EP1588172A2

EP1588172A2 - Method for identifying bhs-specific proteins and fragments thereof

Info

Publication number: EP1588172A2
Application number: EP03757804A
Authority: EP
Inventors: Sabine Wolf; Martina JÄGER; Thorsten Bangsow; Carmen Bangsow; Dominik Jordan; Bernhard Pelzer; Thomas Oppolzer
Original assignee: Frankgen Biotechnologie AG
Current assignee: Frankgen Biotechnologie AG
Priority date: 2002-09-11
Filing date: 2003-09-08
Publication date: 2005-10-26
Also published as: AU2003273841A8; AU2003273841A1; JP2006517181A; WO2004029631A3; WO2004029631A2; US20060172300A1; DE10242016A1

Abstract

The invention relates to a method for identifying the presence of a BHS-specific protein/fragment in cerebral capillary endothelial cells, characterized in that a) cerebral endothelial cells freshly isolated from the brain are pre-cleaned by enzymatic decomposition in the usual manner; b) the decomposition obtained in step a) is treated with a lysis buffer which substantially destroys the erythrocytes and apoptic cells and enables at least 70 % of the cerebral capillary endothelial cells to be obtained in a vital form; c) optionally, the product obtained in step b) is further cleaned; d) a substractive cDNA bank is produced from the cerebral capillary endothelial cells and a substraction tissue; e) a cDNA substraction is performed by means of differential hybridization(s); f) clones from the substractive cDNA bank are verified with regard to the respective expression thereof by differential hybrization; g) a full cDNA sequence is established for the BHS-specific clones from the substractive cDNA bank and h) the expression pattern of the analysed clone is compared in relation to fresh and cultivated cerebral capillary endothelial cells, whereupon the presence of BHS-specific proteins/fragments are identified in addition to the proteins and fragments thereof identified according to said method.

Description

Methods for identifying BBB-specific proteins and

Fragments of it

The invention relates to a method for identifying the presence of a BBB-specific protein or fragment thereof in brain capillary endothelial cells and to the proteins or fragments thereof obtained with this method (BBB = blood-brain barrier). Furthermore, the invention also relates to the genes or transcripts obtained with this method.

The endothelial cells of cerebral capillaries form a selective permeability barrier between the blood and the brain of an organism, the so-called blood-brain barrier (BBB). Individual endothelial cells are arranged around the lumen within the capillaries and form a cylindrical, tubular cavity. Close connections between the individual endothelial cells and other cell types associated with the endothelial cells prevent the uncontrolled passive passage of a large number of substances through this cell layer.

To maintain its function, the brain relies heavily on a constant internal environment, which is ensured by the blood-brain barrier. This also regulates the exchange of substances between the blood and the brain. Specific transport systems mediate this exchange. Training This barrier in the endothelial cells of the brain capillaries (brain microvessel endothelial cells, BMEC) is due to the expression of specific proteins in this highly differentiated cell type compared to other endothelial cells. Some proteins specific to the blood-brain barrier are already known, for example the glucose transporter GLUT-1, which is specific for the BMEC and ensures the energy supply to the brain.

Due to the selective permeability properties of the blood-brain barrier, it is difficult to treat various diseases of the central nervous system, since numerous drugs hardly penetrate the blood-brain barrier and therefore only arrive at their place of action in the brain in low concentrations. For the development of drugs that act in the brain, it would therefore be very important to know how the blood-brain barrier and the proteins involved work. In particular, it would be important to gain knowledge of those proteins which, compared to other cell types in the brain capillary endothelial cells, are produced to a particularly high or particularly low degree or from certain splice variants or which have specific post-translational modifications.

The study of brain capillary endothelial cells has various problems. On the one hand, there is in particular the study of protein expression in humans

The brain does not have enough brain material available, which also includes ethical reasons. Furthermore, the individual individuals from whom the brain mass is derived are usually very different in terms of their genetic information. There are differences, for example, in terms of age, gender, weight, race, etc. Furthermore, the examination material must be within the first hours after entry of death, because after this period there is already a significant change in the protein composition in the cells due to enzymatic degradation and remodeling processes. Previous methods for examining protein expression in brain capillary endothelial cells also have the problem that the test material cannot be obtained in sufficient purity for direct examinations. When brain capillary endothelial cells are isolated by the known method, a mixture with other cell types is usually obtained, so that studies of the protein expression pattern on these samples do not allow adequate assignment exclusively to the brain capillary endothelial cells.

The object of the present invention is therefore to provide a method with which BBB-specific proteins or fragments thereof can be uniquely identified. The method is said to be particularly suitable for identifying BBB-specific proteins or genes in brain capillary endothelial cells. Furthermore, the method should be easy and gentle to carry out. Furthermore, the method according to the invention is intended to be selective for proteins or fragments thereof which are amplified or formed exclusively in brain capillary endothelial cells and not in a comparative tissue or related cell type. Furthermore, the proteins or fragments thereof identified with the method according to the invention are to be suitable as diagnostic markers for diseases which are associated with a dysfunction of the blood-brain barrier. Furthermore, the proteins identified by the method according to the invention are said to be suitable for the manufacture of medicaments for the treatment of diseases which are associated with a dysfunction of the blood-brain barrier.

This object is achieved according to the invention by a method for identifying the presence of a BBB-specific Proteins or fragments thereof in brain capillary endothelial cells, characterized in that a) brain capillary endothelial cells freshly isolated from the brain are pre-cleaned in the usual way by enzymatic digestion, b) the digestion obtained in stage a) is treated with a lysis buffer which contains the erythrocytes and apoptotic cells are essentially destroyed and at least 70% of the brain capillary endothelial cells are kept in vital form, c) if necessary, the product obtained in step b) is further purified, d) a subtractive cDNA library is produced from the brain capillary endothelial cells and a subtraction tissue, e) performs a cDNA subtraction by means of one or more differential hybridization steps, f) clones from the subtractive cDNA bank are verified with respect to their respective expression by differential hybridization, g) the cDNA sequence for the BHS-specific clones from the subtractive cDNA bank added and h) the expression pattern of the examined Comparing clones between fresh and cultured brain capillary endothelial cells and thus identifying the presence of BBB-specific proteins or fragments thereof.

The invention further relates to a method for identifying the presence of a BBB-specific protein or fragment thereof in brain capillary endothelial cells, characterized in that a) brain capillary endothelial cells freshly isolated from the brain are pre-purified in the usual manner by enzymatic digestion, b) the in Step a) of the disruption obtained is treated with a lysis buffer which essentially destroys existing erythrocytes and apoptotic cells and at least 70% of the brain capillary endothelial cells is kept in vital form, c) if appropriate, that obtained in step b)

Product further purified, d) solubilizing the product obtained in step c) in a suitable buffer, e) an iso carries out electrical focusing, f) separates the samples from the isoelectric focusing in the second dimension according to molecular weight, g) identifies and isolates differential spots, h) carries out a mass spectrometric analysis with the isolate from g), and i) evaluates this by means of targeted database analysis.

With the method according to the invention, BBB-specific proteins or fragments thereof can be uniquely and reliably identified and the invention also relates to the proteins isolated with this method and the transcripts or genes coding for these proteins. In particular, the invention also relates to the proteins with the sequences SEQ ID NO: 5, 14, 19, 23, 27, 33, 53 isolated by this method.

The invention further relates to the use of the proteins or

Fragments thereof for the production of agents or medicines for the diagnosis or therapy of diseases which are based on a dysfunction of the blood-brain barrier.

Surprisingly, it was found that the combination of the above-mentioned process steps allows the unique identification of BBB-specific proteins in brain capillary endothelial cells. The proteins isolated with the method according to the invention are specific for the BBB. Because of their specificity for the BBB, the proteins isolated with the method according to the invention have a function in or on the BBB. This function can be, for example, a barrier function, a transport function, a function in connection with the nutrient supply of the BBB, a function as a tight junction protein, an enzymatic activity, etc. It is thus possible, based on the identification of the presence of these proteins, to derive specific functions from them in the BBB. This derive fish functions from it in the BHS. This opens up the possibility of completely new therapy concepts, which are based on the fact that substances can be targeted through the BHS. Furthermore, the proteins identified with the method according to the invention can be the subject of therapeutic interventions in a targeted manner. The method according to the invention allows for the first time the development of therapeutic concepts for diseases affecting the brain. Furthermore, the detection of changes in the proteins identified by the described method can be used to diagnose diseases based on a dysfunction of the BBB.

Of particular importance in the method according to the invention is the use of freshly isolated BMEC (primary cells) instead of cultivated BMEC. Surprisingly, it was found that BMEC dedifferentiate very quickly in culture, i.e. lose their bras properties very quickly. Furthermore, it was also found that the expression of the proteins specific for the blood-brain barrier in cultured brain capillary endothelial cells is strongly downregulated and disappears completely after only a few passages, as a result of which reliable isolation and identification of BMEC-specific proteins is not possible. In addition, pure and vital cells must be isolated in order to ensure cell specificity and to prevent negative or falsifying effects through apoptosis.

In the method according to the invention, brain material can be removed from the respective organism by surgical intervention in the living organism. In this way, for example, brain samples can also be obtained from the human organism during brain operations. However, it is more advantageous to remove the whole brain or parts of it of which from the organism as soon as possible after death. The brain is preferably removed within a period of at most one hour, more preferably about at most 30 minutes, more preferably about at most 15 minutes or even more preferably about 5 minutes after the onset of death. The brain can be taken from any living being, for example humans, cattle, sheep, goats, horses etc. It has now been found that pig brains are a good model for the human brain with a view to examining the brain capillary endothelial cells for BBB-specific Represent proteins and the transferability of the results to humans.

The pig brain is very similar to the human brain in terms of both anatomy and morphology. Furthermore, sequence homologies between humans and pigs are generally very high, both at the protein and at the nucleic acid level, so that results obtained from pig material can be reliably transferred to humans and vice versa. This is due to the fact that humans and pigs are phylogenetically more closely related than humans and classic model organisms such as mice or rats.

Surprisingly, it was found that a hypotonic lysis buffer not only lyses erythrocytes, but also bursts dead and apoptotic cells in general through hypotonic shock. The lysis buffer to be used in the method according to the invention receives at least 70%, preferably 80%, more preferably 90%, still more preferably 95% of the brain capillary endothelial cells in vital form. Furthermore, the lysis buffer must be non-toxic and have a pH in the physiological range. The hypotonic buffer used according to the invention should have an ionic strength of 0.1-0.2 M, contain mono- and divalent anions or cations and have a pH of Buffer range of 7.0-8.0. All substances contained must be non-toxic to the cells so that healthy cells in the buffer are not damaged for a short time. The hypotonic buffer preferably contains an ionic strength of 0.1-0.2 M sodium, potassium, ammonium, calcium, magnesium,

Chloride and sulfate ions as well as glucose and buffers in a pH range of 7.0-8.0. This allows selective enrichment of vital brain capillary endothelial cells from a mixture of erythrocytes and other cells of varying vitality. The buffer used according to the invention preferably has the following composition at a pH of 7.5:

The lysis buffer used more preferably has the following composition:

NaCl 30mM to 50mM

KC1 4.5mM to 5.5mM

NH ₄ C1 80mM to 100mM

MgCl ₂ 0.6 mM to 0.8 mM

MgS0 ₄ 0.3 M to 0.6 mM NaHC0 ₃ 4.5 M to 6.5 mM

NaH ₂ P0 ₄ 0.2 mM to 0.45 mM

Na ₂ HP0 ₄ 0.4 M to 0.65 mM

KH ₂ P0 0.1 mM to 0.15 M

Glucose 1.5mM to 3.0mM

The buffer particularly preferably has the following composition:

NaCl 39 mM

KC1 5.1 mM

NH ₄ C1 88 mM

CaCl ₂ 1.6 mM

MgCl ₂ 0.69 mM

MgSO ₄ 0.46 mM

NaHC0 ₃ 5, 6 mM

NaH ₂ P0 ₄ 0.33 M

Na ₂ HP0 ₄ 0.53 mM

KH ₂ P0 ₄ 0.12 mM

Glucose 2.24 mM

Such lysis buffers are normally used to isolate lymphocytes or RNA from lymphocytes by first lysing the erythrocytes. Neither the composition of the buffer used according to the invention nor the use of such a buffer for the lysis of apoptotic cells has hitherto been described. The selective lysis of apoptotic cells is of essential importance in the method according to the invention in order to enrich BHS-specific transcripts without simultaneously enriching transcripts of genes which are increasingly expressed during apoptosis. In other cell isolation methods, the problem of apoptosis is avoided by culturing the isolated cells. However, brain capillary endothelial cells change their properties in culture, which leads to a change in the gene expression pattern. Thus, the method for cell preparation according to the invention allows for the first time and in a targeted manner the isolation of sufficient quantities of fresh brain capillary endothelial cells by the final lysis step.

After the brain has been removed from the organism, it is expediently transferred to a suitable buffer and transported to the laboratory as soon as possible, cooled on ice, for further processing. The brain capillary endothelial cells to be isolated are essentially in the gray matter of the brain. Before the cells are further purified, the gray matter of the brain is therefore preferably mechanically prepared from the other parts of the brain. For this purpose, the meninges are first removed and the gray matter is scraped off, crushed and transferred to a suitable medium. A suitable medium is, for example, M199 medium (Gibco / BRL, Grand Island, NY) or Earle's buffer. Before further purification, it is advisable to determine the mass of the gray matter obtained. Earle's buffer: NaCl 117.2 mM

(pH 7.3) KC1 5.3 mM

NaH ₂ PO ₄ x 2 H ₂ 0 1.0 mM

MgSO ₄ x 7 H ₂ 0 0.81 mM

CaCl ₂ x 2 H ₂ 0 1.8 mM

Glucose x H ₂ 0 5.6 mM

According to the invention, the brain capillary endothelial cells are pre-cleaned by disrupting the brain substance in at least two successive enzymatic steps. In a first enzymatic step, the brain substance is digested with the enzyme dispase. The dispase digestion causes the nerve tissue to dissolve. An amount of 5 mg dispase per gram of gray brain substance has proven to be particularly suitable. Dispase digestion is expediently carried out in M199 medium, but other media and buffers are also suitable for this reaction. An appropriately prepared Dispas solution is added to the sample of the gray matter, and the suspension is incubated at 37 ° with stirring. Incubation times of two to four hours, preferably about three hours, have proven to be particularly advantageous. The enzyme concentrations, the solvents or media used and the incubation period must be selected so that as much of the material as possible that surrounds or binds the brain capillaries is broken down or dissolved. At the same time, however, the conditions must be set such that the smallest possible part of the brain capillary endothelial cells to be isolated is attacked or killed in the respective enzymatic step and the cells are exposed to the lowest possible load. It is essential that the resulting shear forces are kept as low as possible. This is achieved, for example, by slowly and continuously mixing the enzymatic digestion of the brain mass in spinner bottles.

