AKAP9 is required for integrin-mediated adhesion, and Epac increases barrier properties through RGD-binding integrins at lateral cell borders. (A) Fluorescently labeled control siRNA, AKAP9 siRNA, or Epac1siRNA transfected cells were incubated on RGD-coated plates and cell adhesion was quantified by fluorometric analysis. Data, averaged from 3 experiments, are plotted relative to control siRNA (set at 100%). P < .005. Cells were stained with rhodamine-phalloidin to visualize actin. Actin was present in a cortical ring (arrow, Con^si and AKAP^si) or membrane edge (arrowhead, Epac^si). (B) Control and AKAP9 siRNA-treated cells were stained with antibody to β₁ (blue), α_v (green), or α₅ (red) and analyzed by flow cytometry; mean fluorescence intensities for each are in brackets. Data using isotype IgG are shown in black. One of 2 representative experiments is shown. (C-D) HUVECs were plated on fibronectin or collagen type IV. TER was evaluated in HUVECs treated with RGD or control (RGE) peptides (C) or functional blocking integrin antibodies (D) 1 hour before the addition of vehicle control or O-Me-cAMP (O-Me) at the time points indicated by the arrows. RGD treatment decreased basal permeability over time and reduced O-Me-cAMP-induced enhancement of barrier properties compared with RGE peptide treatment. α_v antibody attenuated O-Me-cAMP-induced effects on cells plated on fibronectin, while α₅ antibody blocked O-Me-cAMP effects on collagen. n = 3-5 experiments. (E) ECs plated on collagen (top panels) or fibronectin (bottom panels) were stimulated with vehicle control (Control) or O-Me-cAMP (O-Me) and stained with antibody to α₅ or α_v. α₅ and α_v were absent at lateral borders of cells plated on collagen, but α₅ appeared at cell-cell contacts (arrow) following O-Me-cAMP treatment. α_v was present at cell-cell contacts of cells plated on fibronectin (arrow) and increased there after O-Me-cAMP treatment (arrow), while α₅ remained absent even after O-Me-cAMP stimulation. One of 2 representative experiments is shown.