Transactivation of Sphingosine 1-Phosphate Receptors Is Essential for Vascular Barrier Regulation

The role for hyaluronan (HA) and CD44 in vascular barrier regulation is unknown. We examined high and low molecular weight HA (HMW-HA, ∼1,000 kDa; LMW-HA, ∼2.5 kDa) effects on human transendothelial monolayer electrical resistance (TER). HMW-HA increased TER, whereas LMW-HA induced biphasic TER changes ultimately resulting in EC barrier disruption. HMW-HA induced the association of the CD44s isoform with, and AKT-mediated phosphorylation of, the barrier-promoting sphingosine 1-phosphate receptor (S1P1) within caveolin-enriched lipid raft microdomains, whereas LMW-HA induced brief CD44s association with S1P1 followed by sustained association of the CD44v10 isoform with, and Src and ROCK 1/2-mediated phosphorylation of, the barrier-disrupting S1P3 receptor. HA-induced EC cytoskeletal reorganization and TER alterations were abolished by either disruption of lipid raft formation, CD44 blocking antibody or siRNA-mediated reductions in expression of CD44 isoforms. Silencing S1P1, AKT1, or Rac1 blocked the barrier enhancing effects of HA whereas silencing S1P3, Src, ROCK1/2, or RhoA blocked the barrier disruption induced by LMW-HA. In summary, HA regulates EC barrier function through novel differential CD44 isoform interaction with S1P receptors, S1P receptor transactivation, and RhoA/Rac1 signaling to the EC cytoskeleton.

Endothelial cells (EC) 2 constitute an inner lining of blood vessels to regulate the interface between the blood and the vessel wall including vascular barrier regulation, passive diffusion, and active transport of substances from the blood, regulation of vascular smooth muscle tone, and blood clotting (1,2). Disruption of this semi-selective cellular barrier is a critical feature of inflammation as well as an important contributing factor to atherosclerosis and tumor angiogenesis (3,4). A number of bioactive agonists contribute to EC barrier regulation via direct effects on the integrity of EC tight junctions, cell-cell, and cellmatrix adhesions. One important extracellular matrix component, hyaluronan (HA), and its cell surface receptor, CD44, has been implicated in normal EC function and angiogenesis (5,6).
Hyaluronan (HA) is a major glycosaminoglycan (GAG) component of the extracellular matrix of many tissues. Structurally, high molecular weight (HMW) HA (Ͼ500,000 daltons) is composed of repeating disaccharide units of D-glucuronic acid and N-acetylglucosamine, which exists as a random coil structure that can expand in aqueous solutions (6,7). Aqueous HA is highly viscous and elastic, properties which contribute to its space filling and filtering functions (7) and is synthesized by at least three hyaluronan synthases (HAS1, HAS2, and HAS3) (8). Studies utilizing HAS gene knock-out mice reveal that only HAS2 is required for viability with HAS2 deletion resulting in lethal abnormalities in cardiac development, which is rescued by addition of exogenous HA (9,10). Proinflammatory cytokines (TNF␣, IL-1␤) and LPS induce HA production in EC in vitro (11) and increased HA levels are observed in bronchoalveolar lavage fluid (BALF) from patients with inflammatory lung disorders such as pulmonary fibrosis, acute lung injury, and chronic obstructive pulmonary disease (12)(13)(14)(15). Further, intratracheal administration of nebulized high MW HA has been used to prevent injury in experimental emphysema (16).
HA is degraded by hyaluronidases, under certain pathological inflammatory conditions, to produce lower molecular weight fragments found in tissue injury and serum of patients with certain malignancies (17,18). Further, low MW fragments of HA (LMW, 1,350 -4,500 Da) are potent inducers of angiogenesis in vitro and in vivo (19,20). Six hyaluronidase genes encode Hyal-1,2,3,4, PHYAL1 (a pseudogene) and PH-20 with high MW HA and its fragments binding hyaladherin proteins including CD44, a major HA receptor (5,8).
