- Email: [email protected]
Cadherin-10 is a novel blood–brain barrier adhesion molecule in human and mouse
Brain Research 1058 (2005) 62 – 72 www.elsevier.com/locate/brainres
Research Report
Cadherin-10 is a novel blood–brain barrier adhesion molecule in human and mouse Matthew J. Williams, Margaret B. Lowrie, Jonathan P. Bennett 1, J. Anthony Firth, Peter Clark* Division of Biomedical Sciences, Imperial College London, Sir Alexander Fleming Building, South Kensington, London SW7 2AZ, UK Accepted 27 July 2005 Available online 21 September 2005
Abstract Maintenance of the specialised environment of the central nervous system requires barriers provided by the endothelium of brain microvessels (the blood – brain barrier (BBB)) or the epithelium lining the ventricles (CSF – brain barrier) or the choroid plexus (blood – CSF barrier). Inter-endothelial junctions are more extensive in the BBB than in other tissues, with elaborate tight junctions. However, few differences in the molecular composition of these junctions have been described. Here, we show, in both human and mouse brain, that the type II classical cadherin, cadherin-10, is expressed in BBB and retinal endothelia, but not in the leaky microvessels of brain circumventricular organs (CVO), or in those of non-CNS tissues. This expression pattern is distinct from, and reciprocal to, VE-cadherin, which is reduced or absent in tight cortical microvessels, but present in leaky CVO vessels. In CVO, the barrier function is switched from the microvasculature to the adjacent ventricular epithelium, which we also find to express cadherin-10. In the vessels of gliobastoma multiforme tumours, where BBB is lost, cadherin-10 is not detected. This demonstration of a distinctive expression pattern of cadherin-10 suggests that it has a pivotal role in the development and maintenance of brain barriers. D 2005 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Blood – brain barrier Keywords: Endothelial cell; Cadherin; Capillary; Cell junction
1. Introduction Cadherins are a large family of cell – cell adhesion molecules involved in inter-cellular adhesion in a wide variety of cell types [4,40]. In the nervous system, cadherins are known to be crucial to all stages of development, including the early separation of the neural tube from the ectoderm, the segregation of neurones and axons, and the formation of synapses [32,33]. These molecules mediate calcium ion-dependent, homotypic
Abbreviations: BBB, blood – brain barrier; CVO, circumventricular organ; UEA-1, Ulex europaeus agglutinin-1; Glut-1, glucose transporter-1 * Corresponding author. E-mail address: [email protected] (P. Clark). 1 Present address: Hull York Medical School, University of Hull, HU6 7RX. 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.07.078
adhesion, and have roles in cell – cell recognition, the segregation of cells and tissues, and the formation of intercellular junctions [21]. Cadherins are typically integralmembrane glycoproteins containing multiple, tandemrepeated cadherin ectodomains, a transmembrane domain, and a cytoplasmic domain that, in many instances, interacts with the actin cytoskeleton. Large numbers of different cadherins from various cadherin sub-families are expressed in the central nervous system, and are believed to be key in the development of the complex neuronal circuitry and patterns of synapses [32,40]. However, many of the cadherins that have been identified in the CNS have not been fully characterised. While studying the materno – foetal barrier, using a culture model of Human Placental Microvascular Endothelial Cells developed in our laboratory [14], we screened for classical cadherin expression using RT-PCR with degenerate
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
primers designed using two highly conserved amino acid sequences in the cytoplasmic domain. One of the cadherins identified was cadherin-10 (unpublished observations). Cadherin-10 was first partially cloned from human brain [31], but its mRNA has also been shown to be present in mouse thymus (designated T2-cadherin), mouse testis, and in the developing brain and eye of mouse, rat, and chick [5,15,17,18,25]. The full-length human cDNA of cadherin10 was sequenced, and the predicted protein was shown to be a type-2, classical cadherin whose mRNA is expressed predominantly in brain, with lesser expression in kidney [22]. The studies of the expression and distribution of cadherin-10 mRNA have led to it being generally regarded as a brain-cadherin, and have suggested that it is one of a battery of cadherins involved in the establishment and maintenance of axon tracts [5,18]. To date, no examination of the distribution of cadherin-10 protein has been undertaken. In this study, we have raised antiserum to cadherin10 and examined the expression of cadherin-10 protein in both human and mouse brain.
2. Materials and methods 2.1. Polymerase chain reaction (PCR) PCR amplification was performed on cDNAs from a range of human and mouse tissues (First Strand, Origene, US). Amplification was carried out with a GeneE thermal cycler (Techne) using Taq DNA polymerase (Qiagen, UK). Primers were synthesised by Invitrogen (UK). Primer sequences were as follows: human cadherin-10 forward 5V-CGACTCACTTGCAACCTA-3V and reverse 5V-GGGCAGGACATGTACCTA-3V; human GAPDH forward 5VGTTGAAGGTCGGAGTCAACG-3V and reverse 5V-CA AAGTTGTCATGGATGACC-3V; mouse cadherin-10 forward 5V-AAGGAGATGGATCGCTCA-3V and reverse 5VTTCACTGTGGTGGTTCCA-3V; and mouse GAPDH forward 5V-TGCTGAGTATGTCGTGGAGTCTA-3V and reverse 5V-AGTGGGAGTTGCTGTTGAAGTCG-3V. PCRamplified products were analysed and visualised on a 1.5% agarose gel using standard procedures. 2.2. Generation of cadherin-10 antibodies Cadherin-10-specific antigens were identified by comparison of amino acid sequences of related cadherins, and a peptide sequence in the third extracellular domain (QNTIHLRVLESSPV) was selected for being unique to cadherin-10, and for having minimal identity with other cadherins, using the Human Genome Mapping Project BLAST database (http://www.hgmp.mrc.ac.uk/) (Fig. 1c). Antiserum was obtained using the rabbit-inoculation service provided by Sigma-Genosys (UK), and cadherin10-specific antibodies affinity purified using a peptide-
63
conjugated column (Sulfo-Link affinity column, PerBio, UK). 2.3. Western blotting Brains were dissected from adult female ML1 mice freshly killed in accordance with UK Home Office regulations. Frozen, unfixed human brain tissue, from individuals unaffected by neurological disease, was kindly provided by the UK Multiple Sclerosis Tissue Bank, Imperial College London. Brain samples were placed in ice-cold Laemmli Buffer supplemented with protease inhibitors and phosphatase inhibitors, and prepared using standard procedures. Protein samples (20 Ag) were electrophoretically separated on an 8% acrylamide gel, and transferred by semi-dry electrophoresis onto Immobilon PVDF nitrocellulose membrane (Millipore, USA). Membranes were blocked with 5% non-fat dried milk in TBS/Tween (blocking buffer), incubated with rabbit cadherin-10 polyclonal antibodies overnight at 4 -C, and incubated with horseradish peroxidase-conjugated antirabbit secondary antibody (Santa Cruz, USA), 1 h at room temperature. Bands were visualised using Pierce Pico-West Super-Signal enhanced chemiluminescence (ECL) reagents (PerBio, UK), and were developed as autoradiographs using ECL photographic film (Amersham, UK). 2.4. Immunohistochemistry Brains and eyes were dissected from adult female MF1 mice transcardially perfused with 4% paraformaldehyde in PBS in accordance with UK Home Office regulations. Fixed tissues were embedded in paraffin wax, for immunoperoxidase staining, or embedded in OCT compound (BDH, UK) and snap-frozen for immunofluorescence staining, using standard procedures. Fixed human brain samples from individuals unaffected by neurological pathology were kindly provided by the UK Multiple Sclerosis Tissue Bank, Imperial College London. These samples were processed for wax-embedding or frozen as for mouse tissues above. Serial sections of human brain cortex, and serial coronal and sagittal sections of mouse brain, of 7 Am thickness were obtained using standard histological procedures. Sections of wax-embedded archival biopsies from three patients diagnosed with glioblastoma multiforme, and from one patient with low-grade astrocytoma, were kindly supplied by the Department of Neuropathology, Imperial College London, Charing Cross Hospital, London, UK. Wax-embedded human and mouse brain sections were re-hydrated using a series of ethanol solutions, and labelled using standard protocols. Endogenous peroxidases were inhibited by incubation in 3% hydrogen peroxide in methanol. Tissue sections were blocked in 5% normal goat serum (NGS), and incubated with 5 Ag ml 1 rabbit anti-
64
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
Fig. 1. Expression of cadherin-10 mRNA and protein in human and mouse tissues. (a) PCR products of human tissue templates using (i) cadherin-10 primers and (ii) GAPDH primers. (b) PCR products of mouse tissue templates using (i) cadherin-10 primers and (ii) GAPDH primers. (c) Table showing cadherin-10specific antigen peptide alignment with homologous regions of type I and II classical cadherins. Sequences from the National Library of Medicine Peptide database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi). (d) Western blots of human and mouse brain lysates probed for cadherin-10.
cadherin-10 affinity-purified antibody, rabbit IgG control, or no-primary control, in 5% NGS overnight at 4 -C with humidity. Sections were subsequently incubated in HRPconjugated secondary antibody (Santa Cruz, USA) 1 h at room temperature and incubated with DAB substrate solution (Vector Labs, UK), until colour development had occurred. One set of brain tumour biopsy sections was stained with anti-cadherin-10 antibodies, as above, and a second set, serial to the first, was stained with HRPconjugated Ulex europaeus agglutenin-1 (UEA-1) (Sigma, UK). Both sets were counter-stained with haematoxylin. Tissue sections were dehydrated and coverslips mounted onto slides using DPX (Histomount, National Diagnostics). Tissues were observed using a Nikon light microscope (Nikon, Japan), and images were captured using a Nikon digital camera and Nikon computer software. Frozen sections of mouse brain were stained with 5 Ag ml 1 affinity-purified rabbit polyclonal anti-cadherin-10 antibodies alone, or together with rat monoclonal anti-VEcadherin/CD144 (BD Pharmingen) antibodies at 1:50 dilution, or rat monoclonal anti-Platelet Endothelial Cell Adhesion Molecule-1 (PECAM-1)/CD31 (BD Pharmingen) antibodies at 1:50 dilution. Frozen human sections were
stained with rabbit polyclonal anti-cadherin-10 antibodies together with mouse monoclonal anti-human VE-cadherin (BV6) antibodies (kindly supplied by Drs. Stephan Leibner and E. Dejana, University of Milan, Italy) or FITCconjugated UAE-1 (Vector Labs, UK). Following standard protocols, specimens were permeabilised with 0.5% Triton X-100, blocked with 5% NGS, and incubated overnight at 4 -C in primary antibodies, purified immunoglobulin G (IgG), or without antibody. After washing in PBS, sections were incubated for 1 h at room temperature with 2 Ag ml 1 Alexa Fluor fluorescently-labelled secondary antibodies (Molecular Probes, Netherlands) or fluoroscein-conjugated UEA-1 (2 Ag ml 1, Vector Labs, UK). After further washing in PBS and water, sections were mounted onto coverslips using Vectashield Hard-Set Mounting Medium (Vector Labs, UK). Confocal images were collected using a Leica TCS NT microscope system (Leica Microsystems, Germany) with a krypton-argon laser. When comparing staining intensity in different regions of a tissue, conditions were kept constant during image collection, which were saved in digital format using TCS NT software. To test the specificity of the cadherin-10 antibodies, pre-adsorption of the affinity-purified cadherin-10 antibody with its peptide
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
antigen resulted in the abolition of any specific staining (not shown).
