Opposing activities of Sema3A and VEGF₁₆₅ on endothelial-cell lamellipodia spreading and retraction, and extracellular matrix–dependent cord formation. (A) Representative images depicting Sema3A-induced inhibition of lamellipodia spreading in endothelial cells. Endothelial cells were incubated for 30 minutes on poly-l-lysine–coated glass slides in medium alone or with Sema3A (2 μg/mL). Cells were visualized as described for Figure 1A, except for the use of a 10 ×/0.25 PhC objective lens. Original magnification, × 100. Arrows point to cells with lamellipodia. (B) Dose dependency of Sema3A-induced lamellipodia spreading inhibition. Results reflect the mean ratio (± SD triplicate wells) of cells with lamellipodia spreading/total number of attached cells. Representative of 3 experiments. (C) Effects of VEGF₁₆₅ and VEGF₁₂₁ on inhibition of lamellipodia spreading by Sema3A. Endothelial cells were incubated in medium alone, with Sema3A (0.5 μg/mL) alone, or with Sema3A (0.5 μg/mL) plus VEGF₁₆₅ (25 ng/mL) or VEGF₁₂₁ (25 ng/mL). The results reflect the mean (± SD of triplicate determinations). (D) Representative images depicting Sema3A-induced retraction of lamellipodia. After endothelial cells were allowed to attach on fibronectin-coated slides for 16 hours and nonadherent cells were removed, the adherent cells were further incubated (1 hour) in medium only or with Sema3A (2 μg/mL). Retraction scores: score 0 indicates cells have intact lamellipodia around; score 1, one side of the lamellipodia is retracted; score 2, 2 sides are retracted; and score 3, 3 sides are retracted or the cell has shrunk and is round. Arrows point to the retracted sides. Cells were visualized as described for Figure 1A, except for the use of a 10 ×/0.25 PhC objective lens. Original magnification: top panels, × 40; bottom panels, × 100 (bottom images enlarged with Photoshop [Adobe Systems, San Jose, CA]). (E) Concentration and (F) time dependency of Sema3A-induced lamellipodia retraction in endothelial cells. Adherent cells were incubated in medium alone or with Sema3A for 1 hour or for the indicated times. The results reflect the average retraction scores (± SD of triplicate determinations). (G) VEGF₁₆₅ blocks lamellipodia retraction induced by Sema3A, but VEGF₁₂₁ and FGF-2 do not. Adherent cells were incubated (1 hour) with Sema3A (2 μg/mL) alone or in the presence of VEGF₁₆₅, VEGF₁₂₁, or FGF-2. (H) Sema3A inhibits extracellular matrix–dependent cord formation by endothelial cells. Cells were incubated (16 hours) onto Matrigel-coated wells in medium alone, with Sema3A, VEGF₁₆₅, VEGF₁₂₁, or with Sema3A plus VEGF₁₆₅ or VEGF₁₂₁. Original magnification, × 40. (I) VEGF₁₆₅ reconstitutes cord formation disrupted by Sema3A. Endothelial cells were cultured (16 hours) on Matrigel-coated wells in medium only, with Sema3A, VEGF₁₆₅, with VEGF₁₂₁, or with Sema3A plus VEGF₁₆₅ or VEGF₁₂₁. The results reflect the average number of branch points formed by intersecting endothelial cords (mean ± SD of triplicate wells). Representative experiment of 4 performed. *P < .05; **P < .01. For all micrographic images, photography, acquisition, and processing were performed as described for Figure 1A.