Understanding the blood–brain barrier using gene and protein expression profiling technologies

Understanding the blood–brain barrier using gene and protein expression profiling technologies

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8 available at www.sciencedirect.com www.elsevier.com/locate/brainresrev Review Understandi...

941KB Sizes 0 Downloads 20 Views

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

available at www.sciencedirect.com

www.elsevier.com/locate/brainresrev

Review

Understanding the blood–brain barrier using gene and protein expression profiling technologies Gwënaël Pottiez, Christophe Flahaut, Roméo Cecchelli, Yannis Karamanos⁎ Univ Lille Nord de France, F-59000 Lille, France UArtois, LBHE, F-62307 Lens, France IMPRT-IFR114, F-59000 Lille, France

A R T I C LE I N FO

AB S T R A C T

Article history:

The blood–brain barrier (BBB) contributes to the brain homeostasis by regulating the passage of

Accepted 12 September 2009

endogenous and exogenous compounds. This function is in part due to well-known proteins

Available online 19 September 2009

such as tight junction proteins, plasma membrane transporters and metabolic barrier proteins. Over the last decade, genomics and proteomics have emerged as supplementary tools for BBB

Keywords:

research. The development of genomic and proteomic technologies has provided several

Blood–brain barrier

means to extend the BBB knowledge and to investigate additional routes for the bypass of this

BBB

barrier. These profiling technologies have been used on BBB models to decipher the

Endothelial cells

physiological characteristics and, under stress conditions, to understand the molecular

Cerebral endothelial cells

mechanisms of brain diseases. In this review, we will report and discuss the genomic and

Proteomics

proteomic studies recently carried out to enhance the understanding of BBB features.

Genomics

© 2009 Elsevier B.V. All rights reserved.

Contents 1. 2. 3. 4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . The expression profiling tools. . . . . . . . . . . . . . . . . . BBB in physiological conditions . . . . . . . . . . . . . . . . . BBB diseases and dysfunction. . . . . . . . . . . . . . . . . . 4.1. New target discovery for stroke with proteomic tools . 4.2. Immune reaction regulation. . . . . . . . . . . . . . . 4.3. Toxicological effect. . . . . . . . . . . . . . . . . . . . 4.4. Electromagnetic fields effects . . . . . . . . . . . . . . 4.5. Expression profiling methods in clinical research . . . 4.6. Concluding remarks . . . . . . . . . . . . . . . . . . . Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

84 86 88 89 89 91 92 93 93 95 96 96

⁎ Corresponding author. Fax: +33 3 21 79 17 36. E-mail address: [email protected] (Y. Karamanos). Abbreviations: EC, endothelial cells; BCEC, brain capillary endothelial cells; BMEC, brain microvascular endothelial cells; BBB, blood– brain barrier; CNS, central nervous system; CSF, cerebro-spinal fluid; 2-DE, 2-dimensional gel electrophoresis; LC, liquid chromatography; MS, mass spectrometry 0165-0173/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.brainresrev.2009.09.004

84

1.

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

Introduction

To preserve its vital function, the central nervous system (CNS) is isolated from the bloodstream by “barriers.” The cerebral capillaries (referred to as the blood–brain barrier, BBB) comprising approximately 95% of the total area of barriers between blood and brain are the main entry route for molecules into the CNS (Cecchelli et al., 2007). The other barriers between blood and brain are the epithelia of the choroid plexus and the other circumventricular organs (referred to as the blood–cerebrospinal fluid barrier) and the arachnoid membranes, separated from the pia mater by the subarachnoid space in which the cerebrospinal fluid flows (Abbott et al., 2006); (Cecchelli et al., 2007). The BBB is recognized as a dynamic interface that, by regulating the exchange of substances between blood and brain, maintains optimal conditions for neuronal and glial functions. This unique selective barrier (Cecchelli et al., 2007) is supported by the endothelial cells (EC) of the brain capillaries. The capillaries are surrounded by a tubular sheath of astrocytic end-feet, which induces BBB properties to the EC (Fig. 1). In the brain, only blood capillaries are endowed with a full-blown BBB phenotype. Vessels of larger diameter have comparably increased levels of leakiness and tend to have an intermediate barrier function (Marchi et al., 2004). The brain capillary endothelial cells (BCEC) maintain brain homeostasis by filtering exogenous compounds and transporting nutrients, ions, hormones and recruit immune cells in view of their transfer from blood to brain. The main physiological features of the BCEC are a lack of fenestration and pinocytic vesicles, a reinforcement of tight junctions (TJs) which form a physical barrier (Reese and Karnovsky, 1967) and many more mito-

chondria than in the peripheral EC. These features give a high trans-endothelial electrical resistance (TEER) due to low paracellular and transcellular permeabilities of the EC-monolayer towards hydrophilic and hydrophobic compounds (Ballabh et al., 2004). Besides this function of physical barrier, the BBB also acts as a metabolic barrier due to the presence of enzymes such as γ-glutamyl-transpeptidase (γ-GT) (Lawrenson et al., 1999), monoamine oxidase (MAO) (Lasbennes et al., 1988) and alkaline phosphatase (AP), which show a rather higher activity than in their peripheral counterparts. Furthermore, the presence of efflux pumps promotes the selectivity of the BBB, since various hydrophobic compounds such as certain therapeutic compounds, like anti-cancer drugs, are unable to reach therapeutic concentrations into the brain (Begley, 2004) (Fig. 2). Extensive knowledge of constitutive membrane receptors and transporters at the BBB, or a metabolic pathway associated to the barrier function, would allow a better understanding of this particular tissue and therefore offer, in the context of drug delivery, new therapeutic targets for the pharmaceutical industries. The emergence of genomic and proteomic methods appears as additional and complementary tools to complete our knowledge of the BBB. Over the last 10 years the molecular properties of the brain endothelium have been increasingly studied using expression profiling technologies. These methods were successfully used in cancer research to provide biomarkers for diagnosis and to identify genes or proteins responsible for cell deregulation during the cancer (Baak et al., 2005). Proteomics were used for the determination of disease mechanisms in complex tissues such as the whole brain (Moron and Devi, 2007). To study BBB features or drug transport to the brain one needs in vivo and in vitro models. The drug transport is studied

Fig. 1 – Brain capillary endothelial cells support the BBB. The blood–brain barrier is formed by endothelial cells at the level of the cerebral capillaries, which are surrounded by a tubular sheath of astrocytic end-feet. In the basal lamina, between endothelium and astrocytic end-feet, pericytes are inserted.

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

85

Fig. 2 – Schematic drawing of main features of brain capillary endothelial cells. Differentiated brain capillary endothelial cells present less paracellular pathway with the reinforcement of tight junctions and a lack of fenestration. These cells contain many mitochondria. Brain protection is provided by efflux pumps and efficient exogenous compound degradation enzymes such as γ-glutamyl-transpeptidase, monoamine oxidase.

by quantification of compounds which reach the brain parenchyma and the cerebro-spinal fluid (CSF). In in vitro models, the abluminal compartment mimics the brain parenchyma (Cecchelli et al., 2007). In addition, in vivo models which generally use mice or rats, give information similar to the situation in humans. However, in vivo experiments require high numbers of animals since the microvasculature must be extracted from the brain. Given that capillaries represent only one thousandth of the brain volume, laser microdissection or enzymatic capillary extraction methods are necessary, and could provoke endothelial cell stress. With in vitro models, tissue complexity cannot be replicated, but these models are more flexible, the culture medium can be changed, supplemented with various compounds, and larger quantities of cells can be easily obtained. Many in vitro models require specific conditions of isolation, purification and growing. The term brain capillary endothelial cells (BCEC) is used when pure capillary endothelial cells are obtained by mechanical homogenization and filtration (Dehouck et al., 1990). In contrast the term brain microvascular endothelial cells (BMEC) is used, when the endothelial cells are prepared by enzymatic digestion and the culture consists of a mixture of endothelial cells from capillaries, arterioles and veinules and sometimes even contains other cells, such as pericytes. Culture conditions can vary according to the cell harvesting methods (Dehouck et al., 1990); (Perriere et al., 2005), the co-culture with perivascular brain cells (Dehouck et al., 1990); (Nakagawa et al., 2009), the addition of cell growing and differentiation compounds such as corticoids (Calabria et al., 2006; Perriere et al., 2007) and the presence of flow (Stanness et al., 1999). The quality of the samples issued from in vivo and in vitro models will be greatly influenced by culture and sample preparation conditions. Therefore the expression profiling methodology used should be adapted to each individual situation.

Genomic and proteomic approaches are based on the sequence of the entire genome. The first complete DNA sequence was from phi X 174 bacteriophage (Sanger et al., 1977). This was the starting point of what is known today as genomics. Currently, the genomes of approximately eight species of mammals are completely sequenced. From the DNA sequence of organisms, information can be obtained on all proteins or genes in this organism but unfortunately not about tissue or cell protein expression. Therefore, quantification of transcript and of protein expression would seem to be essential, as the cells are strongly affected by neighboring ones and tissues, and stress or disease may also modulate the transcript and protein expression. The general term “genomic” includes genome sequencing, gene expression analysis and transcript quantification, called “transcriptomics.” In eukaryote cells, mRNAs are maturated by alternative splicing which produces various isoforms of proteins. Moreover, direct factors of the regulation of transcription, elongation and mRNA turnover influence the cellular mRNA levels. On the one hand, the protein expression depends on the mRNA regulation and on the other hand on the translation regulation, the post-translational modifications (PTM) and the protein turnover. This explains why gene and protein expression have a poor correlation. Wilkins et al. (1996) was the first to use the term “proteome.” Literally, in a cell, a tissue or an organism, the proteome represents all proteins coded by the genome and proteomics correspond then to the study of this proteome. This concept includes a dynamic evolution of the proteome, indeed, environment, stress, and cellular state lead to the modification of the proteome. Therefore, proteomic study highlights the proteome state of a given cell type at a specific time and under specific conditions. Moreover, the first difficulties encountered in the study of proteins are the

86

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

sample complexity, the low level of protein expression in the cell and the availability of the biological samples. Many genomic and proteomic studies have been carried out to further the understanding of BBB features. The upsurge in their use has stimulated us to discuss their application in this review. As pointed out recently (Ge et al., 2005) the identity of vascular segments from which the endothelial cells are derived is, typically speaking, not elaborated, and, in most cases, the starting material is a mixture of the different microvascular branches. Since variations may also occur between species, in the following sections of this review, in addition to the technique used for expression profiling, information will be given on the origin of the biological material used (species, cells, models, …).

2.

