Sphingosine-1-phosphate prevents permeability increases via activation of endothelial sphingosine-1-phosphate receptor 1 in rat venules

Abstract

Sphingosine-1-phosphate (S1P) has been demonstrated to enhance endothelial barrier function in vivo and in vitro. However, different S1P receptor subtypes have been indicated to play different or even opposing roles in the regulation of vascular barrier function. This study aims to differentiate the roles of endogenous endothelial S1P subtype receptors in the regulation of permeability in intact microvessels using specific receptor agonist and antagonists. Microvessel permeability was measured with hydraulic conductivity (L(p)) in individually perfused rat mesenteric venules. S1P-mediated changes in endothelial intracellular Ca(2+) concentration ([Ca(2+)](i)) was measured in fura-2-loaded venules. Confocal images of fluorescent immunostaining illustrated the spatial expressions of three S1P subtype receptors (S1P(R1-3)) in rat venules. The application of S1P (1 μM) in the presence of S1P(R1-3) inhibited platelet-activating factor- or bradykinin-induced permeability increase. This S1P effect was reversed only with a selective S1P(R1) antagonist, W-146, and was not affected by S1P(R2) or S1P(R3) antagonists JTE-013 and CAY-10444, respectively. S1P(R1) was also identified as the sole receptor responsible for S1P-mediated increases in endothelial [Ca(2+)](i). S1P(R2) or S1P(R3) antagonist alone affected neither basal L(p) nor platelet-activating factor-induced permeability increase. The selective S1P(R1) agonist, SEW-2871, showed similar [Ca(2+)](i) and permeability effect to that of S1P. These results indicate that, despite the presence of S1P(R1-3) in the intact venules, only the activation of endothelial S1P(R1) is responsible for the protective action of S1P on microvessel permeability and that endogenous S1P(R2) or S1P(R3) did not exhibit functional roles in the regulation of permeability under basal or acutely stimulated conditions.

Fig. 1.

Confocal images of fluorescent immunostaining of 3 sphingosine-1-phosphate (S1P) subtype receptors (S1P_R1, S1P_R2, and S1P_R3) in rat mesenteric venules. Top: S1P_R1 staining. Top, left: single image section with focus on the center of endothelium at lower half of the vessel wall. Top, middle: low fluorescence areas are the locations of endothelial nuclei that are colocalized with the nuclear dye DRAQ5 staining. The nuclear staining shown on both sides of the vessel wall (indicated by arrows) are the pericyte nuclei. Top, right: projection of the S1P_R1 staining from the lower half image stack of the vessel wall. The horizontally oriented fluorescence pattern represents the staining of pericytes. Bottom: projection of lower half image stack of the vessel wall with S1P_R2, S1P_R3, and the secondary antibody alone (control) staining, respectively. All the images are collected with ×63 objective with ×2 electronic zoom. Each image represents the staining pattern of multiple vessel segments from 3 rats.

Fig. 2.

VPC-23019, a S1P_R1 and S1P_R3 antagonist, abolished the protective role of S1P in platelet-activating factor (PAF)-induced increases in hydraulic conductivity (L_p). Perfusion of VPC-23019 (10 μM) alone and VPC-23019 plus S1P (1 μM) did not affect the baseline L_p but abolished the S1P effect on PAF-stimulated permeability increase. A: individual experiment shows time course of L_p changes with sequential perfusions of different reagents in the same vessel. B: summary results of 5 microvessels. *Significant increase from the control.

Fig. 3.

Inhibition of S1P_R1 abolished the protective role of S1P in PAF-induced L_p increases. When S1P_R1 was inhibited by the selective antagonist W-146 (10 μM), S1P (1 μM) showed no effect on baseline L_p and PAF-induced L_p increases. A: representative experiment shows the time course of the L_p changes. B: summary results of 5 experiments. *Significant increase from the control.

Fig. 4.

Selective inhibition of S1P_R3 with CAY-10444 (CAY) did not affect the protective role of S1P in PAF-induced L_p increase. A: an individual experiment shows the time course of the L_p changes. After inhibition of S1P_R3 by perfusion of the vessel with CAY (10 μM), S1P still effectively inhibited PAF-induced L_p increase. Removal of S1P and CAY from the vessel lumen with Ringer-albumin perfusion for 40 min recovered the PAF response in the same vessel. Ctrl, control. B: summary figure shows that S1P remains effective in inhibiting PAF-induced L_p increase after selective inhibition of S1P_R3 with CAY at 10 (n = 5) and 100 (n = 4) μM. Perfusion of vessels with CAY (10 μM) alone showed no effect on baseline L_p and PAF-induced L_p increases (n = 4). *Significant increase from the control.

