Sphingosine 1-Phosphate-induced Cell Rounding and Neurite Retraction Are Mediated by the G Protein-coupled Receptor H218*

Sphingosine 1-phosphate (SPP) is a lipid second messenger that also acts as a first messenger through the G protein-coupled receptor Edg-1. Here we show that SPP also binds to the related receptors H218 and Edg-3 with high affinity and specificity. SPP and sphinganine 1-phosphate bind to these receptors, whereas neither sphingosylphosphorylcholine nor lysophosphatidic acid compete with SPP for binding to either receptor. Transfection of HEK293 cells with H218 or edg-3, but not edg-1, induces rounded cell morphology in the presence of serum, which contains high levels of SPP. SPP treatment of cells overexpressing H218 cultured in delipidated serum causes cell rounding. A similar but less dramatic effect was observed in cells overexpressing Edg-3 but not with Edg-1. Cell rounding was correlated with apoptotic cell death, probably as a result of loss of attachment. Nerve growth factor-induced neuritogenesis in PC12 cells was inhibited by overexpression of H218 and to a lesser extent Edg-3. SPP treatment rapidly enhanced neurite retraction in PC12 cells overexpressing Edg-1, Edg-3, or H218. Thus, H218, and possibly Edg-3, may be the cell surface receptors responsible for cell rounding and neurite retraction induced by SPP. Moreover, the identification of these two additional SPP receptors indicates that a family of highly specific receptors exists that mediate different responses to SPP.

Several other responses to SPP are mediated through cell surface receptors, including platelet activation (21), inhibition of melanoma cell motility (22), activation of G i protein-gated inward rectifying K ϩ channels in atrial myocytes (23), and Rho-dependent neurite retraction and cell rounding of N1E-115 neurons (24) and PC12 cells (25). We recently identified the G protein-coupled receptor endothelial differentiation gene-1 (Edg-1) as a high affinity receptor for SPP (26). In response to SPP, Edg-1 inhibits adenylyl cyclase (20), activates the mitogen-activated protein kinase Erk2 (26) through a pertussis toxin-sensitive mechanism, and causes morphogenetic differentiation through a pertussis toxin-insensitive mechanism that requires the small GTPase Rho (26). Several other responses to SPP, including mobilization of intracellular Ca 2ϩ , activation of phospholipase D, and tyrosine phosphorylation of p125 FAK (20) are not mediated through Edg-1. Collectively, these data led us to suggest that SPP is a prototype of a new class of lipid second messengers that can also act as first messengers (20).
Recently, Edg-3 and H218 were shown by An et al. (34) to confer responsiveness to SPP of a serum response elementdriven reporter gene when expressed in Jurkat cells and to allow SPP-stimulated 45 Ca 2ϩ efflux in Xenopus oocytes, suggesting that Edg-3 and H218 may also be functional receptors for SPP. However, no direct binding data were presented in this study, and therefore it is not clear at present whether Edg-3 and H218 are bona fide SPP receptors. Moreover, sphingosine and sphingosylphosphorylcholine (SPC) were almost as effective as SPP in Jurkat T cells. Since SPP and sphingosine may also act intracellularly, it was of interest to determine the specificities and affinities of Edg-3 and H218 for these lipids using our recently developed binding assay (20). In this report, we show that both Edg-3 and H218 bind SPP with high affinity and remarkable specificity. Moreover, we have identified H218 as the receptor that mediates cell rounding and neurite retraction in response to SPP.
