In Vitro Models of the Blood–Brain Barrier for the ... - ACS Publications

Mar 18, 2014 - Institute of Life Sciences, Vasile Goldis Western University of Arad, Romania. ABSTRACT: The most important obstacle to the drug delive...
0 downloads 0 Views 2MB Size
Review pubs.acs.org/molecularpharmaceutics

In Vitro Models of the Blood−Brain Barrier for the Study of Drug Delivery to the Brain Imola Wilhelm† and István A. Krizbai*,†,‡ †

Institute of Biophysics, Biological Research Centre, Hungarian Academy of Sciences, Temesvári krt. 62, 6726 Szeged, Hungary Institute of Life Sciences, Vasile Goldis Western University of Arad, Romania



ABSTRACT: The most important obstacle to the drug delivery into the brain is the presence of the blood−brain barrier, which limits the traffic of substances between the blood and the nervous tissue. Therefore, adequate in vitro models need to be developed in order to characterize the penetration properties of drug candidates into the central nervous system. This review article summarizes the presently used and the most promising in vitro BBB models based on the culture of brain endothelial cells. Robust models can be obtained using primary porcine brain endothelial cells and rodent coculture models, which have low paracellular permeability and express functional efflux transporters, showing good correlation of drug penetration data with in vivo results. Models mimicking the in vivo anatomophysiological complexity of the BBB are also available, including triple coculture (culture of brain endothelial cells in the presence of pericytes and astrocytes), dynamic, and microfluidic models; however, these are not suitable for rapid, high throughput studies. Potent human cell lines would be needed for easily available and reproducible models which avoid interspecies differences. KEYWORDS: blood−brain barrier (BBB), in vitro model, cerebral endothelial cell, tight junction, ABC transporter, SLC transporter, drug delivery



INTRODUCTION Disorders of the central nervous system (CNS) are among the diseases with the most severe consequences, from both individual and social points of view. However, our therapeutic possibilities are still limited. Recent advances in understanding the causes of CNS disorders opened the way to new therapeutic strategies and attracted the interest of the pharmaceutical industry. The worldwide market for CNS therapeutics was about 42 billion EUR (approximately 57 billion USD) in 2005,1 it is steadily increasing, and it is forecast that in the near future it could reach the value of 100 billion EUR/year (approximately 136 billion USD/year). This rapid increase may have among others the following causes:2 • The incidence of many CNS disorders (e.g., Alzheimer’s disease, stroke and Parkinson’s disease) increases exponentially after the age of 65. • The number of people in the world over 65 is increasing sharply. • It takes longer to get a CNS drug to market (12−16 years) compared with a non-CNS drug (10−12 years). The reason for this may lie in the complexity of the brain, the liability of CNS drugs to cause CNS side effects, and the requirement of CNS drugs to cross the blood−brain barrier (BBB). A significant number of potential drugs fail because of the poor penetration through the BBB. Therefore, it is of crucial importance to characterize the BBB permeability of a drug candidate as early as possible in the drug developmental © XXXX American Chemical Society

pipeline. In order to meet this need in vitro models of the BBB have been developed. This review will summarize the latest and most promising ones.



THE BLOOD−BRAIN BARRIER

The BBB (Figure 1) is an active interface between the circulation and the CNS, which strictly controls the molecular and cellular traffic between the blood and the brain, thus taking an important share in providing a steady state environment needed for the proper neuronal function. The BBB is formed by endothelial cells of cerebral capillaries and microvessels, coming in close contact with neighboring pericytes and glial cells. The surface of the strictly controlled material exchange is estimated to be around 10−20 m2. From the point of view of CNS drug development, the BBB is one of the major obstacles: due to its relative impermeability, it prevents therapeutic substances from reaching relevant concentrations in the CNS. Cellular Composition of the BBB. Endothelial Cells (ECs). The advent of high resolution electron microscopy techniques clearly identified the cerebral endothelium as the site of the barrier. This is made possible by the continuous line Special Issue: Engineered Biomimetic Tissue Platforms for in Vitro Drug Evaluation Received: January 17, 2014 Revised: March 11, 2014 Accepted: March 18, 2014

A

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

endothelial cells are far from being completely understood. Besides a physical contact, astrocytes synthesize a considerable number of biologically active molecules which may influence endothelial cells. These include TGF-β (transforming growth factor-β), GDNF (glial-derived neurotrophic factor), bFGF (basic fibroblast growth factor), IL-6,3 and Sonic hedgehog.4 Pericytes. Pericytes are localized in the duplication of the basement membrane and cover approximately 22−32%5 of the endothelial surface. Pericytes can secrete a large number of substances which may influence endothelial function including TGF-β, angiopoetin-1, and VEGF. It seems that the differentiation stage of pericytes determines their effect on the endothelium as well.6 The role of pericytes in the formation of the BBB is supported by the finding that the absence of pericytes leads to endothelial hyperplasia, abnormal vasculogenesis,7 and an increased BBB permeability.8 Molecular Structure of the Junctional Complex of the BBB. The principal structures determining the paracellular permeability of the brain endothelium are tight junctions and adherens junctions. Interestingly, although adherens junctions do not form a paracellular seal, they are indispensable for a functional barrier. Tight junctions (TJs). Tight junctions form a network of strands composed of a row of transmembrane proteins with extracellular domains joining together and sealing the intercellular cleft (Figure 1B). By preventing the free diffusion of ions, the continuous line of tight junctions interconnecting brain endothelial cells defines a transendothelial electrical resistance (TEER), which is considered the most sensitive marker of the tightness of the junctions. Transmembrane proteins of tight junctions can be classified in the members of the tetraspan Marvel and claudin families, immunoglobulin-like single span molecules, and non-immunoglobulin-like single span molecules. The TJ-associated Marvel family consists of occludin, MarvelD2 (tricellulin), and MarvelD3. Occludin was the first discovered integral transmembrane TJ protein (for review see ref 9). Since then several alternative splice variants have been discovered, but their exact role has not yet been completely unraveled. Occludin 1B is a longer form of occludin with an additional N-terminal sequence of 56 amino acids with similar distribution to normal occludin.10 An occludin isoform lacking the fourth transmembrane domain (occludin TM4(−)) has also been detected, and interestingly it is expressed in both epithelia and endothelia but only in primates and at low levels.11 Four splice variants were detected by Mankertz et al. designated type I, type II (identical to occludin TM4(−)), type III, and type IV in human epithelial tissues.12 More recently an exon 9 deleted occludin was discovered in liver cells involved in cell apoptosis and invasion, but their role in BBB function has not been studied so far.13 A special role in the regulation of occludin may have a highly conserved phosphorylation hotspot described between Tyr398 and Ser408 which is the target of several protein kinases and plays a crucial role in occludin oligomerization.14 Initially tricellulin, also known as MarvelD2, was described in epithelial cells concentrated at tricellular contacts, but it is also present in bicellular contacts. The protein has structural similarities and 32% homology with the C-terminus of occludin, and it has recently been detected in cerebral endothelial cells as well. Interestingly, besides occludin tricellulin is also expressed by astrocytes.15 MarvelD3 was described a few years ago and is expressed in both epithelial and endothelial cells.16 Although data originates

Figure 1. The cellular and molecular composition of the BBB. A: Cerebral capillaries are lined by endothelial cells coming in close contact with pericytes, astrocytes, and other cells of the neurovascular unit. B: Brain endothelial cells are interconnected by a continuous line of tight and adherens junctions. The transmembrane proteins of interendothelial tight junctions include claudin, Marvel, and JAM proteins. C: The most important ABC and SLC transporters of the human BBB.

of tight junctions interconnecting adjacent endothelial cells. From a morphofunctional point of view cerebral endothelial cells have characteristics of peripheral endothelial cells and epithelial cells as well. Besides expression of classical endothelial markers (factor VIII-related antigen, alkaline phosphatase, γglutamyl transpeptidase, uptake of acetylated-LDL) cerebral endothelial cells have special characteristics similar to epithelia: presence of continuous strands of tight junctions sealing the intercellular cleft between adjacent ECs, very low pinocytotic and transcytotic activity, expression of specific transporter systems, and presence of a high number of mitochondria suggesting high energy requirements of these cells. Due to their critical role in the formation of the barrier, cerebral endothelial cells are crucial and determining components of in vitro BBB models. Astrocytes. Although their role in the formation of the physical barrier is limited, due to their influence on cerebral endothelial cells they play an important role in the maintenance of the BBB. The coverage of the capillaries by astrocytic endfeetdepending on the brain regioncan be almost complete. Astrocytic endfeet express high levels of several specific proteins at their perivascular surface, like glucose transporter 1 (GLUT-1/SLC2A1), ABCB1, aquaporin-4, connexin-43, and the Kir 4.1 K + channel. Molecular mechanisms of the interaction between astrocytes and B

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

Table 1. Functions of the BBBa

a

The barrier function of the BBB is determined by the 4-fold defense line. The pictograms indicate the direction of transport through the brain endothelial cells (light pink squares) with the apical surface upward and the basolateral surface downward.

essential for proper epithelial development,21 and BVES/Pop1 cells colocalizes with ZO-1 and occludin in polarized epithelial cells. Peripheral Proteins of Tight Junctions. The transmembrane proteins of tight junctions are connected to the cellular cytoskeleton by peripheral tight junction proteins. These include proteins containing PDZ domains, like the zonula occludens proteins ZO-1 and ZO-2, and non-PDZ proteins, like cingulin or JACOP (junction-associated coiled-coil protein)/paracingulin. Interestingly, some of the peripheral TJ proteins not only have structural functions but actively participate in cellular signaling as well.22 In recent years a considerable number of proteins located to the junctional complex have been identified, but their role still needs to be clarified. Being a dynamic structure able to respond to environmental stimuli, the junctional complex is regulated by a sophisticated signaling network. Indeed, cerebral endothelial cells are equipped with complex signaling machinery,23 which is responsible for the fine-tuning of cellular permeability. Adherens Junctions. Adherens junctions are present in many cell types, but adherens junctions in endothelial cells have some specific features.24 The transmembrane protein of endothelial adherens junctions is VE-cadherin, a single span transmembrane protein capable of homophilic Ca2+-dependent binding. It is linked to the cytoskeleton through catenins (α, β, γ and p120). Another major protein complex of adherens junctions consists of the nectins linked to afadin/AF6. PECAM-

from epithelial cells, it seems that MarvelD3, tricellulin, and occludin have overlapping but nonredundant functions.17 Claudins also belong to the tetraspan family, however, they do not share sequence homology with occludin. They form a family of 24 members with molecular weights ranging from 20 to 27 kDa. Depending on the composition of the extracellular loops, claudins have distinct functions: some of them are tightening the paracellular barrier, while others are pore forming. In the cerebral endothelium the principal claudin is claudin-5, and its loss opens the barrier for molecules smaller than 800 Da.18 Further claudins detected in brain endothelial cells are claudin-1, -3, and -12; however, mRNAs of claudin-8, -10, -15, -17, -19, -20, -22, and -23 were also detected in low amounts.19 The best investigated family of immunoglobulin-like single span molecules are the junctional adhesion molecules (JAMs). JAM proteins are characterized by a single transmembrane domain and two extracellular loops. The best characterized JAM is JAM-A in cerebral endothelial cells participating in the formation of tight junctions and thus playing a role in the determination of paracellular permeability. JAM-B and JAM-C are also expressed in endothelial cells, and JAM-B has been shown to play a role in the transmigration of leukocytes.20 The role of ESAM (endothelial cell-selective adhesion molecule) in the formation of the BBB is largely unknown. The role of non-immunoglobulin-like single span molecules like CRB3 or BVES/Pop1 (blood vessel/epicardial substance) in the formation of the BBB is less well understood. CRB3 is C

