Role of EPAC/Rap in thrombin-induced permeability. (A) Pull-down assay for active Rap in HUVECs treated with thrombin (2 U/mL) for the times indicated or O-Me-cAMP for 30 minutes. Rap1 activation by thrombin was reproducibly observed at different time points ranging from 2 to 60 minutes (n = 6), a representative blot of which is shown. Densitometric analysis of experiments in which activation of Epac was included as a positive control was undertaken. Thrombin led to a 2- to 3-fold activation of Rap1 (P < .05 for all time points except 60 minutes). On the other hand, only a modest increase in Rap2 was observed under the same conditions (P > .05) (data not shown). (B) Permeability assay for HUVEC response to F + Ror O-Me-cAMP (100 μM) in the presence of thrombin (2 U/mL). Results are the normalized mean and standard error of 6 independent experiments (asterisks indicate P < .01 as compared to thrombin control). (C) Immunofluorescence of HUVECs treated with thrombin alone or in combination with F + R or O-Me-cAMP (100 μM) for 30 minutes. Cells were fixed and stained for actin with FITC-phalloidin (green) and VE-cadherin (red). Right column shows the overlay of actin and VE-cadherin. Thrombin-induced stress fibers and gaps in VE-cadherin (arrows), which were partially reversed by F + R or O-Me-cAMP. The number of gaps were quantitated in 3 independent experiments. Gaps were identified as significant discontinuities in VE-cadherin staining as indicated by the white arrows. The results, expressed as percent compared to thrombin (Thr)–treated cells, are as follows: Thr, 100%; Thr plus F + R, 22.2% ± 6.7%; Thr plus O-Me-cAMP, 37% ± 22%. (D) Pull-down assay for active RhoA (RhoA-GTP) in HUVECs. Rho assays were performed on endothelial monolayers that were stimultaneously treated with vehicle (–, ctrl), thrombin (2 U/mL) alone, thrombin plus O-Me-cAMP, or thrombin plus cAMP for 30 minutes. Results of densitometric analysis of 3 independent experiments are shown in right panel. Asterisk indicates P < .05.