CD44 belongs to a family of transmembrane glycoproteins, which are expressed in a variety of cells including EC (21,22). Multiple CD44 isoforms result from extensive, alternative exon splicing events (23,24) with the alternative splicing often occurring between exons 5 and 15 leading to a tandem insertion of one or more variant exons (v1-v10, or exon 6 through exon 14) within the membrane proximal region of the extracellular domain (25,26). The variable primary amino acid sequence of different CD44 isoforms is further modified by extensive N-and O-glycosylations and glycosaminoglycan (GAG) additions (5,26). The extracellular domain of CD44, containing clusters of conserved basic residues, plays an important role in HA binding, whereas the cytoplasmic domain is both structurally and functionally linked to cytoskeletal elements and signaling molecules (5,26). In particular, HA binding to CD44 isoforms can activate several downstream events including the PI 3-kinase/AKT pathway, the tyrosine kinase Src and the serine/threonine kinase, ROCK (22,27,28). The signaling properties of CD44 are required for a variety of cellular activities including EC adhesion, proliferation, migration, and angiogenesis (5, 20 -22, 26). The effects of HA and CD44 on human vascular barrier regulation, however, are unknown but are explored in the present study.
In this study, we examined the effects of low MW HA (ϳ2,500 Da) and high MW HA (ϳ1 million Da) on selective CD44 isoform-specific S1P receptor transactivation leading to EC barrier regulation. We further examined the role of caveolin-enriched microdomains (CEM) or lipid rafts and cytoskeletal regulatory GTPases (i.e. RhoA and Rac1) on HA-induced regulation of EC barrier integrity.
Preparation and Quantitation of Low and High MW HA-The method of preparation is similar to that described previously (19). For HMW-HA, 500 mg of rooster comb HA (ϳ1million Da polymers (34) was dissolved in distilled water and centrifuged in an Ultrafree-MC TM Millipore 100,000 Da MW cutoff filter and the flow-through (less than 100,000 Da) was discarded. For LMW-HA, 500 mg of rooster comb HA was digested with 20,000 units of bovine testicular hyaluronidase in digestion buffer (0.1 M sodium acetate, pH 5.4, 0.15 M NaCl) for 24 h, and the reaction stopped with 10% trichloroacetic acid. The resulting solution was centrifuged in an Ultrafree-MC TM Millipore 5,000 Da MW cutoff filter and the flow-through (less than 5,000 Da) was dialyzed against distilled water for 24 h at 4°C in 500-Da cutoff Spectra-Por tubing (Pierce-Warriner, Chester, UK). Low and High MW HA were quantitated using an ELISA-like competitive binding assay with a known amount of fixed HA and biotintylated HA-binding peptide (HABP) as the indicator (35). In some cases, both Low and High MW HA were subject to boiling, proteinase K (50 g/ml) digestion, hyaluronidase SD digestion (100 milliunits/ml) or addition of boiled (inactivated) hyaluronidase SD to test for possible protein/lipid contaminants (36). LMW and HMW-HA with DNA standards were run on 4 -20% SDS-PAGE gels and stained with combined Alcian blue and silver staining to further determine HA purity and size (37).
Lipid Raft Isolation-Caveolin-enriched microdomain known as lipid rafts were isolated from human lung EC as we previously described (29). Triton X-100-insoluble materials were mixed with 0.6 ml of cold 60% Optiprep TM and overlaid with 0.6 ml of 40 -20% Optiprep TM and the gradients centrifuged (35,000 rpm) in SW60 rotor for 12 h at 4°C and different fractions were collected and analyzed. In some cases, different fractions were were analyzed for total cholesterol content using the Amplex Red TM cholesterol assay kit (Invitrogen (Molecular Probes).
Total RNA Isolation-Total RNA was isolated using Trizol LS (Invitrogen) followed by RNeasy column (Qiagen, Valencia, CA) for further purification.

HA/CD44/S1P Receptor Signaling in Human EC
Determination of Tyrosine/Serine/Threonine Phosphorylation of the S1P 1 and S1P 3 Receptor-Solubilized proteins in IP buffer B (see above) were immunoprecipitated with either rabbit anti-S1P 1 receptor or mouse anti-S1P 3 receptor antibody followed by SDS-PAGE in 4 -15% polyacrylamide gels and transfer onto Immobilon TM membranes (Millipore Corp.). After blocking nonspecific sites with 5% bovine serum albumin, the blots were incubated with either rabbit anti-S1P 1 antibody, mouse anti-S1P 3 antibody, mouse anti-phosphotyrosine, rabbit anti-phosphoserine antibody, or rabbit anti-phosphothreonine antibody followed by incubation with horseradish peroxidaselabeled goat anti-rabbit or goat anti-mouse IgG. Visualization of immunoreactive bands was achieved using enhanced chemiluminescence (Amersham Biosciences).