Cloning and Expression of edg-1, edg-3, and H218 -edg-1, edg-3, and H218 genes were amplified by PCR from human and rat genomic DNA, respectively. The primers were designed to add a BamHI site at the 5Ј-end and a XhoI site at the 3Ј-end as follows: 5Ј-GAGGGATCCGGG-CCCACCAGCGTCCCGCTG-3Ј and 5Ј-GAGCTCGAGCTAGGAAGAA-GAGTTGACGTTTCC-3Ј for edg-1; 5Ј-GAGGGATCCGCAACTGCCCT-CCCGCCGCGT-3Ј and 5Ј-GAGCTCGAGTCAGTTGCAGAAGATCCCA-TTCTG-3Ј for edg-3; 5Ј-GAGGGATCCGGCGGTTTATACTCAGAGTA-C-3Ј and 5Ј-GAGCTCGAGTCAGACCACTGTGTTGCCCTC-3Ј for H218. PCR products were cloned into the pcDNA3 vector (Invitrogen, Carlsbad, CA) containing a myc epitope tag at the 5Ј-end (a generous gift of Dr. Peter Burbelo). The resulting plasmids were transfected into HEK293 cells using Lipofectamine Plus (Life Technologies, Inc.) according to the manufacturer's instructions at a 4:1 ratio with pCEFL GFP, which encodes green fluorescent protein (a generous gift of Dr. Silvio Gutkind). The cells were then grown for 2 days to allow expression of receptors before experiments were performed. Transfection efficiencies were typically 30 -35% for HEK293 cells and 10% for PC12 cells.
Western Blotting-HEK293 cells were transfected as described above and lysed in PBS containing 1% CHAPS, 1 mM phenylmethylsulfonyl fluoride, and a 10 g/ml concentration each of leupeptin and aprotinin for 1 h at 4°C. Lysates were centrifuged, and equal amounts of protein from the supernatant were separated by SDS-polyacrylamide gel electrophoresis. Proteins were transferred to nitrocellulose membranes (Bio-Rad) and probed with monoclonal anti-c-Myc 9E10 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Bands were visualized with Super Signal chemiluminescent reagent (Pierce) using horseradish peroxidase-conjugated anti-mouse IgG.
SPP Binding Assay-[ 32 P]SPP was synthesized enzymatically using partially purified sphingosine kinase (36) as described previously (20). The specific activity of [ 32 P]SPP was 6 ϫ 10 6 cpm/pmol. Cells were incubated with the indicated concentration of [ 32 P]SPP in 200 l of binding buffer (20 mM Tris-HCl, pH 7.4, 100 mM NaCl, 15 mM NaF, 2 mM deoxypyridoxine, 0.2 mM phenylmethylsulfonyl fluoride, and 1 g/ml aprotinin and leupeptin) for 30 min at 4°C. Unlabeled lipid competitors were added as 4 mg/ml fatty acid-free bovine serum albumin complexes (3). Cells were washed twice with 200 l of ice-cold binding buffer containing 0.4 mg/ml fatty acid-free bovine serum albumin and resuspended in phosphate-buffered saline, and bound [ 32 P]SPP was quantitated by scintillation counting (20).
Cell Rounding Assay-HEK293 cells were plated at 2 ϫ 10 5 cells/well in 12-well dishes coated with polylysine and transfected 2 days later with the indicated receptor expression plasmids together with pCEFL GFP, as described above. Following transfections, cells were incubated in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum or 10% serum that had been stripped with activated charcoal to remove lipids (26). Cells were then washed and placed in serum-free medium for the indicated times and then treated with vehicle or with 100 nM SPP. Cells were fixed in 4% paraformaldehyde containing 5% sucrose for 20 min at room temperature and photographed using a Nikon Eclipse TE200 inverted fluorescence microscope connected to a Sony DKC5000 digital camera. Cells expressing GFP and GFP-expressing cells displaying rounded morphology were counted. At least three different fields were scored with a minimum of 300 cells scored.
For neurite retraction assays, PC12 cells were transfected as above, incubated overnight, and then split into 6-well dishes in RPMI contain-ing 10% heat-inactivated horse serum and 5% fetal bovine serum. Cells were transferred to serum-free medium, and NGF (100 ng/ml) was added to induce differentiation. After 48 h, vehicle or 100 nM SPP was added for the indicated time, cells were fixed with 4% paraformaldehyde containing 5% sucrose, and the morphology of transfected cells was evaluated by a blinded observer using a fluorescence microscope. Cells that had flattened, irregularly shaped cell bodies with neurite outgrowths were scored as differentiated. Cells with spherical shape lacking any neurite extensions and filopodia were scored as rounded. At least three different fields were scored for differentiated and rounded cells with a minimum of 100 cells scored. Cell rounding was expressed as the percentage of round shaped cells among total green fluorescent cells.