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

or rat capillaries are approximately 3−4-fold higher than the ABCB1 levels of capillaries isolated from human or monkey brain.27,29,30 The greater expression of P-gp at the rodent BBB is accompanied by a relatively low expression of Abcg2/ Bcrp,31,32 the amount of which is rather closer to that of Mrp4. Instead, the human BBB abundantly expresses ABCG2/BCRP in an amount quite similar to that of ABCB1/MDR1/P-gp, while the expression level of ABCC4/MRP4 is 20−30-fold lower.30 Further differences have been detected in the expression of ABCC transporters. ABCC2 (Abcc2), ABCC4 (Abcc4), and ABCC5 (Abcc5) seem to be more abundant in rodent, porcine, and/or bovine brain capillaries in comparison to humans, which may influence permeability results obtained with drugs which are substrates of these transporters.30,33 These profiles of drug efflux transporter abundance and stoichiometry at the human BBB are in good agreement with those of the cynomolgus monkey and marmoset BBB rather than those of mouse or rat BBB.27,29,30,34 These discrepancies may influence drug distribution in the brain, and in this way it is possible to underestimate CNS concentration of drugs which are ABCB1 substrates by using rodent BBB models. Indeed, it has been shown in vivo that the brain/plasma ratio of verapamil (P-gp substrate) was 4.1-fold greater in monkeys than in rodents.34 The situation is further complicated by the fact that many drug candidates are substrates of both ABCB1 and ABCG2 transporters. Moreover, there exists a species difference not only in the expression level but in the substrate affinity of P-gp as well.35 Taken together, although there are a few contradictory datalike those of Enerson et al., who have found similar amounts of Abcg2 and Abcb1 in the rat BBB,36 and the results of Warren et al., who have described lower amounts of ABCG2 in human microvessels compared to rodent33the majority of studies indicate an overrepresentation of P-gp and an underrepresentation of Bcrp in rodent in comparison to human brain capillaries (Table 2). However, in a study comparing human and rat cell-based BBB models a similar compound ranking of 20 drugs between rat and human models

1 also participates in the formation of endothelial adherens junctions. Functions of the BBB. The main functions of the BBB are to protect the homeostasis and to serve the nutritional demands of the CNS. Thus, the BBB has a dual role: first, it is a barrier for cells and solutes, and second, it selectively transports various substances, e.g., nutrients essential to the brain (in the bloodto-brain direction) and potentially harmful metabolic products (in the brain-to-blood direction). An ideal biomimetic platform of the BBB has to mimic both the barrier and carrier functions. The main functions of the BBB are presented in Table 1. ATP-Binding Cassette (ABC) and Solute-like Carrier (SLC) Transporters of the BBB. Due to the continuous line of tight junctions interconnecting brain endothelial cells, the molecular traffic is forced to take the transcellular route through individual endothelial cells. Here, however, only small lipophilic molecules can freely pass, and several potential drug candidates fail to overcome the enzymatic barrier or the efflux transporters of brain endothelial cells. Importantly, transporters of the BBB may represent the rate-determining step of drug delivery into the brain. Therefore, in vitro models are needed to determine whether the drug is a substrate of key transporters. In human brain microvessels the main ABC transporters are ABCG2/ BCRP, ABCB1/MDR1, ABCC1, ABCC4, ABCC5, ABCA2, and ABCA8, while the most abundantly expressed SLC transporters are SLC2A1/GLUT1, SLC7A5/LAT1, SLC16A1/MCT1, SLC1A3/EAAT1, and SLC1A2/EAAT225 (Figure 1C). The human drug transporters of the BBB proposed by the International Transporter Consortiumbeing important for evaluation during drug development are the following: ABCB1/MDR1(multidrug resistance protein-1)/Pgp (P-glycoprotein), ABCG2/BCRP (breast cancer resistance protein), MRP4 (multidrug resistance related protein 4), MRP5, OATP1A2 (organic anion-transporting polypeptide 1A2), OATP2B1, ENT1 (equilibrative nucleoside transporter 1)/SLC29A1, and ENT2/SLC29A2.26



DIFFERENCES IN THE BBB OF DIFFERENT SPECIES The most widely used in vitro models are based on the culture of rodent or other non-human mammalian cells. Extrapolation of animal model data to humans requires an understanding of interspecies differences of tight junction proteins, transporters, enzymes, and specific receptors which influence BBB permeability. It has been shown that the expression level of claudin-5 in rats, cynomolgus monkeys, and marmosets is more than 2-fold higher than in humans.27 Since claudin-5 is considered the backbone of brain endothelial tight junctions, this might have important relevance. ABC transporters are also crucial players in CNS drug delivery. Unfortunately a large number of drugs or potential drug candidates are substrates of efflux transporters, which severely limits their usefulness in the therapy of CNS disorders. Therefore, even in the initial phase of drug development it is of primordial importance to investigate possible interactions between the drug candidates and efflux transporters. Here, in vitro systems could be very useful; however, species differences should also be taken into consideration (for review see 28). Detailed analyses on the mRNA and protein expression levels of several transporters in different species have revealed that there are pronounced differences between rodents and primates, with high degrees of similarity between mouse and rat, and among marmoset, cynomolgus monkey, and human, respectively. The Abcb1/Mdr1a/P-gp levels of isolated mouse

Table 2. Differences in the BBB of Primates and Rodents

D

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

was found, although with a 2-fold higher permeability in rat.37 On the other hand, efflux transporters have been probably evolved to transport dietary components as well, therefore, although not tested in detail, porcine efflux transporters might be closer to the human than the rodent or bovine ones. SLC transporters are also abundantly expressed at the level of the BBB. Besides transport of important nutrients to the brainlike glucose, amino acids, and ionsthey are able to transport different drugs as well. Therefore, a proper expression of these transporters is important for a BBB model meant to investigate drug transport through the BBB. Interestingly, recent investigations have revealed considerable species differences in this respect. SLC2A1 (glucose transporter 1, GLUT-1) is considered a specific marker of endothelial cells in the brain; however, it has been detected in human and monkey astrocytes as well.38 Proteomic analysis of isolated microvessels showed that SLC2A1 is expressed in approximately the same amount in the human and monkey,27,30,34 while its rodent counterpart is present in similar or faintly lower quantities in mouse or rat capillaries.27,29 More pronounced differences have been reported in the expression of amino acid transporters. For example SLC7A5 (Slc7a5), which is an important subunit of the large amino acid transporter 1 (LAT1, Lat1), and SLC16A1 (Slc16a1) (monocarboxylate transporter 1, MCT1, Mct1) are 5 times more abundant in mouse and rat capillaries than in primates.27 Interspecies discrepancies have been observed in the expression of excitatory amino acid transporters as well; however, mRNA and protein expression profile data are still contradictory.39 Data regarding interspecies differences in other SLC transporters are not well understood, especially in case of the OATP/SLCO (Oatp/Slco) superfamily, where the gene duplications especially in rodents complicate the direct extrapolation of the data to human. Recent investigations have shown that expression profiles of BBB-related metabolizing enzymes may also differ between humans and rodents, e.g., in case of the cytochrome P450 and glutathione-S-transferase families; however, few quantitative data are available.39 Our knowledge regarding the expression of BBB-specific receptors in different species is also very limited. It has been shown that the protein expression level of transferrin receptor 1 is more than 2-fold higher in rats than in humans, while the levels of insulin receptor and low-density lipoprotein receptorrelated protein 1 (LRP1) are similar.27 All these data indicate that interspecies differences might influence the results of drug permeability data in different models. In addition, several pathophysiological conditions and transporter polymorphisms might affect barrier properties of the BBB and transporter activities. Therefore, no model can be ideal. However, a careful choice of the modeldepending on the scope of the experimentscan make different BBB models very useful tools in drug delivery studies.

early stage drug discovery. In vitro models have the potential to be useful techniques which fulfill the most important criteria required by the pharmaceutical industry. A “perfect” model would mimic all characteristics of the in vivo situation; however, it is easy to understand that some reasonable compromises between costs, capacity, time, and predictive value have to be accepted.40 A useful BBB model should meet the following criteria:41 reproducibility of solute permeability, display of a restrictive paracellular pathway and physiologically realistic architecture, functional expression of transporters, and ease of culture. The tightness of the interendothelial junctions can be estimated from the transendothelial electrical resistance (TEER) (Figure 2A) and the apparent permeability of marker

RECENT ADVANCES IN CELL-BASED IN VITRO BBB MODELS What Do We Expect from a BBB Model? One of the most accurate methods to measure in vivo permeability of a drug is the in situ brain perfusion technique, which can give even more accurate prediction on human permeability if humanized transporter-expressing transgenic animals are used. However, in vivo experiments even using rodents are becoming more and more expensive and are not suitable for medium or high throughput screening; therefore, they are rarely used in

molecules (e.g., mannitol, sucrose, inulin), which should be as close as possible to the values measured in vivo. The TEER in the rat brain is considered to be about 2000 Ω·cm2,42 while the apparent in vivo endothelial permeability of sucrose can be as low as 0.03 × 10−6 cm/s.43 The function of the transporters can be characterized by the measurement of the efflux ratio of different substances. The efflux ratio is calculated by performing bidirectional (i.e., apical-to-basolateral and basolateral-to-apical) permeability assays, as indicated in Figure 2B. If the efflux ratio is greater than 1.5−2 or lower than 0.5, this indicates an active

Figure 2. Measurement of TEER and efflux ratio. Cerebral endothelial cells are cultured on semipermeable inserts. A: The transendothelial electrical resistance is measured by placing one electrode into the apical and another into the basolateral chamber. B: By measuring the concentration of the test compound in the acceptor chamber one can calculate the permeability of the test compound. The efflux ratio is the ratio between the basolateral-to-apical and the apical-to-basolateral permeability.



