A large-scale electrophoresis- and chromatography-based determination of gene expression profiles in bovine brain capillary endothelial cells after the re-induction of blood-brain barrier properties
© Pottiez et al; licensee BioMed Central Ltd. 2010
Received: 22 July 2010
Accepted: 15 November 2010
Published: 15 November 2010
Brain capillary endothelial cells (BCECs) form the physiological basis of the blood-brain barrier (BBB). The barrier function is (at least in part) due to well-known proteins such as transporters, tight junctions and metabolic barrier proteins (e.g. monoamine oxidase, gamma glutamyltranspeptidase and P-glycoprotein). Our previous 2-dimensional gel proteome analysis had identified a large number of proteins and revealed the major role of dynamic cytoskeletal remodelling in the differentiation of bovine BCECs. The aim of the present study was to elaborate a reference proteome of Triton X-100-soluble species from bovine BCECs cultured in the well-established in vitro BBB model developed in our laboratory.
A total of 215 protein spots (corresponding to 130 distinct proteins) were identified by 2-dimensional gel electrophoresis, whereas over 350 proteins were identified by a shotgun approach. We classified around 430 distinct proteins expressed by bovine BCECs. Our large-scale gene expression analysis enabled the correction of mistakes referenced into protein databases (e.g. bovine vinculin) and constitutes valuable evidence for predictions based on genome annotation.
Elaboration of a reference proteome constitutes the first step in creating a gene expression database dedicated to capillary endothelial cells displaying BBB characteristics. It improves of our knowledge of the BBB and the key proteins in cell structures, cytoskeleton organization, metabolism, detoxification and drug resistance. Moreover, our results emphasize the need for both appropriate experimental design and correct interpretation of proteome datasets.
The endothelia of different organs are remarkably heterogeneous but do present many common functional and morphological features. Given the endothelium's strategic position between the blood and the tissues, this cell layer (i) closely controls the transport of plasma molecules (via bidirectional receptor-mediated and receptor-independent transcytosis and endocytosis), (ii) regulates vascular tone, (iii) synthesises and secretes a wide variety of factors and (iv) is involved in the regulation of inflammation, haemostasis, thrombosis and immunity. It is now also generally accepted that the specific ultrastructure of capillaries in the brain, retina, kidney and liver governs the specialized physiological properties of these respective endothelia . In the brain, the blood-brain barrier (BBB) separates the brain microvasculature from the peripheral microvasculature. The BBB constitutes a physical and metabolic barrier which tightly regulates blood-brain exchanges of ions, small molecules and proteins and is involved in the recruitment of immune cells prior to transfer to the brain during inflammation [2–4].
In brain capillaries, the BBB is formed by endothelial cells, which are surrounded by a tubular sheath of astrocytic end-feet. Pericytes are inserted into the basal membrane (between the endothelium and the astrocytic end-feet) . This spatial cell layout and the resulting astrocyte-endothelium communication induce the establishment and maintenance of the BBB [5–7]. Dysregulation of these processes has been linked to the pathogenesis of several human diseases .
In the brain, only blood capillaries are endowed with a complete BBB phenotype . Under physiological conditions, the barrier function is performed by a number of unique endothelial features, including (i) the lack of fenestration, (ii) a decrease in the number of pinocytic vesicles, (iii) the reinforcement of complex tight junctions and (iv) the upregulated expression of metabolic enzymes and plasma membrane transporters and receptors . The physiological consequences of endothelial cell differentiation include an increase in the transendothelial electrical resistance (due to a decrease in the para- and transcellular endothelial permeability of ions and low-molecular-weight hydrophilic compounds) and are associated with marked polarization of the cerebral endothelium [10, 11]. In brain endothelial cells, the plasma membrane acts as the controlling interface for intracellular molecular signalling, the reinforcement of tight junctions and molecular and cell transport between the brain and the blood. The plasma membrane of brain capillary endothelial cells (BCECs) has been extensively studied and its membrane protein expression pattern has been well defined . The intracellular location of certain proteins was shown to be essential for the establishment and maintenance of the BBB-related features of BCECs. These intracellular locations are frequently used as quality control criteria for in vitro BBB models. Furthermore, it is known that the protein distribution changes under pathological conditions [13, 14]. Paradoxically, no dedicated studies in this field have been reported. Moreover, the BBB's metabolic proteome is not well known and the cytosolic, nuclear and mitochondrial protein expression profiles have yet to be extensively characterized. Therefore, the use of Triton X-100 (known to poorly solubilise sparingly soluble proteins ) appeared as the best way to select the BCECs' cytosolic subproteome in the present study.
Proteomics deals with the direct, large-scale determination of gene and cellular function at the protein level. Recent successes have emphasized the role of mass spectrometry-based proteomics as an essential tool in molecular and cellular biology. Two-dimensional polyacrylamide gel electrophoresis (2-DE) has become the core technology for building proteomic databases, for several reasons. Firstly, 2-DE can display thousands of proteins and allows the relative quantification of any given polypeptide. Secondly, the high-mass shifts caused by one or more posttranslational modifications can be characterized. Thirdly, digitized 2-DE images are convenient, informative visual media for creating a web-based database. Although 2-DE can (in theory) separate all proteins, the technique has a number of drawbacks in practice: (i) it is difficult to separate proteins with low (< 10 kDa) or high (> 200 kDa) molecular weights, (ii) the truly operational p I range is 3-11 and (iii) protein spots are frequently superposed. Complementary approaches are often needed to overcome incompatibilities between the 2-DE separation technique and the physical-chemical properties of certain proteins of interest. Liquid chromatography (LC) is an alternative (or a complementary counterpart) to 2-DE.
