WO2014203087A1

WO2014203087A1 - A human blood-brain barrier model derived from stem cells

Info

Publication number: WO2014203087A1
Application number: PCT/IB2014/060186
Authority: WO
Inventors: Lino Da Silva Ferreira; Emmanuel Bernard SEVIN; Roméo Paul CECCHELLI; Sezin ADAY
Original assignee: Biocant - Centro De Inovação Em Biotecnologia; University Of Artois
Priority date: 2013-03-26
Filing date: 2014-03-26
Publication date: 2014-12-24
Also published as: EP2978840B1; US20160040125A1; PT2978840T; EP2978840A1; US20210380938A1; DK2978840T3; ES2841124T3

Abstract

The present disclosure relates to a method for obtaining human brain-like endothelial cells by contacting a population of cells isolated from stem cells with a differentiation medium to obtain endothelial cells and co- culturing said endothelial cells with pericytes, with cells of the neurovascular unit or with a pericytes conditioned medium, to obtain brain-like endothelial cells. The present disclosure also relates to the use of the brain-like endothelial cells as an in vitro model of human blood-brain barrier and a kit for measuring blood-brain barrier permeability of a substance, comprising in vitro human endothelial cells.

Description

DESCRIPTION

"A HUMAN BLOOD-BRAIN BARRIER MODEL DERIVED FROM STEM CELLS" Technical Field

The present disclosure relates to a method to generate a population of endothelial cells showing a brain endothelial cells-like phenotype.

The present disclosure further relates to a population of endothelial cells, said cells showing phenotype similar to brain endothelial cells.

The present disclosure also relates to a stable and reproducible in vitro human brain-blood barrier model.

Background

Blood-brain barrier (BBB) models can provide a valuable tool for studying mechanistic aspects related to the transport of drugs at the brain, as well as biological and pathological processes related to the BBB (Cecchelli, R., et al . Modelling of the blood-brain barrier in drug discovery and development; Nat Rev Drug Discov 6, 650-661 (2007)) . Although several in vitro models were established using murine and bovine cells, the establishment of a stable human BBB model is very important to account for differences between species (Cecchelli, R., et al . Modelling of the blood-brain barrier in drug discovery and development; Nat Rev Drug Discov 6, 650-661 (2007)) . Primary human brain endothelial cells (hBECs) and immortalized human cells have been used as in vitro models; however, several issues prevent their general use including constraints in obtaining human tissue, loss of hBEC phenotype during immortalized cell culture, or lack of important tight junctions and low TEER values as shown in human cell lines (Weksler, B.B., et al . Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J 19, 1872-1874 (2005) ; Sano, Y., et al . Establishment of a new conditionally immortalized human brain microvascular endothelial cell line retaining an in vivo blood-brain barrier function. J Cell Physiol 225, 519-528 (2010) ) . Recently, hBECs have been differentiated from induced pluripotent stem cells (iPSCs); however, the low stability and high variability in the BBB system formed by different iPSC lines might preclude its general use (Lippmann, E.S., et al . Derivation of blood-brain barrier endothelial cells from human pluripotent stem cells. Nat Biotechnol (2012)) .

For example, document WO/2006/056879 Al relates to an immortalized human brain endothelial cell line that is useful as an in vitro model of the blood brain barrier. The document describes the generation of an immortalized human brain endothelial cell line (hCMEC/D3) . The cell line was derived from primary brain endothelial cells transfected with a lentiviral vector system leading to the production of telomerase and SV40 Large T antigen. The cell line retained morphological characteristics of primary brain endothelial cells and expressed specific brain endothelial markers and cell surface adhesion molecules. Furthermore, the cell line expressed chemokine receptors and ATP binding cassette (ABC) - transporters. However, the expression level of ABCA2, MDR1, MRP4 , BCRP, GUT1, 4F2hc, MCT1 and insulin receptor is 4 times lower in hCMEC/D3 cell line than in human brain microvessels (Ohtsoki et al . , Molecular Pharmaceutics 2013, 10(1), 289-296) . Furthermore, the cells show deficiency in typical and important brain endothelial properties such as low TEER value and relative high permeability towards small tracer molecules indicating paracellular leakiness and suboptimal formation of tight j unctions .

Document WO/2007/072953 Al intends to provide a screening system for a drug which passes the blood-brain barrier and acts on the center, a drug which acts on the blood-brain barrier per se, or a drug which is not expected to act in the center but migrates into the brain. This in vitro BBB model is formed by a co-culture of primary brain endothelial cells, pericytes and astrocytes in a three- dimensional culture device. The invention is not related to the use of human brain endothelial cells and therefore does not take into account with inter-species differences in terms of metabolism and physiology. Furthermore, even if a human BBB model could be proposed, it requires the isolation of human brain endothelial cells from an autopsy tissue or freshly resected brain specimens derived from brain tumor or epilepsy patients. The issue here is that brain endotheial cells are not available in enough number for this purpose, and do not have the enough stability to act as a reliable in vitro BBB model.

Document US/2012/0015395 Al describes a method for producing brain specific endothelial cells, preferably comprising the steps of growing human pluripotent stem cells inducing differentiation of the cells by culturing the cells in unconditioned medium, and further expanding the endothelial cells in endothelial cell medium, wherein the expanded cells are GLUT-1⁺, PECAM-1⁺, claudin-5⁺, occludin⁺, ZO-l⁺, and p-glycoprotein⁺ . The invention also claimed a method comprising the step of co-culturing the cells with a cell type selected from the group of astrocytes, neural progenitor cells, and pericytes. The brain endothelial cells derived from human pluripotent stem cells yielded TEER with an average value of 860 ± 260 Ω cm2 (Lippmann et. al . , Nature Biotechnology 2012) . Yet, the TEER values fluctuated over time. For example, the TEER values in a co-culture of brain endothelial cells with astrocytes changed 200% during the first 50 h. Furthermore, it is unclear the stability of the system overtime and its reproducibility .

Document WO/2007/140340 A2 related to methods for providing CD34+ cells from embryoid bodies and stimulating these cells to give rise to endothelial-like and/or smooth muscle-like cells. However, the object of the invention was not linked to the specification of the endothelial cells into brain endothelial cells.

Summary

De initions

"CD34⁺ cells" refers to cells expressing CD34 antigen. This antigen is a single-chain transmembrane glycoprotein expressed in several cells including human hematopoietic stem and progenitor cells, vascular endothelial cells, embryonic fibroblasts and some cells in fetal and adult nervous tissue.

"Brain-like endothelial cells" refers to cells that share properties (gene, protein and functional) of fully functional brain endothelial cells, including the expression of at least one of the following markers, low density lipoprotein receptor, insulin receptor, leptin receptor, transferrin receptor, receptor for advanced glycation endproducts, retinol binding protein, SLC7A5, SLC2A1, SLC38A5, SLC16A1, ABCB1, ABCG2, ABCC1, ABCC2, ABCC4, ABCC5, claudin 1, claudin 3, ZO-1, occludin, JAM-A and claudin-5.

"Hematopoietic stem cells" refers to cells that can themselves or whose progeny can form myeloid, erythroid, and/or megakaryocyte colonies as described in Eaves, et al . , Atlas of Human Hematopoietic Colonies, 1995, StemCell Technologies, Vancouver; Coutinho, et al, in Hematopoiesis : A Practical Approach, Testa, et al, eds . , 1993, Oxford Univ. Press, NY, pp 75-106, and Kaufman, et al . , PNAS, 2001, 98:10716-10721.

"Pericytes" refers to cells that express one of the following markers: vimentin, neuro-glial 2 (NG2), platelet- derived growth factor receptor beta (PDGFR-β), -smooth muscle actin ( -SMA) , cells that express one of the following markers: viet al . , Current Neurovascular Research 2011, Modelling the neurovascular unit and the BBB with the unique function of Pericytes) .

In view of the drawbacks to the pior art, one of the problems was to develop an in vitro human BBB system that could be stable for more than 15 days and be reproducible for different stem cells and could be implemented in different laboratories. The system described is the first human in vitro BBB system with high reproducibility evaluated in three different laboratories; and in stem cells collected from more than 4 different donors - and stable - more than 20 days. This has not ever been addressed by previous technologies described in the literature .

An aspect of disclosed subject matter relates to human brain-like endothelial cells wherein at least a portion of the cells express at least one of the following markers: ZO-1, occludin, JAM-A, claudin-5, claudin-3, claudin-1, preferably express ZO-1 and claudin-1.

In embodiments of the disclosure, the brain-like endothelial cells further express at least one of the following transporters or receptors: aminoacid - SLC7A5, SLC16A1, glucose - SLC2A1.

In embodiments of the disclosure, the brain-like endothelial cells further express a portion of at least one of the following molecules: CD40, VCAM-1.

Others embodiments of the disclosure, at least a portion of the brain-like endothelial cells express at least one of the following transcripts of key efflux transporters as P- glycoprotein, breast cancer resistance protein and multidrug resistance protein.

In others embodiments of the disclosure, at least a portion of the brain-like endothelial cells expresses at least one of the following genes up-regulated: SLC44A5, SLC25A27 the endothelial cells, SLC23A3.

In others embodiments of the disclosure, at least a portion of the brain-like endothelial cells further express at least one of the following markers: lipoprotein receptor, insulin receptor, leptin receptor, transferrin receptor, receptor for advanced glycation endproducts, retinol binding protein, SLC38A5, ABCB1, ABCG2, ABCCl, ABCC2, ABCC4, ABCC5.

In others embodiments of the disclosure, endothelial cells derived from stem cells, can be preferably from hematopoietic stem/progenitor cells (CD34⁺ cells) derived from human cord blood, but can be extended to endothelial cells derived from hematopoietic stem/progenitor cells derived from human peripheral blood, or other type of endothelial cells.