The expression profiling tools

The goal of genomics and especially the transcriptomic study is the global assessment of variations in gene expression by the relative quantification of mRNA produced by cells. The first steps consist in extracting mRNA, retro-transcribing it into cDNA to enhance its stability and amplifying it by polymerase chain reaction (PCR). Historically, the relative quantification of the transcript was carried out by northern blot. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) became the “gold standard” for mRNA quantification. The large scale study of gene expression was carried out on the advent of DNA microarray which consists in the covalent linkage of many thousands of genes (genomic DNA) or expressed genes (expressed sequence tags) on a solid and inert matrix (nylon membrane, glass slide, plastic, …). The hybridization of the fluorescent cDNA samples with the cDNA probes of the microarray allows relative quantification. Supplemental techniques have been developed to improve gene analysis. Serial analysis of gene expression (SAGE) provides an analysis of the frequency of expression of mRNA by the sequential analysis of a large number of short cDNA fragments, each being the signature of a gene (Velculescu et al., 1995). The main advantage of SAGE remains that the quantitative analysis of all transcripts is performed without a prior selection of known genes, unlike systems based on the hybridization of nucleic probe collections which require the analyzed genes to be previously characterized. Concomitantly, suppressive subtractive hybridization (SSH), relies on the removal of double strands of DNA from samples of interest formed by hybridization with a control sample, thus eliminating cDNA or genomic DNA of similar abundance, and retaining differentially expressed, or variable in sequence, transcripts or genomic sequences (Diatchenko et al., 1996). Contrary to the SAGE method, SSH has been recently adapted to the microarray analyses (Munir et al., 2004). These techniques have been used for genomic analysis of BBB features (see for review: (Calabria and Shusta, 2006; Shusta, 2005)). The proteomic studies first begin with an extraction step of molecules of interest. Compared to transcriptomics, proteomics allows access to a subset of proteome by combining cellular fractionation and protein extraction protocols. Even under these conditions, the extracted proteins form a complex

sample which, in most cases, must be separated for their quantification and identification (Aebersold and Mann, 2003). Originally the method used for protein separation was the 2dimensional gel electrophoresis (2-DE). In that case, the separation, visualization and quantification of proteins is done using a polyacrylamide meshwork where proteins move according to their percentage of ionization in an electrolytic medium subjected to an electric field (O'Farrell, 1975). The first dimension separates, via a pH gradient, proteins according to their respective isoelectric point (pI). This separation relies on the net charges of proteins at a given pH. Therefore, when the pI equals the pH, the proteins stop migrating since they become electrically neutral. The second separation is classically based on the molecular weight (MW) of the proteins (Fig. 3). After electrophoretic separation, the in-mesh trapped proteins are revealed by a quantitative staining step based onto the ionic and hydrophobic affinity of some chromophore or fluorescent dyes. The stained proteins appear as dots or spots and the gels can then be digitized using a densitometerscanner and images qualitatively and quantitatively compared by dedicated image analysis software which provide an estimation of the relative expression of proteins under different conditions (Miller et al., 2006) accompanied by appropriate statistics when experiments are repeated. Although, in theory, 2-DE can separate all proteins, in practice this technique has many drawbacks. The meshwork does not allow a separation of proteins with a low (<10 kDa) and high molecular weight (>200 kDa), the observable pIs are between 3 and 11 and the superposition of protein spots is frequently observed (Campostrini et al., 2005). Due to in-gel diffusion and limitations in the threshold of protein staining, the visualization of the proteins requires a significant amount of proteins. In addition variations in protein migration between gels occur frequently. The recent improvement of 2-DE designated under the acronym DIGE (for Differential Gel Electrophoresis) relies on the distinctive fluorescent labeling of proteins from two distinct samples, followed by their mixing together with a quantitative normalization standard before being subjected to a co-separation by 2-DE (Unlu et al., 1997). DIGE leads to a better comparability and a more accurate relative quantification of samples. Overall, 2-DE based approaches display the main benefit of producing separation patterns where the heterogeneity and dynamic range of proteins are clearly observable through the gel image. The second way of quantifying and identifying most proteins in a complex mixture is based on high performance liquid chromatography (HPLC) coupled with mass spectrometry. Although the technology of intact protein fractionation in two dimensions (PF-2D) is available, it has not been developed well in the proteomics field as demonstrated by the very small number of PF-2D-based publications. Concomitantly, shotgun proteomics has rapidly emerged as a complementary approach to 2-DE. Shotgun proteomics relies on the quantification and identification of proteins in complex mixtures using high performance liquid chromatography combined with mass spectrometry. In shotgun proteomics, the proteins are enzymatically cleaved and the resulting peptides are separated either by one-dimensional liquid chromatography (1D-LC) where peptides are often separated according to their

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

87

Fig. 3 – Picture of a 2-dimensional electrophoresis gel. 2-dimensional electrophoresis gel provides, on the first dimension, a separation of the proteins according to their isoelectric point (pI) and on the second dimension, a separation of the proteins according to their molecular weight (MW).

hydrophobicity on a reverse phase column, or by two dimensional liquid chromatography (2D-LC) where peptides are firstly separated according to their charges on an ion exchange column before being resolved according to their hydrophobicity (Fournier et al., 2007; Washburn et al., 2001). Tandem mass spectrometry is then used to identify the peptides and multiple peptide identifications assembled to allow protein identification. Through these methods, large amounts of heterogeneous proteins are cleaved in multiple peptides, increasing dramatically the complexity of the samples which constitutes one of the main drawbacks of shotgun proteomics, the second and third being mainly the perfectible quantification based on mass spectrometry and the absence of protein map reflecting the heterogeneity of the sample, respectively. To overcome the bias of quantification, a first method of chemical derivatization called “isotope coded affinity tags” (ICAT) was developed to be used prior to the HPLC-separation (Gygi et al., 1999). This technique uses the derivatization of cysteine residues with biotin or biotin labelled with stable isotopes. The biotinylated peptides can then be specifically isolated by avidin–biotin affinity chromatography while the relative quantification is performed by mass spectrometry (MS) on the basis of the isotopic ratio of co-eluting peptides. By this approach purification and relative quantification of the modified peptides is possible. Several variants based on this principle of isotopic labeling have been developed over the years (Watt et al., 2003). Proteins can also be labelled earlier during the cell culture with stable isotope labelled amino acids. This metabolic labeling technique called “stable isotope labeling with amino acids in cell culture” (SILAC) (Ong et al., 2002), provides several benefits that have been recently

reviewed (Ong and Mann, 2007). However, although the SILAC literature is rich, SILAC cannot be applied on in vivo systems. Previously confined to the comparison of only two distinct samples, isotope-based MS quantification methodologies have reached a breakthrough with the advent of the “isobaric tag for relative and absolute quantitation” (iTRAQ) (Ross et al., 2004) where 4 and recently 8 distinct samples can now be compared together. With both advantages and drawbacks, the revolution of iTRAQ mainly comes from its MSquantification process that imposes the MS fragmentation of peptides (Phanstiel et al., 2009). Finally, “label free” shotgun proteomics re-emerged in 2004 due to its low cost and its ability to compare large scale samples (Wiener et al., 2004). However, the label free approach requires a multiplication of experiments to limit the bias due to the low accuracy of MSquantification. In addition to the relative quantification, the MS-absolute quantification of peptides from trypsin digestion, called multiple reaction monitoring (MRM), uses synthesized stable isotope-labelled peptides as the internal reference (Baty and Robinson, 1977; Padieu and Maume, 1976). The strategy employs in silico peptide selection for MS analysis. The comparison of the mass signal intensity of a peptide with its corresponding internal reference provides absolute quantification of this peptide. Given that the internal reference amount is known, the quantity ratio with the sample peptide allows absolute peptide quantification. To conclude, ICAT and MRM are the only techniques used for the BBB features analysis (Haqqani et al., 2005, 2007; Kamiie et al., 2008). Peptides issued from in-gel (SDS-PAGE, 2-DE) or in-solution (HPLC) digestion are then analyzed by mass spectrometry (MS) and/or tandem mass spectrometry (MS/MS). MS consist, with extreme accuracy, in determining, following ionization, the

88

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

molecular mass of peptides by measuring of their mass to charge ratio (m/z) while MS/MS determines the molecular mass of fragments generated by dissociation or collision of these ionized peptides. Therefore, due to its rapidity and accuracy and concomitantly to the development of bioinformatics, mass spectrometry became at the beginning of this century the essential analytical technique for protein identification (Aebersold and Mann, 2003). Indeed, although the measurement of the mass of only one peptide of a given protein is not sufficiently informative to specifically identify it, the determination of a sub-set of peptide masses, called peptide mass fingerprint (PMF), of this protein can be sufficient to do so. Inversely, as it is related to the amino acid sequence of a peptide, the MS/MS fragmentation pattern also called peptide fragmentation fingerprint (PFF) of only one peptide from a given protein allows the unambiguous identification of the protein. However, protein identification needs confirming by other methods such as immunodetection with western blot or PCR-based quantifying of gene transcripts. Moreover, the identification of proteins is essential to monitor the variations of their concentration in tissues and the pathways in which these proteins are involved.

3.

BBB in physiological conditions

The transcript analysis of brain endothelial cells in vivo and in vitro, gives information about the loss of differentiation incurred by cell extraction and artefacts due to cell culture. The genomic analysis of in vitro and in vivo BMEC would therefore increase our understanding of the advantages and drawbacks of these two complementary approaches. Extraction of microvessel EC from rat brain for in vitro culture and culture conditions induces cell dedifferentiation, the comparison to isolated microvessels from rat brain show, for example, loss of BBB specific markers such as multidrug resistant protein 1 (MDR1) and transferrin receptor (Calabria and Shusta, 2008). However, some loss of gene expression such as the glucose transporter type 1 (Glut-1) could be a culture artefact since the amount of glucose in the culture medium is generally increased many-fold when compared to the one in vivo, therefore transport of glucose from the blood to the brain compartment appears unnecessary. Moreover, the endothelium was shown to participate in neuronal differentiation (Shen et al., 2004), but the genes involved in brain development and maturation, expressed by BMEC, are also lost in vitro (Calabria and Shusta, 2008). Furthermore, the purification of BMEC cells is difficult, but the application of puromycin, a P-glycoprotein (P-gp) substrate, allows BMEC to be cultured specifically (Perriere et al., 2007). This treatment causes an up-regulation of three BBB specific genes, which suggests that contaminating cells are negative regulators of BMEC gene products. Hydrocortisone (HC) (Hoheisel et al., 1998) treatment induces an increase in genes such as Serpin 1 and Reck, two antiproteases, which reveals that BMEC are important factors in angiogenesis and proliferation (Calabria and Shusta, 2008). MDR1a gene increases with HC treatment. Other specific BBB markers remain unchanged, so HC alone is not sufficient to restore a whole BBB phenotype. Furthermore, HC provides BBB features such as MDR1a gene transcription

and more reduced monolayer permeability (Calabria and Shusta, 2008). The transcript analysis of the human cerebral endothelial cells (HCEC) compared to human umbilical vein endothelial cells (HUVEC) with Human Cytokine Expression Array highlights a neuroprotective function as well as angiogenic and immunoregulatory activities (Kallmann et al., 2002). The factors identified are involved in neurone protection or neurotrophic activity, for example brain-derived neurotrophic factor (BDNF), which promotes the survival of many neural populations could have a therapeutic potential in the treatment of the CNS (Pan et al., 1998). Finally, interleukin-6 (IL-6), an inflammatory signalling molecule, is strongly represented in brain endothelium and is involved in the recruitment of immune cells in the inflammatory area as well as in the activation of astrocytes following brain injury. It also seems to be a neuroprotector since it induces the expression of metallothionein, a protein responsible for metal detoxification during cell death (Penkowa et al., 1999). This transcript analysis presents preferentially released factors and shows a close correlation between endothelia and their environment (Fig. 4) (Kallmann et al., 2002). A comparative proteomic analysis of rat brain microvascular endothelial cells and coronary microvascular endothelial cells highlights BBB characteristic proteins. BMEC, due to their specificity and their singular structures, have a particularity in their metabolic and structural proteins. Indeed, it has been shown (Lu et al., 2007) that the cytoskeleton associated proteins are involved in BMEC differentiation, for example the macrophage-capping protein G (CapG), a gelsolin/villin family protein, interacts with actin to reduce actin filaments and cap end-barbed actin filaments. The interaction of this protein with the cytoskeleton and DNA may play a role in the