On the basis of the in vitro BBB co-culture model developed in our laboratory , we have initiated a comprehensive proteomic approach based on the extractability of proteins with Triton X-100. It uses a 2-DE/1D-LC combination to provide a detailed expression profile for the Triton X-100-soluble proteins expressed by bovine BCECs (BBCECs) with the BBB phenotype. Here, we report on the use of a high-throughput proteomics platform in the 2-DE identification of the most abundant Triton X-100-soluble proteins, together with additional 1D-LC identification. In addition to investigation of the correlation between the identified proteins and BBB functions, the present study was also designed to prime a comprehensive, interactive, web-based proteomics database for BCECs (2-DE images, comparative results, mass spectra, identification scores, protein identities, protein functions, etc.), in accordance with published guidelines [17, 18].
Verification of the re-induction of BBB properties
When isolated in vitro, primary BBCECs dedifferentiate and lose their BBB properties. The latter can be restored by 12 days of co-culture with glial cells . We confirmed the re-induction of BBB properties by (i) measuring an optimal value for the paracellular permeability coefficient (Pe) and (ii) immunostaining the main tight junction proteins (ZO-1, occludin and claudin-5). As reported in the literature [19, 20], the markers display a pericellular distribution in differentiated BBCECs, which contrasts with the uniform pattern for dedifferentiated cells (data not shown). The BBCECs were harvested and lysed according to the procedure described by Pottiez et al. . The Triton X-100-solubilised protein fraction (representing 85% of the cell's total protein) was then subjected to either 2-DE or off-line 1D-LC tandem mass spectrometry (MS/MS) analysis with a C18 column.
Two-dimensional polyacrylamide gel electrophoresis
One-dimensional liquid chromatography
Molecular functions of the identified proteins and their involvement in biological processes
The protein interference problem in a gel-free approach
Evidence for the existence of a 1134-amino acid bovine vinculin
The battle against the relentless progression of neurodegenerative disease, brain tumours and brain damage is a public health priority. Although effective drugs have been reported on the basis of in vivo research results, drug targeting and delivery to the brain are still constitute a challenge  one century after the discovery of the BBB. In fact, the BBB is often the rate-limiting factor in drug transport into the brain. Models of the BBB are of major importance in efforts to expand our understanding of how the brain maintains its integrity and controls the inward and outward fluxes of endogenous substances and drug compounds . The development of genomic and proteomic technologies has provided several means of extending our knowledge of the BBB and investigating additional routes for bypassing this barrier .
Several BCEC-dedicated whole-cell or plasma membrane proteomic approaches have been reported. Most have been comparative studies designed to detect and identify candidate proteins which are either differentially expressed in two distinct conditions [25, 26], involved in a given physiological process [21, 27] or expressed in response to a particular stimulus [14, 28–31]. Very few studies have focused on sub-proteomes such as the protein content of the plasma membrane  or that of the caveolae . Although this type of targeted proteomic study is essential for unravelling the molecular mechanisms governing the physiology and physiopathology of the BBB's endothelial cells, only small sets of protein identities tend to be reported. Hence, large-scale determinations of expression profiles for a given protein class or cell compartment-related protein subset in BBB-differentiated BCECs (relative to non-brain capillary endothelial cells) have yet to be performed in any species. The large-scale determination of BCEC protein contents is a essential analytical step in gaining a better understanding of the physiopathology of the BBB .
In the field of BBB research, proteomic analyses of BCECs displaying BBB features and functions often focus on proteins related to the tight junction, the adherens junction, transport, drug resistance and the regulation of the latter functions . Most of the aforementioned proteins are transmembrane proteins anchored to the actin cytoskeleton by cytosolic proteins. The importance of the intracellular location has been well demonstrated . In general, cytosolic proteins are ignored - despite the fact that they are known to play a role in many BBB functions. Our present study provided a partial list of non-characterized cytosolic proteins with possible relevance in BBB function. Firstly, many of the identified proteins (27%) are related to metabolic processes, such as the primary metabolism of proteins, carbohydrates, nucleobases/nucleosides/nucleotides, lipids, amino acids and their derivatives. We consider that knowledge of the BCECs' enzymatic equipment constitutes a breakthrough. By way of an example, levels of nitric oxide (known to modulate BBB permeability) are related to the metabolism of arginine . Oxidative stress is also an important modulator of BBB permeability. In the present study, we showed that 31 oxidoreductase proteins are expressed by BCECs. Secondly, binding proteins represent around 30% of the identified set. Of these, 32 are known to interact with Ca2+. It has been reported that both increases and depletion of intracellular Ca2+ may result in disruption of intercellular junctions . We also detected 104 nucleotide-binding proteins including 34 ribonucleoproteins and 38 RNA-binding proteins (ribosomal proteins, heterogeneous nuclear ribonucleoproteins, splicing factors, eukaryotic translation initiation factors, tRNA synthetase, etc.). Furthermore, 73 and 30 of the identified proteins are respectively ATP- and GTP-binding proteins. Sixty-five proteins are described as having molecular transport activities. Thirdly, structural components account for around 20% of the proteins identified: half of the latter (n = 44) are cytoskeleton components or cytoskeleton-associated proteins. In addition to the presence of proteins like ERM complexes, filamins, heat shock proteins, plectin, plastin 3, alpha-actinins, cofilins, actin-capping proteins, spectrins and tropomyosins, we report the expression of (i) junction plakoglobin (gamma catenin), a cytoplasmic protein in soluble and membrane-associated forms and (ii) Talin 1, a crucial protein for reinforcement of integrin-cytoskeleton bond and which binds vinculin with high affinity . Fourthly, 45 and 37 proteins reported in this study are from mitochondria and endoplasmic reticulum and which are involved not only in metabolism but also in cell function. One example is the tumour necrosis factor receptor-associated protein 1 (TRAP-1) which protects cells from oxidative stress and apoptosis .
