Functions of fibroblast growth factor (FGF)-2 and FGF-5 in astroglial differentiation and blood-brain barrier permeability: evidence from mouse mutants.

Multiple evidence suggests that fibroblast growth factors (FGFs), most prominently FGF-2, affect astroglial proliferation, maturation, and transition to a reactive phenotype in vitro, and after exogenous administration, in vivo. Whether this reflects a physiological role of endogenous FGF is unknown. Using FGF-2 and FGF-5 single- and double mutant mice we show now a region-specific reduction of glial fibrillary acidic protein (GFAP), but not of S100 in gray matter astrocytes. FGF-2 is apparently the major regulator of GFAP, because in mice deficient for FGF-2, GFAP is distinctly reduced in cortex and striatum, whereas in FGF-5-/- animals only a reduction in the midbrain tegmentum can be observed. In FGF-2-/-/FGF-5-/- double mutant animals, GFAP-immunoreactivity is reduced in all three brain regions. Cortical astrocytes cultured from FGF-2-/-/FGF-5-/- double mutant mice revealed reduced levels of GFAP, but not S100 as compared with wild-type littermates. This phenotype could be rescued by exogenous FGF-2 but not FGF-5 (10 ng/ml). Electron microscopy revealed reduced levels of intermediate filaments in perivascular astroglial endfeet. This defect was accompanied by enhanced permeability of the blood-brain barrier (BBB), as detected by albumin extravasation. Levels of the tight junction proteins Occludin and ZO-1 were reduced in blood vessels of FGF-2-/-/FGF-5-/- double mutant mice as compared with wild-type littermates. Our data support the notion that endogenous FGF-2 and FGF-5 regulate GFAP expression in a region-specific manner. The observed defect in astroglial differentiation is accompanied by a defect in BBB function arguing for an indirect or direct role of FGFs in the regulation of BBB permeability in vivo.

Holey barrier: claudins and the regulation of brain endothelial permeability.

Endothelial tight junctions (TJs)* are an important functional part of the blood-brain barrier (BBB). In this issue, Nitta et al. (2003) demonstrate that claudin-5, a transmembrane protein of TJs, is a critical determinant of BBB permeability in mice. Unexpectedly, knockout of claudin-5 did not result in a general breakdown of TJs but in a selective increase in paracellular permeability of small molecules. This suggests that the BBB can be manipulated to allow selective diffusion of small molecules and makes claudin-5 a possible target for the development of drugs for this purpose.

Size-selective loosening of the blood-brain barrier in claudin-5-deficient mice.

Tight junctions are well-developed between adjacent endothelial cells of blood vessels in the central nervous system, and play a central role in establishing the blood-brain barrier (BBB). Claudin-5 is a major cell adhesion molecule of tight junctions in brain endothelial cells. To examine its possible involvement in the BBB, claudin-5-deficient mice were generated. In the brains of these mice, the development and morphology of blood vessels were not altered, showing no bleeding or edema. However, tracer experiments and magnetic resonance imaging revealed that in these mice, the BBB against small molecules (<800 D), but not larger molecules, was selectively affected. This unexpected finding (i.e., the size-selective loosening of the BBB) not only provides new insight into the basic molecular physiology of BBB but also opens a new way to deliver potential drugs across the BBB into the central nervous system.

Severe alterations of endothelial and glial cells in the blood-brain barrier of dystrophic mdx mice.

