- Email: [email protected]
Growth and characterisation of a cell culture model of the feline blood–brain barrier
Veterinary Immunology and Immunopathology 109 (2006) 233–244 www.elsevier.com/locate/vetimm
Growth and characterisation of a cell culture model of the feline blood–brain barrier Nicola F. Fletcher a,c,*, David J. Brayden b,c, Brenda Brankin c, Sheila Worrall a, John J. Callanan a,c a b
Department of Veterinary Pathology, University College Dublin, Belfield, Dublin 4, Ireland Small Animal Clinical Studies, Faculty of Veterinary Medicine, University College Dublin, Belfield, Dublin 4, Ireland c Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland Received 1 March 2005; received in revised form 15 July 2005; accepted 15 August 2005
Abstract An in vitro model of the feline blood–brain barrier was developed using primary cultures of brain capillary endothelial cells derived from adult cats. They were grown in the presence of astrocytes obtained from newborn kittens. Feline endothelial cell cultures were characterised by uptake of DiI-acetylated low-density lipoprotein (DiI-Ac-LDL) and expression of von Willebrand factor. Astrocytes were characterised based on their expression of glial fibrillary acidic protein (GFAP). Electron microscopy revealed junctional specialisation between endothelial cells. Occludin and ZO-1 expression by the endothelial cell cultures was detected by Western blot analysis. Barrier function of co-cultured endothelial cells and astrocytes was confirmed by a transendothelial electrical resistance (TEER) value of 30–35 V cm2 and apparent permeability coefficients (Papp) for FD-40 (FITC-dextran, 40 kDa) of 4 106 cm/s and for FD-4 (4 kDa) of 1.92 105 cm/s. In endothelial cell monolayers grown with astrocyte-conditioned medium, the TEER value was lower (20–25 V cm2), and Papp of FD-40 and FD-4 was higher at 6.27 106 and 3.96 105 cm/s, respectively. This model should have useful applications in the examination of events occurring at the BBB early in FIV infection, and may provide knowledge applicable to HIV infection. # 2005 Elsevier B.V. All rights reserved. Keywords: Blood–brain barrier; Endothelial cell monolayers; FIV neuropathogenesis; Endothelial cell tight junctions
Abbreviations: ACM, astrocyte-conditioned medium; BBB, blood–brain barrier; CNS, central nervous system; DiI-Ac-LDL, DiI-acetylated low-density lipoprotein; DMEM, Dulbecco’s modified eagle’s medium; ECGF, endothelial cell growth factor; FBEC, feline brain endothelial cells; FIV, feline immunodeficiency virus; FD, FITC-dextran; GFAP, glial fibrillary acidic protein; HRP, horseradish peroxidase; Papp, apparent permeability; PET, polyethylene terephthalate; SIV, simian immunodeficiency virus; TEER, transendothelial electrical resistance; TEM, transmission electron microscopy; TBS, tris buffered saline; ZO-1, zonula occludens-1 * Corresponding author. Tel.: +353 1 7166150; fax: +353 1 7166165. E-mail address: [email protected] (N.F. Fletcher). 0165-2427/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.vetimm.2005.08.025
234
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
1. Introduction The blood–brain barrier (BBB) regulates the passage of molecules between the CNS and the peripheral circulation (Broman, 1955). Tight junctions between cerebral capillary endothelial cells represent the structural basis of this barrier. Under normal physiological conditions, tight junctions form a continuous, almost impenetrable barrier that prevents the passive influx of a variety of substances with the exception of some small hydrophilic molecules (Reese and Karnovsky, 1967). Transcellular absorption for lipophilic agents, and carrier-mediated transport across endothelial cells, are the predominant routes of entry of substances to the brain. Astrocytes take an active part in the formation of the BBB by influencing differentiation of endothelial cells, by promoting high resistance tight junction formation in vivo (Butt, 1995). They also induce decreased pinocytotic activity and fenestrations in endothelial cells (Bauer and Bauer, 2000). Astrocytic endfeet ensheath the brain capillaries, and chemical signals released by these cells are in part responsible for the unique characteristics of the vertebrate BBB (Brightman, 1991; Rubin et al., 1991; Demeuse et al., 2002). The earliest in vitro BBB tissue culture systems were cultured endothelial cells of cerebral or noncerebral origin (Bowman et al., 1983; Audus and Borchardt, 1986; Kondo et al., 1994; Stanness et al., 1996). Currently, the most widely used in vitro BBB models across a range of species are brain capillary endothelial cells cultured with astrocyte-conditioned medium (ACM) (Rubin et al., 1991; Gaillard et al., 2001) or co-cultured with astrocytes (Dehouck et al., 1990; Gaillard et al., 2001). These models result in the formation of endothelial tight junctions and induce the expression of specific BBB markers on the apical side of endothelial cells, including g-glutamyl transpeptidase, the GLUT-1 glucose transporter and P-glycoprotein (Abbott and Revest, 1991). These models have been grown on permeable filters (Gaillard et al., 2001; Rubin et al., 1991; Demeuse et al., 2002) or cultured under flow conditions, where endothelial cells are cultured in the lumen of hollow fibres while astrocytes are seeded in the extraluminal compartment (Cucullo et al., 2002; Marroni et al., 2003). Growth of in vitro BBB models has been described for various mammalian species. Most commonly,
bovine endothelial cells have been cultured with rat astrocytes or rat ACM (Dehouck et al., 1990; Rubin et al., 1991; Gaillard et al., 2001), but other models utilising autologous cell cultures have been developed from the rat (Demeuse et al., 2002), mouse (Song and Pachter, 2003), human (Kim et al., 2003), porcine (Jeliazkova-Mecheva and Bobilya, 2003) and rhesus macaque (MacLean et al., 2002). Many of these models have been developed in order to study drug transport to the CNS (Audus and Borchardt, 1986; Rubin et al., 1991; Gaillard et al., 2001), or have been developed to study aspects of viral neuropathogenesis, including HIV-1 (Megard et al., 2002) and SIV infections (MacLean et al., 2002). HIV-1 infection in man is associated with neuropathology both in the early and late stages of infection (Annunziata, 2003). However, due to poor access to significant quantities of brain tissue in early infection, there is a paucity of information on how virus enters the brain and how it interacts with neurological tissue to induce significant neurological impairment. Animal models of HIV infection provide a useful way to study the pathogenesis of HIV, and the feline lentiviral model, FIV of the domestic cat, has been established as an excellent model of HIV-1 infection (Bendinelli et al., 1995; Callanan, 1995; Henriksen et al., 1995). FIV resembles HIV in many respects, but in particular by its ability to infect the CNS early after experimental infection. It is known that FIV enters the brain in the first week following infection (Boche et al., 1996; Ryan et al., 2003). Changes in the CNS, including perivascular mononuclear infiltrates, have been extensively described in HIV, SIV and FIV infections (Ringler et al., 1988; Dow et al., 1990; Connor and Ho, 1992; Sharer, 1992). However, the mechanism of entry of HIV, SIV and FIV to the brain is still a matter of debate. One mechanism of lentiviral entry to the CNS may be via breaches in the BBB. In support of this hypothesis, endothelial cell tight junction disruption has been demonstrated as a key feature in HIV and SIV encephalitis (Dallasta et al., 1999; Luabeya et al., 2000). In order to study the interaction of FIV with cells of the BBB, we have developed and characterised an in vitro model of the feline BBB using autologous cultures of primary feline brain capillary endothelial
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
cells and feline astrocytes. The endothelial cells were grown either as co-cultures on permeable filters with astrocytes or alternatively, as monolayers in the presence of astrocyte-conditioned medium. This model should be useful in the examination of immunological events at the BBB early in FIV infection, and may provide knowledge applicable to HIV infection.
2. Materials and methods 2.1. Materials Disposable sterile plasticware was purchased from Sarstedt (Sarstedt AG & Co., Nu¨mbrecht, Germany), Sigma (Sigma Chemical Co., St. Louis, MO, USA), Corning Costar (Cambridge, MA, USA) and BectonDickinson (Boston, MA, USA). Cell culture media and supplements, PBS, HBSS, bovine fibronectin, Trypsin–EDTA and Foetal Calf Serum were obtained from Gibco BRL Life Technologies Inc. (Grand Island, NY, USA). Collagenase-dispase was obtained from Roche. DiI-acetylated low-density lipoprotein (DiI-Ac-LDL) was obtained from Molecular Probes (Eugene, OR, USA), rabbit anti-GFAP antibody, rabbit anti-von Willebrand Factor antibody and swine anti-rabbit FITC-conjugated secondary antibody from Dako (Cambridge, UK), rabbit anti-ZO-1 antibody from Santa Cruz Biotechnology (Santa Cruz, CA, USA), rabbit anti-occludin antibody from Zymed (San Francisco, CA, USA) and Vectastain Elite ABC Rabbit IgG Kit and Vectashield mounting medium from Vector Laboratories (Burlingame, CA, USA). Nylon mesh for the isolation of primary cultures was obtained from Clarcor (Rockford, IL, USA). Thermanox coverslips were obtained from Nunc (Rochester, NY, USA). Fluorescein isothiocyanate-dextran-4 (FD-4), FD-40 and Triton X-100 were obtained from Sigma (St. Louis, MO). 2.2. Cell culture Feline brain capillary endothelial cells (FBEC) were isolated from euthanased adult cats, and cultured within 2 h of death according to the method of Rubin et al. (1991), but with some modifications. Freshly harvested feline cortical tissue was immersed
235
in ice-cold Hanks balanced salt solution (HBSS) supplemented with Penicillin/Streptomycin and Amphotericin B. Meninges and white matter were removed with a scalpel and grey matter was fragmented and collected in 10 ml DMEM with 20% heat inactivated foetal calf serum (FCS). The brain tissue was homogenised using a Polytron homogeniser (Kinematica GmbH, Switzerland), fitted with a 7-mm probe, and the resulting homogenate was then filtered through a 125 mm nylon mesh, and digested in collagenase and dispase for 1 h at 37 8C. The cells were cultured in vented 25 cm2 tissue culture flasks (Sarstedt AG & Co., Nu¨mbrecht, Germany) with DMEM/20% heat inactivated FCS and incubated at 37 8C with 5% CO2. The media was replaced every other day. After 4 days in culture, the media was supplemented with 10 ml/ml endothelial cell growth factor (ECGF; Sigma Chemical Co.). After approximately 7 days in culture, several large colonies of endothelial cells were visible in the culture flasks. These colonies were isolated, replated on collagen-1 coated tissue culture flasks and cultured with DMEM/ 20% heat inactivated FCS supplemented with 10 ml/ ml ECGF for approximately 48 h or until the cells were 80–90% confluent. Cells were then passaged with Trypsin–EDTA and stored in liquid nitrogen. Feline astrocytes were isolated from newborn specific pathogen free (SPF) kittens. Isolated cerebral cortices of each brain were fragmented and filtered first through 80 mm and then through 10 mm nylon meshes, and cultured for 4 days in DMEM/10% heat inactivated FCS. After 7 days in culture, the flasks were passaged with Trypsin–EDTA in a split ratio of 1:4 and grown to confluence. Some cultures possessed contaminating cells, and these were purified according to the method of McCarthy and de Vellis (1980). In brief, cultures containing contaminating cells were shaken for 12 h on a plate shaker at 37 8C. This removed the loosely adherent contaminating cells, such as oligodendrocytes, while leaving the astrocyte layer intact. The medium was then replaced and the remaining astrocytes were cultured until they reached confluence. For collection of ACM, cultures that were confluent were cultured with DMEM/10% heat inactivated FCS and the medium was collected twice a week, sterile filtered, and either used immediately or stored at 20 8C. After 2 weeks of conditioned medium
