RGD-dependent vacuolation and lumen formation observed during endothelial cell morphogenesis in three-dimensional fibrin matrices involves the alpha(v)beta(3) and alpha(5)beta(1) integrins

Abstract

Recent data have revealed the involvement of the alpha(v)beta(3) integrin in angiogenesis. However, few studies to date have provided a convincing role for this receptor in in vitro assays of endothelial cell morphogenesis where defined steps can be examined. Here, we present data showing that two integrins, alpha(v)beta(3) and alpha(5)beta(1), regulate human endothelial cell vacuolation and lumen formation in three-dimensional fibrin matrices. Cells resuspended in fibrin formed intracellular vacuoles that coalesced into lumenal structures. These morphogenic events were completely inhibited by the simultaneous addition of anti-alpha(v)beta(3) and anti-alpha(5) integrin antibodies. Complete blockade was also accomplished with a combination of the cyclic Arg-Gly-Asp (cRGD) peptide and anti-alpha(5) integrin antibodies. No blockade was observed with the control Arg-Gly-Glu (RGE) peptide alone or in combination with control antibodies. Finally, we were able to demonstrate regression of vacuoles and lumens several hours after the addition of cRGD peptides combined with anti-alpha(5) integrin antibodies. These effects were not observed with control peptides alone or in combination with control antibodies. We report here the novel involvement of both the alpha(v)beta(3) and alpha(5)beta(1) integrins in vacuolation and lumen formation in a fibrin matrix, implicating a role for multiple integrins in endothelial cell morphogenesis.

Figure 1.

Time course of endothelial cell morphogenesis in a three-dimensional fibrin matrix. ECs (25,000 per well) were resuspended in 25 μl of 10 mg/ml fibrinogen and added to 1 μg thrombin to solidify the matrix. Medium was added and photographs were taken under transmitted light of the same culture and at an identical focal plane at the various time points (0.5, 5, 7, 12, and 24 hours) listed on each figure (upper right). Arrowheads indicate markers present at all time points to assist in orientation. Scale bar, 100 μm.

Figure 2.

Endothelial cell lumen development in a fibrin matrix is regulated by intracellular vacuole formation. Fibrin gels were embedded in plastic, cross-sectioned, stained with toluidine blue, and photographed at the various time points indicated (0, 5, 10, and 24 hours). Scale bar, 20 μm.

Figure 3.

Time-course experiments demonstrating that the development of vacuoles and lumens by endothelial cells in fibrin matrices is an RGD- and α₅ integrin-dependent process. Cultures were prepared in the presence of either cRGD or RGE peptides (500 μg/ml) alone or combined with anti-α₅ or anti-α₂ integrin monoclonal antibodies (20 μg/ml). Control indicates the absence of peptides or antibodies. Cultures were fixed, stained, and quantitated at the various time points shown. Values represent the percentage of cells forming vacuoles and lumens. In each data set, four groups of 50 cells each were analyzed. Data shown are mean values ± SEM.

Figure 4.

Complete blockade of endothelial cell vacuole and lumen formation by a combination of cRGD peptides and anti-α₅ integrin antibodies. Cultures (cRGD/anti-α₅) were prepared in the presence of cRGD peptides (500 μg/ml) combined with anti-α₅ integrin monoclonal antibodies (20 μg/ml). Control indicates the absence of peptides or antibodies. Cultures were fixed at the time points shown (3, 5, 7, 12, and 24 hours), stained, and photographed. Scale bar, 50 μm.

Figure 5.

Endothelial cell vacuolation and morphogenesis are dependent on the α_vβ₃ and α₅β₁ integrins. Endothelial cells suspended in a fibrin matrix were allowed to undergo morphogenesis in the presence of various monoclonal anti-integrin antibodies (20 μg/ml) for 24 hours. Antibodies used included MAb16 (α₅), clone A2-IIE10 (α₂), clone LM609 (α_vβ₃), and clone P1F6 (α_vβ₅). Control indicates the absence of antibodies (None). Values represent the percentage of cells forming vacuoles and lumens. For each data set, four groups of 50 cells each were analyzed to determine the extent of vacuolation/lumen formation. Data shown are mean values ± SEM.

Figure 6.

Regression of preformed endothelial cell lumens follows the addition of α_vβ₃ and α₅β₁ integrin antagonists. Collapse of developed vacuolar and lumenal structures is initiated at both the 12- and 24-hour time points by adding cRGD peptides combined with anti-α₅ integrin antibodies. Endothelial cells suspended in a fibrin matrix were allowed to undergo morphogenesis for either a 12- or 24-hour period, at which time 500 μg/ml of the peptides indicated and/or 20 μg/ml of the various antibodies were added. Cultures were fixed 3 hours later and stained, and vacuolation was quantitated. Values represent the percentage of cells forming vacuoles and lumens. For each data set, four groups of 50 cells each were analyzed to determine the extent of vacuolation/lumen formation. Control indicates the absence of antibodies. Data shown are mean values ± SEM.

Figure 7.

Antagonists of α_vβ₃ and α₅β₁ induce regression of preformed endothelial cell vacuoles and lumens. Preformed lumenal structures (24 hours) were allowed to collapse for 3 hours in the presence of no antibodies or peptides (control), cRGD and anti-α₅ integrin antibodies (cRGD/anti-α₅), or RGE and anti-α₂ integrin antibodies (RGE/anti-α₂). The cultures were then fixed, stained, and photographed. Scale bar, 50 μm.