3. Results 3.1. Cadherin-10 mRNA expression in human and mouse PCR analysis of cDNAs derived from human and mouse showed that cadherin-10 mRNA can be amplified from a number of tissues (Figs. 1a, b), with the strongest signal, compared to the housekeeping gene loading-control GAPDH, being found in brain of both human (Fig. 1a) and mouse (Fig. 1b). These findings are in agreement with previous studies showing that cadherin-10 mRNA is expressed in mouse brain, testis, and thymus and in human brain, with lesser expression in kidney and lung [5,15,17, 18,22,25,31]. In addition, we show that cadherin-10 mRNA is also present in tissues of the gastro-intestinal and urogenital tracts. 3.2. Characterisation of cadherin-10 antibodies The peptide antigen used to immunise rabbits to produce anti-cadherin-10 antiserum was designed from a region of the third cadherin extracellular domain (EC3) of the predicted amino acid sequence of the previously cloned human cadherin-10 cDNA. This sequence is identical in mouse and rat cadherin-10, with three amino acid differences in the chicken (Fig. 1c). The sequence is highly divergent in other human cadherins, allowing us to be confident that the antiserum would not cross-react with
65
other known cadherins. The staining pattern of cadherin-10 in a variety of cells and tissues, that express a number of characterised cadherins, suggests that our antiserum does not cross-react with other known classical cadherins (unpublished observations). When we probed Western blots of human and mouse brain lysates, run in SDS-PAGE, with the affinity-purified rabbit antibodies to cadherin-10, we obtained single major bands at approximately 120 kDa (Fig. 1d). This size is consistent with the recognition of a classical cadherin, which is typically in the range 115 –130 kDa. Fainter bands observed in the mouse brain blot could indicate a higher molecular mass pre-protein and proteolysis products. A previous study immunoprecipitated transfected human cadherin-10, via h-catenin, from mouse L-cells, which was found to run as a band at 116 kDa [30]. 3.3. Cadherin-10 protein expression in human and mouse brain vasculature Immunohistochemical localisation of cadherin-10 in normal adult human brain revealed a pattern of distribution consistent with expression in microvessels of the neuropil, with little or no neural expression (Figs. 2a, b). The microvessel-specific nature of this expression can be seen in Fig. 2c where larger, pial vessels are negative, while positive structures in the adjacent sub-pial cortex have a microvascular morphology. The microvascular nature of the cadherin-10-positive structures in the cortex was confirmed by double-labelling of cadherin-10 with the UEA-1 lectin, which is known to bind specifically to human endothelial cells (Figs. 2e – h). The localisation of cadherin-10 relative
Fig. 2. Cadherin-10 distribution in human brain. Bright-field micrographs of immunoperoxidase-labelled sections of temporal cerebral cortex (a) and cerebellum (b) showing cadherin-10-positive microvessels in the neuropil. In the temporal cerebral cortex (c), cadherin-10 distribution within the vasculature is restricted to microvessels (MV), and is absent from larger pial blood vessels (PV). (d) IgG control; region shown in panel (c), no specific staining was found when tissue sections were incubated with an equivalent concentration of rabbit IgG. (e – g) Confocal micrographs of cortical microvessels in the neuropil showing distribution of cadherin-10 (e), UEA-1 (f), merged in panel (g), and corresponding phase contrast micrograph (h). Scale bar: 500 Am (a, b); 200 Am (c, d); 8 Am (e – h).
66
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
to UEA-1, where cadherin-10 staining is partially internal to UEA-1 staining, shows that cadherin-10 is expressed by the microvascular endothelium. In all microvessels examined in the cortical and cerebellar neuropil, cadherin-10 and UEA-1 were closely associated, with partial colocalisation. In mouse brain, immunohistochemical localisation of cadherin-10 resulted in a more complex pattern than in the human brain. Cadherin-10 was present in a microvascular distribution in cortex (Figs. 3a, b), but was also found to be present in larger intraventricular vessels and in pial vessels
(Figs. 3c, d). In addition, the lepto-meningeal surface of the cortex, but not the adjacent cerebellar surface, was positively stained (Fig. 3d). In the choroid plexus, capillaries were not stained, while the epithelium was strongly labelled (Fig. 3e). Double immunolabelling with cadherin10 and the endothelial marker, PECAM-1, showed that, in the majority of the larger vessels, cadherin-10 was absent from the endothelium, but present in peri-endothelial cells which we have found are a-smooth muscle actin positive (data not shown). In cortical microvessels, cadherin-10 and
Fig. 3. Cadherin-10 distribution in mouse brain. Bright-field micrographs of immunoperoxidase-labelled brain sections (a – f). Sub-pial cortex (a and b), with cadherin-10-positive microvessels (MV) within the neuropil. (c) 3rd ventricle, with cadherin-10-positive ependyma (E) and blood vessels (BV). (d) Pial surface and transverse fissure, lined with cadherin-10-positive cells on the cortical surface, but not on the adjacent cerebellar surface. (e) The epithelial cells of the choroid plexus of the 4th ventricle are strongly stained for cadherin-10. (f) IgG control; same region as shown in panel (e) with equivalent concentrations of rabbit IgG. (g) Diagram of mouse brain in sagittal section, showing the regions from which micrographs (a) – (f) were obtained. (h – p) Confocal micrographs of brain cortex showing localisation of cadherin-10 and PECAM-1. (h – j) Sub-pial cortex; cadherin-10 (h) and PECAM-1 (i) colocalise in microvessels (open arrows), but not in larger blood vessels (BV) (j). In cortical microvessels (k – m), cadherin-10 (k) and PECAM-1 (l) both co-localised in the merge (m). (n – p) Pial blood vessel stained for cadherin-10 (n), PECAM-1 (o), merged in panel (p); cadherin-10 was observed at inter-endothelial junctions (closed arrows), and in vascular smooth muscle (SM) cells. Scale bars: 100 Am (a, b), 500 Am (c – f), 50 Am (h – j), 10 Am (k – p).
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
PECAM-1 co-localised, suggesting that, as in the human brain, microvascular endothelial cells express cadherin-10 (Figs. 3k –m). In a small proportion (estimated to be 5% of pial vessels) of larger pial veins and venules, cadherin-10 could be seen to co-localise with PECAM-1 in a pattern suggesting that both are concentrated at inter-endothelial junctions (Figs. 3n –p). It is known that, in rodents, a proportion of pial vessels express blood – brain barrier markers and possess barrier characteristics similar to cortical microvessels [3,9]. 3.4. Cadherin-10 expression in circumventricular organs The microvessels of the circumventricular organs (CVOs) are known to lack blood –brain barrier characteristics. Barrier function in these areas of the brain is transferred from the microvessels to the adjacent ventricular ependymal cells that act as the local CSF –brain barrier, or the blood – CSF barrier in the choroid plexus [7]. Similarly, markers of blood – brain barrier (e.g. glucose transporter-1 (Glut-1)), which are present in the majority of brain microvascular endothelial cells, are absent from CVO
67
microvascular endothelium, but present in the adjacent ventricular ependymal cells [27]. When we examined the expression of cadherin-10 in CVO of both human and mouse, we observed that while epithelial structures were positively stained, microvessels were not. In human brain, we observed cadherin-10 in the ependymal lining cells of the floor of the 3rd ventricle (median eminence) (Figs. 4a, b), and in the epithelium of the choroid plexus (Figs. 5c, d). The UEA-1-positive choroid plexus endothelia are negative for cadherin-10 (Fig. 4c). In mouse brain, cadherin-10 is present in the ependymal lining of the third ventricle (Fig. 4e). More detailed examination of cadherin-10 distribution in the infundibular recess showed that the cadherin is present in the cells lining the lateral recesses (a2 and h1 tanycytes), with significantly lesser expression in the medial h2 tanycytes. This pattern of expression is identical to the known pattern of expression of the BBB marker, Glut-1 [27]. Strands of cadherin-10 positive material radiating from the epithelial lining of the infundibular process (Fig. 4f) are consistent with the tanycyte basal processes that project to the median eminence. Cadherin-10 and PECAM-1 do not co-localise
Fig. 4. Cadherin-10 in circumventricular organs in human and mouse brain. (a) Cadherin-10 in the human median eminence with (b) the corresponding phase contrast image. Ependymal cells (E) lining third ventricle (3rd Vent) are positive for cadherin-10, whereas few microvessels are stained for cadherin-10. (c) Cadherin-10 and UEA-1 localisation in the human choroid plexus with (d) the corresponding phase contrast image. Epithelial cells of choroid plexus + VE for cadherin-10 (red), while the UEA-1-positive endothelial cells (green) are negative for cadherin-10. (e – k) Cadherin-10 and PECAM-1 localisation in the mouse medial basal hypothalamus. (e) Medial basal hypothalamus (MBH), stained for cadherin-10. Cadherin-10-positive cells line the third ventricle (3rd Vent). (f – h) Enlarged region highlighted in (e) median eminence (M.E.), stained for (f) cadherin-10 (arrowheads indicate tanycyte basal processes), (g) PECAM-1, and (h) merge. (i – k) Enlarged image of region highlighted (f – g), stained for (i) cadherin-10, (j) PECAM-1, and (k) merge. Cuffs of cadherin-10-positive structures (as indicated) were observed surrounding PECAM-1-positive microvessels (MV). Scale bar: 100 Am (a – d); 250 Am (e); 40 Am (f – h), 15 Am (i – k).
