This Article Abstract Full Text (PDF) Data Supplement Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Panetti, T. S. Articles by Mosher, D. F. Search for Related Content PubMed PubMed Citation Articles by Panetti, T. S. Articles by Mosher, D. F. Related Collections Angiogenesis

(Arteriosclerosis, Thrombosis, and Vascular Biology. 2000;20:1013.)
© 2000 American Heart Association, Inc.

Vascular Biology

Sphingosine-1-Phosphate and Lysophosphatidic Acid Stimulate Endothelial Cell Migration

Tracee Scalise Panetti; Julie Nowlen; Deane F. Mosher

From the Departments of Medicine and Biomolecular Chemistry, University of Wisconsin-Madison, Madison, Wis.

Correspondence to Tracee Scalise Panetti, MD, Departments of Medicine and Biomolecular Chemistry, University of Wisconsin-Madison, 1300 University Ave, Madison, WI 53706. E-mail tpanetti{at}facstaff.wisc.edu

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract—Endothelial cell migration is necessary for the formation of new blood vessels. We investigated the effects of 2 lysophospholipid mediators, sphingosine-1-phosphate (S1P) and lysophosphatidic acid (LPA), on endothelial cell migration. S1P and LPA stimulated migration of fetal bovine heart endothelial cells (FBHEs) in a 3D-modified Boyden chamber assay with concentrations as low as 15 nmol/L stimulating a 2-fold change and concentrations in the 1- to 2-µmol/L range stimulating 14- to 20-fold changes. S1P specifically stimulated the migration of several endothelial cell strains but did not stimulate the migration of tumor cells or smooth muscle cells. LPA stimulated some endothelial and nonendothelial cell types to migrate. For FBHEs, S1P and LPA were mostly chemokinetic in checkerboard assays. S1P and LPA stimulated extracellular signal–regulated kinase 1/2 phosphorylation and enhanced paxillin localization to focal contacts, with no discernible change in the actin cytoskeleton in FBHEs. To characterize responsible receptor-dependent signaling pathways, we investigated the involvement of G_i, Rho, and phosphoinositide 3-OH kinase in S1P- and LPA-stimulated migration. Although perturbation of all 3 signaling molecules resulted in decreased migration, the mechanisms underlying the decreased migration were different. Pertussis toxin treatment, to target G_i, caused endothelial cells to develop dense bundles of F-actin and distribute paxillin staining to the cell periphery in response to S1P or LPA. Modification of Rho with C3 toxin disrupted the actin cytoskeleton. Inhibition of phosphoinositide 3-OH kinase decreased S1P- or LPA-induced endothelial cell migration with only minor disruption of the actin cytoskeleton. Inhibition of extracellular signal-regulated kinase kinase with PD98059 caused a loss of phosphorylation of extracellular signal–regulated kinase 1/2, similar to pertussis toxin, but only a minimal decrease in migration. These results indicate that S1P and, for some cells, LPA stimulate migration of endothelial cells through a mechanism that likely requires a balance between G_i and Rho signaling to achieve the cytoskeletal remodeling necessary for cell migration.

Key Words: sphingosine-1-phosphate • lysophosphatidic acid • endothelial cells • cell migration

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Sphingosine-1-phosphate (S1P) is a lysophospholipid mediator released on platelet activation with a serum concentration of 0.5 µmol/L, 2-fold higher than the plasma concentration.¹ S1P stimulates DNA synthesis by fibroblasts, neurite retraction, stress fiber formation,^{1 2} fibronectin deposition,³ and suppression of T-cell apoptosis.⁴ Lysophosphatidic acid (LPA) is the major lysophospholipid growth factor in serum generated as a result of platelet activation.² The concentration of LPA in serum is estimated at 1 to 10 µmol/L, whereas the concentration in plasma is considerably less.¹ LPA has numerous biological effects on cells, including stimulation of cell proliferation, increased tight junction permeability in brain endothelial cells, stress fiber formation, increased binding and deposition of fibronectin by fibroblasts, smooth muscle cell contraction, and platelet aggregation.²

Several G-protein–coupled receptors of the endothelial cell differentiation gene (Edg) family that bind and mediate biological responses to S1P and/or LPA have been identified (reviewed in References 1 and 5^{1 5} ). LPA and S1P stimulate the Rho, Ras/mitogen-activated protein kinase, and phospholipase C pathways and perhaps others through G proteins.^{1 2 5} Rho is a small GTPase downstream from G_12/13 that stimulates actin reorganization¹ and is inactivated by C3 toxin from Clostridium botulinum.⁶ G_i, the pertussis toxin–sensitive heterotrimeric G protein, stimulates the Ras pathway that leads to extracellular signal–regulated kinase (Erk) 1/2 activation through multiple intermediates, including extracellular signal-regulated kinase kinase (Mek-1).^{1 2} G_i activation of Ras may include phosphoinositide 3-OH kinase (PI3-kinase) as an intermediate.² Stimulation of the G_i pathway may also inhibit cAMP generation by cells.¹ Finally, lysophospholipids stimulate phospholipase C downstream from G_q.¹

