Regionally specific properties of midbrain glia: I. Interactions with midbrain neurons

Regional astrocyte cultures were obtained by dissecting and dissociating medial and lateral sectors of the midbrain from 14-day Swiss mouse embryos. Once confluent, these cultures were tested by glial fibrillary acidic protein (GFAP) immunocytochemistry to confirm their astrocyte composition and for 2'-3' cyclic nucleotide 3'-phosphohydrolase (CNPase) and microtubule-associated protein 2 (MAP2) immunocytochemistry to rule out oligodendroglial and neuronal components, respectively. In confluent astrocyte cultures from either sector, virtually all cells were GFAP-positive elements, most of which were flat cells accompanied by smaller numbers of flat cells with processes. Confluent astrocyte cultures, derived from medial (M) or lateral (L) sectors, were used as substrata for culturing dissociated cells from medial (m) or lateral (l) sectors of 14-day embryonic midbrains. Fixed cocultures (Ll, Lm, Mm, Ml) were stained with an anti-MAP2 antibody to verify neuronal aggregation and neuritic morphology. In spite of the morphological constancy of glial substrata at plating, MAP2-positive cells in cocultures showed differences in the aggregation of somata and in the length, caliber, and branching of neurites. These differences, which depend mostly on the sector of origin of astrocytes, suggest that the substrata may differ in adhesiveness and/or growth-promoting vs. growth-interfering properties. Together with evidence for sectorial heterogeneity in brainstem radial glia, the present results raise the possibility that cultured astrocytes have properties that reflect the roles played by their parent radial glia in the developing brain.