After the dispase digestion, the brain capillaries are obtained in a first cleaning stage by centrifugation in dextran solution. Methods known from the prior art can be used for this. It has proven particularly suitable to mix an amount of the cell suspension from the dispase digestion with the same amount of a 15% dextran solution, 10 min. shake and centrifuge for about ten minutes at 10 ° C at 8650 x g in a fixed angle rotor. After centrifugation, the supernatant is removed and the sediment is fed to the second enzymatic step.

In the second enzymatic step, the sediment from the centrifugation is digested with Collagenase D. Collagenase D dissolves the basement membrane, among other things. One or more protease inhibitors are expediently added to the second enzymatic step. The protease inhibitor Na-p-tosyl-L-lysine chloromethyl ketone (TLCK) is particularly suitable for this. The second enzymatic step is expediently carried out with stirring at 37 ° C. for about one hour. It has also proven particularly suitable to use one or more DNAses, such as benzonase, in the second enzymatic step. As a result, the DNA released when digestion of dead cells is broken down, which otherwise increases the viscosity of the suspension.

After the second enzymatic step, a second cleaning step is carried out by centrifugation in the Percoll density gradient. The density gradient is prepared by, for example, 9.91 ml Percoll, 0.72 ml 10 times concentrated Mix M199 medium and 19.37 ml Earle's buffer and centrifuge in the ultra centrifuge at 37200 xg, 4 ° C in a fixed angle rotor for one hour. The cell suspension from the second enzymatic step is washed by repeated centrifugation at low speed, withdrawing the supernatant and resuspending the centrifugation sediment, ie freed from the added enzymes. After the last centrifugation step, the sediment is taken up in a small amount of liquid, such as 6 ml of M199 medium, applied to the prepared Percoll density gradient and centrifuged in the swing-out rotor in the ultracentrifuge at 1400 xg, 4 ° C for ten minutes. Percoll density gradient centrifugation causes the suspended cell material to be separated according to its density, with three discrete bands usually occurring. A first upper band with the lowest density contains cell debris or cell fragments. A second middle band contains the brain capillary endothelial cells to be isolated. Erythrocytes, among other things, collect in a third lower band with the highest density.

The second band, which contains the brain capillary endothelial cells, is isolated and, according to the invention, fed to a further purification. The isolation can be carried out by pulling off the band with the aid of a cannula or, preferably, by pipetting.

In addition to the brain capillary endothelial cells, the material from the second band obtained from Percoll density gradient centrifugation also contains a large number of other cell types, essentially erythrocytes and apoptotic cells. So far it has not been possible to separate these contaminating cells sufficiently from the brain capillary endothelial cells under gentle conditions. Surprisingly, it has now been found that this problem can be solved if the further purification of the brain capillary endothelial cells is carried out with a lysis buffer, as is usually used for the isolation of lymphocytes, the composition of the lysis buffer, the treatment duration and the treatment temperature being selected such that existing ones Erythrocytes and apoptotic cells are essentially completely destroyed and much of the brain capillary endothelial cells survive. The advantages and properties of this buffer have been set out above.

A lysis buffer which contains the following constituents has proven to be suitable according to the invention:

NaCl 30 M to 50 mM

KC1 4.5mM to 5.5mM

NH ₄ C1 80mM to 100mM

CaCl ₂ 1.0 mM to 2.0 mM

MgCl ₂ 0.6 mM to 0.8 mM

MgS0 ₄ 0.3 mM to 0.6 mM

NaHC0 ₃ 4.5 mM to 6.5 mM

NaH ₂ P0 ₄ 0.2 M to 0.45 M

Na ₂ HP0 ₄ 0.4 mM to 0.65 mM

KH ₂ P0 0.1 mM to 0.15 mM

Glucose 1.5 M to 3.0 mM

A lysis buffer with the following composition is particularly suitable: NaCl 39 mM

KC1 5.1 mM

NH ₄ C1 88 mM

CaCl ₂ 1.6 mM

MgCl ₂ 0.69 mM

MgSO ₄ 0.46 mM

NaHC0 ₃ 5, 6 mM

NaH ₂ PO ₄ 0.33 mM

Na ₂ HP0 ₄ 0.53 mM

KH ₂ P0 ₄ 0.12 mM

Glucose 2.24 mM

After addition of the lysis buffer, the suspension is mixed and washed several times by centrifugation at low speed and resuspension in a suitable medium or buffer, such as M199 or Earle's buffer. The cleaned brain capillary endothelial cells collect in the centrifugate.

The purified brain capillary endothelial cells can now be processed in two different ways to identify the presence of BBB-specific proteins or fragments thereof. For example, different proteins or fragments thereof or transcripts can be identified and isolated using the proteomics approach and the genomics approach. Both approaches are described in more detail below.

The following figures explain the subject matter of the present invention in more detail:

Figure la: Northern blot analysis of Itm2A Figure lb: Expression of Itm2A in BMEC under ischemia

Figure 2: Expression pattern of Itm2A in cultured BMEC (M: 100 bp marker)

Figure 3: Expression pattern of S231 (M: 100 bp ladder)

Figure 4: Expression pattern of ssEMPl (M: 100 bp ladder)

Figure 5: Northern blot analysis hybridized with S231 (A) or EMP1 (B) as a probe

Figure 6: Western blot analysis of S231

Figure 7: Homology comparison of human and murine EMP1 and porcine S231. The membrane domain is highlighted in light, the N-glycosylation site in light gray.

Figure 8: Expression pattern of S231 in cultured cells (M: 100 bp marker)

Figure 9: Northern blot hybridizes with full-length FLJ13448 / S012 as a probe

Figure 10: Homology comparison of human, murine and porcine FLJ13448 / S012. The peptides that serve as signal peptides and are split off are shown in italics.

Figure 11: Expression pattern of porcine FLJ13448 / S012 in cultured cells (M: 100 bp marker)

Figure 12: NSE2 amino acid sequence of the human protein. The bold, underlined font identifies the peptides identified in the mass fingerprint. Figure 13: Northern blot for NSE2 hybridized with SEQ ID NO: 22 as a probe

Figure 14: Expression pattern of NSE2 in cultured cells (M: 100 bp marker)

Figure 15: Homology comparison of human NSE2 and NSEl. Potential phosphorilization sites are shown in a bright font. A possible tyrosine kinase domain (ProSite Pattern Match PS00109) is underlined, the active remainder being shown in bold.

Figure 16: Distribution of PEST domains in NSE2. PEST sequences are regions rich in Pro, Glu, Ser and Thr in proteins that are responsible for a short half-life of such proteins in the cell by controlling the ubiquitinization of these proteins. Phosphorilation of certain Ser or Thr residues in the PEST regions (light) is important for the detection and processing by the Ubiquitin-Proteaso way.

Figure 17: The expression of NSE2 in BMEC under ischemia

Figure 18: Amino acid sequence of the human protein DRG-1

(CAB66619). The bold, underlined font identifies the peptides identified in the mass fingerprint.

Figure 19: The homology comparison of human and murine DRG-1 shows 90% identity and 94% homology. Potential phosphorilization sites, a non-conserved potential glycosylation site and the transmembrane domain are shown in a bright font. The N-terminus is located intracellularly. Figure 20: Expression pattern of DRG-1 (M: 100 bp marker)

Figure 21: Expression pattern of DRG-1 in cultured cells (M: 100 bp marker)

Figure 22: TKA-1 amino acid sequence of the human protein. The bold, underlined font identifies the peptides identified in the mass fingerprint.

Figure 23: Northern blot hybridizes with ssTKA-l.ctg as a probe

FIG. 24: Expression pattern of TKA-1 in cultivated cells (M: 100 bp marker)

Figure 25: Expression of TKA-1 in BMEC under ischemia

Figure 26: Western blot analysis of TKA-1

Figure 27: Expression pattern of S064

Figure 28: Expression pattern of ARL8

Figure 29: Multiple tissue blot hybridized with S064 as a probe

Figure 30: Expression of S064 / ARL8 in cultivated BMEC

Figure 31: Expression pattern of 5G9

Figure 32: Homology comparison between HSNOV1 and PNOV1

Figure 33: Prediction of transmembrane domains within the sequence of the protein HSNOV1

Figure 34: Multiple tissue blot hybridizes with 5E7 as a probe

Figure 35: Expression pattern of TSC-22 in cultured BMEC

Figure 36: Reduced expression rate of TSC-22 in BMEC in ischemia Identification of BBB-specific proteins by differential 2D gel electrophoresis

The direct two-dimensional comparison of the gene products enables a complete picture of the brain capillary endothelial cells to be obtained. According to the invention, a comparative tissue is used in all electrophoresis. The reference tissue is a tissue that allows targeted identification of transcripts or proteins that are specific for the blood-brain barrier. In principle, any endothelial cells can be used as reference tissue, for example macro- and microvascular endothelial cells of the same tissue or also endothelial cells from other organs, e.g. Heart, lungs, kidneys, liver, aorta etc. Dedifferentiated BMEC obtained from culture can also be used. However, it is preferred to use another type of endothelial cell as a reference tissue against brain capillary endothelial cells. Endothelial cells from aorta that do not have a barrier function are preferably used. This has the additional advantage that microvessels can be compared to macrovessels. Other microvascular endothelial cells can also be used.

Brain capillary endothelial cells cultured under other conditions are also suitable as comparison tissue, for example under other conditions with regard to pH value, growth matrix, growth factors, for example cytokines. The physiological importance of the identified proteins results from the known properties of the brain capillary endothelial cells compared to the respective comparative tissue. According to the invention, two defined cell types are preferably used: freshly isolated BMEC as the cell type with barrier function and endothelial cells from aorta, which, like BMEC, are also endothelial cells, but have no barrier function. Especially through the Using pig tissue, it is possible for the first time to create such a detailed proteome map of these cells.

sample preparation

The vitality of the prepared cells and the proportion of erythrocytes contained in the preparation must first be determined. To determine the vitality, 20 μl of the suspended cells are removed and 4 μl of fluorescine diacetate working solution (24 μM in Earle's buffer) and 2 μl propidium iodide working solution (70 μM in Earle's buffer) are added. The suspension is mixed and incubated at 37 ° C for 10 min. The cells are documented under a fluorescence microscope and the ratio of vital to damaged cells is determined. Living cells can be recognized by a green fluorescence (excitation 450 nm and emission 515 nm), damaged cells on the other hand by a red fluorescence located in the nucleus (exitation 488 nm and emission 615 nm). The proportion of erythrocytes is determined by adding 20 μl of benzidine working solution (15 mM benzidine hydrochloride, 12% (v / v) acetic acid, 2% (v / v) H ₂ 0 ₂ ) to 20 μl cell suspension. The sample is mixed and incubated for 5 min at 25 ° C. A drop of the cells was then pipetted onto a slide and covered with a coverslip. Erythrocytes appear in this test due to the addition of blue crystals in the transmitted light microscope. The ratio of endothelial cells to erythrocytes is determined by counting. With a vitality ratio of 95% vital cells and an erythrocyte contamination of less than 10%, the cells can be used for the subsequent two-dimensional gel electrophoresis.

The wet weight of the freshly isolated, sedimented cells is determined and with five times the volume (e.g. 100 mg

Cells with 500 μl buffer) buffer A pH 6.8 (10 mM PIPES, 100 mM NaCl, 3 mM MgCl ₂ , 300 mM sucrose, 5 mM EDTA, 1 mM PMSF, 150 μM digitonin) carefully resuspended. The cell suspension was then incubated on ice for 20 min with gentle shaking. Centrifugation (480 g, 4 ° C, 10 min) is then carried out to sediment the cells. The supernatant is removed and stored at -20 ° C until further use.

The sediment is then again in five times the volume of the original wet weight in buffer B pH 7. (10 mM PIPES, 100 M NaCl, 3 mM MgCl ₂ , 300 mM sucrose, 5 mM EDTA, 1 mM PMSF, 0.5% (v / v) Triton X-100) resuspended and incubated on ice for 30 min with vigorous shaking. The sample is then sedimented by centrifugation (5000 g, 4 ° C.) for 10 min, the supernatant is drawn off and stored at −20 ° C. until further use.

The sediment is now in 1.7 times the original wet weight in buffer C pH 7.4 (10 mM PIPES, 10 mM NaCl, 1 mM MgCl ₂ , 1 mM PMSF, 1% (v / v) TWEEN-40, 0.5% ( w / v) deoxycholate) resuspended, transferred to a dounce homogenizer and digested with five strokes. The sample is then transferred back to a 2 ml reaction vessel and incubated for 1 min in an ultrasound bath. The sample is then sedimented by centrifugation (6780 g, 4 ° C, 10 min) and the supernatant is stored at -20 ° C until further use.

Depending on the size, the sediment is to be resuspended in 200 - 500 μl buffer D pH 8.0 (50 mM Tris, 1 mM MgCl ₂ ) and is snap frozen in nitrogen. Then the sample is thawed in an ultrasonic bath and then incubated at 37 ° C with 5 - 10 μl benzonase (25 U / μl) until a homogeneous, no longer viscous liquid forms. Then 7 times the volume of a 5% (w / v) SDS solution is added and the sample is heated to 90 ° C. for 20 min. Centrifugation for 10 min (7000g, 20 ° C) follows to remove insoluble components. The supernatant was removed and stored at -20 ° C until further use. Any sediment that may be present is discarded.

The supernatants are thawed, proportionally combined and mixed. In order to remove the detergents contained in the sample, the sample is mixed with 100% acetone (stored at -30 ° C) in a ratio of 20 to 80. After thorough mixing, the precipitation is incubated at -30 ° C for at least 1 h. The precipitated proteins are then sedimented for 15 min at 10,000 g and 4 ° C. The supernatant is decanted and discarded.