In fine, 293 of the identified proteins are phosphoproteins. Hence, Triton X-100-dependent protein extraction may constitute a way of enriching the phosphoproteome subset.
The creation of a dedicated proteomic database is the best way of gathering and storing proteome information but does require perfect sample traceability and strict compliance with protein identification guidelines  in order to avoid human error and, with a view to high throughputs, guarantee the identity of proteins by reducing the false positive rate as much as possible. Likewise, the repeatability of protein identifications (at least triple mass measurement of the same spot) must be carefully checked and not postponed within the database to assure the latter's clarity and avoid obstruction (especially in the case of publicly available databases). Due to the inherent drawbacks of each analytical technique, the use of several different approaches is the best way of ensuring the broadest possible determination of proteins from a given proteome. The recent development of commercially available integrated solutions for high-throughput protein identification (combining bio-informatics tools and automated sample handling with one or several mass spectrometers) constitutes a true source of progress; it saves time, reduces the number of errors, increases the analytical throughput, centralizes the proteomic data and enables more efficient comparison of datasets. Nevertheless, most in-house 2-DE, 1D-LC or 2D-LC databases are not publicly available (for various reasons). As stated in a recent review article [39, 40], this situation can easily be resolved by public, community-based solutions for the long-term storage, management, sharing and comparison of proteomic data, such as MIAPEGelDB , the World-2DPAGE repository , PRIDE [43, 44], GPMDB , Peptideatlas  and PrestOMIC . However, we believe that a number of limitations remain.
The expansion of the LC-MS proteomic data storage is significantly greater than that of the 2-DE proteomic data. We believe that this can be explained by several factors. Firstly, shotgun strategies are remarkably automated and powerful and clearly increase the overall data throughput compared with 2-DE gel. Secondly, the web-submission process in the proteomic repositories of LC-MS proteomic data is now automatized, convenient and secure (mainly thanks to the use of converters, such as the PRIDE converter). Thirdly, web submission of 2D-PAGE proteomic data is still a manual, fastidious, time-consuming step and is prone to human error. It therefore requires more preparation steps (for the 2-DE gel pictures, spot location data, MS data, MS protein identification data, etc.); this situation clearly weakens the commitment to communal sharing and prompts researchers to overuse shotgun approaches.
The diversity of MS file formats is now handled well by almost all proteomic data repositories, although some proprietary file formats developed by MS manufacturers are not yet supported. Therefore, proteomic data can be submitted over the web via protein identification result files, such as Mascot .dat files and SEQUEST result files. However, whereas the MS/MS files used to create the PFF are fully supported, the files corresponding to protein identifications based on PMF or combined PMF/PFF protocols are not supported and the automatic web-submission of the latter two files types is still not possible. Overall, these are probably the main limiting factors on the expansion of public submission of 2-DE proteomic datasets. We have contacted Dr Juan Antonio Vizcaíno in the PRIDE group with a view to achieving the parsing of PMF and combined PMF/PFF Mascot data files from 2-DE experiments.
Protein databases are now so large that syntactic errors are now acknowledged to be one of the reasons for absent or false protein identifications. For example, bovine vinculin is reported in UniprotKB under two accession numbers Q2PQT6 and Q0VCE6 with 240 and 80 amino acids respectively. However, unmatured vinculin displays an average of 1134 amino acids for isoform-1 and 1066 for isoform-2 from distinct species (human, mouse, rat and pig). The NCBInr database reports a bovine vinculin sequence (NW_001501727) derived from a gene prediction process. Hence, on the basis of NCBInr database and our LC-MS/MS data, we identified a 1134- amino acid bovine vinculin; this shows that vinculin is expressed in BBCECs in a form similar to those reported in mammals. This type of de novo sequencing information (i) provides direct evidence of the existence of listed proteins in a given cell type in a given species, (ii) enables the correction of certain mistakes reported in protein databases (e.g. bovine vinculin) and (iii) can confirm predictions generated by genome annotation methods.
Vimentin identification is a classic example of the technological bias inherent in each proteomic strategy. Whereas 2-DE is able to distinguish between several vimentin isoforms (resulting from post-translational modifications), 1D-LC identifies only one protein. Vimentin is one of the most prominent phosphoproteins in various mesenchymal cells in which phosphorylation is enhanced during cell division , when vimentin filaments are significantly reorganized. Phosphorylation also determines the assembly dynamics of vimentin intermediate filaments . Vimentin is also subject to acetylation. In all, more than 50 isoforms are listed in the protein databases. Consequently, vimentin cannot be simply considered as the single protein reported by 1D-LC. The 1D-LC identification of proteins with very similar primary amino-acid sequences is far from perfect because it is based on the PFF corresponding to protein-specific peptides. This is known as the protein interference problem  and, unfortunately, is still ignored in too many cases. For soluble proteins > 20 kDa, the PMF obtained from 2-DE is more likely to unambiguously distinguish between paralogous proteins. Ultimately, combined PFF/PMF datasets are more discriminant and have a greater certainty of identification because they are derived from in-gel separated proteins.