E

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

coculture with rat glial cells;55 and a later study was mainly confirmatory.56 Interestingly, in these two studies differences in junctional proteins and efflux transporters were not revealed; however, it is well-known that membrane proteins are often overseen by proteomic analyses. Different experimental setups were designed to mimic the astrocytic influence on the cerebral endothelium. One of the possibilities is to culture endothelial cells on the top of a porous filter insert with astrocytes grown on the bottom of the filter. This allows a direct contact of astrocytic processes with endothelial cells; however, due to the relatively low density of the pores, the coverage of the endothelial surface is only a fraction of what can be observed in vivo. A further major difference of this setup compared to the in vivo anatomical conditions is the thickness of the porous membrane of about 10−20 μm, which far exceeds the thickness of the basal lamina. A second possibility is to grow astrocytes in the bottom of the well, which will enable soluble factors to exert their effect. An alternative approach is to use astrocyte-conditioned medium which contains soluble factors released by astrocytes. However, in this latter case the influence of endothelial cells on the release of different factors by astrocytes is excluded. Widely used are rodent endothelial cells in combination with rodent glial cells.57 Regarding the nature of glial cells, often mixed cultures are used containing, besides astrocytes (usually a large majority), oligodendrocytes or even microglia as well.58 These models can be useful in the investigation of different pathological aspects of BBB function as well. Recently a similar time course of BBB opening was observed in vivo and in vitro following stroke and oxygen/glucose deprivation using a mouse model, suggesting a good correlation between in vitro and in vivo models.59 By optimizing culture conditions robust, stable rat BBB models can be obtained with TEER as high as 600 Ω· cm2.57 Even the use of cells from different rodent species (e.g., mouse endothelial cells in combination with rat astrocytes) led to BBB models with good barrier properties, and the correlation coefficient between in vivo data and in vitro data concerning permeability of different drugs is as high as 0.98.60 It is noteworthy that in some comparative studies better barrier permeability can be obtained when astrocytes are cultured at the bottom of the well and not on the lower side of the filter. An explanation of this phenomenon could be that the usual pore size of 0.4 μm of the semiporous membranes is not large enough to enable complete extension of astrocyte foot processes to the endothelial side.61 Very stable BBB models can be established by the use of bovine brain endothelial cells in coculture with rat glial cells. Using a slightly modified method of Dehouck et al.62 long-term maintenance of the BBB models (up to 2 weeks) is now possible, which makes them suitable for chronic studies as well.63 The effect of glial cells can be mimicked by the use of a “BBB-inducing medium” containing coculture-conditioned medium, which allows the preparation of ready-to-use models significantly reducing the effort to establish a BBB model by the enduser.64 Coculture Models Using Pericytes. Although pericytes are the closest neighbors of endothelial cells in vivo, their effect is far less well characterized than that of the astrocytes. They have contractile, immune, phagocytic, and angiogenic functions. Pericytes also contribute to the formation of the basement membrane by synthesizing laminin, type IV collagen, and glycosaminoglycans. Besides being important elements of the BBB they are a source of adult pluripotent stem cells as well.65

transport (i.e., drug efflux or uptake, respectively). By using selective inhibitors of specific transporters one can determine the individual transporter responsible for the drug efflux/uptake (Figure 2B). In the past few years we have witnessed an impressive increase of studies dealing with a large variation in the BBB models. The variations included choice of species, types of cocultured cells, different cell lines, and culture conditions. Although most of these cannot be considered new models but improvements of already established ones, some new trends reflecting technological advances are also emerging. In recent years several excellent reviews summarized the gathered knowledge;44−46 here we will focus only on the developments of the past few years. Although cerebral endothelial monolayers cultured in dishes may be very useful for different investigations, like druginduced intracellular signaling or toxicity, they cannot be considered BBB models simply because they do not provide the two compartments separated by the BBB: the circulation side and the brain side. Endothelial Monocultures. The simplest models are monocultures of cerebral endothelial cells on microporous membranes. The pores of the membrane allow exchange of solutes between the apical and basolateral compartments and depending on the pore sizeeven cellular trafficking can be investigated. The model allows the use of endothelial cells of different origin, the most widely used being murine, rat, porcine, bovine, monkey, or human cerebral endothelial cells. The major advantage of the model is its simplicity, which allows for relative high throughput screenings at moderate costs.47 The major disadvantage of the model is that, in the absence of stimulating factors derived from other cellular components of the neurovascular unit (NVU) (astrocytes, pericytes), these models have a relatively low transendothelial electrical resistance, which means that the paracellular barrier properties are not adequate. Nevertheless, this may be sufficient for many applications investigating the BBB permeability of potential drug candidates, especially because it has been shown that TEER values above 150 Ω·cm2 are not accompanied by a decrease in permeability.48 Similar results were obtained with epithelial cells where the threshold values were around 200 Ω· cm2.49 However, in order to maintain the specific brain endothelial phenotype, the presence of the cerebral environment (e.g., pericytes, astrocytes, etc.) is needed. Therefore, coculture models have been developed which mimic the in vivo anatomy of the neurovascular unit and induce or maintain the barrier properties of cerebral endothelial cells. Coculture Models. Coculture of Cerebral Microvascular Endothelial Cells with Astrocytes. Althoughdue to the presence of the basal membraneastrocytes are not in direct contact with endothelial cells in the neurovascular unit, the interaction between these two cells is the best characterized (for review see refs 50 and 51). Especially important is the effect of astrocytes during adulthood on the interendothelial junctionswhich largely determines permeabilityand also on the modulation of transporter expression. Astrocytes play a crucial role in the development of the complexity of tight junctions52 and upregulation of the efflux transporters ABCB153 and ABCG2.54 Recent proteomic analysis revealed changes in 55 proteinsinvolved mainly in cell structure and motility and protein metabolism and modificationafter induction of bovine brain endothelial BBB functions by F

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

It is now generally accepted that pericytes are able to regulate paracellular permeability as well. Studies using endothelial cell/ pericyte cocultures revealed that pericyte derived angiopoietin1 induces Tie-2 mediated occludin gene expression in cerebral endothelial cells.66 Another pericyte-derived factor able to regulate endothelial functions is TGF-β, which has been shown to reduce permeability of the BBB in a rodent coculture model.67 The permeability reducing effect of pericytes in in vitro BBB models is comparable to or even better than that of astrocytes.68 Besides improving barrier properties pericytes induce the functional activity of ABC transporters as well and are required to maintain optimal barrier characteristics in hypoxia.69,70 Despite these results the interaction of pericytes and endothelial cells may be more complex. It has been shown that pericytes are also able to increase MMP-9 activity which acts toward a deterioration of the barrier properties.71 These not completely consistent results may have the explanation in the high level of heterogeneity among pericytes or different functional state of these cells. This is supported by a recent study in which normal pericytes performed as well as astrocytes in promoting barrier characteristics; however, pericytes from stress-susceptible pigs (carrying a point mutation in ryr1/hal gene encoding the ryanodine receptor) increased blood−brain barrier permeability in vitro.72 Triple Cell Coculture Models. The recognition that not only astrocytes but also pericytes are major players in the modulation of endothelial function led to the development of triple coculture models using the Transwell system (Figure 3). Regarding the localization of the three cell types, different setups are possible. A common feature is the culture of endothelial cells on the top of a porous filter insert. Pericytes and astrocytes then can be cultured on the lower side of the filter and at the bottom of the well, respectively. In an initial study using primary rat cells, the best barrier properties were obtained when pericytes were cultured on the bottom side of the filter and astrocytes at the bottom of the wells. Interestingly, in this study pericytes alone were more effective in inducing barrier properties than astrocytes alone indifferent of the location.68 The model was thoroughly characterized later, demonstrating continuous membrane staining of junctional proteins, active P-gp, and drug permeability with very good in vivo correlation73 (Figure 3A). The same model was used to study age dependent changes of BBB functions. It has been demonstrated that an in vitro BBB model based on cerebral endothelial cells isolated from 2-week old rats showed inferior barrier properties compared to a model based on endothelial cells isolated from adult (8-week old) rats. This suggests that age-related barrier properties of cerebral endothelial cells are retained in vitro, making it possible to investigate differences in BBB functions between adulthood and childhood.74 In a further adaptation of the model, endothelial cells were grown on the bottom of filters with 8 μm pore size and a mixture of pericytes and astrocytes were cultured on the top to study the interaction of BBB with glioblastoma multiforme. TEER values over 200 Ω·cm2 in monoculture and over 300 Ω·cm2 in coculture were measured75 (Figure 3B). The model was transferred to a primate system by using primary monkey brain endothelial cells in coculture with rat astrocytes and pericytes. TEER values of 350 Ω·cm2 were achieved accompanied by low permeability and continuous lines of tight junctions. The model is commercially available (http://www.pharmacocell.co.jp/en/products/mbt24h_e.

Figure 3. In vitro BBB models using three types of cells. Cerebral endothelial cells are cultured on semiporous membranes in the presence of pericytes, astrocytes, and/or neurons in different arrangements.

html). The great promise of the model is that it uses primate endothelial cells, which could show a relatively low level of species differences to human compared to rodent, porcine, or bovine models. In further triple cell culture models neurons or neural precursor cells have also been introduced. A novel triple cell neurovascular unit model coculturing rat cerebral endothelial cells, neurons, and astrocytes was recently established. TEER values of the triple cell coculture model increased with 35.9% compared with endothelial culture only and reached 268 Ω·cm2 (Figure 3C). Increased P-gp protein levels as assessed by immunocytochemistry and improved permeability to sodium fluorescein were also observed, although in the case of pemeability only relative values were presented.76 In a recent elegant study the role of direct contact between endothelial cells and pericytes was investigated. In one configuration pericytes were cultured on the bottom of the well together with astrocytes while endothelial cells were on the filter (noncontact mode), while in another setup astrocytes were cultured in the wells and pericytes together with endothelial cells were cultured on the top of the filter allowing close interactions between these two cell types (Figure 3D). There were no differences in transporter expression of endothelial cells between the two models, and the permeability of the barrier was even lower in the noncontact mode.77 Coculture of Brain Endothelial Cells with Neuronal Precursors. It is well established that astrocytes and pericytes have a BBB inductive effect; however, much less is known G

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

claudin-5. In consequence of the transfection, the permeability for inulin decreased from 11.6 × 10−3 cm/min to 1.14 × 10−3 cm/min, which is comparable to values obtained in primary cells.87 When mouse cerebral endothelial cell lines were compared, bEND.3 cells were shown to have relatively low permeability to sodium fluorescein and an obvious staining of tight junction proteins (claudin-5, occludin and ZO-1) was observed, while the barrier properties of MBEC4 and bEND.5 cells were low.88 A murine endothelial cell line (cEND) with remarkable barrier properties has recently been described. Mouse cerebral and cerebellar endothelial cells were immortalized with oncoprotein of murine polyomavirus, polyoma middle T antigen. High TEER values were reported even in monoculture under low serum conditions (500 Ω·cm2 in 2% serum), which was potentiated by the addition of hydrocortisone or insulin reaching 1000 Ω·cm2. It is noteworthy that endothelial cells isolated from the cerebral cortex and the cerebellar cortex showed distinct characteristics, drawing the attention to the heterogeneity of brain capillaries.89 An alternative approach has been recently used in another study. Here, in order to maintain barrier properties, endothelial cell lines (the murine bEnd5 and the human hCMEC/D3) were cultured in the presence of Wnt3 or drugs that stabilize βcatenin. Transcriptional activation of β-catenin led to an increase of TEER; however, the obtained values were lower than 100 Ω·cm2.90 Human Endothelial Cell Lines. Due to species differences which could lead to misinterpretations during the development of new human drugs, considerable effort has been made to establish human cerebral endothelial cell lines. One of the best characterized and most widely used human cerebral endothelial cell lines is the hCMEC/D3.91 Since its development, over 100 studies on different aspects of cerebral endothelial biology and pharmacology have been published using this cell line.92 In the large majority of the studies the barrier properties of the endothelial monolayer were unsuitable for permeability studies. Apparently cerebral endothelial cells lose their protein expression pattern when removed from their native environment. In a recent transcriptional profiling claudin-5, occludin, JAM2, SLC2A1, ABCB1, LRP1, RAGE, and the insulin receptor expression were shown to be dramatically reduced in cell lines compared to primary cells. On the other hand, genes involved in RNA processing, DNA repair, immune and virus response, and mitosis are highly and significantly upregulated in the hCMEC/D3 cell line.93 However, by optimizing culture conditions considerable progress could be made, and by coculturing hCMEC/D3 cells with primary human astrocytes TEER values over 140 Ω·cm2 were reported.94 Unfortunately, the hCMEC/D3 cell lineat least in monoculturedoes not completely reflect the ABCG2 transporter dominant phenotype observed in human capillaries. In two different studies ABCB1 was shown to exhibit the highest mRNA or protein expression among ABC transporters,19,95 while in another study the mRNA levels of ABCC4 and ABCC5 exceeded the levels of ABCB1 and ABCG296 in the hCMEC/D3 cell line. The human endothelial cell line BB19 was shown to have high permeability to sucrose, suggesting an inadequate tightness for permeability studies.97 Another human cerebral endothelial cell line was established by transfection of early passage endothelial cells with a pBR322-based plasmid containing simian virus 40 large T antigen. The cell line hBMEC expresses key endothelial markers and shows a TEER of more than 300 Ω·cm2 even in monoculture.98 To our knowledge this cell line