New blood vessel formation, angiogenesis, is required for normal development and wound healing.^{7 8} Aberrant angiogenesis contributes to many diseases, including diabetic retinopathy, psoriasis, tumor growth, and arthritis.⁸ Directed migration of endothelial cells is one of the critical steps in angiogenesis.^{7 8} The present studies were initiated to examine the migratory response of endothelial cells in response to S1P and LPA. LPA has been shown to induce cell migration of nonendothelial cells by pathways that use G_i,^{9 10} Rho activation,^{10 11} Ca²⁺ mobilization,¹² Rac activation,¹³ and Ras activation¹⁴ but do not require activation of Erk1/2.^{13 14} S1P, in contrast, has been shown to inhibit cell migration of smooth muscle cells,¹⁵ mouse melanoma cells,¹⁶ breast cancer cells,¹⁷ neutrophils,¹⁸ and fibroblasts.¹⁰ We found that endothelial cells derived from several sources migrated in response to S1P. The responses to LPA were more variable. Inactivation of G_i, inactivation of Rho, and inhibition of PI3-kinase all decreased S1P- and LPA-induced endothelial cell migration independent of their effects on the activation of Erk1/2 but with characteristic effects on the cytoskeleton.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Materials
S1P was purchased from Alexis Corp and solubilized in methanol. LPA (1-oleoyl-LPA) was purchased from Avanti Polar Lipids and solubilized as previously described.¹⁹ Lysophosphatidylcholine, lysophosphatidylserine, and sphingomyelin were purchased from Sigma Chemical Co. Vascular endothelial growth factor (VEGF) and epidermal growth factor were purchased from R&D Systems and Upstate Biotechnology, respectively. Cells were pretreated with pertussis toxin (List Biologicals) or C3 toxin overnight at the indicated concentrations. C3 toxin was expressed in the pGex vector (generous gift of Dr Larry Feig, Tufts University Medical School, Boston, Mass) as a glutathione S-transferase fusion protein, and purification of the fusion protein was performed by glutathione-agarose affinity chromatography as described.⁶ The fusion protein was cleaved with thrombin while bound to glutathione-agarose, and C3 toxin was recovered in the supernatant after removal of thrombin by benzamidine-Sepharose. LY294002 (Sigma) and wortmannin (Sigma) were dissolved in dimethyl sulfoxide at 5 to 10 mmol/L and stored at -20°C. The Mek-1 inhibitor, PD98059 (Calbiochem), was dissolved in dimethyl sulfoxide at 6.5 mg/mL and stored at -20°C. The inhibitors were preincubated with cells for 30 to 60 minutes before the cells were used in experiments.

Cell Culture and Cell Migration Assays
Fetal bovine heart endothelial cells (FBHEs) and ECV (T-24 variant) bladder carcinoma cells were obtained from American Type Culture Collection and cultured in DME containing 10% FBS. Bovine vascular smooth muscle cells and bovine aortic endothelial cells (BAEs) were obtained from the Coriell Institute for Medical Research. FBHEs were supplemented with 20 ng/mL recombinant basic fibroblast growth factor (bFGF, generous gift from Scios, Mountain View, Calif). BAEs were cultured in DME containing 20% FBS. Human umbilical vein endothelial cells (HUVECs), bovine adrenal microvascular endothelial cells (BAMECs), bovine lung microvessel endothelial cells (BLMVECs), and bovine pulmonary artery endothelial cells (BPAEs) were obtained from VEC Technologies, Inc, and cultured in MCDB-131 complete media (VEC Technologies, Inc). Cells were grown in a humidified incubator at 37°C with 8% CO₂ (FBHEs, ECVs, and BAEs) or 5% CO₂ (HUVECs, BAMECs, BLMVECs, BPAEs, and vascular smooth muscle cells) and passaged twice a week. For migration assays, newly confluent cells in growth media were lifted with trypsinization in the presence of EDTA; the trypsin was inhibited with 10% FBS, and cells were washed 3 times in DME containing 0.2% fatty acid–free BSA (FAF-BSA). There was no serum starvation. The migration assays were performed using a 48-well chemotaxis chamber (Nucleopore) with cells and mediators in DME with 0.2% FAF-BSA. Polyvinylpyrrolidone-free polycarbonate membranes with 5-µm pores (Corning/Costar) were coated with fibronectin (10 µg/mL) or vitronectin (10 µg/mL) overnight, rinsed, and air-dried before use. Fibronectin and vitronectin were purified from human plasma, free of platelet-derived growth factors, as described previously.²⁰ The chemotactic agents were added to the lower wells, and cells (1x10⁵ cells per 50 µL) were added to the upper wells. After 6 hours at 37°C, the chamber was disassembled, and the top of the filter was scraped to remove nonmigrated cells. The filter was fixed, stained with DiffQuick (Fisher Scientific), and air-dried on a slide. Each condition was performed in triplicate, and three 0.16-mm² fields from each well were counted at x400 magnification. Cell motility was determined as described²⁰ on a fibronectin substrate with cells and stimulators added to the wells in DME with 0.2% FAF-BSA.

Fluorescence Microscopy
Cells were plated in DME containing 10% FBS at a concentration of 1.5x10⁵ cells per well in a 2-cm² well containing a glass coverslip. After 4 to 6 hours, the cells were rinsed and left overnight in DME containing 0.2% FAF-BSA. The cells were treated with stimulators and inhibitors as described above, fixed with 3% paraformaldehyde, and permeabilized with 0.2% Triton X-100. For visualization of actin stress fibers, cells were incubated with rhodamine-phalloidin (100 µg/mL, Sigma). For visualization of paxillin, cells were incubated with anti-paxillin (Transduction Laboratories) at 5 µg/mL, followed by fluorescein-conjugated goat anti-mouse IgG (Cappel/Organon Teknika) at 1:100 dilution. Images were obtained with an Olympus camera or Photometrix CCD camera mounted on an Olympus BX-60 epifluorescence microscope.