236
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
collection, astrocytes were passaged and stored in freeze mix (10% DMSO in FCS) in liquid nitrogen for co-culture purposes. The in vitro co-culture BBB model was prepared on rat-tail collagen and bovine fibronectin coated polyethylene terephthalate (PET) filters (surface area 0.9 cm2; pore size 1 mm; Becton-Dickinson). Rat-tail collagen was prepared from the tails of adult Wistar rats according to the method of Strom and Michalopoulos (1982). Astrocytes were seeded on the bottom of the filter at a density of 5 105 cells per filter. The filters were inverted and astrocytes were allowed to adhere to the bottom of the filter for 2 h. The filters were then placed in 12 well plates, and the astrocytes were fed with DMEM/20% FCS. After 72 h, endothelial cells were seeded at a density of 1.5 105 cells per filter onto the apical side of each filter. The medium was replaced every 2 or 3 days. Endothelial cell monolayers were also cultured on porous filters as described above, but with 50% (v/v) astrocyte-conditioned medium added to the culture medium. Experiments with the in vitro BBB models were performed 7 days after the endothelial cells were seeded. 2.3. Immunocytochemical analysis of BBB model After 7 days in culture, filters containing endothelial cell cultures were fixed in ice-cold methanol for immunocytochemistry and immunofluorescence. Cells were permeabilised with 0.05% Triton X-100 for 30 min. Non-specific binding of antibody was blocked by incubation with 5% normal goat serum for 10 min. Primary antibody was incubated at 4 8C for 12 h and washed with PBS. Primary antibodies used were polyclonal anti-rabbit antibody specific for von Willebrand factor (1:1000; Dako, Cambridge, UK) for FBEC, and polyclonal anti-rabbit antibody specific for GFAP (Undiluted; Dako) for feline astrocytes. For immunofluorescence, secondary antibody (swine antirabbit FITC) was applied for 1 h at 37 8C. For immunocytochemistry, detection of the primary antibodies was performed using the Vectastain Elite ABC kit (Rabbit IgG) as per manufacturer’s instructions. Diaminobenzidine was used as the chromagen. Filters were counterstained with haematoxylin, dehydrated and mounted in DPX (Sigma Chemical Co.) for immunocytochemistry, or Vectashield mounting med-
ium for immunofluorescence. Negative controls were treated as above, but an irrelevant isotype-matched antibody was used in place of the primary antibody. 2.4. Uptake of DiI-Ac-LDL Cultures of FBEC were cultured on collagen and fibronectin coated Thermanox1 coverslips for 2–3 days. The coverslips were then washed with DMEM, and the medium was replaced with DMEM containing 10 mg/ml DiI-Ac-LDL, and cultured for 4 h at 37 8C with 5% CO2. The coverslips were then washed with PBS, fixed with 2% paraformaldehyde for 10 min at room temperature and mounted with Vectashield mounting medium. 2.5. Western blot analysis The presence of occludin and ZO-1 proteins in the in vitro BBB was determined by Western blot analysis. Occludin and ZO-1 protein was isolated from filters containing confluent monolayers of FBEC. ARPE-19 cells (human retinal pigment epithelial cells) and Mya-1 cells (feline T-lymphoblastoid cells) were used as positive and negative controls, respectively. Three confluent monolayers of FBEC or ARPE-19 cells (approximately 5 105 cells), or 5 105 Mya-1 cells, were used for each experiment. The filters were removed from the filter inserts and washed with PBS. The cell monolayers were scraped and homogenised with a 21-gauge needle in boiling buffer containing 62.5 mM Tris, 2% SDS, 10 mM dithiothreitol, 10 ml protease inhibitor cocktail/100 ml (Sigma Chemical Co.). The homogenate was centrifuged at 10,000 g for 20 min at 4 8C, and the resulting pellet was used for occludin or ZO-1 analysis. Protein concentrations were determined using the Micro-BCA1 assay (Pierce, MSC, Dublin, Ireland) assay. A BSA standard curve and equal amounts of protein were loaded onto SDS-PAGE gels. 10% and 6% SDS-PAGE was used for occludin and ZO-1 analyses, respectively. Protein samples were separated by electrophoresis, and transferred to nitrocellulose membrane for 2 h using a semi-dry electroblot apparatus (Bio-Rad). Efficiency of protein transfer was determined using Ponceau-S solution (Sigma Chemical Co.). Non-specific binding sites were blocked by incubating the membrane at room
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
temperature with 5% non-fat dry skimmed milk in TBS (0.05 M Tris, 150 mM NaCl, pH 7.5) for 2 h. Membranes were briefly washed with TBS, and incubated with polyclonal anti-rabbit antibody (Zymed) specific for occludin (1:1000) or polyclonal anti-rabbit antibody (Santa-Cruz Biotech, CA) specific for ZO-1 (1:1000). Antibodies were incubated with membranes overnight at 4 8C. The membranes were washed with TBS, and incubated with a secondary anti-rabbit (IgG) antibody with a HRP conjugate (1:2500), for 3 h at room temperature. Immune complexes were detected using enhanced chemiluminescence (ECL). 2.6. Transmission electron microscopy Filters containing confluent monolayers of endothelial cells with or without astrocyte co-cultures, or cultured with astrocyte-conditioned medium were fixed at room temperature for 1 h with glutaraldehyde (2.5%, v/v) in phosphate buffer, and postfixed in osmium tetroxide (2%, w/v) in 0.05 M potassium phosphate buffer for 30 min. The filters were dehydrated through a graded series of ethanols to 100% and embedded in epoxy resin. Sections were cut and observed through a Hitachi H-7000 transmission electron microscope. 2.7. Transendothelial electrical resistance measurements Restriction of paracellular transport of small ions was determined indirectly by calculation of transendothelial electrical resistance (TEER). TEER was measured using an Endohm Chamber connected to an EVOM Voltohmeter (World Precision Instruments Inc., New Haven, CT). TEER (V cm2) was calculated by subtraction of the electrical resistance of an empty filter and a correction was made for filter surface area. 2.8. Permeability assay Restriction of the paracellular transport of FITClabelled dextran FD-4 (4 kDa) and FD-40 (40 kDa) through the in vitro BBB models was determined directly by analysis of the apparent permeability coefficient (Papp) across the in vitro BBB. FD-4 or FD40 was added to the apical side of the filters at a
237
concentration of 250 mg/ml. The media in the basolateral compartment was sampled every 15 min for 2 h. The concentration of FD-4 and FD-40 was determined using a fluorimeter (Spectra Max Gemini, Molecular Devices) with an excitation wavelength of 490 nm and an emission wavelength of 515 nm. Papp (cm/s) was calculated according to the method of Artursson (1990) using the following equation: Papp ¼
dQ 1 dt A C0 60
where dQ/dt is the amount of FD-4 or FD-40 transported per minute (ng/min), A the surface area of the filter (cm2), C0 the initial concentration of FD-4 (ng/ml) and 60 is the conversion from minutes to seconds.
3. Results 3.1. Cell culture of FBEC and astrocytes Feline brain capillary endothelial cells were isolated and cultured based largely on procedures described for bovine endothelial cells by Rubin et al. (1991), with some modifications. Colonies of endothelial cells grew from the isolated capillaries approximately 2 days after isolation. The endothelial cells spread around the capillaries and proliferated (Fig. 1a). When viewed using phase-contrast microscopy, endothelial cells grown on collagen-1 substrate displayed characteristic oval nuclei and were spindle shaped when confluent (Fig. 1b). By TEM analysis, endothelial cells were rich in mitochondria, and Weibel-Palade bodies were not observed in these cells. Astrocytes were isolated from neonatal felines according to published protocols described by Hertz et al. (1982) for rat astrocytes. Isolation procedures for feline astrocytes were similar to those described in other species (Gaillard et al., 2001; MacLean et al., 2002). Type-1 astrocytes proliferated at the first passage, as assessed by their characteristic flattened morphology when observed using phase-contrast microscopy (Fig. 1c and d). To develop the in vitro blood brain barrier, endothelial cells were cultured with astrocyte-conditioned medium, or alternatively, endothelial cells were co-cultured with astrocytes on the basolateral side of PET filters. After the first passage, endothelial cells
238
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
Fig. 1. FBEC 4 days after isolation, just before replating (a) and a confluent monolayer of FBEC (b). Feline astrocytes 5 days after isolation (c) and a confluent monolayer at passage 1 (d) (original magnification 10).
were seeded on the apical side of rat-tail collagen and fibronectin coated PET filters. Rat-tail collagen and fibronectin was found to be the best coating medium for the growth and morphology of the endothelial cells, when compared to Matrigel1, collagen-1 or TranswellCOL filters (data not shown). PET filters with 1 mm pores were used, as these had previously been reported to optimise contacts between astrocyte foot processes and endothelial cells (Demeuse et al., 2002). 3.2. Characterisation of the in vitro BBB model Cultures of FBECs expressed von Willebrand factor (Fig. 2a) and took up DiI-Ac-LDL (Fig. 2b), and cultures of astrocytes expressed GFAP (Fig. 2c and d). Cultures of endothelial cells were routinely >95% pure. When endothelial cells were co-cultured with astrocytes, junctional specialisation between the endothelial cells was visible by TEM (Fig. 3b). The cells were more closely associated with each other than cells which had been cultured without astrocytes. The presence of the tight junction proteins occludin and ZO-1 in the endothelial cell cultures was qualitatively measured using Western blots, and
endothelial cells were found to contain these proteins (Fig. 4). Tight junction development was quantified using TEER measurements and FITC-dextran (FD4 and FD40) permeability assays. Mean TEER values of endothelial cells cultured on PET filters without astrocytes or astrocyte-conditioned medium were 18 V cm2 (n = 5). This increased significantly to 20–25 V cm2 for endothelial cells grown with astrocyte-conditioned medium (n = 21), and 30–35 V cm2 for the co-culture model (n = 4) (Fig. 5). Tight junctions between the endothelial cells in the filter grown culture maintained barrier function as demonstrated by restricted transport of the hydrophilic paracellular transport markers FD-4 and FD-40. The Papp for FD-4 of an empty collagen/fibronectin coated filter was 7.79 105 cm/s (n = 4). In the presence of a monolayer of feline endothelial cells cultured with astrocyte-conditioned medium, the Papp decreased significantly to 3.96 105 cm/s (n = 4). When endothelial cells were co-cultured with astrocytes, the Papp was significantly decreased to 1.92 105 cm/s (n = 4) (Fig. 6a). Similarly, the Papp for FD-40 of an empty collagen/fibronectin coated filter was
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
239
Fig. 2. Characterisation of cell cultures. Primary feline brain capillary endothelial cells were positive for von Willebrand Factor (brown staining) by immunocytochemistry (a), and also phagocytosed DiI-Ac-LDL (red cytoplasmic staining) (b). Primary feline astrocytes were positive for GFAP by immunofluorescence (c). Original magnifications, 10 (a and c); 40 (b). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
3 105 cm/s (n = 4). In the presence of a monolayer of endothelial cells cultured with astrocyte-conditioned medium, the Papp decreased significantly to 6.27 106 cm/s (n = 6), and to 4.02 106 cm/s (n = 5) when endothelial cells were co-cultured with astrocytes (Fig. 6b).
