Ј-CAGGATGCAGAAGGAGATC-ACTG-3Ј; antisense primer, 5Ј-TACTCCTGCTTGCTGATC-CACAT-3Ј) were designed to amplify a 153-bp fragment. The cDNA level of each sample was quantified using the second derivative maximum method in LightCycler analysis software.
Statistical Analysis-To determine statistical significance, the values were compared by unpaired two-tailed t test, analysis of variance followed by Tukey-Kramer test, or two-way analysis of variance using Prism 4 software (GraphPad Software, San Diego, CA). The differences were considered significant if p values were less than 0.05.

Cell Aggregation of B103-p2y5 Cells in Serum-containing
Medium-An orphan receptor p2y5 shares the highest amino acid sequence homology with LPA 4 among all GPCRs (12), which led us to hypothesize that p2y5 is an additional LPA receptor. At first, we tested this hypothesis by detecting Ca 2ϩ mobilization or the changes of intracellular cAMP level using RH7777 cells and B103 cells stably overexpressing p2y5 (RH7777-p2y5 cells and B103-p2y5 cells, respectively) These cell lines were selected because they lack endogenous responses to LPA in these assays (18,21). Although flow cytometry analysis proved the high expression levels of p2y5 in these cells (Fig.  1A), we could not observe any p2y5-dependent signaling by 1-oleoyl-LPA (data not shown). Reporter gene assays detected no p2y5-dependent activation of the zif268 or the CRE promoter (data not shown).
In the course of the study of p2y5, we noticed that B103-p2y5 cells formed aggregates in serum-containing medium (Fig. 1B). We previously showed that B103 cells expressing LPA 4 formed aggregates through Rho in serum-containing medium, which is abundant in LPA (18). Thus, the similar cell morphology observed in B103-p2y5 cells implied the activation of the LPA-p2y5-Rho signaling axis in these cells.
Neurite Retraction in B103-p2y5 Cells and Membrane Blebbing in RH7777-p2y5 Cells by LPA-Since Rho activation in B103 cells results in neurite retraction (14,18,21,22), we examined whether LPA can induce neurite retraction in B103-p2y5 cells. Indeed, 1-oleoyl-LPA at a concentration of 1 M induced rapid neurite retraction in B103-p2y5 cells but not in B103vector cells (Fig. 2A, left and middle). Other related lysophospholipids with glycerol, choline, serine, inositol, or ethanolamine as headgroup (at 10 M) did not induce neurite retraction at all (data not shown). This morphological change in B103-p2y5 cells was Rho-associated kinase-dependent, because the LPA effect was inhibited by the pretreatment with the Rhoassociated kinase inhibitor Y27632 ( Fig. 2A, right). To confirm the Rho-activating property of p2y5 in another cell line, RH7777-p2y5 cells were stimulated with 1 M LPA. LPA induced membrane blebbing in RH7777-p2y5 cells much more potently than in RH7777-vector cells (Fig. 2B, left and middle). It has been reported that membrane blebbing is regulated by Rho (23)(24)(25). Consistently, the membrane blebbing in RH7777-p2y5 cells was severely attenuated by Y27632 (Fig. 2B, right).
[ 3 H]LPA Binding to RH7777-p2y5 Cell Membrane-Next, we performed the LPA binding assay using membrane fractions from RH7777 cells transiently expressing p2y5 or each of the known five LPA receptors (LPA 1 -LPA 5 ). Protein expression of these receptors was confirmed by Western blotting analysis (Fig. 3A, top). Under our experimental conditions, we could not detect any specific binding of [ 3 H]1-oleoyl-LPA to transiently expressed p2y5 or LPA 3 (Fig. 3A,  bottom). Other four LPA receptors displayed specific binding to [ 3 H]1-oleoyl-LPA (Fig. 3, A (bottom) and B). Functional LPA 3 receptors were actually expressed at the cell surface, because treatment with LPA induced Ca 2ϩ mobilization in the cells expressing LPA 3 (data not shown). The failure to detect the specific binding to p2y5 and LPA 3 might be due to the low affinities or the rapid off rates of LPA (see "Discussion").