Immunoblots
Cells were plated in 2-cm² plates with 2x10⁵ cells per well in DME containing 10% FBS, allowed to attach for 4 to 6 hours, and serum-starved overnight in DME containing 0.2% FAF-BSA. The cells were treated with stimulators and inhibitors as described above. Immediately after stimulation, cells were placed on ice, washed, solubilized in lysis buffer (2% SDS and 10% glycerol in 50 mmol/L Tris-HCl, pH 6.8), and boiled. Protein concentration was determined by use of a BCA kit (Pierce), and samples were treated with ß-mercaptoethanol and boiled. Samples (10 µg per lane) were run on 8% SDS-PAGE and transferred to nitrocellulose. Transferred protein was reversibly stained with Ponceau S (Sigma) to ensure equivalent loading of protein in each lane. Nitrocellulose was incubated with polyclonal antibodies to doubly phosphorylated active Erk1/2 (Promega) at a 1:20 000 dilution, followed by a secondary antibody, horseradish peroxidase–conjugated goat anti-rabbit (Cappel/Organon Teknika) at 1:5000, and detected by use of the Renaissance Chemiluminescence kit (NEN Life Sciences).

	Results

Top Abstract Introduction Methods Results Discussion References

 
S1P or LPA stimulated FBHE migration in a dose-dependent fashion over a wide concentration range (Figure 1). Migration, 5- to 8-fold greater than baseline, was stimulated with S1P or LPA concentrations as low as 15 nmol/L S1P, with a 14- to 20-fold increase above baseline at 0.5 to 2 µmol/L S1P or LPA. Maximum migration was somewhat higher for S1P than for LPA. The migration was not stimulated above baseline by 1 µmol/L lysophosphatidylcholine, lysophosphatidylserine, or sphingomyelin (data not shown). The S1P and LPA responses were similar on fibronectin- or vitronectin-coated filters for FBHEs, although background migration in the absence of growth factor was higher on vitronectin than on fibronectin (data not shown). VEGF (10 ng/mL) stimulated FBHE migration to 50% of the value seen with 1 µmol/L S1P, whereas epidermal growth factor did not stimulate migration above baseline (data not shown). Endothelial cells from other sources, including BAEs, HUVECs, BLMVECs, and BPAEs, migrated to S1P, whereas vascular smooth muscle cells and ECV bladder carcinoma cells did not migrate to S1P (Table). BAMECs were tested in a 2D cell motility assay and showed migration in response to S1P. LPA stimulated the migration of BAEs but not HUVECs, BAMECs, BPAEs, or BLMVECs. LPA caused bladder carcinoma cells (but not vascular smooth muscle cells) to migrate.

View larger version (16K): [in this window] [in a new window]   Figure 1. Dose-response of S1P and LPA stimulation of FBHE migration. Cells (1x105 cells per 50 µL) were added to the top of the fibronectin-coated polycarbonate filter and migrated through in response to S1P or LPA added to the bottom. After 6 hours, nonmigrated cells were scraped from the top of the filter, and migrated cells were fixed and stained. Each condition was performed in triplicate, and 3 fields from each well were counted at x400 magnification (n=9). Data are expressed as mean±SEM. Data are an average of 3 experiments.

View this table: [in this window] [in a new window]   Table 1. Cell Migration to Lysophospholipids

To examine the directional component of migration, we examined lysophospholipid-induced migration of FBHEs by using a checkerboard analysis. S1P or LPA was required in the bottom chamber for maximum migration of cells to the bottom of the filter (Figure 2). High concentrations of S1P in the top, however, stimulated some cell migration to the bottom of the filter, albeit there was 4-fold less migration than with S1P in the bottom chamber. S1P and LPA enhanced migration in the absence of a gradient (equal concentrations in the top and bottom), showing that there is a large chemokinetic component to the migration.

View larger version (43K): [in this window] [in a new window]   Figure 2. Checkerboard assay of FBHE migration in response to LPA (A) or S1P (B). The migration assay was performed as described for Figure 1, except that LPA or S1P was added to the top and/or bottom wells as indicated. The data are expressed as mean of 9 determinations, with the SEM being <5% of the mean. The data are from a representative experiment that was repeated 2 times.

To understand the downstream signaling pathways stimulated by S1P or LPA that lead to FBHE migration, we tested agents that perturb signal transduction. Pertussis toxin modification of G_i is known to block stimulation of Ras and the induction of mitogenesis in cells treated with S1P or LPA.^{1 2} Overnight pretreatment with pertussis toxin, at doses as low as 2.5 ng/mL (Figure IA, published online only at http://atvb.ahajournals.org/cgi/content/full/20/4/1013/DC1), caused a loss of ability of FBHEs to migrate in response to S1P or LPA (Figure 3). Pertussis toxin was not globally deleterious to the endothelial cells, inasmuch as the cells attached and spread on tissue culture plastic in the presence of 10% FBS (data not shown) and assembled a fibronectin matrix (data not shown). To learn whether the effects of pertussis toxin are mediated downstream from mitogen-activated protein kinase, a Mek-1 inhibitor, PD98059, was tested. A 25 µmol/L concentration of PD98059 had little effect on S1P- or LPA-induced endothelial cell migration (Figure 3) and was not dose dependent (Figure IB). Treatment with either pertussis toxin or PD98059, however, caused equivalent loss of Erk1/2 activation in response to S1P or LPA (Figure 4). These results suggest that the G_i-mediated pathway important for cell migration diverges upstream from Mek-1.

View larger version (38K): [in this window] [in a new window]   Figure 3. Comparison of effects of modifiers on S1P- and LPA-induced FBHE migration. The data are expressed as percentage of control±SEM (n=9), where control indicates cell migration in the absence of the modifiers (100%). Background migration, in the absence of LPA or S1P, was subtracted from all values. The concentrations of the modifiers were as follows: pertussis toxin (PTX), 5 ng/mL; C3 toxin (C3), 50 µg/mL; LY294002 (LY), 50 µmol/L; and PD98059 (PD), 25 µmol/L. Data are expressed as mean±SD (n=9) and are from a representative experiment that was repeated 4 times. All mediators are statistically different from control, as determined by t test.