4. Discussion This study describes the isolation procedures and primary culture methods for a novel in vitro model of the feline blood–brain barrier. This is the first time a cell culture model of the feline BBB has been described. The methods of isolating and characterising the feline in vitro BBB model are similar to the isolation techniques previously described for in vitro BBB models where cells have been isolated from other species, such as porcine (Jeliazkova-Mecheva and
Bobilya, 2003), rhesus macaque (MacLean et al., 2002) and in heterologous BBB systems which combine bovine endothelial cells with rat astrocytes (Rubin et al., 1991; Gaillard et al., 2001). FBEC and astrocytes were never used after the first passage, as they have been reported to lose the specific characteristics of these cells in vivo following repeated passaging. We found that FBEC could be isolated more readily from adult felines than from neonates. One adult brain could be used to generate one confluent 25 cm2 flask of primary FBECs. Studies in the macaque, feline and rat have used adult or adolescent animals for the isolation of brain endothelial cells (Steffan et al., 1994; Demeuse et al., 2002; Jeliazkova-Mecheva and Bobilya, 2003). Alternative methods, such as bovine models, have used neonatal or newborn animals (Gaillard et al., 2001). The isolation technique used for endothelial cells was based on the initial plating of a heterologous culture of
240
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
Fig. 5. TEER measurements of feline in vitro BBB models. Significant differences ( p < 0.0001) were found between all BBB models (endothelial cell monolayers, endothelial cells with astrocyte-conditioned medium and endothelial cells co-cultured with astrocytes) using two-tailed Mann–Whitney test.
Fig. 3. TEM analysis of (a) co-culture of endothelial cells and astrocytes. Cross-section of filter confirmed that endothelial cells (ec) were present on the apical surface of the filter (F) and astrocytes (as) were growing on the basolateral surface of the filter. Note the pore in the filter (P). (b) Co-culture of endothelial cells and astrocytes. Junction between two endothelial cells (arrow). Horizontal bars indicate 200 nm in (a and b).
Fig. 4. Western blot to identify the tight junction proteins occludin and ZO-1 in primary feline brain capillary endothelial cell cultures, 7 days after seeding of cells.
cells with the subsequent isolation and replating of pure endothelial cell colonies. When neonatal animals were used to isolate endothelial cells, all isolated cells grew very quickly. It was, therefore, difficult to isolate pure endothelial cell colonies as other cell types, for example fibroblasts, grew over these endothelial cell colonies. As cells isolated from adult animals grew more slowly than cells from neonates, and fewer cells attached to the culture dish, the isolation of endothelial cell colonies could be achieved with minimal contamination of other cell types because the replating of endothelial cells could be performed before any contaminating cells had grown. In contrast to FBEC, neonatal cats were used for astrocyte isolations, as it was found to be extremely difficult to isolate sufficient quantities of astrocytes from adult cats. One neonatal brain could be used to generate three confluent 175 cm2 flasks of primary feline astrocytes. This finding is in agreement with methods described for the isolation of astrocytes from rats (Rubin et al., 1991; Gaillard et al., 2001; Demeuse et al., 2002), humans (Fiala et al., 1997; Persidsky et al., 1997) and also feline astrocytes (Dow et al., 1992). Neonatal animals have relatively large populations of proliferating astrocyte precursor cells when compared to oligodendrocyte precursor cells, which do not develop until 3 months of age in rodents, and so astrocyte cultures are more homogenous when obtained from newborn animals (Hertz et al., 1982). The endothelial cell and astrocyte cultures were characterised based on morphology by means of light and electron microscopic analysis, immunocytochemically using von Willebrand factor antibody for
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
241
Fig. 6. Measurement of the apparent permeability (Papp) of the paracellular markers FD-4 (4 kDa) and FD-40 (40 kDa) across feline in vitro BBB models. In (a), permeability of FD-4, significant differences ( p < 0.05) were found between an empty filter compared with endothelial cells with astrocyte-conditioned medium, and between an empty filter compared with endothelial cells co-cultured with astrocytes, using the unpaired t-test. In (b), permeability of FD-40, significant differences ( p < 0.0001) were also found between an empty filter compared with endothelial cells with astrocyte-conditioned medium, and between an empty filter compared with endothelial cells co-cultured with astrocytes, using the unpaired t-test.
endothelial cells and GFAP antibody for astrocytes, and the endothelial cell cultures were also characterised by their ability to phagocytose DiI-Ac-LDL. The cell cultures were characterised using methods similar to those described in other studies (Dow et al., 1992; Demeuse et al., 2002; Jeliazkova-Mecheva and Bobilya, 2003). The von Willebrand factor and GFAP antibodies are commonly used in studies of other BBB models to characterise brain endothelial and astrocyte cell cultures (Rubin et al., 1991; Dow et al., 1992; de Boer et al., 1999; Gaillard et al., 2001; Demeuse et al., 2002; Jeliazkova-Mecheva and Bobilya, 2003). The development of specific BBB characteristics in the in vitro BBB models was assessed by the measurement of tight junction formation between endothelial cells. Tight junction complexes contain the tight junction proteins occludin and ZO-1, and as in studies of other animal models (Rubin et al., 1991), these proteins were also found to be present in feline endothelial cell cultures by Western blot analysis. TEER measurements and permeability assays were used to quantify tight junction formation between endothelial cells in the in vitro BBB models, and they were also used to compare with barrier strength of endothelial cells grown with no astrocytes. In comparing both TEER and paracellular permeability parameters between different model systems, the coculture model developed significantly higher TEER (30–35 V cm2) and was less permeable to FD-4 (Papp, 1.92 105 cm/s) and FD-40 (Papp, 4 106 cm/s) than the model where endothelial cells were cultured only in the presence of astrocyte-conditioned medium
(TEER of 20–25 V cm2, Papp of FD-4 of 3.96 105 cm/s and Papp of FD-40 of 6.27 106 cm/s). Astrocytes cultured without FBEC on coated filters did not develop TEER (data not shown). Many in vitro BBB studies have not included TEER measurements, but when compared to BBB models described for other species where TEER was measured, TEER of feline endothelial cells was considerably lower than that of the rat co-culture model which has been reported to develop TEER of 108–139 V cm2 (Jeliazkova-Mecheva and Bobilya, 2003). Bovine brain endothelial cells developed TEER of 115 V cm2, and human brain endothelial cells developed TEER of 67 V cm2 when cultured with astrocyte-conditioned medium (Rubin et al., 1991). The BBB in vivo has been reported to develop TEER of approximately 2000 V cm2, and to date no in vitro model has achieved TEER levels close to this in an in vitro BBB model. The paracellular permeability levels of FD-4 and FD-40, however, are comparable with the permeability of similar molecular weight markers described in other studies (Fiala et al., 1997; Franke et al., 2000; To¨ro¨k et al., 2003). One possibility to explain how these two parameters can become dissociated has been described by Balda et al. (1996). According to this hypothesis, TEER is an instantaneous measurement that reflects the permeability of tight junctions at a given point in time, and paracellular permeability (of FD-4, for example) is an indicator of permeability over a period of time. Tight junctions are not always tightly sealed but open and close in a fluctuating fashion. Thus, TEER measurements reflect the number of closed barriers at a given point in time,
242
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
whereas paracellular permeability gives a measurement of the permeability of the endothelial cell monolayer over a longer period of time. The dissociation of TEER and paracellular permeability has also been explained by Gaillard and de Boer (2000). This non-linearity can be explained by the fact that solute transport (e.g. Papp) is dependent on the sum of transport across all junctional pathways, whereas TEER is essentially dependent on areas with the lowest electrical resistance between single cells (Gaillard and de Boer, 2000). It has previously been reported that paracellular permeability to hydrophilic marker molecules affords a more accurate description than TEER of selective barrier function in tight epithelial cells (Artursson et al., 1993; Balda et al., 1996). An additional comparison between feline BBB model systems described in the present study included TEM analysis where endothelial cells which had been co-cultured with astrocytes possessed more junctional specialisation and were more closely associated with each other than endothelial cells which had been cultured with astrocyte-conditioned medium. Such observations supported TEER and paracellular permeability findings in the present study, and in others, such as the rat model system (Jeliazkova-Mecheva and Bobilya, 2003). A well-evaluated model of the BBB has major relevance in the study of species related issues in aspects of pathogenesis of neuroinflammation and neurodegeneration (Grant et al., 1998). While many viral agents may induce neuroinflammation in the cat (Fenner, 1994), one virus of direct relevance to the BBB is feline immunodeficiency virus. FIV is a model of HIV-1 infection in man. Like HIV-1, FIV enters the feline brain early in infection, at a time when there is considerable trafficking of lymphocytes through the BBB and choroid plexus (Bragg et al., 2002; Ryan et al., 2003). Therefore, use of this model has value in the dissection of interactions between FIV infected lymphocytes and the BBB, and lymphocyte trafficking through the BBB. In conclusion, we have developed and characterised an in vitro model of the feline BBB, which uses primary cultures of feline brain capillary endothelial cells and astrocytes grown as co-cultures on permeable filters. This model may have many future applications in studies related to neurodegeneration and neuroinflammation.
Acknowledgements We acknowledge the technical assistance of Brian Cloak and Christy King in the Department of Veterinary Pathology, University College Dublin, Matthew Campbell in the Conway Institute, University College Dublin for assistance with the Western blot technique and Dr. David John of Trinity College Dublin for assistance with Transmission Electron Microscopy. This work was funded by the Irish Higher Education Authority through the National Neuroscience Network (PRTLI-3).