Rho Family Activation Assay-RhoA and Rac activities in human lung EC were performed as described previously (38).  Measurement of TransEC Electrical Resistance (TER)-EC were grown to confluence in polycarbonate wells containing evaporated gold microelectrodes, and TER measurements performed using an electrical cell substrate impedance sensing system (Applied Biophysics) as previously described (4). TER values from each microelectrode were pooled at discrete time points and plotted versus time as the mean Ϯ S.E.
In Vitro S1P Receptor Phosphorylation-The S1P receptor phosphorylation reaction was carried out in 50 l of the reaction mixture containing 40 mM Tris-HCl (pH 7.5), 2 mM EDTA, 1 mM dithiothreitol, 7 mM MgCl 2 , 0.1% CHAPS, 0.1 M calyculin A, 100 M ATP, purified enzymes (i.e. 100 ng of recombinant active Src, ROCK1 or ROCK2) with or without immunoprecipitated S1P 1 or S1P 3 receptor obtained from human pulmonary EC that were serum-starved for 1 h. After incubation for 30 min at 30°C, the reaction mixtures were boiled in SDS sample buffer and subjected to SDS-PAGE. Immunoblots were performed using mouse anti-phosphotyrosine, rabbit anti-phosphoserine, rabbit anti-phosphothreonine, rabbit anti-S1P 1 , or mouse anti-S1P 3 antibody followed by incubation with horseradish peroxidase-labeled goat anti-rabbit or goat anti-mouse IgG. Visualization of immunoreactive bands was achieved using enhanced chemiluminescence.
Immunofluorescence Microscopy and Cortical Actin Quantitation-Polymerized actin rearrangement was assessed with Texas Red-conjugated phalloidin and analyzed using a Nikon Eclipse TE 300 microscope as we have described (4). Computer-recorded tiff images were analyzed with ImageQuant TM software from Amersham Biosciences. A standardized average gray value (SAGV) was generated for total phalloidin staining versus cortical phalloidin staining for each cell (29). To calculate percent cortical actin staining, the following equation was used: ((cortical actin SAGV ϫ area) divided by (total actin SAGV ϫ area)) ϫ 100. At least ten cells per sample were analyzed. Experiments were performed in triplicate.
Statistical Analysis-Student's t test was used to compare the means of data from two or more different experimental groups. Results are expressed as means Ϯ S.E.  1 (B, panel h) antibody. Experiments were performed in triplicate with highly reproducible findings (representative data shown). C, in vitro S1P receptor phosphorylation reaction was carried out in 50 l of the reaction mixture containing 40 mM Tris-HCl (pH 7.5), 2 mM EDTA, 1 mM dithiothreitol, 7 mM MgCl 2 , 0.1% CHAPS, 0.1 M calyculin A, 100 M ATP, purified enzymes (i.e. 100 ng of recombinant active Src, ROCK1, or ROCK2) with or without immunoprecipitated S1P 1 or S1P 3 receptor obtained from human pulmonary EC that were serum-starved for 1 h. After incubation for 30 min at 30°C, the reaction mixtures were boiled in SDS sample buffer and subjected to SDS-PAGE. Immunoblots were performed using anti-phosphotyrosine (C, panels a and e), anti-phosphoserine (C, panels b and f), anti-phosphothreonine (C, panels c and g), anti-S1P 1 (C, panel d), or anti-S1P 3 (C, panel h) antibody.

Divergent Effects of Low and High Molecular Weight Hyaluronan on Human Lung Endothelial Cell Barrier Function: Role of Caveolin-enriched Microdomains (Lipid Rafts)-Initial
experiments examined the effects of low and high MW-HA on human lung EC barrier function and the role of CD44 and lipid rafts in this process. Lipid rafts isolated from human lung EC contain specific markers (caveolin-1 and flotillin-1), are enriched in cholesterol and exclude other subcellular organelle markers (Fig. 1, A and B). These results demonstrate the purity and specificity of our lipid raft isolation procedure. Next, RT-PCR and isoform-specific immunoblot analysis were performed to explore whether CD44 isoforms, a major cell surface HA receptor family, were present in human lung EC. Fig. 1, C and D demonstrate that human pulmonary EC express at least two major CD44 isoforms, CD44s (standard form, ϳ85 kDa) and CD44v10 (ϳ116 kDa).
