Although the precise role played by each of these factors is not clear, their existence suggests yet another level of control during vessel maturation (Fig. 1). Thus, vessel stabilization results from a balance between stimulators (such as VEGF) and inhibitors (such as TGF-β), all acting in the context of the vessel microenvironment. The idea that cell–cell interactions mediate differentiation and stabilization of developing tissues is not restricted to the vasculature. Similar interactions have been described in the developing kidney (15), lung (16), mammary gland (17), and heart (18) where there are heterotypic interactions between epithelial and mesenchymal cells, associated in many cases, with the local production of members of the TGF-β superfamily. An additional parallel can be drawn to the developing nervous system, where heterotypic interactions between an innervating neuron and its target cell effect the survival and/or differentiation of both cell types (19). Thus, heterotypic cell–cell interactions may be a universal mechanism for the local regulation of cell-type specification and differentiation. Whereas the earliest studies of the vasculature described cell behavior in situ, and the past two decades have focused on dissecting the function of isolated vascular cells in culture, we are now at a transition point which recognizes the complexity of the microvasculature (20). Most importantly, the importance of heterotypic cell–cell interactions, local control of growth factor expression (eg, hypoxic regulation of VEGF, local activation of TGF-β), and the signaling potential of the basement membrane are acknowledged and being investigated. As a result, new questions continue to arise. Is vessel assembly controlled similarly in all vascular beds? Studies involving PDGF B-deficient mice suggest there are tissue-specific differences, since not all microvascular beds in these mice lacked pericytes.