Paracrine Stimulation of Capillary Endothelial Cell Migration by Endometrial Tissue Involves Epidermal Growth Factor and Is Mediated Via Up-Regulation of the Urokinase Plasminogen Activator Receptor

Abstract

Endometrial angiogenesis is not well studied, but has potential as a model for studies of physiological angiogenesis. Migration as well as proliferation of vascular endothelial cells are modulated by other endometrial cells. This study analyzes the chemotactic signal released from endometrial tissue in a wound assay using human microvascular endothelial cells. Endometrial tissue explants stimulate migration, and this effect is significantly weaker with explants taken at midcycle than those obtained earlier or later in the cycle. Migration is inhibited more than 50% by either blocking antibodies to the urokinase plasminogen activator receptor (uPAR) or enzymatic removal of uPAR from the cell surface. Also, migration is inhibited more than 50% by antibodies to epidermal growth factor (EGF), but not by antibodies to vascular endothelial growth factor or basic fibroblast growth factor. The combination of anti-EGF and anti-uPAR antibodies does not further reduce the response, suggesting that these antibodies target a common pathway. Conditioned medium from endometrial explants contains EGF, and EGF stimulates the migration of endothelial cells in a dose-dependent way. This effect is completely blocked by antibodies to uPAR. These data suggest up-regulation of the uPA system by EGF. Conditioned medium from EGF-treated cells contains less uPA than medium from control cells. In contrast, binding of radiolabeled uPA reveals an increased number of uPA-binding sites in EGF-treated cells. Increased expression of uPAR potentially increases the activation and assembly of focal adhesion sites, a prerequisite for cell migration. We conclude that the endometrial migratory signal has two components. The major part of the signal is blocked by antibodies to EGF, and the response is mediated via up-regulation of uPAR in the endothelial cells. The other part of the signal is unknown, and the response does not involve uPAR. Decreased endometrial chemotactic signal at midcycle is probably related to decreased release of EGF, which is secondary to increased binding to endometrial cell membranes.