HMW-HA (ϳ1 million Da) consistently produced a gradual and sustained rise in transmonolayer electrical resistance (TER) in dose-dependent fashion whereas LMW-HA (ϳ2,500 Da) induced biphasic changes in TER with an initial rapid increase in barrier enhancement followed by significant and prolonged barrier disruption (Fig. 2, A and B). The dose response was significant when comparing equal nanomolar concentrations (but not equal concentrations in the range of 1.0 -100 g/ml (Fig. 2C)) of Low and High MW HA. Depleting cholesterol with methyl-␤-cyclodextrin (M␤CD) treatment (Fig. 2D) or using a pan-CD44 blocking antibody, which blocks HA binding to all CD44 isoforms, abolished both HMW-and LMW-HA-induced changes in TER (Fig. 2E). These results demonstrate that cholesterol-enriched microdomains regulate HA-mediated EC barrier function. Further, CD44 is the major HA receptor responsible for HA-mediated EC barrier alterations.
We previously demonstrated that CD44 localizes in activated EC to specialized cholesterol-and caveolin-enriched lipid rafts, plasma membrane microdomains implicated in a variety of cellular functions including potocytosis, cholesterol, and calcium regulation as well as signal transduction (21,22,39). Lipid rafts are biochemically defined by insolubility in 4°C Triton X-100 and light buoyant density after discontinuous gradient centrifugation (40). Both HMW-HA and LMW-HA rapidly (5 min) recruit CD44s to the lipid raft fraction whereas LMW-HA promotes robust but delayed recruitment of CD44v10 (after 15 min) (Fig. 3A).  panels a and d), anti-S1P 1 receptor (B, panel b), anti-phosphotyrosine (B, panel c), or anti-S1P 3 receptor (B, panel e) antibody. Experiments were performed in triplicate with highly reproducible findings (representative data shown). C, graphical representation of normalized resistance (TER) with scramble siRNA (siRNA that does not target any known human mRNA), Src siRNA, AKT1 siRNA, ROCK1 siRNA, or ROCK2 siRNA treatment of EC. EC were plated on gold microelectrodes and treated with scramble siRNA (siRNA that does not target any known human mRNA), Src siRNA, AKT1 siRNA, ROCK1 siRNA, or ROCK2 siRNA for 48 h. EC were then serum-starved for 1 h followed by either no treatment (scramble control) or addition of 100 nM High or Low MW HA. The bar graphs represent pooled TER data Ϯ S.E. at 1 h after agonist addition from three independent experiments as described under "Experimental Procedures." Asterisks indicate significant abolition of the High HA-inducible induction or Low HA-inducible repression of TER, respectively.

HA/CD44/S1P Receptor Signaling in Human EC
Transactivation of S1P Receptors Are Involved in HA-mediated Lung Vascular Barrier Regulation in a CD44 Isoform-specific Manner-We next explored whether HA induces physical and/or functional associations between CD44 and S1P receptors, which may be involved in HA-mediated vascular barrier responses. HMW-HA (100 nM) induced CD44s association in lipid rafts with S1P 1 , the known barrier-promoting S1P receptor (Fig. 3B). In contrast, LMW-HA initially recruited the S1P 1 receptor followed by recruitment of S1P 3 receptors. Immunoprecipitation followed by immunoblotting from lipid raft fractions revealed that HMW-HA promotes S1P 1 receptor association with CD44s. In contrast, LMW-HA (100 nM) induced an initial CD44s association with S1P 1 followed by CD44v10 association with S1P 3 receptor in lipid raft fractions. Both the spatially specific actin cytoskeletal reorganization and TER alterations evoked by HMW-HA and LMW-HA were abolished by either M␤CD (to inhibit lipid raft formation), by anti-CD44 blocking antibody or by siRNAs specific for CD44 (Figs. 2 and 4, Table 1). Silencing S1P 1 receptor blocked the EC barrier enhancing effects of High MW HA while silencing S1P 3 receptor blocked the EC barrier disruptive effects of Low MW HA (Fig. 4). Consistent with HA-mediated S1P transactivation, HMW-HA promoted AKT1-mediated threonine phosphorylation of S1P 1 receptor whereas LMW-HA induced sequential AKT1-mediated S1P 1 and Src/ROCK1/2-mediated S1P 3 receptor phosphorylation/activation (Figs. 5 and 6). These results were confirmed by using in vitro phosphorylation of S1P receptors with recombinant AKT1, Src, ROCK1 and ROCK2 (Fig.  5C). Further, silencing AKT1 expression blocks HWM-HAmediated EC barrier enhancement while silencing Src or both ROCK 1 and 2 expression blocks LMW-HA-mediated EC barrier disruption (Fig. 6C). Thus, low and high MW HA promote differential CD44 isoform-specific association with and activation of S1P receptors in lipid rafts. Activation of S1P 1 receptor is required for HA-induced EC barrier enhancement while S1P 3 receptor activation promotes barrier disruption.
