Activation of RhoA, but Not Rac1, Mediates Early Stages of S1P-Induced Endothelial Barrier Enhancement

Abstract

Compromised endothelial barrier function is a hallmark of inflammation. Rho family GTPases are critical in regulating endothelial barrier function, yet their precise roles, particularly in sphingosine-1-phosphate (S1P)-induced endothelial barrier enhancement, remain elusive. Confluent cultures of human umbilical vein endothelial cells (HUVEC) or human dermal microvascular endothelial cells (HDMEC) were used to model the endothelial barrier. Barrier function was assessed by determining the transendothelial electrical resistance (TER) using an electrical cell-substrate impedance sensor (ECIS). The roles of Rac1 and RhoA were tested in S1P-induced barrier enhancement. The results show that pharmacologic inhibition of Rac1 with Z62954982 failed to block S1P-induced barrier enhancement. Likewise, expression of a dominant negative form of Rac1, or knockdown of native Rac1 with siRNA, failed to block S1P-induced elevations in TER. In contrast, blockade of RhoA with the combination of the inhibitors Rhosin and Y16 significantly reduced S1P-induced increases in TER. Assessment of RhoA activation in real time using a fluorescence resonance energy transfer (FRET) biosensor showed that S1P increased RhoA activation primarily at the edges of cells, near junctions. This was complemented by myosin light chain-2 phosphorylation at cell edges, and increased F-actin and vinculin near intercellular junctions, which could all be blocked with pharmacologic inhibition of RhoA. The results suggest that S1P causes activation of RhoA at the cell periphery, stimulating local activation of the actin cytoskeleton and focal adhesions, and resulting in endothelial barrier enhancement. S1P-induced Rac1 activation, however, does not appear to have a significant role in this process.

Fig 1. Pharmacological inhibition or siRNA-mediated knockdown of Rac1 impaired baseline endothelial barrier integrity.

A. Treatment with the selective Rac1 inhibitor Z62954982 reduces TER in a concentration-dependent manner in HDMEC monolayers. B. Comparison of the mean maximum decreases in TER in the 30-min time window for each concentration of Z62954982 in HDMEC monolayers. Panels C & D show that Z62954982 produces a similar concentration-dependent TER in HUVEC monolayers. E. Western blot confirming knockdown (KD) with Rac1-specific siRNA, compared to sham and scrambled RNA (Scr) control groups. Bands for β-actin from re-probed blots confirmed equivalent loading of protein for each lane. F. Mean baseline TER values of HDMEC and HUVEC monolayers in Rac1 knockdown, scrambled control, and sham-transfected groups. *P<0.05 versus vehicle treated group. †P<0.05 versus other concentrations.

Fig 2. S1P caused initial endothelial barrier enhancement is concentration-dependent manner.

A. Maximal change in TER (%) from baseline within the initial 10-min time window after addition of the indicated concentrations of S1P. *P<0.05 versus vehicle treated group. B. Comparison of representative tracings of TER of HUVEC monolayers treated with 1 μM or 10 μM S1P. After the initial increase in TER, with higher concentrations of S1P (10 μM shown here), this elevated TER is typically short-lived and often decreases to a level below the initial baseline.

Fig 3. Pharmacologic inhibition of Rac1 failed to block S1P-induced endothelial barrier enhancement in HUVEC and HDMEC monolayers.

A. The time course of changes in of TER of HUVEC monolayers pretreated with the 30 min with Rac1 inhibitor Z62954982 or vehicle control, followed by treatment with 2 μM S1P are shown (N = 8 for each group). B. Mean maximal change in TER (%) of HUVEC monolayers after S1P treatment within the first 10-min window. Panels C & D show corresponding results for HDMEC monolayers (N = 8 each group). *P<0.05 vs. S1P Vehicle pretreated groups.

Fig 4. Overexpression of WT or DN Rac1 in HUVEC or HDMEC did not alter S1P-induced endothelial barrier enhancement.

A. Time course of TER of HUVEC monolayers transfected with WT and DN Rac1 plasmids, treated with either 2 μM S1P or vehicle (N = 8 each group). The TER is normalized to the time point just prior to addition of S1P for a more direct comparison of the magnitude of the response. B. The mean maximal change in TER (%) of HUVEC monolayers in the 10-min window immediately following S1P treatment. The corresponding data for HDMEC monolayers are shown in panels C & D (N = 8 each group). *P<0.05 vs. Vehicle treated groups.

Fig 5. Knockdown of Rac1 expression with siRNA did not diminish S1P-induced barrier enhancement of HUVEC or HDMEC monolayers.

A. Time course of changes in TER of HUVEC monolayers before and after treatment with 2 μM S1P or vehicle control, for the Rac1 knockdown and scrambled RNA transfected groups (N = 8 each group). The TER is normalized to the time point just prior to the addition of S1P, for more direct comparisons of the responses to S1P between the groups. B. The mean maximal change in TER of HUVEC monolayers (%) during the first 10 min after S1P was added. The corresponding results for HDMEC monolayers are shown in panels C & D (N = 8 each group). *P<0.05 vs. vehicle control groups; †P<0.05 vs. scrambled RNA sequence group.

Fig 6. Inhibition of RhoA attenuated S1P-induced barrier enhancement of endothelial monolayers.

A. The time course of changes in of TER of HUVEC monolayers pretreated with the 30 min with the combination of Rhosin and Y16 (5 μM of each) or vehicle control, followed by treatment with 2 μM S1P or vehicle (N = 8 for each group). B. Comparison of the mean maximal changes in TER of HUVEC monolayers (%) within the first 10 min after S1P or vehicle. The corresponding results for HDMEC monolayers are shown in panels C & D (N = 8 each group). *P<0.05, S1P vs. vehicle treated group. †P<0.05, inhibitor vs. vehicle pretreatments.

Fig 7. S1P activated RhoA primarily at cell periphery.

A. Representative images of HUVEC expressing the pTriEx-RhoA FLARE.sc Biosensor, showing the CFP and YFP channels, and the ratio (FRET) indicating RhoA activation, during baseline and after the treatment of 2 μM S1P. The entire time course can be viewed in S1 Movie. B. Normalized mean intensity of the whole cell before and after S1P treatment (N = 5 cells studied). *P<0.05, before vs. after S1P treatment.

Fig 8. Inhibition of RhoA abrogated S1P-induced phosphorylation of MLC-2 on Thr-18/Ser-19 that is primarily near cell borders.

A. Z-projection confocal immunofluorescence microscopy images of phosphorylated MLC-2 on HUVEC monolayers are shown. Each image represents three replicates for each time point. S1P was applied at 2 μM. Combined Rhosin and Y16 pretreatment was for 30 min, at 5 μM each. B. Quantification of phosphorylated MLC-2 intensity for each time point. *P<0.05, S1P treatment compared with baseline.

Fig 9. RhoA inhibition abrogated S1P-induced F-actin formation and recruitment of vinculin near the cell periphery.

The results showed that S1P increases F-actin and vinculin labeling in the peripheral areas of cells (10 min after the addition of S1P). This was inhibited after pretreatment with combined Rhosin and Y-16 (Inh; 5 μM each, 30 min). The inhibitors alone had no impact. All images are representative of 3 separate experiments.