about the cell types responsible for early embryonic BBB induction. It has been shown that soluble factors released by neuronal precursor cells in their early stage of differentiation are able to increase TEER by about 50% with accompanying decrease in permeability. Since low TEER values and relatively high permeability values were measured, the model is mainly suitable for the study of the interaction between NVU cells during development. However, it may also be useful in the study of drug effects during embryonic life.78 The model was further developed by using neuronal progenitor cell-derived astrocytes and neurons. Best results were obtained with cocultures containing roughly a 3:1 mixture of astrocytes and neurons with varying degrees of cellular maturation. TEER values reached 250 Ω·cm2, which is comparable to the values induced by pure astrocytes.79 Recently, a fully human model has been described using neural progenitor cell-derived astrocytes and neurons, reaching very high TEER values.80 Stem cell-derived brain endothelial cells were sequentially cocultured with primary human brain pericytes, and later with human astrocytes and neurons derived from human neural progenitor cells (Figure 3E). Endothelial Cells Used for Preparing BBB Models. Primary Microvascular Brain Endothelial Cells Used As BBB Models. Brain endothelial cells tend to lose their BBB properties in vitro even in the presence of astrocytes, pericytes, or other cells of the neurovascular unit. Therefore, the barrier properties of cerebral endothelial monolayers are altered by long-term cultures or repeated passages. In parallel, cell lines are inferior to primary cells in respect to the tightness of the barrier. Therefore, primary cultures are the first choice for drug permeability studies. The most widely accepted models use primary cells of rat, murine, porcine, and bovine origin in mono- or coculture.45 With respect to the tightness of the barrier, primary porcine brain endothelial cells seem to be the best, reaching average TEER values of 700−800 Ω·cm2 even in monoculture;81,82 however, bovine models and rodent models in coculture with astrocytes and/or pericytes can also reach good TEER values of 200−400 or even up to 600−700 Ω· cm2.45,57,58,73,83 For drug delivery studies human primary cells would be the ideal model; however, the restricted availability of the experimental material hinders the high throughput tests. Moreover, the human brain tissue used for the isolation of cells usually originates from surgical or cryopreserved material and not “healthy” tissue.84−86 Since isolation and culture of primary cells is often complicated and costly and the cells might differ from batch to batch and are usually not enough for high throughput studies, establishment of reliable brain endothelial cell lines recapitulating the most important BBB characteristics are needed. Non-Human Endothelial Cell Lines As BBB Models. A considerable number of endothelial cell lines of mouse, rat, porcine, bovine, and human origin were established in the past 15−20 years, recently reviewed by Toth et al.44 While useful in the study of different aspects of the BBB function, one of the major problems of these cell lines is the poor barrier property (low TEER and high permeability values), which seriously limits their usefulness in drug testing studies. However, some recent publications report about endothelial cell line-based models worth being considered for permeability studies as well. One such model is based on the conditionally immortalized rat brain capillary endothelial cell line TR-BBB overexpressing H

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

in vivo blood flow by culturing endothelial cells in hollow fibers and circulating culture media creating a tunable shear stress. One of the best characterized dynamic in vitro models was developed in the laboratory of Dr. Janigro. As in the case of static models, the cells are cultured on a polypropylene support which is folded to form a hollow fiber mimicking a microvessel. The support is a porous material with a pore size of 200 nm, which in the case of coculture allows a direct contact between cells grown on the opposite sides of the membrane. However, the relative thickness of the membrane (>10 μm) limits the possibility of direct contact. Using bovine aortic endothelial cells in coculture with astrocytes, TEER values of above 500 Ω· cm2 were obtained after 20 days in culture.109 The model was later humanized using human cerebral endothelial cells and astrocytes originating from a surgical specimen. TEER values over 1000 Ω·cm2 were obtained with a sucrose permeability below 2 × 10−7 cm/s.110 Similar TEER values were obtained by using the human cerebral endothelial cell line hCMEC/D3. It is noteworthy that the same cells developed TEER of 60−70 Ω· cm2 in a static Transwell system, underlining the importance of the shear stress. Interestingly, astrocytes in this case were not able to further improve the barrier properties.111 A further adaptation of the humanized dynamic in vitro BBB model enabled the study of transendothelial trafficking of immune cells as well by using hollow fibers with pores of 2−4 μm.112 Recently the model was further developed to mimic venous conditions as well. This was achieved by reducing the pressure and coculturing endothelial cells with smooth muscle cells.113 Despite the numerous advantages of this system some limitations also have to be mentioned. The DIV-BBB system is not suitable for high throughput pharmacological studies, and the establishment of the model requires specific technical skills. Furthermore, initially a large number of cells are needed to load the capillaries and cell morphology cannot be controlled. In another model immortalized porcine brain endothelial cells PBMEC/C1-2 were seeded onto the inner surface of the hollow fibers (polypropylene capillaries coated with ProNectin F with an inner diameter of 330 μm, wall thickness of 150 μm, and average pore size of 0.5 μm), whereas C6 glioma cells were seeded in the extracapillary space. The model does not allow for measurement of TEER but showed low permeability properties for several tested substances.114 Microfluidic Models. Microfluidic methods allow considerable downsizing of BBB models with the advantage of lower required cell number. In these models cells of the neurovascular unit are also cultured on porous membranes but the membrane is placed at the interface of two microchannels, which allows flow of the culture medium (Figure 4). For the formation of the

has been mainly used for basic research; however, in a comparative study this cell line proved to be superior to the hCMEC/D3, the TY10, and the BB1999 cell lines concerning barrier tightness.100 In recent years several other human brain endothelial cells were established. A conditionally immortalized endothelial cell line is the HBMEC/ciβ, which expresses tight junction proteins and efflux transporters as well. A permeability of 1.8 × 10−3 cm/min for Na-fluorescein and an A-B/B-A ratio of 1.8 ± 0.3 for rhodamine 123 were reported.101 Stem Cell-Derived Brain Endothelial Cells As BBB Models. An intriguing possibility is to use pluripotent stem cells, especially human ones to differentiate them into different cellular components of the neurovascular unit. These approachesif successfulcould resolve the problem of limited accessibility to human material and the poor barrier characteristics of endothelial cell lines. A first step in this direction was recently made by developing a BBB model based on human endothelial cells derived from pluripotent stem cells. Endothelial monolayers obtained in this way achieved TEER of 150−175 Ω·cm2. Optimization of the seeding density and coculture with astrocytes led to values as high as 1450 ± 140 Ω· cm2, and the permeability values for sucrose were as low as 3.4 × 10−5 cm/min, which are substantially better than those obtained with hCMEC/D3. In addition, a functional ABCB1 was also present in these cells.102 This promising new development opens the way toward a BBB model in which not only endothelial cells but all cellular elements of the neurovascular unit are derived from pluripotent stem cells.103 The recently developed fully human BBB model has been prepared using retinoic acid-treated human pluripotent stem cell-derived cerebral microvascular endothelial cells.80 The highest TEER (∼3500 Ω·cm2) was achieved when subconfluent brain endothelial cells were cocultured with pericytes immediately following purification and then the confluent endothelial layer was transferred into coculture with differentiated neural progenitor cells 24 h later (Figure 3E). Primary Glia and Glial Cell Lines Used for Coculture Models. When coculturing brain endothelial cells with astrocytesalthough the use of primary glial cells usually gives better resultsthe advantages of primary gial cells over the use of cell lines is not obvious, especially taking into account the relative simplicity of the use of a cell line. While for example in a study using porcine brain endothelial and rat astrocytes no differences between the effect of primary astrocytes and C6 glioma cells were observed,104 in a comparative study using bovine endothelial cells in coculture with rat astrocytes and C6 glioma cells, Boveri et al. have shown that although C6 glioma cells are able to increase the TEER, the effect was significantly weaker than that observed with primary glial cells.83 The barrier increasing effect of C6 cells has been demonstrated in other studies as well using endothelial cells from different species.105 Due to the tumor origin of C6 glioma cells the coculture of cerebral endothelial cells with these cells is often regarded as a blood−tumor barrier, which is an oversimplification of the in vivo situation. In a comparative study the effect of an immortalized rat astrocyte cell line (CTXTNA2) was evaluated. Both primary rat astrocytes and CTXTNA2 cells induced a 13−14-fold increase in TEER in a BBB model based on porcine endothelial cells.106 Dynamic in Vitro (DIV) Models. Shear stress (approximately 5 dyn/cm2) induced by flow has been demonstrated to significantly influence endothelial properties including barrier properties as well.107,108 Dynamic BBB models try to mimic the

Figure 4. Schematic representation of microfluidic models. Cerebral endothelial and glial cells are cultured on the two sides of a semiporous membrane placed at the interface of two microchannels. I

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

lines, several attempts have been made to use modified epithelial cell lines for the investigation of brain uptake of different drugs (Table 3). Although these cannot be considered

channel system polydimethylsiloxane (PDMS) is the material of choice, and built in electrodes allow a continuous TEER measurement. Several new developments have been published with different outcomes. In a model developed by Griep et al. hCMEC/D3 cells were cultured in the microfluidic device up to 7 days grown on Transwell polycarbonate membranes with a thickness of 10 μm and a pore size of 0.4 μm. The perpendicular channels were 1 cm long, 500 μm wide, and 100 μm deep. Under static conditions the obtained TEER was 36.9 Ω·cm2, a value comparable to the values obtained with these cells in Transwell systems. However, when flow was applied (a shear stress of 5.8 × 10−1 Pa) TEER increased to 120 Ω·cm2. The cells expressed ZO-1 protein at the membranes, but no further characteristics were measured.115 Another, even more comprehensively characterized model was developed by Booth and Kim. The authors used a perpendicular channel system fabricated from PDMS with a height of 200 μm, and widths of 2 mm (lumen) and 5 mm (ablumen) to allow laminar flow. To build up the model, b.END3 brain endothelial cells and C8-D1A astrocytes were cocultured on polycarbonate sheets with a surface of 10 mm2 (400 nm pores, 10 μm thickness). Coculture with an astrocyte cell line increased the TEER from approximately 15−20 Ω·cm2 to 20−25 Ω·cm2 under static conditions, whereas under dynamic conditions the basal TEER of well over 100 Ω·cm2 increased to over 250 Ω·cm2 after 3 days in culture. Permeability for propidium iodide was 1−2 × 10−6 cm/s and for 70 kDa dextran 1−2 × 10−7 cm/s, which indicates a tight barrier. ZO-1 staining showed continuous membrane localization; however, unfortunately no data about efflux transporters were published.116 A model of two side-by-side compartments made of PDMS was designed by Prabhakarpandian et al. Here the apical and basolateral compartments are separated by a continuous line of pillars with 3 μm gaps, the apical side hosting RBE4 endothelial cells, the basolateral side containing astrocyte-conditioned medium. The conditioned medium significantly improved TJ protein and Abcb1a expression. Unfortunately the device in its present form does not allow for TEER measurements.117 In a different microfluidic setup also based on PDMS, a microhole structure for trapping cells was designed. HUVECs were trapped on microholes in the microfluidic device and incubated with astrocyte-conditioned medium. The permeability for antipyrine, carbamazepine, and atenolol were comparable to the results obtained with the same cells in the classical Transwell system; however, permeability of propranolol and verapamil was higher, indicating a low ABCB1 function. Immunofluorescence staining of ZO-1 showed a junctional expression with some discontinuities. Since the device was not designed for TEER measurements, no data about the tightness of the junctions is available.118 Finally, a new three-dimensional model is under development which will recapitulate the whole neurovascular unit in a physiological geometry. The planned “brain-on-a-chip” will utilize two microfabricated compartments representing the brain and the cerebrospinal fluid, separated by an ependymal layer. The upper chamber will contain a hollow fiber capillary lined by endothelial cells, carrying a blood surrogate and surrounded by neurons, pericytes, and astrocytes.119 Epithelial Cell Lines Used for the Study of Drug Delivery. Due to the intensive human resources required to work with primary cerebral endothelial cells and to the generally inferior barrier characteristics of endothelial cell