Our combination of 2-DE and 1D-LC approaches enabled the first ever identification of about 430 Triton X-100-soluble proteins (see Additional File 3) from BBCECs displaying BBB characteristics. The information on these protein identifications is now stored in in-house databases and will be soon shared through the PRIDE database for convenient comparison with proteomic datasets from non-brain vascular endothelial cells. Due to the extraction conditions, fewer than 5% of the proteins corresponded to membrane-associated proteins. More than 75% of the identified proteins display binding, catalytic or structural functions. Most identified proteins were involved in metabolic and cellular processes but transport and cell-cell communication process accounted for almost 25% of the identified species. The present study also emphasized the advantage of inter-species sequence homology comparisons for protein identification in non-completely sequenced genomes and highlighted the drifts and drawbacks generated by the rapid, gel-free proteomic methods (which nevertheless provide most of today's data). In the absence of a physiological, human in vitro BBB model, the proteomics expression profile of BBCECs displaying BBB properties is the first step towards the creation of an anti-BBB antibody library which will greatly facilitate large-scale, array-based screening of protein expression.
Heat-inactivated calf serum and horse serum were bought from Hyclone Laboratories (Logan, UT, USA). Glutamine and gentamycin were purchased from Biochrom AG (Berlin, Germany). Serum, basic fibroblast growth factor and Dulbecco's Modified Eagle's Medium (DMEM) were from GIBCO (Invitrogen Corporation, Carlsbad, CA, USA). Six-well plates and Transwell inserts were from Corning Inc. (New York, NY, USA). The duracryl/bis-acrylamide solution was from Genomic Solution (Proteomic Solutions, Saint-Marcel, France). Immobilized pH gradient-ready strips were from GE Healthcare (Amersham Bioscience, Orsay, France). Trypsin digestion kits for the DP Proteineer™robot and alpha-cyano-4-hydroxycinnamic acid were from Bruker Daltonik GmbH (Bremen, Germany). All other reagents were of analytical or electrophoresis grades.
Cell culture and the blood-brain barrier model
Primary cultures of mixed glial cells were made from newborn rat cerebral cortex, as described by Booher and Sensenbrenner . Briefly, glial cells were cultured in DMEM supplemented with 10% (v/v) heat-inactivated foetal calf serum, 2 mM glutamine and 50 μg/ml gentamycin. Three weeks after seeding, the glial cell cultures were stabilized and used for co-culture. Bovine brain capillary endothelial cells were isolated and characterized as described by Méresse et al. . The BBCECs were co-cultured until confluence (12 days) with glial cells on an extracellular matrix (rat tail collagen) in DMEM supplemented with 10% (v/v) heat-inactivated calf serum, 10% (v/v) horse serum, 2 mM glutamine, 50 μg/ml gentamycin and 1 ng/ml basic fibroblast growth factor.
Cell harvesting and protein extraction
The endothelial cells (8 × 105 cells) were harvested by collagenase treatment (Clostridium histolyticum, Sigma, Lyon, France) . Briefly, the cells were treated at 37°C for 40 min with 1.5 ml of a collagenase solution (0.1% w/v). The collected cell materials were rinsed 3 times in phosphate buffered saline (PBS), centrifuged 10 min at 500 g. The cell pellets were lysed in 200 μl of lysis buffer [Tris/HCl 10 mM, EDTA 1 mM, Triton X-100 1% (v/v), 2-mercaptoethanol 0.1% (v/v) and protease inhibitors (Roche Biomoleculars, Meylan, France)] and centrifuged (13,500 g, 4°C, 45 min). The protein content of the Triton X-100 soluble and insoluble fractions was assessed . The supernatants were concentrated, desalted and delipidated by overnight acid precipitation at -20°C.
Two-dimensional gel electrophoresis
After resolubilization in an isoelectrofocusing buffer (7 M urea, 2 M thiourea, 4% (v/v) CHAPS and 2% (v/v) ampholytes), 300 μg of cytosolic proteins were subjected to 2-DE on 24-cm length pH 4-7 IPG strips at 100,000 V and in the 10-100 kDa molecular mass range. The IPG strips were passively and actively rehydrated for 7 h and 9 h at 50 V. The pre-focusing and focusing procedures were carried out at 50 mA/strip in 4 steps: 200 V for 1 h, a gradient up to 1000 V for 1 h, a gradient up to 10000 V for 6 h and, lastly, 10000 V for 4.5 h. The IPG strips were wiped up and successively equilibrated for 15 min with gentle shaking in 6 M urea, 20% (v/v) glycerol, 2% (w/v) SDS, 93 mM TRIS-HCl pH 8.8 buffers supplemented with 20 mM DTT and 100 mM iodoacetamide and a trace of BPB, respectively. The equilibrated strips were sealed on the top of the second-dimension duracrylamide/bis-acrylamide gel (12% T, 2.6% C) with 0.5% (w/v) low-melting point agarose (Biorad, Marnes-la-Coquette, France) in SDS running buffer. Migration as a function of molecular weight was performed in the Ettan DALTsix electrophoresis unit (Amersham Bioscience) at 16 mA/gel for 30 min and then at 32 mA/gel until the tracking dye reached the anodic end. The proteins were stained with silver nitrate  for image acquisition (with a freshly calibrated Umax scanner (Amersham Biosciences, Orsay, France) at 300 dpi using Labscan 3.0 software) and with colloidal Coomassie Brilliant Blue for protein identification by MALDI-TOF/TOF mass spectrometry.