In this study, we investigated the involvement of the blood-brain barrier (BBB) in the brain of the dystrophin-deficient mdx mouse, an experimental model of Duchenne muscular dystrophy (DMD). To this purpose, we used two tight junction markers, the Zonula occludens (ZO-1) and claudin-1 proteins, and a glial marker, the aquaporin-4 (AQP4) protein, whose expression is correlated with BBB differentiation and integrity. Results showed that most of the brain microvessels in mdx mice were lined by altered endothelial cells that showed open tight junctions and were surrounded by swollen glial processes. Moreover, 18% of the perivascular glial endfeet contained electron-dense cellular debris and were enveloped by degenerating microvessels. Western blot showed a 60% reduction in the ZO-1 protein content in mdx mice and a similar reduction in AQP4 content compared with the control brain. ZO-1 immunocytochemistry and claudin-1 immunofluorescence in mdx mice revealed a diffuse staining of microvessels as compared with the control ones, which displayed a banded staining pattern. ZO-1 immunogold electron microscopy showed unlabeled tight junctions and the presence of gold particles scattered in the endothelial cytoplasm in the mdx mice, whereas ZO-1 gold particles were exclusively located at the endothelial tight junctions in the controls. Dual immunofluorescence staining of alpha-actin and ZO-1 revealed colocalization of these proteins. As in ZO-1 staining, the pattern of immunolabeling with anti-alpha-actin antibody was diffuse in the mdx vessels and pointed or banded in the controls. alpha-actin immunogold electron microscopy showed gold particles in the cytoplasms of endothelial cells and pericytes in the mdx mice, whereas alpha-actin gold particles were revealed on the endothelial tight junctions and the cytoskeletal microfilaments of pericytes in the controls. Perivascular glial processes of the mdx mice appeared faintly stained by anti-AQP4 antibody, while in the controls a strong AQP4 labeling of glial processes was detected at light and electron microscope level. The vascular permeability of the mdx brain microvessels was investigated by means of the horseradish peroxidase (HRP). After HRP injection, extensive perivascular areas of marker escape were observed in mdx mice, whereas HRP was exclusively intravascularly localized in the controls. Inflammatory cells, CD4-, CD8-, CD20-, and CD68-positive cells, were not revealed in the perivascular stroma of the mdx brain. These findings indicate that dystrophin deficiency in the mdx brain leads to severe injury of the endothelial and glial cells with disturbance in alpha-actin cytoskeleton, ZO-1, claudin-1, and AQP4 assembly, as well as BBB breakdown. The BBB alterations suggest that changes in vascular permeability are involved in the pathogenesis of the neurological dysfunction associated with DMD.

Different presentations of late-detected phenylketonuria in two brothers with the same R408W/R111X genotype in the PAH gene.

Although the clinical heterogeneity of phenylketonuria (PKU) is well established, some questions about this condition remain. Subjects from the same family who share the same mutations in the phenylalanine hydroxylase (PAH) gene are expected to display similar disease courses, and therefore, when blood phenylalanine (Phe) levels, genotype and dietary treatment are all similar, differences in patient outcomes require additional explanations. The present authors describe two entirely different courses of late-detected PKU in two brothers with the same R408W/R111X genotype in the PAH gene. The older sibling was diagnosed with PKU at the age of 4 years and given treatment. His IQ was 97 at 26 years of age and moderate involvement of periventricular white matter was detected. The younger brother was diagnosed with PKU at the age of 11 months and given treatment. His IQ was < 25 at 22 years of age and severe dysmyelination changes were found by magnetic resonance imaging. The differences in the courses of the disease between these two brothers appear to be related to variations in their blood-brain barriers.

Toll-like receptor 2 participates in mediation of immune response in experimental pneumococcal meningitis.

Heterologous expression of Toll-like receptor (TLR)2 and CD14 in Chinese hamster ovary fibroblasts was reported to confer responsiveness to pneumococcal peptidoglycan. The present study characterized the role of TLR2 in the host immune response and clinical course of pneumococcal meningitis. Pneumococcal infection of mice caused a significant increase in brain TLR2 mRNA expression at both 4 and 24 h postchallenge. Mice with a targeted disruption of the TLR2 gene (TLR2-/-) showed a moderate increase in disease severity, as evidenced by an aggravation of meningitis-induced intracranial complications, a more pronounced reduction in body weight and temperature, and a deterioration of motor impairment. These symptoms were associated with significantly higher cerebellar and blood bacterial titers. Brain expression of the complement inhibitor complement receptor-related protein y was significantly higher in infected TLR2-/- than in wild-type mice, while the expression of the meningitis-relevant inflammatory mediators IL-1beta, TNF-alpha, IL-6, macrophage-inflammatory protein (MIP)-2, inducible NO synthase, and C3 was similar in both genotypes. We first ectopically expressed single candidate receptors in HEK293 cells and then applied peritoneal macrophages from mice lacking TLR2 and/or functional TLR4 for further analysis. Overexpression of TLR2 and TLR4/MD-2 conferred activation of NF-kappaB in response to pneumococcal exposure. However, pneumococci-induced TNF-alpha release from peritoneal macrophages of wild-type and TLR2/functional TLR4/double-deficient mice did not differ. Thus, while TLR2 plays a significant role in vivo, yet undefined pattern recognition receptors contribute to the recognition of and initiation of the host immune defense toward Streptococcus pneumoniae infection.