Staining of Apoptotic Nuclei-Fixed cells were washed with phosphate-buffered saline and then treated with bisbenzimide trihydrochloride (24 g/ml in phosphate-buffered saline; Hoechst dye 33258; Calbiochem) for 10 min. Stained cells were examined with an inverted fluorescence microscope. Transfected cells were marked by the expression of green fluorescent protein, and apoptotic cells were distinguished by condensed, fragmented nuclear regions using an ultraviolet filter. A minimum of 300 cells were scored.

Sphingosine 1-Phosphate Binds to H218 and Edg-3-Quan-
titative binding studies with lysolipid agonists such as SPP and LPA have been hampered by the lipophilic nature of the ligand, resulting in very high levels of nonspecific binding (for example, see Ref. 37). However, recently we have developed a sensitive assay for binding of SPP to cell surface receptors that overcomes these problems (20). To determine if the Edg-3 and H218 G protein-coupled receptors, which share 48 and 45% homology, respectively, with Edg-1, were capable of specifically binding SPP, HEK293 cells were transiently transfected with pcDNA3 expression plasmids containing H218 or edg-3 open reading frames with c-myc tags fused in frame at the N terminus. HEK293 cells were selected for these studies, since they have no specific SPP binding. Although no edg-1 mRNA expression can be detected by Northern analysis, barely detectable expression can be seen by RT-PCR. Additionally, a low level of edg-3 mRNA was detected by RT-PCR; however, no product was seen using primers specific for H218 (data not shown). Western analysis confirmed that an approximately 43-kDa protein containing the c-Myc tag was expressed in cells transfected with edg-3 or H218 but not with the pcDNA3myc vector or in untransfected cells (Fig. 1A). H218-and Edg-3-expressing HEK293 cells display dramatically increased specific binding of [ 32 P]SPP in comparison with untransfected and vector-transfected cells (Fig. 1B). HEK293 cells expressing H218 and Edg-3 had similar binding affinities for [ 32 P]SPP with K D values of 27 and 23 nM, respectively (Fig. 2).
Several other lipids that are structurally related to SPP, especially SPC and LPA, often have similar effects to SPP, and it has been suggested that in some instances they may share the same receptor (21,38). Moreover, recently it has been suggested that LPA is a low affinity agonist for Edg-1 (39). Therefore, the specificity of binding to H218 and Edg-3 for a variety of lipid analogs of SPP was tested. Similar to our previous results with Edg-1 (20), only unlabeled SPP and dihydro-SPP, which lacks the 4-trans-double bond present in SPP, effectively competed with [ 32 P]SPP for binding to cells expressing either H218 or Edg-3 (Fig. 3, A and B). Sphingosine had a small but statistically significant effect on SPP binding to both Edg-3 and H218, while ceramide had a similarly small effect on binding to H218 only. Interestingly, the homophosphonate analog of SPP, which differs from SPP in that the oxygen atom at the 1-position is replaced by a carbon atom, competed for binding to Edg-1 as effectively as did unlabeled SPP (Fig. 3C), although it had only a slight effect on binding of SPP to Edg-3 and no effect on binding to H218 (Fig. 3, A and B). The other lipids tested, including both LPA and SPC, do not compete for binding of SPP to any of the three Edg family receptors ( Fig. 3 and Ref. 20).
H218 Mediates Sphingosine 1-Phosphate-induced Cell Rounding-Cell rounding is one of the most profound biological responses induced by low (nanomolar) concentrations of SPP that is thought to be mediated by a cell surface receptor (24). HEK293 cells transfected with H218 and edg-3 exhibited marked changes in cell morphology and increased numbers of rounded cells in comparison with vector-and edg-1-transfected cells (Fig. 4). To quantitate the magnitude of this response, the percentage of transfected cells, indicated by expression of green fluorescent protein (GFP), which displayed rounded morphology, was determined. Expression of H218 or Edg-3 increased rounding of HEK293 cells, while Edg-1 did not have a significant effect (Fig. 5A). Since the loss of cell attachment can lead to death by apoptosis (40 -42), the percentage of transfected cells that displayed apoptotic characteristics was also determined. Expression of H218 or Edg-3 also increased the percentage of transfected cells HEK293 cells that displayed fragmented nuclei characteristic of apoptosis (Fig. 5B). Edg-1 expression led to a significant but much smaller increase. Although this response was seen in the absence of exogenously added SPP, these cells were grown in the presence of serum, which has been shown to contain high levels of SPP (43).