The sediment is then washed with 80% (v / v) acetone (-30 ° C cold) and incubated again at -30 ° C. After renewed centrifugation (15 min, 10,000 g, 4 ° C), the supernatant is discarded and the sediment in the smallest possible amount of solubilization buffer I (7 M urea, 2 M thiourea, 4% (w / v) CHAPS) or II (8 M urea, 4% (w / v) CHAPS) and the protein content of the samples was determined. For the protein determination, 1 part Rotiquant (Roth) and 4 parts double-distilled water are mixed in a 50 ml reaction vessel to a finished working solution and insoluble components are removed using a pleated filter. For the calibration line, dilutions of bovine serum albumin (BSA) are made in solubilization buffer I or II. Concentrations of 0.2 mg / ml, 0.4 mg / ml, 0.6 mg / ml, 0.8 mg / ml and 1.0 mg / ml were set for the calibration solutions. 20 μl each of the calibration solutions, the sample and the reference (solubilization buffer I or II) are placed in a 1.5 ml reaction vessel and 1 ml of the Rotiquant working solution is added. Mixing is carried out in the respective reaction vessel by immediate inverting, after which the sample is incubated at 25 ° C. for 20 minutes. After transferring the sample to a 1 ml cuvette is measured in a spectrophotometer at 560 nm the absorption. The protein content of the samples can be determined by creating a calibration line.

The remaining supernatants of the samples are then stored at -80 ° C until further use.

Isoelectric focusing

For 12 focusing gels, 2.5 mg sample dissolved in 4.5 ml solubilization buffer I or II (for pH gradient 4.5-5.5) are mixed with 1.125 ml focusing buffer I (7 M urea, 2 M thiourea, 4% (w / v) CHAPS, 91 mM DTT, 2.5% IPG buffer) or II (8 M urea, 4% (w / v) CHAPS, 91 mM, 2.5% IPG buffer; for pH gradient 4.5-5.5), treated in an ultrasonic bath for 1 min and then centrifuged for 5 min in a table centrifuge (20,000 g). For the pH gradients used (3.5 - 4.5; 4.0 - 5.0; 4.5 - 5.5; 5.0 - 6.0; 5.5 - 6.7; 6.0 - 9.0) which are used as 24 cm long gels (Immobiline DryStrip; Amersham Biosciences), the appropriate IPG buffers are used.

Then 450 μl samples are pipetted into the rehydration device and the Immobiline DryStrip-

Focus gel with two tweezers, gel side down, placed on the solution with no air bubbles. Then the gel is covered with paraffin oil. The rehydration time is at least 12 h to a maximum of 16 h. At the pH gradient 6.0 - 9.0, in which the sample is applied by cup loading, the corresponding mixture of solubilizing buffer I (4.5 ml) with the corresponding focusing buffer I (1.125 ml) is used instead of the sample Gel strips rehydrated. After rehydration, each gel strip is transferred to a 24 cm strip holder and, in the case of "cup loading", the sample cup directly in front of the Cathode placed. The sample, each with 200 μg ^' protein, is pipetted into the sample cup and, together with the Immobiline DryStrip-Gel, covered with paraffin oil.

The electrode strips moistened with double-distilled water are positioned on the respective ends. The electrodes are then placed on these strips. Then 6 loaded stripholders are focused in an ETTAN IPGphor focusing apparatus (Amersham Biosciences) with a program that corresponds to the pH gradient (see Table 1). After focusing, the strips are removed with tweezers and stored at - 80 ° C until further use.

Program 1 is used for the pH gradients 3.5 - 4.5, 4.0 - 5.0, 4.5 - 5.5, 5.0 - 6.0 and 5.5 - 6.7, while Program 2 is used for gels with a pH gradient of 6.0 - 9.0.

SDS gel electrophoresis

For the second dimension, the required SDS-polyacrylamide gels with a concentration of acrylic id of 12.5% (w / v) are produced in-house.

The gel casting apparatus is assembled according to the operating instructions (Amersham Biosciences) and filled with displacement buffer pH 8.8 (0.375 M Tris, 50% (v / v) glycerin, 0.002% (w / v) bromophenol blue) into the intended reservoir.

The gel polymerization batch pH 8.8 (12.17% (w / v) acrylamide, 0.33% (w / v) bisacrylamide, 0.375 M Tris, 0.1% (w / v) SDS, 0.05% (w / v) ammonium peroxodisulfate) is mixed in a vessel with a spout and then degassed in an ultrasonic bath for 5 min. The polymerization reaction is then started by adding 0.04% (v / v) TEMED. The vessel is immediately mounted on a tripod and connected to the gel casting apparatus via a hose. The gel solution is allowed to flow into the apparatus until it is about 3 cm below the lower edge of the gel cassettes. Then the plug of the reservoir for displacement Solution buffer loosened and the buffer displaces the gel solution until it has risen about 1 cm below the glass edge of the cassette. The cast gels are overlaid with water-saturated n-butanol until the polymerization is complete.

One focusing gel each is removed from the - 80 ° C freezer and transferred to an equilibration tube. By adding 15 ml of reduction buffer pH 8.8 (6 M urea, 50 mM Tris, 30% (v / v) glycerol, 4% (w / v) SDS, 65 mM DTT) the proteins focused in the gels are shaken for Reduced for 15 min at 25 ° C. The reduction buffer is then discarded and the proteins are alkylated with iodoacetamide by adding 15 ml of alkylation buffer (6 M urea, 50 mM Tris, 30% (v / v) glycerol, 4% (w / v) SDS, 260 mM iodoacetamide) , Incubation is also carried out for 15 min at 25 ° C with shaking. The buffer is then also discarded and the gel strip is removed from the tube. Using a pair of tweezers, the gel is placed on the SDS gels and mixed with 2 ml of liquid agarose solution pH 8.3 (0.5% (w / v) agarose, 25 mM Tris, 192 mM glycine, 0.1% (w / v) SDS) overlaid and thereby fixed. The electrophoresis chamber ETTAN DALT II (Amersham

Biosciences) is filled with 10 1 2D running buffer pH 8.3 (25 mM Tris, 192 mM glycine, 0.1% (w / v) SDS), the PAA gels are installed in the device and the electrophoresis run is carried out. A constant output of 5 W per PAA gel is initially set for 50 min. The temperature is constantly 20 ° C.

Then the power is increased to 55 W per gel, but to a maximum of 180 W and the electrophoresis is continued until the blue control dye (bromophenol blue) has reached the lower end of the gels. The electrophoresis is stopped and the gels removed. 400 ml (7% (v / v) acetic acid, 10% (v / v) methanol) are placed in a bowl per gel, the gel is removed from the glass plates and transferred to the bowls. For fixie The gels are incubated for 30 min at 25 ° C with shaking. In the meantime, 400 ml of SyproRuby staining solution are placed in a black bowl and the fixed gels are transferred to the staining solution after the incubation period. After staining for 16 h while shaking, the gels are decolorized in 400 ml fixation for 15 min and for documentation in the FLA 5000 scanner (Fuji) at an excitation wavelength of 473 nm and an emission wavelength of 575 nm with a resolution of 100 μm and a 16 bit Gradiation scanned. The gels are then sealed in plastic wrap and stored at 4 ° C until further use.

Identification of differential spots

In order to identify differential protein spots, the pH gradient was first (3.5 - 4.5; 4.0 - 5.0; 4.5 - 5.5; 5.0 - 6.0; 5.5 - 6.7 ; 6.0 - 9.0) 10 gels each with freshly prepared BMECs and 10 gels from AOECs (Aorta Endothelial Cells) created and scanned. The gels were then compared with Z3 evaluation software (Compugen) and differential spots were annotated. A minimum spot size of 100 pixels was assumed as the filter. The protein spots, which were detected in BMECs higher (three times the amount or more) or uniquely, were cut out with a 1000 μl tip on a blue light table (MoBiTec) and transferred to a 0.2 ml reaction vessel. The cut spots were labeled and stored at -80 ° C until further use.

Hydrolysis and mass spectrometric analysis of the protein samples

The cut-out protein fixed in the gel matrix was removed from the -80 ° C. refrigerator and washed by adding 100 μl of bidistilled water. For this purpose, the respective batch was incubated for 20 min at 25 ° C. with shaking and then the supernatant was pipetted off and discarded. The process was still repeated two more times. It was overlaid twice with 100 μl of 50% (v / v) acetonitrile and incubated for 15 min at 25 ° C. with shaking. The supernatants were discarded again. By adding 100 μl of 100% acetonitrile and incubating for 15 minutes at 25 ° C. with shaking, the gel piece was completely dehydrated. After removing the supernatant, the gel piece was air-dried for 5 minutes. The gel piece was then rehydrated in 15 μl hydrolysis buffer (50 mM (NH ₄ ) ₂ CO ₃ , 25 ng-50 ng / 15 μl trypsin V) and swollen. The proteins were hydrolyzed by

Incubate at 37 ° C for 18 h. To create a peptide "fingerprint" using matrix-assisted laser desorption ionization (MALDI), the hydrolyses are acidified with 15 μl 0.1% (v / v) trifluoroacetic acid. The ZipTip-C18 pipette tips used are rehydrated three times prepared with 10 μl 50% (v / v) acetonitrile and then equilibrated three times with 10 μl 0.1% (v / v) trifluoroacetic acid The sample is applied by drawing up the supernatant of the hydrolysis batch seven to ten times the ZipTips then with 10 μl 0.1% (v / v) trifluoroacetic acid The peptides are directly mixed with the matrix (α-cyanocinnamic acid, 50% (v / v) acetonitrile, 0.1% (v / v) trifluoroacetic acid) Pipette up and down three or four times onto the MALDI measuring carrier. After the samples have dried on the carrier, the samples are first measured and analyzed by mass spectrometry using a MALDI fingerprint. For this purpose, the Voyager DE PRO (PerSeptive Biosyste ms) MALDI mass spectrometer measured the samples in positive reflector mode with an acceleration voltage of 20,000 V, a grid voltage of 75%, a turnout wire of 0.02% and a delay time of 220 ns. A mass window is used that takes masses between 700 - 3500 Da into account. The mass lists obtained for each protein spot are used in a data query. Three different programs are used: Mascot, MSFit and Profound.

Protein spots, in which no database identification is possible despite a good mass fingerprint, are used with ESI mass spectrometry to generate amino acid sequence information.

After the hydrolysis, the hydrolysis batch is acidified with 15 μl of 0.2% (v / v) formic acid and incubated for 30 min with shaking. ZipTip-C18 pipette tips (MilliPore) are rehydrated three times with 10 μl 50% (v / v) acetonitrile and then equilibrated by washing three times with 10 μl 0.1% (v / v) trifluoroacetic acid. The sample is applied by drawing up the supernatant of the hydrolysis batch seven to ten times. The ZipTips are then washed with 10 μl 0.1% (v / v) trifluoroacetic acid, then buffered by washing twice with 0.1% (v / v) formic acid and then the peptides by drawing up 2 μl 50 five to seven times % (v / v) methanol eluted.

The peptide mixtures obtained can be analyzed either directly or by means of liquid chromatography (LC) coupled with ESI mass spectrometry.

For a direct measurement, 1 μl of the eluted sample is filled into a hollow needle (Protana). First, an overview spectrum with an ion spray voltage of 850 - 1000 V, a curtain gas pressure of 20 psi, a "declustering" potential of 40 - 50 V, a focusing potential of 245 V and the voltage on the multi-channel plate from 2000 - 2100 V measured in "positive mode". The scan range is 100 - 1600 or 410 - 1600 Th. The peptides detected in the spectrum are subjected to shock-induced fragmentation. For this purpose, production spectra of each peptide at an ion spray voltage of 850-1000 V, a curtain gas pressure of 20-40 psi, a “declustering” potential of 40-50 V, a focusing potential of 245 V, a quadrupole resolution of 0.7-1 , 0 amu, a collision energy of 15 - 50 V and a voltage on the multi-channel plate of 2100 - 2400 V. The scanning range is 50 - 1600 Th.

When the nanoHPLC is coupled to the ESI mass spectrometer, the reversed phase (RP) precolumn is first loaded with 2 μl sample at a flow of 20 μl / in 0.1% (v / v) trifluoroacetic acid. The peptides are chromatographically separated on a RP-C18 column (LC packings) with a gradient over 35 min from the initial conditions (0.05% (v / v) formic acid, 10% (v / v) acetonitrile) to Final conditions (0.05% (v / v) formic acid, 76% (v / v) acetonitrile). The HPLC is coupled to the mass spectrometer via a hollow needle (New Objective). The settings of the mass spectrometer are chosen so that two experiments can be carried out during the LC run. In addition to the ion spray voltage (1800 - 2200 V), the parameters set correspond to those already listed above. Both overview spectra and production spectra are recorded alternately during the run. The settings for the production spectra are selected so that the two most intense signals in the overview spectrum, which are loaded twice, three times or four times and whose intensity is greater than 10 cps, are then analyzed using a shock-induced fragmentation. The scan range is 450 - 1600 Th. The evaluation of the spectra obtained is carried out in three stages:

A) The production spectra, the information about the

Amino acid sequence of the corresponding peptide are first determined using the software program es MASCOT (Mat- rix Sciences) completely compared with public databases. If the peptide cannot be assigned to a protein,

B) the production spectra are first automatically sequenced using a software tool from the device manufacturer. The amino acid sequences thus obtained were according to Shevchenko et al. compared to the public databases via MSBlast. The protein could not be identified

C) the production spectra were evaluated manually and the amino acid sequences obtained were compared with the public databases using Blast or FASTA.

According to the method described above, BHS-specific proteins or fragments thereof can be specifically identified in brain capillary endothelial cells. The foregoing description, of course, allows for routine variations that will be apparent to those skilled in the art. For example, the following process steps can be varied:

Other methods known to those skilled in the art for obtaining the proteins described in the standard literature can also be used as the digestion method (for example “2D-Proteome Analysis Protocols”)

Corresponding focusing gels from other manufacturers can of course be used for isoelectric focusing. Different lengths and pH gradients can also be used.

Corresponding gel systems from other manufacturers can of course be used for the separation in the second dimension. The use of other gel sizes is also possible. As a standard, other proteases known to those skilled in the art can also be used to produce the peptide pattern.