References Abbott, N.J., Revest, P.A., 1991. Control of brain endothelial permeability. Cerebrovasc. Brain Metab. Rev. 3, 39–72. Annunziata, P., 2003. Blood–brain barrier changes during invasion of the central nervous system by HIV-1. Old and new insights into the mechanism. J. Neurol. 250, 901–906. Artursson, P., 1990. Epithelial transport of drugs in cell culture. I. A model for studying the passive diffusion of drugs over intestinal absorptive (Caco-2) cells. J. Pharm. Sci. 79, 476–482. Artursson, P., Ungell, A.L., Lofroth, J.E., 1993. Selective paracellular permeability in two models of intestinal absorption: cultured monolayers of human intestinal epithelial cells and rat intestinal segments. Pharm. Res. 10, 1123–1129. Audus, K.L., Borchardt, R.T., 1986. Characteristics of the large neutral amino acid transport system of bovine brain microvessel endothelial cell monolayers. J. Neurochem. 47, 484–488. Balda, M.S., Whitney, J.A., Flores, C., Gonzalez, S., Cereijido, M., Matter, K., 1996. Functional dissociation of paracellular permeability and transepithelial electrical resistance and disruption of the apical–basolateral intramembrane diffusion barrier by expression of a mutant tight junction membrane protein. J. Cell Biol. 134, 1031–1049. Bauer, H.C., Bauer, H., 2000. Neural induction of the blood–brain barrier: still an enigma. Cell Mol. Neurobiol. 20, 13–28. Bendinelli, M., Pistello, M., Lombardi, S., Poli, A., Garzelli, C., Matteucci, D., Ceccherini-Nelli, L., Malvaldi, G., Tozzini, F., 1995. Feline immunodeficiency virus: an interesting model for AIDS studies and an important cat pathogen. Clin. Microbiol. Rev. 8, 87–112. Boche, D., Hurtrel, M., Gray, F., Claessens-Maire, M.A., Ganiere, J.P., Montagnier, L., Hurtrel, B., 1996. Virus load and neuropathology in the FIV model. J. Neurovirol. 2, 377–387. Bowman, P.D., Ennis, S.R., Rarey, K.E., Betz, A.L., Goldstein, G.W., 1983. Brain microvessel endothelial cells in tissue culture: a model for study of blood–brain barrier permeability. Ann. Neurol. 14, 396–402. Bragg, D.C., Childers, T.A., Tompkins, M.B., Tompkins, W.A., Meeker, R.B., 2002. Infection of the choroid plexus by feline immunodeficiency virus. J. Neurovirol. 8, 211–224.
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244 Brightman, M., 1991. Implication of astroglia in the blood–brain barrier. Ann. N.Y. Acad. Sci. 633, 343–347. Broman, T., 1955. On basic aspects of the blood–brain barrier. Acta. Psychiatr. Neurol. Scand. 30, 115–124. Butt, A.M., 1995. Effect of inflammatory agents on electrical resistance across the blood–brain barrier in pial vessels of anaesthetized rats. Brain Res. 696, 145–150. Callanan, J.J., 1995. Feline immunodeficiency virus infection: a clinical and pathological perspective. In: Willett, B., Jarrett, O. (Eds.), Feline Immunology and Immunodeficiency. Oxford University Press, Oxford, pp. 111–130. Connor, R., Ho, D., 1992. Pathogenesis of human immunodeficiency virus. Semin. Virol. 3, 213–224. Cucullo, L., McAllister, M.S., Kight, K., Krizanac-Bengez, L., Marroni, M., Mayberg, M.R., Stanness, K.A., Janigro, D., 2002. A new dynamic in vitro model for the multidimensional study of astrocyte–endothelial cell interactions at the blood– brain barrier. Brain Res. 951, 243–254. Dallasta, L.M., Pisarov, L.A., Esplen, J.E., Werley, J.V., Moses, A.V., Nelson, J.A., Achim, C.L., 1999. Blood–brain barrier tight junction disruption in human immunodeficiency virus-1 encephalitis. Am. J. Pathol. 155, 1915–1927. de Boer, A.G., Gaillard, P.J., Breimer, D.D., 1999. The transference of results between blood–brain barrier cell culture systems. Eur. J. Pharm. Sci. 8, 1–4. Dehouck, M.P., Meresse, S., Delorme, P., Fruchart, J.C., Cecchelli, R., 1990. An easier, reproducible, and mass-production method to study the blood–brain barrier in vitro. J. Neurochem. 54, 1798–1801. Demeuse, P., Kerkhofs, A., Struys-Ponsar, C., Knoops, B., Remacle, C., van den Bosch de Aguilar, P., 2002. Compartmentalized coculture of rat brain endothelial cells and astrocytes: a syngenic model to study the blood–brain barrier. J. Neurosci. Methods 121, 21–31. Dow, S.W., Poss, M.L., Hoover, E.A., 1990. Feline immunodeficiency virus: a neurotropic lentivirus. J. Acquir. Immune Defic. Syndr. 3, 658–668. Dow, S.W., Dreitz, M.J., Hoover, E.A., 1992. Feline immunodeficiency virus neurotropism: evidence that astrocytes and microglia are the primary target cells. In: Dandekar, S., Pedersen, N. (Eds.), Feline Immunodeficiency Virus, vol. 35. Vet. Immunol. Immunopathol, pp. 23–35. Fenner, W., 1994. Diseases of the brain, spinal cord and peripheral nerves. In: Sherding, R. (Ed.), The Cat: Diseases and Clinical Management, vol. 2. WB Saunders Company, pp. 1507–1568. Fiala, M., Looney, D.J., Stins, M., Way, D.D., Zhang, L., Gan, X., Chiappelli, F., Schweitzer, E.S., Shapshak, P., Weinand, W., Graves, M.C., Witte, M., Kim, K.S., 1997. TNF-a opens a paracellular route for HIV-1 invasion across the blood–brain barrier. Mol. Med. 3, 553–564. Franke, H., Galla, H., Beuckmann, C.T., 2000. Primary cultures of brain microvessel endothelial cells: a valid and flexible model to study drug transport through the blood–brain barrier in vitro. Brain Res. Brain Res. Protoc. 5, 248–256. Gaillard, P.J., de Boer, A.G., 2000. Relationship between permeability status of the blood–brain barrier and in vitro permeability coefficient of a drug. Eur. J. Pharm. Sci. 12 (2), 95–102.
243
Gaillard, P.J., Voorwinden, L.H., Nielsen, J.L., Ivanov, A., Atsumi, R., Engman, H., Ringbom, C., de Boer, A.G., Breimer, D.D., 2001. Establishment and functional characterization of an in vitro model of the blood–brain barrier, comprising a co-culture of brain capillary endothelial cells and astrocytes. Eur. J. Pharm. Sci. 12, 215–222. Grant, G.A., Abbott, N.J., Janigro, D., 1998. Understanding the physiology of the blood–brain barrier: in vitro models. News Physiol. Sci. 13, 287–293. Henriksen, S.J., Prospero-Garcia, O., Phillips, T.R., Fox, H.S., Bloom, F.E., Elder, J.H., 1995. Feline immunodeficiency virus as a model for study of lentivirus infection of the central nervous system. Curr. Top. Microbiol. Immunol. 202, 167–186. Hertz, L., Juurlink, B., Fosmark, H., Schousboe, A., 1982. Astrocytes in Primary Cultures. In: Pfeiffer, S. (Ed.), Neuroscience Approached through Cell Culture. CRC Press, pp. 175–186. Jeliazkova-Mecheva, V.V., Bobilya, D.J., 2003. A porcine astrocyte/ endothelial cell co-culture model of the blood–brain barrier. Brain Res. Brain Res. Protoc. 12, 91–98. Kim, J.A., Tran, N.D., Wang, S.J., Fisher, M.J., 2003. Astrocyte regulation of human brain capillary endothelial fibrinolysis. Thromb. Res. 112, 159–165. Kondo, T., Imaizumi, S., Kato, I., Yoshimoto, T., 1994. Isolation and culture of brain endothelial cells and establishment of in vitro blood–brain barrier model. Cell. Transplant. (Suppl. 1), S35– S37. Luabeya, M.K., Dallasta, L.M., Achim, C.L., Pauza, C.D., Hamilton, R.L., 2000. Blood–brain barrier disruption in simian immunodeficiency virus encephalitis. Neuropathol. Appl. Neurobiol. 26, 454–462. MacLean, A.G., Orandle, M.S., MacKey, J., Williams, K.C., Alvarez, X., Lackner, A.A., 2002. Characterization of an in vitro rhesus macaque blood–brain barrier. J. Neuroimmunol. 131, 98–103. Marroni, M., Kight, K.M., Hossain, M., Cucullo, L., Desai, S.Y., Janigro, D., 2003. Dynamic in vitro model of the blood–brain barrier. Gene profiling using cDNA microarray analysis. Methods Mol. Med. 89, 419–434. McCarthy, K.D., de Vellis, J., 1980. Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J. Cell. Biol. 85, 890–902. Megard, I., Garrigues, A., Orlowski, S., Jorajuria, S., Clayette, P., Ezan, E., Mabondzo, A., 2002. A co-culture-based model of human blood–brain barrier: application to active transport of indinavir and in vivo–in vitro correlation. Brain Res. 927, 153– 167. Persidsky, Y., Stins, M., Way, D., Witte, M.H., Weinand, M., Kim, K.S., Bock, P., Gendelman, H.E., Fiala, M., 1997. A model for monocyte migration through the blood–brain barrier during HIV-1 encephalitis. J. Immunol. 158, 3499–3510. Reese, T.S., Karnovsky, M.J., 1967. Fine structural localization of a blood–brain barrier to exogenous peroxidase. J. Cell. Biol. 34, 207–217. Ringler, D.J., Hunt, R.D., Desrosiers, R.C., Daniel, M.D., Chalifoux, L.V., King, N.W., 1988. Simian immunodeficiency virusinduced meningoencephalitis: natural history and retrospective study. Ann. Neurol. 23 (Suppl.), S101–S107.
244
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
Rubin, L.L., Hall, D.E., Porter, S., Barbu, K., Cannon, C., Horner, H.C., Janatpour, M., Liaw, C.W., Manning, K., Morales, J., Tanner, L.I., Tomaselli, K.J., Bard, F., 1991. A cell culture model of the blood–brain barrier. J. Cell. Biol. 115, 1725–1735. Ryan, G., Klein, D., Knapp, E., Hosie, M.J., Grimes, T., Mabruk, M.J., Jarrett, O., Callanan, J.J., 2003. Dynamics of viral and proviral loads of feline immunodeficiency virus within the feline central nervous system during the acute phase following intravenous infection. J. Virol. 77, 7477–7485. Sharer, L.R., 1992. Pathology of HIV-1 infection of the central nervous system: a review. J. Neuropathol. Exp. Neurol. 51, 3–11. Song, L., Pachter, J.S., 2003. Culture of murine brain microvascular endothelial cells that maintain expression and cytoskeletal association of tight junction-associated proteins. In Vitro Cell Dev. Biol. Anim. 39, 313–320.
Stanness, K.A., Guatteo, E., Janigro, D., 1996. A dynamic model of the blood–brain barrier ‘‘in vitro’’. Neurotoxicology 17, 481– 496. Steffan, A.M., Lafon, M.E., Gendrault, J.L., Koehren, F., De Monte, M., Royer, C., Kirn, A., Gut, J.P., 1994. Feline immunodeficiency virus can productively infect cultured endothelial cells from cat brain microvessels. J. Gen. Virol. 75, 3647–3653. Strom, S.C., Michalopoulos, G., 1982. Collagen as a substrate for cell growth and differentiation. Methods Enzymol. 82, 544– 555. To¨ro¨k, M., Huwyler, J., Gutmann, H., Fricker, G., Drewe, J., 2003. Modulation of transendothelial permeability and expression of ATP-binding cassette transporters in cultured brain capillary endothelial cells by astrocytic factors and cell-culture conditions. Exp. Brain Res. 221, 356–365.