Overexpression of bradykinin type 2 receptors on glioma cells enhances bradykinin-mediated blood-brain tumor barrier permeability increase.

Variations in the expression levels of bradykinin (BK) type 2 receptors (B2R) in different brain tumors may explain variable increases in BK-mediated blood-brain tumor barrier (BTB) permeability. This study investigated whether elevation of the B2R expression levels on glioma cells enhances BK-mediated BTB permeability increases. Stable transfectants of C6 rat glioma cells overexpressing B2R were established by transfection with recombinant vectors harboring rat B2R cDNA sequence. Elevated B2R expression levels in transfectants were confirmed by quantitative real-time PCR, Western blots, and [3H]-BK binding studies. BTB permeability was quantified with autoradiography and expressed as a unidirectional transport constant, Ki, for [14C]-alpha-aminoisobutyric acid (AIB: Mr 103), using a rat brain tumor model. Baseline Ki values in tumors overexpressing B2R were not significantly higher than in control tumors. Ki values after BK treatment in tumors overexpressing B2R, however, were significantly higher than in control tumors. Western blots confirmed that B2R expression levels in vivo in tumors overexpressing B2R remained higher than in control tumors. These results suggested that alteration of B2R expression levels on tumor cells could modulate BK-mediated BTB permeability. Therefore, B2R expression levels in human glioma could be used to analyze the treatment results of patients undergoing treatment involving BK-modulated BTB permeability.

Rat blood-brain barrier genomics. II.

The tissue-specific gene expression at the brain microvasculature, which forms the blood-brain barrier (BBB) can be elucidated with a brain vascular genomics program, which starts with the isolation of gene products derived from purified brain microvessels. Genes commonly expressed in peripheral organs are subtracted with the suppression subtractive hybridization method using driver cDNA produced from a pool of rat liver/kidney-derived poly A+RNA. From a screen of 480 clones in the subtracted tester cDNA library, 156 clones were sequenced. The clones fell into 3 groups: known genes (51%), rat expressed sequence tags (31%), and novel rat genes not found in databases (18%). The known genes could be categorized into families of common function including vascular remodeling, signal transduction, transcription factors, biologic transport, vascular amyloid, hemostasis, myelin, lipids, secretion, cytoskeleton, and junctional complexes. Brain vascular genomics, or BBB genomics, allows for an accelerated discovery of the gene families that are differentially expressed at the microvasculature in brain.

Expression polymorphism of the blood-brain barrier component P-glycoprotein (MDR1) in relation to Parkinson's disease.

Because drug transporters such as P-glycoprotein, the product of the multidrug resistance (MDR1 ) gene, contribute to the function of the blood-brain barrier, we hypothesized that differences in their expression could affect the uptake of neurotoxic xenobiotics, thereby modulating interindividual susceptibility for neurological disorders such as Parkinson's disease. In a pilot case-control study comprising 95 Parkinson's disease patients (25 early-onset patients with onset age < or = 45 years) and 106 controls we analysed the three common polymorphisms, 3435C >T in exon 26, 2677G > T,A in exon 21, and -129T > C in exon 1b. There were no statistically significant associations between any of these polymorphisms and Parkinson's disease. However, a distribution pattern consistent with our hypothesis was observed in that the frequency of the 3435T/T genotype, which had previously been associated with decreased P-glycoprotein expression and function, was highest in the early-onset Parkinson's disease group (36.0%), second-highest in the late-onset Parkinson's disease group (22.9%), and lowest in the control group (18.9%). Furthermore, we confirmed that the MDR1 exon 21 and exon 26 polymorphisms are in significant linkage disequilibrium since the [2677G, 3435C] and [2677T, 3435T] haplotypes were far more frequently observed than expected. In conclusion, MDR1 and other drug transporters represent plausible candidates as Parkinson's disease risk genes. Larger studies are required to confirm this role in the etiology of Parkinson's disease.

Presence of claudins mRNA in the brain. Selective modulation of expression by kindling epilepsy.