Mass spectrometers of other types and from other manufacturers can also be used to determine the peptide masses and de novo amino acid sequences.

To determine the peptide masses and the de novo amino acid sequencing, the mass spectrometric conditions can be varied both in terms of equipment and function, according to the sample.

Western blot of the proteins

First, the proteins were separated on 12.5% polyacrylamide gels as described above and then transferred to nitrocellulose membrane.

For this purpose, seven Whatman papers (Schleicher & Schüll) were cut to size of separating gel and each soaked with different buffers.

Two papers in anode buffer I (300 mM Tris base, 20% (v / v) methanol) were placed on the anodes of the blot apparatus (BioRad) without air bubbles, followed by two papers in anode buffer II (25 mM Tris base, 20 % (v / v) methanol). The nitrocellulose membrane, also soaked in anode buffer II, was placed on top, followed by the polyacrylamide gel. Finally, three papers were soaked in cathode buffer (25 mM Tris, 40 mM aminocaproic acid, 0.1% (w / v) SDS, 20% methanol) and placed on top. The apparatus was closed and the proteins were transferred for one hour at a maximum of 25 V and 2.5 mA / cm ² gel. The proteins were then immunochemically stained with rabbit polyclonal antisera.

For this purpose, the membranes were in TBST buffer (10 mM Tris base,

150 mM sodium chloride, 0.05% (v / v) Tween 20; pH 8.0) and then free for 30 min with blotto (10 mM Tris base, 150 mM sodium chloride, 0.05% (v / v) Tween 20, 5% (w / v) skimmed milk powder; pH 8.0) Saturation places saturated. Incubation with the 1st antibody was carried out for 2 h at RT in TBST buffer [anti-EMPl antibody (rabbit) 1: 4000; Anti-TKA-1 antibody (rabbit) 1: 4000], followed by washing three times with TBST buffer. The bound antibodies were detected by incubation with a second antibody conjugated with alkaline phosphatase for 1 h at RT in TBST buffer [anti-rabbit IgG antibody (goat) 1: 5000]. After washing twice with TBST buffer, the membrane was buffered to an alkaline pH by incubation with AP buffer (100 mM Tris base, 100 mM sodium chloride, 5 mM magnesium chloride; pH 9.5). 0.016% (w / v) nitrotetrazolium blue chloride and 0.033% (w / v) 5-bromo-4-chloro-3-indolylphosphate disodium salt in AP buffer were used as substrates for the color reaction.

The identification of BBB-specific proteins or fragments thereof in brain capillary endothelial cells using the genomics approach is now described below.

Identification of BBB-specific transcripts by cDNA subtraction

The targeted identification of cell- or tissue-specific proteins is carried out using differential methods. This can be done at the protein level by comparing 2D gels from digestions of different tissues or cells and then determining the proteins specific for a tissue or cell type. To from the physical properties of To be independent of proteins (size, solubility), differential methods can also be carried out at the transcript level to identify specific proteins. Such subtractive RNA techniques also have the advantage of requiring less tissue or cell material.

The use of freshly isolated BMEC as a starting material is crucial for the identification of BBB-specific proteins. The methods described so far have at best been based on the subtraction of RNA from brain capillaries against RNA from kidney (Li et al., 2001). The problem here is that brain capillaries also contain other cell types such as pericytes and astrocytes in addition to BMEC. Furthermore, the kidney subtraction tissue is very heterogeneous, since it consists of different cell types, of which endothelial cells make up only a small part. According to the invention, a subtraction tissue is to be used which permits targeted identification of transcripts or proteins which are specific for the blood-brain barrier. In principle, any endothelial cells can be used as reference tissue, for example macro- and microvascular endothelial cells of the same tissue or also endothelial cells from other organs, for example heart, lung, kidney, liver, aorta etc. Dedifferentiated BMEC obtained from culture can also be used. However, it is preferred to use another type of endothelial cell as a reference tissue against brain capillary endothelial cells. Endothelial cells from aorta that do not have a barrier function are preferably used. This has the additional advantage that microvessels can be compared to macrovessels. Other microvascular endothelial cells can also be used. Brain capillary endothelial cells cultured under other conditions are also suitable as comparison tissue, for example under different conditions with regard to pH value, growth matrix, growth factors (for example Cytokines etc.). From the known properties of the brain capillary endothelial cells in relation to the respective comparative tissue, the physiological significance of the identified targets results. According to the invention, two defined cell types are preferably used: freshly isolated BMEC as the cell type with barrier function and endothelial cells from aorta, which, like BMEC, are also endothelial cells, but have no barrier function. This approach allows much more targeted transcripts or proteins to be identified that contribute to the formation of the blood-brain barrier.

Preparation of the subtractive cDNA library

Total RNA is isolated from the cells with Trizol (Invitrogen) according to the manufacturer's instructions. The total RNA is then checked for intactness on a denaturing agarose gel. For RNA isolation, 100 mg of tissue or 10 cm ^{2 of} confluently grown cells are mechanically homogenized in 1 ml of trizole and the homogenate is then incubated for 5 min at RT. Then 0.2 ml of chloroform / 1 ml of Trizol (Invitrogen) are added, mixed by vortexing for 15 seconds and incubated at RT for 3 minutes. For phase separation, centrifugation is carried out at 4 ° C. and 12,000 × g for 15 min and then the upper, aqueous phase is transferred to a fresh vessel. For this purpose, 0.5 ml isopropanol / 1 ml trizole is added, mixed and incubated at RT for 10 min. The RNA is sedimented by centrifugation at 4 ° C. and 12,000 × g for 10 min, washed twice with 75% EtOH, air-dried and dissolved in DEPC-treated water. The concentration is determined spectrophotometrically and the quality checked in a denaturing agarose gel.

Starting from total RNA, the mRNA is enriched using Dynabeads (Dynal) according to the manufacturer's instructions. RNA enrichment: 75 μg of total RNA is denatured for 2 min at 65 ° C., immediately added to 200 μl Dynabeads 01igo (dT) ₂₅ (Dynal) in double binding buffer and incubated for 5 min with mixing. The supernatant from the magnetic separation is discarded and the Dynabeads are washed twice with washing buffer. The polyA ⁺ RNA is finally eluted with 20 ul 10 M Tris-HCl pH 7.5 for 2 min at 85 ° C.

The subtractive cDNA library can be produced using commercially available PCR subtraction kits. For example, the PCR-Select cDNA subtraction kit from Clontech can be used according to the manufacturer's instructions.

For this purpose, 2 μg mRNA from BMEC (tester) and AOEC (driver) are transcribed from a oligo (dT) adapter primer with the enzyme AMV reverse transcriptase into single-stranded cDNA. Immediately afterwards, the second strand synthesis is carried out with an enzyme mixture (DNA polymerase I, RNase H and DNA ligase) for two hours at 16 ° C. and then with the addition of T4 DNA polymerase and further incubation at 16 ° C. for 30 minutes. The double-stranded cDNA thus produced is purified by phenol / chloroform extraction and ethanol precipitation.

In order to introduce suitable ends for the later adapter ligation and to produce a more uniform size distribution of the cDNA fragments, a restriction with Rsa I now takes place. The double-stranded cDNA fragments thus produced are purified by phenol / chloroform extraction and ethanol precipitation. The products of the cDNA syntheses and the restrictions are checked for purity by gel electrophoresis.

For later amplification by PCR, the adapters 1 and 2R are now connected to the tester cDNA via the Rsa I ends Enzyme T4 DNA ligase attached. The ligation is checked by PCR.

The actual subtraction is carried out by two hybridizations. For the first hybridization, cDNA from BMEC adapter 1 is hybridized with AOEC cDNA in one approach, and cDNA from BMEC adapter 2R with AOEC cDNA in another approach. In the second hybridization, the two approaches from the first hybridization are combined and hybridized with freshly denatured cDNA from AOEC.

The products from the hybridization are eventually called

Template used in a first PCR reaction; An oligonucleotide from the common area of the two adapters 1 and 2R serves as the primer.

The product mixture from this first PCR was then used as a template in a nested PCR (nested PCR), the two primers arranged one inside the other coming from the unique area of the two adapters 1 and 2R. This second PCR increases the specificity.

The efficiency of the subtraction was checked by comparative PCR on a household gene (GAPDH): in comparison to the two non-subtracted cDNAs from BMEC and AOEC, the cDNA from the subtraction can only be used to produce products after significantly more PCR cycles. GAPDH is expressed as a typical household gene in all tissues and cell types of comparable strength. Therefore, in a subtractive hybridization, it should not be enriched like differentially expressed genes, but the amount of transcript in the subtracted cDNAs (both forward and reverse subtraction) should decrease significantly compared to the cDNAs from BMEC or AOEC before subtraction. This is confirmed experimentally by the two cDNAs before the subtraction or the respectively subtracted cDNAs are used in a PCR with GAPDH-specific primers. Since with the subtracted cDNAs an initial product formation is only achieved after an additional 16 cycles compared to the two non-subtracted cDNAs, the hybridization therefore results in an enrichment by at least a factor of 50,000. This enrichment enables the targeted identification of BBB-specific transcripts and also represents a first validation of the isolated sequences.

The products of the second PCR are cloned into the vector pT-Adv (Clontech) and into TOP10F '(Clontech) chemocompetent E. coli transformed. The products of the second PCR are cloned into the plasmid vector pT-Adv (Clontech). This vector has protruding dT residues at the 5 'ends which are compatible with the 3' dA residues e.g. attached to PCR products by Taq DNA polymerase. This or comparable systems allow the direct cloning of PCR products with high efficiency. The transformation takes place in chemocompetent E. col i TOP10F '(Clontech) as described in the literature (Sambrook et al., 1989).

Differential hybridization

Clones from the subtractive cDNA library are expressed by differential hybridization with respect to their expression BMEC vs. AOEC verified. The PCR-Select Differential Screening Kit (Clontech) is used for this. The reverse subtracted probe was produced using the PCR-Select cDNA subtraction kit from Clontech according to the manufacturer's instructions as described above, with BMEC as driver and AOEC as tester.

According to the manufacturer's instructions, liquid cultures of the clones are inoculated in microtiter plates with 96 cavities. These are used as templates for amplifying the insertions with the Primers adapter 1 and 2R used. The rest of the liquid cultures are mixed with glycerin and frozen as a permanent culture. The PCR products are checked by gel electrophoresis. For products that were larger than 200 bp, 1 μl is spotted on two identical HybondN membranes and fixed on them with UV light. Contrary to the manufacturer's instructions, only two filters with 92 clones are hybridized: one filter with the forward subtracted probe, in which BMEC-specific transcripts are enriched, and the other filter with the reverse subtracted probe, in which AOEC-specific transcripts are enriched , The hybridization of two additional filters with cDNA from BMEC or AOEC was dispensed with, since no relevant additional information can be derived from this. Instead, RNA from BMEC and AOEC is used for later verification in Northern blot analyzes or RT-PCR experiments to create expression patterns. PCR products from 92 clones from the subtractive cDNA library and two negative controls from the manufacturer are applied to each filter. In addition to the manufacturer's information, a PCR product of a household gene is spotted, which is in BMEC and

AOEC is expressed equally strongly, and as a positive control a PCR product to a BBB marker (apolipoprotein AI), which is more strongly expressed in BMEC than in AOEC.

The hybridizations are carried out as described by the manufacturer with probes of the same activity at 72 ° C. with “ExpressHyb” solution

(Clontech) and the filters are then washed stringently. Standard stringency conditions can be used. The filters are advantageously washed 2 × 20 min at 68 ° C. up to a stringency of 0.2 × SSC / 0.5% SDS. The signal intensities are determined by exposing the film to different lengths of time using a phosphoimager (FLA-5000, Fuji). Clones that are about five times stronger Signals shown in BMEC as in AOEC are classified as differentially expressed and processed.

Liquid cultures are inoculated from positive clones from the permanent culture and the plasmid DNA is isolated using standard methods (Birnboim and Doly, 1979) using Qiagen columns. The plasmid insertions are sequenced with universal primers and optionally additional gene-specific primers. Databases are obtained with the DNA sequences obtained using the BLAST (http://www.ncbi.nlm.nih.gov/BLAST) and FASTA algorithms

(http://www.ebi.ac.uk/fasta33) searched for homologies.

Further verification of BHS-specific transcripts: expression pattern

Expression patterns are created from the positive clones of interest in BMEC, AOEC and nine other tissues. This is done by RT-PCR and / or Northern blot analysis. •

For RT-PCR experiments, cDNAs are made from total RNA by random priming. All enzymes used and the random hexa ere come from Invitrogen. For this purpose, 10 μg total RNA in 40 μl nuclease-free water are mixed with 5 μl DNase I 10x buffer and 5 μl DNase I and incubated at 25 ° C. for 15 minutes. 5 μl of 25 M EDTA are then added and the enzyme is heat-deactivated at 65 ° C. for 15 minutes. 25 μl are removed from the batch, made up to 100 μl with nuclease-free water and stored at -80 ° C. as an RT control. 8 μl of random primer (100 ng / μl), 3 μl of dNTP mix (10 M each) and 2 μl of nuclease-free water are added to the remaining 25 μl of total RNA from the DNase I digestion. Now the RNA secondary structures are dissolved for 5 minutes at 65 ° C and the sample is then immediately placed on ice. There are 10 ul 5x Ist Strand buff r, 6 ul DTT (100 mM) and 3 μl RNaseOUT was added, incubated for 10 minutes at 25 ° C for priming and then tempered at 42 ° C for 2 minutes. Now 3 μl SuperScript II Reverse Transcriptase are added and incubated at 42 ° C for 50 minutes. The enzyme is then heat deactivated at 70 ° C. for 15 minutes. In order to break down the total RNA from the cDNA, 3 μl RNase H are added and incubated at 37 ° C. for 20 minutes. Finally, make up to 100 μl with nuclease-free water and store the cDNA at -80 ° C. The quality of the cDNAs is checked by PCR with primers for a household gene (GAPDH) or for the 18S rRNA. It is to be expected here that comparable product quantities with the cDNAs arise from the different tissues or cells. The cDNAs produced in this way are each used to create expression patterns for the transcripts to be examined.