Growth and characterisation of a cell culture model of the feline blood–brain barrier Nicola F. Fletcher a,c,*, David J. Brayden b,c, Brenda Brankin c, Sheila Worrall a, John J. Callanan a,c a b
Department of Veterinary Pathology, University College Dublin, Belfield, Dublin 4, Ireland Small Animal Clinical Studies, Faculty of Veterinary Medicine, University College Dublin, Belfield, Dublin 4, Ireland c Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland Received 1 March 2005; received in revised form 15 July 2005; accepted 15 August 2005
Abstract An in vitro model of the feline blood–brain barrier was developed using primary cultures of brain capillary endothelial cells derived from adult cats. They were grown in the presence of astrocytes obtained from newborn kittens. Feline endothelial cell cultures were characterised by uptake of DiI-acetylated low-density lipoprotein (DiI-Ac-LDL) and expression of von Willebrand factor. Astrocytes were characterised based on their expression of glial fibrillary acidic protein (GFAP). Electron microscopy revealed junctional specialisation between endothelial cells. Occludin and ZO-1 expression by the endothelial cell cultures was detected by Western blot analysis. Barrier function of co-cultured endothelial cells and astrocytes was confirmed by a transendothelial electrical resistance (TEER) value of 30–35 V cm2 and apparent permeability coefficients (Papp) for FD-40 (FITC-dextran, 40 kDa) of 4 106 cm/s and for FD-4 (4 kDa) of 1.92 105 cm/s. In endothelial cell monolayers grown with astrocyte-conditioned medium, the TEER value was lower (20–25 V cm2), and Papp of FD-40 and FD-4 was higher at 6.27 106 and 3.96 105 cm/s, respectively. This model should have useful applications in the examination of events occurring at the BBB early in FIV infection, and may provide knowledge applicable to HIV infection. # 2005 Elsevier B.V. All rights reserved. Keywords: Blood–brain barrier; Endothelial cell monolayers; FIV neuropathogenesis; Endothelial cell tight junctions
Abbreviations: ACM, astrocyte-conditioned medium; BBB, blood–brain barrier; CNS, central nervous system; DiI-Ac-LDL, DiI-acetylated low-density lipoprotein; DMEM, Dulbecco’s modified eagle’s medium; ECGF, endothelial cell growth factor; FBEC, feline brain endothelial cells; FIV, feline immunodeficiency virus; FD, FITC-dextran; GFAP, glial fibrillary acidic protein; HRP, horseradish peroxidase; Papp, apparent permeability; PET, polyethylene terephthalate; SIV, simian immunodeficiency virus; TEER, transendothelial electrical resistance; TEM, transmission electron microscopy; TBS, tris buffered saline; ZO-1, zonula occludens-1 * Corresponding author. Tel.: +353 1 7166150; fax: +353 1 7166165. E-mail address: [email protected] (N.F. Fletcher). 0165-2427/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.vetimm.2005.08.025
234
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
1. Introduction The blood–brain barrier (BBB) regulates the passage of molecules between the CNS and the peripheral circulation (Broman, 1955). Tight junctions between cerebral capillary endothelial cells represent the structural basis of this barrier. Under normal physiological conditions, tight junctions form a continuous, almost impenetrable barrier that prevents the passive influx of a variety of substances with the exception of some small hydrophilic molecules (Reese and Karnovsky, 1967). Transcellular absorption for lipophilic agents, and carrier-mediated transport across endothelial cells, are the predominant routes of entry of substances to the brain. Astrocytes take an active part in the formation of the BBB by influencing differentiation of endothelial cells, by promoting high resistance tight junction formation in vivo (Butt, 1995). They also induce decreased pinocytotic activity and fenestrations in endothelial cells (Bauer and Bauer, 2000). Astrocytic endfeet ensheath the brain capillaries, and chemical signals released by these cells are in part responsible for the unique characteristics of the vertebrate BBB (Brightman, 1991; Rubin et al., 1991; Demeuse et al., 2002). The earliest in vitro BBB tissue culture systems were cultured endothelial cells of cerebral or noncerebral origin (Bowman et al., 1983; Audus and Borchardt, 1986; Kondo et al., 1994; Stanness et al., 1996). Currently, the most widely used in vitro BBB models across a range of species are brain capillary endothelial cells cultured with astrocyte-conditioned medium (ACM) (Rubin et al., 1991; Gaillard et al., 2001) or co-cultured with astrocytes (Dehouck et al., 1990; Gaillard et al., 2001). These models result in the formation of endothelial tight junctions and induce the expression of specific BBB markers on the apical side of endothelial cells, including g-glutamyl transpeptidase, the GLUT-1 glucose transporter and P-glycoprotein (Abbott and Revest, 1991). These models have been grown on permeable filters (Gaillard et al., 2001; Rubin et al., 1991; Demeuse et al., 2002) or cultured under flow conditions, where endothelial cells are cultured in the lumen of hollow fibres while astrocytes are seeded in the extraluminal compartment (Cucullo et al., 2002; Marroni et al., 2003). Growth of in vitro BBB models has been described for various mammalian species. Most commonly,
bovine endothelial cells have been cultured with rat astrocytes or rat ACM (Dehouck et al., 1990; Rubin et al., 1991; Gaillard et al., 2001), but other models utilising autologous cell cultures have been developed from the rat (Demeuse et al., 2002), mouse (Song and Pachter, 2003), human (Kim et al., 2003), porcine (Jeliazkova-Mecheva and Bobilya, 2003) and rhesus macaque (MacLean et al., 2002). Many of these models have been developed in order to study drug transport to the CNS (Audus and Borchardt, 1986; Rubin et al., 1991; Gaillard et al., 2001), or have been developed to study aspects of viral neuropathogenesis, including HIV-1 (Megard et al., 2002) and SIV infections (MacLean et al., 2002). HIV-1 infection in man is associated with neuropathology both in the early and late stages of infection (Annunziata, 2003). However, due to poor access to significant quantities of brain tissue in early infection, there is a paucity of information on how virus enters the brain and how it interacts with neurological tissue to induce significant neurological impairment. Animal models of HIV infection provide a useful way to study the pathogenesis of HIV, and the feline lentiviral model, FIV of the domestic cat, has been established as an excellent model of HIV-1 infection (Bendinelli et al., 1995; Callanan, 1995; Henriksen et al., 1995). FIV resembles HIV in many respects, but in particular by its ability to infect the CNS early after experimental infection. It is known that FIV enters the brain in the first week following infection (Boche et al., 1996; Ryan et al., 2003). Changes in the CNS, including perivascular mononuclear infiltrates, have been extensively described in HIV, SIV and FIV infections (Ringler et al., 1988; Dow et al., 1990; Connor and Ho, 1992; Sharer, 1992). However, the mechanism of entry of HIV, SIV and FIV to the brain is still a matter of debate. One mechanism of lentiviral entry to the CNS may be via breaches in the BBB. In support of this hypothesis, endothelial cell tight junction disruption has been demonstrated as a key feature in HIV and SIV encephalitis (Dallasta et al., 1999; Luabeya et al., 2000). In order to study the interaction of FIV with cells of the BBB, we have developed and characterised an in vitro model of the feline BBB using autologous cultures of primary feline brain capillary endothelial
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
cells and feline astrocytes. The endothelial cells were grown either as co-cultures on permeable filters with astrocytes or alternatively, as monolayers in the presence of astrocyte-conditioned medium. This model should be useful in the examination of immunological events at the BBB early in FIV infection, and may provide knowledge applicable to HIV infection.
2. Materials and methods 2.1. Materials Disposable sterile plasticware was purchased from Sarstedt (Sarstedt AG & Co., Nu¨mbrecht, Germany), Sigma (Sigma Chemical Co., St. Louis, MO, USA), Corning Costar (Cambridge, MA, USA) and BectonDickinson (Boston, MA, USA). Cell culture media and supplements, PBS, HBSS, bovine fibronectin, Trypsin–EDTA and Foetal Calf Serum were obtained from Gibco BRL Life Technologies Inc. (Grand Island, NY, USA). Collagenase-dispase was obtained from Roche. DiI-acetylated low-density lipoprotein (DiI-Ac-LDL) was obtained from Molecular Probes (Eugene, OR, USA), rabbit anti-GFAP antibody, rabbit anti-von Willebrand Factor antibody and swine anti-rabbit FITC-conjugated secondary antibody from Dako (Cambridge, UK), rabbit anti-ZO-1 antibody from Santa Cruz Biotechnology (Santa Cruz, CA, USA), rabbit anti-occludin antibody from Zymed (San Francisco, CA, USA) and Vectastain Elite ABC Rabbit IgG Kit and Vectashield mounting medium from Vector Laboratories (Burlingame, CA, USA). Nylon mesh for the isolation of primary cultures was obtained from Clarcor (Rockford, IL, USA). Thermanox coverslips were obtained from Nunc (Rochester, NY, USA). Fluorescein isothiocyanate-dextran-4 (FD-4), FD-40 and Triton X-100 were obtained from Sigma (St. Louis, MO). 2.2. Cell culture Feline brain capillary endothelial cells (FBEC) were isolated from euthanased adult cats, and cultured within 2 h of death according to the method of Rubin et al. (1991), but with some modifications. Freshly harvested feline cortical tissue was immersed
235
in ice-cold Hanks balanced salt solution (HBSS) supplemented with Penicillin/Streptomycin and Amphotericin B. Meninges and white matter were removed with a scalpel and grey matter was fragmented and collected in 10 ml DMEM with 20% heat inactivated foetal calf serum (FCS). The brain tissue was homogenised using a Polytron homogeniser (Kinematica GmbH, Switzerland), fitted with a 7-mm probe, and the resulting homogenate was then filtered through a 125 mm nylon mesh, and digested in collagenase and dispase for 1 h at 37 8C. The cells were cultured in vented 25 cm2 tissue culture flasks (Sarstedt AG & Co., Nu¨mbrecht, Germany) with DMEM/20% heat inactivated FCS and incubated at 37 8C with 5% CO2. The media was replaced every other day. After 4 days in culture, the media was supplemented with 10 ml/ml endothelial cell growth factor (ECGF; Sigma Chemical Co.). After approximately 7 days in culture, several large colonies of endothelial cells were visible in the culture flasks. These colonies were isolated, replated on collagen-1 coated tissue culture flasks and cultured with DMEM/ 20% heat inactivated FCS supplemented with 10 ml/ ml ECGF for approximately 48 h or until the cells were 80–90% confluent. Cells were then passaged with Trypsin–EDTA and stored in liquid nitrogen. Feline astrocytes were isolated from newborn specific pathogen free (SPF) kittens. Isolated cerebral cortices of each brain were fragmented and filtered first through 80 mm and then through 10 mm nylon meshes, and cultured for 4 days in DMEM/10% heat inactivated FCS. After 7 days in culture, the flasks were passaged with Trypsin–EDTA in a split ratio of 1:4 and grown to confluence. Some cultures possessed contaminating cells, and these were purified according to the method of McCarthy and de Vellis (1980). In brief, cultures containing contaminating cells were shaken for 12 h on a plate shaker at 37 8C. This removed the loosely adherent contaminating cells, such as oligodendrocytes, while leaving the astrocyte layer intact. The medium was then replaced and the remaining astrocytes were cultured until they reached confluence. For collection of ACM, cultures that were confluent were cultured with DMEM/10% heat inactivated FCS and the medium was collected twice a week, sterile filtered, and either used immediately or stored at 20 8C. After 2 weeks of conditioned medium