68
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
Fig. 5. Comparison of the distributions of cadherin-10 and VE-cadherin in mouse and human brain. (a – c) Sub-pial cortical neuropil from mouse brain. Microvessels (MV) are positive for cadherin-10 (a), but negative for VE-cadherin (b), merged in panel (c). (d – f) 4th Ventricular choroid plexus and adjacent brainstem (BrSt) in the same tissue-section as panels (a) – (c). The epithelial cells of the choroid plexus (CP) are positive for cadherin-10, while the microvessels (MV) within the peri-ventricular neuropil and in the choroid plexus were negative for cadherin-10 (d). These endothelial cells were strongly stained for VEcadherin (e), merged in panel (f). (g – l) Confocal images double-labelled for cadherin-10 (red) and VE-cadherin (green) from the same section of human brain. (g – i) Cortical microvessel, cadherin-10 strongly localised to the cell – cell junctions in the MV endothelial cells (g, indicated by closed arrows), with VEcadherin being found at low levels (h), merged in panel (i). (j – l) Pial vessel taken from same tissue-section as panels (g) – (l). While the endothelial cells of the pial vessel are positive for VE-cadherin (k), with a strongly junctional distribution (arrows), cadherin-10 is not detected (j), merged in panel (l). Scale bar: 50 Am (a – f); 8 Am (g – l).
in the capillaries of the median eminence. PECAM-1positive endothelia of microvessels in the median eminence appear to be surrounded by a loose cuff of cadherin-10 staining (Figs. 4f –k) that may represent tanycyte basal processes wrapping capillaries. Cadherin-10 is expressed in choroid plexus epithelium, but not in its microvessels (Figs. 3e, 5d –f), again reflecting the known pattern of expression of the BBB marker, Glut-1 [27]. 3.5. Comparison of cadherin-10 and VE-cadherin distribution VE-cadherin is well established as the adhesion molecule of adherens junctions between endothelial cells [12,23,24], with the possible exception of those in brain vessels [6]. When we compared the distributions of cadherin-10 and
VE-cadherin in double-labelled sections of mouse brain, we found that, in mouse cerebral cortex, cadherin-10, but not VE-cadherin, is present in microvessels (Figs. 5a– c). In other areas of the same section, i.e. the choroid plexus of the fourth ventricle and the adjacent brainstem examined using identical settings in the confocal system, microvessels express VE-cadherin, but not cadherin-10 (Figs. 5d – f). It should be noted that cadherin-10 is present in the choroid plexus epithelial cells, which form the blood – CSF barrier, but not in its leaky endothelium. In human cerebral cortex, we found that the expression of VE-cadherin is consistent with previous findings [26]: VEcadherin staining is less intense in the microvessels than in the junctions of larger lepto-meningeal vessels in the same section (Figs. 5k– l). The expression of cadherin-10 and VEcadherin appears to be inversely related: lepto-meningeal
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
69
vessels with strong junctional labelling of VE-cadherin are negative for cadherin-10; cadherin-10-expressing cortical microvessels have significantly reduced labelling intensity. 3.6. Cadherin-10 expression in retinal microvessels In the mouse retina, cadherin-10 is found in two distinct locations: the retinal microvessel endothelial cells and the retinal pigmented epithelium (RPE) (Figs. 6a, d). The retinal microvessels and the RPE are known to be relatively impermeable, constituting the inner blood – retinal barrier (together with pericytes and retinal glial cells) and outer retinal barrier, respectively [11]. In retinal microvessels, cadherin-10 is co-expressed with PECAM-1 (Figs. 6a – c). However, unlike the mouse brain, where cadherin-10 and VE-cadherin are expressed in distinct sets of microvessels (Fig. 5), retinal microvessels express both cadherin-10 and VE-cadherin (Figs. 6d – f). While the microvessels of the choroid, the choroicapillaris, express both PECAM-1 and VE-cadherin more strongly than in the adjacent retina (Figs. 6b, e), cadherin-10 is absent in these microvessels (Figs. 6a, d). In contrast to the retinal vessels, the choroicapillaris is a highly fenestrated, leaky capillary bed.
Fig. 7. Comparison of cadherin-10 expression in normal cortex and glioblastoma multiforme (GBM) tumour in human. Normal cortex (a – b) and GBM tumour (c – d) in the same section of biopsy. Microvessels are positive for cadherin-10 in the unaffected region (arrows in panel a), but not in the GBM tumour (c). The presence of microvessels in both normal cortex and tumour is shown by UEA-1 staining (arrows in panels b and d). Scale bar: 150 Am.
3.7. Cadherin-10 expression in glioblastoma multiforme It is known that BBB properties are absent in the vessels of both primary and metastatic brain tumours [13,28,36,39]. We examined the expression of cadherin-10 in three cases of human glioblastoma multiforme (GBM). Immunohistochemical localisation of cadherin-10 in each of these tumour biopsies was compared to endothelial labelling by UEA-1 in
serial sections. In all cases, cadherin-10 was absent from GBM vessels. Cadherin-10 is present in the microvessels of normal cortex (Fig. 7a), but absent from vessels in adjacent tumour tissue in the same section (Fig. 7c). In a serial section of this case, UEA-1 labelling shows the presence of numerous vessels in both normal cortex and adjacent tumour tissue (Figs. 7b, d). In a single case of low-grade
Fig. 6. Cadherin-10 in mouse retina, showing distributions of cadherin-10 with PECAM-1 (a – c), and with VE-cadherin (d – f). (a – c) Microvessels (MV) within the retina are positive for both cadherin-10 (a) and PECAM-1 (b), as shown in merge (c), while cadherin-10 was absent from the PECAM-1-positive microvessels in the choroid. (d – f) Both cadherin-10 (d) and VE-cadherin (e) co-localise in retinal microvessels, (f) merge. Cadherin-10 is not expressed in VEcadherin-positive choroid microvessels. Retinal pigmented epithelia (RPE) are positive for cadherin-10, as shown in panels (a) and (d). Scale bar: 50 Am.
70
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
astrocytoma, where BBB properties are typically retained by the vessels [13], microvessels were positive for cadherin-10 (not shown). 3.8. Cadherin-10 in non-brain blood vessels We have examined cadherin-10 protein expression in a number of human (prostate, kidney, bladder, placenta) and mouse (kidney, bladder, prostate, heart, skin, skeletal muscle, placenta, GI-tract) tissues, but have never observed significant expression of cadherin-10 in the endothelia of any vessels (data not shown). In mouse, we observe that peri-endothelial cells express cadherin-10 in both the brain (Figs. 3h –j) and in vascular beds outside of the CNS (data not shown). We have never observed cadherin-10 staining in human vascular smooth muscle (data not shown).
4. Discussion Localisation of cadherin-10 protein shows that it is primarily expressed in vascular and epithelial structures of adult human and mouse brain. Previous studies have examined mRNA expression in the developing brain of mouse, rat, and chicken embryos, and implicated cadherin10 in inter-neuronal adhesion and the segregation of neural tracts [5,17,18]. However, we find that, in adult human and mouse brain, cadherin-10 is expressed by the endothelial cells of microvessels of the cerebral cortex, cerebellum and retina, the epithelial cells of the choroid plexuses, and the ependymal cells lining some ventricular regions. These structures form the bases of the brain barriers. In addition, the RPE cells, which act as the barrier between the retina and the outer layers of the eye [11], also express cadherin10. The pattern of cadherin-10 distribution is essentially identical to that of known markers of brain barriers, such as Glut-1. These markers are generally expressed in microvessel endothelium, choroid plexus epithelium, and in the ependymal cells adjacent to CVOs, while the leaky microvessels of CVOs lack BBB marker expression [7,8]. Cadherin-10 is therefore a rare example of a cell adhesion molecule present in BBB endothelia, but absent from the endothelia of leaky brain vessels and non-brain vessels. The absence of cadherin-10 from the vessels of GBM tumours, where the BBB is known to be compromised, provides further evidence of a key role for this cell adhesion molecule in BBB organisation. In vertebrates, the structural basis of the blood – brain barrier (BBB) is the endothelial lining of brain capillaries [2]. Under the influence of local factors in the CNS, particularly astrocytes, the endothelial cells of the BBB form tighter and more extensive inter-cellular junctions than those found in non-brain capillary beds [1,29]. While it is clear that tight junctions (TJ) are the structures that restrict paracellular passage across epithelial and endothelial cell
layers [19,38], their formation, organisation, and maintenance are believed to depend on cadherin-mediated adhesion, and on the formation of adherens junctions [10,16, 20,34,35,37,41]. In spite of the difference in the complexity of their inter-endothelial junctions, few differences in the repertoire of junction adhesion molecules expressed by BBB and non-brain endothelial cells have been found [26]. It has been shown, in the mouse, that the general endothelial cadherin found at inter-cellular junctions, VE-cadherin (cadherin-5) [12,23,24], is lost from brain microvessels, as the BBB is being established [6]. Recently, it has been found that the tight junction component, claudin-3, which has not been previously reported in endothelia, is present in BBB endothelia [39]. VE-cadherin is generally regarded as the major cadherin of inter-endothelial junctions [12,23,24], but it would appear that this is not the case in the brain. It has been shown that both VE-cadherin mRNA and protein are down-regulated or absent in mouse brain vessels, at a point in development when the BBB is being established [6]. In mouse brain endothelium, we find that cadherin-10 and VE-cadherin are expressed in a mutually exclusive manner: cadherin-10 is present in vessels of known tight BBB vessels, while VEcadherin is expressed in vessels that are known to be leaky. Interestingly, cadherin-10 and VE-cadherin are co-expressed in mouse retinal microvessels, which may indicate that these vessels are an intermediate phenotype. In the human brain, although no qualitative differences in the expression of a wide range of cell adhesion molecules have been observed between cerebral cortical microvessels and non-brain microvessels, VE-cadherin expression in brain microvessels was relatively reduced compared to that in larger vessels [26], despite the known increase in the complexity of BBB microvessel junctions. We also observed this difference in staining intensity between cortical microvessels and the endothelium of lepto-meningeal vessels, and we found that there is an inverse relationship between the expression of cadherin-10 and VE-cadherin. Cadherin-10 expression by BBB endothelia and its absence from CVO vessels indicate that this molecule might be responsible for replacing VEcadherin (at least partially) in inter-endothelial junctions, and may be part of the mechanism of inducing the elaboration of BBB junctions. Cadherin-10 expression in the epithelial barriers of the CNS (ependyma of CVO, the choroid plexus epithelium, and in the RPE of the outer retinal barrier) suggests a more general barrier function for this molecule. Our observation that cadherin-10 is present at the surface of mouse brain cortex, but not the surface of the cerebellum, has revealed possible differences in their organisation. We, as yet, do not know the significance of this, but might speculate that there is heterogeneity of brain surface barriers. In conclusion, we find that the pattern of expression of cadherin-10 is consistent with a key role in forming and maintaining the barriers that are crucial to the normal functioning of central nervous system. The loss of BBB is a
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
key contribution to the pathology of many brain tumours, where oedema is a major complication. The loss of cadherin-10 in the vessels of GBM tumours indicates the involvement of this cell adhesion molecule in pathological processes. It will be interesting to examine cadherin-10 expression in other pathologies where BBB is affected. The tight correlation between the presence of cadherin-10 and known barrier function strongly suggests that cadherin-10 may have unique properties.