View larger version (18K): [in this window] [in a new window]   Figure 4. Western blot detection of active Erk1/2 in FBHEs after treatment with modifiers of Rho, Gi, PI3-kinase, and Mek-1. Cells were plated in DME with 10% FBS, attached for 4 to 6 hours, serum-starved overnight, and pretreated with the different substances at the indicated concentrations: untreated (control); C3 toxin, 50 µg/mL; pertussis toxin, 5 ng/mL; LY294002, 50 µmol/L; and PD98059, 25 µmol/L. Cells were stimulated for 5 minutes with 1 µmol/L LPA or S1P as indicated, put on ice, and solubilized in lysis buffer. Erk1 and Erk2 are designated as p44 and p42, respectively. Data are from a representative experiment that was repeated 2 times.

Cell movement requires the reorganization of the actin cytoskeleton as the cell establishes a leading edge and migrates forward.²¹ Therefore, we examined the effect of modifiers of S1P- and LPA-induced cell migration on the actin cytoskeleton and paxillin staining in focal contacts. In untreated cells, there was no detectable change in the actin cytoskeleton after lysophospholipid stimulation (Figure 5A); however, paxillin immunofluorescence was altered (Figure 5B). Paxillin staining was found in fine fibrillar streaks in serum-starved cells but was more intense, with shorter thicker patches, after lysophospholipid stimulation. Inhibition of G_i with pertussis toxin did not have a dramatic effect on the actin cytoskeleton or paxillin accumulation in the focal contacts in nonstimulated cells. S1P or LPA, however, induced cellular contraction of pertussis toxin–treated cells: the actin cytoskeleton was more cortical, and the cells were less flattened so that the cytoskeleton was present in several planes with only a small portion in focus in any single plane (Figure 5A). The paxillin staining was also rearranged and present only at the cell periphery and was not distributed throughout the cell, as in the non–pertussis toxin-treated cells (Figure 5B).

View larger version (130K): [in this window] [in a new window]   Figure 5. Effect of S1P and LPA on actin stress fiber formation and paxillin localization. FBHEs were plated on fibronectin-coated glass coverslips, attached for 4 to 6 hours, and serum-starved overnight in the absence or presence of PTX (5 ng/mL). Cells were untreated (control) or stimulated for 15 minutes with 1 µmol/L LPA or S1P, fixed, and permeabilized. A, Actin cytoskeleton was visualized by rhodamine-phalloidin staining. B, Paxillin localization was examined by anti-paxillin antibodies followed by a fluorescein-labeled secondary antibody. Cells are representative of changes observed in 3 different experiments.

Two additional signaling molecules, Rho and PI3-kinase, were perturbed to characterize better the processes involved in endothelial cell migration stimulated by lysophospholipids. PI3-kinase links the activation of G_i to Ras² and the activation of Ras to other small G proteins, particularly Rac, with downstream activation of Rho.²² Furthermore, S1P or LPA, signaling through G_12/13, causes direct activation of Rho.¹ FBHE migration in response to S1P or LPA was decreased after an overnight preincubation with C3 toxin, the Rho inactivator, to 25% of control at a concentration of 10 µg/mL (Figure 3). Preincubation with increasing doses of toxin was not associated with further inhibition (Figure II, published online only at http://atvb.ahajournals.org/cgi/content/full/20/4/1013/DC1). The toxin inhibited migration with the same efficacy regardless of the LPA dose used to stimulate cell migration (Figure II). The toxin did not have deleterious effects on the cells, as shown by cell attachment to tissue culture plastic in DME containing 10% FBS (data not shown). Pretreatment with wortmannin (10 µmol/L) or LY294002 (50 µmol/L), 2 inhibitors of PI3-kinase,²³ resulted in a decrease in migration in response to S1P, LPA, or serum to 30% in the absence of the inhibitor (Figure III, published online only at http://atvb.ahajournals.org/cgi/content/full/20/4/1013/DC1), and Figure 3). C3 toxin and the PI3-kinase inhibitor LY294002 each caused a modest inhibition of Erk1/2 phosphorylation in response to S1P or LPA (Figure 4).

The effects of C3 toxin, wortmannin, LY294002, or PD98059 on actin cytoskeleton were compared with the effects of pertussis toxin (Figure 6). Despite serum starvation, the cells exhibited an extensive actin cytoskeleton, as detected by rhodamine-phalloidin fluorescence in the absence of modifiers (Figure 6). C3 toxin caused loss of the actin cytoskeleton, such that in the majority of C3 toxin–treated cells, no filamentous actin was detected. Wortmannin also disrupted the actin cytoskeleton with an intermediate phenotype of fewer stress fibers than control cells but not a complete loss of actin cytoskeleton, as was caused by C3 toxin. LY294002 had the same effect as wortmannin, with some loss of stress fibers (data not shown). PD98059 did not effect the actin cytoskeleton. The actin cytoskeleton remained unchanged after stimulation with S1P or LPA in the presence of these modifiers (data not shown).