A further aspect of the disclosure relates to a method of inducing a blood brain barrier phenotype in endothelial cells derived from stem cells, preferably from hematopoietic stem/progenitor cells (CD34⁺ cells) derived from human cord blood, but can be extended to endothelial cells derived from hematopoietic stem/progenitor cells derived from human peripheral blood, or other type of endothelial cells.

An aspect of disclosed subject matter relates to a method for obtaining in vitro human brain-like endothelial cells comprising the following steps:

contacting a population of stem cells with a differentiation medium to form endothelial cells;

co-culturing the said endothelial cells with pericytes or with cells of the neurovascular unit or with a pericytes conditioned medium, to obtain brain like endothelial cells, preferably during at least 4 days, more preferably during 5-7 days, namely 5,6,7 days. In others embodiments of the method for obtaining in vitro human brain-like endothelial cells wherein the said stem cells may be CD34+ derived from human cord blood, or cells from human peripheral blood.

In others embodiments of the a method for obtaining in vitro human brain-like endothelial cells wherein the cells are grown on solid support, preferably transwell systems or well plates. In preferred embodiments, the pericytes can be placed in the bottom of each plate.

In others embodiments of the method for obtaining in vitro human brain-like endothelial the differentiation medium may be EGM-2 medium with 20% (v/v) FBS and 50 ng/mL of VEGF165.

In others embodiments of the method for obtaining in vitro human brain-like endothelial cells the pericytes may express at least one of the following markers: vimentin, PDGF-β, NG-2, a-SMA; γ-GT.

In others embodiments of the method for obtaining in vitro human brain-like endothelial cells the pericytes may be seeded at a density of 40xl0³- 50xl0³, preferably 45xl0³ cells .

In others embodiments of the method, the pericytes might be replaced by a cell line that secrete Wnt3a or Wnt7a.

In others embodiments of the method for obtaining in vitro human brain-like endothelial cells the pericytes conditioned medium (obtained from 45 xlO³ cells of pericytes cultured for 6 days in a well of a 12-well plate) may be replaced every day. A further aspect of aspect of disclosed subject matter relates to a method for evaluating blood-brain barrier permeability of a substance, cell or protein comprising exposing the said test substance, cell or protein to the in vitro endothelial cells, said substance can be any synthetic or natural compound.

In others embodiments a method for evaluating blood-brain barrier permeability of a substance, cell or protein by the measurement of efflux transport, preferably in the presence or absence of inhibitors of the efflux pumps. Preferably, the inhibitors may be at least one of the following: cyclosporin-A, PSC-833, MK-571, KO-143, verapamil or elacridar .

A further aspect of aspect of disclosed subject matter relates to a method for evaluating blood-brain barrier metabolism of a test substance, cell or protein which comprises the following steps:

contacting a test substance, cell or protein to the brain endothelial cells previously described;

analysing the metabolic degradation of the said test substance, cell or protein.

In others embodiments of the method for evaluating blood- brain barrier toxicity of a test substance comprises the culturing the said brain endothelial cells in the presence of the said test substance.

The viability can be determined by a live/dead assay, preferably using calcein and propidium iodide as reagents, ATP production, cell membrane damage by the release of lactate dehydrogenase, cell replication by a BrdU assay.

Any changes in the BBB functionality (e.g: permeability to a non permeant marker) in vitro could be used as an alternative toxicological endpoint.

In others embodiments of the method for evaluating blood- brain barrier metabolism of a test substance comprises the culturing the said brain endothelial cells in the presence of the said test substance.

A further aspect of the disclosure relates to a kit for measuring blood-brain barrier permeability of a substance, comprising the in vitro human endothelial cells previously described .

In various embodiments, pericytes are preferably derived from bovine, but they can be isolated from any other species. They are characterized by a set of different markers including PDGF-β in other species. They are characterized by a set of different markers including PDGF- β isolated from endothelial cell culture, -smooth muscle actin ( oi-SMA) , γ-glutamyl-transpeptidase and P-glycoprotein (P-gp) and others that someone skilled in the art will identify. So far, there is no single marker that differentiates pericytes from other cells. In one embodiment, the pericytes are placed in the bottom of the plate, but they can be also applied in one of the sides of the transwell filter. In a preferable embodiment, the pericytes are seeded at a density of 45xl0³ cells into each well of 12-well plates. The cells are cultured on Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 20% (v/v) fetal bovine serum (FBS), 2 mM L-glutamine, 50 μg/mL gentamycin and 1 ng/mL basic fibroblast growth factor.

In a preferred embodiment, the pericytes are cultured in the presence of the endothelial cells to induce BBB properties. Conditioned medium obtained from pericytes might have the same inductive effect on endothelial cells if medium is replaced every day, and at suitable concentrations .

In one embodiment, the BBB properties of endothelial cells can be achieved by supplementing the culture medium with Wnt3a, Wnt7a or a mixture of Wnt3a and Wnt7a. In a preferable embodiment, the concentration of Wnt3a and Wnt7a is 6.25 ng/mL.

Brief description of drawings

Without intent to limit the disclosure herein, this application presents attached drawings of illustrated embodiments for an easier understanding.

Figure 1: Expression of BBB markers, stability and functional properties of a monolayer of human BLECs;

(A) BLECs were obtained by the co-culture of CD34+- derived ECs with pericytes for 6 days in a transwell™ system.

(B-D) Paracellular permeability to lucifer yellow of EC monolayers either cultured alone or with pericytes. Results are Mean ± SEM (n>4) ;

(E) Expression of endothelial and BBB markers in BLECs as obtained by immunofluorescence;

(F) Electron micrographs of ECs cultured alone (2,3) or with pericytes (1) ; (1) In the intercellular cleft, WGA-HRP penetrates from the luminal compartment (asterisks) to the tight junction, which occludes the cleft (arrows) . From this point, the intercellular space is free of the electron-dense reaction product ;

(2) When ECs are cultured alone, there is no occlusion of the intercellular space between the ECs in 84% of the cases, and the tracer penetrates from the luminal compartment

(asterisks) trough the entire intercellular cleft and is deposited in the underlying matrix

( arrowheads ) ;

BLEC gene expression of tight junctions and influx transporters. Results are Mean ± SEM (n=3);

BLEC gene expression of efflux transporters and large molecule receptors;

Expression of P-gp and RAGE as evaluated by immunofluorescence. In E and I, bar corresponds to 50 μπι. *P<0.05, **P<0.01, ***P<0.001.

Figure 2: Functional properties of BLECs and mechanism for the in the induction of BBB properties in CD34+-derived ECs;

(A) Effect of P-gp protein inhibition on active transport of drugs;

(B) Efflux ratio of small (sucrose) and large (HSA and IgG) molecules. In A and B: results ± SEM (n=3-7);

(C) Unbound brain-to-plasma or CSF-to-plasma concentration ratio^'s for human and rat; (D) Expression of BBB markers as evaluated by whole genome microarrays of monocultures or co-cultures of CD34+-derived ECs with pericytes at day 3 and 6;

(E) Effect of Wnt3a, Wnt7a and BIO in the expression of β-catenin (after 1 day) as well as in the paracellular permeability (at days 1 and 5) of monocultures of CD34+-derived ECs. Results are Mean ± SEM (n=3-6) . The dashed line represents the paracellular permeability of ECs in co-culture with pericytes for 6 days. For permeability results the concentrations of Wnt3a, Wnt7a and BIO were 6.25 ng/mL, 6.25 ng/mL, and 0.5 μΜ;

(F) Expression and localization of claudin-1 (at day 6) and total β-catenin (day 3) in monoculture of CD34+- derived ECs cultured in medium supplemented with BIO (0.5 uM) or Wnt3a (6.25 ng/mL) . Arrow-heads indicate nuclear accumulation of β-catenin. Bar corresponds to 50 μπι. *P<0.05, **P<0.01, ***P<0.001.

Figure 3: Differentiation of human umbilical cord CD34+ cells into ECs and evaluation of their paracellular permeability;

(A) Schematic representation of the differentiation of hematopoietic stem cells (CD34+CD45+CD31+KDR-vWF- CD14-) into ECs (2-3 weeks of differentiation) and evaluation of their paracellular permeability (Pe) using a Transwell™ system;

(B) ECs immediately after differentiation (before culture in the Transwell™ system) express typical EC markers including CD31, VE-cadherin (VECAD) , vWF and are able to incorporate AcLDL;

(C) ECs after culture in the Transwell™ system have typical cobblestone morphology, express vWF and markers associated to hBECs such as claudin-5, ZO-1, and occludin; however, the expression of all these markers is discontinuous and cells do not express claudin-1 at cell-cell contacts. Bar corresponds to 50 μπι;

(D) Paracellular permeability of human ECs in monoculture and bovine ECs in co-culture with astrocytes for 12 days .

Figure 4: (A) Paracellular permeability of CD34+-derived ECs after co-culture with different types of cells in EGM-2 supplemented with 2% fetal calf serum FCS) . Results are Mean ± SEM (n=6) ; (B) Characterization of bovine pericytes by phase contrast and Immunocytochemistry for the expression of vimentin, neuro-glial 2 (NG2), platelet- derived growth factor receptor beta (PDGFR-β), and -smooth muscle actin ( -SMA) . Scale bar corresponds to 50 μπι. CM stands for conditioned media; (C) The induction of BBB properties on CD34+-derived ECs requires the presence of pericytes in the co-culture system since pericyte- conditioned medium does not have the same BBB-inductive properties .