In contrast to the transient expression system, small but significant specific binding was successfully detected when we used membranes from RH7777 cells stably expressing p2y5 (Fig. 3C). The specific binding did not reach saturation even at 200 nM [ 3 H]LPA (data not shown), suggesting that the affinity of 1-oleoyl-LPA to the receptor is too low for Scatchard analysis under our experimental conditions. The pretreatment with 100 M GTP␥S severely decreased the [ 3 H]LPA binding of p2y5-expressing membranes (Fig. 3D), suggesting that the specific binding to [ 3 H]LPA involves GPCRs.
LPA-induced GTP␥S Binding to Cell Membrane via p2y5-To demonstrate that LPA signals directly through p2y5, we performed a [ 35 S]GTP␥S binding assay using membrane fractions from RH7777 cells expressing p2y5. After the optimization of GDP concentration (0.3 M), we found that 1-oleoyl-LPA increased [ 35 S]GTP␥S binding in a dose-dependent manner both in transient and stable p2y5-expressing cells (Fig. 4A).
As described earlier, p2y5 induced striking morphological changes (i.e. neurite retraction and membrane blebbing) in a Rho-associated kinase-dependent manner (Fig. 2, A  and B). This strongly suggested that p2y5 coupled to G 12/13 protein. To verify the G 13 coupling of p2y5, we examined [ 35 S]GTP␥S incorporation into G␣ 13 protein by immunoprecipitation with anti-G␣ 13 protein antibody. With membranes from RH7777 cells transiently expressing p2y5, 10 M 1-oleoyl-LPA induced an ϳ4-fold increase in [ 35 S]GTP␥S binding to G 13 proteins, whereas LPA had no effect on vector transfectants (Fig. 4B).

Adenylyl Cyclase Activation by p2y5 through a G s/13 Chimeric
Protein-For further confirmation of G 12/13 proteins coupling of p2y5, we applied G s/13 , a chimeric ␣ subunit of G s and G 13 proteins that enables G 13 -coupling GPCRs to activate adenylyl cyclases (19). 1-Oleoyl-LPA again failed to induce any significant change in the intercellular cAMP level in B103-p2y5 cells with or without transfection of a G s/13 expression vector. However, when the cells transfected with G s/13 were treated with PTX, LPA significantly increased the cAMP level in a dose-dependent manner (Fig. 5). In B103-vector cells, there was no significant change in cAMP levels even after the transfection of G s/13 and treatment with PTX (Fig. 5).
Recently, p2y5 was identified as a causative gene for human hair growth deficiency (29 -33). Of note, the 2-acyl-LPA-producing enzyme, membrane-bound PA-selective phospholipase A 1 ␣ (mPA-PLA 1 ␣ or LIPH), was also encoded by a gene responsible for a similar hair disease (34). Thus, it was strongly suggested that the endogenous and authentic ligand for p2y5 is 2-acyl-LPA. We next examined whether p2y5 shows a preference for 2-acyl-LPA over 1-acyl-LPA. Commercially unavailable 2-acyl-LPA species were prepared using the R. miehei lipase (15) and various PA species as described under "Experimental Procedures." When 1-palmitoyl-2-oleoyl PA was digested with this enzyme, mass spectrometry analysis demonstrated that the products mainly consisted of oleoyl-LPA, whereas a very small amount of palmitoyl-LPA was produced (data not shown), indicating the selective digestion at the sn-1position of PA. Next, 2-acyl-LPA species from dioleoyl-PA, dilinoleoyl-PA, and diarachidonoyl-PA were prepared. Contrary to LPA 1 , LPA 3 showed a preference for 2-acyl-LPA (35). We consistently observed that the 2-oleoyl-LPA prepared by our method showed higher activity than 1-oleoyl-LPA in the Ca 2ϩ mobilization of RH7777 cells stably expressing LPA 3 but not in the cells expressing LPA 1 (Fig. S2). cAMP assays gave consistent results (Fig. 6C, left and middle), thus we examined the effect of the position of fatty acid chain on p2y5 activation in the cAMP assay with G s/13 -transfected B103 cells after PTX treatment. As expected, 2-oleoyl-LPA showed higher activity in forming cAMP via G s/13 than 1-oleoyl-LPA (Fig. 6C, right). LPA species containing linoleic acid and arachidonic acid at the sn-2-position also showed higher activity than those at the sn-1 position (Fig. 6D). Like 1-acyl LPA species, 2-linoleoyl-LPA was more potent than 2-oleoyl-LPA, whereas 2-arachidonoyl-LPA was less potent than 2-oleoyl-LPA (data not shown). The preference of p2y5 for 2-acyl-LPA was consistently observed in the [ 35 S]GTP␥S binding assay using membrane fractions from RH7777-p2y5 cells (Fig. 6E). Taken together, these results suggest that 2-acyl-LPA is a potent ligand for p2y5.