Fig. 4 – Relevant proteins for the endothelial cells deduced from transcript analysis (adapted from (Kallmann et al., 2002)). Transcript analyses of brain capillary endothelial cells reveal that endothelium contribute to neuroprotection, angiogenesis, growth-supporting and immunoregulation. Due to the localization of the BBB endothelium in close proximity to the central nervous system function (CNS), this tissue may have unknown influence on the CNS.

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

regulation of the cytoplasmic and nuclear structure of actin. Moreover, in the aortic endothelial cells, this protein contributes to the increase in blood flow (Pellieux et al., 2003). The chloride intracellular channel protein 4 (CLIC-4) is up-regulated in the BMEC. This protein, which is to a small extent localized in the caveolae membranes, is associated to the dynamin I, actin and α-tubulin which suggests a link with transport and fusion of the vesicle to the membrane (Suginta et al., 2001). CLIC-4 is also involved in the proliferative phenotype of the EC (Bohman et al., 2005). Cytosol aminopeptidase, increased in brain endothelium, is involved in proliferative phenotype (Sato, 2004). Its function is the hydrolysis or maturation of peptides and could have a barrier function preventing transmitters from entering the BBB. Interestingly, proteins associated with the tight junctions are up-regulated in the BMEC. The protein called calcium/calmodulin-dependent serine protein kinase membrane-associated guanylate kinase (CASK), possess a PDZ domain, like zonula occludens-1 (ZO-1), which is a junctional adhesion molecule (JAM) binding domain and was shown to co-localize with JAM (MartinezEstrada et al., 2001). VAMP associated protein of 33 kDa (VAP33) was shown to co-localize with occludin, a transmembrane tight junction protein, and is also associated with the vesicular membrane (Lapierre et al., 1999). Finally, IL-6 is up-regulated in the brain endothelial cells and is also involved in the stress response. Recently, a differential proteomic approach was initiated, using the in vitro BBB model developed in our laboratory, made up of pure bovine brain capillary endothelial cells (BCEC) cocultured with glial cells. These particular culture conditions induce differentiation of endothelial cells which provides an in vivo BBB phenotype (Dehouck et al., 1992). Therefore, the first challenge was to adapt the methodology to the in vitro BBB model. It was demonstrated that culture conditions strongly affected the results of the proteome analysis using 2-DE. To obtain informative gels with no interference from the culture medium proteins, the BCEC had to be harvested by collagenase which allows conservation of the cell integrity and provides isolated BCEC for proteome analysis (Pottiez et al., 2009). The second part of the investigation was to carry out the first differential proteomic analysis of the differentiated BCEC. To identify the main pathways involved in the dynamic regulation of BBB function, we compared the phenotypic differences between fully differentiated BCEC, cultured with glial cells, and undifferentiated cells, cultured without glial cells. The results indicated that actin cytoskeleton remodeling is closely involved in BBB differentiation. The use of Triton X100 for protein extraction, after cell harvesting, provides supplemental information. Keeping in mind that this detergent partially solubilizes the membranes and cortical actin, it was shown that peripheral localization of actin filaments, associated to the differentiation, implicates proteins like gelsolin and filamin-A and actin-bundling proteins (Pottiez et al., 2009). This in vitro model has already been used for other applications as well, such as inflammation and stroke and could be used, in the future, for proteome analysis during patho-physiology. The absolute quantification of multiple membrane proteins in very complex samples such as mammalian tissues can be studied by liquid chromatography and identified by

89

mass spectrometry (LC-MS/MS) combined with in silico peptide selection criteria, using multiple reaction monitoring (MRM). By applying this focused proteomic to plasma membrane transporter proteins it will be possible, for the first time (Fig. 5) (Kamiie et al., 2008), to determine simultaneously the expression level of multiple membrane transporters such as multi drug resistant protein 1 A (Mdr1a), breast cancer resistant protein (BCRP), and Glut-1 amongst others, and provide a quantitative atlas of membrane transporter proteins. Transporters and efflux pumps are numerous and the understanding of the regulation of the transport through the BBB could indicate alternative ways of reaching the brain. Therefore, this technique could be a new area for pharmaceutical research, providing the quantification of known proteins localized at the membrane of the brain microvessels (Ohtsuki and Terasaki, 2007). Metabolomics is a method for analysis of metabolites derived from peptides or other compounds. One study has focused on the analysis of metabolites of substance P, a neuropeptide involved in inflammation and edema. This peptide seems to play a role in the increase in permeability of the BBB during traumatic brain injury and ischemia (Donkin et al., 2007). In addition, substance P could be cut by the peptidases synthesized by the BBB endothelial cells. The degradation of substance P was studied by LC-MS/MS (Chappa et al., 2007). The analysis revealed seven peptides produced in the presence of bovine BMEC. The comparison of in vitro and in vivo results showed a high correlation and demonstrated that the in vitro bovine BMEC model could be used to investigate the metabolism of neuropeptides and offer new fields for pharmacoproteomics. Furthermore, identification of peptides produced by the BBB has led to the hypothesis of the existence of a peptidase synthesized by this particular endothelium. The previously presented studies revealed a large number of applications for proteomic and genomic analysis. Indeed, these methods provide information on cell behavior under physiological conditions or in cell cultures. Genomics and proteomics normally provide information about specific markers; unfortunately in the case of BBB, it has not been possible to measure the γ-GT, MAO and AP protein expression levels by proteomics, probably due to their relatively low abundance and their particular properties, which implies difficulties in solubilization and necessitate further treatment for proteomic research. These BBB markers were never detected on 2-DE gels and no data are available in relevant literature on their identification using liquid chromatography. Nevertheless, proteomics includes techniques for analyzing compounds, degradation products, receptors and transporters and opens a new route for understanding the BBB and brain targeting.

4.

BBB diseases and dysfunction

4.1.

New target discovery for stroke with proteomic tools

Stroke and BBB dysfunction have often been studied in order to perform specific therapeutic targeting to avoid damage of the brain area concerned and also to design rapid diagnostic tools. Stroke is blood-flow failure in a particular brain area.

90

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

Fig. 5 – Schematic representation of the protein quantification strategy developed by Kamiie et al. (2008). Tissue preparation followed by protein digestion (multiple channel MRM analysis), and theoretical protein digestion (peptide probe selection) are carried out in parallel. Selected peptides, in silico, are synthesized with stable isotope amino-acids. Sample and stable isotopelabelled peptides are mixed and analyzed by liquid chromatography coupled with tandem mass spectrometry. The quantity of target proteins is calculated from the peak area ratio (endogenous peptide/stable isotope-labelled peptide).

This suggests a lack of nutrients and oxygen. Previously, effects of this pathology on BBB integrity was studied e.g. endothelial permeability, cell shape was observed by microscopy and immunochemistry. The results of those methods have pointed out some mechanisms of BBB dysfunction and could be beneficial for brain targeting with new therapeutic components or stroke treatments. To mimic a stroke, various models have been used. The application of in vivo models is based on an artificial obstruction of blood flow in the brain leading to a lack of nutrients and oxygen. The in vitro model involves cell oxygen deprivation named hypoxia, and oxygen and glucose deprivation (OGD or ischemia). The variation in protein expression during hypoxia, and posthypoxic reoxygenation was recently studied using rat BCEC transfected with immortalizing genes and analyzed with 2-DE (Haseloff et al., 2006). The proteins with a variable expression could be sorted into three categories, proteins from mitochondria and endoplasmic reticulum, proteins associated with the cytoskeleton, and the proteins of the glycolysis pathway. After hypoxia, protein expression and the enzymatic activity of the glycolysis pathway were up-regulated. These results suggest that endothelial cells respond to the hypoxic stress with an increase in glucose degradation. The modulation of cytoskeleton-associated proteins implies a cell structure rearrangement. Finally, during posthypoxic reoxygenation, the up-regulation of most proteins was reduced towards control levels, indicating that, under the used conditions, hypoxia induced metabolic over-expression is reversible (Haseloff et al., 2006). The use of an in vivo BBB model is the most efficient way of studying a stroke response. However, capillary localization induces a low sample quantity for a proteomic approach. The

ICAT method (Haqqani et al., 2005) was used to demonstrate the variation in protein expression after 20 min of transient global cerebral ischemia, and 1, 6 and 24 h after reperfusion of rats. The microvessels were obtained by laser capture dissection, providing approximately 300 captured microvessels. The results presented 50 proteins with a significant expression variation. The up-regulation of the proteins in the early step (1 h) reveals a modification of the cells for an inflammatory and a proliferative phenotype, with proteins as transcription factors and signal transduction molecules. Most of them returned to a basal level after 6 h of reperfusion. After 24 h of reperfusion a second wave of up-regulation appears with proteins as inflammatory cytokines and metalloproteases. These variations seem to be correlated with the BBB disruption observed in this pathology (Haqqani et al., 2005). Recently, the 2-DE and ICAT methods, have been compared, to show their complementarities and to complete our knowledge of ischemia/reperfusion (Haqqani et al., 2007). This study, carried out with rat BMEC immortalized with a plasmid from the simian virus 40 (SV40) (Zhang et al., 2003) presents approximately 200 identified proteins with a significant variation, sorted by their functions. On the one hand, glucose metabolism, stress proteins and antioxidative defense are involved in the endothelial response to brain ischemia, indeed, after reperfusion proteins from these groups were increased. This emphasizes an endothelial cell survival. On the other hand, a new protein category appears with free radical detoxification proteins such as superoxide-dismutase (SOD), formaldehyde dehydrogenase, and thioredoxin and metallothionein. Interestingly, in another study using immortalized rat BMEC (Roux et al., 1994) SOD has been shown to reduce BBB disruption after hypoxia/reoxygenation