In the central nervous system, the junctional types that establish and maintain tissue architecture include gap junctions, for cytoplasmic connectivity, and tight junctions, for paracellular and/or cell polarity barriers. Connexins are the integral membrane proteins of gap junctions, whereas occludin and members of the multigene family of claudins form tight junctions. In the brain, there are no transendothelial pathways, as continuous tight junctions are present between the endothelial cells. Thus, they provide a continuous cellular barrier between the blood and the insterstitial fluid. However, several brain pathologies, including epilepsy, are known to alter the permeability of the blood-brain barrier and to cause edema. Therefore, since claudins, as constitutive proteins of tight junctions are likely candidates for modulation under pathological states, we explored their normal pattern of expression in the brain and its modulation by seizures. We found that several members of this family are normally expressed in the hippocampus and cortex. Interestingly, claudin-7 is expressed in the hippocampus but not in the cortex. On the other hand, the expression of claudin-8 is selectively down-regulated in the hippocampus as kindling evolves. These results link for the first time the modulation of expression of a tight junction protein to abnormal neuronal synchronization that could probably be reflected in permeability changes of the blood-brain barrier or edema.

Circulating gp120 alters the blood-brain barrier permeability in HIV-1 gp120 transgenic mice.

The mechanism underlying invasion of the central nervous system by HIV-1 is unclear. We recently demonstrated blood-brain barrier changes in a model of HIV-1 gp120 transgenic mice. To test whether this alteration was intrinsic to the brain endothelium of transgenic mice or depended on circulating gp120, we used brain endothelial cultures from gp120 transgenic and non-transgenic mice and exposed them to serum from gp120 transgenic or non-transgenic mice. We measured permeability to albumin as a marker of functional endothelial integrity. A significant increase in permeability (up to 47%) was observed in transgenic and non-transgenic cultures exposed to serum samples from transgenic but not to those from non-transgenic mice. This permeability was neutralized after immunoabsorption of sera with anti-gp120 monoclonal antibody. These findings demonstrate that the blood-brain barrier alteration in HIV-1 gp120 transgenic mice is due to circulating gp120.

Isolating vessels from the mouse brain for gene expression analysis using laser capture microdissection.

Studies of gene expression often examine a pool of RNA extracted from the diverse cell types making up a tissue. We have developed a method for isolating vessels from the brain in order to understand the changes occurring in the vessels during the pathogenesis of cerebral malaria. Vessels were visualised by incubating sections of mouse brain with a substrate for alkaline phosphatase. Vessels were collected by laser capture microdissection and the specificity was monitored by measuring the expression of cell-specific markers. RNA from the captured vessels was highly enriched in mRNA for genes associated with endothelial cells and pericytes. Measurement of indoleamine 2,3-dioxygenase mRNA indicated it was possible to detect changes in gene expression, due to malaria infection, occurring specifically within the vessels. Laser capture microdissection can be used to study changes in gene expression occurring at the blood-brain barrier.

Subtractive expression cloning reveals high expression of CD46 at the blood-brain barrier.

A subtractive expression cloning methodology was used to identify proteins having enriched expression at the blood-brain barrier (BBB) in comparison to liver and kidney tissues. A bovine brain capillary COS-1 cell cDNA expression library was screened with a BBB-specific antiserum. This strategy revealed that the membrane cofactor protein CD46, which is a regulator of complement activation in vivo and is also a potential measles virus receptor, is highly expressed at the BBB. The selective CD46 expression in brain at the BBB was confirmed by Northern blot analysis and confocal microscopy. The finding of selective expression of CD46 at the BBB is consistent with an important role played by the microvasculature in the immune surveillance of the brain.

TNFalpha transport across the blood-brain barrier is abolished in receptor knockout mice.

The presence of transport systems at the blood-brain barrier (BBB) enables some cytokines in blood to reach specific targets in the brain and spinal cord. The "transporters" function in a way different from conventional receptors, in that cytokines are chaperoned from blood to the CNS rather than being degraded in the specialized endothelial cells composing the BBB. Here we present the first study to determine whether the transporter for tumor necrosis factor-alpha (TNFalpha) is identical to its receptors. Three types of TNFalpha receptor knockout mice were used, and the influx of (125)I-TNFalpha from blood to brain and blood to spinal cord was measured. In either p55 or p75 receptor knockout mice, the influx of (125)I-TNFalpha was significantly, but not completely, decreased in spinal cord, whereas the decrease in brain was not statistically significant. This indicates that both receptors are partially involved in the transport of TNFalpha across the BBB but that neither receptor is the sole transporter. By contrast, in double knockout mice lacking both p55 and p75 receptors, the entry of (125)I-TNFalpha into brain and spinal cord was completely abolished. Therefore, both receptors are necessary for transporting TNFalpha across the BBB. The results clearly demonstrate that the transport of TNFalpha across the BBB is a complicated process involving additive or even synergistic activities of both receptors, thus differing from typical ligand-receptor binding and downstream signal transduction.