For Northern blot analyzes to create expression patterns, total RNA from the cells or tissues in denaturing gels is separated according to size, transferred to a nylon membrane and hybridized there with radioactively labeled, gene-specific probes. 6.0 g of agarose is heated in

Dissolved 290 ml DEPC-treated water. The mixture is then cooled to 60 ° C. in a water bath and 40 ml of 10 × MOPS buffer (200 mM MOPS, 50 mM sodium acetate, 10 mM EDTA) and 70 ml of formaldehyde are added. Finally, a maxi gel with a large pocket former (12 tracks) is poured into the trigger and allowed to solidify there. 15 μg total RNA in 10 μl are denatured with 40 μl sample buffer (500 μl deionized formamide, 160 μl formaldehyde, 100 μl lOx MOPS, 240 μl DEPC-treated water) for 15 minutes at 65 ° C and then transferred to ice leads. Now 10 μl loading buffer (500 μl glycerol, 2 μl 500 mM EDTA, 25 μl 10% bromophenol blue, 473 μl DEPC-treated water) are added and the sample is loaded with 1 × MOPS buffer layered gel applied. The electrophoresis is carried out at 250 V for 3-4 hours. Then the gel is first swirled in water for 10 minutes, then in lOx SSC for 30 minutes. A Hybond XL filter cut to gel size is swirled in 10X SSC for 15 minutes.

Structure of the blot (from bottom to top): salt bridge (immersed in buffer reservoir with lOx SSC), gel, filter, 5 3MM (previously inserted in lOx SSC), approx. 7 cm green cloths (cellulose cloths), glass plate, weight of approx 0.5 kg. The blotting is carried out for 16-20 h. Then the blot is dismantled and the

Hybond XL filter washed in 2x SSC for 10 minutes. The RNA is now fixed in a UV crosslinker at 70,000 μJ / cm ² on the Hybond filter. The filter is then stained for 1 minute in staining solution (300 mg methylene blue in 1 L 0.3 M Na acetate) to make the RNA visible and then washed with water for 2 minutes to decolorize the background. The colored filters are documented photographically. The filter is then dried between 3MM paper, wrapped in Saran wrap and stored at -20 ° C.

The hybridizations are carried out using radioactively labeled cDNA probes (Rediprime II, Amersham), which were purified using ProbeQuant G-50 columns (Amersham), using ExpressHyb solution (Clontech) according to the manufacturer's instructions. After a first check of the hybridization using the FLA-5000 phosphoimager (Fuji), autoradiograms are made on Biomax MS films (Kodak).

Completion of cDNA sequences

The complete cDNA sequences of the BHS-specific clones of interest from the subtractive cDNA bank are determined by screening various cDNA banks and RACE-PCR experiments. A cDNA bank is created from BMEC from Schwein with the SMART cDNA Library Construction Kit (Clontech) in the vector λTriplEx2 according to the manufacturer's instructions. For this purpose, total RNA is first isolated with Trizol (Invitrogen) as described above, and polyA ⁺ RNA is enriched therefrom using Dynabeads (Dynal). 2 μg of polyA ⁺ -RNÄ from BMEC are used to manufacture the bank. Finally, the ligations are packaged in vitro with the Gigapack III Gold phage extract (Stratagene) according to the manufacturer's instructions. The number of independent phages from the BMEC cDNA library is 1.3 million pfu, more than 99% of which were recombinant when a blue / white test was carried out (cf. Sambrook et al., 1989). At least half of the inserts have a size of more than 1 kb. After amplification of the complete bank, the titer is approx. 2 x 10 ¹⁰ pfu / ml with a total volume of approx. 150 ml. This phage lysate is reduced to 7

(v / v)% DMSO adjusted and stored at -80 ° C. The phage bank described is converted into a plasmid bank according to the manufacturer's instructions (Clontech ClonCapture cDNA Selection Kit) by infecting E. coli BM25.8 with 2 million pfu of the phage bank. This bacterial strain expresses Cre-

Recombinase, which recognizes the loxP sites in the vector λTriplEx2 and thus enables the conversion. The conversion of the lamda phages into plasmids takes place by in vivo excision and subsequent circularization of the complete plasmid. The plasmids obtained are then stable in E. coli passed. The plasmid preparation is carried out from plate cultures of infected BM25.8 with the NucleoBond Plasmid Kit (Clontech).

Biotinylated cDNA probes are used to screen cDNA plasmid banks with ClonCapture. In a RecA-mediated reaction, these form DNA triplex structures with homologous sequences of the plasmid insertions. The so select Plasmids can be isolated via streptavidin coupled to magnetic beads and used in a transformation. Clones from such an enrichment are then screened by colony hybridization, and the plasmid DNA is isolated and sequenced from the resulting positive clones.

The isolation of positive clones by ClonCapture is carried out exactly according to the manufacturer's instructions (Clontech). For probe production, a PCR with gene-specific primers was first optimized on a suitable plasmid, so that only one product was created. For this purpose, the primers are designed in such a way that they have melting temperatures that differ from one another by a maximum of 1 ° C, and do not form any primer dimers or stable loops. The Annealinςr temperature is calculated at 2-5 ° C below that according to the formula T _m = [(G + C) x 4] + [(A + T) x 2]

Melting temperature chosen relatively high. This leads to specific product formation, which manifests itself in only one band when checked by gel electrophoresis. A piece of this product is removed from an agarose gel with a sterile Pasteur pipette and transferred to 200 μl sterile water. The DNA is eluted from the gel piece by swirling and incubating at 70 ° C. for 30 minutes and serves as a template for probe production. Control reactions with and without biotin-21-dUTP are carried out with this template and analyzed in an agarose gel, since biotin can inhibit the PCR. If the control reaction is successful, the biotinylated probe is then prepared in a preparative PCR with the simultaneous addition of 10 μCi [α ³² P] dCTP. After 20 cycles, 5 μl from the mixture are checked on an agarose gel and 5 further cycles may be connected. The PCR product is then cleaned with the NucleoSpin Extraction Kit (Clontech) according to the manufacturer's instructions and with 35 μl elution buffer eluted. 2 μl of this are analyzed by gel electrophoresis and the product quantified spectrophotometrically. To check the biotinylation, 2 μl of the purified PCR product are added to 15 μl magnetic beads and the pre-incubation signal is determined using a Geiger counter. After 30

Minutes of incubation with gentle shaking, the magnetic spheres in the magnet are separated and the supernatant is again quantified using the Geiger counter (post-incubation signal). If biotinylation is successful, the pre-incubation signal is 2-4 times stronger than the post-incubation signal.

For capturing, 50 (200 bp) - 100 (600 bp) ng of biotinylated PCR product are denatured in water for 5 minutes at 100 ° C and then immediately transferred to ice. Now all components except the plasmid DNA are added, 2 μg RecA protein per 50 ng probe being used. After 15 minutes of incubation at 37 ° C., 1 μg of the plasmid bank is added and incubation is continued at 37 ° C. for 20 minutes. In the meantime, 15 μl magnetic beads are unsaturated with herring sperm DNA and prepared for the cleaning of the capturing. For capturing, EcoR V-cut λ DNA is added and, after adding SDS, proteinase K digestion is carried out for 10 minutes at 37 ° C. This reaction is finally stopped by adding PMSF and the capturing approach is cleaned up via the magnetic beads. The isolated plasmids are eluted with 100 μl elution buffer, precipitated and then dissolved in 10 μl water.

2 μl of the plasmid bank enriched by ClonCapture are placed in electrocompetent E. coli DH5α transformed and plated on LB-Amp plates. Positive clones are generated by colony hybridization with radioactively labeled probes (same

Amplicon as in biotinylation) identified by standard methods. The clones thus obtained are by colony PCR further verified, using a primer from the amplicon mentioned and a further primer located downstream. Avoid choosing both primers from the amplicon that was used as a probe for ClonCapture to use in colony PCR [a PCR approach is performed in which bacteria from a single colony are placed instead of DNA] to avoid that the product formation does not take place on the plasmids contained in the bacteria, but rather by means of a contaminating probe. Therefore, at least 1 primer should be outside the amplicon, ideally 3 'to it, since this sequence is both known and is contained in all positive clones of the cDNA library. The plasmid DNA was isolated from positive clones by standard methods (Birnboim and Doly, 1979) using Qiagen columns and sequenced using the chain termination method (Sanger et al., 1977). The "ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit, Version 2.0" (Applied Biosystems) according to the manufacturer's instructions can be used for sequencing. The products of the sequencing reactions are analyzed on the "ABI Prism 310 Genetic Analyzer" (Applied Biosystems).

RACE-PCR (Froh an et al., 1988) is used to determine unknown cDNA sequences from a known sequence section by cDNA synthesis, followed by the introduction of known, synthetic ends for attachment of the second PCR primer.

The 5 'RACE-PCR is carried out with the 5' RACE System for Rapid Amplification of cDNA Ends, Version 2.0 (Invitrogen) according to the manufacturer's instructions. First, a cDNA first-strand synthesis with a gene-specific primer (GSP1) and 1 μg total RNA from BMEC takes place. At the 3 'end of this cDNA, after cleaning the cDNA over GlassMAX columns in a second step with the aid of the enzyme terminal deoxynucleotide transfer. an oligo dC tail attached. The first PCR is carried out on 5 μl of thawed cDNA with a further gene-specific primer (GSP2) and the abridged anchor primer, which attaches to the oligo-dC tail. The specificity of the PCR was increased with the help of a second, nested PCR, which is carried out with the abridged universal amplification primer and a third gene-specific primer (GSP3) on 5 μl 1: 100 diluted PCR product from the first PCR. The second PCR batches with only one primer and a water control serve as controls. After gel electrophoretic analysis, the product of the second PCR may be cloned, for which purpose a ligation with the pGEM-Teasy System II (Promega) and transformation into electrocompetent DH5α is carried out. The clones obtained are examined using colony PCR, the plasmid DNA is prepared and finally sequenced in a manner known per se.

The 3 'RACE-PCR can be carried out with the 3' RACE System for Rapid Amplification of cDNA Ends (Invitrogen) according to the manufacturer's instructions. In the 3 'RACE-PCR, the cDNA first strand synthesis is carried out on 5 μg total RNA from BMEC using the Oligo-dT adapter primer. For the first PCR, 2 μl cDNA with a gene-specific primer (GSP1) and the abridged universal amplification primer are used. A semi-nested second PCR is carried out as described for 5'RACE with a gene-specific primer (GSP2) and the abridged universal amplification primer including the controls. The products are cloned and sequenced as described.

Using the method described above, BHS-specific proteins or fragments thereof can be specifically identified in brain capillary endothelial cells. The above description of course allows routine variations that one Expert are obvious. For example, the following process steps can be varied:

• Isolated BMECs can be sown and a primary culture can be used as a "tester" for the subtraction instead of the fresh BMECs.

Another subtraction tissue ("driver"), e.g. dedifferentiated BMEC from the culture (min. passage 2) can be selected.

• RNA or mRNA can be prepared by any other method known to a person skilled in the art, with the proviso that the RNA is intact or the mRNA can be transcribed into cDNA by reverse transcription.

• The PCR products from the subtraction can be cloned into any suitable vector system, both via polyerase-related 3 'dA residues, as well as over blunt ends or after restriction. The transformation can be in different E. coli strains occur in both chemically and electro-competent cells, as is well known in the art.

• The differential hybridization step is optional, but recommended. Other suitable membranes (for example positively charged or uncharged nylon membranes) and other hybridization solutions can also be used here. Stringent washing of the membranes can also be achieved at other temperatures or with other solutions (for example low temperatures and low salt content or higher temperature and higher salt content, for example T = 50-70 ° C., 0.5-0.05 times) SSC / 0.1% 5DS). • Expression patterns can also be determined by quantitative PCR (real-time PCR) with the corresponding cDNAs. The quantitative PCR is expediently carried out using the Opticon (MJ Research). The "QuantiTect SYBR Green PCR Kit" from Qiagen is used to carry out the reaction, PCR conditions being used as described above. A series of dilutions is made from BMEC cDNA for quantification, the information is given in picograms of RNA equivalents In the relative quantification, the calculated amounts of target are divided by the calculated amounts of 18S rRNA, then a sample, eg BMEC, is set as 100% and all other samples are related to it.

• The cDNAs can also be produced using other systems. Northern blot analyzes can also be carried out with other suitable probes and hybridization / washing solutions.

• cDNAs can also be expanded through database mining with the help of known, overlapping sequences. Any other cDNA banks from cells or tissues in which the searched transcript is found can also be experimentally screened with various systems or RNA from cells or tissues in which the searched transcript occurs can be used in RACE-PCR (see Sambrock, 1989). Any other suitable systems known to a person skilled in the art can be used for RACE-PCRs.

The proteins or fragments identified by this method have a specificity for the blood-brain barrier and are also the subject of the present invention. Knowing the specificity of a protein or fragment thereof for the The blood-brain barrier now allows the protein to be found in a targeted manner. As a rule, the function is determined by comparison with known sequence data in available databases, for example using the BLAST algorithm. Knowing the specificity of the identified proteins also allows a targeted modulation of their expression in the blood-brain barrier, whereby pathological conditions can be treated in a targeted manner.

In this way, agonists or antagonists can be developed to the respective BBB-specific proteins that selectively modulate their activity. The expression of such proteins can also be modulated directly, e.g. by gene transfer or anti-sense RNA. The development of "Trojan horses" is particularly attractive for therapeutic approaches - drugs that are linked to molecules that are actively transported by identified transporters via the BBB. So-called pro drugs, substances that are derived from BBB-specific enzymes in the endothelial cells, are also possible be modified and thus achieve their therapeutic effect.

BHS-specific proteins perform a variety of functions. For example, they serve to supply nutrients (example glucose transporter GLUT1) or serve as contact proteins (e.g. ZO-1 as a tight junction protein). They also have enzymatic activity (e.g. glutamyl transpeptidase GGT) or act as a transport vehicle for amino acids.