Table 3. Differences between Brain Endothelial and Epithelial Junctions

BBB models, they can be useful tools in initial phases of drug discovery. In this respect the Madin−Darby canine kidney (MDCK) cell line is one of the most important. Using MDCKII and MDCK-ABCB1 cell monolayers and mathematical analysis the BBB parameters obtained were predictive of the in vivo behavior of the test compounds.120 Similarly, efflux ratio measurements using porcine kidney epithelial cells LLCPK1 stably transfected with Abcb1a proved to be useful in predicting in vivo cerebrospinal fluid/plasma fractions for central nervous system-targeting drugs.121 In another study the vinblastine-resistant Caco-2 and the MDCK-ABCB1 models identified more P-glycoprotein drug substrates than a primary rat brain endothelial-based BBB model; therefore, it has been concluded that, although not reaching the predictive value of the BBB model, these cell lines may be useful in determining whether a drug candidate is an efflux transporter substrate or not.122 Optimization of in Vitro BBB Models. Isolation of primary brain endothelial cells requires difficult techniques: enzymatic digestion steps combined with separation of the microvessels using gradient centrifugation or filter meshes. Several protocols have been described.45,58,81 When isolating primary cerebral endothelial cells, removal of contaminating cells is also an important issue. The most widespread method is to use puromycin-containing medium, which kills cells with low or no P-gp activity.123 In order to obtain a model suitable for drug delivery studies, optimization of culture conditions of individual models seems to be of crucial importance. Regarding the material on which endothelial cells are grown, most models use polycarbonate or polyethylene filter inserts with pore sizes ranging from 0.4 to 8 μm. Generally, before seeding endothelial cells, a coating of the surface is needed with a material which mimics the basement membrane; usually type IV collagen and fibronectin are the materials of choice. In a recent comparative study highest TEER values were obtained on 0.4 μm pore size filters,124 and this pore size is well suitable for drug testing. For investigations of cellular transmigration, filters with larger pore size are required.125 Regarding the material of the filter, better barrier properties were seen using polyethylene terephthalate than polycarbonate.124 Moreover, a relatively high minimal seeding density (4 × 105 cells/cm2) was found to be critical.126 Components of the culture medium may also contribute to the formation of a tight barrier. In this respect elevated J

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

microfluidic models. However, in their present stage of development they are relatively expensive, and it is still not granted that the extra effortin order to more precisely mimic the in vivo situationwill be rewarded by a significantly better correlation, especially at initial stages of CNS drug development. Presently, the most widely used models are based on culture of primary brain endothelial cells on filter inserts in mono- or coculture. Independent of the setup, if the model gives high TEER and expresses functional transporters, the obtained permeability results will be very useful and in good correlation with in vivo results. However, interpretation of the data needs care, and several issues (i.e., interspecies differences, pathological conditions affecting BBB permeability) should be considered. Development of more potent human cell lines or more optimal cell culture conditions would be extremely helpful in the study of drug delivery to the central nervous system.

intraendothelial cAMP levels have been shown to be important;127 therefore, culture media often contain cell permeable cAMP analogues and substances which inhibit phosphodiesterases. Hydrocortisone in physiological concentration is also often used to improve barrier properties.128 It has been recently shown that activation of the Wnt signaling pathway is also beneficial for the formation of the barrier.129 An increase in the buffer capacity of the medium also significantly increases the tightness of the BBB model.130 Another important question is the addition of serum. Usually plasma-derived serum (PDS)containing lower amounts of VEGFis preferred to fetal bovine serum (FBS) in a concentration between 1 and 20% (in general 10%); however, in the case of porcine models serum-free differentiation medium may also be used.131,132 Other factors proved to be useful in tightening the barrier are insulin, transferrin, sodium selenite, putrescine, and progesterone.





CONCLUSIONS AND OUTLOOK Unfortunately, as the large number of different models suggests, there is no ideal model: each has its advantages and disadvantages (Table 4). A clearly defined consensus about

*Tel: +36-62-599602. Fax: +36-62-433133. E-mail: krizbai. [email protected]. Notes

Table 4. Advantages and Disadvantages of Different in Vitro BBB Models model

advantages

epithelial cells overexpressing transporters

cheap

Transwell monoculture

uses brain endothelial cells

coculture

easy to standardize

inexpensive takes into account the influence of other elements of theNVU

dynamic

mimics in vivo situation possibility of coculture

microfluidic

mimics in vivo situation possibility of coculture

AUTHOR INFORMATION

Corresponding Author

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The work of I.W. and I.A.K. is supported by grants from the Hungarian Research Fund (OTKA PD-100958, K-100807), the National Development Agency (Hungary-Romania CrossBorder Co-operation Programme 2007-2013: HURO/1101/ 173/2.2.1; and the TÁ MOP-4.2.2.A-11/1/KONV-2012-0052 project), and the János Bolyai Research Fellowship of the Hungarian Academy of Sciences (BO/00320/12/8).

disadvantages differences between epithelial and endothelial cells nonphysiologically high levels of transporters effect of other cellular components of the NVU is neglected no shear stress relatively expensive and time-consuming



REFERENCES

(1) Pathan, S. A.; Iqbal, Z.; Zaidi, S. M.; Talegaonkar, S.; Vohra, D.; Jain, G. K.; Azeem, A.; Jain, N.; Lalani, J. R.; Khar, R. K.; Ahmad, F. J. CNS drug delivery systems: novel approaches. Recent Pat. Drug Delivery Formulation 2009, 3, 71−89. (2) Alavijeh, M. S.; Chishty, M.; Qaiser, M. Z.; Palmer, A. M. Drug metabolism and pharmacokinetics, the blood-brain barrier, and central nervous system drug discovery. NeuroRx 2005, 2, 554−571. (3) Haseloff, R. F.; Blasig, I. E.; Bauer, H. C.; Bauer, H. In search of the astrocytic factor(s) modulating blood-brain barrier functions in brain capillary endothelial cells in vitro. Cell. Mol. Neurobiol. 2005, 25, 25−39. (4) Alvarez, J. I.; Dodelet-Devillers, A.; Kebir, H.; Ifergan, I.; Fabre, P. J.; Terouz, S.; Sabbagh, M.; Wosik, K.; Bourbonniere, L.; Bernard, M.; van Horssen, J.; de Vries, H. E.; Charron, F.; Prat, A. The Hedgehog pathway promotes blood-brain barrier integrity and CNS immune quiescence. Science 2011, 334, 1727−1731. (5) Sims, D. E. Recent advances in pericyte biologyimplications for health and disease. Can. J. Cardiol. 1991, 7, 431−443. (6) Thanabalasundaram, G.; Schneidewind, J.; Pieper, C.; Galla, H. J. The impact of pericytes on the blood-brain barrier integrity depends critically on the pericyte differentiation stage. Int. J. Biochem. Cell Biol. 2011, 43, 1284−1293. (7) Hellstrom, M.; Gerhardt, H.; Kalen, M.; Li, X.; Eriksson, U.; Wolburg, H.; Betsholtz, C. Lack of pericytes leads to endothelial hyperplasia and abnormal vascular morphogenesis. J. Cell Biol. 2001, 153, 543−553. (8) Armulik, A.; Genove, G.; Mae, M.; Nisancioglu, M. H.; Wallgard, E.; Niaudet, C.; He, L.; Norlin, J.; Lindblom, P.; Strittmatter, K.; Johansson, B. R.; Betsholtz, C. Pericytes regulate the blood-brain barrier. Nature 2010, 468, 557−561.

no shear stress expensive no possibility to optically monitor the cells special skills required to culture cells in these conditions not well established models presently expensive

which characteristics or parameters a model should possess to be suitable for drug testing is also lacking. In this respect low paracellular permeability/high TEER values and physiological expression and function of ABC and SLC transporters and metabolizing enzymes seem to be the most essential. The right model has to be chosen depending on the information expected to be obtained from the study. For initial large scale screenings (i.e., for the determination if a substance is substrate of an efflux transporter) modified epithelial cells may be useful; however, it has to be taken into account that overexpression of a transporter in a cell line will lead to much higher expression levels than that found under physiological conditions in cerebral endothelial cells. Regarding the species of choice, the most robust models are based on bovine or porcine endothelial cells, and porcine vascular physiology is the closest to humans. However, if a specific disease is targeted for which transgenic models exist, rodent models may be more useful. An emerging direction is the use of triple coculture, dynamic, and K