View larger version (120K): [in this window] [in a new window]   Figure 6. Rhodamine-phalloidin staining of FBHEs after treatment with modifiers of Rho, PI3-kinase, and Mek-1. Cells were plated on glass coverslips in DME with 10% FBS, attached for 4 to 6 hours, serum-starved overnight, and pretreated with C3 toxin (50 µg/mL), wortmannin (10 µmol/L), or PD98059 (25 µmol/L), as indicated in Methods. Cells are representative of changes observed in 2 different experiments.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
We found that S1P stimulates the migration of FBHEs, BAEs, HUVECs, BLMVECs, BPAEs, and BAMECs. This response was specific inasmuch as S1P did not stimulate the migration of vascular smooth muscle cells or bladder carcinoma cells. S1P inhibits migration or does not stimulate the migration of a number of other non–endothelial cell types, including vascular smooth muscle cells,¹⁵ mouse melanoma cells, BALB/c3T3 fibroblasts, human HT1080 fibrosarcoma cells, MG63 osteosarcoma cells,¹⁶ breast cancer cells,¹⁷ ß1A-GD25 fibroblasts,¹⁰ and human neutrophils.¹⁸ T-lymphoma cells have been shown to invade a fibroblast monolayer after S1P treatment but only after transfection of activated Tiam1, a Rac guanine nucleotide exchange factor.¹² Endothelial cells, therefore, are unique in that S1P stimulates cell migration in nontransformed cells. LPA stimulated the migration of FBHEs and BAEs but not HUVECs, BPAEs, BLMVECs, or BAMECs. Related phospholipids, lysophosphatidylcholine, lysophosphatidylserine, or sphingomyelin did not stimulate endothelial cell migration. Other cell types have been shown to migrate in response to LPA, including human Rat-1 fibroblasts,¹³ ß1A-GD25 fibroblasts,¹⁰ hepatoma cells,²⁴ breast carcinoma cells,^{17 25} and Dictyostelium discoideum slime mold.²⁶ However, LPA does not stimulate the migration of vascular smooth muscle cells or osteosarcoma cells (M. Magnusson and D.F. Mosher, unpublished data, 2000), suggesting that whether a cell migrates in response to LPA is also cell type specific.

Endothelial cell migration to lysophospholipids can be contrasted with VEGF- and bFGF-stimulated migration of endothelial cells in 3 ways: First, S1P and LPA are primarily chemokinetic, with a small chemotactic component, similar to bFGF but unlike VEGF, which is primarily chemotactic.²⁷ Second, S1P and LPA stimulate migration over a broad dose range (200-fold) that is distinct from the narrow dose response (10-fold) that results in the bell-shaped curve characteristic of VEGF or bFGF.²⁷ Third, the magnitude of migration is greater, with a >10-fold increase found with lysophospholipids compared with the few-fold change found with VEGF (herein) or bFGF.²⁷

As evidence that S1P and LPA are acting through G-protein–coupled receptors, we found that treatment with pertussis toxin ablates S1P- or LPA-stimulated endothelial cell migration. Pertussis toxin and the Mek-1 inhibitor PD98059 caused Erk1/2 activation by S1P or LPA to fall to undetectable levels; however, PD98059 had minimal effects on S1P- and LPA-stimulated endothelial cell migration. Therefore, Erk1/2 activation is not required for S1P- and LPA-stimulated endothelial cell migration, consistent with the previous report indicating that PD98059 does not affect LPA-stimulated migration of Rat-1 fibroblasts.¹³ Examination of the actin cytoskeleton and paxillin localization in pertussis toxin–treated cells suggests that inactivation of G_i uncovers a lysophospholipid-induced contractile phenotype (ie, cortical localization of actin and peripheral localization of paxillin) that is incompatible with cell migration. The phenotype is consistent with the contractile response of fibroblasts and osteosarcoma cells to LPA, which has been shown to be dependent on Rho activation.^{28 29} Nobes and Hall¹¹ have shown that microinjection of activated Rho does indeed prevent fibroblast cell migration in a wound assay. Treatment of endothelial cells with C3 toxin to inactivate Rho had a deleterious effect on cell migration that was probably due to disruption of the actin cytoskeleton. C3 toxin disrupted the actin cytoskeleton and resulted in loss of migration by the majority of the cells. C3 toxin has been shown to decrease hepatoma cell invasion in response to LPA²⁴ and closure of wounds in fibroblast monolayers.¹¹ Therefore, a balance between the activation of Rho, presumably by G_12/13,¹ and activation of G_i is important for endothelial cell migration stimulated by S1P and LPA.

Rac, a member of the Rho family, stimulates the formation of lamellipodia and may activate Rho.²⁹ PI3-kinase is an intermediate in the activation of Rac by Ras.²² T-lymphoma cells with activated Rac or activated Tiam1, a Rac guanine nucleotide exchange factor, require an additional signal from LPA or S1P to migrate.¹² We found that PI3-kinase inhibitors decreased lysophospholipid-stimulated endothelial cell migration by 50% to 70% and partially disrupted the actin cytoskeleton, with a minimal inhibition of Erk1/2 phosphorylation. The effect of PI3-kinase inhibitors may be explained by the ability of PI3-kinase to activate Ras downstream from G_i,²² of PI3-kinase to activate Rac downstream from Ras,²² and of integrin receptors to activate Rac through PI3-kinase.²³ Our data do not distinguish among these possibilities, although we have demonstrated that disruption of the actin cytoskeleton is the important end point, rather than disruption of Erk phosphorylation, for inhibition of cell migration.