Figure 5: (A) Double immunostaining for anti-human receptor for advanced glycation endproducts (RAGE) and anti-human organic cation/carnitine transporter (OCTN2; also known as SLC22A5) in monoculture (A.1 ) or in co- culture of CD34+-derived ECs with pericytes (A.2) at day 6. In the co-culture system, RAGE is present essentially in the luminal side of endothelial cells and OCTN2 in the abluminal side, while in mono-culture, both markers seem to be located in the same plane. Bar corresponds to 10 μπι; (B) Paracellular permeability in a co-culture of CD34+- derived ECs with pericytes at day 6 obtained from different donors. Results are Mean ± SEM (n=4);

(C) Interlaboratory reproducibility. The BBB was generated in two different laboratories. The paracellular permeability to lucifer yellow was not statistical significant. Results are Mean ± SEM (n>4) ;

(D) Stability of the BBB properties after removal of the pericytes. CD34+-derived ECs were in co-culture with pericytes for 14 days (1) or in co-culture for 6 days and then 8 days in monoculture (2) .

Figure 6: (A) Transendothelial electrical resistance (TEER) of monocultures of CD34+-derived ECs or co-cultures of ECs with pericytes for 6 days . The TEER of the co-culture of ECs was compared with the gold standard of bovine brain microvascular endothelial cells co-cultured with bovine astrocytes for 12 days on insert filters 30 mm diameter. Values are Mean ± SEM, n=4. *** P<0.001; ns means P>0.05; (B) Expression of adhesion molecules by ECs in co-culture with pericytes. The expression of the adhesion molecules was assessed by flow cytometry analysis on untreated and treated ECs by TNF (10 ng/mL) for 24 h.

Figure 7: (A-C) Western blot for the expression of Shh (A), Wnt7a (B) , Wnt3a (C) and total β-catenin (D) in CD34+- derived ECs in monoculture (1) or in co-culture with pericytes (2), or pericytes in monoculture (3) or pericytes in co-culture with ECs (4), for 6 days. Human recombinant Wnt3a, Wnt7a and Shh were used as a positive control. Data shown are representative of n=2. In D: results ± SEM, n=2. Figure 8: qRT-PCR results showing changes on Wnt signaling (A-C) , tight junctions (D) and BBB transporters (E) genes on CD34+-derived ECs co-cultured with pericytes for 1,3 and 6 days. Values are Mean ± SEM, n=4. Our results show that Wnt3a transcript increased significantly at day 1 followed by a decrease at day 6 to baseline levels. Genes of canonical Wnt ligands Wnt7a and Wnt7b, which have been reported to be involved in BBB development, increased slightly at day 1 and then decreased at day 3 to baseline levels. The expression of genes encoding Wnt receptor frizzled 4 (FZD4) and frizzled 6 (FZD6) were not affected by the co-culture system; however, Wnt receptor frizzled 7 (FZD7) was significantly up-regulated up to 6 days. The expression of LEF1, the β-catenin-associated transcription factor, peaked at day 1 matching the profile observed for Wnt3a and FZD7. The expression of APCDD1, an antagonist of Wnt signaling and highly expressed in adult brain endothelial cells, peaked at day 3, at the time that Wnt3a drops significantly. Finally, genes related to tight junctions such as claudin 1 and ZO-1 and the transporters SLC7A5 and SLC16A1 are upregulated overtime.

Figure 9: Modulation of Wnt signaling activates the barrier properties of ECs in monoculture;

(A) Schematic representation of the methodology used to assess the modulation of Wnt signaling. CD34+-derived ECs were seeded in a transwell™ insert coated with Matrigel at a density of 80,000 cells. Wnt ligands were added in the culture medium at the basolateral side ;

(B) qRT-PCR results showing differences in expression of claudin-1 and Lefl genes on CD34+-derived ECs cultured with or without Wnt3a. Values are Mean ± SEM, n=4;

(C-D) Paracellular permeability of untreated ECs or ECs treated with different concentrations of human recombinant protein Wnt3a (C) or Wnt7a (D) for 5 days. Results are Mean ± SEM (n=4) .

Figure 10: Abrogation of Wnt signaling in ECs during co- culture with pericytes affect their paracellular permeability;

(A) Schematic representation of the methodology used to assess the effect of abrogation of Wnt signaling. CD34+-derived ECs were seeded in a transwell™ insert coated with Matrigel at a density of 80,000 cells and cultured in medium supplemented with XAV 939 (0.1 and 1 uM) . In the bottom of the transwell™ was seeded 45,000 bovine pericytes. After 4 days of coculture, the paracellular permeability and cell organization were evaluated;

(B) Fluorescence microscopy images showing the expression of ZO-1 in untreated ECs or ECs treated with XAV 939 (1 uM) for 4 days;

(C) Paracellular permeability of untreated ECs or ECs treated with 0.1 or 1 μΜ XAV939 for 4 days. Results are Mean ± SEM (n=4) .

Detailed description

In the present disclosure is described a method to generate a human blood-brain barrier model using cord blood-derived hematopoietic stem cells. The cells were initially differentiated into endothelial cells followed by the induction of blood-brain barrier (BBB) properties by co- culture with pericytes. The brain-like endothelial cells (BLECs) express tight junctions and transporters typically observed in brain endothelium and maintain expression of most in vivo BBB properties for at least 20 days.

To differentiate stem cells into endothelial cells, CD34+CD45+CD31+KDR-vWF-CDl 4- cells isolated from cord blood were initially cultured for 15-20 days in EGM-2 medium with 20% (v/v) FBS and 50 ng/mL of VEGF165 (Supplementary Fig. 1A) . At this stage, cells have a cobblestone-like morphology and express high levels of endothelial cells markers, including CD31, VE-cadherin and vWF (Supplementary Fig. IB) . When these cells were grown to confluence on filters for 6 days they show discontinuous expression of ZO-1, occludin and claudin-5, do not express claudin-1 at cell-cell contacts and have high permeability to Lucifer yellow (2.0 x 10-3 cm/min) as compared to bovine BECs (Supplementary Figs. 1C and ID) .

To induce BBB properties in CD34+-derived endothelial cells, cells were seeded in a transwell™ system and co- cultured with pericytes (Fig. 1A) . Pericytes were selected after a screening of different cell types from the neurovascular unit (Supplementary Figs. 2A and 2B) and because of their role in the stabilization/maturation of BBB (Armulik, A., et al . Pericytes regulate the blood-brain barrier. Nature 468, 557-561 (2010); Daneman, R., Zhou, L., Kebede, A.A. & Barres, B.A. Pericytes are required for blood-brain barrier integrity during embryogenesis . Nature 468, 562-566 (2010) ) . Under these conditions, the permeability of endothelial cells decreases during the first 3 days until it reaches a stationary phase at day 4 (Fig. IB), maintaining its stability up to 20 days (Fig. 1C) . At day 6, the cells had low permeability values (0.61 x 10-3 cm/min) similarly to the values found in other BBB models (Deli, M.A., et al . Permeability studies on in vitro blood-brain barrier models: physiology, pathology, and pharmacology. Cell Mol Neurobiol 25, 59-127 (2005) ) (Fig. ID), they showed a continuous expression of ZO-1, occludin, JAM-A and claudin-5 at cell-cell contacts (Fig. IE) and they were able to block the passage of wheat germ agglutinin (WGA) - horseradish peroxidase (HRP) in contrast with monolayers of CD34+-derived endothelial cells where WGA-HRP reached the underlying matrix (Fig. IF) . Importantly, the induction of BBB properties in CD34+- derived endothelial cells is highly reproducible since similar permeability results were obtained for cells derived from multiple human donors (Supplementary Fig. 3B) and in 3 different laboratories (Supplementary Fig. 3C) . Furthermore, the BBB properties of CD34+-derived endothelial cells are lost if the pericytes are removed from the co-culture system (Supplementary Fig. 3D) showing that the crosstalk between the two cells is important to maintain the BBB properties. Cells co-cultured with pericytes for 6 days express transcripts encoding tight junctions such as ZO-1 and claudin-1 higher than in endothelial cells in monoculture, while the expression of claudin-3 and occludin was similar (Fig. 1G) . Importantly, the expression of influx transporters, specifically the expression of aminoacid (SLC7A5, SLC16A1) and glucose (SLC2A1) transporters and receptors (e.g. transferrin receptor; TFRC) was increased when the cells were co- cultured with pericytes relatively to cells cultured alone. In addition, endothelial cells co-cultured with pericytes for 6 days express transcripts of key efflux transporters such as P-glycoprotein (P-gp) , breast cancer resistance protein (BCRP) and multidrug resistance protein (MRP; subfamily of the ATP-binding cassette (ABC) transporters) family (Fig. 1H) . As in hBECs, the receptor for advanced glycation end products (RAGE) and P-gp protein were expressed as confirmed by immunofluorescence (Fig. II), being RAGE located at the luminal side of cells (Supplementary Fig. 3A) . Overall, endothelial cells co- cultured with pericytes for 6 days have BBB properties at gene, protein and permeability levels, and from now on are named as brain endothelial-like cells (BLECs) .

BLECs have the ability to act as an active barrier. The inhibition of P-gp protein by verapamil or elacridar, and the concomitant blocking of the active transport of drugs to outside the cell, it leads to a significant increase in the accumulation of the antitumor drug vincristine (Fig. 2A) . This result demonstrates that P-gp is functionally active in BLECs. The higher efflux ratio of IgG as compared to human serum albumin shows receptor-mediated transport of macromolecules across the polarized monolayer (Fig. 2B) . In addition, BLECs have the ability to form a monolayer that has a transendothelial electric resistance (TEER) similar to monolayers of bovine BECs (Supplementary Fig. 4A) and higher than monolayers of human hCMEC/D3 cell line (<40 Qcm2 ) (Weksler, B.B., et al . Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J 19, 1872-1874 (2005) ) . Moreover, BLECs express constitutively the adhesion molecule ICAM-2, typically found in hBECs (Bo, L., et al . Distribution of immunoglobulin superfamily members ICAM-1, -2, -3, and the beta 2 integrin LFA-1 in multiple sclerosis lesions. J Neuropathol Exp Neurol 55, 1060-1072 (1996)), and show an up-regulation in the expression of ICAM-1, ICAM-2, CD40 and VECAM-1 after stimulation with 10 ng/mL TNF- for 24 h, as hBECs (Weksler, B.B., et al . Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J 19, 1872-1874 (2005) (Supplementary Fig. 4B) . Finally, the in vitro ratio of concentrations of unbound drug in brain and plasma for atenolol, bupropion, rifampicin and verapamil were closer to the in vivo ratio of concentrations of unbound drugs in cerebrospinal fluid (CSF) and plasma reported in humans than in rats(Friden, M., Gupta, A., Antonsson, M., Bredberg, U. & Hammarlund- Udenaes, M. In vitro methods for estimating unbound drug concentrations in the brain interstitial and intracellular fluids. Drug Metab Dispos 35, 1711-1719 (2007)) (Fig. 2C) . Together, these results indicate that brain-like endothelial cells can be used to predict accurately in humans the in vivo transport of drugs with different properties .