Acknowledgments We thank Lorraine Lawrence for histology technical support; the UK Multiple Sclerosis Tissue Bank, Imperial College London, for providing human brain material; Dr Federico Roncaroli, Department of Neuropathology, Imperial College London, for providing material and advice on human brain tumours; and Stefan Liebner and Elisabetta Dejana for anti-human VE-cadherin antibodies. This work was supported by the Wellcome Trust.
References [1] N.J. Abbott, Astrocyte-endothelial interactions and blood – brain barrier permeability, J. Anat. 200 (2002) 629 – 638. [2] N.J. Abbott, Dynamics of CNS barriers: evolution, differentiation, and modulation, Cell Mol. Neurobiol. 25 (2005) 5 – 23. [3] G. Allt, J.G. Lawrenson, Is the pial microvessel a good model for blood – brain barrier studies? Brain Res. Brain Res. Rev. 24 (1997) 67 – 76. [4] B.D. Angst, C. Marcozzi, A.I. Magee, The cadherin superfamily: diversity in form and function, J. Cell Sci. 114 (2001) 629 – 641. [5] I.H. Bekirov, L.A. Needleman, W. Zhang, D.L. Benson, Identification and localization of multiple classic cadherins in developing rat limbic system, Neuroscience 115 (2002) 213 – 227. [6] G. Breier, F. Breviario, L. Caveda, R. Berthier, H. Schnurch, U. Gotsch, D. Vestweber, W. Risau, E. Dejana, Molecular cloning and expression of murine vascular endothelial-cadherin in early stage development of cardiovascular system, Blood 87 (1996) 630 – 641. [7] J.E. Bruni, Ependymal development, proliferation, and functions: a review, Microsc. Res. Tech. 41 (1998) 2 – 13. [8] J.P. Cassella, J.G. Lawrenson, G. Allt, J.A. Firth, Ontogeny of four blood – brain barrier markers: an immunocytochemical comparison of pial and cerebral cortical microvessels, J. Anat. 189 (Pt. 2) (1996) 407 – 415. [9] J.P. Cassella, J.G. Lawrenson, L. Lawrence, J.A. Firth, Differential distribution of an endothelial barrier antigen between the pial and cortical microvessels of the rat, Brain Res. 744 (1997) 335 – 338. [10] R.G. Contreras, L. Shoshani, C. Flores-Maldonado, A. Lazaro, A.O. Monroy, M.L. Roldan, R. Fiorentino, M. Cereijido, E-Cadherin and tight junctions between epithelial cells of different animal species, Pflugers Arch. 444 (2002) 467 – 475. [11] J.G. Cunha-Vaz, The blood – retinal barriers system. Basic concepts and clinical evaluation, Exp. Eye Res. 78 (2004) 715 – 721. [12] E. Dejana, R. Spagnuolo, G. Bazzoni, Interendothelial junctions and their role in the control of angiogenesis, vascular permeability and leukocyte transmigration, Thromb. Haemostasis 86 (2001) 308 – 315. [13] M. Demeule, A. Regina, B. Annabi, Y. Bertrand, M.W. Bojanowski, R. Beliveau, Brain endothelial cells as pharmacological targets in brain tumors, Mol. Neurobiol. 30 (2004) 157 – 183.
71
[14] J.F. Dye, L. Leach, P. Clark, J.A. Firth, Cyclic AMP and acidic fibroblast growth factor have opposing effects on tight and adherens junctions in microvascular endothelial cells in vitro, Microvasc. Res. 62 (2001) 94 – 113. [15] B.E. Faulkner-Jones, L.N. Godinho, S.S. Tan, Multiple cadherin mRNA expression and developmental regulation of a novel cadherin in the developing mouse eye, Exp. Neurol. 156 (1999) 316 – 325. [16] T.P. Fleming, T. Papenbrock, I. Fesenko, P. Hausen, B. Sheth, Assembly of tight junctions during early vertebrate development, Semin. Cell Dev. Biol. 11 (2000) 291 – 299. [17] D. Fushimi, K. Arndt, M. Takeichi, C. Redies, Cloning and expression analysis of cadherin-10 in the CNS of the chicken embryo, Dev. Dyn. 209 (1997) 269 – 285. [18] O.D. Gil, L. Needleman, G.W. Huntley, Developmental patterns of cadherin expression and localization in relation to compartmentalized thalamocortical terminations in rat barrel cortex, J. Comp. Neurol. 453 (2002) 372 – 388. [19] L. Gonzalez-Mariscal, A. Betanzos, P. Nava, B.E. Jaramillo, Tight junction proteins, Prog. Biophys. Mol. Biol. 81 (2003) 1 – 44. [20] B. Gumbiner, B. Stevenson, A. Grimaldi, The role of the cell adhesion molecule uvomorulin in the formation and maintenance of the epithelial junctional complex, J. Cell. Biol. 107 (1988) 1575 – 1587. [21] K.A. Knudsen, A.P. Soler, Cadherin-mediated cell – cell interactions, Methods Mol. Biol. 137 (2000) 409 – 440. [22] P. Kools, K. Vanhalst, E. Van den Eynde, F. van Roy, The human cadherin-10 gene: complete coding sequence, predominant expression in the brain, and mapping on chromosome 5p13 – 14, FEBS Lett. 452 (1999) 328 – 334. [23] M.G. Lampugnani, M. Resnati, M. Raiteri, R. Pigott, A. Pisacane, G. Houen, L.P. Ruco, E. Dejana, A novel endothelial-specific membrane protein is a marker of cell – cell contacts, J. Cell Biol. 118 (1992) 1511 – 1522. [24] L. Leach, P. Clark, M.G. Lampugnani, A.G. Arroyo, E. Dejana, J.A. Firth, Immunoelectron characterisation of the inter-endothelial junctions of human term placenta, J. Cell. Sci. 104 (Pt. 4) (1993) 1073 – 1081. [25] S.B. Munro, A.J. Duclos, A.R. Jackson, M.G. Baines, O.W. Blaschuk, Characterization of cadherins expressed by murine thymocytes, Cell. Immunol. 169 (1996) 309 – 312. [26] E. Navratil, A. Couvelard, A. Rey, D. Henin, J.Y. Scoazec, Expression of cell adhesion molecules by microvascular endothelial cells in the cortical and subcortical regions of the normal human brain: an immunohistochemical analysis, Neuropathol. Appl. Neurobiol. 23 (1997) 68 – 80. [27] B. Peruzzo, F.E. Pastor, J.L. Blazquez, K. Schobitz, B. Pelaez, P. Amat, E.M. Rodriguez, A second look at the barriers of the medial basal hypothalamus, Exp. Brain Res. 132 (2000) 10 – 26. [28] J.C. Reijneveld, E.E. Voest, M.J. Taphoorn, Angiogenesis in malignant primary and metastatic brain tumors, J. Neurol. 247 (2000) 597 – 608. [29] W. Risau, S. Esser, B. Engelhardt, Differentiation of blood – brain barrier endothelial cells, Pathol. Biol. (Paris) 46 (1998) 171 – 175. [30] Y. Shimoyama, G. Tsujimoto, M. Kitajima, M. Natori, Identification of three human type-II classic cadherins and frequent heterophilic interactions between different subclasses of type-II classic cadherins, J. Biochem. 349 (2000) 159 – 167. [31] S. Suzuki, K. Sano, H. Tanihara, Diversity of the cadherin family: evidence for eight new cadherins in nervous tissue, Cell Regul. 2 (1991) 261 – 270. [32] M. Takeichi, H. Matsunami, T. Inoue, Y. Kimura, S. Suzuki, T. Tanaka, Roles of cadherins in patterning of the developing brain, Dev. Neurosci. 19 (1997) 86 – 87. [33] M. Takeichi, T. Uemura, Y. Iwai, N. Uchida, T. Inoue, T. Tanaka, S.C. Suzuki, Cadherins in brain patterning and neural network formation, Cold Spring Harbor Symp. Quant. Biol. 62 (1997) 505 – 510.
72
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
[34] M.L. Troxell, Y.T. Chen, N. Cobb, W.J. Nelson, J.A. Marrs, Cadherin function in junctional complex rearrangement and posttranslational control of cadherin expression, Am. J. Physiol. 276 (1999) C404 – C418. [35] M.L. Troxell, S. Gopalakrishnan, J. McCormack, B.A. Poteat, J. Pennington, S.M. Garringer, E.E. Schneeberger, W.J. Nelson, J.A. Marrs, Inhibiting cadherin function by dominant mutant E-cadherin expression increases the extent of tight junction assembly, J. Cell. Sci. 113 (Pt. 6) (2000) 985 – 996. [36] P. Vajkoczy, M.D. Menger, Vascular microenvironment in gliomas, Cancer Treat. Res. (2004) 249 – 262. [37] M. Watabe-Uchida, N. Uchida, Y. Imamura, A. Nagafuchi, K. Fujimoto, T. Uemura, S. Vermeulen, F. van Roy, E.D. Adamson, M. Takeichi, Alpha-catenin – vinculin interaction functions to organize the apical junctional complex in epithelial cells, J. Cell Biol. 142 (1998) 847 – 857.
[38] H. Wolburg, A. Lippoldt, Tight junctions of the blood – brain barrier: development, composition and regulation, Vasc. Pharmacol. 38 (2002) 323 – 337. [39] H. Wolburg, K. Wolburg-Buchholz, J. Kraus, G. Rascher-Eggstein, S. Liebner, S. Hamm, F. Duffner, E.H. Grote, W. Risau, B. Engelhardt, Localization of claudin-3 in tight junctions of the blood – brain barrier is selectively lost during experimental autoimmune encephalomyelitis and human glioblastoma multiforme, Acta Neuropathol. (Berl.) 105 (2003) 586 – 592. [40] T. Yagi, M. Takeichi, Cadherin superfamily genes: functions, genomic organization, and neurologic diversity, Genes Dev. 14 (2000) 1169 – 1180. [41] A.S. Yap, B.R. Stevenson, J.R. Keast, S.W. Manley, Cadherinmediated adhesion and apical membrane assembly define distinct steps during thyroid epithelial polarization and lumen formation, Endocrinology 136 (1995) 4672 – 4680.