Protein identification experiments after 2D-PAGE
Protein identifications from 2-DE gels were performed using a Proteineer™workflow from Bruker Daltonik GmbH. Colloidal Coomassie-blue-stained spots were excised from gels with a spot picker (the Proteineer™spII) equipped with a 2 mm needle and placed into 96-well microtitre plates. In-gel digestion and sample preparation for MALDI analysis were performed according to the manufacturer's instructions using a digester/spotter robot (the Proteineer™dp) and tryptic digest kits (the DP 384 standard kit from Bruker Daltonik). Briefly, after destaining of the gel plugs with 10 mM ammonium bicarbonate and 50% acetonitrile in 10 mM ammonium bicarbonate, protein spots were digested essentially according to Shevshenko et al., 1996 . Peptide were extracted with acetonitrile: 0.1% TFA-acidified water (1:1) and then mixed with an α-cyano-4-hydroxycinnamic acid matrix (0.3 mg/ml in acetone:ethanol, 3:6 v/v) on the MALDI target plate (AnchorChip™, Bruker Daltonics). The molecular mass measurements were performed in automatic mode using FlexControl™ 2.2 software on an Ultraflex™ II TOF/TOF instrument and in reflectron mode for MALDI-TOF PMF or LIFT mode for MALDI-TOF/TOF PFF. External calibration over the 1000-3500 mass range was performed using the [M+H]+ monoisotopic ions of bradykinin 1-7, angiotensin I, angiotensin II, substance P, bombesin and adrenocorticotropic hormone (clips 1-17 and clips 18-39) from a peptide calibration standard kit (Bruker Daltonik). Briefly, a 25 kV accelerating voltage, a 26.3 kV reflector voltage and a 160 ns pulsed ion extraction were used to obtain the MS spectrum. Each spectrum was produced by accumulating data from 500 laser shots. Two precursor ions per sample at most were chosen for LIFT-TOF/TOF MS/MS analysis. Precursor ions were accelerated to 8 kV and selected in a timed ion gate. Metastable ions generated by laser-induced decomposition (LID) were further accelerated by 19 kV in the LIFT cell and their masses were measured in reflectron mode. Peak lists were generated from MS and MS/MS spectra using Flexanalysis™ 2.4 software (Bruker Daltonik). Database searches with Mascot 2.2 (Matrix Science Ltd, London, UK) using combined PMF and PFF datasets were performed in the UniProt 57.7 and NCBInr databases via ProteinScape 1.3 (Bruker Daltonik). A mass tolerance of 75 ppm and 1 missing cleavage site for PMF and an MS/MS tolerance of 0.5 Da and 1 missing cleavage site for MS/MS searching were allowed. Variable cysteine carbamidomethylation and methionine oxidation were also considered. The relevance of protein identities was judged according to the probability-based Mowse score, calculated with p < 0.05.
Triton-soluble proteins from differentiated BBCECs are extremely heterogeneous. To attenuate this phenomenon, we fractionated the samples into 5 fractions of increasing concentration in acetonitrile (0%, 25%, 50%, 75% and 100%). Following the tryptic digestion, nanoseparations were performed on an U3000 nanoHPLC system (Dionex-LC-Packings, Sunnyvale, CA, USA). After a conventional pre-concentration step (C18 cartridge, 300 μm, 1 mm), the peptide samples were separated on a Pepmap C18 column (75 μm, 15 cm) using an acetonitrile gradient from 5% to 12% over 20 minutes, 12% to 50% over 140 minutes and 50% to 70% over 15 minutes and, lastly, 15 minutes in 70% of acetonitrile. The flow was set to 300 nl/min and 380 fractions were automatically collected every 30 seconds on an AnchorChip™ MALDI target using a Proteineer™FC fraction collector (Bruker Daltonik). 2 μl of CHCA matrix (0.3 mg/ml in acetone:ethanol:0.1% TFA-acidified water, 3:6:1 v/v/v) were added during the collection process. The MS and MS/MS mass measurements were performed off-line using the Ultraflex™ II TOF/TOF mass spectrometer, as described above. The apparatus parameters were set to the values given above. Peptide fragmentation was driven by Warp LC software (Bruker Daltonik) according to the following parameters: signal-to-noise ratio > 15, more than 3 MS/MS by fraction if the MS signal was available, 0.15 Da of MS tolerance for peak merge and the elimination of peaks which appears in more than 35% of the fractions. The protein identification was performed as described above.
Sorting protein lists and bioinformatics resources
The protein lists were compared using nwCompare software . All identified proteins were converted into gene names with the DAVID bioinformatics resources  before to be classification by the Protein Analysis Through Evolutionary Relationships (PANTHER) classification system. PANTHER is a unique resource that classifies genes by their functions, using published scientific experimental evidence and evolutionary relationships. Proteins are classified by expert biologists into families and subfamilies of shared function, which are then categorized by molecular function and biological process ontology terms [57, 58].
This research was funded by the Ministère de la Recherche et de l'Enseignement Supérieur and a grant from Oseo-Anvar. The mass spectrometry facilities used for this study were funded by the European Union (FEDER), the Fonds d'Industrialisation du Bassin Minier (FIBM), the Ministère de l'Education Nationale, de l'Enseignement Supérieur et de la Recherche and the Université d'Artois. We are also grateful to Johan Hachani for his excellent technical assistance in the course of this work.