To determine whether binding of SPP to these receptors mediates cell rounding, transfected HEK293 cells were grown in charcoal-stripped serum, which contains no detectable SPP. 2 Under these conditions, HEK293 cell morphology was similar in all cells in the absence of SPP (Fig. 6). Treatment with 100 nM SPP induced modest cell rounding in vector-transfected and Edg-1-expressing HEK293 cells. However, SPP induction of cell rounding was dramatically increased by expression of H218, whereas expression of Edg-3 appeared to only slightly enhance the rounding effect of SPP. To quantify this response, the number of rounded cells, expressed as a percentage of the transfected cells identified by expression of GFP, was determined (Fig. 7A). Expression of all three receptors modestly enhanced the percentage of rounded cells in comparison with vector-transfected controls. Treatment with SPP further enhanced cell rounding to a small degree in Edg-3-expressing cells and by approximately 2-fold in H218-expressing cells. Expression of these Edg receptors had a similar effect on apoptosis (Fig. 7B). Edg-1 expression slightly increased apoptosis, and this effect was not enhanced by SPP. Edg-3 expression led to a more pronounced increase in apoptosis. H218 expression was again the most effective, leading to a significant increase in apoptosis, which was further enhanced by SPP treatment.
H218 Mediates Sphingosine 1-Phosphate-induced Neurite Retraction-SPP has previously been shown to induce rapid retraction of developing neurites and transient rounding of the cell body in PC12 cells (25) and in N1E-115 neuronal cells (24). However, the putative cell surface receptor mediating these 2 L. Edsall and S. Spiegel, unpublished observations.

FIG. 1. SPP binds to H218 and Edg-3 transiently expressed in HEK293 cells.
A, HEK293 cells were either not transfected (untrans.) or transiently transfected with pcDNA3myc (vector), pcDNA3myc containing the edg-3 coding sequence, or the H218 coding sequence as described under "Experimental Procedures." Western blotting was performed using an antibody specific for the c-Myc epitope tag. B, binding of 1 nM [ 32 P]SPP to untransfected HEK293 cells and cells transiently transfected with pcDNA3myc vector alone or containing H218 or edg-3 coding sequence was determined as described under "Experimental Procedures." Total binding is in the absence of unlabeled competitor, and nonspecific binding is in the presence of 1 M unlabeled SPP. Results are means Ϯ S.D. of triplicate determinations. effects has not yet been identified. PC12 cells express H218 (33), and it has been suggested that H218 is involved in regulation of some of the early steps in neuronal differentiation, including axonal outgrowth (44). In agreement, we found by RT-PCR that PC12 cells express H218 but not edg-3 or edg-1 (data not shown). Thus, it was of interest to examine whether H218 might be the cell surface receptor responsible for neurite retraction induced by SPP. Approximately 50% of PC12 cells transfected with vector alone have long processes when cultured in serum-free medium in the presence of NGF. This morphology remained unchanged when cells were treated with SPP for 10 min. However, expression of H218 and Edg-3, but not Edg-1, decreased NGF-induced differentiation in the absence of added SPP, with the strongest effect observed with H218 (Fig. 8). Expression of these receptors results in changes in morphology leading to retraction of neurites and soma rounding, producing cells with a spherical appearance (Fig. 8). As in HEK293 cells, H218 was the most potent receptor for this response. The addition of nanomolar concentrations of SPP to NGF-differentiated PC12 cells overexpressing the different Edg family receptors further enhanced this response. This is in contrast to the lack of effect of Edg-1 expression on cell rounding in HEK293 cells. Flattened cells start to round up rapidly after the addition of SPP, with rounding being complete within 10 min. As in HEK293 cells, rounding of Edg-expressing