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

(Utepbergenov et al., 1998). Finally, the EC suffer from an altered expression of proteins involved in polarization of brain endothelial cells, such as ion and neurotransmitter transporters, which could be involved in the changed permeability of the BBB during ischemia/reperfusion (Haqqani et al., 2007). In addition, the increase in heavy and light chains of clathrin indicates increased endocytosis. This could provide additional information on BBB endothelial cells behavior during stroke and is in agreement with the results from our laboratory (Brillault et al., 2002) demonstrating that endothelial ischemia induces an increase in trans-endothelial transport of compounds mediated by vesicles. Some inflammatory cytokines and growth factors reveal remodeling and an angiogenic cell cycle (Haqqani et al., 2007). Finally, the up-regulation of transforming growth factor beta-1 (TGF-β1) and metalloproteases, and a decrease in secreted extracellular matrix components suggest a proliferation and migration phenotype. In addition, TGF-β1 is an activator of the BBB properties by inducing permeability and increase in P-gp activity in immortalized mouse BMEC (Dohgu et al., 2004). To conclude, brain hypoxia and ischemia studied by proteomic methods suggest, firstly, that glucose metabolism is up-regulated early during an oxygen stress, and secondly, that EC respond to the lack of oxygen with free radical detoxification proteins. Finally, these results also show a strong communication between the BMEC and their environment and tissue modeling following inflammation.

4.2.

Immune reaction regulation

The BCEC participates in the recruitment of immune cells in view of their transfer from blood to the brain during inflammation. Indeed, for these immune cells, the post capillary veinules are the link between blood and the brain. During inflammation or infection, immune cells can cross the BBB to reach the CSF or activate the endothelial cell wall in combination with cytokines such as TNFα. This factor is a powerful activator of endothelial cell inflammatory responses. It induces a reversible reduction of TEER in the BBB (de Vries et al., 1996). Concurrently, TNFα toxicity is linked to the NO level and an enhanced efflux of lactate dehydrogenase (LDH) in BCEC (Zhu et al., 2000). Genomic and proteomic studies, with DNA microarray (Affymetrix) and 2-DE, respectively, allow the determination of the specific brain EC response to TNFα activation of the endothelium (HCEC and HUVEC (Franzén et al., 2003)). The DNA analysis provides high amounts of quantified genes (8000 genes), and requires low quantities of sample. Among other results, it has been shown, that after TNFα activation, the genes of chemotaxis molecules are up-regulated, which supposes a recruitment of bloodstream immune cells. After recruitment, proteins such as intercellular adhesion molecules (ICAM), participate in the trans-endothelial lymphocyte transport. The adhesion molecules are also up-regulated in HCEC and HUVEC treated with TNFα. This suggests an increase in immune cell adhesion and a modification of endothelial cell-cell adhesion. The 2-DE identification of these adhesion proteins would have been difficult, due to their hydrophobic properties. Chemotaxis molecules are generally secreted and could not be observed in 2-DE gels (Franzén et al.,

91

2003). Interestingly, a protein that is up-regulated in BMEC ischemia (Haqqani et al., 2007) has been identified also following TNFα activation (Franzén et al., 2003). The mitochondrial isoform of SOD, an enzyme which catalyses the conversion of superoxide to hydrogen peroxide and oxygen, was up-regulated in human HCEC and HUVEC for their mRNA and protein expression. This supposes the involvement of free radical detoxification. The proteome analysis highlights a down-regulation of cofilin and heat shock protein beta-1 (HSPβ1), two proteins involved in actin polymerization. However, 2-DE separations not only show proteins expression, but also indicate modification of the proteins (post-translational modifications and hydrolysis). For example, cofilin identified in the gel was unchanged at mRNA level, which suggests that cofilin could contain post-translational modifications. Other molecules are brain specific, the signalling system, urokinase plasminogen activator and its receptor, urokinase plasminogen activator receptor (uPA/uPAR) which has a role in cell migration. The signal transcription factor Nuclear factor kappa-B (NFκB) is also involved in brain response, this factor has several stress response gene binding sites, as in inducible nitric oxide synthase (iNOS), SOD, phospholipase C, vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF) and enzymes such as ceramide glycosyltransferase, hemeoxygenase, metalloproteinase, collagenase and gelatinase (Franzén et al., 2003). Moreover, the TNFα mediated-increase in cerebrovascular permeability could be linked to NFκB (Trickler et al., 2005). NFκB was also shown to be involved in the hypoxia and hypoxia/reoxygenation stress response (Brown and Davis, 2002; Witt et al., 2005). Moreover, two proteins involved in vesicle transport, synaptic vesicles membrane protein (VAT1) and Rab GDP dissociation inhibitor beta (GDIB) were down-regulated in TNFα stimulated HCEC. These results indicate that molecules, such as NFκB and SOD, are potential targets for the modulation of inflammation. During progressive HIV-1 disease, infection is associated with dementia. The neuropathogenesis of this infection is underlined by the trans-endothelial migration of HIV-infected leukocytes into the central nervous system. Physiological analysis has revealed a modification of tight junction proteins, but this phenomenon is not clearly understood (Dallasta et al., 1999). Significant studies have demonstrated that cellular and viral proteins can alter vascular BBB permeability and disrupt the BBB. For example, the glycoprotein of the HIV envelope, gp-120, is involved in blood–brain barrier damage and neuronal cell death (Kanmogne et al., 2002). As previously described, TNFα is a factor involved in this infection and inflammation. Another factor, IL-6, is increased in the CSF of patients with HIV-1-associated dementia. Both, HIV-1 and IL6, synergicaly increase monocyte migration across brain endothelium (Chaudhuri et al., 2008). Furthermore, Tat released by HIV infected cells, induces activation of NFκB and induces an oxidative stress (Toborek et al., 2003). The coculture of human BMEC with HIV-infected monocytes-derived macrophages (MDM) induces up-regulation of metabolic proteins, voltage-gated ion channels, heat shock, transport, cytoskeletal, and calcium regulatory and binding proteins in brain endothelial cells. Finally, LDH is also upregulated with infected MDM (Ricardo-Dukelow et al., 2007). Amongst the proteins that are involved in cytoskeleton regulation following

92

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

TNFα treatment, cofilin and HSP-β1 were identified again. Moreover, numerous proteins are associated with the structure of the cytoskeleton. The soluble isoform of the SOD has been identified also and the up-regulation of this protein presents an interaction between inflammation and free radical detoxification. Finally, the up-regulated proteins are also associated with calcium regulation (Ricardo-Dukelow et al., 2007). To conclude, the proteomic approach provides more details of the mechanisms of patho-physiological response. In the case of HIV, the cytoskeleton was modified as well as the calcium regulation, which is one of the most important ions in BBB physiology (Deli et al., 2005). Indeed, intracellular calcium ion concentration is increased following BBB treatment with compounds able to induce permeabilization of the EC monolayer (Revest et al., 1991), such as bradykinin (Bartus et al., 1996). Calcium induces an increase of NO, cGMP and the number of pinocytotic vesicles (Liu et al., 2008), a severing of actin filaments (Pottiez et al., 2009) and a modulation of the tight junctions (Lai et al., 2005). The pertussis toxin (PTX) is used as an adjuvant that promotes development of tissue specific experimental autoimmune diseases such as experimental allergic encephalomyelitis (EAE). However the mechanism of activation of this toxin is not well understood. PTX increases BBB permeability and enhances the infiltration of T-cells and macrophages, that are, amongst others, responsible for inflammation (Lu et al., 2008). Analysis of gene expression, from mouse BMEC, with DNA microarray (Affymetrix) reveals 34 genes that were up- or down-regulated 6 h after activation by PTX. It was shown that 13 genes were associated with inflammation and angiogenesis, including genes encoding chemokine ligands, adhesion molecules and cytokines (Lu et al., 2008). This gene expression highlights two different actions of PTX. On the one hand, PTX contributes to BBB disruption, and on the other, it participates in immune cell recruitment into the CNS. The Kinex antibody microarray analysis underlines a strong involvement of the phosphorylation cascade, such as mitogen-activated protein kinase (MAPK) and protein kinase A and C (PKA and PKC), in the PTX response; the protein variations, 2 h after the treatment, already present various phosphorylated and un-phosphorylated kinases. A few receptors and proteins involved in apoptosis were also identified among the proteins with a significant expression. Interestingly, the proteins SOD and NFkB inhibitor (IκB) were found in this study, which shows the importance of these proteins during stress (Lu et al., 2008). The vascular cell adhesion molecule (VCAM) evidenced in this study contributes to T-cell migration across the BBB (Engelhardt, 2006). Other EAE models have been designed, for example, C57Bl/ 6 mice immunized with a peptide of the myelin oligodendrocyte glycoprotein, which induces infiltration of leukocytes and demyelination, while immunization of SJL/J mice with a peptide of proteolipid protein induces massive infiltration of mononuclear cells and little or no demyelination within CNS. A DNA microarray (Affymetrix) and a 2-DE analysis of the mouse brain microvasculature study has been carried out on these two animal models (Alt et al., 2005). In this study it was reported that proteins involved in cell adhesion, such as ICAM-1, were upregulated as occurs in EAE. Moreover, the recruitment of immune cells could be explained by an

increased gene expression of the inflammatory chemokines CCL2, CCL5, CCL6, CXCL9 and CXCL10. Gene expression analysis of the cytoskeletal and extracellular matrix proteins reveals alteration in both categories. This underlies changes in tissue architecture at the BBB during EAE beyond the protein level. The increased expression of T-complex protein 1 (TCP1), which is a chaperon involved in actin and tubulin folding, suggests a cytoskeleton remodeling in the inflamed microvascular compartment during EAE (Alt et al., 2005). Furthermore, altered gene expression of extracellular matrix proteins found in the microvessel preparation of both EAE models emphasizes that molecular alteration at the BBB basal membrane, favoring inflammatory cells recruitment and BBB breakdown, cannot simply be visualized as a consequence of the degradation of certain extracellular matrix proteins due to enzymatic digestion. This may additionally be influenced by reduced gene expression in cerebral endothelial cells (Alt et al., 2005). Finally, the enhancement of the interferon response in EAE models supports the recent finding that a systemic lack of interferon-β (IFN-β) leads to increased severity of EAE. Moreover, IFN-β, which is a commonly approved and therapeutically effective treatment for multiple sclerosis, has been shown to stabilize the BBB (Kraus et al., 2004). Recently, in a 2-DE study the mechanisms of multiple sclerosis were investigated by treating a human brain capillary endothelial cell transfected with a plasmid from SV40 cultured with serum of multiple sclerosis patients or with this serum and interferon-β 1b (IFNβ-1b) (Alexander et al., 2007). Several 14-3-3 isoform proteins were involved in the BBB multiple sclerosis response. Nevertheless, according to the isoform there is variation in regulation, for example, 14-3-3 epsilon increases with serum while 14-3-3 zeta/delta decreases. Some proteins are up- or down-regulated with multiple sclerosis serum and are also modulated by IFNβ-1b treatment. For example, annexin 1 decreases with serum treatment and this phenomenon is reversed with INFβ-1b (Alexander et al., 2007). Other proteins are regulated by these treatments, as Rasrelated protein, plasminogen and ribonuclease/angiogenin inhibitor 1. This shows that angiogenesis may play a role in the pathogenesis of multiple sclerosis. Lastly, IFNβ-1b induces the expression of heat shock protein 70 kDa (HSP70), possibly linked to an immunomodulatory effect of the multiple sclerosis treatment (Alexander et al., 2007).