Expression of BBB-specific proteins in ischemia

The expression behavior of the BBB-specific proteins identified according to the invention in the case of ischemia was examined. For this purpose, the prepared endothelial cells were resuspended after washing and, as in Franke et al. (2000) described in cell culture bottles coated with collagen. The The cells were cultivated at 37 ° C. in CO ₂ incubators with a constant CO ₂ content of 5%. After the cells had reached confluence, they were detached by treatment with trypsin solution and split into prepared transwell dishes (44 cm ² , Corning). After culturing the cells for 3 days under the conditions already described, the Transwell batch was converted into a dish, on the bottom of which C6-glioma cells (commercially available, for example obtainable from the ATCC) had grown. The two cell types were cultivated for two days in coculture with the addition of hydrocortisone. A medium change was carried out for the experiment for expression under ischemia. The new medium was previously gassed with 0.2% 0 ₂ , 94.2% N ₂ and 5% CO ₂ and contained no glucose. The cells were then stored for 24 h at 37 ° C. in CO ₂ incubators with 0.2% 0 ₂ , 94.2% N ₂ and 5% CO ₂ . A medium change was also carried out during the control. The medium was previously gassed with 21% 0 ₂ , 74% N ₂ and 5% CO ₂ and contained glucose. The cells were further cultured for 24 hours under these conditions. The expression of the respective protein was then determined quantitatively, as described above.

The following proteins have now been identified at the blood-brain barrier using the method according to the invention.

Example 1: Identification of S129 = ITM2A

BMEC freshly isolated and purified from the brain of pigs as described above were cultured in M199 medium (Sigma) with 10 (v / v)% ox serum (PAA) on collagen G (biochrom) and passaged by trypsinization. Total RNA was isolated from cultured BMEC from the primary culture (PO) and from passages 1-3 (PI-3) from a T75 cell culture bottle as described above. This became like already described cDNA produced and examined for their quality. Expression patterns were created with the respective gene-specific primers comparing fresh BMEC and PO-3, each with reference to GAPDH or 18S rRNA. The clones described in this and the following examples were obtained.

The subtractive clone S129 showed a> 5 times stronger signal in the differential screen with the forward probe compared to the reverse probe and was therefore selected for sequencing. The sequence of clone S129 is given as SEQ ID N0: 1.

Based on this sequence, S129 was clearly identified as Itm2A.

First, as described, an expression pattern for Itm2A with the primers Itm2a.s2 (5 'ACC TCC ATT GTT ATG CCT CCT A 3' = SEQ ID NO 2) and Itm2a.as2 (5 'GTT GCC TCT CAC TCT TGA CAG A 3' ' = SEQ ID NO 3), GAPDH was used as a control. The expression pattern was obtained by RT-PCR (not shown).

The semi-quantitative expression pattern shows that Itm2A is more strongly expressed in BMEC than in AOEC and thus confirms the result of the differential hybridization. In addition, the expression in BMEC is also significantly stronger than in Cortex (brain), which is an indication of the specificity for BMEC in the brain. A strong expression can only be seen in the heart, which can possibly be correlated with the expression described in muscle. The expression pattern was verified by Northern blot analysis, the coding region of Itm2A from pig (FIG. 1a) being used as the probe. The specificity for BMEC becomes even clearer in the Northern blot. This expression in BMEC and thus at the BHS has not yet been described.

Furthermore, a second, smaller transcript in BMEC can be seen in the Northern blot. In order to characterize this, the coding region, as well as 5 'and 3' non-coding region were examined with RT-PCR or RACE-PCR in BMEC. It was found that there are two 3 'non-coding regions for Itm2A, the shorter of which results from an alternative polyadenylation signal, as was shown by sequencing. This was also not previously described for Itm2A. Probably two different 3 'regions regulate the transcript frequency via different stabilities and thus also the amount of protein. The experiments described also provided the complete cDNA sequence (complete CDS 119-910) for Itm2A from porcine (SEQ ID NO 4 + SEQ ID NO 5).

The expression of the Itm2A / S129 target under ischemic conditions was examined according to the general experimental protocol given above. It was shown that the target S129 in BMEC under ischemia is greatly reduced in expression. This suggests that Itm2A is involved in diseases that are associated with ischemic conditions, such as stroke, heart attack and tumor-associated conditions, such as those that occur with a glioblastoma. The expression pattern found proves on the one hand the usability of the target as a diagnostic marker for these diseases, and on the other hand the therapeutic usability of the target for the causal treatment of the above-mentioned diseases. The expression pattern of Itm2A in BMEC under ischemia compared to a control set up as 100% is shown in FIG. 1b. BMEC were once under ischemia conditions as described in the method section ("Ischemia") and cultivated once under normal conditions ("control"). The expression of targets in both samples was then measured relative to the 18S rRNA. The value obtained was set as 100% for the control and the ischemia sample related to it.

In order to obtain indications of a possible role of Itm2A on the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH. The result is shown in Figure 2. These data show a rapid decrease in expression before Itm2A, as is also described for known BBB markers. However, a household gene like GAPDH shows no regulation.

The data clearly indicate a function of Itm2A at the BHS. Taking into account the role of the protein in the differentiation in chondrocytes and T cells, it can be concluded that Itm2A is also responsible for the special differentiation state of endothelial cells at the BBB. Since Itm2A has been shown to be located in certain states of cells in the plasma membrane, it appears to be a receptor here by possibly forming homo- or hetero-multimers. The extracellular part of such a receptor would bind secreted molecules or surface molecules of other cells, the intracellular part of the receptor complex could in such a model - e.g. by changing the conformation due to the binding - forwarding signals that trigger a response of the cell within signal cascades and thus change its properties.

Itm2A was first developed in a differential screen of a cDNA bank from condyles (condyle) from mouse by Delersnijder et al. (1996) found. The encoded protein consists of 263 amino acids and is an integral membrane protein of the type II. It has a potential glycosylation site and a possible leucine zipper. The gene, which consists of six exons, is most strongly expressed in bone-forming tissues and is a marker for the differentiation of cartilage / bone. Itm2A is a member of a new gene family consisting of three members. The individual members of the family are highly conserved between humans and mice. Conservation among the individual members is only approx. 40%, with the C-terminus in particular being conserved, but not the N-terminus. The leucine zipper motif can only be found at Itm2A, otherwise the family proteins do not contain any known sequence motifs.

Example 2: Identification of S231

The subtractive clone S231 showed a> 5 times stronger signal in the differential screen with the forward probe compared to the reverse probe and was therefore selected for sequencing. The sequence of clone S231 is given as SEQ ID NO: 6. In BLAST homology searches, sequence S231 showed the highest homology to EMP1.

First, as described, an expression pattern for S231 with the primers S231.1 (5 'CCA TAA CTC TTT CAC GCA ACT G 3' = SEQ ID NO 7) and Ξ231.1R (5 'ACA ACA GAG GAG TTG GCT GTT T 3' ' = SEQ ID NO 8), GAPDH was used as a control (see FIG. 3).

This semi-quantitative expression pattern shows that S231 is more strongly expressed in BMEC than in AOEC and thus confirms the result of the differential hybridization. In addition, the expression in BMEC is also significantly stronger than in Cortex (brain), which is an indication of the specificity for BMEC in the brain. A strong expression can only be seen in the heart, but only weakly in the lungs, colon or brain, although strong expression for these tissues is described in the literature (brain only for rats). This raises the question of whether S231 is really porcine EMP1 or is another member of this gene family.

In order to clarify this, the cDNA bank (λTriplEx2) from BMEC was screened with S231 as a probe (radioactively marked, standard method). Several clones were isolated, of which the two largest clones were each 5 'sequenced. Both sequences again showed the greatest homologies to EMP1, the overlaps each being in the 3 'non-coding area.

In order to investigate whether S231 is actually pork EMP1, the primers hsEMPl.sl (5 'GGT ATT GCT GGC TGG TAT CTT T 3' = SEQ ID NO 9) and hsEMPl.asl (5 'ATG TAG GAA TAG CCG TGG TGA T 3 '= SEQ ID NO 10), which were derived from the coding region of human EMP1, carried out an RT-PCR with BMEC. The product obtained (ssEMPl) was cloned and sequenced. The primer ssEMPl.1 (5 'GGT CTT TGT GTT CCA GCT CTT C 3' = SEQ ID NO 11) was derived from this sequence. A second primer ssEMPl. IR (5 'TTC TCA GGA CCA GAT AGA GAA CG 3' = SEQ ID NO 12) was derived from a section of absolute agreement between the coding sequence of the human EMP1 and the EST F23116 from pigs. With these two primers ssEMPl .1 / ssEMPl. An expression pattern was created as described above (see FIG

⁴ > ^•

The expression patterns with the primers from clone S231 and from ssEMPl are similar, but in no way identical. It is therefore postulated that S231 is another member of the pmp-22 / emp / mp20 gene family. Both expression patterns have now been verified by Northern blot analyzes, the clone S231 (FIG. 5A) or the PCR product hsEMPl. sl / hsEMPl. asl (EMP1) (Fig. 5B) was used (see Fig. 5).

The specificity for BMEC becomes even clearer in the Northern blot. This expression in BMEC and thus at the BHS has not yet been described.

What is striking is the stronger expression in the Northern blot in the plexus (however, the 2-3-fold amount of RNA was also applied here) and colon, whereas the expression by RT-PCR was stronger in the heart. Furthermore, the ratio of the two transcripts in BMEC is significantly different, depending on the probe used. With S231 the ratio of larger to smaller transcript is approximately the same, whereas with EMP1 as a probe the smaller transcript appears significantly stronger.

In comparison to the expression data of EMP1 in the literature, it is noticeable that S231 from pigs has different transcript sizes than EMP1 from humans and mice and that the expression pattern continues to some extent. deviates significantly from the literature data on EMP1 from different species. These deviations at the transcript level show that the clone S231 described here does not represent EMPl, but is another member of this gene family as S231. There may be only one EMPL gene in humans that is regulated by two promoters, and in pigs this task is performed by two separate genes - EMPl and S231.

In order to obtain the complete coding region of S231 from porcine, the cDNA library from BMEC was screened in pTriplEx2 with EMPl as a ClonCapture probe. Several positive clones were isolated, which encode the complete Region included. This has now been sequenced and the protein sequence derived therefrom (SEQ ID NO 13 and SEQ ID NO 14).

The identity of S231 from pig to human EMPl is only 78% at the amino acid level, and to mouse it is 76%. This further strengthens the thesis that S231 does not represent EMPl, since proteins between humans and pigs are normally 85-95% identical (see FIG. 7).

In order to obtain indications of a possible role of S231 in the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH as described in Example 1 (cf. FIG. 8). These data show a rapid decrease in the expression of S231, as is also described for known BBB markers. However, a household gene like GAPDH shows no regulation.

Western blot analysis for S231 performed according to the above procedure confirms the results obtained at the RNA level. A strong expression of the protein can be seen in BMEC, but not in AOEC. Furthermore, the Western blot shows that S231 occurs mainly in the membrane fraction. Expression decreases in cultured BMEC, however, a protein with a lower molecular weight can be increasingly detected. It is possible that these are two different homodimers or heterodimers of S231. The Western blot is shown in Figure 6.

The data clearly indicate a function of S231 at the BHS. Taking into account the described role of the protein in the differentiation of other cell types, it can be concluded that S231 is responsible for the special differentiation state of endothelial cells at the BBB and may represent a cell adhesion molecule or a channel (membrane domain most conserved). Example 3: Identification of S012

The subtractive clone S012 showed a> 5 times stronger signal in the differential screen with the forward probe compared to the reverse probe and thus selected for sequencing. The sequence of clone S012 is listed in SEQ ID NO 15. Using this sequence, S012 was clearly assigned to the human hypothetical protein FLJ13448.

First, as described above, an expression pattern for S012 with the primers S012.sl (5 'GTA TCG GGA GTG GAG GAT TAC A 3' = SEQ ID NO 16) and S012.asl (5 'CCC GAG GTA TAT TTG TTT CTG G 3 '= SEQ ID NO 17), GAPDH was used as a control (expression pattern not shown).

This semi-quantitative expression pattern shows that S012 is more strongly expressed in BMEC than in AOEC and thus confirms the result of the differential hybridization. In addition, expression in BMEC is also significantly stronger than in cortex (brain), which is an indication of the specificity for BMEC in the brain. A strong expression can only be seen in the heart. The full-length cDNA of porcine S012 / FLJ13448 was obtained by overlapping 5 'and 3' RACE-PCR and is shown together with the protein sequence in SEQ ID NO 18 and SEQ ID NO 19.

The expression pattern was verified by Northern blot analysis; the full-length clone FLJ13448 / S012 (SEQ ID NO 18) was used as the probe (see FIG. 9).

The specificity for BMEC becomes even clearer in the Northern blot. This expression in BMEC and thus at the BHS has not yet been described. S012 is homologous to the human hypothetical protein FLJ13448 and the corresponding homologue from mouse (XM_129724). A homology comparison of human, murine and procine FLJ13448 / S012 is shown in FIG. 10. The peptides that serve as signal peptides and are split off are each printed in italics.

The low conservation of the N-terminal 60 amino acids and the high homology of the C-terminus are striking. The N-terminus is probably a signal peptide that is responsible for the correct localization of the protein in the cell. Bioinformatics studies show mitochondrial localization of the protein in the cell. The function of the protein can be assigned to the highly conserved C-terminus.

In order to obtain indications of a possible role of FLJ13448 / S012 in the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH as described in Example 1 (cf. FIG. 11).

These data clearly indicate a role of FLJ13448 / S012 at the BHS. The strong decrease in expression in cultured BMEC suggests that FLJ13448 / S012 is related to the differentiation state of the cells.

Example: Identification of NSE2

The sample material was prepared as described above under the section “Identification of BBB-Specific Proteins by Differential 2D Gel Electrophoresis”.

The differential spot _. 1.1.0.1.10.37 showed the following peptide masses in the MALDI-TOF analysis: 861,499; 878.47; 975.50; 1,056.61; 1,132.53; 1,198.71; 1,216.71; 1,227.53; 1,347.69; 1,430.76; 1,438.69; 1,516.71; 1,623.79; 1,790.87; 1,796.81; 1,935.93; 1,954.05; 2,081.02; 2,231.07; 2,375.08; 2,577.09; 2613.1.

Spot 1.1.0.1.10.37 was identified as NSE2 by the database queries with Profound in the NCBI database. The human NSE2 has a calculated molecular weight of 34.5 kDa and a pI value of 5.4, both of which agree very well with the observed position of the spot 1.1.0.1.10.37 in the 2D gel. The fat-highlighted, underlined peptide masses could be assigned as identical to the human sequence. In Fig. 12 the coverage of the peptide masses on the human protein sequence is shown.