236
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
collection, astrocytes were passaged and stored in freeze mix (10% DMSO in FCS) in liquid nitrogen for co-culture purposes. The in vitro co-culture BBB model was prepared on rat-tail collagen and bovine fibronectin coated polyethylene terephthalate (PET) filters (surface area 0.9 cm2; pore size 1 mm; Becton-Dickinson). Rat-tail collagen was prepared from the tails of adult Wistar rats according to the method of Strom and Michalopoulos (1982). Astrocytes were seeded on the bottom of the filter at a density of 5 105 cells per filter. The filters were inverted and astrocytes were allowed to adhere to the bottom of the filter for 2 h. The filters were then placed in 12 well plates, and the astrocytes were fed with DMEM/20% FCS. After 72 h, endothelial cells were seeded at a density of 1.5 105 cells per filter onto the apical side of each filter. The medium was replaced every 2 or 3 days. Endothelial cell monolayers were also cultured on porous filters as described above, but with 50% (v/v) astrocyte-conditioned medium added to the culture medium. Experiments with the in vitro BBB models were performed 7 days after the endothelial cells were seeded. 2.3. Immunocytochemical analysis of BBB model After 7 days in culture, filters containing endothelial cell cultures were fixed in ice-cold methanol for immunocytochemistry and immunofluorescence. Cells were permeabilised with 0.05% Triton X-100 for 30 min. Non-specific binding of antibody was blocked by incubation with 5% normal goat serum for 10 min. Primary antibody was incubated at 4 8C for 12 h and washed with PBS. Primary antibodies used were polyclonal anti-rabbit antibody specific for von Willebrand factor (1:1000; Dako, Cambridge, UK) for FBEC, and polyclonal anti-rabbit antibody specific for GFAP (Undiluted; Dako) for feline astrocytes. For immunofluorescence, secondary antibody (swine antirabbit FITC) was applied for 1 h at 37 8C. For immunocytochemistry, detection of the primary antibodies was performed using the Vectastain Elite ABC kit (Rabbit IgG) as per manufacturer’s instructions. Diaminobenzidine was used as the chromagen. Filters were counterstained with haematoxylin, dehydrated and mounted in DPX (Sigma Chemical Co.) for immunocytochemistry, or Vectashield mounting med-
ium for immunofluorescence. Negative controls were treated as above, but an irrelevant isotype-matched antibody was used in place of the primary antibody. 2.4. Uptake of DiI-Ac-LDL Cultures of FBEC were cultured on collagen and fibronectin coated Thermanox1 coverslips for 2–3 days. The coverslips were then washed with DMEM, and the medium was replaced with DMEM containing 10 mg/ml DiI-Ac-LDL, and cultured for 4 h at 37 8C with 5% CO2. The coverslips were then washed with PBS, fixed with 2% paraformaldehyde for 10 min at room temperature and mounted with Vectashield mounting medium. 2.5. Western blot analysis The presence of occludin and ZO-1 proteins in the in vitro BBB was determined by Western blot analysis. Occludin and ZO-1 protein was isolated from filters containing confluent monolayers of FBEC. ARPE-19 cells (human retinal pigment epithelial cells) and Mya-1 cells (feline T-lymphoblastoid cells) were used as positive and negative controls, respectively. Three confluent monolayers of FBEC or ARPE-19 cells (approximately 5 105 cells), or 5 105 Mya-1 cells, were used for each experiment. The filters were removed from the filter inserts and washed with PBS. The cell monolayers were scraped and homogenised with a 21-gauge needle in boiling buffer containing 62.5 mM Tris, 2% SDS, 10 mM dithiothreitol, 10 ml protease inhibitor cocktail/100 ml (Sigma Chemical Co.). The homogenate was centrifuged at 10,000 g for 20 min at 4 8C, and the resulting pellet was used for occludin or ZO-1 analysis. Protein concentrations were determined using the Micro-BCA1 assay (Pierce, MSC, Dublin, Ireland) assay. A BSA standard curve and equal amounts of protein were loaded onto SDS-PAGE gels. 10% and 6% SDS-PAGE was used for occludin and ZO-1 analyses, respectively. Protein samples were separated by electrophoresis, and transferred to nitrocellulose membrane for 2 h using a semi-dry electroblot apparatus (Bio-Rad). Efficiency of protein transfer was determined using Ponceau-S solution (Sigma Chemical Co.). Non-specific binding sites were blocked by incubating the membrane at room
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
temperature with 5% non-fat dry skimmed milk in TBS (0.05 M Tris, 150 mM NaCl, pH 7.5) for 2 h. Membranes were briefly washed with TBS, and incubated with polyclonal anti-rabbit antibody (Zymed) specific for occludin (1:1000) or polyclonal anti-rabbit antibody (Santa-Cruz Biotech, CA) specific for ZO-1 (1:1000). Antibodies were incubated with membranes overnight at 4 8C. The membranes were washed with TBS, and incubated with a secondary anti-rabbit (IgG) antibody with a HRP conjugate (1:2500), for 3 h at room temperature. Immune complexes were detected using enhanced chemiluminescence (ECL). 2.6. Transmission electron microscopy Filters containing confluent monolayers of endothelial cells with or without astrocyte co-cultures, or cultured with astrocyte-conditioned medium were fixed at room temperature for 1 h with glutaraldehyde (2.5%, v/v) in phosphate buffer, and postfixed in osmium tetroxide (2%, w/v) in 0.05 M potassium phosphate buffer for 30 min. The filters were dehydrated through a graded series of ethanols to 100% and embedded in epoxy resin. Sections were cut and observed through a Hitachi H-7000 transmission electron microscope. 2.7. Transendothelial electrical resistance measurements Restriction of paracellular transport of small ions was determined indirectly by calculation of transendothelial electrical resistance (TEER). TEER was measured using an Endohm Chamber connected to an EVOM Voltohmeter (World Precision Instruments Inc., New Haven, CT). TEER (V cm2) was calculated by subtraction of the electrical resistance of an empty filter and a correction was made for filter surface area. 2.8. Permeability assay Restriction of the paracellular transport of FITClabelled dextran FD-4 (4 kDa) and FD-40 (40 kDa) through the in vitro BBB models was determined directly by analysis of the apparent permeability coefficient (Papp) across the in vitro BBB. FD-4 or FD40 was added to the apical side of the filters at a
237
concentration of 250 mg/ml. The media in the basolateral compartment was sampled every 15 min for 2 h. The concentration of FD-4 and FD-40 was determined using a fluorimeter (Spectra Max Gemini, Molecular Devices) with an excitation wavelength of 490 nm and an emission wavelength of 515 nm. Papp (cm/s) was calculated according to the method of Artursson (1990) using the following equation: Papp ¼
dQ 1 dt A C0 60
where dQ/dt is the amount of FD-4 or FD-40 transported per minute (ng/min), A the surface area of the filter (cm2), C0 the initial concentration of FD-4 (ng/ml) and 60 is the conversion from minutes to seconds.
3. Results 3.1. Cell culture of FBEC and astrocytes Feline brain capillary endothelial cells were isolated and cultured based largely on procedures described for bovine endothelial cells by Rubin et al. (1991), with some modifications. Colonies of endothelial cells grew from the isolated capillaries approximately 2 days after isolation. The endothelial cells spread around the capillaries and proliferated (Fig. 1a). When viewed using phase-contrast microscopy, endothelial cells grown on collagen-1 substrate displayed characteristic oval nuclei and were spindle shaped when confluent (Fig. 1b). By TEM analysis, endothelial cells were rich in mitochondria, and Weibel-Palade bodies were not observed in these cells. Astrocytes were isolated from neonatal felines according to published protocols described by Hertz et al. (1982) for rat astrocytes. Isolation procedures for feline astrocytes were similar to those described in other species (Gaillard et al., 2001; MacLean et al., 2002). Type-1 astrocytes proliferated at the first passage, as assessed by their characteristic flattened morphology when observed using phase-contrast microscopy (Fig. 1c and d). To develop the in vitro blood brain barrier, endothelial cells were cultured with astrocyte-conditioned medium, or alternatively, endothelial cells were co-cultured with astrocytes on the basolateral side of PET filters. After the first passage, endothelial cells
238
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
Fig. 1. FBEC 4 days after isolation, just before replating (a) and a confluent monolayer of FBEC (b). Feline astrocytes 5 days after isolation (c) and a confluent monolayer at passage 1 (d) (original magnification 10).
were seeded on the apical side of rat-tail collagen and fibronectin coated PET filters. Rat-tail collagen and fibronectin was found to be the best coating medium for the growth and morphology of the endothelial cells, when compared to Matrigel1, collagen-1 or TranswellCOL filters (data not shown). PET filters with 1 mm pores were used, as these had previously been reported to optimise contacts between astrocyte foot processes and endothelial cells (Demeuse et al., 2002). 3.2. Characterisation of the in vitro BBB model Cultures of FBECs expressed von Willebrand factor (Fig. 2a) and took up DiI-Ac-LDL (Fig. 2b), and cultures of astrocytes expressed GFAP (Fig. 2c and d). Cultures of endothelial cells were routinely >95% pure. When endothelial cells were co-cultured with astrocytes, junctional specialisation between the endothelial cells was visible by TEM (Fig. 3b). The cells were more closely associated with each other than cells which had been cultured without astrocytes. The presence of the tight junction proteins occludin and ZO-1 in the endothelial cell cultures was qualitatively measured using Western blots, and
endothelial cells were found to contain these proteins (Fig. 4). Tight junction development was quantified using TEER measurements and FITC-dextran (FD4 and FD40) permeability assays. Mean TEER values of endothelial cells cultured on PET filters without astrocytes or astrocyte-conditioned medium were 18 V cm2 (n = 5). This increased significantly to 20–25 V cm2 for endothelial cells grown with astrocyte-conditioned medium (n = 21), and 30–35 V cm2 for the co-culture model (n = 4) (Fig. 5). Tight junctions between the endothelial cells in the filter grown culture maintained barrier function as demonstrated by restricted transport of the hydrophilic paracellular transport markers FD-4 and FD-40. The Papp for FD-4 of an empty collagen/fibronectin coated filter was 7.79 105 cm/s (n = 4). In the presence of a monolayer of feline endothelial cells cultured with astrocyte-conditioned medium, the Papp decreased significantly to 3.96 105 cm/s (n = 4). When endothelial cells were co-cultured with astrocytes, the Papp was significantly decreased to 1.92 105 cm/s (n = 4) (Fig. 6a). Similarly, the Papp for FD-40 of an empty collagen/fibronectin coated filter was
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
239
Fig. 2. Characterisation of cell cultures. Primary feline brain capillary endothelial cells were positive for von Willebrand Factor (brown staining) by immunocytochemistry (a), and also phagocytosed DiI-Ac-LDL (red cytoplasmic staining) (b). Primary feline astrocytes were positive for GFAP by immunofluorescence (c). Original magnifications, 10 (a and c); 40 (b). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
3 105 cm/s (n = 4). In the presence of a monolayer of endothelial cells cultured with astrocyte-conditioned medium, the Papp decreased significantly to 6.27 106 cm/s (n = 6), and to 4.02 106 cm/s (n = 5) when endothelial cells were co-cultured with astrocytes (Fig. 6b).