Fig. 5.

Selective inhibition of S1P_R2 with JTE-013 (JTE) did not affect the protective role of S1P in PAF-perfused microvessels. A: representative experiment shows the time course of the L_p changes. S1P effectively inhibited PAF-induced L_p increase after preperfusion of the vessel with JTE (10 μM). The PAF response was recovered in the same vessel after Ringer-albumin perfusion for 40 min, indicating the vessel responsiveness to PAF. B: summary that shows that the effect of S1P on inhibition of PAF-induced L_p increases was not altered after inhibition of S1P_R2 with JTE at 10 (n = 6) and 30 (n = 6) μM. JTE (10 μM) alone also showed no effect on baseline L_p and PAF-induced L_p increases (n = 3). *Significant increase from the control.

Fig. 6.

S1P inhibited bradykinin (BK)-induced L_p increases. A: paired measurements of L_p in response to BK (1 nM) in the presence and absence of S1P (1 μM) in a single venule. In the presence of S1P, BK-induced L_p increase was completely abolished. After S1P and BK were washed from the vessel lumen with BSA-Ringer perfusion, the L_p response to BK was recovered with a peak value of 6.6 times that of the control. B: summary results compare BK-induced L_p increases in the absence (n = 6) and presence (n = 3) of S1P. *Significant increase from the control.

Fig. 7.

The activation of S1P_R1, not S1P_R2, is required for the inhibitory effect of S1P on BK-induced L_p increases. A: single experiment shows that addition of BK to S1P_R1 antagonist W-146 (10 μM)- and W-146 plus S1P-perfused vessel abolished the S1P effect and transiently increased L_p in a pattern similar to that of exposure to BK alone (Fig. 6). B: another experiment shows that blocking S1P_R2 by selective antagonist JTE did not affect the inhibitory effect of S1P on BK-induced L_p increase, and BK response was recovered in the same vessel when the preperfused agents were removed from the microvessel lumen with albumin-Ringer perfusion for 40 min. C: summary results show the effects of blocking S1P_R1 (n = 3) and S1P_R2 (n = 3) on S1P-mediated inhibition of BK-induced L_p increases. *Significant increase from the control.

Fig. 8.

SEW-2871 (SEW), a selective S1P_R1 agonist, attenuated the PAF-induced increases in L_p. A: single experiment shows that S1P_R1 agonist significantly attenuated PAF-induced L_p increases, an effect similar to that of S1P. The second application of PAF in the same vessel after perfusion of albumin-Ringer solution restored the PAF response. B: blocking S1P_R1 with selective S1P_R1 antagonist W-146 abolished the inhibitory effect of SEW on PAF-induced L_p increases, indicating that the effect of SEW is S1P_R1 specific. C: summary results show the effect of SEW on PAF-induced L_p increases with (n = 3, right) and without (n = 5, left) blocking S1P_R1 by antagonist W-146. *Significant increase from the control.

Fig. 9.

S1P-induced increase in endothelial intracellular Ca²⁺ concentration ([Ca²⁺]_i) is subtype receptor specific. A: time course of changes in endothelial [Ca²⁺]_i in an individual experiment. W-146 (10 μM), a selective S1P_R1 antagonist, had no effect on baseline endothelial [Ca²⁺]_i but abolished S1P-induced endothelial [Ca²⁺]_i increase. The S1P effect was recovered after 40 min perfusion of albumin-Ringer solution. B: superimposed time courses of 2 individual experiments showing that selective inhibition of S1P_R2 or S1P_R3 by antagonist JTE (10 μM) or CAY (10 μM), respectively, showed no effect on S1P-induced endothelial [Ca²⁺]_i increase. C: SEW, a selective S1P_R1 agonist, induced similar transient increases in endothelial [Ca²⁺]_i to that of S1P, further supporting the S1P_R1-mediated Ca²⁺ response. D: LaCl₃ (50 μM), a Ca²⁺ channel blocker, completely abolished S1P-induced endothelial [Ca²⁺]_i increase, and removal of LaCl₃ restored the S1P effect in the same vessel, indicating Ca²⁺ influx is the main source contributing to the S1P-induced increase in endothelial [Ca²⁺]_i. E: summary results show the changes in endothelial [Ca²⁺]_i in each group of 5 experiments. *Significant increase from the control.