First, as described, an expression pattern for NSE2 with the primers ssNSE2.sl (5 'CGC GTG GTG AAT GAT CTG TA 3' = SEQ ID NO 20) and ssNSE2.asl (5 'CTC CAT GAT CAG GTC CTC CAG 3' = SEQ ID NO 21), GAPDH was used as a control.

This semi-quantitative expression pattern shows that the expression of NSE2 is highest in the heart, followed by BMEC and Cortex (not shown). This result was confirmed by Northern blot analysis (see FIG. 13). The partial cDNA sequence of porcine NSE2 was used for hybridization (SEQ ID NO 22 and SEQ ID NO 23) (partial CDS 1-192, coded C-terminus), which was obtained by 3'RACE-PCR.

In order to obtain indications of a possible role of NSE2 in the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH as described in Example 1. The result is shown in FIG. 14. These data show a rapid decrease in the expression of NSE2 and thus indicate a function of NSE2 at the BHS. FIG. 15 shows a homology comparison of human NSE2 and NSE1.

Potential phosphorilization sites are shown in a bright font. A possible tyrosine kinase domain (ProSite Pattern Match PS00109) is underlined, the active remainder being shown in bold. Figure 16 shows the distribution of PEST domains in NSE2. PEST sequences are regions rich in Pro, Glu, Ser and Thr in proteins that are responsible for a short half-life of such proteins in the cell by controlling the ubiquitination of these proteins.

Phosphorilation of certain Ser or Thr residues in the PEST regions (light gray) are important for the detection of processing by the ubiquitin proteasome pathway.

Position 81-163 in human NSE2 shows homologies to the NLP / P60 family (pfam domain 00877.4), which was found in several lipoproteins but was not assigned a function.

This target was also examined under ischemic conditions. It shows that the expression of NSE2 is reduced in BMEC under ischemia (cf. FIG. 17). This suggests NSE2 involvement in diseases associated with ischemic conditions such as stroke, heart attack and tumor associated conditions such as glioblastoma. The expression pattern of NSE2 can thus be used as a diagnostic marker for such diseases. A causal therapy can also start with the modulation of the expression of NSE2.

Example 5: Identification of DRG-1

The sample material was prepared as described above under the section “Identification of BBB-Specific Proteins by Differential 2D Gel Electrophoresis”. The differential spot 1.1.0.1.11.12 showed the following peptide masses in the MALDI-TOF analysis: 789.45; 880.47; 890.50; 948.49; 1,204.68; 1,217.64; 1,289.58; 1,428.70; 1,517.79; 1,573.73; 1,753.91; 2,017.08.

The database queries with Profound in the NCBI database identified Spot 1.1.0.1.11.12 as a hypothetical protein with the accession number CAB66619. In other database entries, the identical protein is also referred to as dopamine-responsive protein DRG-1, as LYST-interacting protein LIP5 and as HSPC228. The hypothetical protein

CAB66619 / DRG-1 has a calculated molecular weight of 33.8 kDa and a pI value of 6.1, both of which correspond very well with the observed position of the spot 1.1.0.1.11.12 in the 2D gel. The fat-labeled peptide masses could be assigned as identical to the human sequence. 18 shows the coverage of the peptide masses on the human protein sequence.

A homology comparison between human (CAB66619) and mouse (XP_125508) shows very large homologies, especially in the area of AS 1-180. Bioinformatic approaches show a transmembrane domain and suggest that the N-terminus is located intracellularly. The intracellular domain shows a conserved phosphorilization site, extracellularly a glycosylation site is predicted in the human sequence (cf. FIG. 19).

First, as described, an expression pattern for DRG-1 with the CAB66619.sl (5 'CGA GAC CCT GTG GTG GCT TAT TAC 3' = SEQ ID NO 24) and CAB66619.asl (5 'CTG GTG TAT TAG CTG GAG CGT GTG 3 '= SEQ ID NO 25), GAPDH was used as control. This semi-quantitative expression pattern (FIG. 20; which was confirmed by Northern blot analysis) shows that DRG-1 from pigs is less expressed in BMEC than in AOEC and thus contradicts the result of the 2D gel. In general, DRG-1 is expressed to varying degrees but is ubiquitously expressed. The difference found in the 2D gel must therefore be due to a specific post-translational modification of DRG-1 in BMEC. Such a difference can occur, for example, due to the predicted phosphorilation site. Cell-specific phosphorilations can thus determine the activity of the protein.

In order to obtain indications of a possible role of DRG-1 in the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH as described in Example 1 (cf. FIG. 21). These data show a clear decrease in the expression of DRG-1 and thus indicate a function of DRG-1 at the BHS.

SEQ ID NO 26 + 27 show the partial cDNA sequence of DRG-1 from porcine (CDS1-585, internal section).

Example 6: Identification of TKA-1

The differential spot 1.1.0.1.6.30 resulted in the following peptide masses in the MALDI-TOF analysis: 776.44; 847.47; 900.50; 916.46; 976.52; 1,048.58; 1,085.61; 1,127.66; 1,137.55; 1,167.67; 1,180.68; 1,212.69; 1,234.69; 1,291.67; 1,301.67; 1,303.69; 1,338.72; 1350.70 1370.65; 1,419.70; 1,423.77; 1,434.79; 1,440.79; 1,456.76; 1466.76 1467.71; 1,483.77; 1,547.78; 1,558.85; 1,665.90; 1,714.96; 1,716.90 1740, 80; 1762, 90; 1838, 92; 1897, 99; 2025, 11; 2054, 06; 2234, 15; 2243, 20; 2244, 18th

Spot 1.1.0.1.6.30 was identified as TKA-1 by the database queries with MSFIT in the NCBI database. The fat-marked, underlined peptide masses could be assigned identically to the human sequence. 22 shows the coverage of the peptide masses on the human protein sequence.

The database contains 3 isoforms of TKA-1, which have the following calculated masses and pl values: CAA90511 with 49.3 kDa / pl 6.7, BAA33216 with 37.4 kDa / pl 7.9,

AAB53042 with 36.2 kDa / pl 8.2. The position in the 2D gel clearly speaks against the large isoform. Thus, the BAA33216 isoform was found experimentally unambiguously, since the peptide DGSAWKQΣXPFQ (italics in FIG. 22) is missing in the protein with the accession number AAB53042, but was partially (bold) detected within a trypsin fragment in the MALDI analysis ,

The alignment of TKA-1 between humans, mice and rats shows a very high level of conservation. TKA-1 has two PDZ domains that mediate protein-protein interactions. There are several potential phosphorilization sites in these PDZ domains, which may regulate the interactions with other proteins. A potential N-glycosylation site is also preserved.

First, as described, an expression pattern for TKA-1 with the primers ssSLC9A3R2. sl (5 'AAA AGG CCC CCA GGG TTA CG 3' = SEQ ID NO 28) and ssSLC9A3R2. asl (5 'GGA GTG GGC AGC AGG TGA GC 3' = SEQ ID NO 29), GAPDH was used as a control. The expression pattern was verified by Northern blot analysis, the 550 bp PCR product ssTKA-l.ctg between the two primers ssTKA-lctg was used as the probe. sl (5 'TTA ACC TGC ACA GCG ACA AGT 3' = SEQ ID NO 30) and ssTKA-lctg. asl (5 'TTG CTG AAG ATC TCA CGC TTC 3' = SEQ ID NO 31).

The Northern blot (FIG. 23) shows that TKA-1 expresses the most in BMEC, t is and that three different transcripts occur in BMEC. The expression is comparatively strong in the lungs, but the small transcript is completely missing here. So far, no connection between TKA-1 and the BBB and also not with endothelial cells has been described in the literature.

In order to obtain indications of a possible role of TKA-1 in the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH as described in Example 1 (cf. FIG. 24). These data show a clear decrease in the expression of TKA-1 and thus indicate a function of TKA-1 at the BHS.

The target TKA-1 was also examined with regard to its expression under ischemia in accordance with the instructions given above. It is shown that this target in BMEC is strongly reduced in expression under ischemia. This suggests a functional involvement of TKA-1 in diseases that are associated with ischemic conditions, such as stroke, heart attack and tumor-associated conditions, for example in glioblastoma. The study of the expression of TKA-1 can therefore be used as a diagnostic marker for such diseases. The TKA-1 target is also a suitable starting point for causal therapies against the abovementioned diseases.

The expression pattern of TKA-1 in BMEC under ischemia compared to a control is shown in Figure 25. BMEC were cultivated as described in the method section once under ischemia conditions ("ischemia") and once under normal conditions ("control"). The expression of targets in both samples was then measured relative to the 18S rRNA. The value obtained was set as 100% for the control and the ischemia sample related to it.

Western blot analysis for TKA-1 confirms the results obtained at the RNA level. A strong expression of the protein can be seen in BMEC, but hardly in AOEC. Furthermore, the Western blot shows that TKA-1 is mainly membrane-associated and occurs in the cell nucleus. In cultured BMEC, expression decreases very quickly and is no longer detectable in the first passage. Western blot analysis of TKA-1 is shown in Figure 26.

SEQ ID NO 32 and SEQ ID NO 33 show the partial cDNA

Sequence of porcine TKA-1 (partial CDS 1-741, encoding the C-terminus).

Example 7: Identification of S064 / ARL8

The subtractive clone S064 showed a> 5 times stronger signal in the differential screen with the forward probe compared to the reverse probe and was therefore selected for sequencing. The sequence of clone S064 is listed as SEQ ID NO 35. Based on this sequence, S064 could not be assigned to any known gene. BLAST searches revealed a significant homology to the DKFZ cDNA clone p43401317, which apparently does not contain a coding region.

To identify the corresponding protein, a cDNA library from porcine BMEC was screened with the subtractive clone S064. Two independent clones were identified. The sequence of the longest clone S064.3 is as SEQ ID NO 36 listed. This sequence could also not be assigned to any known gene by BLAST searches.

However, the sequence of clone S064.3 could be localized in the 10pl2 region by homology comparisons in the human genome. The next gene in the same orientation on this locus is ADP-ribosylation-like factor 8 (ARL8). A link-PCR was carried out to check whether S064 represents a new 3 'end of ARL8. For this purpose, the primers hsARLδ.sl (5 'TAA TGC AGG GAA AAC CAC CAT TCT 3', SEQ ID NO 37) and S064.3R (5 'AAC CAA GAG ACA TGT TGG CAC T 3', SEQ ID NO 38) were also used RNA from BMEC used in a OneStep RT-PCR. To check the product specificity, the product was diluted 1: 1000 from the OneStep RT-PCR and into a nested PCR with the primers hsARL8. s2 (5 'ATA GCA TTG ACA GGG AAC GAC T 3', SEQ ID NO 39) and S064.GSP2 (5 'CTG CTA GAT TCA AGT CAT CAT GC 3', SEQ ID NO 40). The product obtained was cloned and sequenced. The sequence obtained clearly confirmed that the subtractive clone S064 represents the ARL8 gene.

The complete coding cDNA sequence of ARL8 was determined using OneStep RT-PCR with RNA from BMEC and the primers S064cds.sl (5 'CTC GTG ATG GGG CTG ATC TTC 3', SEQ ID NO 41) and S064cds.asl (5 'ATC TCA CAC CAA TCC GGG AGG T 3', SEQ ID NO 42) received. The coding sequence ARL8 from porcine is given as SEQ ID NO 43, the protein encoded thereby is shown in SEQ ID NO 44. The ARL8 protein is 100% identical to human and mouse ARL8. This high degree of conservation speaks for the important role of this protein. The cDNA sequence of ARL8 (porcine) has 95% and 92% homology in the coding region to the corresponding sequence from human or mouse. As described, an expression pattern for S064 with the primers S064.sl (5 'AAG CCT GAA GCT TGA TGG ATA A 3', SEQ ID NO 45) and S064.asl (5 'CAA TTA CAG CTT TGC TCC TGT G 3' , SEQ ID NO 46), 18Ξ rRNA was used as reference. The two primers S064cds. sl / asl were derived from the human sequence due to the high homology between humans and pigs (eg the product of the link-PCR). In addition to the general criteria of the primer design, care was taken to ensure that the two primers flank the complete coding sequence: Primer S064cds.sl contains the ATG start codon in position 7-9 and position 22 in primer S064cds.asl provides the first base of the stop codon. The expression pattern is shown in FIG.

The expression pattern was repeated with another pair of primers from the coding region of ARL8: ARLδcds.sl (5 'ATA GCA TTG ACA GGG AAC GAC T 3', SEQ ID NO 47) and ARLδcds.asl (5 'GAA CTG AGG GTG AGG TAT TTG G 3 ', SEQ ID NO 48). The expression pattern is shown in Figure 28.

Furthermore, the expression pattern was verified by Northern blot analysis, the clone S064 was used here. The result is shown in FIG. 29.

All three experiments show the very high specificity of ARL8 for BMEC and thus for the blood-brain barrier. This expression in BMEC or at the BHS has not yet been described. This high specificity indicates a very important role of ARL8 at the BHS.

In order to obtain information on the type of function of ARL8 on the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or GAPDH. The result is shown in FIG. 30. The data show a rapid decrease in the expression of ARL8 in cultivated BMEC and thus clearly indicate that ARL8 actually functions at the BHS.

ARL8 belongs to the RAS superfamily of regulatory GTPases. These are involved in a variety of processes, such as Cell growth, signal transduction, organization of the cytoskeleton and regulation of membrane trafficking (exo- or endocytosis). ARL8 was first developed by Sebald et al. Described in 2003, but these were unable to show expression in the adult brain. The present example shows for the first time the actual expression of ARL8 at the BHS. This confirms the high BBB specificity of this protein. This means that ARL8 is responsible for the special state of differentiation of endothelial cells at the BBB and thus contributes to the functionality of the BBB.

Example 8: Identification of 5G9 / PNOV1

The subtractive clone 5G9 showed a> 5 times stronger signal in the differential screen with the forward probe compared to the reverse probe and was therefore selected for sequencing. The sequence of clone 5G9 is listed as SEQ ID NO 49.