4. Discussion This study describes the isolation procedures and primary culture methods for a novel in vitro model of the feline blood–brain barrier. This is the first time a cell culture model of the feline BBB has been described. The methods of isolating and characterising the feline in vitro BBB model are similar to the isolation techniques previously described for in vitro BBB models where cells have been isolated from other species, such as porcine (Jeliazkova-Mecheva and
Bobilya, 2003), rhesus macaque (MacLean et al., 2002) and in heterologous BBB systems which combine bovine endothelial cells with rat astrocytes (Rubin et al., 1991; Gaillard et al., 2001). FBEC and astrocytes were never used after the first passage, as they have been reported to lose the specific characteristics of these cells in vivo following repeated passaging. We found that FBEC could be isolated more readily from adult felines than from neonates. One adult brain could be used to generate one confluent 25 cm2 flask of primary FBECs. Studies in the macaque, feline and rat have used adult or adolescent animals for the isolation of brain endothelial cells (Steffan et al., 1994; Demeuse et al., 2002; Jeliazkova-Mecheva and Bobilya, 2003). Alternative methods, such as bovine models, have used neonatal or newborn animals (Gaillard et al., 2001). The isolation technique used for endothelial cells was based on the initial plating of a heterologous culture of
240
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
Fig. 5. TEER measurements of feline in vitro BBB models. Significant differences ( p < 0.0001) were found between all BBB models (endothelial cell monolayers, endothelial cells with astrocyte-conditioned medium and endothelial cells co-cultured with astrocytes) using two-tailed Mann–Whitney test.
Fig. 3. TEM analysis of (a) co-culture of endothelial cells and astrocytes. Cross-section of filter confirmed that endothelial cells (ec) were present on the apical surface of the filter (F) and astrocytes (as) were growing on the basolateral surface of the filter. Note the pore in the filter (P). (b) Co-culture of endothelial cells and astrocytes. Junction between two endothelial cells (arrow). Horizontal bars indicate 200 nm in (a and b).
Fig. 4. Western blot to identify the tight junction proteins occludin and ZO-1 in primary feline brain capillary endothelial cell cultures, 7 days after seeding of cells.
cells with the subsequent isolation and replating of pure endothelial cell colonies. When neonatal animals were used to isolate endothelial cells, all isolated cells grew very quickly. It was, therefore, difficult to isolate pure endothelial cell colonies as other cell types, for example fibroblasts, grew over these endothelial cell colonies. As cells isolated from adult animals grew more slowly than cells from neonates, and fewer cells attached to the culture dish, the isolation of endothelial cell colonies could be achieved with minimal contamination of other cell types because the replating of endothelial cells could be performed before any contaminating cells had grown. In contrast to FBEC, neonatal cats were used for astrocyte isolations, as it was found to be extremely difficult to isolate sufficient quantities of astrocytes from adult cats. One neonatal brain could be used to generate three confluent 175 cm2 flasks of primary feline astrocytes. This finding is in agreement with methods described for the isolation of astrocytes from rats (Rubin et al., 1991; Gaillard et al., 2001; Demeuse et al., 2002), humans (Fiala et al., 1997; Persidsky et al., 1997) and also feline astrocytes (Dow et al., 1992). Neonatal animals have relatively large populations of proliferating astrocyte precursor cells when compared to oligodendrocyte precursor cells, which do not develop until 3 months of age in rodents, and so astrocyte cultures are more homogenous when obtained from newborn animals (Hertz et al., 1982). The endothelial cell and astrocyte cultures were characterised based on morphology by means of light and electron microscopic analysis, immunocytochemically using von Willebrand factor antibody for
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
241
Fig. 6. Measurement of the apparent permeability (Papp) of the paracellular markers FD-4 (4 kDa) and FD-40 (40 kDa) across feline in vitro BBB models. In (a), permeability of FD-4, significant differences ( p < 0.05) were found between an empty filter compared with endothelial cells with astrocyte-conditioned medium, and between an empty filter compared with endothelial cells co-cultured with astrocytes, using the unpaired t-test. In (b), permeability of FD-40, significant differences ( p < 0.0001) were also found between an empty filter compared with endothelial cells with astrocyte-conditioned medium, and between an empty filter compared with endothelial cells co-cultured with astrocytes, using the unpaired t-test.
endothelial cells and GFAP antibody for astrocytes, and the endothelial cell cultures were also characterised by their ability to phagocytose DiI-Ac-LDL. The cell cultures were characterised using methods similar to those described in other studies (Dow et al., 1992; Demeuse et al., 2002; Jeliazkova-Mecheva and Bobilya, 2003). The von Willebrand factor and GFAP antibodies are commonly used in studies of other BBB models to characterise brain endothelial and astrocyte cell cultures (Rubin et al., 1991; Dow et al., 1992; de Boer et al., 1999; Gaillard et al., 2001; Demeuse et al., 2002; Jeliazkova-Mecheva and Bobilya, 2003). The development of specific BBB characteristics in the in vitro BBB models was assessed by the measurement of tight junction formation between endothelial cells. Tight junction complexes contain the tight junction proteins occludin and ZO-1, and as in studies of other animal models (Rubin et al., 1991), these proteins were also found to be present in feline endothelial cell cultures by Western blot analysis. TEER measurements and permeability assays were used to quantify tight junction formation between endothelial cells in the in vitro BBB models, and they were also used to compare with barrier strength of endothelial cells grown with no astrocytes. In comparing both TEER and paracellular permeability parameters between different model systems, the coculture model developed significantly higher TEER (30–35 V cm2) and was less permeable to FD-4 (Papp, 1.92 105 cm/s) and FD-40 (Papp, 4 106 cm/s) than the model where endothelial cells were cultured only in the presence of astrocyte-conditioned medium
(TEER of 20–25 V cm2, Papp of FD-4 of 3.96 105 cm/s and Papp of FD-40 of 6.27 106 cm/s). Astrocytes cultured without FBEC on coated filters did not develop TEER (data not shown). Many in vitro BBB studies have not included TEER measurements, but when compared to BBB models described for other species where TEER was measured, TEER of feline endothelial cells was considerably lower than that of the rat co-culture model which has been reported to develop TEER of 108–139 V cm2 (Jeliazkova-Mecheva and Bobilya, 2003). Bovine brain endothelial cells developed TEER of 115 V cm2, and human brain endothelial cells developed TEER of 67 V cm2 when cultured with astrocyte-conditioned medium (Rubin et al., 1991). The BBB in vivo has been reported to develop TEER of approximately 2000 V cm2, and to date no in vitro model has achieved TEER levels close to this in an in vitro BBB model. The paracellular permeability levels of FD-4 and FD-40, however, are comparable with the permeability of similar molecular weight markers described in other studies (Fiala et al., 1997; Franke et al., 2000; To¨ro¨k et al., 2003). One possibility to explain how these two parameters can become dissociated has been described by Balda et al. (1996). According to this hypothesis, TEER is an instantaneous measurement that reflects the permeability of tight junctions at a given point in time, and paracellular permeability (of FD-4, for example) is an indicator of permeability over a period of time. Tight junctions are not always tightly sealed but open and close in a fluctuating fashion. Thus, TEER measurements reflect the number of closed barriers at a given point in time,
242
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
whereas paracellular permeability gives a measurement of the permeability of the endothelial cell monolayer over a longer period of time. The dissociation of TEER and paracellular permeability has also been explained by Gaillard and de Boer (2000). This non-linearity can be explained by the fact that solute transport (e.g. Papp) is dependent on the sum of transport across all junctional pathways, whereas TEER is essentially dependent on areas with the lowest electrical resistance between single cells (Gaillard and de Boer, 2000). It has previously been reported that paracellular permeability to hydrophilic marker molecules affords a more accurate description than TEER of selective barrier function in tight epithelial cells (Artursson et al., 1993; Balda et al., 1996). An additional comparison between feline BBB model systems described in the present study included TEM analysis where endothelial cells which had been co-cultured with astrocytes possessed more junctional specialisation and were more closely associated with each other than endothelial cells which had been cultured with astrocyte-conditioned medium. Such observations supported TEER and paracellular permeability findings in the present study, and in others, such as the rat model system (Jeliazkova-Mecheva and Bobilya, 2003). A well-evaluated model of the BBB has major relevance in the study of species related issues in aspects of pathogenesis of neuroinflammation and neurodegeneration (Grant et al., 1998). While many viral agents may induce neuroinflammation in the cat (Fenner, 1994), one virus of direct relevance to the BBB is feline immunodeficiency virus. FIV is a model of HIV-1 infection in man. Like HIV-1, FIV enters the feline brain early in infection, at a time when there is considerable trafficking of lymphocytes through the BBB and choroid plexus (Bragg et al., 2002; Ryan et al., 2003). Therefore, use of this model has value in the dissection of interactions between FIV infected lymphocytes and the BBB, and lymphocyte trafficking through the BBB. In conclusion, we have developed and characterised an in vitro model of the feline BBB, which uses primary cultures of feline brain capillary endothelial cells and astrocytes grown as co-cultures on permeable filters. This model may have many future applications in studies related to neurodegeneration and neuroinflammation.
Acknowledgements We acknowledge the technical assistance of Brian Cloak and Christy King in the Department of Veterinary Pathology, University College Dublin, Matthew Campbell in the Conway Institute, University College Dublin for assistance with the Western blot technique and Dr. David John of Trinity College Dublin for assistance with Transmission Electron Microscopy. This work was funded by the Irish Higher Education Authority through the National Neuroscience Network (PRTLI-3).