PC12 cells in response to SPP was correlated with death by apoptosis; however, increased apoptosis in SPP treated Edg-expressing cells was not seen until 24 h of treatment (Fig. 8C). Since SPP treatment did not lead to increased apoptosis during the first 3 h, SPP-induced cell rounding is unlikely to be a result of cell death. DISCUSSION The G protein-coupled Edg family of receptors is now known to include numerous members. Edg-3 and H218 are nearly 50% homologous to the known SPP receptor Edg-1. In contrast, Edg-2/Vzg-1 and Edg-4, both of which have been shown to act as receptors for LPA (27)(28)(29)(30), are only 31 and 27% homologous to Edg-1, respectively. The binding data presented in this report establish that, similar to our previous studies on Edg-1 (26), Edg-3 and H218 are also bona fide receptors for SPP. The affinities of these receptors for SPP are similar: 8 nM (26), 23 nM, and 27 nM for Edg-1, Edg-3, and H218, respectively. Moreover, our data showing lack of competition for SPP binding to Edg-1, Edg-3, and H218 by LPA suggest that there may be two subfamilies of Edg receptors, one specific for SPP and the other for LPA (27)(28)(29). We propose that these receptors should thus be named based on their ligand binding specificities. Thus, Edg-1, H218, and Edg-3 should be named SPPR1, SPPR2, and SPPR3, respectively, and Edg-2/Vzg-1 and Edg-4 should be named LPAR1 and LPAR2. SPC and LPA are structurally similar to SPP and often induce similar biological responses. Binding of SPP to platelets is inhibited by LPA, and LPA-induced platelet aggregation is prevented by pretreatment with SPP, suggesting that in platelets, LPA and SPP bind to a common receptor (21). SPC and SPP apparently share a receptor in atrial myocytes, since they both activate a K ϩ conductance in these cells and cause heterologous desensitization (38). However, neither SPC nor LPA compete for binding of radioactively labeled SPP to any of the three receptors investigated in this study. Therefore, it is likely that there are other, as yet unidentified, Edg family members that do not show such high ligand specificity and are more promiscuous. Interestingly, expression of Edg-3 or H218 in Jurkat cells allows activation of SRE-driven gene transcription by SPP, SPC, and sphingosine (34). Since neither SPC nor sphingosine compete for binding of SPP to Edg-3 or H218, the reason for this is unclear. However, because these sphingolipid metabolites also have intracellular actions (20,45,46), the possibility exists that their effects are partially mediated via intracellular targets. A recent study demonstrated that LPA can bind to the Edg-1 receptor with an apparent K D of 2.3 M, resulting in receptor phosphorylation and extracellular signalregulated kinase activation as well as Rho-dependent morphogenesis and P-cadherin expression (39). Moreover, the binding of labeled LPA was competed only poorly by unlabeled SPP. However, it should be noted that SPP desensitizes [ 3 H]LPA binding to Edg-1 (39), while our studies indicate that LPA does not desensitize [ 32 P]SPP binding to Edg-1. These results suggest that there may be two different binding sites on Edg-1: one that binds SPP with high specificity in the nanomolar concentration range and one that binds LPA with low affinity in the micromolar concentration range.
Although Edg-1, Edg-3, and H218 bind SPP with high specificity, there are some differences in the binding of SPP analogs. The homophosphonate analog of SPP binds to Edg-1 as well as does SPP, but it only competes weakly for SPP binding to Edg-3 and not at all for H218. In addition, the short chain C8 analog of SPP slightly inhibits SPP binding to Edg-1 but is completely ineffective for Edg-3 and H218. Thus, although these receptors are highly homologous and bind SPP with similar affinities, their binding sites may not be identical. SPP analogs with different specificities for the different SPP receptors should be useful to determine which receptors mediate specific biological responses to SPP.