Research Report
Cadherin-10 is a novel blood–brain barrier adhesion molecule in human and mouse Matthew J. Williams, Margaret B. Lowrie, Jonathan P. Bennett 1, J. Anthony Firth, Peter Clark* Division of Biomedical Sciences, Imperial College London, Sir Alexander Fleming Building, South Kensington, London SW7 2AZ, UK Accepted 27 July 2005 Available online 21 September 2005
Abstract Maintenance of the specialised environment of the central nervous system requires barriers provided by the endothelium of brain microvessels (the blood – brain barrier (BBB)) or the epithelium lining the ventricles (CSF – brain barrier) or the choroid plexus (blood – CSF barrier). Inter-endothelial junctions are more extensive in the BBB than in other tissues, with elaborate tight junctions. However, few differences in the molecular composition of these junctions have been described. Here, we show, in both human and mouse brain, that the type II classical cadherin, cadherin-10, is expressed in BBB and retinal endothelia, but not in the leaky microvessels of brain circumventricular organs (CVO), or in those of non-CNS tissues. This expression pattern is distinct from, and reciprocal to, VE-cadherin, which is reduced or absent in tight cortical microvessels, but present in leaky CVO vessels. In CVO, the barrier function is switched from the microvasculature to the adjacent ventricular epithelium, which we also find to express cadherin-10. In the vessels of gliobastoma multiforme tumours, where BBB is lost, cadherin-10 is not detected. This demonstration of a distinctive expression pattern of cadherin-10 suggests that it has a pivotal role in the development and maintenance of brain barriers. D 2005 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Blood – brain barrier Keywords: Endothelial cell; Cadherin; Capillary; Cell junction
1. Introduction Cadherins are a large family of cell – cell adhesion molecules involved in inter-cellular adhesion in a wide variety of cell types [4,40]. In the nervous system, cadherins are known to be crucial to all stages of development, including the early separation of the neural tube from the ectoderm, the segregation of neurones and axons, and the formation of synapses [32,33]. These molecules mediate calcium ion-dependent, homotypic
Abbreviations: BBB, blood – brain barrier; CVO, circumventricular organ; UEA-1, Ulex europaeus agglutinin-1; Glut-1, glucose transporter-1 * Corresponding author. E-mail address: [email protected] (P. Clark). 1 Present address: Hull York Medical School, University of Hull, HU6 7RX. 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.07.078
adhesion, and have roles in cell – cell recognition, the segregation of cells and tissues, and the formation of intercellular junctions [21]. Cadherins are typically integralmembrane glycoproteins containing multiple, tandemrepeated cadherin ectodomains, a transmembrane domain, and a cytoplasmic domain that, in many instances, interacts with the actin cytoskeleton. Large numbers of different cadherins from various cadherin sub-families are expressed in the central nervous system, and are believed to be key in the development of the complex neuronal circuitry and patterns of synapses [32,40]. However, many of the cadherins that have been identified in the CNS have not been fully characterised. While studying the materno – foetal barrier, using a culture model of Human Placental Microvascular Endothelial Cells developed in our laboratory [14], we screened for classical cadherin expression using RT-PCR with degenerate
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
primers designed using two highly conserved amino acid sequences in the cytoplasmic domain. One of the cadherins identified was cadherin-10 (unpublished observations). Cadherin-10 was first partially cloned from human brain [31], but its mRNA has also been shown to be present in mouse thymus (designated T2-cadherin), mouse testis, and in the developing brain and eye of mouse, rat, and chick [5,15,17,18,25]. The full-length human cDNA of cadherin10 was sequenced, and the predicted protein was shown to be a type-2, classical cadherin whose mRNA is expressed predominantly in brain, with lesser expression in kidney [22]. The studies of the expression and distribution of cadherin-10 mRNA have led to it being generally regarded as a brain-cadherin, and have suggested that it is one of a battery of cadherins involved in the establishment and maintenance of axon tracts [5,18]. To date, no examination of the distribution of cadherin-10 protein has been undertaken. In this study, we have raised antiserum to cadherin10 and examined the expression of cadherin-10 protein in both human and mouse brain.
2. Materials and methods 2.1. Polymerase chain reaction (PCR) PCR amplification was performed on cDNAs from a range of human and mouse tissues (First Strand, Origene, US). Amplification was carried out with a GeneE thermal cycler (Techne) using Taq DNA polymerase (Qiagen, UK). Primers were synthesised by Invitrogen (UK). Primer sequences were as follows: human cadherin-10 forward 5V-CGACTCACTTGCAACCTA-3V and reverse 5V-GGGCAGGACATGTACCTA-3V; human GAPDH forward 5VGTTGAAGGTCGGAGTCAACG-3V and reverse 5V-CA AAGTTGTCATGGATGACC-3V; mouse cadherin-10 forward 5V-AAGGAGATGGATCGCTCA-3V and reverse 5VTTCACTGTGGTGGTTCCA-3V; and mouse GAPDH forward 5V-TGCTGAGTATGTCGTGGAGTCTA-3V and reverse 5V-AGTGGGAGTTGCTGTTGAAGTCG-3V. PCRamplified products were analysed and visualised on a 1.5% agarose gel using standard procedures. 2.2. Generation of cadherin-10 antibodies Cadherin-10-specific antigens were identified by comparison of amino acid sequences of related cadherins, and a peptide sequence in the third extracellular domain (QNTIHLRVLESSPV) was selected for being unique to cadherin-10, and for having minimal identity with other cadherins, using the Human Genome Mapping Project BLAST database (http://www.hgmp.mrc.ac.uk/) (Fig. 1c). Antiserum was obtained using the rabbit-inoculation service provided by Sigma-Genosys (UK), and cadherin10-specific antibodies affinity purified using a peptide-
63
conjugated column (Sulfo-Link affinity column, PerBio, UK). 2.3. Western blotting Brains were dissected from adult female ML1 mice freshly killed in accordance with UK Home Office regulations. Frozen, unfixed human brain tissue, from individuals unaffected by neurological disease, was kindly provided by the UK Multiple Sclerosis Tissue Bank, Imperial College London. Brain samples were placed in ice-cold Laemmli Buffer supplemented with protease inhibitors and phosphatase inhibitors, and prepared using standard procedures. Protein samples (20 Ag) were electrophoretically separated on an 8% acrylamide gel, and transferred by semi-dry electrophoresis onto Immobilon PVDF nitrocellulose membrane (Millipore, USA). Membranes were blocked with 5% non-fat dried milk in TBS/Tween (blocking buffer), incubated with rabbit cadherin-10 polyclonal antibodies overnight at 4 -C, and incubated with horseradish peroxidase-conjugated antirabbit secondary antibody (Santa Cruz, USA), 1 h at room temperature. Bands were visualised using Pierce Pico-West Super-Signal enhanced chemiluminescence (ECL) reagents (PerBio, UK), and were developed as autoradiographs using ECL photographic film (Amersham, UK). 2.4. Immunohistochemistry Brains and eyes were dissected from adult female MF1 mice transcardially perfused with 4% paraformaldehyde in PBS in accordance with UK Home Office regulations. Fixed tissues were embedded in paraffin wax, for immunoperoxidase staining, or embedded in OCT compound (BDH, UK) and snap-frozen for immunofluorescence staining, using standard procedures. Fixed human brain samples from individuals unaffected by neurological pathology were kindly provided by the UK Multiple Sclerosis Tissue Bank, Imperial College London. These samples were processed for wax-embedding or frozen as for mouse tissues above. Serial sections of human brain cortex, and serial coronal and sagittal sections of mouse brain, of 7 Am thickness were obtained using standard histological procedures. Sections of wax-embedded archival biopsies from three patients diagnosed with glioblastoma multiforme, and from one patient with low-grade astrocytoma, were kindly supplied by the Department of Neuropathology, Imperial College London, Charing Cross Hospital, London, UK. Wax-embedded human and mouse brain sections were re-hydrated using a series of ethanol solutions, and labelled using standard protocols. Endogenous peroxidases were inhibited by incubation in 3% hydrogen peroxide in methanol. Tissue sections were blocked in 5% normal goat serum (NGS), and incubated with 5 Ag ml 1 rabbit anti-
64
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
Fig. 1. Expression of cadherin-10 mRNA and protein in human and mouse tissues. (a) PCR products of human tissue templates using (i) cadherin-10 primers and (ii) GAPDH primers. (b) PCR products of mouse tissue templates using (i) cadherin-10 primers and (ii) GAPDH primers. (c) Table showing cadherin-10specific antigen peptide alignment with homologous regions of type I and II classical cadherins. Sequences from the National Library of Medicine Peptide database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi). (d) Western blots of human and mouse brain lysates probed for cadherin-10.
cadherin-10 affinity-purified antibody, rabbit IgG control, or no-primary control, in 5% NGS overnight at 4 -C with humidity. Sections were subsequently incubated in HRPconjugated secondary antibody (Santa Cruz, USA) 1 h at room temperature and incubated with DAB substrate solution (Vector Labs, UK), until colour development had occurred. One set of brain tumour biopsy sections was stained with anti-cadherin-10 antibodies, as above, and a second set, serial to the first, was stained with HRPconjugated Ulex europaeus agglutenin-1 (UEA-1) (Sigma, UK). Both sets were counter-stained with haematoxylin. Tissue sections were dehydrated and coverslips mounted onto slides using DPX (Histomount, National Diagnostics). Tissues were observed using a Nikon light microscope (Nikon, Japan), and images were captured using a Nikon digital camera and Nikon computer software. Frozen sections of mouse brain were stained with 5 Ag ml 1 affinity-purified rabbit polyclonal anti-cadherin-10 antibodies alone, or together with rat monoclonal anti-VEcadherin/CD144 (BD Pharmingen) antibodies at 1:50 dilution, or rat monoclonal anti-Platelet Endothelial Cell Adhesion Molecule-1 (PECAM-1)/CD31 (BD Pharmingen) antibodies at 1:50 dilution. Frozen human sections were
stained with rabbit polyclonal anti-cadherin-10 antibodies together with mouse monoclonal anti-human VE-cadherin (BV6) antibodies (kindly supplied by Drs. Stephan Leibner and E. Dejana, University of Milan, Italy) or FITCconjugated UAE-1 (Vector Labs, UK). Following standard protocols, specimens were permeabilised with 0.5% Triton X-100, blocked with 5% NGS, and incubated overnight at 4 -C in primary antibodies, purified immunoglobulin G (IgG), or without antibody. After washing in PBS, sections were incubated for 1 h at room temperature with 2 Ag ml 1 Alexa Fluor fluorescently-labelled secondary antibodies (Molecular Probes, Netherlands) or fluoroscein-conjugated UEA-1 (2 Ag ml 1, Vector Labs, UK). After further washing in PBS and water, sections were mounted onto coverslips using Vectashield Hard-Set Mounting Medium (Vector Labs, UK). Confocal images were collected using a Leica TCS NT microscope system (Leica Microsystems, Germany) with a krypton-argon laser. When comparing staining intensity in different regions of a tissue, conditions were kept constant during image collection, which were saved in digital format using TCS NT software. To test the specificity of the cadherin-10 antibodies, pre-adsorption of the affinity-purified cadherin-10 antibody with its peptide
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
antigen resulted in the abolition of any specific staining (not shown).