4.3.

Toxicological effect

A cellular toxicity analysis of compounds in cigarette smoke was carried out with rat BMEC activated by nicotine and polyaromatic hydrocarbon (PAH). The results show transendothelial transport variation with nicotine and PAH. Immunochemistry also shows influence of these molecules on cell adhesion proteins (Hutamekalin et al., 2008). This study shows variations in protein abundance between two fractions called the Triton X-100 soluble fraction (TSF) and the Triton X100 insoluble fraction (TIF) of tight junction proteins (ZO-1, Occludin, Claudin-5) and catenin. The notion of the Triton fraction is due to the Triton-X100's solubilization ability, indeed, the latter is a non-ionic, non-denaturant detergent, which partially solubilizes the membranes and the cortical cytoskeleton. To further these results, a 2-DE study was

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

carried out highlighting that heat shock proteins (HSP) are involved in this stress response. Proteins which link the cytoskeleton to the adhesion plaques of the membrane, such as metavinculin and zyxin, and proteins related to the translation were also identified (Hutamekalin et al., 2008). Toxic compounds found in cigarette smoke could induce brain capillary dysfunctions linked to the cytoskeleton organization, tight junction modifications and modulation of the cell–cell adhesion with proteins of the adhesion plaques. To conclude, these compounds appear as potentially dangerous molecules for brain homeostasis.

4.4.

Electromagnetic fields effects

The exponential increase in the use of mobile phones, in recent years, has raised some public health questions. Analysis of protein expression of non cerebral human endothelial cell lines (EA.hy926) exposed to mobile phonelike radiation has revealed the activation of the P38 MAP kinase stress signalling pathway, leading to phosphorylation of the small stress protein HSP-27, and cytoskeleton expression with vimentin expression and F-actin stress fibers. However, this study was not carried out with brain endothelial cells (Nylund and Leszczynski, 2004). Nevertheless, the endothelial response looks like an inflammatory response associated with a cell shape modification with up-regulation of cytoskeletal proteins. Moreover, cellular expression of HSP27, also called HSP-β1, was visualized by immunochemistry and was upregulated. It is possible to extrapolate this effect to brain endothelial cells exposed to radiation suggesting that radiations could have an impact on the BBB functions of controlling CNS homeostasis (Nylund and Leszczynski, 2004). Moreover, this was reported to increase the stability of F-actin stress fibers leading to changes in cell size and shape. The proper functioning of the tight junctions of endothelial cells lining brain capillary blood vessels is crucial for the proper functioning of the BBB. The cytoskeleton of endothelial cells plays an important role in the regulation of endothelial cell contacts through the tight junctions (Lai et al., 2005). Effects on the cytoskeleton and on the stress protein HSP-27 might be a part of the underlying regulatory mechanism. The observed effects on the expression of vimentin and on the distribution pattern of vimentin filaments give further support to the notion that the cytoskeleton might be one of the mobile phone radiation-responding cytoplasmic structures (Nylund and Leszczynski, 2004). To conclude, electromagnetic fields induce endothelium modifications that are related to the cytoskeleton, but it is reassuring to know that the in vitro model of BBB exposed to electromagnetic fields did not show barrier disruption (Franke et al., 2005). Endothelial cells stressed with cytokines, infected immune cells, toxic compounds or radiations reveal that the cytoskeleton takes part in the BBB stress response. Indeed, numerous studies of the BBB during stress present an involvement of actin and cytoskeleton associated proteins. To summarize, cell shape, tight junctions and cytoskeleton structure are widely linked to the endothelial permeability (Hawkins and Davis, 2005). Moreover, vascular barrier functions from different tissues are strongly correlated to the cytoskeleton organization (Bogatcheva and Verin, 2008). In some cases

93

stress response involves detoxification enzymes, with proteins such as SOD and NFκB or associated proteins that activate many stress response genes.

4.5.

Expression profiling methods in clinical research

Nowadays, the diagnosis of stroke and BBB disruption is not easy. Only invasive and expensive techniques such as contrast enhanced magnetic resonance imaging, computed tomography scanning and lumbar puncture are available to clinically assess BBB integrity. Patients treated with mannitol have a transient hyperosmotic disruption of the BBB. In particular the plasma protein S-100β is upregulated in these patients; furthermore this marker is also increased in the serum of brain tumor patients (Kanner et al., 2003; Kapural et al., 2002; Vogelbaum et al., 2005). These original studies were done using immunochemistry and opened a highway for clinical investigation of BBB disruption. Proteomic studies offer therefore new markers of BBB disruption. Analysis of the plasma proteins, by 2-DE, of patients treated with mannitol reveals an increase in transthyretin (Marchi et al., 2003a). This protein is a secreted hormone binding protein, more abundant in the choroids plexus. Chronologically, protein S-100β represents BBB disruption, while transthyretin, should be a potential marker of the blood–CSF barrier opening (Marchi et al., 2003b, 2004). Plasma level of α2-macroglobulin, a protease inhibitor, is increased during BBB disruption also (Cucullo et al., 2003). In this study both in patients treated with mannitol and in vitro culture medium of endothelial cells treated with mannitol reveal increased plasma levels and release of α2-macroglobulin respectively. An increase in the plasma level of a proteinase inhibitor reveals a correlation between BBB damage and extracellular matrix remodeling. Indeed, capillary EC are strongly associated with the basal lamina, therefore a decrease in protease activity could be involved in the endothelial stress response or a protective effect of the surrounding tissues (Cucullo et al., 2003). These results suggest the use of plasma markers to monitor the progression of BBB disruption, and moreover, could be analyzed by blood tests which are less invasive and expensive than imaging and lumbar puncture methods. When a patient presents high plasma levels of one of these markers, this could be linked to a blood–brain barrier dysfunction; more investigations could highlight the release mechanism. These studies open new fields for the determination of BBB disruption, using peripheral blood analysis as is currently being done to estimate creatine phophokinase and LDH in blood as a marker of myocardial infarcts. Generally, as previously described, clinical proteomic studies comprise the 2-DE separation of the serum or plasma proteins. In the case of the analysis of membrane proteins, such as multidrug resistant proteins (MRP), the 2-DE approach should be avoided. For example, 10% to 20% of patients with epilepsy, treated with antiepileptics, are pharmacoresistant. A study, focused on cDNA by human GeneFilters from isolated capillaries from surgical brain resection, has investigated specifically MRP expressed genes (Dombrowski et al., 2001). The results have demonstrated that epilepsy induced overexpression of multidrug resistant protein 1 (MDR1) and

94

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

Table 1 – The main proteins associated to BBB as deduced by expression profiling technologies. Protein name 14-3-3 isoforms

Function Signaling pathway

a2-macroglobulin Annexin 1

Metalloprotease inhibitor Vesicle associated protein and calcium associated protein brain-derived neutrophic factor (BDNF) Neuro protection C–C motif chemokine 2 (CCL2) Chemokine C–C motif chemokine 5 (CCL5) Chemokine C–C motif chemokine 6 (CCL6) Chemokine C–X–C motif chemokine 10 (CCL10) Chemokine C-X-C motif chemokine 9 (CCL9) Chemokine Calcium/calmodulin-dependent serine Associated with tight junctions protein kinase (CASK) Clathrin heavy chain Vesicle associated protein Clathrin light chain

Vesicle associated protein

Chloride intracellular channel (CLIC-4) Cofilin

Ion channel Cytoskeleton associated protein

Cytosol aminopeptidase Formaldehyde dehydrogenase

Endopeptidase Detoxification enzyme

Glucose transporter 1 (Glut-1)

Transporter

Heat shock protein β1 (HSP-β1)

Cytoskeleton associated protein

Heat shock protein of 70 kDa (HSP-70) Inhibitor of NFkB (IkB) Intercellular adhesion molecule ICAM

Heat shock protein Transcription factor regulation protein Adhesion molecule

Interleukin-6 Interleukin-7 Interleukin-8

Cytokine Cytokine Cytokine

Lactate dehydrogenase (LDH)

Metabolism

Macrophage-capping protein G (Cap G) Metallothionein

Cytoskeleton associated protein Detoxification enzyme

Metavinculin

Binding to adhesion plaques

Multidrug associated protein type 2 (MRP2)

Efflux pump

Multidrug resistant protein 1 (MDR1)

Efflux pump

Nuclear factor kappa-B (NFkB) Plasminogen

Transcription factor Angiogenesis

Protein S100b Rab GDP dissociatin inhibitor beta (GDIB) Ras-related protein (RAP-1)

Ion binding protein Vesicle associated protein Signaling pathway

Reck

Antiprotease

Ribonuclease/angiogenin inhibitor 1

Angiogenesis

Serpin

Antiprotease

Involvement Induced in multiple sclerosis response Marker of BBB disruption Induced in multiple sclerosis response Specific protein of the BBB Induced by EAE a Induced by EAE a Induced by EAE a Induced by EAE a Induced by EAE a Specific protein of the BBB Induced by ischemia/ reperfusion Induced by ischemia/ reperfusion Specific protein of the BBB Reduced by the TNFα Induced by HIV-infected MDM b Specific protein of the BBB Induced by ischemia/ reperfusion Reduced in in vitro culture Reduced by the TNFα Electromagnetic field response Induced by HIV-infected MDM b Response to interferon-β 1b Induced by pertussis toxin Induced by EAE a Induced by the TNFα Specific protein of the BBB Specific protein of the BBB Induced by HIV-infected MDM b Induced by HIV-infected MDM b Specific protein of the BBB Induced by Ischemia/ reperfusion Response to smoke toxic compounds c Epilepsia pharmacoresistance Reduced in in vitro culture Epilepsia pharmacoresistance Induced by the TNFα Induced in multiple sclerosis response Marker of BBB disruption Reduced by the TNFα Induced in multiple sclerosis response Induced by hydrocortison Induced in multiple sclerosis response Induced by hydrocortison