Role of RhoA and Rac1 Signaling on HA-induced EC Barrier Function-We have previously demonstrated that the Rho family GTPase, Rac1, regulates S1P-mediated EC barrier enhancement (4). We examined whether Rho family GTPases could play a role in the HA-specific regulatory responses and identified that either LMW-HA (5 min.) or HMW-HA (5, 15, 30 min) induced Rac1 activation in concert with recruitment of the Rac1-specific exchange factor, Tiam1, to EC lipid rafts (Fig.  7). Rac1 activation was inhibited by siRNA for S1P 1 (but not S1P 3 ) to reduce receptor expression. The HA-induced EC barrier enhancement was inhibited by silencing Rac1 (but not RhoA) expression. In contrast, LMW-HA (but not HMW-HA) recruited the RhoA exchange factor, p115 Rho-GEF, to EC lipid rafts at 15-30 min. and promoted RhoA activation. LMW-HA-induced RhoA activation was inhibited by siRNA for S1P 3 (but not S1P 1 ) and LMW-HA-induced EC barrier disruption was inhibited by silencing RhoA (but not Rac1) expression.
Finally, silencing either S1P 1 or Rac1 expression attenuated EC barrier-enhancing effects of HMW-HA and LMW-HA whereas silencing S1P 3 or RhoA expression diminished the EC barrier-disruptive response to LMW-HA (Figs. 4 and 7). These results suggest that HA promotes cytoskeletal reorganization and EC barrier regulation through differential CD44 isoform interaction with S1P receptors via RhoA/Rac1 signaling in lipid rafts. Transactivation of S1P 1 receptor may represent a common mechanism for receptor-mediated vascular barrier regulation.
HA-induced, CD44, and S1P Receptor-dependent, Cytoskeletal Reorganization in EC-TER measurements of EC barrier function in vitro revealed that reduction in expression of either S1P 1 or Rac1 attenuated the barrier-enhancing effects of low and high MW HA; whereas reduction in S1P 3 or RhoA expression attenuated the delayed barrier-disruptive response to low MW HA on EC (Figs. 4 and 7). As cytoskeletal reorganization is a fundamental element of virtually all EC barrier-regulatory responses, we compared phalloidin staining of HA-and S1Pchallenged EC to visualize cellular F-actin localization ( Fig. 8 and Table 1). At early time points (5 min) both low and high MW HA induced prominent cortical actin ring formation, which was attenuated by reduction of S1P 1 (but not S1P 3 ), AKT, or Rac1 (but not RhoA) expression, findings similar to that reported for S1P (4,29). Low MW HA challenge for 30 min., however, resulted in a loss of cortical actin staining with increased F-actin stress fiber formation, which was significantly attenuated by silencing S1P 3 (but not S1P 1 ) or RhoA (but not Rac1) expression. Role of S1P 1 Receptor as a Central Regulator of EC Permeability-We recently demonstrated that the S1P 1 receptor regulates activated protein C (APC)/endothelial cell protein C receptor (EPCR)-mediated EC barrier protection against edemagenic agents such as thrombin (41). As silencing the S1P 1 receptor reduces the barrier enhancement induced by HA, and both LMW-HA and HMW-HA promote transactivation of S1P 1 receptor during the EC barrier-enhancing stages of these agonists, we next explored whether S1P 1 receptor serves as a central regulator of EC barrier function (Fig. 4D). Reductions in S1P 1 receptor expression significantly modulated the barrierregulatory effects of human lung EC challenged with HGF, PDGF, VEGF, or ATP (4,42). In contrast, thrombin, a known EC barrier-disruptive agent, was unaffected by S1P 1 receptor silencing, suggesting that the S1P 1 receptor serves as a critical and central regulator of EC barrier function.