3. Results 3.1. Cadherin-10 mRNA expression in human and mouse PCR analysis of cDNAs derived from human and mouse showed that cadherin-10 mRNA can be amplified from a number of tissues (Figs. 1a, b), with the strongest signal, compared to the housekeeping gene loading-control GAPDH, being found in brain of both human (Fig. 1a) and mouse (Fig. 1b). These findings are in agreement with previous studies showing that cadherin-10 mRNA is expressed in mouse brain, testis, and thymus and in human brain, with lesser expression in kidney and lung [5,15,17, 18,22,25,31]. In addition, we show that cadherin-10 mRNA is also present in tissues of the gastro-intestinal and urogenital tracts. 3.2. Characterisation of cadherin-10 antibodies The peptide antigen used to immunise rabbits to produce anti-cadherin-10 antiserum was designed from a region of the third cadherin extracellular domain (EC3) of the predicted amino acid sequence of the previously cloned human cadherin-10 cDNA. This sequence is identical in mouse and rat cadherin-10, with three amino acid differences in the chicken (Fig. 1c). The sequence is highly divergent in other human cadherins, allowing us to be confident that the antiserum would not cross-react with
65
other known cadherins. The staining pattern of cadherin-10 in a variety of cells and tissues, that express a number of characterised cadherins, suggests that our antiserum does not cross-react with other known classical cadherins (unpublished observations). When we probed Western blots of human and mouse brain lysates, run in SDS-PAGE, with the affinity-purified rabbit antibodies to cadherin-10, we obtained single major bands at approximately 120 kDa (Fig. 1d). This size is consistent with the recognition of a classical cadherin, which is typically in the range 115 –130 kDa. Fainter bands observed in the mouse brain blot could indicate a higher molecular mass pre-protein and proteolysis products. A previous study immunoprecipitated transfected human cadherin-10, via h-catenin, from mouse L-cells, which was found to run as a band at 116 kDa [30]. 3.3. Cadherin-10 protein expression in human and mouse brain vasculature Immunohistochemical localisation of cadherin-10 in normal adult human brain revealed a pattern of distribution consistent with expression in microvessels of the neuropil, with little or no neural expression (Figs. 2a, b). The microvessel-specific nature of this expression can be seen in Fig. 2c where larger, pial vessels are negative, while positive structures in the adjacent sub-pial cortex have a microvascular morphology. The microvascular nature of the cadherin-10-positive structures in the cortex was confirmed by double-labelling of cadherin-10 with the UEA-1 lectin, which is known to bind specifically to human endothelial cells (Figs. 2e – h). The localisation of cadherin-10 relative
Fig. 2. Cadherin-10 distribution in human brain. Bright-field micrographs of immunoperoxidase-labelled sections of temporal cerebral cortex (a) and cerebellum (b) showing cadherin-10-positive microvessels in the neuropil. In the temporal cerebral cortex (c), cadherin-10 distribution within the vasculature is restricted to microvessels (MV), and is absent from larger pial blood vessels (PV). (d) IgG control; region shown in panel (c), no specific staining was found when tissue sections were incubated with an equivalent concentration of rabbit IgG. (e – g) Confocal micrographs of cortical microvessels in the neuropil showing distribution of cadherin-10 (e), UEA-1 (f), merged in panel (g), and corresponding phase contrast micrograph (h). Scale bar: 500 Am (a, b); 200 Am (c, d); 8 Am (e – h).
66
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
to UEA-1, where cadherin-10 staining is partially internal to UEA-1 staining, shows that cadherin-10 is expressed by the microvascular endothelium. In all microvessels examined in the cortical and cerebellar neuropil, cadherin-10 and UEA-1 were closely associated, with partial colocalisation. In mouse brain, immunohistochemical localisation of cadherin-10 resulted in a more complex pattern than in the human brain. Cadherin-10 was present in a microvascular distribution in cortex (Figs. 3a, b), but was also found to be present in larger intraventricular vessels and in pial vessels
(Figs. 3c, d). In addition, the lepto-meningeal surface of the cortex, but not the adjacent cerebellar surface, was positively stained (Fig. 3d). In the choroid plexus, capillaries were not stained, while the epithelium was strongly labelled (Fig. 3e). Double immunolabelling with cadherin10 and the endothelial marker, PECAM-1, showed that, in the majority of the larger vessels, cadherin-10 was absent from the endothelium, but present in peri-endothelial cells which we have found are a-smooth muscle actin positive (data not shown). In cortical microvessels, cadherin-10 and
Fig. 3. Cadherin-10 distribution in mouse brain. Bright-field micrographs of immunoperoxidase-labelled brain sections (a – f). Sub-pial cortex (a and b), with cadherin-10-positive microvessels (MV) within the neuropil. (c) 3rd ventricle, with cadherin-10-positive ependyma (E) and blood vessels (BV). (d) Pial surface and transverse fissure, lined with cadherin-10-positive cells on the cortical surface, but not on the adjacent cerebellar surface. (e) The epithelial cells of the choroid plexus of the 4th ventricle are strongly stained for cadherin-10. (f) IgG control; same region as shown in panel (e) with equivalent concentrations of rabbit IgG. (g) Diagram of mouse brain in sagittal section, showing the regions from which micrographs (a) – (f) were obtained. (h – p) Confocal micrographs of brain cortex showing localisation of cadherin-10 and PECAM-1. (h – j) Sub-pial cortex; cadherin-10 (h) and PECAM-1 (i) colocalise in microvessels (open arrows), but not in larger blood vessels (BV) (j). In cortical microvessels (k – m), cadherin-10 (k) and PECAM-1 (l) both co-localised in the merge (m). (n – p) Pial blood vessel stained for cadherin-10 (n), PECAM-1 (o), merged in panel (p); cadherin-10 was observed at inter-endothelial junctions (closed arrows), and in vascular smooth muscle (SM) cells. Scale bars: 100 Am (a, b), 500 Am (c – f), 50 Am (h – j), 10 Am (k – p).
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
PECAM-1 co-localised, suggesting that, as in the human brain, microvascular endothelial cells express cadherin-10 (Figs. 3k –m). In a small proportion (estimated to be 5% of pial vessels) of larger pial veins and venules, cadherin-10 could be seen to co-localise with PECAM-1 in a pattern suggesting that both are concentrated at inter-endothelial junctions (Figs. 3n –p). It is known that, in rodents, a proportion of pial vessels express blood – brain barrier markers and possess barrier characteristics similar to cortical microvessels [3,9]. 3.4. Cadherin-10 expression in circumventricular organs The microvessels of the circumventricular organs (CVOs) are known to lack blood –brain barrier characteristics. Barrier function in these areas of the brain is transferred from the microvessels to the adjacent ventricular ependymal cells that act as the local CSF –brain barrier, or the blood – CSF barrier in the choroid plexus [7]. Similarly, markers of blood – brain barrier (e.g. glucose transporter-1 (Glut-1)), which are present in the majority of brain microvascular endothelial cells, are absent from CVO
67
microvascular endothelium, but present in the adjacent ventricular ependymal cells [27]. When we examined the expression of cadherin-10 in CVO of both human and mouse, we observed that while epithelial structures were positively stained, microvessels were not. In human brain, we observed cadherin-10 in the ependymal lining cells of the floor of the 3rd ventricle (median eminence) (Figs. 4a, b), and in the epithelium of the choroid plexus (Figs. 5c, d). The UEA-1-positive choroid plexus endothelia are negative for cadherin-10 (Fig. 4c). In mouse brain, cadherin-10 is present in the ependymal lining of the third ventricle (Fig. 4e). More detailed examination of cadherin-10 distribution in the infundibular recess showed that the cadherin is present in the cells lining the lateral recesses (a2 and h1 tanycytes), with significantly lesser expression in the medial h2 tanycytes. This pattern of expression is identical to the known pattern of expression of the BBB marker, Glut-1 [27]. Strands of cadherin-10 positive material radiating from the epithelial lining of the infundibular process (Fig. 4f) are consistent with the tanycyte basal processes that project to the median eminence. Cadherin-10 and PECAM-1 do not co-localise
Fig. 4. Cadherin-10 in circumventricular organs in human and mouse brain. (a) Cadherin-10 in the human median eminence with (b) the corresponding phase contrast image. Ependymal cells (E) lining third ventricle (3rd Vent) are positive for cadherin-10, whereas few microvessels are stained for cadherin-10. (c) Cadherin-10 and UEA-1 localisation in the human choroid plexus with (d) the corresponding phase contrast image. Epithelial cells of choroid plexus + VE for cadherin-10 (red), while the UEA-1-positive endothelial cells (green) are negative for cadherin-10. (e – k) Cadherin-10 and PECAM-1 localisation in the mouse medial basal hypothalamus. (e) Medial basal hypothalamus (MBH), stained for cadherin-10. Cadherin-10-positive cells line the third ventricle (3rd Vent). (f – h) Enlarged region highlighted in (e) median eminence (M.E.), stained for (f) cadherin-10 (arrowheads indicate tanycyte basal processes), (g) PECAM-1, and (h) merge. (i – k) Enlarged image of region highlighted (f – g), stained for (i) cadherin-10, (j) PECAM-1, and (k) merge. Cuffs of cadherin-10-positive structures (as indicated) were observed surrounding PECAM-1-positive microvessels (MV). Scale bar: 100 Am (a – d); 250 Am (e); 40 Am (f – h), 15 Am (i – k).