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

(9) Cummins, P. M. Occludin: one protein, many forms. Mol. Cell. Biol. 2012, 32, 242−250. (10) Muresan, Z.; Paul, D. L.; Goodenough, D. A. Occludin 1B, a variant of the tight junction protein occludin. Mol. Biol. Cell 2000, 11, 627−634. (11) Ghassemifar, M. R.; Sheth, B.; Papenbrock, T.; Leese, H. J.; Houghton, F. D.; Fleming, T. P. Occludin TM4−: an isoform of the tight junction protein present in primates lacking the fourth transmembrane domain. J. Cell. Sci. 2002, 115, 3171−3180. (12) Mankertz, J.; Waller, J. S.; Hillenbrand, B.; Tavalali, S.; Florian, P.; Schoneberg, T.; Fromm, M.; Schulzke, J. D. Gene expression of the tight junction protein occludin includes differential splicing and alternative promoter usage. Biochem. Biophys. Res. Commun. 2002, 298, 657−666. (13) Gu, J. M.; Lim, S. O.; Park, Y. M.; Jung, G. A novel splice variant of occludin deleted in exon 9 and its role in cell apoptosis and invasion. FEBS J. 2008, 275, 3145−3156. (14) Dorfel, M. J.; Huber, O. A phosphorylation hotspot within the occludin C-terminal domain. Ann. N.Y. Acad. Sci. 2012, 1257, 38−44. (15) Mariano, C.; Palmela, I.; Pereira, P.; Fernandes, A.; Falcao, A. S.; Cardoso, F. L.; Vaz, A. R.; Campos, A. R.; Goncalves-Ferreira, A.; Kim, K. S.; Brites, D.; Brito, M. A. Tricellulin expression in brain endothelial and neural cells. Cell Tissue Res. 2013, 351, 397−407. (16) Steed, E.; Rodrigues, N. T.; Balda, M. S.; Matter, K. Identification of MarvelD3 as a tight junction-associated transmembrane protein of the occludin family. BMC Cell Biol. 2009, 10, 95−2121−10−959. (17) Raleigh, D. R.; Marchiando, A. M.; Zhang, Y.; Shen, L.; Sasaki, H.; Wang, Y.; Long, M.; Turner, J. R. Tight junction-associated MARVEL proteins marveld3, tricellulin, and occludin have distinct but overlapping functions. Mol. Biol. Cell 2010, 21, 1200−1213. (18) Nitta, T.; Hata, M.; Gotoh, S.; Seo, Y.; Sasaki, H.; Hashimoto, N.; Furuse, M.; Tsukita, S. Size-selective loosening of the blood-brain barrier in claudin-5-deficient mice. J. Cell Biol. 2003, 161, 653−660. (19) Ohtsuki, S.; Yamaguchi, H.; Katsukura, Y.; Asashima, T.; Terasaki, T. mRNA expression levels of tight junction protein genes in mouse brain capillary endothelial cells highly purified by magnetic cell sorting. J. Neurochem. 2008, 104, 147−154. (20) Jia, W.; Martin, T. A.; Zhang, G.; Jiang, W. G. Junctional adhesion molecules in cerebral endothelial tight junction and brain metastasis. Anticancer Res. 2013, 33, 2353−2359. (21) Whiteman, E. L.; Fan, S.; Harder, J. L.; Walton, K. D.; Liu, C. J.; Soofi, A.; Fogg, V. C.; Hershenson, M. B.; Dressler, G. R.; Deutsch, G. H.; Gumucio, D. L.; Margolis, B. Crumbs3 is Essential for Proper Epithelial Development and Viability. Mol. Cell. Biol. 2014, 34 (1), 43−56. (22) Traweger, A.; Toepfer, S.; Wagner, R.; Zweimueller-Mayer, J.; Gehwolf, R.; Lehner, C.; Tempfer, H.; Krizbai, I.; Wilhelm, I.; Bauer, H.; Bauer, H. Beyond cell-cell adhesion: Emerging roles of the tight junction scaffold ZO-2. Tissue Barriers 2013, 1, e25039. (23) Fabian, G.; Szabo, C. A.; Bozo, B.; Greenwood, J.; Adamson, P.; Deli, M. A.; Joo, F.; Krizbai, I. A.; Szucs, M. Expression of G-protein subtypes in cultured cerebral endothelial cells. Neurochem. Int. 1998, 33, 179−185. (24) Dejana, E.; Orsenigo, F. Endothelial adherens junctions at a glance. J. Cell. Sci. 2013, 126, 2545−2549. (25) Shawahna, R.; Uchida, Y.; Decleves, X.; Ohtsuki, S.; Yousif, S.; Dauchy, S.; Jacob, A.; Chassoux, F.; Daumas-Duport, C.; Couraud, P. O.; Terasaki, T.; Scherrmann, J. M. Transcriptomic and quantitative proteomic analysis of transporters and drug metabolizing enzymes in freshly isolated human brain microvessels. Mol. Pharmaceutics 2011, 8, 1332−1341. (26) Giacomini, K. M.; Huang, S. M. Transporters in drug development and clinical pharmacology. Clin. Pharmacol. Ther. 2013, 94, 3−9. (27) Hoshi, Y.; Uchida, Y.; Tachikawa, M.; Inoue, T.; Ohtsuki, S.; Terasaki, T. Quantitative atlas of blood-brain barrier transporters, receptors, and tight junction proteins in rats and common marmoset. J. Pharm. Sci. 2013, 102, 3343−3355.

(28) Deo, A. K.; Theil, F. P.; Nicolas, J. M. Confounding parameters in preclinical assessment of blood-brain barrier permeation: an overview with emphasis on species differences and effect of disease states. Mol. Pharmaceutics 2013, 10, 1581−1595. (29) Kamiie, J.; Ohtsuki, S.; Iwase, R.; Ohmine, K.; Katsukura, Y.; Yanai, K.; Sekine, Y.; Uchida, Y.; Ito, S.; Terasaki, T. Quantitative atlas of membrane transporter proteins: development and application of a highly sensitive simultaneous LC/MS/MS method combined with novel in-silico peptide selection criteria. Pharm. Res. 2008, 25, 1469− 1483. (30) Uchida, Y.; Ohtsuki, S.; Katsukura, Y.; Ikeda, C.; Suzuki, T.; Kamiie, J.; Terasaki, T. Quantitative targeted absolute proteomics of human blood-brain barrier transporters and receptors. J. Neurochem. 2011, 117, 333−345. (31) Daneman, R.; Zhou, L.; Agalliu, D.; Cahoy, J. D.; Kaushal, A.; Barres, B. A. The mouse blood-brain barrier transcriptome: a new resource for understanding the development and function of brain endothelial cells. PLoS One 2010, 5, e13741. (32) Chun, H. B.; Scott, M.; Niessen, S.; Hoover, H.; Baird, A.; Yates, J., 3rd; Torbett, B. E.; Eliceiri, B. P. The proteome of mouse brain microvessel membranes and basal lamina. J. Cereb. Blood Flow Metab. 2011, 31, 2267−2281. (33) Warren, M. S.; Zerangue, N.; Woodford, K.; Roberts, L. M.; Tate, E. H.; Feng, B.; Li, C.; Feuerstein, T. J.; Gibbs, J.; Smith, B.; de Morais, S. M.; Dower, W. J.; Koller, K. J. Comparative gene expression profiles of ABC transporters in brain microvessel endothelial cells and brain in five species including human. Pharmacol. Res. 2009, 59, 404− 413. (34) Ito, K.; Uchida, Y.; Ohtsuki, S.; Aizawa, S.; Kawakami, H.; Katsukura, Y.; Kamiie, J.; Terasaki, T. Quantitative membrane protein expression at the blood-brain barrier of adult and younger cynomolgus monkeys. J. Pharm. Sci. 2011, 100, 3939−3950. (35) Katoh, M.; Suzuyama, N.; Takeuchi, T.; Yoshitomi, S.; Asahi, S.; Yokoi, T. Kinetic analyses for species differences in P-glycoproteinmediated drug transport. J. Pharm. Sci. 2006, 95, 2673−2683. (36) Enerson, B. E.; Drewes, L. R. The rat blood-brain barrier transcriptome. J. Cereb. Blood Flow Metab. 2006, 26, 959−973. (37) Lacombe, O.; Videau, O.; Chevillon, D.; Guyot, A. C.; Contreras, C.; Blondel, S.; Nicolas, L.; Ghettas, A.; Benech, H.; Thevenot, E.; Pruvost, A.; Bolze, S.; Krzaczkowski, L.; Prevost, C.; Mabondzo, A. In vitro primary human and animal cell-based bloodbrain barrier models as a screening tool in drug discovery. Mol. Pharmaceutics 2011, 8, 651−663. (38) Morgello, S.; Uson, R. R.; Schwartz, E. J.; Haber, R. S. The human blood-brain barrier glucose transporter (GLUT1) is a glucose transporter of gray matter astrocytes. Glia 1995, 14, 43−54. (39) Shawahna, R.; Decleves, X.; Scherrmann, J. M. Hurdles with using in vitro models to predict human blood-brain barrier drug permeability: a special focus on transporters and metabolizing enzymes. Curr. Drug Metab. 2013, 14, 120−136. (40) Reichel, A. The role of blood-brain barrier studies in the pharmaceutical industry. Curr. Drug Metab. 2006, 7, 183−203. (41) Gumbleton, M.; Audus, K. L. Progress and limitations in the use of in vitro cell cultures to serve as a permeability screen for the bloodbrain barrier. J. Pharm. Sci. 2001, 90, 1681−1698. (42) Butt, A. M.; Jones, H. C.; Abbott, N. J. Electrical resistance across the blood-brain barrier in anaesthetized rats: a developmental study. J. Physiol. 1990, 429, 47−62. (43) Bickel, U. How to measure drug transport across the bloodbrain barrier. NeuroRx 2005, 2, 15−26. (44) Toth, A.; Veszelka, S.; Nakagawa, S.; Niwa, M.; Deli, M. A. Patented in vitro blood-brain barrier models in CNS drug discovery. Recent Pat. CNS Drug Discovery 2011, 6, 107−118. (45) Wilhelm, I.; Fazakas, C.; Krizbai, I. A. In vitro models of the blood-brain barrier. Acta Neurobiol. Exp. 2011, 71, 113−128. (46) Abbott, N. J. Blood-brain barrier structure and function and the challenges for CNS drug delivery. J. Inherited Metab. Dis. 2013, 36, 437−449. L

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

repeated-dose toxicological screening. Toxicol. In Vitro 2013, 27, 1944−1953. (64) Vandenhaute, E.; Sevin, E.; Hallier-Vanuxeem, D.; Dehouck, M. P.; Cecchelli, R. Case study: adapting in vitro blood-brain barrier models for use in early-stage drug discovery. Drug Discovery Today 2012, 17, 285−290. (65) Sa-Pereira, I.; Brites, D.; Brito, M. A. Neurovascular unit: a focus on pericytes. Mol. Neurobiol. 2012, 45, 327−347. (66) Hori, S.; Ohtsuki, S.; Hosoya, K.; Nakashima, E.; Terasaki, T. A pericyte-derived angiopoietin-1 multimeric complex induces occludin gene expression in brain capillary endothelial cells through Tie-2 activation in vitro. J. Neurochem. 2004, 89, 503−513. (67) Dohgu, S.; Takata, F.; Yamauchi, A.; Nakagawa, S.; Egawa, T.; Naito, M.; Tsuruo, T.; Sawada, Y.; Niwa, M.; Kataoka, Y. Brain pericytes contribute to the induction and up-regulation of blood-brain barrier functions through transforming growth factor-beta production. Brain Res. 2005, 1038, 208−215. (68) Nakagawa, S.; Deli, M. A.; Nakao, S.; Honda, M.; Hayashi, K.; Nakaoke, R.; Kataoka, Y.; Niwa, M. Pericytes from brain microvessels strengthen the barrier integrity in primary cultures of rat brain endothelial cells. Cell. Mol. Neurobiol. 2007, 27, 687−694. (69) Berezowski, V.; Landry, C.; Dehouck, M. P.; Cecchelli, R.; Fenart, L. Contribution of glial cells and pericytes to the mRNA profiles of P-glycoprotein and multidrug resistance-associated proteins in an in vitro model of the blood-brain barrier. Brain Res. 2004, 1018, 1−9. (70) Al Ahmad, A.; Taboada, C. B.; Gassmann, M.; Ogunshola, O. O. Astrocytes and pericytes differentially modulate blood-brain barrier characteristics during development and hypoxic insult. J. Cereb. Blood Flow Metab. 2011, 31, 693−705. (71) Zozulya, A.; Weidenfeller, C.; Galla, H. J. Pericyte-endothelial cell interaction increases MMP-9 secretion at the blood-brain barrier in vitro. Brain Res. 2008, 1189, 1−11. (72) Vandenhaute, E.; Culot, M.; Gosselet, F.; Dehouck, L.; Godfraind, C.; Franck, M.; Plouet, J.; Cecchelli, R.; Dehouck, M. P.; Ruchoux, M. M. Brain pericytes from stress-susceptible pigs increase blood-brain barrier permeability in vitro. Fluids Barriers CNS 2012, DOI: 10.1186/2045-8118-9-11. (73) Nakagawa, S.; Deli, M. A.; Kawaguchi, H.; Shimizudani, T.; Shimono, T.; Kittel, A.; Tanaka, K.; Niwa, M. A new blood-brain barrier model using primary rat brain endothelial cells, pericytes and astrocytes. Neurochem. Int. 2009, 54, 253−263. (74) Takata, F.; Dohgu, S.; Yamauchi, A.; Matsumoto, J.; Machida, T.; Fujishita, K.; Shibata, K.; Shinozaki, Y.; Sato, K.; Kataoka, Y.; Koizumi, S. In vitro blood-brain barrier models using brain capillary endothelial cells isolated from neonatal and adult rats retain agerelated barrier properties. PLoS One 2013, 8, e55166. (75) Toyoda, K.; Tanaka, K.; Nakagawa, S.; Thuy, D. H.; Ujifuku, K.; Kamada, K.; Hayashi, K.; Matsuo, T.; Nagata, I.; Niwa, M. Initial contact of glioblastoma cells with existing normal brain endothelial cells strengthen the barrier function via fibroblast growth factor 2 secretion: a new in vitro blood-brain barrier model. Cell. Mol. Neurobiol. 2013, 33, 489−501. (76) Xue, Q.; Liu, Y.; Qi, H.; Ma, Q.; Xu, L.; Chen, W.; Chen, G.; Xu, X. A novel brain neurovascular unit model with neurons, astrocytes and microvascular endothelial cells of rat. Int. J. Biol. Sci. 2013, 9, 174− 189. (77) Vandenhaute, E.; Dehouck, L.; Boucau, M. C.; Sevin, E.; Uzbekov, R.; Tardivel, M.; Gosselet, F.; Fenart, L.; Cecchelli, R.; Dehouck, M. P. Modelling the neurovascular unit and the blood-brain barrier with the unique function of pericytes. Curr. Neurovasc. Res. 2011, 8, 258−269. (78) Weidenfeller, C.; Svendsen, C. N.; Shusta, E. V. Differentiating embryonic neural progenitor cells induce blood-brain barrier properties. J. Neurochem. 2007, 101, 555−565. (79) Lippmann, E. S.; Weidenfeller, C.; Svendsen, C. N.; Shusta, E. V. Blood-brain barrier modeling with co-cultured neural progenitor cell-derived astrocytes and neurons. J. Neurochem. 2011, 119, 507− 520.