References Abbott, N.J., Revest, P.A., 1991. Control of brain endothelial permeability. Cerebrovasc. Brain Metab. Rev. 3, 39–72. Annunziata, P., 2003. Blood–brain barrier changes during invasion of the central nervous system by HIV-1. Old and new insights into the mechanism. J. Neurol. 250, 901–906. Artursson, P., 1990. Epithelial transport of drugs in cell culture. I. A model for studying the passive diffusion of drugs over intestinal absorptive (Caco-2) cells. J. Pharm. Sci. 79, 476–482. Artursson, P., Ungell, A.L., Lofroth, J.E., 1993. Selective paracellular permeability in two models of intestinal absorption: cultured monolayers of human intestinal epithelial cells and rat intestinal segments. Pharm. Res. 10, 1123–1129. Audus, K.L., Borchardt, R.T., 1986. Characteristics of the large neutral amino acid transport system of bovine brain microvessel endothelial cell monolayers. J. Neurochem. 47, 484–488. Balda, M.S., Whitney, J.A., Flores, C., Gonzalez, S., Cereijido, M., Matter, K., 1996. Functional dissociation of paracellular permeability and transepithelial electrical resistance and disruption of the apical–basolateral intramembrane diffusion barrier by expression of a mutant tight junction membrane protein. J. Cell Biol. 134, 1031–1049. Bauer, H.C., Bauer, H., 2000. Neural induction of the blood–brain barrier: still an enigma. Cell Mol. Neurobiol. 20, 13–28. Bendinelli, M., Pistello, M., Lombardi, S., Poli, A., Garzelli, C., Matteucci, D., Ceccherini-Nelli, L., Malvaldi, G., Tozzini, F., 1995. Feline immunodeficiency virus: an interesting model for AIDS studies and an important cat pathogen. Clin. Microbiol. Rev. 8, 87–112. Boche, D., Hurtrel, M., Gray, F., Claessens-Maire, M.A., Ganiere, J.P., Montagnier, L., Hurtrel, B., 1996. Virus load and neuropathology in the FIV model. J. Neurovirol. 2, 377–387. Bowman, P.D., Ennis, S.R., Rarey, K.E., Betz, A.L., Goldstein, G.W., 1983. Brain microvessel endothelial cells in tissue culture: a model for study of blood–brain barrier permeability. Ann. Neurol. 14, 396–402. Bragg, D.C., Childers, T.A., Tompkins, M.B., Tompkins, W.A., Meeker, R.B., 2002. Infection of the choroid plexus by feline immunodeficiency virus. J. Neurovirol. 8, 211–224.
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244 Brightman, M., 1991. Implication of astroglia in the blood–brain barrier. Ann. N.Y. Acad. Sci. 633, 343–347. Broman, T., 1955. On basic aspects of the blood–brain barrier. Acta. Psychiatr. Neurol. Scand. 30, 115–124. Butt, A.M., 1995. Effect of inflammatory agents on electrical resistance across the blood–brain barrier in pial vessels of anaesthetized rats. Brain Res. 696, 145–150. Callanan, J.J., 1995. Feline immunodeficiency virus infection: a clinical and pathological perspective. In: Willett, B., Jarrett, O. (Eds.), Feline Immunology and Immunodeficiency. Oxford University Press, Oxford, pp. 111–130. Connor, R., Ho, D., 1992. Pathogenesis of human immunodeficiency virus. Semin. Virol. 3, 213–224. Cucullo, L., McAllister, M.S., Kight, K., Krizanac-Bengez, L., Marroni, M., Mayberg, M.R., Stanness, K.A., Janigro, D., 2002. A new dynamic in vitro model for the multidimensional study of astrocyte–endothelial cell interactions at the blood– brain barrier. Brain Res. 951, 243–254. Dallasta, L.M., Pisarov, L.A., Esplen, J.E., Werley, J.V., Moses, A.V., Nelson, J.A., Achim, C.L., 1999. Blood–brain barrier tight junction disruption in human immunodeficiency virus-1 encephalitis. Am. J. Pathol. 155, 1915–1927. de Boer, A.G., Gaillard, P.J., Breimer, D.D., 1999. The transference of results between blood–brain barrier cell culture systems. Eur. J. Pharm. Sci. 8, 1–4. Dehouck, M.P., Meresse, S., Delorme, P., Fruchart, J.C., Cecchelli, R., 1990. An easier, reproducible, and mass-production method to study the blood–brain barrier in vitro. J. Neurochem. 54, 1798–1801. Demeuse, P., Kerkhofs, A., Struys-Ponsar, C., Knoops, B., Remacle, C., van den Bosch de Aguilar, P., 2002. Compartmentalized coculture of rat brain endothelial cells and astrocytes: a syngenic model to study the blood–brain barrier. J. Neurosci. Methods 121, 21–31. Dow, S.W., Poss, M.L., Hoover, E.A., 1990. Feline immunodeficiency virus: a neurotropic lentivirus. J. Acquir. Immune Defic. Syndr. 3, 658–668. Dow, S.W., Dreitz, M.J., Hoover, E.A., 1992. Feline immunodeficiency virus neurotropism: evidence that astrocytes and microglia are the primary target cells. In: Dandekar, S., Pedersen, N. (Eds.), Feline Immunodeficiency Virus, vol. 35. Vet. Immunol. Immunopathol, pp. 23–35. Fenner, W., 1994. Diseases of the brain, spinal cord and peripheral nerves. In: Sherding, R. (Ed.), The Cat: Diseases and Clinical Management, vol. 2. WB Saunders Company, pp. 1507–1568. Fiala, M., Looney, D.J., Stins, M., Way, D.D., Zhang, L., Gan, X., Chiappelli, F., Schweitzer, E.S., Shapshak, P., Weinand, W., Graves, M.C., Witte, M., Kim, K.S., 1997. TNF-a opens a paracellular route for HIV-1 invasion across the blood–brain barrier. Mol. Med. 3, 553–564. Franke, H., Galla, H., Beuckmann, C.T., 2000. Primary cultures of brain microvessel endothelial cells: a valid and flexible model to study drug transport through the blood–brain barrier in vitro. Brain Res. Brain Res. Protoc. 5, 248–256. Gaillard, P.J., de Boer, A.G., 2000. Relationship between permeability status of the blood–brain barrier and in vitro permeability coefficient of a drug. Eur. J. Pharm. Sci. 12 (2), 95–102.
243
Gaillard, P.J., Voorwinden, L.H., Nielsen, J.L., Ivanov, A., Atsumi, R., Engman, H., Ringbom, C., de Boer, A.G., Breimer, D.D., 2001. Establishment and functional characterization of an in vitro model of the blood–brain barrier, comprising a co-culture of brain capillary endothelial cells and astrocytes. Eur. J. Pharm. Sci. 12, 215–222. Grant, G.A., Abbott, N.J., Janigro, D., 1998. Understanding the physiology of the blood–brain barrier: in vitro models. News Physiol. Sci. 13, 287–293. Henriksen, S.J., Prospero-Garcia, O., Phillips, T.R., Fox, H.S., Bloom, F.E., Elder, J.H., 1995. Feline immunodeficiency virus as a model for study of lentivirus infection of the central nervous system. Curr. Top. Microbiol. Immunol. 202, 167–186. Hertz, L., Juurlink, B., Fosmark, H., Schousboe, A., 1982. Astrocytes in Primary Cultures. In: Pfeiffer, S. (Ed.), Neuroscience Approached through Cell Culture. CRC Press, pp. 175–186. Jeliazkova-Mecheva, V.V., Bobilya, D.J., 2003. A porcine astrocyte/ endothelial cell co-culture model of the blood–brain barrier. Brain Res. Brain Res. Protoc. 12, 91–98. Kim, J.A., Tran, N.D., Wang, S.J., Fisher, M.J., 2003. Astrocyte regulation of human brain capillary endothelial fibrinolysis. Thromb. Res. 112, 159–165. Kondo, T., Imaizumi, S., Kato, I., Yoshimoto, T., 1994. Isolation and culture of brain endothelial cells and establishment of in vitro blood–brain barrier model. Cell. Transplant. (Suppl. 1), S35– S37. Luabeya, M.K., Dallasta, L.M., Achim, C.L., Pauza, C.D., Hamilton, R.L., 2000. Blood–brain barrier disruption in simian immunodeficiency virus encephalitis. Neuropathol. Appl. Neurobiol. 26, 454–462. MacLean, A.G., Orandle, M.S., MacKey, J., Williams, K.C., Alvarez, X., Lackner, A.A., 2002. Characterization of an in vitro rhesus macaque blood–brain barrier. J. Neuroimmunol. 131, 98–103. Marroni, M., Kight, K.M., Hossain, M., Cucullo, L., Desai, S.Y., Janigro, D., 2003. Dynamic in vitro model of the blood–brain barrier. Gene profiling using cDNA microarray analysis. Methods Mol. Med. 89, 419–434. McCarthy, K.D., de Vellis, J., 1980. Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J. Cell. Biol. 85, 890–902. Megard, I., Garrigues, A., Orlowski, S., Jorajuria, S., Clayette, P., Ezan, E., Mabondzo, A., 2002. A co-culture-based model of human blood–brain barrier: application to active transport of indinavir and in vivo–in vitro correlation. Brain Res. 927, 153– 167. Persidsky, Y., Stins, M., Way, D., Witte, M.H., Weinand, M., Kim, K.S., Bock, P., Gendelman, H.E., Fiala, M., 1997. A model for monocyte migration through the blood–brain barrier during HIV-1 encephalitis. J. Immunol. 158, 3499–3510. Reese, T.S., Karnovsky, M.J., 1967. Fine structural localization of a blood–brain barrier to exogenous peroxidase. J. Cell. Biol. 34, 207–217. Ringler, D.J., Hunt, R.D., Desrosiers, R.C., Daniel, M.D., Chalifoux, L.V., King, N.W., 1988. Simian immunodeficiency virusinduced meningoencephalitis: natural history and retrospective study. Ann. Neurol. 23 (Suppl.), S101–S107.
244
N.F. Fletcher et al. / Veterinary Immunology and Immunopathology 109 (2006) 233–244
Rubin, L.L., Hall, D.E., Porter, S., Barbu, K., Cannon, C., Horner, H.C., Janatpour, M., Liaw, C.W., Manning, K., Morales, J., Tanner, L.I., Tomaselli, K.J., Bard, F., 1991. A cell culture model of the blood–brain barrier. J. Cell. Biol. 115, 1725–1735. Ryan, G., Klein, D., Knapp, E., Hosie, M.J., Grimes, T., Mabruk, M.J., Jarrett, O., Callanan, J.J., 2003. Dynamics of viral and proviral loads of feline immunodeficiency virus within the feline central nervous system during the acute phase following intravenous infection. J. Virol. 77, 7477–7485. Sharer, L.R., 1992. Pathology of HIV-1 infection of the central nervous system: a review. J. Neuropathol. Exp. Neurol. 51, 3–11. Song, L., Pachter, J.S., 2003. Culture of murine brain microvascular endothelial cells that maintain expression and cytoskeletal association of tight junction-associated proteins. In Vitro Cell Dev. Biol. Anim. 39, 313–320.
Stanness, K.A., Guatteo, E., Janigro, D., 1996. A dynamic model of the blood–brain barrier ‘‘in vitro’’. Neurotoxicology 17, 481– 496. Steffan, A.M., Lafon, M.E., Gendrault, J.L., Koehren, F., De Monte, M., Royer, C., Kirn, A., Gut, J.P., 1994. Feline immunodeficiency virus can productively infect cultured endothelial cells from cat brain microvessels. J. Gen. Virol. 75, 3647–3653. Strom, S.C., Michalopoulos, G., 1982. Collagen as a substrate for cell growth and differentiation. Methods Enzymol. 82, 544– 555. To¨ro¨k, M., Huwyler, J., Gutmann, H., Fricker, G., Drewe, J., 2003. Modulation of transendothelial permeability and expression of ATP-binding cassette transporters in cultured brain capillary endothelial cells by astrocytic factors and cell-culture conditions. Exp. Brain Res. 221, 356–365.