LPA-induced Cell Contraction via Endogenous p2y5 in HUVECs-Finally, we examined the function of the endogenously expressed p2y5. For this purpose, we used HUVECs, which are shown to express p2y5 mRNA at a high level by a comprehensive data base of gene expression profiles (available on the World Wide Web). We also confirmed the high  expression of p2y5 mRNA in HUVECs by reverse transcription-PCR (Fig. 7A). Consistent with previous reports (36,37), LPA 1 mRNA was also highly expressed in these cells. No mRNA expression of LPA 2 , LPA 3 , LPA 4 , and LPA 5 was detected under our experimental conditions. It has been reported that HUVECs show Rho-dependent actin reorganization in response to LPA (38). When these cells were seeded on poly-D-lysine-coated glass bottom dishes and serumstarved, 1-oleoyl-LPA induced rapid contraction and rounding of these cells (Fig. 7B). Because similar morphological changes were observed in LPA-stimulated B103-p2y5 cells ( Fig. 2A), it was probable that the contraction of HUVECs was modulated by p2y5.
For the objective evaluation and quantification of the morphological changes, we utilized the xCELLigence system from Roche Applied Science, which detects the cell-substrate impedance (20). For example, a decrease in cell adhesion or cell contraction, i.e., loss of the cell-electrode contact area, leads to a decrease in the impedance (displayed as "cell index"). Consistent with the rapid contraction and rounding of HUVECs as shown in Fig. 7B, a decrease of cell index was observed upon LPA-stimulation (Fig. 7C) in a dose-dependent manner (data not shown). We next examined the role of p2y5 in this effect by utilizing siRNA-mediated gene silencing. The selective and effective suppression of p2y5 gene expression by siRNAs was confirmed by quantitative real time PCR (Fig. 7D). The cells treated with p2y5 siRNA showed severe defects in the LPA-induced cell index decrease compared with control siRNA-treated cells (Fig.  7E, left). Another p2y5-siRNA with a different target sequence gave consistent results (Fig. S3). On the other hand, thrombin decreased the cell index equally in control siRNA-treated and p2y5 siRNA-treated cells (Fig. 7E,  right). These results demonstrate that not only overexpressed p2y5 but also endogenously expressed p2y5 can function as an LPA receptor that changes cell morphologies.

DISCUSSION
In the current study, we have identified LPA as a ligand for p2y5, which shares the highest sequence homology with LPA 4 (12). The morphological characteristics of p2y5overexpressing cells led us to hypothesize that p2y5 is an LPA receptor coupling with G 12/13 proteins. This hypothesis was validated by the following experiments: [ 3 H]LPA binding assay, [ 35 S]GTP␥S binding assay, observation of cellular morphology, cAMP assay with G s/13 proteins, and gene knockdown study using HUVECs. Furthermore, we have shown that 2-acyl-LPA has higher potency to p2y5 than 1-acyl-LPA.