Lysophospholipids activate numerous cell processes through the Edg family of 7 transmembrane G-protein–coupled receptors.^{1 5} The Edg family of receptors fall into 2 homology clusters, with the S1P homology cluster containing Edg-1, Edg-3, and Edg-5 and the LPA homology cluster containing Edg-2 and Edg-4.¹ Edg family members may be coupled to 3 heterotrimeric G proteins, G_i, G_12/13, and G_q.^{1 2} We found that downstream effectors of G_i and G_12/13 have an important role in endothelial cell migration. An important question is how each of the Edg receptors contributes to the effect. Reverse transcriptase–polymerase chain reaction with human primers in a bovine system indicates that at least 2 Edg receptor family members, Edg1 and Edg3, are present on FBHEs to mediate the biological response to S1P (O. Peyruchaud and T. Panetti, unpublished data, 1999). Thus, there likely are multiple readouts for Edg-1, Edg-3, and probably other Edg receptors through G_i, G_12/13, and G_q. A balance of these downstream signaling pathways may account for the different responses of different cells to S1P and LPA as migratory agents over such a wide concentration range.

Endothelial cell migration is a critical component of angiogenesis and wound repair. S1P and LPA are released from platelets at sites of blood coagulation.^{1 2} LPA increases vascular permeability of brain endothelial cells,³⁰ endothelial cell proliferation,¹⁹ and endothelial cell migration as described herein. It may be especially noteworthy that S1P stimulates endothelial cell migration in contrast to its inhibitory effect on platelet-derived growth factor–stimulated smooth muscle cell migration.¹⁵ In vivo, the variable responses of endothelial and smooth muscle cells may aid in tissue repair. For example, after balloon angioplasty, endothelial cell migration and repair of the denuded endothelium are necessary to limit further platelet activation and stimulation of smooth muscle cell proliferation and migration into the site of injury, leading to restenosis and occlusion of the vessel.³¹ Therefore, S1P may provide a way to stimulate endothelial cell migration and repair while limiting the influx of smooth muscle cells, even in the presence of platelet-derived growth factor.

	Acknowledgments

 
This study was supported by grants HL-21644 and HL-54462 from the National Institutes of Health (to Dr Mosher) and a grant from the Wisconsin Affiliate of the American Heart Association (to Dr Panetti). We thank Drs Qinghong Zhang and Olivier Peyruchaud for many helpful discussions. We would like to dedicate this article to the memory of Prof Russell Ross.