68
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
Fig. 5. Comparison of the distributions of cadherin-10 and VE-cadherin in mouse and human brain. (a – c) Sub-pial cortical neuropil from mouse brain. Microvessels (MV) are positive for cadherin-10 (a), but negative for VE-cadherin (b), merged in panel (c). (d – f) 4th Ventricular choroid plexus and adjacent brainstem (BrSt) in the same tissue-section as panels (a) – (c). The epithelial cells of the choroid plexus (CP) are positive for cadherin-10, while the microvessels (MV) within the peri-ventricular neuropil and in the choroid plexus were negative for cadherin-10 (d). These endothelial cells were strongly stained for VEcadherin (e), merged in panel (f). (g – l) Confocal images double-labelled for cadherin-10 (red) and VE-cadherin (green) from the same section of human brain. (g – i) Cortical microvessel, cadherin-10 strongly localised to the cell – cell junctions in the MV endothelial cells (g, indicated by closed arrows), with VEcadherin being found at low levels (h), merged in panel (i). (j – l) Pial vessel taken from same tissue-section as panels (g) – (l). While the endothelial cells of the pial vessel are positive for VE-cadherin (k), with a strongly junctional distribution (arrows), cadherin-10 is not detected (j), merged in panel (l). Scale bar: 50 Am (a – f); 8 Am (g – l).
in the capillaries of the median eminence. PECAM-1positive endothelia of microvessels in the median eminence appear to be surrounded by a loose cuff of cadherin-10 staining (Figs. 4f –k) that may represent tanycyte basal processes wrapping capillaries. Cadherin-10 is expressed in choroid plexus epithelium, but not in its microvessels (Figs. 3e, 5d –f), again reflecting the known pattern of expression of the BBB marker, Glut-1 [27]. 3.5. Comparison of cadherin-10 and VE-cadherin distribution VE-cadherin is well established as the adhesion molecule of adherens junctions between endothelial cells [12,23,24], with the possible exception of those in brain vessels [6]. When we compared the distributions of cadherin-10 and
VE-cadherin in double-labelled sections of mouse brain, we found that, in mouse cerebral cortex, cadherin-10, but not VE-cadherin, is present in microvessels (Figs. 5a– c). In other areas of the same section, i.e. the choroid plexus of the fourth ventricle and the adjacent brainstem examined using identical settings in the confocal system, microvessels express VE-cadherin, but not cadherin-10 (Figs. 5d – f). It should be noted that cadherin-10 is present in the choroid plexus epithelial cells, which form the blood – CSF barrier, but not in its leaky endothelium. In human cerebral cortex, we found that the expression of VE-cadherin is consistent with previous findings [26]: VEcadherin staining is less intense in the microvessels than in the junctions of larger lepto-meningeal vessels in the same section (Figs. 5k– l). The expression of cadherin-10 and VEcadherin appears to be inversely related: lepto-meningeal
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
69
vessels with strong junctional labelling of VE-cadherin are negative for cadherin-10; cadherin-10-expressing cortical microvessels have significantly reduced labelling intensity. 3.6. Cadherin-10 expression in retinal microvessels In the mouse retina, cadherin-10 is found in two distinct locations: the retinal microvessel endothelial cells and the retinal pigmented epithelium (RPE) (Figs. 6a, d). The retinal microvessels and the RPE are known to be relatively impermeable, constituting the inner blood – retinal barrier (together with pericytes and retinal glial cells) and outer retinal barrier, respectively [11]. In retinal microvessels, cadherin-10 is co-expressed with PECAM-1 (Figs. 6a – c). However, unlike the mouse brain, where cadherin-10 and VE-cadherin are expressed in distinct sets of microvessels (Fig. 5), retinal microvessels express both cadherin-10 and VE-cadherin (Figs. 6d – f). While the microvessels of the choroid, the choroicapillaris, express both PECAM-1 and VE-cadherin more strongly than in the adjacent retina (Figs. 6b, e), cadherin-10 is absent in these microvessels (Figs. 6a, d). In contrast to the retinal vessels, the choroicapillaris is a highly fenestrated, leaky capillary bed.
Fig. 7. Comparison of cadherin-10 expression in normal cortex and glioblastoma multiforme (GBM) tumour in human. Normal cortex (a – b) and GBM tumour (c – d) in the same section of biopsy. Microvessels are positive for cadherin-10 in the unaffected region (arrows in panel a), but not in the GBM tumour (c). The presence of microvessels in both normal cortex and tumour is shown by UEA-1 staining (arrows in panels b and d). Scale bar: 150 Am.
3.7. Cadherin-10 expression in glioblastoma multiforme It is known that BBB properties are absent in the vessels of both primary and metastatic brain tumours [13,28,36,39]. We examined the expression of cadherin-10 in three cases of human glioblastoma multiforme (GBM). Immunohistochemical localisation of cadherin-10 in each of these tumour biopsies was compared to endothelial labelling by UEA-1 in
serial sections. In all cases, cadherin-10 was absent from GBM vessels. Cadherin-10 is present in the microvessels of normal cortex (Fig. 7a), but absent from vessels in adjacent tumour tissue in the same section (Fig. 7c). In a serial section of this case, UEA-1 labelling shows the presence of numerous vessels in both normal cortex and adjacent tumour tissue (Figs. 7b, d). In a single case of low-grade
Fig. 6. Cadherin-10 in mouse retina, showing distributions of cadherin-10 with PECAM-1 (a – c), and with VE-cadherin (d – f). (a – c) Microvessels (MV) within the retina are positive for both cadherin-10 (a) and PECAM-1 (b), as shown in merge (c), while cadherin-10 was absent from the PECAM-1-positive microvessels in the choroid. (d – f) Both cadherin-10 (d) and VE-cadherin (e) co-localise in retinal microvessels, (f) merge. Cadherin-10 is not expressed in VEcadherin-positive choroid microvessels. Retinal pigmented epithelia (RPE) are positive for cadherin-10, as shown in panels (a) and (d). Scale bar: 50 Am.
70
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
astrocytoma, where BBB properties are typically retained by the vessels [13], microvessels were positive for cadherin-10 (not shown). 3.8. Cadherin-10 in non-brain blood vessels We have examined cadherin-10 protein expression in a number of human (prostate, kidney, bladder, placenta) and mouse (kidney, bladder, prostate, heart, skin, skeletal muscle, placenta, GI-tract) tissues, but have never observed significant expression of cadherin-10 in the endothelia of any vessels (data not shown). In mouse, we observe that peri-endothelial cells express cadherin-10 in both the brain (Figs. 3h –j) and in vascular beds outside of the CNS (data not shown). We have never observed cadherin-10 staining in human vascular smooth muscle (data not shown).
4. Discussion Localisation of cadherin-10 protein shows that it is primarily expressed in vascular and epithelial structures of adult human and mouse brain. Previous studies have examined mRNA expression in the developing brain of mouse, rat, and chicken embryos, and implicated cadherin10 in inter-neuronal adhesion and the segregation of neural tracts [5,17,18]. However, we find that, in adult human and mouse brain, cadherin-10 is expressed by the endothelial cells of microvessels of the cerebral cortex, cerebellum and retina, the epithelial cells of the choroid plexuses, and the ependymal cells lining some ventricular regions. These structures form the bases of the brain barriers. In addition, the RPE cells, which act as the barrier between the retina and the outer layers of the eye [11], also express cadherin10. The pattern of cadherin-10 distribution is essentially identical to that of known markers of brain barriers, such as Glut-1. These markers are generally expressed in microvessel endothelium, choroid plexus epithelium, and in the ependymal cells adjacent to CVOs, while the leaky microvessels of CVOs lack BBB marker expression [7,8]. Cadherin-10 is therefore a rare example of a cell adhesion molecule present in BBB endothelia, but absent from the endothelia of leaky brain vessels and non-brain vessels. The absence of cadherin-10 from the vessels of GBM tumours, where the BBB is known to be compromised, provides further evidence of a key role for this cell adhesion molecule in BBB organisation. In vertebrates, the structural basis of the blood – brain barrier (BBB) is the endothelial lining of brain capillaries [2]. Under the influence of local factors in the CNS, particularly astrocytes, the endothelial cells of the BBB form tighter and more extensive inter-cellular junctions than those found in non-brain capillary beds [1,29]. While it is clear that tight junctions (TJ) are the structures that restrict paracellular passage across epithelial and endothelial cell
layers [19,38], their formation, organisation, and maintenance are believed to depend on cadherin-mediated adhesion, and on the formation of adherens junctions [10,16, 20,34,35,37,41]. In spite of the difference in the complexity of their inter-endothelial junctions, few differences in the repertoire of junction adhesion molecules expressed by BBB and non-brain endothelial cells have been found [26]. It has been shown, in the mouse, that the general endothelial cadherin found at inter-cellular junctions, VE-cadherin (cadherin-5) [12,23,24], is lost from brain microvessels, as the BBB is being established [6]. Recently, it has been found that the tight junction component, claudin-3, which has not been previously reported in endothelia, is present in BBB endothelia [39]. VE-cadherin is generally regarded as the major cadherin of inter-endothelial junctions [12,23,24], but it would appear that this is not the case in the brain. It has been shown that both VE-cadherin mRNA and protein are down-regulated or absent in mouse brain vessels, at a point in development when the BBB is being established [6]. In mouse brain endothelium, we find that cadherin-10 and VE-cadherin are expressed in a mutually exclusive manner: cadherin-10 is present in vessels of known tight BBB vessels, while VEcadherin is expressed in vessels that are known to be leaky. Interestingly, cadherin-10 and VE-cadherin are co-expressed in mouse retinal microvessels, which may indicate that these vessels are an intermediate phenotype. In the human brain, although no qualitative differences in the expression of a wide range of cell adhesion molecules have been observed between cerebral cortical microvessels and non-brain microvessels, VE-cadherin expression in brain microvessels was relatively reduced compared to that in larger vessels [26], despite the known increase in the complexity of BBB microvessel junctions. We also observed this difference in staining intensity between cortical microvessels and the endothelium of lepto-meningeal vessels, and we found that there is an inverse relationship between the expression of cadherin-10 and VE-cadherin. Cadherin-10 expression by BBB endothelia and its absence from CVO vessels indicate that this molecule might be responsible for replacing VEcadherin (at least partially) in inter-endothelial junctions, and may be part of the mechanism of inducing the elaboration of BBB junctions. Cadherin-10 expression in the epithelial barriers of the CNS (ependyma of CVO, the choroid plexus epithelium, and in the RPE of the outer retinal barrier) suggests a more general barrier function for this molecule. Our observation that cadherin-10 is present at the surface of mouse brain cortex, but not the surface of the cerebellum, has revealed possible differences in their organisation. We, as yet, do not know the significance of this, but might speculate that there is heterogeneity of brain surface barriers. In conclusion, we find that the pattern of expression of cadherin-10 is consistent with a key role in forming and maintaining the barriers that are crucial to the normal functioning of central nervous system. The loss of BBB is a
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
key contribution to the pathology of many brain tumours, where oedema is a major complication. The loss of cadherin-10 in the vessels of GBM tumours indicates the involvement of this cell adhesion molecule in pathological processes. It will be interesting to examine cadherin-10 expression in other pathologies where BBB is affected. The tight correlation between the presence of cadherin-10 and known barrier function strongly suggests that cadherin-10 may have unique properties.