Psychomotor retardation, spastic paraplegia, cerebellar ataxia and dyskinesia associated with low 5-methyltetrahydrofolate in cerebrospinal fluid: a novel neurometabolic condition responding to folinic acid substitution.

INTRODUCTION: Normal brain development and function depend on the active transport of folates across the blood-brain barrier. The folate receptor-1 (FR 1) protein is localized at the basolateral surface of the choroid plexus, which is characterized by a high binding affinity for circulating 5-methyltetrahydrofolate (5-MTHF). PATIENTS AND METHODS: We report on the clinical and metabolic findings among five children with normal neurodevelopmental progress during the first four to six months followed by the acquisition of a neurological condition which includes marked irritability, decelerating head growth, psychomotor retardation, cerebellar ataxia, dyskinesias (choreoathetosis, ballism), pyramidal signs in the lower limbs and occasional seizures. After the age of six years the two oldest patients also manifested a central visual disorder. Known disorders have been ruled out by extensive investigations. Cerebrospinal fluid (CSF) analysis included determination of biogenic monoamines, pterins and 5-MTHF. RESULTS: Despite normal folate levels in serum and red blood cells with normal homocysteine, analysis of CSF revealed a decline towards very low values for 5-methyltetrahydrofolate (5-MTHF), which suggested disturbed transport of folates across the blood-brain barrier. Genetic analysis of the FR 1 gene revealed normal coding sequences. Oral treatment with doses of the stable compound folinic acid (0.5-1 mg/kg/day Leucovorin(R)) resulted in clinical amelioration and normalization of 5-MTHF values in CSF. CONCLUSION: Our findings identified a new condition manifesting after the age of 6 months which was accompanied by low 5-MTHF in cerebrospinal fluid and responded to oral supplements with folinic acid. However, the cause of disturbed folate transfer across the blood-brain barrier remains unknown.

Interleukin-1 beta is required for the early evolution of reactive astrogliosis following CNS lesion.

The CNS response to injury is characterized by the rapid activation of astrocytes in a process known as astrogliosis. The function of reactive astrocytes is controversial, in that both beneficial and detrimental properties are postulated. Identification of the molecules involved in regulating astrogliosis is an important step towards understanding astrocyte functions and establishing suitable conditions for CNS regeneration. We previously reported that inflammatory cytokines are regulators of astrogliosis but the key cytokine involved in initiating astrogliosis was unclear. We describe here that the elevation of glial fibrillary acid protein (GFAP) transcripts follows the very early rise of interleukin (IL)-1beta mRNA in a murine corticectomy model of CNS lesion. Furthermore, the injury-induced upregulation of GFAP mRNA and protein did not occur in mice genetically deficient for IL-1beta compared to wild-type animals. This was correlated with an absence of an increase in GFAP-immunoreactivity (GFAP-ir) in IL-1beta-null mice at 2 and 3 days of injury. However, by 5 to 7 days after the lesion, GFAP-ir was not different between cytokine-deficient and wild-type controls. Functionally, mice lacking IL-1beta exhibited a significant impairment in reformation of the blood-brain barrier (BBB) following corticectomy compared to wild-type controls. These findings suggest that the rapid production of IL-1beta following trauma plays a beneficial role in initiating astrogliosis in an attempt to restore the integrity of the BBB and seal off the wound site.

Effect of mdr1a P-glycoprotein gene disruption, gender, and substrate concentration on brain uptake of selected compounds.

PURPOSE: This study assessed the influence of mdr1a P-glycoprotein (P-gp) gene disruption, gender and concentration on initial brain uptake clearance (Clup) of morphine, quinidine and verapamil. METHODS: Clup of radiolabeled substrates was determined in P-gp-competent and deficient [mdr1a(-/-)] mice by in situ brain perfusion. Brain:plasma distribution of substrates after i.v. administration was determined in both strains. RESULTS: Genetic disruption of mdr1a P-gp resulted in 1.3-, 6.6- and 14-fold increases in Clup for morphine, verapamil and quinidine, respectively. With the exception of small differences for verapamil, gender did not affect Clup. Saturable transport of verapamil and quinidine was observed only in P-gp-competent mice, with apparent IC50 values for efflux of 8.6 +/- 2.3 microM and 36 +/- 2 microM, respectively. Verapamil Clup was approximately 50% higher in mdr1a(+/-) vs. mdr1a(+/+) mice; no such difference was observed for quinidine. In P-gp-competent mice, uptake of verapamil and quinidine was unaffected by organic vehicles. Plasma decreased VER Clup to a greater extent in the presence of P-gp. The influence of P-gp in situ was lower than, but correlated with, the effect in vivo. CONCLUSIONS: P-gp decreases Clup of morphine, verapamil and quinidine in situ with little or no influence of gender, but this effect cannot fully account for the effects of P-gp in vivo. P-gp is the only saturable transport mechanism for verapamil and quinidine at the murine blood-brain barrier. The influence of protein binding on Clup may be enhanced by P-gp-mediated efflux.