Ligand binding has been recognized to be most important for the identification of receptor ligands (39). We could not detect specific binding of [ 3 H]1-oleoyl-LPA to RH7777 cells transiently expressing p2y5. However, RH7777 cells stably expressing p2y5 showed small but significant specific binding (Fig. 3C). Considering the sensitivity to GTP␥S (Fig. 3D), the specific binding of LPA can be ascribed to p2y5. The reasons for the discrepancy between the transient and stable expression systems are currently unknown. With the use of the HA epitope and a magnetic cell sorting system, the expression level of p2y5 was found to be extremely high in the stable cell line (Fig. 1A), which might enable us to detect the low specific [ 3 H]1-oleoyl-LPA binding. Indeed, the specific binding of p2y5 was lower compared with other LPA receptors (i.e. LPA 1 , LPA 2 , LPA 4 , and LPA 5 ). Even in the stable expression system, we failed to detect the saturation of the binding at 200 nM [ 3 H]1-oleoyl-LPA (data not shown). In the [ 35 S]GTP␥S binding assay and cAMP assay, the EC 50 values of 1-oleoyl-LPA were about 1 M, which was consistent with the specific binding too low to detect the saturation of the binding under our experimental conditions. Thus, radiolabeled ligands with much higher affinities to p2y5 than 1-oleoyl-LPA will be required to determine an accurate K d value for this receptor. Another possible reason for the poor detection of the specific binding might be the rapid offrate of the receptor. In our assay system, filter-trapped membrane proteins were washed 10 times with a buffer. The receptor-bound ligand might be released from the receptors if the off-rate is much faster than the time spent in washing. We note that an assay with three-time washing resulted in nonspecific binding too high to allow detection of the specific binding (data not shown). The receptor purification may overcome the problem of high nonspecific binding, as recently reported in the binding assay for free fatty acid receptor (40). Alternatively, a binding assay without washing steps, such as the binding assay with scintillation proximity assay technology and surface plasmon resonance biosensor (41), may be needed for more reliable LPA binding assays. The [ 35 S]GTP␥S binding assay detects direct receptor-G protein interactions (42) and is another important assay for the demonstration of the direct receptor activation by ligands. 1-Oleoyl-LPA increased the incorporation of [ 35 S]GTP␥S into the membrane fractions both from the cells transiently and stably expressing p2y5 in a dose-dependent manner (Fig. 4A). Although the signal-to-noise ratio is relatively low in the standard [ 35 S]GTP␥S binding assay (about a 1.3-fold increase), we could detect an ϳ4-fold increase of [ 35 S]GTP␥S incorporation to G␣ 13 protein (Fig. 4B) by another binding assay employing immunoprecipitation steps (43). Taken together, these results demonstrated that p2y5 is directly activated by LPA.
In the first course of our screening of lipid ligands for p2y5 by the Ca 2ϩ assay and cAMP assay, no LPA-dependent signaling of p2y5 was observed in "LPA receptor-null" RH7777 and B103 cells (data not shown). Besides this, neither the zif268 nor the CRE promoter was activated by LPA in reporter gene assays using p2y5-expressing PC12h rat pheochromocytoma cells (data not shown). Pasternack et al. (30) showed LPA-induced CRE-directed luciferase activation in Chinese hamster ovary cells expressing p2y5. This inconsistency might be due to the fact that different cells, Chinese hamster ovary cells, were used. Under our experimental conditions, LPA strongly enhanced luciferase activity in PC12h cells transiently expressing LPA 4 in the CRE-driven reporter gene assay (data not shown). Thus, at least the potency of p2y5 to couple with G s protein might be much lower than that of LPA 4 in this cell line. G q family proteins did not seem to be involved in p2y5 signaling, because B103 cells and RH7777 cells overexpressing p2y5 did not show any Ca 2ϩ mobilization in response to LPA. In accordance with this observation, in Chinese hamster ovary cells that have endogenous LPA receptors, p2y5-overexpression had no effect on the Ca 2ϩ signaling by LPA (12). Even in the presence of the promiscuous G protein, G 16 (44), p2y5-dependent Ca 2ϩ signaling was not observed in RH7777 cells (data not shown).  (Fig. S3).
