The blood-brain barrier (BBB) is a unique feature of the human body, preserving brain homeostasis and preventing toxic substances to enter the brain. However, in various neurodegenerative diseases, the function of the BBB is disturbed. Mechanisms of the breakdown of the BBB are incompletely understood and therefore a realistic model of the BBB is essential. We present here the smallest model of the BBB yet, using a microfluidic chip, and the immortalized human brain endothelial cell line hCMEC/D3. Barrier function is modulated both mechanically, by exposure to fluid shear stress, and biochemically, by stimulation with tumor necrosis factor alpha (TNF-α), in one single device. The device has integrated electrodes to analyze barrier tightness by measuring the transendothelial electrical resistance (TEER). We demonstrate that hCMEC/D3 cells could be cultured in the microfluidic device up to 7 days, and that these cultures showed comparable TEER values with the well-established Transwell assay, with an average (± SEM) of 36.9 Ω.cm(2) (± 0.9 Ω.cm(2)) and 28.2 Ω.cm(2) (± 1.3 Ω.cm(2)) respectively. Moreover, hCMEC/D3 cells on chip expressed the tight junction protein Zonula Occludens-1 (ZO-1) at day 4. Furthermore, shear stress positively influenced barrier tightness and increased TEER values with a factor 3, up to 120 Ω.cm(2). Subsequent addition of TNF-α decreased the TEER with a factor of 10, down to 12 Ω.cm(2). This realistic microfluidic platform of the BBB is very well suited to study barrier function in detail and evaluate drug passage to finally gain more insight into the treatment of neurodegenerative diseases.

One of the main difficulties with primary rat brain endothelial cell (RBEC) cultures is obtaining pure cultures. The variation in purity limits the achievement of in vitro models of the rat blood-brain barrier. As P-glycoprotein expression is known to be much higher in RBECs than in any contaminating cells, we have tested the effect of five P-glycoprotein substrates (vincristine, vinblastine, colchicine, puromycin and doxorubicin) on RBEC cultures, assuming that RBECs would resist the treatment with these toxic compounds whereas contaminating cells would not. Treatment with either 4 lg/mL puromycin for the first 2 days of culture or 3 lg/mL puromycin for the first 3 days showed the best results without causing toxicity to the cells. Transendothelial electrical resistance was significantly increased in cell monolayers treated with puromycin compared with untreated cell monolayers. When cocultured with astrocytes in the presence of cAMP, the puromycin-treated RBEC monolayer showed a highly reduced permeability to sodium fluorescein (down to 0.75 · 10 )6 cm/s) and a high electrical resistance (up to 500 W · cm 2 ). In conclusion, this method of RBEC purification will allow the production of in vitro models of the rat blood-brain barrier for cellular and molecular biology studies as well as pharmacological investigations. Keywords: blood-brain barrier, in vitro model, P-glycoprotein, puromycin, rat brain microvessel endothelium. In the last decade, many efforts have been made to produce reliable in vitro models in order to study the blood-brain barrier (BBB). It is indeed important to better understand the complex cellular and molecular interactions at the interface between blood and brain. The BBB regulates the passage of physiological substances into and out of the CNS and protects it against potentially harmful substances present in the blood. It also prevents the passage of pharmacological substances into the CNS. In order to optimize drug delivery to the CNS, it is important to gain knowledge about the passage of drug candidates through the BBB, especially their effects on the CNS and their toxicity to this barrier (Begley 1996;Tsuji and Tamai 1997). The better we understand BBB regulation, the better we will be able to conceive treatments for CNS pathologies, including neurodegenerative diseases and brain tumours