(47) Berezowski, V.; Landry, C.; Lundquist, S.; Dehouck, L.; Cecchelli, R.; Dehouck, M. P.; Fenart, L. Transport screening of drug cocktails through an in vitro blood-brain barrier: is it a good strategy for increasing the throughput of the discovery pipeline? Pharm. Res. 2004, 21, 756−760. (48) Gaillard, P. J.; de Boer, A. G. Relationship between permeability status of the blood-brain barrier and in vitro permeability coefficient of a drug. Eur. J. Pharm. Sci. 2000, 12, 95−102. (49) Madara, J. L. Regulation of the movement of solutes across tight junctions. Annu. Rev. Physiol. 1998, 60, 143−159. (50) Abbott, N. J.; Ronnback, L.; Hansson, E. Astrocyte-endothelial interactions at the blood-brain barrier. Nat. Rev. Neurosci. 2006, 7, 41− 53. (51) Alvarez, J. I.; Katayama, T.; Prat, A. Glial influence on the blood brain barrier. Glia 2013, 61, 1939−1958. (52) Wolburg, H.; Neuhaus, J.; Kniesel, U.; Krauss, B.; Schmid, E. M.; Ocalan, M.; Farrell, C.; Risau, W. Modulation of tight junction structure in blood-brain barrier endothelial cells. Effects of tissue culture, second messengers and cocultured astrocytes. J. Cell. Sci. 1994, 107 (Part 5), 1347−1357. (53) Gaillard, P. J.; van der Sandt, I. C.; Voorwinden, L. H.; Vu, D.; Nielsen, J. L.; de Boer, A. G.; Breimer, D. D. Astrocytes increase the functional expression of P-glycoprotein in an in vitro model of the blood-brain barrier. Pharm. Res. 2000, 17, 1198−1205. (54) Hori, S.; Ohtsuki, S.; Tachikawa, M.; Kimura, N.; Kondo, T.; Watanabe, M.; Nakashima, E.; Terasaki, T. Functional expression of rat ABCG2 on the luminal side of brain capillaries and its enhancement by astrocyte-derived soluble factor(s). J. Neurochem. 2004, 90, 526−536. (55) Pottiez, G.; Duban-Deweer, S.; Deracinois, B.; Gosselet, F.; Camoin, L.; Hachani, J.; Couraud, P. O.; Cecchelli, R.; Dehouck, M. P.; Fenart, L.; Karamanos, Y.; Flahaut, C. A differential proteomic approach identifies structural and functional components that contribute to the differentiation of brain capillary endothelial cells. J. Proteomics 2011, 75, 628−641. (56) Deracinois, B.; Pottiez, G.; Chafey, P.; Teerlink, T.; Camoin, L.; Davids, M.; Broussard, C.; Couraud, P. O.; Dehouck, M. P.; Cecchelli, R.; Karamanos, Y.; Flahaut, C. Glial-cell-mediated re-induction of the blood-brain barrier phenotype in brain capillary endothelial cells: a differential gel electrophoresis study. Proteomics 2013, 13, 1185−1199. (57) Abbott, N. J.; Dolman, D. E.; Drndarski, S.; Fredriksson, S. M. An improved in vitro blood-brain barrier model: rat brain endothelial cells co-cultured with astrocytes. Methods Mol. Biol. 2012, 814, 415− 430. (58) Coisne, C.; Dehouck, L.; Faveeuw, C.; Delplace, Y.; Miller, F.; Landry, C.; Morissette, C.; Fenart, L.; Cecchelli, R.; Tremblay, P.; Dehouck, B. Mouse syngenic in vitro blood-brain barrier model: a new tool to examine inflammatory events in cerebral endothelium. Lab. Invest. 2005, 85, 734−746. (59) Kuntz, M.; Mysiorek, C.; Petrault, O.; Petrault, M.; Uzbekov, R.; Bordet, R.; Fenart, L.; Cecchelli, R.; Berezowski, V. Stroke-induced brain parenchymal injury drives blood-brain barrier early leakage kinetics: a combined in vivo/in vitro study. J. Cereb. Blood Flow Metab. 2014, 34, 95−107. (60) Shayan, G.; Choi, Y. S.; Shusta, E. V.; Shuler, M. L.; Lee, K. H. Murine in vitro model of the blood-brain barrier for evaluating drug transport. Eur. J. Pharm. Sci. 2011, 42, 148−155. (61) Demeuse, P.; Kerkhofs, A.; Struys-Ponsar, C.; Knoops, B.; Remacle, C.; van den Bosch de Aguilar, P. Compartmentalized coculture of rat brain endothelial cells and astrocytes: a syngenic model to study the blood-brain barrier. J. Neurosci. Methods 2002, 121, 21−31. (62) Dehouck, M. P.; Meresse, S.; Delorme, P.; Fruchart, J. C.; Cecchelli, R. An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J. Neurochem. 1990, 54, 1798− 1801. (63) Fabulas-da Costa, A.; Aijjou, R.; Hachani, J.; Landry, C.; Cecchelli, R.; Culot, M. In vitro blood-brain barrier model adapted to M

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

(80) Lippmann, E. S.; Al-Ahmad, A.; Azarin, S. M.; Palecek, S. P.; Shusta, E. V. A retinoic acid-enhanced, multicellular human bloodbrain barrier model derived from stem cell sources. Sci. Rep. 2014, 4, 4160. (81) Franke, H.; Galla, H.; Beuckmann, C. T. Primary cultures of brain microvessel endothelial cells: a valid and flexible model to study drug transport through the blood-brain barrier in vitro. Brain Res. Brain Res. Protoc. 2000, 5, 248−256. (82) Patabendige, A.; Skinner, R. A.; Abbott, N. J. Establishment of a simplified in vitro porcine blood-brain barrier model with high transendothelial electrical resistance. Brain Res. 2013, 1521, 1−15. (83) Boveri, M.; Berezowski, V.; Price, A.; Slupek, S.; Lenfant, A. M.; Benaud, C.; Hartung, T.; Cecchelli, R.; Prieto, P.; Dehouck, M. P. Induction of blood-brain barrier properties in cultured brain capillary endothelial cells: comparison between primary glial cells and C6 cell line. Glia 2005, 51, 187−198. (84) Persidsky, Y.; Stins, M.; Way, D.; Witte, M. H.; Weinand, M.; Kim, K. S.; Bock, P.; Gendelman, H. E.; Fiala, M. A model for monocyte migration through the blood-brain barrier during HIV-1 encephalitis. J. Immunol. 1997, 158, 3499−3510. (85) Bernas, M. J.; Cardoso, F. L.; Daley, S. K.; Weinand, M. E.; Campos, A. R.; Ferreira, A. J.; Hoying, J. B.; Witte, M. H.; Brites, D.; Persidsky, Y.; Ramirez, S. H.; Brito, M. A. Establishment of primary cultures of human brain microvascular endothelial cells to provide an in vitro cellular model of the blood-brain barrier. Nat. Protoc. 2010, 5, 1265−1272. (86) Cioni, C.; Turlizzi, E.; Zanelli, U.; Oliveri, G.; Annunziata, P. Expression of Tight Junction and Drug Efflux Transporter Proteins in an in vitro Model of Human Blood-Brain Barrier. Front. Psychiatry 2012, 3, 47. (87) Ohtsuki, S.; Sato, S.; Yamaguchi, H.; Kamoi, M.; Asashima, T.; Terasaki, T. Exogenous expression of claudin-5 induces barrier properties in cultured rat brain capillary endothelial cells. J. Cell. Physiol. 2007, 210, 81−86. (88) Watanabe, T.; Dohgu, S.; Takata, F.; Nishioku, T.; Nakashima, A.; Futagami, K.; Yamauchi, A.; Kataoka, Y. Paracellular barrier and tight junction protein expression in the immortalized brain endothelial cell lines bEND.3, bEND.5 and mouse brain endothelial cell 4. Biol. Pharm. Bull. 2013, 36, 492−495. (89) Burek, M.; Salvador, E.; Forster, C. Y. Generation of an immortalized murine brain microvascular endothelial cell line as an in vitro blood brain barrier model. J. Visualized Exp. 2012, DOI: 10.3791/4022. (90) Paolinelli, R.; Corada, M.; Ferrarini, L.; Devraj, K.; Artus, C.; Czupalla, C. J.; Rudini, N.; Maddaluno, L.; Papa, E.; Engelhardt, B.; Couraud, P. O.; Liebner, S.; Dejana, E. Wnt activation of immortalized brain endothelial cells as a tool for generating a standardized model of the blood brain barrier in vitro. PLoS One 2013, 8, e70233. (91) Weksler, B. B.; Subileau, E. A.; Perriere, N.; Charneau, P.; Holloway, K.; Leveque, M.; Tricoire-Leignel, H.; Nicotra, A.; Bourdoulous, S.; Turowski, P.; Male, D. K.; Roux, F.; Greenwood, J.; Romero, I. A.; Couraud, P. O. Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J. 2005, 19, 1872−1874. (92) Weksler, B.; Romero, I. A.; Couraud, P. O. The hCMEC/D3 cell line as a model of the human blood brain barrier. Fluids Barriers CNS 2013, DOI: 10.1186/2045-8118-10-16. (93) Urich, E.; Lazic, S. E.; Molnos, J.; Wells, I.; Freskgard, P. O. Transcriptional profiling of human brain endothelial cells reveals key properties crucial for predictive in vitro blood-brain barrier models. PLoS One 2012, 7, e38149. (94) Daniels, B. P.; Cruz-Orengo, L.; Pasieka, T. J.; Couraud, P. O.; Romero, I. A.; Weksler, B.; Cooper, J. A.; Doering, T. L.; Klein, R. S. Immortalized human cerebral microvascular endothelial cells maintain the properties of primary cells in an in vitro model of immune migration across the blood brain barrier. J. Neurosci. Methods 2013, 212, 173−179. (95) Carl, S. M.; Lindley, D. J.; Couraud, P. O.; Weksler, B. B.; Romero, I.; Mowery, S. A.; Knipp, G. T. ABC and SLC transporter