To study the induction of BBB properties in CD34+-derived endothelial cells, these cells cultured alone or with pericytes for 3 or 6 days were characterized by whole genome microarrays. Gene expression analyses at 6 days show that 84 and 2 genes are up- and down-regulated in CD34+- derived endothelial cells in co-culture, respectively, relatively to CD34+-derived endothelial cells in monoculture (Supplementary Tables 3 and 4) . From the overall up-regulated genes, 3 genes were related with influx transporters including SLC44A5, SLC25A27 and SLC23A3, and 2 genes were related with Wnt signaling (Wnt inhibitory factor 1 and disheveled associated activator of morphogenesis (Cecchelli, R., et al . Modelling of the blood-brain barrier in drug discovery and development. Nat Rev Drug Discov 6, 650-661 (2007)) (Daam 1)) (Fig. 2D) . Yet, the expression of most markers associated to BBB (tight junctions and transporters) was not significantly different in CD34+-derived endothelial cells in monoculture and co-culture, which indicates that pericytes exert a discrete influence on endothelial BBB-specific genes. Gene expression on CD34+-derived endothelial cells in co-culture at day 6 and 3 was significantly different regarding BBB markers, specifically for efflux transporters including solute carrier family members SLC2A3, SLC6A6 and SLC47A1

(downregulated at day 6), and members SLC30A3, SLC26A10, SLC13A3 and SLC44A5 (upregulated at day 6) and non-BBB markers such as channels and extracellular matrix

(Supplementary Tables 3 and 4) . Together these results show that the induction process is a dynamic process affecting the expression of transporters, channels and extracellular matrix components.

Two major pathways regulating the formation of BBB are the canonical Wnt/wingless pathway acting via β-catenin stabilization and Sonic hedgehog (Shh) pathway (Liebner, S., et al . Wnt/beta-catenin signaling controls development of the blood-brain barrier. J Cell Biol 183, 409-417 (2008); Alvarez, J. I., et al . The Hedgehog pathway promotes blood-brain barrier integrity and CNS immune quiescence. Science 334, 1727-1731 (2011); Daneman, R., et al . Wnt/beta-catenin signaling is required for CNS, but not non-CNS, angiogenesis . Proc Natl Acad Sci USA 106, 641- 646 (2009) . Protein analyses show that pericytes do not express Shh but do express Wnt ligands such as Wnt3a and Wnt7a (Supplementary Fig. 5) . By other hand, endothelial cells express at gene level Wnt receptors such as frizzled receptor 4, 6 and 7 ( FZD4 , FZD6 and FZD7 ) , and in co- culture with pericytes, they show an up-regulation in the expression of Wnt3a and FZD7 receptor during the first day followed by a decrease in the next 5 days (Supplementary Fig. 6) . This was accompanied by an increase of APCDD1, an antagonist of Wnt signaling, that peaked at day 3, and an increase of the tight junctions ZO-1 and claudin-1

(Supplementary Fig. 6) . To determine whether the activation of Wnt is required for the induction of barrier properties in CD34+-derived endothelial cells, these cells were cultured alone for 5 days and then exposed them to Wnt ligands/agonists . Endothelial cells respond rapidly to BIO, a specific pharmacological inhibitor of glycogen synthase kinase-3 (GSK-3) and thus an activator of Wnt signaling, or Wnt3a by increasing the expression of active β-catenin

(Fig. 2E) . The paracellular permeability of Wnt3a-treated endothelial cells to Lucifer Yellow was statistical lower

(P<0.01, n=4) for short-term (1 day) and long-term (5 days) as compared to untreated cells (Fig. 2E and Supplementary Fig. 7) . The effect of Wnt7a and BIO was only observed at day 5. During the induction process by Wnt 3a or BIO, there is an increase in the expression and nuclear localization of total β-catenin (Fig. 2F and Supplementary Fig. 5) and the localization of claudin-1 at the cell-cell contacts

(Fig. 2F) . The localization of claudin-1 at the periphery of the cells might explain the restrictive permeability of endothelial cells in co-culture with pericytes. Overall, results indicate that Wnt pathway contributes, at least in part, for the induction of BBB properties in CD34+-derived endothelial cells.

To further confirm the role of Wnt pathway in the induction of BBB properties, was abrogated the Wnt signaling in endothelial cells co-cultured with pericytes. Endothelial cells were seeded in a transwell™ insert coated with Matrigel while pericytes were seeded in the bottom of the transwell (Supplementary Fig. 8) . Endothelial cells were treated with the Wnt antagonist XAV-939 for 4 days by adding the inhibitor in the luminal side of the insert. The abrogation of Wnt pathway, in conditions that did not affect cell viability, increased the paracellular permeability of the endothelial monolayer to lucifer yellow. These results again indicate that Wnt signaling is required for the BBB properties in CD34+-derived endothelial cells co-cultured with pericytes.

In summary, was generated a human in vitro BBB model from endothelial cells derived from cord blood hematopoietic stem cells that is highly reproducible and stable for at least 20 days after its derivation. Is provided in vitro evidence for a role of pericytes in the induction of BBB formation through the canonical Wnt pathway. Due to the relative easy access to cord blood stem cells, this model can be adopted by the research community to improve the delivery of therapeutic agents into the central nervous compartment for the treatment of stroke, multiple sclerosis and brain tumors .

In one embodiment of the present disclosure can be used as a method to measure BBB permeability to a test substance. The test substance may be any synthetic or natural compound, with variable molecular weight and hydrophilicity/hydrophobicity ratio. The method of the disclosure can measure passive diffusion or active transport, as appreciated by those skilled in the art. Efflux transport can be measured wherein measuring permeability values is performed in the presence or absence of inhibitors of the efflux pumps such as, but not limited to, cyclosporin-A, PSC-833, MK-571, KO-143. The methods of the present disclosure can also be used to measure blood brain barrier metabolism of a substance by measuring permeability values and profiling the metabolic degradation of compounds of interest as a function of time using quantitative analytical techniques such as high pressure liquid chromatography and mass spectrometry. Test substances that prove to pass our BBB in vitro model may be further analyzed for their pharmacological profile.

In another embodiment, the in vitro BBB model of the present disclosure may be useful as a method for determining the toxicity of a test substance or vector towards the BBB. In this case, the method comprises the culture of the brain endothelial-like cells in the presence of the test substance and assessing its viability after a certain time. A range of concentrations of the test substance can be used to determine the IC50. Cell viability can be determined by a live/dead assay using calcein and propidium iodide as reagents, ATP production, cell membrane damage by the release of lactate dehydrogenase, cell replication by a BrdU assay.

In another embodiment, the in vitro BBB model can be used to design more effective vectors to target or delivery drugs into the brain. This might be useful for the treatment of vascular dysfunction in patients with Alzheimer^'s. Neudegeneration is likely a consequence of altered drug transport across the BBB and abnormal cerebral blood flow due to amyloid peptide deposition. Our in vitro BBB model can be very useful for testing drug candidates for the treatment of Alzheimer.

Isolation and differentiation of CD34+ cells from UCB In a preferred embodiment CD34+ cells may be isolated from human umbilical cord blood and differentiated into endothelial cells according to a protocol previously reported by us (Pedroso, D.C., et al . Improved survival, vascular differentiation and wound healing potential of stem cells co-cultured with endothelial cells. PLoS One 6, el6114 (2011)) Briefly, isolated CD34+ cells were cultured in EGM-2 medium (preferably Lonza) supplemented with 20%

(v/v) fetal bovine serum (preferably FBS; Life Technologies) and 50 ng/mL of VEGF165 (preferably PeproTech Inc.), on 1% gelatin-coated 24-well plates (2xl0⁵ cells/well) . After 15-20 days endothelial cells are seen in the culture dish. For each experiment, the cells were expanded in 1% (w/v) gelatin-coated 100 mm Petri dishes

(preferably BD Falcon) in EGM-2 medium (with all the supplements except FBS and gentamycin/amphotericin) supplemented with 2% (v/v) FBS, 50 μg/mL gentamycin

(preferably Biochrom AG) and 1 ng/mL home-made bFGF.