Received February 2, 1999; accepted November 4, 1999.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Goetzl EJ, An S. Diversity of cellular receptors and functions for the lysophospholipid growth factors lysophosphatidic acid and sphingosine 1-phosphate. FASEB J. 1998;12:1589–1598.[Abstract/Free Full Text]
2. Moolenaar WH, Kranenburg O, Postma FR, Zondag GCM. Lysophosphatidic acid: G-protein signalling and cellular responses. Curr Opin Cell Biol. 1997;9:168–173.[Medline] [Order article via Infotrieve]
3. Zhang Q, Peyruchaud O, French KJ, Magnusson MK, Mosher DF. Sphingosine 1-phosphate stimulates fibronectin matrix assembly through a rho-dependent signal pathway. Blood. 1999;93:2984–2990.[Abstract/Free Full Text]
4. Goetzl EJ, Kong Y, Mei B. Lysophosphatidic acid and sphingosine-1-phosphate protection of T cells from apoptosis in association with suppression of bax. J Immunol. 1999;162:2049–2056.[Abstract/Free Full Text]
5. Chun J, Contos JJA, Munroe D. A growing family of receptor genes for lysophosphatidic acid (LPA) and other lysophospholipids (LPs). Cell Biochem Biophys. 1999;30:213–242.[Medline] [Order article via Infotrieve]
6. Dillon ST, Feig LA. Purification and assay of recombinant C3 transferase. Methods Enzymol. 1995;256:174–184.[Medline] [Order article via Infotrieve]
7. Auerbach R, Auerbach W, Polakaowski I. Assays for angiogenesis: a review. Pharmacol Ther. 1991;51:1–11.[Medline] [Order article via Infotrieve]
8. Folkman J, Shing Y. Angiogenesis. J Biol Chem. 1992; 267: 10931–10934.
9. Pietruck F, Busch S, Virchow S, Brockmeyer N, Siffert W. Signalling properties of lysophosphatidic acid in primary human skin fibroblasts: role of pertussis toxin-sensitive GTP-binding proteins. Naunyn Schmiedebergs Arch Pharmacol. 1997;355:1–7.[Medline] [Order article via Infotrieve]
10. Sakai T, de la Pena JM, Mosher DF. Synergism among lysophosphatidic acid, ß1A integrins and EGF or PDGF in mediation of cell migration. J Biol Chem. 1999;274:15480–15486.[Abstract/Free Full Text]
11. Nobes CD, Hall A. Rho GTPases control polarity, protrusion, and adhesion during cell movement. J Cell Biol. 1999;144:1235–1244.[Abstract/Free Full Text]
12. Stam JC, Michiels F, van der Kammen RA, Moolenaar WH, Collard JG. Invasion of T-lymphoma cells: cooperation between Rho family GTPases and lysophospholipid receptor signaling. EMBO J. 1998;17:4066–4074.[Medline] [Order article via Infotrieve]
13. Anand-Apte B, Zetter BR, Viswanathan A, Qiu RG, Chen J, Ruggieri R, Symons M. Platelet-derived growth factor and fibronectin-stimulated migration are differentially regulated by the Rac and extracellular signal-related kinase pathways. J Biol Chem. 1997;272:30688–30692.[Abstract/Free Full Text]
14. Kundra V, Anand-Apte B, Feig LA, Zetter BR. PDGF-BB. Evidence of a role for Ras. J Cell Biol. 1995;130:725–731.[Abstract/Free Full Text]
15. Bornfeldt KE, Graves LM, Raines EW, Igarashi Y, Wayman G, Yamamura S, Yatomi Y, Sidhu JS, Krebs EG, Hakomori S, Ross R. Sphingosine-1-phosphate inhibits PDGF-induced chemotaxis of human arterial smooth muscle cells: spatial and temporal modulation of PDGF chemotactic signal transduction. J Cell Biol. 1995;130:193–206.[Abstract/Free Full Text]
16. Sadahira Y, Ruan F, Hakomori S, Igarashi Y. Sphingosine-1-phosphate, a specific endogenous signaling molecule controlling cell motility and tumor cell invasiveness. Proc Natl Acad Sci U S A. 1992;89:9686–9690.[Abstract/Free Full Text]
17. Wang F, Nohara K, Olivera A, Thompson EW, Spiegel S. Involvement of focal adhesion kinase in inhibition of motility of human breast cancer cells by sphingosine-1-phosphate. Exp Cell Res. 1999;247:17–28.[Medline] [Order article via Infotrieve]
18. Kawa S, Kimura S, Hakomori S, Igarashi Y. Inhibition of chemotactic motility and trans-endothelial migration of human neutrophils by sphingosine-1-phosphate. FEBS Lett. 1997;420:196–200.[Medline] [Order article via Infotrieve]
19. Panetti TS, Chen H, Misenheimer TM, Getzler SB, Mosher DF. Endothelial cell mitogenesis induced by LPA: inhibition by thrombospondins -1 and -2. J Lab Clin Med. 1997;129:208–216.[Medline] [Order article via Infotrieve]
20. Sakai T, Zhang Q, Fässler R, Mosher DF. Modulation of ß1A integrin functions by tyrosine residues in the ß1 cytoplasmic domain. J Cell Biol. 1998;141:527–538.[Abstract/Free Full Text]
21. Lauffenburger DA, Horwitz AF. Cell migration: a physically integrated molecular process. Cell. 1996;84:359–369.[Medline] [Order article via Infotrieve]
22. Khosravi-Far R, Campbell S, Rossman KL, Der CJ. Increasing complexity of Ras signal transduction: involvement of Rho family proteins. Adv Cancer Res. 1998;72:57–107.[Medline] [Order article via Infotrieve]
23. Keely PJ, Westwick JK, Whitehead IP, Der CJ, Parise LV. Cdc42 and Rac1 induce integrin-mediated cell motility and invasiveness through PI(3)K. Nature. 1997;390:632–636.[Medline] [Order article via Infotrieve]
24. Imamura F, Shinkai K, Mukai M, Yoshioka K, Komagome R, Iwasaki T, Akedo H. Rho-mediated protein tyrosine phosphorylation in lysophosphatidic-acid induced tumor-cell invasion. Int J Cancer. 1996;65:627–632.[Medline] [Order article via Infotrieve]
25. O’Connor KL, Shaw L, Mercurio AM. Release of cAMP gating by the 6ß4 integrin stimulates lamellae formation and the chemotactic migration of invasive carcinoma cells. J Cell Biol. 1998;143:1749–1760.[Abstract/Free Full Text]
26. Jalink K, Moolenaar WH, van Duijn B. Lysophosphatidic acid is a chemoattractant for Dictyostelium discoideum amoebae. Proc Natl Acad Sci U S A. 1993;90:1857–1861.[Abstract/Free Full Text]
27. Yoshida A, Anand-Apte B, Zetter BR. Differential endothelial migration and proliferation to basic fibroblast growth factor and vascular endothelial growth factor. Growth Factors. 1996;13:57–64.[Medline] [Order article via Infotrieve]
28. Zhang Q, Checovich WJ, Peters DM, Albrecht RM, Mosher DF. Modulation of cell surface fibronectin assembly sites by lysophosphatidic acid. J Cell Biol. 1994;127:1447–1459.[Abstract/Free Full Text]
29. Nobes CD, Hall A. Rho, rac, and cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell. 1995;81:53–62.[Medline] [Order article via Infotrieve]
30. Schulze C, Smales C, Rubin LL, Staddon JM. Lysophosphatidic acid increases tight junction permeability in cultured brain endothelial cells. J Neurochem. 1997;68:991–1000.[Medline] [Order article via Infotrieve]
31. Harker LA. Role of platelets and thrombosis in mechanisms of acute occlusion and restenosis after angioplasty. Am J Cardiol. 1987;60:20B–28B.[Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				I. Gorshkova, D. He, E. Berdyshev, P. Usatuyk, M. Burns, S. Kalari, Y. Zhao, S. Pendyala, J. G. N. Garcia, N. J. Pyne, et al. Protein Kinase C-{epsilon} Regulates Sphingosine 1-Phosphate-mediated Migration of Human Lung Endothelial Cells through Activation of Phospholipase D2, Protein Kinase C-{zeta}, and Rac1 J. Biol. Chem., April 25, 2008; 283(17): 11794 - 11806. [Abstract] [Full Text] [PDF]

				C. Avraamides, M. E. Bromberg, J. P. Gaughan, S. M. Thomas, A. Y. Tsygankov, and T. S. Panetti Hic-5 promotes endothelial cell migration to lysophosphatidic acid Am J Physiol Heart Circ Physiol, July 1, 2007; 293(1): H193 - H203. [Abstract] [Full Text] [PDF]

				S. R. Barthel, N. N. Jarjour, D. F. Mosher, and M. W. Johansson Dissection of the Hyperadhesive Phenotype of Airway Eosinophils in Asthma Am. J. Respir. Cell Mol. Biol., September 1, 2006; 35(3): 378 - 386. [Abstract] [Full Text] [PDF]