HEK293 cells transiently expressing Edg-3 or H218 receptors exhibited a rounded morphology when grown in the presence of serum. However, removal of lipids from serum by charcoal stripping prevented this effect on morphology. Serum contains high levels of SPP (400 nM) (43), and charcoal stripping reduces SPP to undetectable levels. 2 In agreement, the HEK293 cells were transiently transfected with pcDNA3myc or pcDNA3myc containing the edg-1, edg-3, or the H218 coding sequences together with pCEFL GFP; grown overnight in the presence of delipidated serum; and then changed to serum-free medium for an additional day. Cells were treated with vehicle or with 100 nM SPP for 3 h and fixed, and phase contrast images of cellular morphology are shown. Similar results were obtained in two separate experiments. addition of SPP markedly increased cell rounding in cells expressing H218 and to a lesser extent in Edg-3-expressing cells. Moreover, cells transiently transfected with H218, when cultured in serum-free medium, also exhibited a small but significant increase in cell rounding. It is possible that H218 is partially activated in the absence of ligand when the receptor is overexpressed, a phenomenon commonly observed for G protein-coupled receptors (47). Our results indicate that H218 may be the unidentified cell surface receptor that was previously suggested in several studies to be responsible for SPP-induced cell morphology alterations and remodeling of the actin cytoskeleton (48,49).
Apoptosis of mammalian cells can occur by default unless the cell receives certain survival signals (50,51). Loss of cell attachment induces apoptosis in certain cell types due to the loss of integrin-mediated signaling, which normally suppresses apoptosis, and this phenomenon has been termed anoikis (40 -42). Thus, detachment of HEK293 cells or PC12 cells following SPP-induced rounding could lead to death by anoikis. The close correlation of apoptotic appearing cells with rounded cells (Figs. 5 and 7) supports this possibility. However, it cannot be concluded that SPP normally induces apoptosis by binding to these cell surface receptors, since overexpression of receptors could result in a robust, nonphysiological signal. It has been reported that nanomolar concentrations of SPP induce neurite retraction and rounding in differentiated PC12 cells (25) and N1E-115 neuroblastoma cells (24). In both of these studies, it was postulated that the SPP effect was mediated through a putative cell surface receptor and not by intracellular actions, since microinjected SPP had no effect (24), whereas SPP immobilized to glass beads was effective (25). In this study, we found that expression of H218, and to a lesser extent Edg-3, in rat pheochromocytoma PC12 cells caused a decrease in NGFinduced neurite outgrowth and increased the fraction of cells with rounded morphology. MacLennan et al. (33) demonstrated that PC12 cells express H218. In agreement, we detected H218 expression in PC12 cells by RT-PCR analysis; however, we were not able to detect expression of edg-1 or edg-3. Thus, it is likely that SPP-induced neurite retraction in PC12 cells is mediated through H218. Interestingly, treatment of NGF-differentiated PC12 cells overexpressing any of the Edg receptors with SPP for as little as 10 min caused further neurite retraction and cell rounding, although the effect was most marked with H218. This is in contrast to the moderate efficacy of Edg-3 and the lack of effect of Edg-1 on rounding in HEK293 cells. Therefore, it appears that all three receptors are capable of coupling to signaling pathways leading to rounding depending on the cellular context.
H218 is expressed in the cardiovascular system (32) and in the brain during embryogenesis, where its expression is temporally regulated such that high levels of expression are found in neuronal cell bodies during early stages of differentiation and in axons during their outgrowth (44). This led to the suggestion that H218 plays an important role in neuronal development and may steer axons by regulating their growth and inhibiting their extension (44). Thus, SPP synthesized by target tissues could help guide axons by regulating axon extension or stabilization through binding to H218.
In summary, we have shown that the G protein-coupled receptors H218 and Edg-3 are high affinity, specific receptors for SPP and that binding of SPP to these receptors induces cell rounding. H218 is clearly the most efficacious mediator of this response, although Edg-3 is also effective. Since these receptors are widely expressed in most cells and tissues, important questions that should be addressed in the future are their roles in mediating various biological responses to SPP. Thus, SPP might play a role during normal brain development or after traumatic injury by acting through H218 and possibly Edg-3 to affect neuritogenesis.