Di ferenciation of CD34+ derived endothelial cells into brain like endothelial cells by pericytes

Isolation of pericytes

In a preferred embodiment pericytes may be extracted from freshly collected bovine brain capillaries. Brain capillaries were collected on a 60 μπι nylon sieve

(preferably Blutex®, Saati, France) as described by Meresse et al . (1989) and suspended in Hanks Balanced Salt Solution

(preferably HBSS, Sigma-Aldrich) containing 10 mM HEPES and 0.1% BSA. This suspension was centrifuged at lOOOg for 7 min at room temperature. The pellet was then digested with 2 mg/mL collagenase-dispase (preferably Roche Diagnostics), 10 g/mL DNasel (preferably Roche Diagnostics) and 0.147 g/mL TLCK (preferably Sigma-Aldrich) , for 30 minutes at 37°C in a shaking water bath. After washes, the digested capillaries were seeded onto growth factor reduced Matrigel

(preferably BD Biosciences) -coated dishes (preferably Corning) containing pericyte growth culture medium: DMEM

(preferably Life Technologies) supplemented with 20% fetal calf serum (preferably Integro) , 2 mM L-glutamine

(preferably Merck Chemicals), 50 μg/mL gentamicin

(preferably Biochrom AG) and 1 ng/mL bFGF (preferably Sigma-Aldrich) . The medium was changed every other day. Pericytes and endothelial cells migrated from the vessels walls. Pericytes rapidly overgrew from capillaries and invaded the whole surface of the dishes. Confluent cultures consisting almost exclusively of pericytes, were dissociated using trypsin/EDTA saline solution (preferably 0.05%/0.02% Biochrom AG), and cells were frozen in liquid nitrogen. For experiments, each pericyte vial was rapidly thawed and seeded in gelatin (preferably sigma-Aldrich) coated 60-mm Petri dishes containing pericyte culture medium. After thawing, there were no endothelial cells left in cultures. Pericytes were subcultured at a split ratio 1/3, and were used at passages ≤3.

Co-culture of CD34+ derived endothelial cells with pericytes .

In a preferred embodiment for co-culture experiments, pericytes may be initially seeded on 60-mm gelatin-coated petri dishes and cultured in Dulbecco's Modified Eagle's Medium (DMEM) (preferably Life Technologies) supplemented with 20% (v/v) fetal bovine serum (FBS) (preferably Life Technologies), 2 mM L-glutamine, 50 μg/πιL gentamycin and 1 ng/mL basic fibroblast growth factor (bFGF) . The cells reached confluency after 2 days. 45xl0³ cells were seeded into each well of 12-well plates (preferably Costar) . CD34+- endothelial cells growing on gelatin-coated 100 mm petri dishes in EGM-2 (with all the supplements except FBS and gentamycin/amphotericin) supplemented with 2% (v/v) FBS, 50 μg/mL gentamycin (preferably Biochrom AG) and 1 ng/mL home-made bFGF were trypsinized and cells were seeded at a density of 8<10⁴/insert onto the Matrigel-coated (preferably BD Biosciences) Transwell™ inserts (preferably Costar) . After 6 days in co-culture, the experiments were carried out.

Reverse transcription and quantitative real time polymerase chain reaction (qRT-PCR) analysis. CD34+-endothelial cells cultured in different conditions were homogenized in Trizol reagent (preferably Life Technologies) and total RNA was extracted using the RNeasy Mini Kit (preferably Qiagen) , according to manufacturer's instructions. In all cases, cDNA was prepared from 1 \iq total RNA using Taqman Reverse transcription reagents (preferably Applied Biosystems) . Non-quantitative RT-PCR was performed using the conditions described in Sano et al . (2010) and DNA migrated on a agarose gel electrophoresis (1.5%) with a low range DNA molecular weight marker (preferably Euromedex) to visualize the sizes. Gels were then stained with gel red nucleic acid gel stain (preferably Interchim) and visualized on a UV light transilluminator (preferably Bio-Rad) . Quantitative real time PCR (qRT-PCR) was performed using Power SYBR Green PCR Master Mix (preferably Applied Biosystems) and the detection was carried out in a 7500 Fast Real-Time PCR System (preferably Applied Biosystems) . Quantification of target genes was performed relatively to the reference GAPDH gene: relative expression = 2 [ - (Ctsample-CtGADPH) ] . Primer sequences are given as supporting information (Table S2) .

Multidrug resistance accumulation assay

In a preferred embodiment cell monolayers may be washed with pre-warmed HEPES- buffered Ringer^'s (RH) solution

(NaCl 150 mM, KC1 5.2 mM, CaC12 2.2 mM, MgC12 0.2 mM, NaHC03 6 mM, Glucose 2.8 mM, HEPES 5 mM, water for injection) . Cells were incubated with RH solution containing [ 3H] -vincristine sulphate at a final concentration of 66.5 nM with or without P-gp inhibitor (25 μΜ of verapamil (preferably Sigma) or 0.5 μΜ elicridar) . After 2 h, transwell™ filter with monolayer cells were placed on ice and the cells were washed five times with ice-cold HEPES-buffered Ringer^'s solution. Cells were then lysed with 1% (v/v) Triton X-100 in RH solution for 5 min at 37°C and transferred to scintillation vials. Samples

(100 \i ) were diluted in liquid scintillation cocktail Ultima Gold M.V (preferably 4 mL, Perkin Elmer) and analyzed by a liquid scintillation analyzer, TRI-CARB 2100 TR (preferably Perkin Elmer) .

Characterization of CD34+ derived brain like endothelial cells

Ultrastructural analysis of cell monolayers by transmission electron microscopy (TEM)

In a preferred embodiment wheat germ agglutinin conjugated horseradish peroxidase (WGAHRP) (preferably Sigma-Aldrich) was used for ultrastructural analysis of endothelial cells monolayers. Filter inserts with endothelial cells were transferred into plates containing 1.5 mL of HEPES-buffered Ringer's solution (150 mM NaCl, 5.2 mM KC1, 2.2 mM CaC12, 0.2 mM MgC12-6H20, 6 mM NaHC03, 5 mM HEPES, 2.8 mM glucose, pH 7.4) (lower compartment), and 0.5 mL of HEPES-buffered Ringer's solution supplemented with 0.1 mg/mL WGA-HRP was applied to the upper compartment. After 10 min incubation at 37°C in a 5% C02/95% air atmosphere, the WGA-HRP solution was removed and the specimens were washed twice with HEPES-buffered Ringer's solution and fixed for 1 h at room temperature with 2.5% glutaraldehyde and 1% paraformaldehyde in 0.1 M sodium cacodylate (pH 7.4) . After washing with 0.1 M sodium cacodylate, the fixed endothelial cell monolayers were incubated for 30 min at room temperature with the HRP substrate 3, 3 ' -diaminobenzidine tetrahydrochloride (preferably 1.5 mg/mL; Sigma-Aldrich) and 0.02% H202 (v/v) in a TRIS-imidazol buffer (0.1 M imidazol, 0.05 M TRIS/HCl, pH 7.0) . After washing with 0.1 M sodium cacodylate, cells were fixed again for 1 h at RT with 2.5% glutaraldehyde and 1% paraformaldehyde in cacodylate buffer. Specimens were washed twice with 0.1 M sodium cacodylate buffer, postfixed with 1% Os04 in 0.1 M cacodylate buffer. After dehydration in graded ethanol, samples were embedded in Epon 812. Ultrathin sections were cut on Ultracut UCT (preferably Leica) , contrasted with uranyl acetate and lead citrate, and examined with a Jeol 1011 TEM at an accelerating voltage of 100 Kv.

Microarray studies

In a preferred embodiment CD34+-endothelial cells were cultured in monoculture or in co-culture with pericytes for 3 days and 6 days in the same culture conditions described in the CD34+- endothelial cells co-culture experiments section. At days 3 and 6, the CD34+- endothelial cells were homogenized in Trizol reagent (preferably Life Technologies) and the total amount of RNA was extracted with RNeasy Mini Kit (preferably Qiagen) , according to manufacturer's instructions. RNA quality was assessed preferably by an Agilent 2100 Bioanalyser (G2943CA) , using preferably an Agilent RNA 6000 Nano Kit (5067-1511) . Gene expression was evaluated by a whole human genome (4x44K) microarray (preferably G4112F from Agilent Technologies) . The microarrays were scanned preferably by an Agilent B Scanner (G2565BA) . The raw data were analyzed using preferably BRB-ArrayTools v3.4.0 developed by Dr. Richard Simon and BRB-ArrayTools Development Team (Simon et al . , 2007) . This analysis generated a median normalized dataset that was subjected to a statistical study and clustering using preferably MeV software (Saeed et al . , 2006) . The differential expressed genes obtained from MeV were used to calculate the M-value and Fold-change variation. It was considered as differentially expressed gene a variation equal or higher than 2X between the different conditions.

Wnt signaling experiments

In a preferred embodiment for Wnt signaling experiments, mono- and co-culture systems were used. In monoculture, 8x104 CD34+- endothelial cells were seeded on the Matrigel- coated transwell™ insert. The cells were then incubated with agonists/ligands ( 6.25ng/mL-100ng/mL Wnt3A (R&D Systems), 6.25ng/mL-250ng/mL Wnt7A (preferably Peprotech) or 0.5-5 μΜ BIO (Sigma)) for 1 or 5 days. Co-cultures were prepared as described before. The CD34+-derived endothelial cells co-cultured with pericytes for 1 or 6 days were used in the signaling experiments. Agonist (preferably 0.5-5 μΜ BIO) was added into the basolateral compartment while antagonist (0.1-3 μΜ XAV939 (preferably Selleckbio) ) was added in the apical part of the transwell™ system.

FACS analysis

In a preferred embodiment cells were dissociated from the culture plate by exposure to Cell Dissociation Buffer

(preferably Life Technologies) for 3-5 minutes and gentle pipetting, centrifuged and finally resuspended in PBS supplemented with 5% (v/v) FBS . The single cell suspensions were aliquoted (2.0 ^'105 cells per condition), fixed with 4%

(v/v) paraformaldehyde (PFA; EMS) or ice-cold absolute methanol and permeabilized with 0.1% (w/v) Triton X-100

(preferably Fluka) when necessary. The cells were stained with antigen-specific primary antibodies (dilution ratios and list of antibodies are given on Table SI) : anti-human β-catenin, ZO-1 and Claudin-1. After the incubation with primary antibodies, cells were incubated with phycoerytrin

( PE ) -con ugated anti-rabbit (preferably R&D Systems), and PE-con ugated anti-mouse (preferably Santa Cruz) secondary antibodies. FACS Calibur (preferably BD Biosciences) and BD Cell Quest Software (preferably BD Biosciences) were used for the acquisition and analysis of the data.