- Pries AR, Kuebler WM: Normal endothelium. Handb Exp Pharmacol 2006, 1–40. full_textGoogle Scholar
- Abbott NJ: Dynamics of CNS barriers: evolution, differentiation, and modulation. Cell Mol Neurobiol 2005, 25: 5–23. 10.1007/s10571-004-1374-yPubMedView ArticleGoogle Scholar
- Ballabh P, Braun A, Nedergaard M: The blood-brain barrier: an overview: structure, regulation, and clinical implications. Neurobiol Dis 2004, 16: 1–13. 10.1016/j.nbd.2003.12.016PubMedView ArticleGoogle Scholar
- Pottiez G, Flahaut C, Cecchelli R, Karamanos Y: Understanding the blood-brain barrier using gene and protein expression profiling technologies. Brain Res Rev 2009, 62: 83–98. 10.1016/j.brainresrev.2009.09.004PubMedView ArticleGoogle Scholar
- Abbott NJ, Ronnback L, Hansson E: Astrocyte-endothelial interactions at the blood-brain barrier. Nat Rev Neurosci 2006, 7: 41–53. 10.1038/nrn1824PubMedView ArticleGoogle Scholar
- Haseloff RF, Blasig IE, Bauer HC, Bauer H: In search of the astrocytic factor(s) modulating blood-brain barrier functions in brain capillary endothelial cells in vitro. Cell Mol Neurobiol 2005, 25: 25–39. 10.1007/s10571-004-1375-xPubMedView ArticleGoogle Scholar
- Abbott N: Astrocyte-endothelial interactions and blood-brain barrier permeability. J Anat 2002, 200: 527. 10.1046/j.1469-7580.2002.00064.xPubMed CentralView ArticleGoogle Scholar
- Weiss N, Miller F, Cazaubon S, Couraud PO: The blood-brain barrier in brain homeostasis and neurological diseases. Biochim Biophys Acta 2009, 1788: 842–857. 10.1016/j.bbamem.2008.10.022PubMedView ArticleGoogle Scholar
- Cecchelli R, Berezowski V, Lundquist S, Culot M, Renftel M, Dehouck MP, Fenart L: Modelling of the blood-brain barrier in drug discovery and development. Nat Rev Drug Discov 2007, 6: 650–661. 10.1038/nrd2368PubMedView ArticleGoogle Scholar
- Cooray HC, Blackmore CG, Maskell L, Barrand MA: Localisation of breast cancer resistance protein in microvessel endothelium of human brain. Neuroreport 2002, 13: 2059–2063. 10.1097/00001756-200211150-00014PubMedView ArticleGoogle Scholar
- Pardridge WM: Molecular biology of the blood-brain barrier. Methods Mol Med 2003, 89: 385–399.PubMedGoogle Scholar
- Kamiie J, Ohtsuki S, Iwase R, Ohmine K, Katsukura Y, Yanai K, Sekine Y, Uchida Y, Ito S, Terasaki T: Quantitative atlas of membrane transporter proteins: development and application of a highly sensitive simultaneous LC/MS/MS method combined with novel in-silico peptide selection criteria. Pharm Res 2008, 25: 1469–1483. 10.1007/s11095-008-9532-4PubMedView ArticleGoogle Scholar
- Chen SH, Stins MF, Huang SH, Chen YH, Kwon-Chung KJ, Chang Y, Kim KS, Suzuki K, Jong AY: Cryptococcus neoformans induces alterations in the cytoskeleton of human brain microvascular endothelial cells. J Med Microbiol 2003, 52: 961–970. 10.1099/jmm.0.05230-0PubMedView ArticleGoogle Scholar
- Hutamekalin P, Farkas AE, Orbok A, Wilhelm I, Nagyoszi P, Veszelka S, Deli MA, Buzas K, Hunyadi-Gulyas E, Medzihradszky KF, et al.: Effect of nicotine and polyaromtic hydrocarbons on cerebral endothelial cells. Cell Biol Int 2008, 32: 198–209. 10.1016/j.cellbi.2007.08.026PubMedView ArticleGoogle Scholar
- Santoni V, Molloy M, Rabilloud T: Membrane proteins and proteomics: un amour impossible? Electrophoresis 2000, 21: 1054–1070. 10.1002/(SICI)1522-2683(20000401)21:6<1054::AID-ELPS1054>3.0.CO;2-8PubMedView ArticleGoogle Scholar
- Dehouck MP, Meresse S, Delorme P, Fruchart JC, Cecchelli R: An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J Neurochem 1990, 54: 1798–1801. 10.1111/j.1471-4159.1990.tb01236.xPubMedView ArticleGoogle Scholar
- Wilkins MR, Appel RD, Van Eyk JE, Chung MC, Gorg A, Hecker M, Huber LA, Langen H, Link AJ, Paik YK, et al.: Guidelines for the next 10 years of proteomics. Proteomics 2006, 6: 4–8. 10.1002/pmic.200500856PubMedView ArticleGoogle Scholar
- Bradshaw RA: Revised draft guidelines for proteomic data publication. Mol Cell Proteomics 2005, 4: 1223–1225.PubMedGoogle Scholar
- Culot M, Lundquist S, Vanuxeem D, Nion S, Landry C, Delplace Y, Dehouck MP, Berezowski V, Fenart L, Cecchelli R: An in vitro blood-brain barrier model for high throughput (HTS) toxicological screening. Toxicol In Vitro 2008, 22: 799–811. 10.1016/j.tiv.2007.12.016PubMedView ArticleGoogle Scholar
- Gosselet F, Candela P, Sevin E, Berezowski V, Cecchelli R, Fenart L: Transcriptional profiles of receptors and transporters involved in brain cholesterol homeostasis at the blood-brain barrier: use of an in vitro model. Brain Res 2009, 1249: 34–42. 10.1016/j.brainres.2008.10.036PubMedView ArticleGoogle Scholar
