Ependymal and choroidal cells in culture: Characterization and functional differentiation

During the past 10 years, our teams developed long-term primary cultures of ependymal cells derived from ventricular walls of telencephalon and hypothalamus or choroidal cells (modified ependymal cells) derived from plexuses dissected out of fetal or newborn mouse or rat brains. Cultures were established in serum-supplemented or chemically defined media after seeding on serum-, fibronectin-, or collagen-laminin-coated plastic dishes or semipermeable inserts. To identify and characterize cell types growing in our cultures, we used morphological features provided by phase contrast, scanning, and transmission electron microscopy. We used antibodies against intermediate filament proteins (vimentin, glial fibrillary acidic protein, cytokeratin, desmin, neurofilament proteins), actin, myosin, ciliary rootlets, laminin, and fibronectin in single or double immunostaining, and monoclonal antibodies against epitopes of ependymal or endothelial cells, to recognize ventricular wall cell types with immunological criteria. Ciliated or nonciliated ependymal cells in telencephalic cultures, tanycytes and ciliated and nonciliated ependymal cells in hypothalamic cultures always exceeded 75% of the cultured cells under the conditions used. These cells were characterized by their cell shape and epithelial organization, by their apical differentiations observed by scanning and transmission electron microscopy, and by specific markers (e.g., glial fibrillary acidic protein, ciliary rootlet proteins, DARPP 32) detected by immunofluorescence. All these cultured ependymal cell types remarkably resembled in vivo ependymocytes in terms of molecular markers and ultrastructural features. Choroidal cells were also maintained for several weeks in culture, and abundantly expressed markers were detected in both choroidal tissue and culture (Na+-K+-dependent ATPase, DARPP 32, G proteins, ANP receptors). In this review, the culture models we developed (defined in terms of biological material, media, substrates, duration, and subculturing) are also compared with those developed by other investigators during the last 10 years. Focusing on morphological and functional approaches, we have shown that these culture models were suitable to investigate and provide new insights on (1) the gap junctional communication of ependymal, choroidal, and astroglial cells in long-term primary cultures by freeze-fracture or dye transfer of Lucifer Yellow CH after intracellular microinjection; (2) some ionic channels; (3) the hormone receptors to tri-iodothyronine or atrial natriuretic peptides; (4) the regulatory effect of tri-iodothyronine on glutamine synthetase expression; (5) the endocytosis and transcytosis of proteins; and (6) the morphogenetic effects of galactosyl-ceramide. We also discuss new insights provided by recent results reported on in vitro ependymal and choroidal expressions of neuropeptide-processing enzymes and neurosecretory proteins or choroidal expression of transferrin regulated through serotoninergic activation. Microsc. Res. Tech. 41:124–157, 1998. © 1998 Wiley-Liss, Inc.