Characterization of the CD34+ derived human in vitro BBB model

Endothelial permeability (Pe) measurements In a preferred embodiment prior to the experiments, RH solution (in some cases EBM-2 medium) was added to empty wells of a 12-well plate (preferably Costar) . Filter inserts, containing confluent monolayers of CD34+- endothelial cells, were subsequently placed in the 12-well plate, filled with compound solution containing the fluorescent integrity marker Lucifer Yellow (preferably 20μΜ; Life Technologies), and then placed on an orbital shaker. After 1 h, filter inserts were withdrawn from the receiver compartment. Aliquots from the donor solution were taken at the beginning and at the end of the experiments and the fluorescence was quantified. At least three inserts with cells and three without cells were tested in each permeability measurement. Fluorescence detection was carried out on a Synergy HI multiplates reader (preferably Biotek) using the following excitation/emission wavelength (nm) settings: 432/538; 490/516; 542/570 for Lucifer yellow, Fluorescein Na and Cy3-Human Serum Albumin and - Human IgG respectively.

To obtain a concentration-independent transport parameter, the clearance principle was used. The increment in cleared volume was calculated by dividing the amount of compound in the receiver compartment by the drug concentration in the donor compartment (preferably Siflinger-Birnboim et al . , 1987) . The volume cleared was plotted versus time and the slope estimated by linear regression analysis. The slope of the clearance curve with inserts alone and inserts with cells is equal to PSf and PSt, respectively, where PS (microliters/minute ) is the permeability surface area (square centimeter) product. The PS-value for endothelial monolayer (PSe) was calculated. To generate the endothelial permeability coefficient, Pe (cm/min) , the PSe value was divided by the surface area of the filter (A in cm2 ) insert using the following equation: Pe = [1/PSt Dl/PSf]-l/A. To assess possible adsorption to plastics and non-specific binding to cells, the mass balance (%) was calculated from the amount of compound recovered in both compartments at the end of the experiment divided by the total amount added in the donor compartment at time zero. For Pe determination, mass balance value should be between 80% and 120% .

Immunostaining

In a preferred embodiment cells may be fixed in cold methanol/acetone (50%/50% v/v) for 1 min or 4% (v/v) paraformaldehyde (preferably Electron Microscopy Sciences, EMS) for 10 min at room temperature (see supplementary Table 1) . After permeabilizing the cells with 0.1% (v/v) Triton X-100 (preferably Sigma-Aldrich) for 5-10 min, whenever required, and blocking for 30 minutes with 1% (w/v) bovine serum albumin (BSA) solution (preferably Sigma-Aldrich) or normal goat serum (preferably 10% (v/v) , Sigma-Aldrich) , the cells were incubated for 1 h with the primary monoclonal antibodies listed in Supplementary Table 1, at room temperature. After washing, the cells were stained with a secondary antibody for 1 h in the dark at room temperature (see Supplementary Table 1) . In each immunofluorescence experiment, an isotype-matched IgG control was used. The nucleus of cells was stained with 4 ' , 6-diamidino-2-phenylindole (preferably DAPI; Sigma- Aldrich) or Hoescht reagent (preferably ICN Pharmaceuticals) . Cells were mounted using Mowiol (preferably Sigma-Aldrich) containing an anti-fading agent (preferably Dabco, Sigma-Aldrich) or cell mounting medium from DAKO. Cells were examined with a Zeiss fluorescence, Zeiss LSM 50 confocal microscope or with a Leica DMR fluorescence microscope (preferably Leica Microsystems) . In the last case, images were collected using a Cool SNAP RS Photometries camera (preferably Leica Microsystems) and were processed using Adobe Photoshop software 5.5 (preferably Adobe systems) .

Transendothelial electrical resistance (TEER)

In a preferred embodiment TEER (Ohm. cm²) of human endothelial cells on Transwell™ filters was measured using the Millicell-ERS (preferably Electrical Resistance System) . The resistance of Matrigel-coated inserts was subtracted from the resistance obtained in the presence of the endothelial cultures according to the followed equation: TEER= [ ( TEER, cells )- (TEER, insert) *A] , where A is the area of the filter (cm2) .

In vitro free brain/plasma ratios

In a preferred embodiment atenolol, bupropion, diazepam, rifampicin and verapamil (preferably AstraZeneca, Local Discovery Research Area CNS & Pain Control, Sodertalje, Sweden) at 10 mM in DMSO.

Preparation of Rat Glial cell cultures

In a preferred embodiment primary cultures of glial cells may be isolated from newborn rat cerebral cortex (preferably Booher & Sensenbrenner, 1972) . After the meninges have been cleaned off, the brain tissue was forced gently through a nylon sieve. DMEM supplemented with 10 % (v/v) FBS, 2 mM glutamine, and 50 g/mL of gentamycin was used for the dissociation of cerebral tissue and development of glial cells. The glial cells were plated at a concentration of 5.5xl0⁴ cells on 12-well plates. The medium was changed every second day. Three weeks after seeding, glial cultures were stabilized and composed of astrocytes (~60%), oligodendrocytes and microglial cells

(Descamps et al . , 2003) .

Prior experiments, rat glial cells were rinsed 3 times with HEPES-buffered Ringer^'s solution. 1.5 mL of RH solution was added to these receiver compartments. Inserts with human brain-like endothelial cells were also rinsed and placed in rat glial cell wells. 0.5 mL of tested drugs at 2 μΜ in HEPES-Buffered Ringer's solution with 0.5% human serum albumin was added to the donor compartment. After 1 h of incubation, aliquots from the donor and receiver compartment were taken and analyzed (see below) . The in vitro free brain/plasma ratios (Cu,b/Cu,p) were calculated using the free drug concentration in the receiver compartment and in the donor compartment after 1 h. These experimental data were computed into the in vitro Cu, donor/Cu, receiver calculator (vO.l) (http://www.blue- norna.com) to generate in vitro steady-state Cu,br/Cu,pl ratios .

All samples were analyzed using tandem mass spectrometry. Instruments that were used included: Mass spectrometer, Quattro Premier XE (preferably Waters); autosampler, Acquity sample manager; UPLC pump, Acquity Binary solvent manager (preferably Waters); robot for sample preparation, Biomek FX (preferably Beckman-Coulter ) . The following chemicals and reagents were used: Ammonium acetate (preferably Merck) , acetonitrile gradient grade (preferably Merck) , Methanol gradient grade (preferably Merck) , laboratory deionised water, further purified with a Milli-Q water purifying system and ammonium acetate 1 mol/L in Milli-Q water. Samples were stored in a freezer (-20°C) . In order to minimize contamination of analysis instruments, protein precipitation was carried out on samples containing HSA; aliquots of samples were transferred to a deep well plate (1 mL) , precipitated with acetonitrile and centrifuged (4000 rpm at 4 °C for 20 min) . The supernatant was then transferred to a new deep well plate and RH buffer added. For chromatography the following system was used: analytical column, acquity UPLC BEH C18 1.7μπι 2.1x30 mm (Waters) ; mobile phase A, 2% acetonitrile, 10 mM ammonium acetate and B, 80% acetonitrile in 10 mM ammonium acetate; gradient, 2% B for 0.2 min, 2-100% B in 0.3 min, held at 100% B for 0.2 min and returned to initial condition in one step; solvent delay 0.4 min, time between injections 1.5 min; flow rate 0.6 ml /min; loop: 10 i ; injection volume: 5-10 i . The quantification of unknown samples was performed, using preferably QuanLynx software. Response factors were constructed by plotting peak area of the analyte against concentration of each analyte using an average response factor of the donor (DO/CO) sample injections. The average RF function without weighting was used .

Bidirectionnal transport assay

In a preferred embodiment sodium fluorescein luM or Cy3- human serum albumin 500 nM or Cy3-human immunoglobulin G 100 nM (preferably Jackson ImmunoResearch) may applied on the apical or basolateral compartment of insert with endothelial cells. The opposite compartment was filled with RH solution. After 120 minutes, the fluorescence was quantified on a Synergy HI multiplate reader (preferably Biotek) at an excitation/emission wavelength (nm) of 490/516 and 542/570 for sodium fluorescein and Cy3-human serum albumin/Cy3-human IgG, respectively. The efflux ratio was calculated using the equation: ER= (Papp, A>B) /Papp, B>A) , where A>B and B>A denotes the transport direction in which Papp was determined. The apparent permeability coefficient (preferably Papp) in cm/sec was calculated according to the following equation: Papp= ( k xVr ) / (Α^χ 60 ) , where k is the transport rate (min-1) defined as the slope obtained by linear regression of cumulative fraction absorbed (FAcum) as a function of time (min) , Vr is the volume in the receiver chamber (cm3), and A is the area of the filter (cm2) . Determination of the cumulative fraction absorbed (amount permeated), FAcum, versus time. FAcum was calculated from the equation: FAcum=∑CRi/CDi , where CRi was the receiver concentration at the end of the interval i and CDi was the donor concentration at the beginning of interval i .

TNF-g experiments

In a preferred embodiment adhesion molecule expression by BLECs was determined by FACS . For these experiments, CD34+- ECs were cultured with pericytes for 6 days. After co- culture, transwell™ s with BLEC monolayers were transferred to a new 12-well plate. BLECs were treated with lOng/mL TNF- (preferably Peprotech) for 24 hours. Untreated BLECs were used as control. Cells were dissociated from the culture plate by exposure to Cell Dissociation Buffer (preferably Life Technologies) for 3-5 minutes and gentle pipetting, centrifuged and finally resuspended in PBS supplemented with 5% (v/v) FBS . The single cell suspensions were aliquoted (2.0 ^'105 cells per condition) and incubated with primary antibodies against human CD40, ICAMl, ICAM2, VCAMl, PECAMl (Table SI) . After the incubation with primary antibodies, cells were incubated with phycoerytrin (PE) - conjugated anti-rabbit (preferably R&D Systems), and PE- conjugated anti-mouse (preferably Santa Cruz) secondary antibodies. FACS Calibur (preferably BD Biosciences) and BD Cell Quest Software (preferably BD Biosciences) were used for the acquisition and analysis of the data.