Using this sequence, 5G9 could be assigned to a human transcript (No. BC039195, NCBI database) which codes for a new protein HSNOV1 (AAH39195). In this database entry, which describes an mRNA molecule, the open reading frame and the resulting hypothetical protein are given as annotation. This is not experimental data, but computer-based predictions. The derived protein showed no similarities to known proteins and was therefore called novel protein.

An expression pattern for 5G9 with the primers 5G9.1 (5 'TGT ATA TGT GGG ACA GCC ATC A 3', SEQ ID NO 50) and 5G9.1R (5 'GTC CGA GCA GGA TAT ACT TCC A 3', SEQ ID NO

51), 18S rRNA was used as reference. The primer pair for determining the expression pattern was derived according to the general rules: melting temperature of the primers from 55-75 ° C; approximately the same melting temperature of the two primers; 18-26 base length; optimal 40-60% GC content; Avoiding hairpins loops; Avoidance of homo- and heterodimer formation; Product size 100-300 bp. The expression pattern is shown in Figure 31.

The expression pattern shows that 5G9 is mainly formed on the BBB, in the colon and in the kidney. Expression in the brain appears to be specific for BMEC. This expression in BMEC or at the BHS has not yet been described.

In order to identify the corresponding protein from pig, the complete cDNA PNOV1 from pig was isolated from the sequence of clone 5G9 by 5 'and 3' RACE-PCR (SEQ ID NO 52). This transcript codes from position 480-1466 for a protein with SEQ ID NO 53. The homology comparison between HSNOV1 and PNOV1 is shown in FIG.

The homology between HSNOV1 and PNOV1 is 94%. It is striking, however, that PNOV1 is N-terminally shortened by 47 amino acids compared to HSNOV1. In HSNOV1, this sequence may represent a signal sequence that will later be split off.

The HSNOV1 protein shows no significant homologies to other known proteins. Bioinformatic analyzes show 8 potential transmembrane domains (see FIG. 33).

In addition, several domains (eg InterPro domain ipr002657) were found that indicate a function as a transporter. These data suggest that PNOV1 / HSNOV1 is a new transporter at the BBB that also occurs in the colon and kidney, two tissues with high transport activities for many substances.

Example 9: Identification of 5E7 / TSC-22

The subtractive clone 5E7 showed a> 5 times stronger signal in the differential screen with the forward probe compared to the reverse probe and was therefore selected for sequencing. The sequence of clone 5E7 is listed as SEQ ID NO 54. Using this sequence, 5E7 was clearly identified as a transforming growth factor beta-stimulated protein TSC-22.

Clone 5E7 represents the 3 'end of the TSC-22 transcript. In order to obtain the complete porcine cDNA, a 5' RACE-PCR was carried out. The product of this PCR was cloned and sequenced. The complete cDNA sequence of porcine TSC-22 is listed as SEQ ID NO 55, the coding region here is from position 243-677. The corresponding protein is listed as SEQ ID NO 56. The porcine protein is 100% identical to the already known human protein TSC-22, which speaks for the special importance of this protein.

As described, an expression pattern for 5E7 was created by Northern blot analysis, the subtractive clone 5E7 serving as the probe (cf. FIG. 34).

The experiment shows the stronger expression of TSC-22 in BMEC compared to the whole brain and thus shows the specificity for the blood-brain barrier. This expression in BMEC or at the BHS has not yet been described. In order to obtain indications of a possible role of TSC-22 in the BBB, the expression pattern in cultured BMEC was examined in comparison to known BBB markers or 18S rRNA. The primers 5E7.1 (5 'AAG AGG TGT GGC TTG TCT TTT A 3', SEQ ID NO 57) and 5E7 were used for the quantitative PCR. IR (5 'TTT TTC AAA GTA TTC AAC CAG CTC 3', SEQ ID NO 58) used. The result is shown in FIG. 35.

The data show a rapid decrease in the expression of TSC-22 in cultivated BMEC and thus clearly indicate a role of TSC-22 in the BHS. The strong decrease in expression in cultured BMEC suggests that TSC-22 is related to the differentiation state of the cells.

The expression of TSC-22 in BMEC under ischemia was examined in the same way. The result is shown in FIG. 36.

The target TSC-22 is greatly reduced in expression in BMEC under ischemia. This demonstrates a possible functional role for TSC-22 in diseases associated with ischemic conditions. These include stroke, heart attack (TSC-22 is also strongly expressed in the heart, see Fig. 34) and the conditions in a tumor, e.g. in glioblastoma. The study of the expression of TSC-22 can therefore also be used as a diagnostic marker for these diseases. Based on these findings, therapy concepts for diseases associated with a dysfunction of the BBB can be developed.

TSC-22 belongs to the class of leucine zipper transcription factors (Kester et al., 1999). It is involved in the signal transduction of TGF-beta and others (Kawamata et al., 1998) and thus plays a role in cell growth and differentiation. It follows that TSC-22 is also responsible for the differentiation state of BMEC.

Literature:

Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J. (1990): "Basic local alignment search tool", J. Mol. Biol. 215, 403-410.

Birnboim, H.C. and Doly, J. (1979): "A rapid alkaline extraction procedure for screening recombinant plasmid DNA", Nucl. Acids Res. 7: 1513-1522.

Cserzo et al. (1997): "Prediction of transmembrane alpha- helices in procariotic membrane proteins: the dense alignment surface method", Prot. Eng. 10: 673-676.

Deleersnijder, W., Hong, G., Cortvrindt, R., Poirier, C, Tylzanowski, P., Pittois, K., Van Marck, E., and Merregaert, J. (1996): “Isolation of markers for chondro -osteogenic differentiation using cDNA library subtraction. Molecular cloning and characterization of a gene belonging to a novel multigene family of integral membrane proteins ", J. Biol. Chem. 271, 19475-19482.

Kawamata, H., Nakashiro, K., Uchida, D., Hino, S., Omotehara, F., Yoshida, H., and Sato M. (1998): “Induction of TSC-22 by treatment with a new anti -cancer drug, vesnarinone, in a human salivary gland cancer cell ", Brit. J. Cancer 77: 71-78.

Kester, HA, Blanchelot, C, den Hertog, J., van der Saag, PT, and van der Burg, B. (1999): “Transforming growth factor-ß-stimulated clone-22 is a member of a family of leucine zipper proteins that can homo- and heterodimerize and has transcriptional repressor activity ", J. Biol. Chem. 274: 27439-27447. Li, JY, Boado, RJ, and Pardridge, WM (2001): "Blood-brain barrier genomics", J. Cereb. Blood Flow Metabol. 21, 61-68.

Marvin, K.W., Fujimoto, W., Jetten, A.M. (1995): "Identification and characterization of a novel squamous cell-associated gene related to PMP22", J. Biol. Chem. 270, 28910-28916.

Pearson, W.R. and Lipman, D.J. (1988): "Improved Tools for Biological Sequence Analysis", PNAS 85, 2444-2448.

Sambrook, J., Fritsch, E.F., and Maniatis, T., in Molecular Clonin: A Laboratory Manual. Cold Spring Harbor Laboratory Press, NY, Vol. 1, 2, 3 (1989).

Sebald, E., Krueger, R., King, L.M., Cohn, D.H., and Krakow,

D. (2003): "Isolation of a new member of the ADP-ribosylation like factor gene family, ARL8, from a cartilage cDNA library", Gene 311: 147-151.

Shevchenko A., Sunyaev S., Loboda A., Shevchenko A., Bork P., Ens W. and Standing K.G. (2001); "Charting the Proteomes of Organisms with Unsequenced Genomes by MALDI-Quadrupole Time- of-Flight Mass Spectrometry and BLAST Ho ology Searching"; Anal. Chem. 73: 1917-1926.

Claims

1. A method for identifying the presence of a BBB-specific protein or fragment thereof in brain capillary endothelial cells, by means of the fact that

a) brain capillary endothelial cells freshly isolated from the brain are pre-cleaned in the usual way by enzymatic digestion,

b) the digestion obtained in stage a) is treated with a lysis buffer which essentially destroys existing erythrocytes and a-poptotic cells and maintains at least 70% of the brain capillary endothelial cells in vital form,

c) optionally further purifying the product obtained in stage b),

d) produces a subtractive cDNA library from the brain capillary endothelial cells and a subtraction tissue,

e) performing a cDNA subtraction using one or more differential hybridizations,

f) clones from the subtractive cDNA library verified by differential hybridization with regard to their respective expression,

g) the cDNA sequence is supplemented to the BBB-specific clones from the subtractive cDNA bank and

h) the expression pattern of the examined clones between fresh and cultivated brain capillary endothelial cells compares and thus identifies the presence of BBB-specific proteins or fragments thereof.

2. The method according to claim 1, characterized in that the lysis buffer in stage b) has the following composition:

Na ⁺ 30.0 mM to 60.0 mM

K ⁺ 5.0mM to 7.5mM

NH ₄ 80.0 mM to 100.0 mM

Mg ²⁺ 6.0 mM to 9.0 mM

Cl ^" 125.0 mM to 175.0 mM

HC0 ₃ ^" 4.5 mM to 6.5 mM

H ₂ P0 ₄ ^" 0.5 mM to 2.5 mM

S0 ₄ ^{2 "} 0.3 mM to 0.6 mM

HP0 ₄ ^{2 "} 0.4 mM to 0.7 mM

Glucose 1.5mM to 3.0mM

3. The method according to claim 2, characterized in that the lysis buffer has the following composition: NaCl 30mM to 50mM

KC1 4.5mM to 5.5mM

NH ₄ C1 80mM to 100mM

CaCl ₂ 1.0 mM to 2.0 mM

MgCl ₂ 0.6 mM to 0.8 mM

MgS0 ₄ 0.3 mM to 0.6 mM

NaHC0 ₃ 4.5 mM to 6.5 mM

NaH ₂ P0 ₄ 0.2 mM to 0.45 mM

Na ₂ HP0 ₄ 0.4 mM to 0.65 mM

KH ₂ P0 ₄ 0.1 mM to 0.15 mM

Glucose 1.5mM to 3.0mM

4. The method according to any one of claims 1 to 3, characterized in that the subtraction tissue in stage f) is aortic endothelial cells.

5. The method according to any one of claims 1 to 4, characterized in that the complete cDNA sequence in step i) is created by screening, cDNA banks and RACE-PCR.

6. The method according to any one of claims 1 to 5, characterized in that the brain capillary endothelial cells come from humans or pigs.

7. Protein with BBB specificity or a fragment thereof, obtainable by a method according to any one of claims 1 to 6.

8. Protein according to claim 7, characterized in that it is a sequence selected from SEQ ID NO: 5, SEQ ID NO: 14, SEQ ID N0: 19, or SEQ ID NO: 53.

9. A method for identifying the presence of a BBB-specific protein or fragment thereof in brain capillary endothelial cells, by means of the fact that

c) optionally further purifying the product obtained in stage b),

d) solubilizing the product obtained in step c) in a suitable buffer,

e) carries out isoelectric focusing,

f) separating the samples from the isoelectric focusing in the second dimension by molecular weight,

g) differential spots identified and isolated,

h) performs a mass spectrometric analysis on the isolate from g), and

i) of which carries out an evaluation by means of targeted database analysis.

0. The method according to claim 9, characterized in that the lysis buffer in stage b) has the following composition:

Na ⁺ 30.0 mM to 60.0 mM

K ⁺ 5.0 mM to 7.5 mM

NH ₄ ⁺ 80.0 mM to 100.0 mM

Ca, ² 2+ 1.0 mM to 2.0 mM

Mg. ² 2+ 6.0 mM to 9.0 mM

Cl ^" 125.0 mM to 175.0 mM

HC0 ₃ ^~ 4.5 mM to 6.5 mM

H ₂ P0 ₄ ^" 0.5 mM to 2.5 mM

S0 ₄ ^{2 "} 0.3 mM to 0.6 mM

HPO ₄ ^{2 "} 0.4 mM to 0.7 mM

Glucose 1.5mM to 3.0mM

11. The method according to claim 10, characterized in that the lysis buffer has the following composition:

NaCl 30mM to 50mM

KC1 4.5 mM to 5.5 mM

NH ₄ CI 80mM to 100mM CaCl ₂ 1.0 mM to 2.0 mM

MgCl ₂ 0.6 mM to 0.8 mM

MgS0 ₄ 0.3 mM to 0.6 mM

NaHC0 ₃ 4.5 mM to 6.5 mM

NaH ₂ P0 ₄ 0.2 mM to 0.45 mM

Na ₂ HP0 0.4 mM to 0.65 mM

KH ₂ P0 ₄ 0.1 mM to 0.15 mM

Glucose 1.5mM to 3.0mM

12. Protein with BBB specificity or a fragment thereof, obtainable by a method according to any one of claims 9 to 11.

13. Protein according to claim 12, characterized in that it has a sequence selected from SEQ ID NO 23, SEQ ID N0: 27, SEQ ID NO: 33.

14. Use of a protein according to any one of claims 7 to 8 or 12 to 13 for the manufacture of a medicament for the transport of substances through the blood-brain barrier.

15. Use of a protein according to any one of claims 7 to

8 or 12 to 13 for the manufacture of an agent or medicament for the diagnosis or therapy of diseases which are based on a dysfunction of the blood-brain barrier.

16. Means for diagnosing diseases based on a dysfunction of the blood-brain barrier, characterized in that it comprises a protein according to one of claims 7 to 8 or 12 to 13.

17. Means for the therapy of diseases which are based on a dysfunction of the blood-brain barrier, characterized in that it comprises a protein according to one of claims 7 to 8 or 12 to 13.

18. Use of one or more DNA sequence (s) selected from SEQ ID N0: 4, SEQ ID N0: 6, SEQ ID NO: 15, SEQ ID NO: 22, SEQ ID NO: 23, SEQ ID NO: 26, SEQ ID NO: 32, SEQ ID NO: 35, SEQ ID NO: 36, SEQ ID NO: 43, SEQ ID NO: 49, SEQ ID NO: 52, SEQ ID NO: 54, SEQ ID NO: 55 a means of diagnosing diseases associated with ischemic conditions.

19. Use according to claim 18 for the diagnosis of stroke, heart attack or tumor-associated conditions.

20. Use according to one of claims 18 or 19, characterized in that the diagnosis is carried out by controlling the expression of the proteins encoded by the said DNA sequences.