Altered gene expression in cerebral capillaries of stroke-prone spontaneously hypertensive rats.

Stroke-prone spontaneously hypertensive rats (SHRSP) are a well-characterized, genetic model for stroke. We showed earlier that the structure and function of the tight junctions in SHRSP blood-brain barrier endothelial cells is disturbed prior to stroke. To investigate the molecular events leading to endothelial dysfunction in SHRSP cerebral capillaries, we carried out suppression subtractive hybridization (SSH) in combination with a cDNA filter screening step. We identified two cDNA fragments that were upregulated in SHRSP, compared to stroke-resistant spontaneously hypertensive rats (SHR), and found open reading frames of 133 and 138 amino acids, respectively. These peptides did not match any known proteins in public databases. A third upregulated SHRSP cDNA fragment was identified as the rat sulfonylurea receptor 2B (SUR2B). We also isolated and cloned the cDNA of the rat homologue for the mouse G-protein signaling 5 (RGS5) regulator. This regulator was downregulated in SHRSP. We used in situ hybridization to show that rat RGS5 is expressed in the brain capillary endothelium and in the choroid plexus. Our findings may lead to the identification of new stroke-related genes.

Intrathecal delivery of IFN-gamma protects C57BL/6 mice from chronic-progressive experimental autoimmune encephalomyelitis by increasing apoptosis of central nervous system-infiltrating lymphocytes.

The exclusive detrimental role of proinflammatory cytokines in demyelinating diseases of the CNS, such as multiple sclerosis, is controversial. Here we show that the intrathecal delivery of an HSV-1-derived vector engineered with the mouse IFN-gamma gene leads to persistent (up to 4 wk) CNS production of IFN-gamma and inhibits the course of a chronic-progressive form of experimental autoimmune encephalomyelitis (EAE) induced in C57BL/6 mice by myelin oligodendrocyte glycoprotein (MOG)(35-55). Mice treated with the IFN-gamma-containing vector before EAE onset showed an earlier onset but a milder course of the disease compared with control mice treated with the empty vector. In addition, 83% of IFN-gamma-treated mice completely recovered within 25 days post immunization, whereas control mice did not recover up to 60 days post immunization. Mice treated with the IFN-gamma-containing vector within 1 wk after EAE onset partially recovered from the disease within 25 days after vector injection, whereas control mice worsened. Recovery from EAE in mice treated with IFN-gamma was associated with a significant increase of CNS-infiltrating lymphocytes undergoing apoptosis. During the recovery phase, the mRNA level of TNFR1 was also significantly increased in CNS-infiltrating cells from IFN-gamma-treated mice compared with controls. Our results further challenge the exclusive detrimental role of IFN-gamma in the CNS during EAE/multiple sclerosis, and indicate that CNS-confined inflammation may induce protective immunological countermechanisms leading to a faster clearance of encephalitogenic T cells by apoptosis, thus restoring the immune privilege of the CNS.

Emerging molecular mechanisms of brain tumour oedema.

A common property of brain tumours is their ability to cause oedema in the surrounding brain. Oedema forms as a result of a leaky blood-tumour barrier and persists when the brain fails to clear the excess fluid. It is a significant source of morbidity and mortality. The principal anatomical component of the blood-brain barrier is the endothelial tight junction which opens in glioma microvessels. Multiple tight junction proteins have recently been identified, such as occludin, claudin, ZO-1, ZO-2 and ZO-3. We propose a model to explain tight junction opening in gliomas based on vascular endothelial growth factor secretion and loss of tight junction inducing factor production by tumour cells. The level of expression of the water channel aquaporin-4 in peritumoural astrocytes may determine the rate of oedema fluid clearance. The identification of the molecular mechanisms of brain tumour oedema may allow the design of novel anti-oedema medications.