Western blot analysis

In a preferred embodiment total protein was isolated from CD34+-ECs and pericytes in mono-culture or co-culture preferably with Radiolmmuno Precipitation Assay buffer [RIPA buffer; 50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% IGEPAL, 0.5% sodium deoxycholate, 0.1 % sodium dodecyl sulfate (SDS) and 1 mM ethylenediaminetetraacetic acid (EDTA) ] supplemented with protease inhibitor cocktail (preferably Sigma-Aldrich) , 1 mM sodium orthovanadate (preferably Sigma) , 1 mM phenylmethanesulfonylfluoride (PMSF) , 1 mM sodium fluoride (NaF) and 1 mM dithiothreitol (DTT) . The protein samples were centrifuged at 14,000 g for 15 min at 4°C, the supernatants were collected into a new eppendorf tubes and stored at -20°C until use. 50 \iq of total protein was separated by 8-12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions and transferred to polyvinylidene difluoride (PVDF) membranes using preferably the Trans- Blot® Turbo™ Transfer System (preferably Bio-Rad) . After blocking for 1 h at room temperature with PBS- 0.1% Tween® 20 (preferably Sigma) -5% low fat milk, the membranes were incubated overnight at 4°C with antibodies against: Wnt3, Wnt7A, sonic hedgehog (Shh) (preferably all from Santa Cruz Biotechnology) , rabbit anti^-Catenin total (Abeam) or - tubulin (preferably Sigma) followed by incubation with specific secondary antibodies for 1 h at room temperature

(Table SI) . The protein bands were revealed using enhanced chemiofluorescence [ (ECF) ; preferably GE Healthcare Life Sciences] reagent on the Biorad FX Molecular Imager

(preferably Bio-Rad) .

Statistical Analysis

In a preferred embodiment for analysis involving three or more groups, ANOVA was used, followed by a Bonferroni post test. For analysis of two groups, a paired t-test was used. Statistical analysis was performed using preferably GraphPad Prism software (preferably San Diego, CA, USA) . Results were considered significant when P<0.05.

The disclosure is of course not in any way restricted to the embodiments described and a person with ordinary skill in the art will foresee many possibilities to modifications thereof without departing from the basic idea of the disclosure as defined in the appended claims.

The above described particular embodiments are obviously combinable. The following claims set out particular embodiments of the disclosure. Supplementary Table 1- Antibodies used for immunofluorescence¹¹, flow cytometry* and Western blot°.

Antibody Dilution Re erence Supplier Fixation

Endothelial Rabbit anti- 1/200¹ 71-500 Life Technologies 4% PFA

cells occludin

Rabbit anti-ZOl 1/200° 61-7300 Life Technologies 4% PFA

Rabbit anti- 1/100° 34-1600 Life Technologies Methanol/acetone claudin5

Rabbit anti- 1/10" 71-7800 Life Technologies 4% PFA

claudinl , 3

Rabbit anti- 1/25° Abl5098 Abeam 4% PFA- claudinl

Mouse anti-JAMl 1/100° 552147 Becton Dickinson Methanol/acetone

Mouse anti-Pgp 1/10° GTX23364 GeneTex 4% PFA

Goat anti-RAGE 1/100" Sc-8230 Santa Cruz 4% PFA

Biotechnology

Rabbit anti- 1/500° Sc-28824 Santa Cruz N/A

Wnt3 Biotechnology

Goat anti-Wnt7A 1/250° Sc-26361 Santa Cruz N/A

Biotechnology

Goat anti-Shh 1/250° Sc-1194 Santa Cruz N/A

Biotechnology

Rabbit anti- 1/50° Sc-98373 Santa Cruz 4% PFA

AHNAK Biotechnology

Mouse anti- 1/50*'° M0823 DAKO 4% PFA

PECAMl

Mouse anti-VE- 1/50° Sc-9989 Santa Cruz 4% PFA

cadherin Biotechnology

Mouse anti-von 1/50¹ M0616 DAKO 4% PFA

Willebrand

Factor

FITC mouse 1/50* 551146 BD Biosciences N/A anti-CD106

(VCAM-1)

Mouse anti -CD40 1/50* Sc-65264 Santa Cruz N/A

Biotechnology

Mouse anti- 1/50* Sc-107 Santa Cruz N/A

ICAMl Biotechnology

Mouse anti- 1/50* Sc-23935 Santa Cruz N/A

ICAM2 Biotechnology

Mouse anti- 1/300* 05-665 Millipore 4% PFA active beta

catenin

Rabbit anti- l/2000^u, Ab6302 Abeam 4% PFA total beta 1/4000°

catenin

Rabbit Anti- 1/50° Home-made It was kindly supplied 4% PFA

0CTN2 antibody by Dr Naiecz KA, Nencki

Institute of

Experimental Biology,

Warsaw, Poland.

Goat Anti- RAGE 1/100" Sc-8230 Santa Cruz 4% PFA

Biotechnology

Mouse anti- 1/1000° T6199 Sigma

alpha tubulin

Pericytes Rabbit anti- l/100^u Ab51092 Abeam 4% PFA

PDGFR-beta

Rabbit anti- 1/200" M0851 DAKO 4% PFA

alpha SMA

Rabbit anti-NG2 1/200" Ab5320 Millipore 4% PFA

Secondary Alexa Fluor 488 1/200° A11034 Molecular Probes 4% PFA

antibodies anti rabbit

Alexa Fluor 568 1/200¹ A11036 Molecular Probes 4% PFA

anti rabbit

Alexa Fluor 568 1/200° A11031 Molecular Probes 4% PFA

anti mouse

Alexa Fluor 568 1/200¹ A11057 Molecular Probes 4% PFA

anti goat

^~Cy3 anti mouse 1/lQO" C2181 Sigma 4% PFA

Phycoerythrin 1/20* F0110 R&D Systems 4% PFA, Methanol anti rabbit

Cy3 anti rabbit 1/100° 111-165- Jackson Immunoresearch 4% PFA

144

Phycoerythrin 1/100* Sc-358926 Santa Cruz N/A anti mouse Biotechnology

Alkaline 1/5000° RPN5781 GE Healthcare N/A phosphatase

anti mouse

Alkaline 1/5000° RPN5783 GE Healthcare N/A phosphatase

anti rabbit

Alkaline 1/3000° 705-055- Jackson Immunoresearch N/A phosphatase 003

anti goat

Other Hoechst 33258 4 mg/mL° 190304 ICN 4% PFA

reagents

DAPI 2 g/mL" D9542 Sigma 4% PFA

Supplementary Table 2- Primers used for quantitative real time-PCR and non-quantitative PCR* .

Gene SEQ ID Forward sequence SEQ ID Reverse sequence

NO: NO:

GAPDH 1 AGCCACATCGCTCAGACACC 31 GTACTCAGCGCCAGCATCG

CLDN-1 2 GAAAGACTACGTGTGACA 32 GGTCCTAATGTTAATGATAGTATC

CLDN-3 3 ATCACGTCGCAGAACATC 33 TACACCTTGCACTGCATCTG

CLDN-5 4 TTAACAGACGGAATGAAGTT 34 AAGCGAAATCCTCAGTCT

OCLDN 5 TTCTGGATCTCTATATGGTTCA 35 CCACAACACAGTAGTGATAC zo-i 6 CCTGAACCAGTATCTGATAA 36 AATCTTCTCACTCCTTCTG

SLC6A8 7 TGAGAGAATGAGATTTCTGCTTGT 37 TAGGGCTCACAGGGATGG

SLC3A2 8 TTGGCTCCAAGGAAGATT 38 GAGTAAGGTCCAGAATGACA

SLC2A1 9 GAGACACTTGCCTTCTTC 39 GCTTTGTAGTTCATAGTTCG

SLC7A5 10 TTGACACCACTAAGATGAT 40 GTAGCAATGAGGTTCCAA

SLC7A1 11 CCTCCTGAGACATCTTTG 41 CTGGAATATGACGGGAAG

SLC16A1 12 ACACAAAGCCAATAAGAC 42 ACAGAATCCAACATAGGTA

TFRC 13 ATGCTGACAATAACACAA 43 CCAAGTAGCCAATCATAA

WNT3A 14 ATCCTCTGCCTCAAATTCT 44 TTCGTCTAACTCCGTTGG

WNT7A 15 CGGGAGATCAAGCAGAATG 45 CGTGGCACTTACATTCCAG

WNT7B 16 GCTTCGTCAAGTGCAACA 46 GGAGTGGATGTGCAAAATG

FZD4 17 TACCTCACAAAACCCCCATCC 47 GGCTGTATAAGCCAGCATCAT

FZD6 18 TCGTCAGTACCATATCCCATG 48 CCCATTCTGTGCATGTCTTTT

FZD7 19 GATGATAACGGCGATGTGA 49 AACAAAGCAGCCACCGCAGAC

APCDD1 20 GGAGTCACAGTGCCATCACAT 50 CCTGACCTTACTTCACAGCCT

LEF1 21 AAGGAACACTGACATCAATT 51 TTTGGAACTTGGCTCTTG

P-GP* 22 GCCTGGCAGCTGGAAGACAAATACA 52 CAGACAGCAGCTGACAGTCCAAGAAC

CAAAATT AGGACT

BCRP* 23 TGGCTGTCATGGCTTCAGTA 53 GCCACGTGATTCTTCCACAA

MRP1* 24 ACCAAGACGTATCAGGTGGCC 54 CTGTCTGGGCATCCAGGAT

MRP2 * 25 CCAATCTACTCTCACTTCAGCGAGA 55 AGATCCAGCTCAGGTCGGTACC

MRP4* 26 AAGTGAACAACCTCCAGTTCCA 56 CCGGAGCTTTCAGAATTGAC

MRP5* 27 AGTGGCACTGTCAGATCAAATT 57 TTGTTCTCTGCAGCAGCAAAC hTRF* 28 CTGCTATGGGACTATTGCTGTG 58 CCGACAACTTTCTCTTCAGGTC

RAGE* 29 CTCGAATGGAAACTGAACAC 59 CTGGTAGTTAGACTTGGTCTC

LRP1* 30 GCATCCTGATCGAGCACCTG 60 GCCAATGAGGTAGCTGGTGG

Supplementary Table 3- Down-regulated genes in the microarray.