Reference Alexander et al., 2007 Cucullo et al., 2003 Alexander et al., 2007 Kallmann et al., 2002 Alt et al., 2005 Alt et al., 2005 Alt et al., 2005 Alt et al., 2005 Alt et al., 2005 Lu et al., 2007 Haqqani et al., 2007 Haqqani et al., 2007 Lu et al., 2007 Franzén et al., 2003 Ricardo-Dukelow et al., 2007 Lu et al., 2007 Haqqani et al., 2007 Calabria and Shusta, 2008 Franzén et al., 2003 Nylund and Leszczynski, 2004 Ricardo-Dukelow et al. 2007 Alexander et al. 2007 Lu et al. 2008 Alt et al. 2005 Franzén et al., 2003 Lu et al. 2007 Kallmann et al. 2002 Ricardo-Dukelow et al. 2007 Ricardo-Dukelow et al. 2007 Lu et al. 2007 Haqqani et al. 2007 Hutamekalin et al. 2008 Dombrowski et al. 2001 Calabria and Shusta, 2008 Dombrowski et al. 2001 Franzén et al., 2003 Alexander et al. 2007 Kapural et al. 2002 Franzén et al., 2003 Alexander et al. 2007 Calabria and Shusta, 2008 Alexander et al. 2007 Calabria and Shusta, 2008

95

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

Table 1 (continued) Protein name

Function

Involvement

Superoxide dismutase (mitochondrial) (SOD) Detoxification enzyme

Superoxide dismutase (soluble) (SOD)

Detoxification enzyme

Synaptic vesicle membrane protein T-complex protein 1 (TCP1) Thioredoxin

Vesicle associated protein Folding of actin and tubulin Detoxification enzyme

Transferin receptor

Membrane receptor

Transforming growth factor β1 (TGF β1)

Growth factor

Transforming growth factor β2 (TGF β2) Transthyretin uPA/uPAR system VAMP associated protein of 33 kDa (VAP-33) Vascular cell adhesion molecule (VCAM) Vimentin

Growth factor Released hormone binding protein Signaling pathway Associated with tight junctions Involved in the T-cell migration Cytoskeleton associated protein

Zyxin

Binding to adhesion plaques

a b c

Induced by the TNFα Induced by ischemia/ reperfusion Induced by pertussis toxin Induced by HIV-infected MDM b Reduced by the TNFα Induced by EAE a Induced by Ischemia/ reperfusion Reduced in in vitro culture Induced by ischemia/ reperfusion Specific protein of the BBB Marker of BBB disruption Induced by the TNFα Specific protein of the BBB Induced by pertussis toxin Electromagnetic field response Response to smoke toxic compounds c

Reference Franzén et al., 2003 Haqqani et al. 2007 Lu et al. 2008 Ricardo-Dukelow et al. 2007 Franzén et al., 2003 Alt et al. 2005 Haqqani et al. 2007 Calabria and Shusta, 2008 Haqqani et al. 2007 Kallmann et al. 2002 Marchi et al. 2003a, b Franzén et al., 2003 Lu et al. 2007 Lu et al. 2008 Nylund and Leszczynski, 2004 Hutamekalin et al. 2008

Experimental allergic encephalomyelitis. Monocyte-derived macrophage. Nicotine and polyaromatic hydrocarbon.

multidrug resistant-associated protein 2 (MRP2), presenting the modulation of gene expression, but this could not completely explain pharmacoresistance (Dombrowski et al., 2001). Currently, clinical proteomics provides diagnostic or prognostic markers, with 2-DE analysis. As presented, BBB disruption could be followed by monitoring the amount of three proteins in serum. In addition, powerful genomic tools give a new impetus to clinical pharmacology.

4.6.

Concluding remarks

Since its discovery 100 years ago, the BBB has been the center of many studies. Over the last 10 years, the BBB has been increasingly analyzed by gene and protein expression profiling, also called “omics.” We have summarized here some of the genomic and proteomic investigations carried out in order to enhance our understanding of the BBB. Primarily, this work has demonstrated that expression profiling technologies are powerful and provide significant information on the brain microvessel endothelium. Genomics and proteomics provide different complementary data on a tissue; genomics shows relative expression of membrane and released proteins, while proteomics indicates protein modifications and isoforme expression. Moreover, the characterization of endothelial cells highlights some brain specific proteins which could explain BBB features. The extent of the involvement of the cytoskeleton is shown by the brain endothelial stress. Modifications induced by stress such as auto-immunity, TNFα treatment and electromagnetic field, demonstrate a (differential) modulation of proteins such as SOD or NFκB, which may be of great interest for brain vascular therapy. Extracellular matrix remodeling is

also implicated in stroke and autoimmunity response. Table 1 enumerates the proteins presented in this review and summarizes their specific implications. To conclude, applying genomic and proteomic tools has led to the discovery of new routes for bypassing the blood–brain barrier but also raise many questions for further research. Life science methodologies and other techniques of the expression profiling such as iTRAQ, SILAC and probably new ones will help us to address all the current questions. It should be noticed that many of the studies presented in this review do not use capillary endothelial cells sensu stricto. They were derived from different microvascular branches and thus were not completely representative of the BBB, which is essentially supported by the brain capillary endothelial cells (Ge et al., 2005). Moreover, laser microdissection or enzymatic capillary extraction provokes stress during in vivo studies. The use of puromycin or hydrocortisone provides artificial BBB properties but this is not sufficient to restore an in vivo BBB phenotype. The in vitro model developed in the authors' laboratory uses brain capillary endothelial cells with BBB properties close to the in vivo phenotype (Dehouck et al., 1992). A preliminary proteomic analysis of differentiated endothelial cells from this model reveals cytoskeletal actors of BBB origin (Pottiez et al., 2009). This model has been successfully used in studying ischemia (Brillault et al., 2002), activation of signalling pathway by TNFα (Miller et al., 2005) and lipopolysaccharide effects (Boveri et al., 2006). By completing the current investigations, a thorough proteomic analysis will provide additional information concerning brain pathologies or BBB metabolism. Indeed new questions will arise from proteomics data which have to be answered by applying the complete life sciences methodology. New fields will emerge and will

96

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

provide new challenges for expression profiling and other technologies in the future.

Acknowledgments We thank Professor Bert de Boer for his critical reading of the manuscript and Alastair Balloch and Gilles Choquet for linguistic advice.

REFERENCES

Abbott, N.J., Ronnback, L., Hansson, E., 2006. Astrocyte-endothelial interactions at the blood–brain barrier. Nat. Rev. Neurosci. 7, 41–53. Aebersold, R., Mann, M., 2003. Mass spectrometry-based proteomics. Nature 422, 198–207. Alexander, J.S., Minagar, A., Harper, M., Robinson-Jackson, S., Jennings, M., Smith, S.J., 2007. Proteomic analysis of human cerebral endothelial cells activated by multiple sclerosis serum and IFNbeta-1b. J. Mol. Neurosci. 32, 169–178. Alt, C., Duvefelt, K., Franzén, B., Yang, Y., Engelhardt, B., 2005. Gene and protein expression profiling of the microvascular compartment in experimental autoimmune encephalomyelitis in C57Bl/6 and SJL mice. Brain Pathol. 15, 1–16. Baak, J.P., Janssen, E.A., Soreide, K., Heikkilae, R., 2005. Genomics and proteomics—the way forward. Ann. Oncol. 16 (Suppl. 2), ii30–ii44. Ballabh, P., Braun, A., Nedergaard, M., 2004. The blood–brain barrier: an overview: structure, regulation, and clinical implications. Neurobiol. Dis. 16, 1–13. Bartus, R.T., Elliott, P.J., Dean, R.L., Hayward, N.J., Nagle, T.L., Huff, M.R., Snodgrass, P.A., Blunt, D.G., 1996. Controlled modulation of BBB permeability using the bradykinin agonist, RMP-7. Exp. Neurol. 142, 14–28. Baty, J.D., Robinson, P.R., 1977. Single and multiple ion recording techniques for the analysis of diphenylhydantoin and its major metabolite in plasma. Biomed. Mass Spectrom. 4, 36–41. Begley, D.J., 2004. Delivery of therapeutic agents to the central nervous system: the problems and the possibilities. Pharmacol. Ther. 104, 29–45. Bogatcheva, N.V., Verin, A.D., 2008. The role of cytoskeleton in the regulation of vascular endothelial barrier function. Microvasc. Res. 76, 202–207. Bohman, S., Matsumoto, T., Suh, K., Dimberg, A., Jakobsson, L., Yuspa, S., Claesson-Welsh, L., 2005. Proteomic analysis of vascular endothelial growth factor-induced endothelial cell differentiation reveals a role for chloride intracellular channel 4 (CLIC4) in tubular morphogenesis. J. Biol. Chem. 280, 42397–42404. Boveri, M., Kinsner, A., Berezowski, V., Lenfant, A.M., Draing, C., Cecchelli, R., Dehouck, M.P., Hartung, T., Prieto, P., Bal-Price, A., 2006. Highly purified lipoteichoic acid from gram-positive bacteria induces in vitro blood–brain barrier disruption through glia activation: role of pro-inflammatory cytokines and nitric oxide. Neuroscience 137, 1193–1209. Brillault, J., Berezowski, V., Cecchelli, R., Dehouck, M.P., 2002. Intercommunications between brain capillary endothelial cells and glial cells increase the transcellular permeability of the blood–brain barrier during ischaemia. J. Neurochem. 83, 807–817. Brown, R.C., Davis, T.P., 2002. Calcium modulation of adherens and tight junction function: a potential mechanism for blood–brain barrier disruption after stroke. Stroke 33, 1706–1711.