Acknowledgments We thank Lorraine Lawrence for histology technical support; the UK Multiple Sclerosis Tissue Bank, Imperial College London, for providing human brain material; Dr Federico Roncaroli, Department of Neuropathology, Imperial College London, for providing material and advice on human brain tumours; and Stefan Liebner and Elisabetta Dejana for anti-human VE-cadherin antibodies. This work was supported by the Wellcome Trust.
References [1] N.J. Abbott, Astrocyte-endothelial interactions and blood – brain barrier permeability, J. Anat. 200 (2002) 629 – 638. [2] N.J. Abbott, Dynamics of CNS barriers: evolution, differentiation, and modulation, Cell Mol. Neurobiol. 25 (2005) 5 – 23. [3] G. Allt, J.G. Lawrenson, Is the pial microvessel a good model for blood – brain barrier studies? Brain Res. Brain Res. Rev. 24 (1997) 67 – 76. [4] B.D. Angst, C. Marcozzi, A.I. Magee, The cadherin superfamily: diversity in form and function, J. Cell Sci. 114 (2001) 629 – 641. [5] I.H. Bekirov, L.A. Needleman, W. Zhang, D.L. Benson, Identification and localization of multiple classic cadherins in developing rat limbic system, Neuroscience 115 (2002) 213 – 227. [6] G. Breier, F. Breviario, L. Caveda, R. Berthier, H. Schnurch, U. Gotsch, D. Vestweber, W. Risau, E. Dejana, Molecular cloning and expression of murine vascular endothelial-cadherin in early stage development of cardiovascular system, Blood 87 (1996) 630 – 641. [7] J.E. Bruni, Ependymal development, proliferation, and functions: a review, Microsc. Res. Tech. 41 (1998) 2 – 13. [8] J.P. Cassella, J.G. Lawrenson, G. Allt, J.A. Firth, Ontogeny of four blood – brain barrier markers: an immunocytochemical comparison of pial and cerebral cortical microvessels, J. Anat. 189 (Pt. 2) (1996) 407 – 415. [9] J.P. Cassella, J.G. Lawrenson, L. Lawrence, J.A. Firth, Differential distribution of an endothelial barrier antigen between the pial and cortical microvessels of the rat, Brain Res. 744 (1997) 335 – 338. [10] R.G. Contreras, L. Shoshani, C. Flores-Maldonado, A. Lazaro, A.O. Monroy, M.L. Roldan, R. Fiorentino, M. Cereijido, E-Cadherin and tight junctions between epithelial cells of different animal species, Pflugers Arch. 444 (2002) 467 – 475. [11] J.G. Cunha-Vaz, The blood – retinal barriers system. Basic concepts and clinical evaluation, Exp. Eye Res. 78 (2004) 715 – 721. [12] E. Dejana, R. Spagnuolo, G. Bazzoni, Interendothelial junctions and their role in the control of angiogenesis, vascular permeability and leukocyte transmigration, Thromb. Haemostasis 86 (2001) 308 – 315. [13] M. Demeule, A. Regina, B. Annabi, Y. Bertrand, M.W. Bojanowski, R. Beliveau, Brain endothelial cells as pharmacological targets in brain tumors, Mol. Neurobiol. 30 (2004) 157 – 183.
71
[14] J.F. Dye, L. Leach, P. Clark, J.A. Firth, Cyclic AMP and acidic fibroblast growth factor have opposing effects on tight and adherens junctions in microvascular endothelial cells in vitro, Microvasc. Res. 62 (2001) 94 – 113. [15] B.E. Faulkner-Jones, L.N. Godinho, S.S. Tan, Multiple cadherin mRNA expression and developmental regulation of a novel cadherin in the developing mouse eye, Exp. Neurol. 156 (1999) 316 – 325. [16] T.P. Fleming, T. Papenbrock, I. Fesenko, P. Hausen, B. Sheth, Assembly of tight junctions during early vertebrate development, Semin. Cell Dev. Biol. 11 (2000) 291 – 299. [17] D. Fushimi, K. Arndt, M. Takeichi, C. Redies, Cloning and expression analysis of cadherin-10 in the CNS of the chicken embryo, Dev. Dyn. 209 (1997) 269 – 285. [18] O.D. Gil, L. Needleman, G.W. Huntley, Developmental patterns of cadherin expression and localization in relation to compartmentalized thalamocortical terminations in rat barrel cortex, J. Comp. Neurol. 453 (2002) 372 – 388. [19] L. Gonzalez-Mariscal, A. Betanzos, P. Nava, B.E. Jaramillo, Tight junction proteins, Prog. Biophys. Mol. Biol. 81 (2003) 1 – 44. [20] B. Gumbiner, B. Stevenson, A. Grimaldi, The role of the cell adhesion molecule uvomorulin in the formation and maintenance of the epithelial junctional complex, J. Cell. Biol. 107 (1988) 1575 – 1587. [21] K.A. Knudsen, A.P. Soler, Cadherin-mediated cell – cell interactions, Methods Mol. Biol. 137 (2000) 409 – 440. [22] P. Kools, K. Vanhalst, E. Van den Eynde, F. van Roy, The human cadherin-10 gene: complete coding sequence, predominant expression in the brain, and mapping on chromosome 5p13 – 14, FEBS Lett. 452 (1999) 328 – 334. [23] M.G. Lampugnani, M. Resnati, M. Raiteri, R. Pigott, A. Pisacane, G. Houen, L.P. Ruco, E. Dejana, A novel endothelial-specific membrane protein is a marker of cell – cell contacts, J. Cell Biol. 118 (1992) 1511 – 1522. [24] L. Leach, P. Clark, M.G. Lampugnani, A.G. Arroyo, E. Dejana, J.A. Firth, Immunoelectron characterisation of the inter-endothelial junctions of human term placenta, J. Cell. Sci. 104 (Pt. 4) (1993) 1073 – 1081. [25] S.B. Munro, A.J. Duclos, A.R. Jackson, M.G. Baines, O.W. Blaschuk, Characterization of cadherins expressed by murine thymocytes, Cell. Immunol. 169 (1996) 309 – 312. [26] E. Navratil, A. Couvelard, A. Rey, D. Henin, J.Y. Scoazec, Expression of cell adhesion molecules by microvascular endothelial cells in the cortical and subcortical regions of the normal human brain: an immunohistochemical analysis, Neuropathol. Appl. Neurobiol. 23 (1997) 68 – 80. [27] B. Peruzzo, F.E. Pastor, J.L. Blazquez, K. Schobitz, B. Pelaez, P. Amat, E.M. Rodriguez, A second look at the barriers of the medial basal hypothalamus, Exp. Brain Res. 132 (2000) 10 – 26. [28] J.C. Reijneveld, E.E. Voest, M.J. Taphoorn, Angiogenesis in malignant primary and metastatic brain tumors, J. Neurol. 247 (2000) 597 – 608. [29] W. Risau, S. Esser, B. Engelhardt, Differentiation of blood – brain barrier endothelial cells, Pathol. Biol. (Paris) 46 (1998) 171 – 175. [30] Y. Shimoyama, G. Tsujimoto, M. Kitajima, M. Natori, Identification of three human type-II classic cadherins and frequent heterophilic interactions between different subclasses of type-II classic cadherins, J. Biochem. 349 (2000) 159 – 167. [31] S. Suzuki, K. Sano, H. Tanihara, Diversity of the cadherin family: evidence for eight new cadherins in nervous tissue, Cell Regul. 2 (1991) 261 – 270. [32] M. Takeichi, H. Matsunami, T. Inoue, Y. Kimura, S. Suzuki, T. Tanaka, Roles of cadherins in patterning of the developing brain, Dev. Neurosci. 19 (1997) 86 – 87. [33] M. Takeichi, T. Uemura, Y. Iwai, N. Uchida, T. Inoue, T. Tanaka, S.C. Suzuki, Cadherins in brain patterning and neural network formation, Cold Spring Harbor Symp. Quant. Biol. 62 (1997) 505 – 510.
72
M.J. Williams et al. / Brain Research 1058 (2005) 62 – 72
[34] M.L. Troxell, Y.T. Chen, N. Cobb, W.J. Nelson, J.A. Marrs, Cadherin function in junctional complex rearrangement and posttranslational control of cadherin expression, Am. J. Physiol. 276 (1999) C404 – C418. [35] M.L. Troxell, S. Gopalakrishnan, J. McCormack, B.A. Poteat, J. Pennington, S.M. Garringer, E.E. Schneeberger, W.J. Nelson, J.A. Marrs, Inhibiting cadherin function by dominant mutant E-cadherin expression increases the extent of tight junction assembly, J. Cell. Sci. 113 (Pt. 6) (2000) 985 – 996. [36] P. Vajkoczy, M.D. Menger, Vascular microenvironment in gliomas, Cancer Treat. Res. (2004) 249 – 262. [37] M. Watabe-Uchida, N. Uchida, Y. Imamura, A. Nagafuchi, K. Fujimoto, T. Uemura, S. Vermeulen, F. van Roy, E.D. Adamson, M. Takeichi, Alpha-catenin – vinculin interaction functions to organize the apical junctional complex in epithelial cells, J. Cell Biol. 142 (1998) 847 – 857.
[38] H. Wolburg, A. Lippoldt, Tight junctions of the blood – brain barrier: development, composition and regulation, Vasc. Pharmacol. 38 (2002) 323 – 337. [39] H. Wolburg, K. Wolburg-Buchholz, J. Kraus, G. Rascher-Eggstein, S. Liebner, S. Hamm, F. Duffner, E.H. Grote, W. Risau, B. Engelhardt, Localization of claudin-3 in tight junctions of the blood – brain barrier is selectively lost during experimental autoimmune encephalomyelitis and human glioblastoma multiforme, Acta Neuropathol. (Berl.) 105 (2003) 586 – 592. [40] T. Yagi, M. Takeichi, Cadherin superfamily genes: functions, genomic organization, and neurologic diversity, Genes Dev. 14 (2000) 1169 – 1180. [41] A.S. Yap, B.R. Stevenson, J.R. Keast, S.W. Manley, Cadherinmediated adhesion and apical membrane assembly define distinct steps during thyroid epithelial polarization and lumen formation, Endocrinology 136 (1995) 4672 – 4680.