DISCUSSION
Agents that exhibit the capacity to restore barrier integrity after periods of increased vascular permeability have obvious therapeutic applications in diverse inflammatory syndromes as well as in conditions such as tumor angiogenesis and atherosclerosis (4,42). We explored the effects of HA, a major glycosaminoglycan, which exists in multiple MW forms (6,43), on EC permeability, We found that the high MW form (ϳ1,000,000 Da) promotes increased EC barrier integrity in vitro and propose that high MW HA plays an important role in providing a protective barrier between endothelial cells and underlying vasculature in vivo. In contrast, hyaluronan fragments of ϳ2,500 Da (low MW HA), previously shown to be angiogenic (19), induces a biphasic effect on EC permeability with a brief, barrier-enhancing phase followed by a prolonged barrier-disruptive phase (Fig. 9).
CD44 is highly likely to be important in lung disease as CD44 Ϫ/Ϫ mice develop lung fibrosis, inflammatory cell recruitment, and hyaluronan fragment accumulation at sites of lung injury (44). Both high MW HA and its fragments bind to CD44, however, high and low MW HA evoke highly specific cellular functions. The high MW HA induces CD44s-mediated transactivation (AKT-dependent threonine phosphorylation) of S1P 1 receptor and consequent Rac1 signaling leading to cortical actin thickening and barrier enhancement in human pulmonary EC. Silencing S1P 1 receptor, AKT1 or Rac1 reverses the barrier-protective effects of high MW HA. In contrast, low MW HA promotes CD44v10-mediated transactivation (Srcdependent tyrosine phosphorylation and ROCK1/2-dependent threonine phosphorylation) of S1P 3 receptor and RhoA signaling leading to EC barrier disruption. Silencing S1P 3 receptor, Src, ROCK1/2, or RhoA reverses the barrier disruptive effects of low MW HA. We previously demonstrated that CD44v10 promotes RhoA activation with consequent Rho kinase (ROCK) activity (22) which regulates Ca 2ϩ signaling and EC  migration. In the present study, low MW HA was found to be a potent inducer of CD44v10 signaling, S1P 3 receptor activation and RhoA-mediated EC barrier disruption. The effects of Ca 2ϩ signaling on HA-induced EC barrier regulation are currently under investigation.
Caveolin-enriched microdomains or lipid rafts, are important plasma membrane microdomains that regulate numerous EC functions (45,46). CD44 localization in lipid rafts is important for HA-mediated signaling (21,30) as cholesterol depletion blocks both low and high MW HA-induced EC barrier responses. Further, CD44 isoform-specific activation of S1P receptors occurs in lipid rafts indicating that these microdomains play an important role in HA-induced EC functions. We have previously demonstrated that PI3 kinase and Rac1 signaling from lipid rafts are important for S1P 1 receptor-mediated EC barrier enhancement (29). The ability to potentially target these microdomains as a means of drug delivery for edemagenic states has significant promise.
HA/CD44-mediated stimulation of Src (pp60Src, c-Src tyrosine kinase) activity has been shown to regulate cytoskeletal function (28). In agreement with our results, researchers have reported that activation of Src promotes cytoskeletalmediated EC barrier disruption (47)(48)(49). In particular, Src regulates EC contraction and vascular permeability (48). Inhibition of Src reduces edema and stabilizes a VEGF receptor 2/cadherin complex after myocardial infarction (50). The role of Src activation on receptor tyrosine kinases and adhesion proteins in EC barrier function are currently being investigated in our laboratory.
Actin cytoskeletal reorganization plays a key role in EC barrier regulatory responses to a variety of agents (3,4). We observed that HA promotes cytoskeletal reorganization and EC barrier regulation via differential CD44 isoform interaction with S1P receptors and RhoA/Rac1 signaling in lipid rafts. In particular, high MW HA induces cortical actin ring formation while low MW HA treatment of EC for 15 or 30 min promotes actin stress fiber formation. These results demonstrate the requirement for S1P 1 receptor transactivation in agonist-induced EC barrier enhancement, and the potential for S1P 1 activation to represent a common mechanism for receptor-mediated vascular barrier regulation.