- Pottiez G, Sevin E, Cecchelli R, Karamanos Y, Flahaut C: Actin, gelsolin and filamin-A are dynamic actors in the cytoskeleton remodelling contributing to the blood brain barrier phenotype. Proteomics 2009, 9: 1207–1219. 10.1002/pmic.200800503PubMedView ArticleGoogle Scholar
- Martens L, Hermjakob H, Jones P, Adamski M, Taylor C, States D, Gevaert K, Vandekerckhove J, Apweiler R: PRIDE: the proteomics identifications database. Proteomics 2005, 5: 3537–3545. 10.1002/pmic.200401303PubMedView ArticleGoogle Scholar
- Barsnes H, Vizcaino JA, Eidhammer I, Martens L: PRIDE Converter: making proteomics data-sharing easy. Nat Biotechnol 2009, 27: 598–599. 10.1038/nbt0709-598PubMedView ArticleGoogle Scholar
- Neuwelt E, Abbott NJ, Abrey L, Banks WA, Blakley B, Davis T, Engelhardt B, Grammas P, Nedergaard M, Nutt J, et al.: Strategies to advance translational research into brain barriers. Lancet Neurol 2008, 7: 84–96. 10.1016/S1474-4422(07)70326-5PubMedView ArticleGoogle Scholar
- Alt C, Duvefelt K, Franzen B, Yang Y, Engelhardt B: Gene and protein expression profiling of the microvascular compartment in experimental autoimmune encephalomyelitis in C57Bl/6 and SJL mice. Brain Pathol 2005, 15: 1–16. 10.1111/j.1750-3639.2005.tb00094.xPubMedView ArticleGoogle Scholar
- Lu L, Yang PY, Rui Y, Kang H, Zhang J, Zhang JP, Feng WH: Comparative proteome analysis of rat brain and coronary microvascular endothelial cells. Physiol Res 2007, 56: 159–168.PubMedGoogle Scholar
- Haseloff RF, Krause E, Bigl M, Mikoteit K, Stanimirovic D, Blasig IE: Differential protein expression in brain capillary endothelial cells induced by hypoxia and posthypoxic reoxygenation. Proteomics 2006, 6: 1803–1809. 10.1002/pmic.200500182PubMedView ArticleGoogle Scholar
- Franzen B, Duvefelt K, Jonsson C, Engelhardt B, Ottervald J, Wickman M, Yang Y, Schuppe-Koistinen I: Gene and protein expression profiling of human cerebral endothelial cells activated with tumor necrosis factor-alpha. Mol Brain Res 2003, 115: 130–146. 10.1016/S0169-328X(03)00185-2PubMedView ArticleGoogle Scholar
- Haqqani AS, Kelly J, Baumann E, Haseloff RF, Blasig IE, Stanimirovic DB: Protein markers of ischemic insult in brain endothelial cells identified using 2 D gel electrophoresis and ICAT-based quantitative proteomics. J Proteome Res 2007, 6: 226–239. 10.1021/pr0603811PubMedView ArticleGoogle Scholar
- Minagar A, Alexander JS, Kelley RE, Harper M, Jennings MH: Proteomic analysis of human cerebral endothelial cells activated by glutamate/MK-801: significance in ischemic stroke injury. J Mol Neurosci 2009, 38: 182–192. 10.1007/s12031-008-9149-4PubMedView ArticleGoogle Scholar
- Alexander JS, Minagar A, Harper M, Robinson-Jackson S, Jennings M, Smith SJ: Proteomic analysis of human cerebral endothelial cells activated by multiple sclerosis serum and IFNbeta-1b. J Mol Neurosci 2007, 32: 169–178. 10.1007/s12031-007-0018-3PubMedView ArticleGoogle Scholar
- Sprenger RR, Horrevoets AJ: Proteomic study of caveolae and rafts isolated from human endothelial cells. Methods Mol Biol 2007, 357: 199–213.PubMedGoogle Scholar
- Hamacher M, Meyer HE: HUPO Brain Proteome Project: aims and needs in proteomics. Expert Rev Proteomics 2005, 2: 1–3. 10.1586/147894220.127.116.11PubMedView ArticleGoogle Scholar
- Cardoso FL, Brites D, Brito MA: Looking at the blood-brain barrier: molecular anatomy and possible investigation approaches. Brain Res Rev 64: 328–363. 10.1016/j.brainresrev.2010.05.003Google Scholar
- Daff S: NO synthase: structures and mechanisms. Nitric Oxide 23: 1–11. 10.1016/j.niox.2010.03.001Google Scholar
- Persidsky Y, Ramirez SH, Haorah J, Kanmogne GD: Blood-brain barrier: structural components and function under physiologic and pathologic conditions. J Neuroimmune Pharmacol 2006, 1: 223–236. 10.1007/s11481-006-9025-3PubMedView ArticleGoogle Scholar
- Giannone G, Jiang G, Sutton DH, Critchley DR, Sheetz MP: Talin1 is critical for force-dependent reinforcement of initial integrin-cytoskeleton bonds but not tyrosine kinase activation. J Cell Biol 2003, 163: 409–419. 10.1083/jcb.200302001PubMed CentralPubMedView ArticleGoogle Scholar
- Landriscina M, Laudiero G, Maddalena F, Amoroso MR, Piscazzi A, Cozzolino F, Monti M, Garbi C, Fersini A, Pucci P, Esposito F: Mitochondrial chaperone Trap1 and the calcium binding protein Sorcin interact and protect cells against apoptosis induced by antiblastic agents. Cancer Res 2010, 70: 6577–6586. 10.1158/0008-5472.CAN-10-1256PubMedView ArticleGoogle Scholar
