Re-evaluating the role of astrocytes in blood-brain barrier induction

Neural tissue induces brain capillary endothelial cells to express a diverse array of characteristics that allow them to regulate the passage of solutes between the blood and the brain; these features are collectively referred to as the blood-brain barrier (BBB). Because astrocytes are intimately associated with brain capillaries, they have been thought to be the cell type responsible for barrier induction. Widely accepted support of this hypothesis has been derived from experiments showing that astrocytes implanted into the anterior chamber of the rat eye, or onto the chorioallantoic membrane of the chicken embryo, remain unstained by circulating Evan's blue, while grafts of fibroblasts in these sites stain intensely. We have found several limitations associated with placing grafts in either site, leading us to believe that previously reported results are inconclusive. Astrocytes implanted into the anterior chamber form grafts that are poorly vascularized, whereas fibroblast grafts are richly vascularized by vessels which are often fenestrated. This likely accounts for apparent differences in vessel permeability reported by others. We have found that iridial vessels associated with astrocyte grafts do not change their ultrastructure to resemble brain capillaries. Grafting of cells to the chorioallantoic membrane elicits an extensive inflammatory response. Inflammation results in poor delivery of tracers to graft vasculature as well as altering vessel permeability. Treatment of hosts with steroidal anti-inflammatory agents in doses compatible with survival of the host does allow improved graft survival. Even after treatment with anti-inflammatory agents, however, astrocyte graft vasculature fails to express high levels of a barrier marker, the GLUT-1 isoform of the glucose transporter. Transplantation of avascular embryonic spinal cord, that induces robust vessel ingrowth and GLUT-1 expression in intra-embryonic vessels, was unable to elicit the ingrowth of more than a few vessels from the chorioallantoic membrane vasculature, and none of these expressed glucose transporter. We conclude that the anterior chamber and chorioallantoic membrane are not suitable sites for studying BBB induction, and that there is, at present, no conclusive evidence that mature astrocytes play a significant role in the initial expression of the BBB.