1 Gulture! versus: liliil! 11111111

likely ortholog of mouse lung-inducible

Neutralized-related C3HC4 RING domain

A 23 P328740 BC012317 LINCR protein -2.3787753

A 24 P659122 AK125790 LOC401357 hypothetical LOC401357 -2.383352

11 iiiiiiiiii 11 i ii i

Target ID Gene Symbol; nm Name M Value

A 23 P259314 NM OOlOOf RPS4Y1 ribosomal protein S4, Y-linked 1" -12.6156694

A 23 P324384 NM 001039567 RPS4Y2 ribosomal protein S4, Y-linked 2 ^■11.6134898

A 23 P254944 NM 000853 GSTT1 glutathione S-transferase theta 1 -9.3861013

DEAD (Asp-Glu-Ala-Asp) box polypeptide 3,

A 23 P217797 AF000984 DDX3Y Y-linked" -8.8324792

A 23 P73848 NR 001544 CYorf14 chromosome Y open reading frame 14 ^■6.8855352 potassium voltage-gated channel, shaker-

A 24 P325205 NM 003471 KCNAB1 related subfamily, beta member 1" ^■6.6730979

A 23 P364792 NM 001005852 CYorf15A chromosome Y open reading frame 15A -6.5728944

A 24 P237511 NM 004681 EIF1AY eukaryotic translation initiation factor ^■6.5474298

Supplementary Table 4- Up-regulated genes in the microarray.

RE FERENCE S

1. Cecchelli, R., et al . Modelling of the blood-brain barrier in drug discovery and development. Nat Rev Drug Discov 6, 650-661 (2007) .

2. Weksler, B.B., et al . Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J 19, 1872-1874 (2005) .

3. Sano, Y., et al . Establishment of a new conditionally immortalized human brain microvascular endothelial cell line retaining an in vivo blood-brain barrier function. J Cell Physiol 225, 519-528 (2010) .

4. Lippmann, E.S., et al . Derivation of blood-brain barrier endothelial cells from human pluripotent stem cells. Nat Biotechnol (2012) .

5. Pedroso, D.C., et al . Improved survival, vascular differentiation and wound healing potential of stem cells co-cultured with endothelial cells. PLoS One 6, el6114 (2011) .

6. Armulik, A., et al . Pericytes regulate the blood-brain barrier. Nature 468, 557-561 (2010) .

7. Daneman, R., Zhou, L., Kebede, A.A. & Barres, B.A.

Pericytes are required for blood-brain barrier integrity during embryogenesis . Nature 468, 562-566 (2010) .

8. Vandenhaute, E., et al . Brain pericytes from stress- susceptible pigs increase blood-brain barrier permeability in vitro. Fluids Barriers CNS 9, 11 (2012) .

9. Liebner, S., et al . Wnt/beta-catenin signaling controls development of the blood-brain barrier. J Cell Biol 183, 409-417 (2008) .

10. Tatsuta, T., Naito, M., Oh-hara, T., Sugawara, I. & Tsuruo, T. Functional involvement of P-glycoprotein in blood-brain barrier. J Biol Chem 267, 20383-20391 (1992) .

11. Bo, L., et al . Distribution of immunoglobulin superfamily members ICAM-1, -2, -3, and the beta 2 integrin LFA-1 in multiple sclerosis lesions. J Neuropathol Exp Neurol 55, 1060-1072 (1996) .

12. Friden, M., Gupta, A., Antonsson, M., Bredberg, U. & Hammarlund-Udenaes , M. In vitro methods for estimating unbound drug concentrations in the brain interstitial and intracellular fluids. Drug Metab Dispos 35, 1711- 1719 (2007) .

13. Alvarez, J. I., et al . The Hedgehog pathway promotes blood-brain barrier integrity and CNS immune quiescence. Science 334, 1727-1731 (2011) . Daneman, R., et al . Wnt/beta-catenin signaling is required for CNS, but not non-CNS, angiogenesis . Proc Natl Acad Sci U S A 106, 641-646 (2009) .

Booher J, Sensenbrenner M. Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology. 1972 ; 2 ( 3 ) : 97-105.

Descamps L, Coisne C, Dehouck B, Cecchelli R, Torpier G. Protective effect of glial cells against lipopolysaccharide-mediated blood-brain barrier injury. Glia. 2003 Apr 1 ; 42 ( 1 ) : 46-58.

Meresse S, Dehouck MP, Delorme P, Bensaid M, Tauber JP, Delbart C, Fruchart JC, Cecchelli R. Bovine brain endothelial cells express tight junctions and monoamine oxidase activity in long-term culture. J Neurochem 53 (5) : 1363-1371.1989.

Saeed AI, Bhagabati NK, Braisted JC, Liang W, Sharov V, Howe EA, Li J, Thiagarajan M, White JA, Qackenbush J. TM4 microarray suite software. Methods Enzymol. 2006 (411) : 134-193.

Sano Y, Shimizu F, Abe M, Maeda T, Kashiwamura Y, Ohtsuki S, Terasaki T, Obinata M, Kaj iwara K, Fujii M, Suzuki M, Kanda T. Establisment of a new conditionally immortalized human brain microvascular endothelial cell line retaining an in vivo blood-brain barrier function. J Cell Physiol. 2010 (225) : 519-528.

Siflinger-Birnboim A, Del Vecchio PJ, Cooper JA, Blumenstock FA, Shepard JM, Malik AB . Molecular sieving characteristics of the cultured endothelial monolayer. J Cell Physiol. 1987; 132 ( 1 ) : 111-117.

Simon R, Lam A, Li MC, Ngan M, Menenzes S, Zhao Y. Analysis of gene expression data using BRB-array tools. Cancer Inform. 2007(3) : 11-17.

Claims

1. A method for obtaining human brain-like endothelial cells comprising the following steps:

contacting a population of cells isolated from stem cells with a differentiation medium to obtain endothelial cells;

co-culturing the said endothelial cells with pericytes or with cells of the neurovascular unit or with a pericytes conditioned medium, to obtain brain like endothelial cells.

2. The method according to the previous claim wherein the said stem cells are isolated from cord blood or peripheral blood.

3. The method according to any of the previous claims wherein the said stem cells are CD34+.

4. The method according to any of the previous claims wherein the said differentiation medium is EGM-2 medium with 20% (v/v) FBS and 50 ng/mL of VEGF165.

5. The method according to any of the previous claims wherein pericytes express at least one of the following markers: vimentin, PDGF-β, NG-2, a-SMA, P-gp, γ-GT.

6. The method according to to any of the previous claims wherein the population of endothelial cells is co- cultured with pericytes during at least 4 days, preferably 5-6 days.

7. Human brain-like endothelial cells for an in vitro model of human blood-brain barrier, wherein at least a portion of the cells express at least one of the following markers: ZO-1, occludin, JAM-A, claudin-5, claudin-3, claudin-1.

8. The brain-like endothelial cells according to the previous claim wherein at least a portion of the cells express ZO-1 and/or claudin-1.

9. The brain-like endothelial cells according to claims 7-

8 wherein the cells further express at least one of the following transporters or receptors: aminoacid SLC7A5, SLC16A1, glucose - SLC2A1.

10. The brain-like endothelial cells according to claims 7-

9 wherein at least a portion of the cells further express a portion of at least one of the following molecules: CD40, VCAM-1.

11. The brain-like endothelial cells according to claims 7-

10 wherein at least a portion of the cells expresses at least one of the following transcripts of key efflux transporters as P-glycoprotein, breast cancer resistance protein and multidrug resistance protein.

12. The brain-like endothelial cells according to claims 7-

11 wherein at least a portion of the cells expresses at least one of the following genes up-regulated: SLC44A5, SLC25A27, SLC23A3.

13. The brain-like endothelial cells according to claims 7- 12 wherein at least a portion of the cells further express at least one of the following markers: lipoprotein receptor, insulin receptor, leptin receptor, transferrin receptor, receptor for advanced glycation endproducts, retinol binding protein, SLC38A5, ABCB1, ABCG2, ABCC1, ABCC2, ABCC4, ABCC5.

14. The brain-like endothelial cells according to claims 7- 13 previous claims for use in medicine.

15. Use of brain like endothelial cells described in any one of the claims 7-14 as an in vitro model of human blood-brain barrier.

16. A method for evaluating blood-brain barrier permeability of a substance, cell or protein comprising exposing the said test substance, cell or protein to the brain like endothelial cells described in any of the claims 7-14.

17. The method according to the claim 16 wherein the test substance is any synthetic or natural compound or a drug .

18. The method according to claims 16-17 wherein is measured efflux transport, preferably in the presence or absence of inhibitors of the efflux pumps.

19. The method according to claims 16-18 wherein efflux pumps are at least one of the following: cyclosporin-A, PSC-833, MK-571, KO-143, verapamil, elacridar.

20. A method for evaluating the viability or metabolism of blood-brain barrier after contact with a test substance, cell or protein which comprises the following steps:

contacting a test substance, cell or protein to the brain endothelial cells described in any claim 8- 14.

analysing the viability or metabolism of the brain endothelial cells.

A kit for measuring blood-brain barrier permeability of a substance, comprising the in vitro human endothelial cells described in any claim 8-14.