Calabria, A.R., Shusta, E.V., 2006. Blood–brain barrier genomics and proteomics: elucidating phenotype, identifying disease targets and enabling brain drug delivery. Drug Discov. Today 11, 792–799. Calabria, A.R., Shusta, E.V., 2008. A genomic comparison of in vivo and in vitro brain microvascular endothelial cells. J. Cereb. Blood Flow Metab. 28, 135–148. Calabria, A.R., Weidenfeller, C., Jones, A.R., de Vries, H.E., Shusta, E.V., 2006. Puromycin-purified rat brain microvascular endothelial cell cultures exhibit improved barrier properties in response to glucocorticoid induction. J. Neurochem. 97, 922–933. Campostrini, N., Areces, L.B., Rappsilber, J., Pietrogrande, M.C., Dondi, F., Pastorino, F., Ponzoni, M., Righetti, P.G., 2005. Spot overlapping in two-dimensional maps: a serious problem ignored for much too long. Proteomics 5, 2385–2395. Cecchelli, R., Berezowski, V., Lundquist, S., Culot, M., Renftel, M., Dehouck, M.P., Fenart, L., 2007. Modelling of the blood–brain barrier in drug discovery and development. Nat. Rev. Drug Discov. 6, 650–661. Chappa, A.K., Cooper, J.D., Audus, K.L., Lunte, S.M., 2007. Investigation of the metabolism of substance P at the blood–brain barrier using LC-MS/MS. J. Pharm. Biomed. Anal. 43, 1409–1415. Chaudhuri, A., Yang, B., Gendelman, H.E., Persidsky, Y., Kanmogne, G.D., 2008. STAT1 signaling modulates HIV-1-induced inflammatory responses and leukocyte transmigration across the blood–brain barrier. Blood 111, 2062–2072. Cucullo, L., Marchi, N., Marroni, M., Fazio, V., Namura, S., Janigro, D., 2003. Blood–brain barrier damage induces release of alpha2-macroglobulin. Mol. Cell Proteomics 2, 234–241. Dallasta, L.M., Pisarov, L.A., Esplen, J.E., Werley, J.V., Moses, A.V., Nelson, J.A., Achim, C.L., 1999. Blood–brain barrier tight junction disruption in human immunodeficiency virus-1 encephalitis. Am. J. Pathol. 155, 1915–1927. de Vries, H.E., Blom-Roosemalen, M.C., van Oosten, M., de Boer, A. G., van Berkel, T.J., Breimer, D.D., Kuiper, J., 1996. The influence of cytokines on the integrity of the blood–brain barrier in vitro. J. Neuroimmunol. 64, 37–43. Dehouck, M.P., Meresse, S., Delorme, P., Fruchart, J.C., Cecchelli, R., 1990. An easier, reproducible, and mass-production method to study the blood–brain barrier in vitro. J. Neurochem. 54, 1798–1801. Dehouck, M.P., Jolliet-Riant, P., Bree, F., Fruchart, J.C., Cecchelli, R., Tillement, J.P., 1992. Drug transfer across the blood–brain barrier: correlation between in vitro and in vivo models. J. Neurochem. 58, 1790–1797. Deli, M.A., Abraham, C.S., Kataoka, Y., Niwa, M., 2005. Permeability studies on in vitro blood–brain barrier models: physiology, pathology, and pharmacology. Cell. Mol. Neurobiol. 25, 59–127. Diatchenko, L., Lau, Y.F., Campbell, A.P., Chenchik, A., Moqadam, F., Huang, B., Lukyanov, S., Lukyanov, K., Gurskaya, N., Sverdlov, E.D., Siebert, P.D., 1996. Suppression subtractive hybridization: a method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc. Natl. Acad. Sci. U. S. A. 93, 6025–6030. Dohgu, S., Yamauchi, A., Takata, F., Naito, M., Tsuruo, T., Higuchi, S., Sawada, Y., Kataoka, Y., 2004. Transforming growth factor-beta1 upregulates the tight junction and P-glycoprotein of brain microvascular endothelial cells. Cell. Mol. Neurobiol. 24, 491–497. Dombrowski, S.M., Desai, S.Y., Marroni, M., Cucullo, L., Goodrich, K., Bingaman, W., Mayberg, M.R., Bengez, L., Janigro, D., 2001. Overexpression of multiple drug resistance genes in endothelial cells from patients with refractory epilepsy. Epilepsia 42, 1501–1506. Donkin, J.J., Turner, R.J., Hassan, I., Vink, R., 2007. Substance P in traumatic brain injury. Prog. Brain Res. 161, 97–109.

B RA I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

Engelhardt, B., 2006. Molecular mechanisms involved in T cell migration across the blood–brain barrier. J. Neural. Transm. 113, 477–485. Fournier, M.L., Gilmore, J.M., Martin-Brown, S.A., Washburn, M.P., 2007. Multidimensional separations-based shotgun proteomics. Chem. Rev. 107, 3654–3686. Franke, H., Streckert, J., Bitz, A., Goeke, J., Hansen, V., Ringelstein, E.B., Nattkamper, H., Galla, H.J., Stogbauer, F., 2005. Effects of Universal Mobile Telecommunications System (UMTS) electromagnetic fields on the blood–brain barrier in vitro. Radiat. Res. 164, 258–269. Franzén, B., Duvefelt, K., Jonsson, C., Engelhardt, B., Ottervald, J., Wickman, M., Yang, Y., Schuppe-Koistinen, I., 2003. Gene and protein expression profiling of human cerebral endothelial cells activated with tumor necrosis factor-alpha. Brain Res. Mol. Brain Res. 115, 130–146. Ge, S., Song, L., Pachter, J.S., 2005. Where is the blood–brain barrier really? J. Neurosci. Res. 79, 421–427. Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., Gelb, M.H., Aebersold, R., 1999. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 17, 994–999. Haqqani, A.S., Nesic, M., Preston, E., Baumann, E., Kelly, J., Stanimirovic, D., 2005. Characterization of vascular protein expression patterns in cerebral ischemia/reperfusion using laser capture microdissection and ICAT-nanoLC-MS/MS. FASEB J. 19, 1809–1821. Haqqani, A.S., Kelly, J., Baumann, E., Haseloff, R.F., Blasig, I.E., Stanimirovic, D.B., 2007. Protein markers of ischemic insult in brain endothelial cells identified using 2D gel electrophoresis and ICAT-based quantitative proteomics. J. Proteome Res. 6, 226–239. Haseloff, R.F., Krause, E., Bigl, M., Mikoteit, K., Stanimirovic, D., Blasig, I.E., 2006. Differential protein expression in brain capillary endothelial cells induced by hypoxia and posthypoxic reoxygenation. Proteomics 6, 1803–1809. Hawkins, B.T., Davis, T.P., 2005. The blood–brain barrier/ neurovascular unit in health and disease. Pharmacol. Rev. 57, 173–185. Hoheisel, D., Nitz, T., Franke, H., Wegener, J., Hakvoort, A., Tilling, T., Galla, H.J., 1998. Hydrocortisone reinforces the blood–brain barrier properties in a serum free cell culture system. Biochem. Biophys. Res. Commun. 244, 312–316. Hutamekalin, P., Farkas, A.E., Orbok, A., Wilhelm, I., Nagyoszi, P., Veszelka, S., Deli, M.A., Buzas, K., Hunyadi-Gulyas, E., Medzihradszky, K.F., Meksuriyen, D., Krizbai, I.A., 2008. Effect of nicotine and polyaromatic hydrocarbons on cerebral endothelial cells. Cell Biol. Int. 32, 198–209. Kallmann, B.A., Wagner, S., Hummel, V., Buttmann, M., Bayas, A., Tonn, J.C., Rieckmann, P., 2002. Characteristic gene expression profile of primary human cerebral endothelial cells. FASEB J. 16, 589–591. Kamiie, J., Ohtsuki, S., Iwase, R., Ohmine, K., Katsukura, Y., Yanai, K., Sekine, Y., Uchida, Y., Ito, S., Terasaki, T., 2008. 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. 25, 1469–1483. Kanmogne, G.D., Kennedy, R.C., Grammas, P., 2002. HIV-1 gp120 proteins and gp160 peptides are toxic to brain endothelial cells and neurons: possible pathway for HIV entry into the brain and HIV-associated dementia. J. Neuropathol. Exp. Neurol. 61, 992–1000. Kanner, A.A., Marchi, N., Fazio, V., Mayberg, M.R., Koltz, M.T., Siomin, V., Stevens, G.H., Masaryk, T., Aumayr, B., Vogelbaum, M.A., Barnett, G.H., Janigro, D., 2003. Serum S100beta: a noninvasive marker of blood–brain barrier function and brain lesions. Cancer 97, 2806–2813. Kapural, M., Krizanac-Bengez, L., Barnett, G., Perl, J., Masaryk, T., Apollo, D., Rasmussen, P., Mayberg, M.R., Janigro, D., 2002.

97

Serum S-100beta as a possible marker of blood–brain barrier disruption. Brain Res. 940, 102–104. Kraus, J., Ling, A.K., Hamm, S., Voigt, K., Oschmann, P., Engelhardt, B., 2004. Interferon-beta stabilizes barrier characteristics of brain endothelial cells in vitro. Ann. Neurol. 56, 192–205. Lai, C.H., Kuo, K.H., Leo, J.M., 2005. Critical role of actin in modulating BBB permeability. Brain Res. Brain Res. Rev. 50, 7–13. Lapierre, L.A., Tuma, P.L., Navarre, J., Goldenring, J.R., Anderson, J.M., 1999. VAP-33 localizes to both an intracellular vesicle population and with occludin at the tight junction. J. Cell. Sci. 112 (Pt 21), 3723–3732. Lasbennes, F., Lacombe, P., Seylaz, J., 1988. Effect of monoamine oxidase inhibition on the regional cerebral blood flow response to circulating noradrenaline. Brain Res. 454, 205–211. Lawrenson, J.G., Reid, A.R., Finn, T.M., Orte, C., Allt, G., 1999. Cerebral and pial microvessels: differential expression of gamma-glutamyl transpeptidase and alkaline phosphatase. Anat. Embryol. (Berl) 199, 29–34. Liu, L.B., Xue, Y.X., Liu, Y.H., Wang, Y.B., 2008. Bradykinin increases blood–tumor barrier permeability by down-regulating the expression levels of ZO-1, occludin, and claudin-5 and rearranging actin\ cytoskeleton. J. Neurosci. Res. 86, 1153–1168. Lu, C., Pelech, S., Zhang, H., Bond, J., Spach, K., Noubade, R., Blankenhorn, E.P., Teuscher, C., 2008. Pertussis toxin induces angiogenesis in brain microvascular endothelial cells. J. Neurosci. Res. 86, 2624–2640. Lu, L., Yang, P.Y., Rui, Y., Kang, H., Zhang, J., Zhang, J.P., Feng, W.H., 2007. Comparative proteome analysis of rat brain and coronary microvascular endothelial cells. Physiol. Res. 56, 159–168. Marchi, N., Fazio, V., Cucullo, L., Kight, K., Masaryk, T., Barnett, G., Vogelbaum, M., Kinter, M., Rasmussen, P., Mayberg, M.R., Janigro, D., 2003a. Serum transthyretin monomer as a possible marker of blood-to-CSF barrier disruption. J. Neurosci. 23, 1949–1955. Marchi, N., Rasmussen, P., Kapural, M., Fazio, V., Kight, K., Mayberg, M.R., Kanner, A., Ayumar, B., Albensi, B., Cavaglia, M., Janigro, D., 2003b. Peripheral markers of brain damage and blood–brain barrier dysfunction. Restor. Neurol. Neurosci. 21, 109–121. Marchi, N., Cavaglia, M., Fazio, V., Bhudia, S., Hallene, K., Janigro, D., 2004. Peripheral markers of blood–brain barrier damage. Clin. Chim. Acta 342, 1–12. Martinez-Estrada, O.M., Villa, A., Breviario, F., Orsenigo, F., Dejana, E., Bazzoni, G., 2001. Association of junctional adhesion molecule with calcium/calmodulin-dependent serine protein kinase (CASK/LIN-2) in human epithelial caco-2 cells. J. Biol. Chem. 276, 9291–9296. Miller, F., Fenart, L., Landry, V., Coisne, C., Cecchelli, R., Dehouck, M.P., Buee-Scherrer, V., 2005. The MAP kinase pathway mediates transcytosis induced by TNF-alpha in an in vitro blood–brain barrier model. Eur. J. Neurosci. 22, 835–844. Miller, I., Crawford, J., Gianazza, E., 2006. Protein stains for proteomic applications: which, when, why? Proteomics 6, 5385–5408. Moron, J.A., Devi, L.A., 2007. Use of proteomics for the identification of novel drug targets in brain diseases. J. Neurochem. 102, 306–315. Munir, S., Singh, S., Kaur, K., Kapur, V., 2004. Suppression subtractive hybridization coupled with microarray analysis to examine differential expression of genes in virus infected cells. Biol. Proced. Online 6, 94–104. Nakagawa, S., Deli, M.A., Kawaguchi, H., Shimizudani, T., Shimono, T., Kittel, A., Tanaka, K., Niwa, M., 2009. A new blood–brain barrier model using primary rat brain endothelial cells, pericytes and astrocytes. Neurochem. Int. 54, 253–263. Nylund, R., Leszczynski, D., 2004. Proteomics analysis of human endothelial cell line EA.hy926 after exposure to GSM 900 radiation. Proteomics 4, 1359–1365.