In contrast to the Ca 2ϩ assay and cAMP assay, LPA induced drastic responses in the morphology of B103 cells and RH7777 cells overexpressing p2y5 (Fig. 2, A and B). Given the susceptibility to Y27632, the morphological changes were Rho-associated kinase-dependent. Furthermore, because the morphological changes were resistant to PTX or the G q/11 inhibitor YM254890 (data not shown), G 12/13 protein seemed to be involved, as previously reported in other LPA receptors (14,18,45). The G 12/13 protein coupling of p2y5 was consistently demonstrated not only by G 13 -specific [ 35 S]GTP␥S binding assay (Fig. 4B) (see above) but also by the increase in intracellular cAMP levels through a G s/13 chimeric G protein (Fig. 5). However, the cAMP elevation was observed only after pretreatment with PTX. Pasternack et al. (30) also showed a similar enhancing effect of PTX in reporter gene assays of p2y5. PTX-sensitive G proteins (possibly G i/o family G proteins) might modulate the signaling of p2y5, although the precise mechanism is unclear and awaits further signaling analysis of this receptor. We also demonstrated that LPA regulates cell morphology via endogenous p2y5 using siRNAs ( Fig. 7 and Fig. S3). It has been reported that LPA induces Rho-dependent morphological changes in endothelial cells, which are supposed to regulate endothelial permeability (46,47). Taken together, our present data of overexpression and knockdown experiments indicate that the G 12/13 -Rho axis is involved in p2y5 signaling.
Autotaxin is an LPA-producing enzyme responsible for the generation of LPA in plasma and serum (48,49). Autotaxin knock-out mice show embryonic lethality mainly because of severe blood vessel malformation (50,51). Of note, G 13 knock-out (52) and G 12/13 double knock-out mice (53) had similar phenotypes, which can be rescued by re-expression of G 13 protein in the endothelial cells (54). These facts suggest that the LPA-G 13 -Rho signaling pathway in endothelial cells is deeply involved in vascular development. However, the responsible LPA receptor(s) has not yet been identified (49). Since p2y5 played a main role in the Rhoinvolved morphological changes induced by LPA in HUVECs ( Fig. 7 and Fig. S3), p2y5 might be the missing link between the functions of autotaxin and G 13 . mPA-PLA 1 ␣ is another enzyme that produces LPA (55). Recent studies showed that human hair growth deficiencies are linked to genetic defects in mPA-PLA 1 ␣ (34, 56 -59) and p2y5 (29 -33). Both p2y5 and mPA-PLA 1 ␣ are highly expressed in the inner root sheaths of hair follicles and are supposed to participate in the development of hair follicles (30,33,34). In this study, we showed that 2-acyl-LPA activated p2y5 more efficiently than 1-acyl-LPA (Fig. 6). Since mPA-PLA 1 ␣ produces 2-acyl-LPA by digesting PA (55), our results are consistent with the similar phenotypes due to defects of mPA-PLA 1 ␣ and p2y5 (48). Unfortunately, no detection method for discrimination between 1-acyl-and 2-acyl-LPA has been established (60). However, it is acceptable to assume that LPA prepared by our method was rich in 2-acyl-LPA for the following reasons. First, we used R. miehei lipase that has 1,3-selective lipase activities (15) to avoid the digestion at the sn-2-position of PA. Actually, when 1-palmitoyl-2-oleoyl PA was digested with the lipase, mass spectrometry analysis demonstrated that the products mainly consisted of oleoyl-LPA (data not shown). Second, to reduce the effect of rapid acyl migration, we used the fraction within 2 h after the column elution. Finally, using the products, we reconfirmed the preference of LPA 3 for 2-oleoyl-LPA over 1-oleoyl-LPA ( Fig. 6C and Fig. S2), which was reported by Bandoh et al. (35). Further analysis, such as molecular species profiling of LPA in the inner root sheaths of hair follicles, will be needed to reveal the role of the mPA-PLA 1 ␣-p2y5 axis proposed in human hair growth.
In conclusion, we report here the identification of p2y5 as a novel sixth LPA receptor (LPA 6 ), which activates the G 12/13 -Rho signaling pathways. 2-Acyl-LPA had higher activity to p2y5 than 1-acyl-LPA, suggesting the functional coupling of mPA-PLA 1 ␣ with p2y5. We also showed the endogenous p2y5 function that affects the morphology of HUVECs. Our present study will help to better understand the physiological and pathological roles of p2y5 as well as detail mechanisms of human hair growth.