- Mead JA, Bianco L, Bessant C: Recent developments in public proteomic MS repositories and pipelines. Proteomics 2009, 9: 861–881. 10.1002/pmic.200800553PubMedView ArticleGoogle Scholar
- Mead JA, Shadforth IP, Bessant C: Public proteomic MS repositories and pipelines: available tools and biological applications. Proteomics 2007, 7: 2769–2786. 10.1002/pmic.200700152PubMedView ArticleGoogle Scholar
- Robin X, Hoogland C, Appel RD, Lisacek F: MIAPEGelDB, a web-based submission tool and public repository for MIAPE gel electrophoresis documents. J Proteomics 2008, 71: 249–251. 10.1016/j.jprot.2008.06.005PubMedView ArticleGoogle Scholar
- Hoogland C, Mostaguir K, Appel RD, Lisacek F: The World-2DPAGE Constellation to promote and publish gel-based proteomics data through the ExPASy server. J Proteomics 2008, 71: 245–248. 10.1016/j.jprot.2008.02.005PubMedView ArticleGoogle Scholar
- Jones P, Cote RG, Martens L, Quinn AF, Taylor CF, Derache W, Hermjakob H, Apweiler R: PRIDE: a public repository of protein and peptide identifications for the proteomics community. Nucleic Acids Res 2006, 34: D659–663. 10.1093/nar/gkj138PubMed CentralPubMedView ArticleGoogle Scholar
- Vizcaino JA, Cote R, Reisinger F, J MF, Mueller M, Rameseder J, Hermjakob H, Martens L: A guide to the Proteomics Identifications Database proteomics data repository. Proteomics 2009, 9: 4276–4283. 10.1002/pmic.200900402PubMed CentralPubMedView ArticleGoogle Scholar
- Craig R, Cortens JP, Beavis RC: Open source system for analyzing, validating, and storing protein identification data. J Proteome Res 2004, 3: 1234–1242. 10.1021/pr049882hPubMedView ArticleGoogle Scholar
- Desiere F, Deutsch EW, King NL, Nesvizhskii AI, Mallick P, Eng J, Chen S, Eddes J, Loevenich SN, Aebersold R: The PeptideAtlas project. Nucleic Acids Res 2006, 34: D655–658. 10.1093/nar/gkj040PubMed CentralPubMedView ArticleGoogle Scholar
- Howes CG, Foster LJ: PrestOMIC, an open source application for dissemination of proteomic datasets by individual laboratories. Proteome Sci 2007, 5: 8. 10.1186/1477-5956-5-8PubMed CentralPubMedView ArticleGoogle Scholar
- Dephoure N, Zhou C, Villen J, Beausoleil SA, Bakalarski CE, Elledge SJ, Gygi SP: A quantitative atlas of mitotic phosphorylation. Proc Natl Acad Sci USA 2008, 105: 10762–10767. 10.1073/pnas.0805139105PubMed CentralPubMedView ArticleGoogle Scholar
- Eriksson JE, He T, Trejo-Skalli AV, Harmala-Brasken AS, Hellman J, Chou YH, Goldman RD: Specific in vivo phosphorylation sites determine the assembly dynamics of vimentin intermediate filaments. J Cell Sci 2004, 117: 919–932. 10.1242/jcs.00906PubMedView ArticleGoogle Scholar
- Nesvizhskii AI, Aebersold R: Interpretation of shotgun proteomic data: the protein inference problem. Mol Cell Proteomics 2005, 4: 1419–1440. 10.1074/mcp.R500012-MCP200PubMedView ArticleGoogle Scholar
- Booher J, Sensenbrenner M: Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology 1972, 2: 97–105.PubMedGoogle Scholar
- Meresse S, Dehouck MP, Delorme P, Bensaid M, Tauber JP, Delbart C, Fruchart JC, Cecchelli R: Bovine brain endothelial cells express tight junctions and monoamine oxidase activity in long-term culture. J Neurochem 1989, 53: 1363–1371. 10.1111/j.1471-4159.1989.tb08526.xPubMedView ArticleGoogle Scholar
- Peterson GL: A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal Biochem 1977, 83: 346–356. 10.1016/0003-2697(77)90043-4PubMedView ArticleGoogle Scholar
- Shevchenko A, Wilm M, Vorm O, Mann M: Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem 1996, 68: 850–858. 10.1021/ac950914hPubMedView ArticleGoogle Scholar
- Pont F, Fournie JJ: Sorting protein lists with nwCompare: a simple and fast algorithm for n-way comparison of proteomic data files. Proteomics 2010, 10: 1091–1094.PubMedGoogle Scholar
- Huang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009, 4: 44–57. 10.1038/nprot.2008.211PubMedView ArticleGoogle Scholar
- Thomas PD, Campbell MJ, Kejariwal A, Mi H, Karlak B, Daverman R, Diemer K, Muruganujan A, Narechania A: PANTHER: a library of protein families and subfamilies indexed by function. Genome Res 2003, 13: 2129–2141. 10.1101/gr.772403PubMed CentralPubMedView ArticleGoogle Scholar
- Mi H, Guo N, Kejariwal A, Thomas PD: PANTHER version 6: protein sequence and function evolution data with expanded representation of biological pathways. Nucleic Acids Res 2007, 35: D247–252. 10.1093/nar/gkl869PubMed CentralPubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.