98

BR A I N R ES E A RC H R EV IE W S 6 2 (2 0 0 9) 8 3–9 8

O'Farrell, P.H., 1975. High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250, 4007–4021. Ohtsuki, S., Terasaki, T., 2007. Contribution of carrier-mediated transport systems to the blood–brain barrier as a supporting and protecting interface for the brain; importance for CNS drug discovery and development. Pharm. Res. 24, 1745–1758. Ong, S.E., Mann, M., 2007. Stable isotope labeling by amino acids in cell culture for quantitative proteomics. Methods Mol. Biol. 359, 37–52. Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., Steen, H., Pandey, A., Mann, M., 2002. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell Proteomics 1, 376–386. Padieu, P., Maume, B.F., 1976. Enzymatic activity assays in the hepatic cell by mass fragmentography associated with gas-liquid chromatography. Ann. Biol. Clin. (Paris). 34, 63–77. Pan, W., Banks, W.A., Fasold, M.B., Bluth, J., Kastin, A.J., 1998. Transport of brain-derived neurotrophic factor across the blood–brain barrier. Neuropharmacology 37, 1553–1561. Pellieux, C., Desgeorges, A., Pigeon, C.H., Chambaz, C., Yin, H., Hayoz, D., Silacci, P., 2003. Cap G, a gelsolin family protein modulating protective effects of unidirectional shear stress. J. Biol. Chem. 278, 29136–29144. Penkowa, M., Moos, T., Carrasco, J., Hadberg, H., Molinero, A., Bluethmann, H., Hidalgo, J., 1999. Strongly compromised inflammatory response to brain injury in interleukin-6deficient mice. Glia 25, 343–357. 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., 2005. Puromycin-based purification of rat brain capillary endothelial cell cultures. Effect on the expression of blood–brain barrier-specific properties. J. Neurochem. 93, 279–289. Perriere, N., Yousif, S., Cazaubon, S., Chaverot, N., Bourasset, F., Cisternino, S., Decleves, X., Hori, S., Terasaki, T., Deli, M., Scherrmann, J.M., Temsamani, J., Roux, F., Couraud, P.O., 2007. A functional in vitro model of rat blood–brain barrier for molecular analysis of efflux transporters. Brain Res. 1150, 1–13. Phanstiel, D., Unwin, R., McAlister, G.C., Coon, J.J., 2009. Peptide quantification using 8-plex isobaric tags and electron transfer dissociation tandem mass spectrometry. Anal. Chem. 81, 1693–1698. Pottiez, G., Sevin, E., Cecchelli, R., Karamanos, Y., Flahaut, C., 2009. Actin, gelsolin and filamin-A are dynamic actors in the cytoskeleton remodelling contributing to the blood brain barrier phenotype. Proteomics 9, 1207–1219. Reese, T.S., Karnovsky, M.J., 1967. Fine structural localization of a blood–brain barrier to exogenous peroxidase. J. Cell Biol. 34, 207–217. Revest, P.A., Abbott, N.J., Gillespie, J.I., 1991. Receptor-mediated changes in intracellular [Ca2+] in cultured rat brain capillary endothelial cells. Brain Res. 549, 159–161. Ricardo-Dukelow, M., Kadiu, I., Rozek, W., Schlautman, J., Persidsky, Y., Ciborowski, P., Kanmogne, G.D., Gendelman, H.E., 2007. HIV-1 infected monocyte-derived macrophages affect the human brain microvascular endothelial cell proteome: new insights into blood–brain barrier dysfunction for HIV-1associated dementia. J. Neuroimmunol. 185, 37–46. Ross, P.L., Huang, Y.N., Marchese, J.N., Williamson, B., Parker, K., Hattan, S., Khainovski, N., Pillai, S., Dey, S., Daniels, S., Purkayastha, S., Juhasz, P., Martin, S., Bartlet-Jones, M., He, F., Jacobson, A., Pappin, D.J., 2004. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol. Cell Proteomics 3, 1154–1169. Roux, F., Durieu-Trautmann, O., Chaverot, N., Claire, M., Mailly, P., Bourre, J.M., Strosberg, A.D., Couraud, P.O., 1994. Regulation of gamma-glutamyl transpeptidase and alkaline phosphatase activities in immortalized rat brain microvessel endothelial cells. J. Cell. Physiol. 159, 101–113.

Sanger, F., Air, G.M., Barrell, B.G., Brown, N.L., Coulson, A.R., Fiddes, C.A., Hutchison, C.A., Slocombe, P.M., Smith, M., 1977. Nucleotide sequence of bacteriophage phi X174 DNA. Nature 265, 687–695. Sato, Y., 2004. Role of aminopeptidase in angiogenesis. Biol. Pharm. Bull. 27, 772–776. Shen, Q., Goderie, S.K., Jin, L., Karanth, N., Sun, Y., Abramova, N., Vincent, P., Pumiglia, K., Temple, S., 2004. Endothelial cells stimulate self-renewal and expand neurogenesis of neural stem cells. Science 304, 1338–1340. Shusta, E.V., 2005. Blood–brain barrier genomics, proteomics, and new transporter discovery. NeuroRx 2, 151–161. Stanness, K.A., Neumaier, J.F., Sexton, T.J., Grant, G.A., Emmi, A., Maris, D.O., Janigro, D., 1999. A new model of the blood–brain barrier: co-culture of neuronal, endothelial and glial cells under dynamic conditions. NeuroReport 10, 3725–3731. Suginta, W., Karoulias, N., Aitken, A., Ashley, R.H., 2001. Chloride intracellular channel protein CLIC4 (p64H1) binds directly to brain dynamin I in a complex containing actin, tubulin and 14-3-3 isoforms. Biochem. J. 359, 55–64. Toborek, M., Lee, Y.W., Pu, H., Malecki, A., Flora, G., Garrido, R., Hennig, B., Bauer, H.C., Nath, A., 2003. HIV-Tat protein induces oxidative and inflammatory pathways in brain endothelium. J. Neurochem. 84, 169–179. Trickler, W.J., Mayhan, W.G., Miller, D.W., 2005. Brain microvessel endothelial cell responses to tumor necrosis factor-alpha involve a nuclear factor kappa B (NF-kappaB) signal transduction pathway. Brain Res. 1048, 24–31. Unlu, M., Morgan, M.E., Minden, J.S., 1997. Difference gel electrophoresis: a single gel method for detecting changes in protein extracts. Electrophoresis 18, 2071–2077. Utepbergenov, D.I., Mertsch, K., Sporbert, A., Tenz, K., Paul, M., Haseloff, R.F., Blasig, I.E., 1998. Nitric oxide protects blood–brain barrier in vitro from hypoxia/reoxygenationmediated injury. FEBS Lett. 424, 197–201. Velculescu, V.E., Zhang, L., Vogelstein, B., Kinzler, K.W., 1995. Serial analysis of gene expression. Science 270, 484–487. Vogelbaum, M.A., Masaryk, T., Mazzone, P., Mekhail, T., Fazio, McCartney, S., Marchi, N., Kanner, A., Janigro, D., 2005. S100beta as a predictor of brain metastases: brain versus cerebrovascular damage. Cancer 104, 817–824. Washburn, M.P., Wolters, D., Yates III, J.R., 2001. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 19, 242–247. Watt, S.A., Patschkowski, T., Kalinowski, J., Niehaus, K., 2003. Qualitative and quantitative proteomics by two-dimensional gel electrophoresis, peptide mass fingerprint and a chemicallycoded affinity tag (CCAT). J. Biotechnol. 106, 287–300. Wiener, M.C., Sachs, J.R., Deyanova, E.G., Yates, N.A., 2004. Differential mass spectrometry: a label-free LC-MS method for finding significant differences in complex peptide and protein mixtures. Anal. Chem. 76, 6085–6096. Wilkins, M.R., Pasquali, C., Appel, R.D., Ou, K., Golaz, O., Sanchez, J.C., Yan, J.X., Gooley, A.A., Hughes, G., Humphery-Smith, I., Williams, K.L., Hochstrasser, D.F., 1996. From proteins to proteomes: large scale protein identification by two-dimensional electrophoresis and amino acid analysis. Biotechnology (NY) 14, 61–65. Witt, K.A., Mark, K.S., Huber, J., Davis, T.P., 2005. Hypoxia-inducible factor and nuclear factor kappa-B activation in blood–brain barrier endothelium under hypoxic/reoxygenation stress. J. Neurochem. 92, 203–214. Zhang, W., Mojsilovic-Petrovic, J., Andrade, M.F., Zhang, H., Ball, M., Stanimirovic, D.B., 2003. The expression and functional characterization of ABCG2 in brain endothelial cells and vessels. FASEB J. 17, 2085–2087. Zhu, D.Y., Li, R., Liu, G.Q., Hua, W.Y., 2000. Tumor necrosis factor alpha enhances the cytotoxicity induced by nitric oxide in cultured cerebral endothelial cells. Life Sci. 66, 1325–1335.