One of the most striking features of developmental biology is the dramatic morphological changes that an embryo must undergo to achieve its final form. Arguably, the most stunning example of this is found in the embryonic vasculature: not only does the vasculature undergo morphological changes, it must continue to do so adaptively as the size and nutritional needs of the embryo change during gestation. As embryonic blood flow starts long before the end of vessel morphogenesis, the vessels must maintain the integrity of their cell-cell contacts while at the same time remodeling into their final state. Receptor tyrosine kinases and their ligands have been implicated in the regulation of blood vessel growth and remodeling during development. Recently, the Eph receptors and their ephrin ligands were found to be expressed in the developing vasculature. While one Eph receptor, EphB4, is restricted to veins, its specific ligand, ephrinB2, is restricted to arteries. Furthermore, the ephrinB2 knockout mice exhibit defects in blood vessel remodeling, angiogenesis. Although the reciprocal expression of ephrinB2 and EphB4 suggested that Eph signaling from arteries to veins was important for blood vessel development, the presence of additional Eph receptors suggested EphB4 might not be required for this process. Additionally, the widespread expression of ephrinB2 outside the vasculature suggested that vascular-specific expression of this ligand might not be the tissue source necessary for angiogenic remodeling.
To determine which Eph receptor was mediating the ephrinB2 signal, I generated a knockout of the EphB4 gene in mouse. A reporter gene replacement in the EphB4 locus confirmed the vein-biased expression of this receptor. Homozygous EphB4 mutant mice exhibit angiogenesis and cardiac defects, and embryonic lethality indistinguishable from those of ephrinB2 knockout mice. This suggests that EphB4 is the main Eph receptor responsible for transducing the angiogenic ephrinB2 signal. To examine the importance of endothelial specific expression of ephrinB2 in angiogenesis, in contrast to its nonvascular expression, I generated a conditional ephrinB2 mouse. These mice carry a functional ephrinB2 gene, which can be inactivated in a tissue specific manner. Mice with endothelial-specific inactivation of ephrinB2 (and intact non-vascular ephrinB2 expression) exhibit severe angiogenesis and cardiac defects identical to those of the conventional ephrinB2 mutant mice. This suggests that vascular ephrinB2 is essential, and cannot be compensated for by non-vascular ephrinB2 from surrounding vessels.
These studies have clarified two important issues. The first is that the ephrinB2 signal is received by EphB4 expressing endothelial cells (of the veins), rather than by perivascular cells that also express Eph receptors. Second, cphrinB2 expression in endothelial cells of the vessels is an essential tissue source of angiogenic ephrin signals. Together, these studies reinforce the original interpretation of the ephrinB2 mutant, that Eph signaling between arteries to veins is essential for angiogenesis in the early embryo.
