In vitro models of BBB: a tool for the analysis of cell to cell communication in the brain

Many researchers have been trying to set in vitro models of the blood-brain barrier (BBB) aimed at analyzing, in simplified terms, the molecular mechanisms responsible for formation, maintenance and functioning of the BBB, as well as the capability of specific drugs and pro-drugs to cross BBB. We did it, starting with a simpler system of co-culture that allowed us to analyze the effects of neurons on differentiation of brain capillary endothelial cells (RBE4.B cells) in culture, and setting then a more complex model, that includes three cell types (endothelial cells, neurons and astrocytes). The reciprocal geometrical organization of brain cells in this model system is similar to the one observable in vivo: neurons are plated on laminin at the bottom of companion wells of the system, astrocytes on fibronectin, on the outside of the inserts, and RBE4.B cells on collagen IV, inside the inserts. The whole system is fed with a synthetic medium (Maat Medium: Cestelli et al., 1985). Thanks to the model, we found that neurons and astrocytes have synergistic effect on the differentiation of the BCEC layer that forms a functional BBB. As cell contacts among neurons, astrocytes and BCECs are not allowed, we hypothesized that neurons and astrocytes released diffusible factors that are able to induce BCEC differentiation. In particular, we tested, by different approaches, the possibility that brain cells shed angiogenic factors via extracellular vesicles, and found that both neurons and astrocytes release indeed VEGF and FGF2, at least in part through this mechanism. The existance of vesicles released by brain cells has been proved by scanner electron microscopy and immunofluorescence. We are now investigating the possibility of using the model to study possible alterations of BBB in pathological conditions. In particular, we have been looking first for the presence in the sera from MS patients of factors able to induce both morphological and molecular modifications in brain cells, and searching then for a direct effect of MS sera on the permeability properties of the model BBB. Primary cultures of rat astrocytes or neurons were treated with either control- or MS- serum and analyzed for possible alterations of cytoskeletal- and angiogenic- proteins: we found that cytoskeletal proteins, such as GFAP in astrocytes, and neurofilaments in neurons, respectively, are altered in cells treated with MS serum. We also studied the effects of control and MS serum on the permeability of the BBB model, by using radiolabeled sucrose as a tracer of paracellular flux, and the measurement of the transendothelial electrical resistance (TEER) as a parameter of BBB alteration. We also looked at occludin localization as a marker of TJ organization. Interestingly, treatment with MS serum induced a TEER decrease and a parallel loss of the normal peripheral localization of occludin, probably indicating a loss of TJs. Cestelli A. et al., 1985, Dev. Brain Res. 22: 219-227. Cestelli A. et al., 2001, J. Controll. Rel. 76: 139-147. Savettieri G. et al., 2000, NeuroReport 11: 1081-1084. Schiera G. et al, 2003, J. Cell. Mol. Med. 7: 165-170. Schiera G. et al., 2005, J. Cell. Mol. Med. 9: 373-379. Proia et al (2008) Int J Mol Med 21:63-7. Schiera et al (2007) J Cell Mol Med 11:1384-94