				S. Li, H. Tanaka, H. H. Wang, S. Yoshiyama, H. Kumagai, A. Nakamura, D. L. Brown, S. E. Thatcher, G. L. Wright, and K. Kohama Intracellular signal transduction for migration and actin remodeling in vascular smooth muscle cells after sphingosylphosphorylcholine stimulation Am J Physiol Heart Circ Physiol, September 1, 2006; 291(3): H1262 - H1272. [Abstract] [Full Text] [PDF]

				T. Clair, J. Aoki, E. Koh, R. W. Bandle, S. W. Nam, M. M. Ptaszynska, G. B. Mills, E. Schiffmann, L. A. Liotta, and M. L. Stracke Autotaxin Hydrolyzes Sphingosylphosphorylcholine to Produce the Regulator of Migration, Sphingosine-1-Phosphate Cancer Res., September 1, 2003; 63(17): 5446 - 5453. [Abstract] [Full Text] [PDF]

				K. L. Schaphorst, E. Chiang, K. N. Jacobs, A. Zaiman, V. Natarajan, F. Wigley, and J. G. N. Garcia Role of sphingosine-1 phosphate in the enhancement of endothelial barrier integrity by platelet-released products Am J Physiol Lung Cell Mol Physiol, July 1, 2003; 285(1): L258 - L267. [Abstract] [Full Text] [PDF]

				D. Shida, J. Kitayama, H. Yamaguchi, Y. Okaji, N. H. Tsuno, T. Watanabe, Y. Takuwa, and H. Nagawa Lysophosphatidic Acid (LPA) Enhances the Metastatic Potential of Human Colon Carcinoma DLD1 Cells through LPA1 Cancer Res., April 1, 2003; 63(7): 1706 - 1711. [Abstract] [Full Text] [PDF]

				I. Shiojima and K. Walsh Role of Akt Signaling in Vascular Homeostasis and Angiogenesis Circ. Res., June 28, 2002; 90(12): 1243 - 1250. [Abstract] [Full Text] [PDF]

				A. Endo, K.-I. Nagashima, H. Kurose, S. Mochizuki, M. Matsuda, and N. Mochizuki Sphingosine 1-Phosphate Induces Membrane Ruffling and Increases Motility of Human Umbilical Vein Endothelial Cells via Vascular Endothelial Growth Factor Receptor and CrkII J. Biol. Chem., June 21, 2002; 277(26): 23747 - 23754. [Abstract] [Full Text] [PDF]

				V. Vouret-Craviari, C. Bourcier, E. Boulter, and E. Van Obberghen-Schilling Distinct signals via Rho GTPases and Src drive shape changes by thrombin and sphingosine-1-phosphate in endothelial cells J. Cell Sci., June 15, 2002; 115(12): 2475 - 2484. [Abstract] [Full Text] [PDF]

				J. Lu, Y.-j. Xiao, L. M. Baudhuin, G. Hong, and Y. Xu Role of ether-linked lysophosphatidic acids in ovarian cancer cells J. Lipid Res., March 1, 2002; 43(3): 463 - 476. [Abstract] [Full Text] [PDF]

				T. Levade, N. Auge, R. J. Veldman, O. Cuvillier, A. Negre-Salvayre, and R. Salvayre Sphingolipid Mediators in Cardiovascular Cell Biology and Pathology Circ. Res., November 23, 2001; 89(11): 957 - 968. [Abstract] [Full Text] [PDF]

				F. Liu, A. D. Verin, P. Wang, R. Day, R. P. Wersto, F. J. Chrest, D. K. English, and J. G. N. Garcia Differential Regulation of Sphingosine-1-Phosphate- and VEGF-Induced Endothelial Cell Chemotaxis . Involvement of Gialpha 2-Linked Rho Kinase Activity Am. J. Respir. Cell Mol. Biol., June 1, 2001; 24(6): 711 - 719. [Abstract] [Full Text] [PDF]

				G. P. van Nieuw Amerongen and V. W.M. van Hinsbergh Cytoskeletal Effects of Rho-Like Small Guanine Nucleotide-Binding Proteins in the Vascular System Arterioscler. Thromb. Vasc. Biol., March 1, 2001; 21(3): 300 - 311. [Abstract] [Full Text] [PDF]

				D. English, J. G.N. Garcia, and D.N. Brindley Platelet-released phospholipids link haemostasis and angiogenesis Cardiovasc Res, February 16, 2001; 49(3): 588 - 599. [Abstract] [Full Text] [PDF]

				L. Balazs, J. Okolicany, M. Ferrebee, B. Tolley, and G. Tigyi Topical application of the phospholipid growth factor lysophosphatidic acid promotes wound healing in vivo Am J Physiol Regulatory Integrative Comp Physiol, February 1, 2001; 280(2): R466 - R472. [Abstract] [Full Text] [PDF]

				J. S. Alexander Rho, Tyrosine Kinase, Ca2+, and Junctions in Endothelial Hyperpermeability Circ. Res., August 18, 2000; 87(4): 268 - 271. [Full Text] [PDF]

				T. M. Misenheimer, K. G. Huwiler, D. S. Annis, and D. F. Mosher Physical Characterization of the Procollagen Module of Human Thrombospondin 1 Expressed in Insect Cells J. Biol. Chem., December 22, 2000; 275(52): 40938 - 40945. [Abstract] [Full Text] [PDF]

				J. H. Paik, S.-s. Chae, M.-J. Lee, S. Thangada, and T. Hla Sphingosine 1-Phosphate-induced Endothelial Cell Migration Requires the Expression of EDG-1 and EDG-3 Receptors and Rho-dependent Activation of alpha vbeta 3- and beta 1-containing Integrins J. Biol. Chem., April 6, 2001; 276(15): 11830 - 11837. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Data Supplement