expression and pot substrate characterization across the human CMEC/D3 blood-brain barrier cell line. Mol. Pharmaceutics 2010, 7, 1057−1068. (96) Poller, B.; Gutmann, H.; Krahenbuhl, S.; Weksler, B.; Romero, I.; Couraud, P. O.; Tuffin, G.; Drewe, J.; Huwyler, J. The human brain endothelial cell line hCMEC/D3 as a human blood-brain barrier model for drug transport studies. J. Neurochem. 2008, 107, 1358−1368. (97) Kusch-Poddar, M.; Drewe, J.; Fux, I.; Gutmann, H. Evaluation of the immortalized human brain capillary endothelial cell line BB19 as a human cell culture model for the blood-brain barrier. Brain Res. 2005, 1064, 21−31. (98) Stins, M. F.; Badger, J.; Sik Kim, K. Bacterial invasion and transcytosis in transfected human brain microvascular endothelial cells. Microb. Pathog. 2001, 30, 19−28. (99) Prudhomme, J. G.; Sherman, I. W.; Land, K. M.; Moses, A. V.; Stenglein, S.; Nelson, J. A. Studies of Plasmodium falciparum cytoadherence using immortalized human brain capillary endothelial cells. Int. J. Parasitol. 1996, 26, 647−655. (100) Eigenmann, D. E.; Xue, G.; Kim, K. S.; Moses, A. V.; Hamburger, M.; Oufir, M. Comparative study of four immortalized human brain capillary endothelial cell lines, hCMEC/D3, hBMEC, TY10, and BB19, and optimization of culture conditions, for an in vitro blood-brain barrier model for drug permeability studies. Fluids Barriers CNS 2013, DOI: 10.1186/2045-8118-10-33. (101) Kamiichi, A.; Furihata, T.; Kishida, S.; Ohta, Y.; Saito, K.; Kawamatsu, S.; Chiba, K. Establishment of a new conditionally immortalized cell line from human brain microvascular endothelial cells: a promising tool for human blood-brain barrier studies. Brain Res. 2012, 1488, 113−122. (102) Lippmann, E. S.; Azarin, S. M.; Kay, J. E.; Nessler, R. A.; Wilson, H. K.; Al-Ahmad, A.; Palecek, S. P.; Shusta, E. V. Derivation of blood-brain barrier endothelial cells from human pluripotent stem cells. Nat. Biotechnol. 2012, 30, 783−791. (103) Lippmann, E. S.; Al-Ahmad, A.; Palecek, S. P.; Shusta, E. V. Modeling the blood-brain barrier using stem cell sources. Fluids Barriers CNS 2013, DOI: 10.1186/2045-8118-10-2. (104) Fischer, S.; Wobben, M.; Kleinstuck, J.; Renz, D.; Schaper, W. Effect of astroglial cells on hypoxia-induced permeability in PBMEC cells. Am. J. Physiol. 2000, 279, C935−44. (105) Nakhlband, A.; Omidi, Y. Barrier functionality of porcine and bovine brain capillary endothelial cells. BioImpacts 2011, 1, 153−159. (106) Cantrill, C. A.; Skinner, R. A.; Rothwell, N. J.; Penny, J. I. An immortalised astrocyte cell line maintains the in vivo phenotype of a primary porcine in vitro blood-brain barrier model. Brain Res. 2012, 1479, 17−30. (107) Cucullo, L.; Hossain, M.; Puvenna, V.; Marchi, N.; Janigro, D. The role of shear stress in Blood-Brain Barrier endothelial physiology. BMC Neurosci. 2011, DOI: 10.1186/1471-2202-12-40. (108) Naik, P.; Cucullo, L. In vitro blood-brain barrier models: current and perspective technologies. J. Pharm. Sci. 2012, 101, 1337− 1354. (109) Cucullo, L.; McAllister, M. S.; Kight, K.; Krizanac-Bengez, L.; Marroni, M.; Mayberg, M. R.; Stanness, K. A.; Janigro, D. A new dynamic in vitro model for the multidimensional study of astrocyteendothelial cell interactions at the blood-brain barrier. Brain Res. 2002, 951, 243−254. (110) Cucullo, L.; Hossain, M.; Rapp, E.; Manders, T.; Marchi, N.; Janigro, D. Development of a humanized in vitro blood-brain barrier model to screen for brain penetration of antiepileptic drugs. Epilepsia 2007, 48, 505−516. (111) Cucullo, L.; Couraud, P. O.; Weksler, B.; Romero, I. A.; Hossain, M.; Rapp, E.; Janigro, D. Immortalized human brain endothelial cells and flow-based vascular modeling: a marriage of convenience for rational neurovascular studies. J. Cereb. Blood Flow Metab. 2008, 28, 312−328. (112) Cucullo, L.; Marchi, N.; Hossain, M.; Janigro, D. A dynamic in vitro BBB model for the study of immune cell trafficking into the central nervous system. J. Cereb. Blood Flow Metab. 2011, 31, 767−777. N

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Molecular Pharmaceutics

Review

(113) Cucullo, L.; Hossain, M.; Tierney, W.; Janigro, D. A new dynamic in vitro modular capillaries-venules modular system: cerebrovascular physiology in a box. BMC Neurosci. 2013, DOI: 10.1186/1471-2202-14-18. (114) Neuhaus, W.; Lauer, R.; Oelzant, S.; Fringeli, U. P.; Ecker, G. F.; Noe, C. R. A novel flow based hollow-fiber blood-brain barrier in vitro model with immortalised cell line PBMEC/C1−2. J. Biotechnol. 2006, 125, 127−141. (115) Griep, L. M.; Wolbers, F.; de Wagenaar, B.; ter Braak, P. M.; Weksler, B. B.; Romero, I. A.; Couraud, P. O.; Vermes, I.; van der Meer, A. D.; van den Berg, A. BBB on chip: microfluidic platform to mechanically and biochemically modulate blood-brain barrier function. Biomed. Microdevices 2013, 15, 145−150. (116) Booth, R.; Kim, H. Characterization of a microfluidic in vitro model of the blood-brain barrier (muBBB). Lab Chip 2012, 12, 1784− 1792. (117) Prabhakarpandian, B.; Shen, M. C.; Nichols, J. B.; Mills, I. R.; Sidoryk-Wegrzynowicz, M.; Aschner, M.; Pant, K. SyM-BBB: a microfluidic Blood Brain Barrier model. Lab Chip 2013, 13, 1093− 1101. (118) Yeon, J. H.; Na, D.; Choi, K.; Ryu, S. W.; Choi, C.; Park, J. K. Reliable permeability assay system in a microfluidic device mimicking cerebral vasculatures. Biomed. Microdevices 2012, 14, 1141−1148. (119) Alcendor, D. J.; Block, F. E., III; Cliffel, D. E.; Daniels, J.; Ellacott, K. L.; Goodwin, C. R.; Hofmeister, L. H.; Li, D.; Markov, D. A.; May, J. C.; McCawley, L. J.; McLaughlin, B.; McLean, J. A.; Niswender, K. D.; Pensabene, V.; Seale, K. T.; Sherrod, S. D.; Sung, H. J.; Tabb, D. L.; Webb, D. J.; Wikswo, J. P. Neurovascular unit on a chip: implications for translational applications. Stem Cell Res. Ther. 2013, 4 (Suppl. 1), S18. (120) Mangas-Sanjuan, V.; Gonzalez-Alvarez, I.; Gonzalez-Alvarez, M.; Casabo, V. G.; Bermejo, M. Innovative in vitro method to predict rate and extent of drug delivery to the brain across the blood-brain barrier. Mol. Pharmaceutics 2013, 10, 3822−3831. (121) Ohe, T.; Sato, M.; Tanaka, S.; Fujino, N.; Hata, M.; Shibata, Y.; Kanatani, A.; Fukami, T.; Yamazaki, M.; Chiba, M.; Ishii, Y. Effect of Pglycoprotein-mediated efflux on cerebrospinal fluid/plasma concentration ratio. Drug Metab. Dispos. 2003, 31, 1251−1254. (122) Hellinger, E.; Veszelka, S.; Toth, A. E.; Walter, F.; Kittel, A.; Bakk, M. L.; Tihanyi, K.; Hada, V.; Nakagawa, S.; Duy, T. D.; Niwa, M.; Deli, M. A.; Vastag, M. Comparison of brain capillary endothelial cell-based and epithelial (MDCK-MDR1, Caco-2, and VB-Caco-2) cell-based surrogate blood-brain barrier penetration models. Eur. J. Pharm. Biopharm. 2012, 82, 340−351. (123) Perriere, N.; Demeuse, P.; Garcia, E.; Regina, A.; Debray, M.; Andreux, J. P.; Couvreur, P.; Scherrmann, J. M.; Temsamani, J.; Couraud, P. O.; Deli, M. A.; Roux, F. Puromycin-based purification of rat brain capillary endothelial cell cultures. Effect on the expression of blood-brain barrier-specific properties. J. Neurochem. 2005, 93, 279− 289. (124) Wuest, D. M.; Wing, A. M.; Lee, K. H. Membrane configuration optimization for a murine in vitro blood-brain barrier model. J. Neurosci. Methods 2013, 212, 211−221. (125) Fazakas, C.; Wilhelm, I.; Nagyoszi, P.; Farkas, A. E.; Hasko, J.; Molnar, J.; Bauer, H.; Bauer, H. C.; Ayaydin, F.; Dung, N. T.; Siklos, L.; Krizbai, I. A. Transmigration of melanoma cells through the bloodbrain barrier: role of endothelial tight junctions and melanomareleased serine proteases. PLoS One 2011, 6, e20758. (126) Wuest, D. M.; Lee, K. H. Optimization of endothelial cell growth in a murine in vitro blood-brain barrier model. Biotechnol. J. 2012, 7, 409−417. (127) Rubin, L. L.; Hall, D. E.; Porter, S.; Barbu, K.; Cannon, C.; Horner, H. C.; Janatpour, M.; Liaw, C. W.; Manning, K.; Morales, J. A cell culture model of the blood-brain barrier. J. Cell Biol. 1991, 115, 1725−1735. (128) Hoheisel, D.; Nitz, T.; Franke, H.; Wegener, J.; Hakvoort, A.; Tilling, T.; Galla, H. J. Hydrocortisone reinforces the blood-brain properties in a serum free cell culture system. Biochem. Biophys. Res. Commun. 1998, 247, 312−315.

(129) Paolinelli, R.; Corada, M.; Ferrarini, L.; Devraj, K.; Artus, C.; Czupalla, C. J.; Rudini, N.; Maddaluno, L.; Papa, E.; Engelhardt, B.; Couraud, P. O.; Liebner, S.; Dejana, E. Wnt activation of immortalized brain endothelial cells as a tool for generating a standardized model of the blood brain barrier in vitro. PLoS One 2013, 8, e70233. (130) Helms, H. C.; Waagepetersen, H. S.; Nielsen, C. U.; Brodin, B. Paracellular tightness and claudin-5 expression is increased in the BCEC/astrocyte blood-brain barrier model by increasing media buffer capacity during growth. AAPS J. 2010, 12, 759−770. (131) Nitz, T.; Eisenblatter, T.; Psathaki, K.; Galla, H. J. Serumderived factors weaken the barrier properties of cultured porcine brain capillary endothelial cells in vitro. Brain Res. 2003, 981, 30−40. (132) Patabendige, A.; Skinner, R. A.; Morgan, L.; Abbott, N. J. A detailed method for preparation of a functional and flexible bloodbrain barrier model using porcine brain endothelial cells. Brain Res. 2013, 1521, 16−30.

O

dx.doi.org/10.1021/mp500046f | Mol. Pharmaceutics XXXX, XXX, XXX−XXX