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

https://doi.org/10.1016/S0928-0987(00)00123-8Get rights and content

Abstract

Objective: The aim was to establish a flexible, abundantly available, reproducible and functionally characterized in vitro model of the blood–brain barrier (BBB). Methods: In a first step, bovine brain capillaries and newborn rat astrocytes were isolated. Subsequently, a co-culture of primary brain capillary endothelial cells (BCEC) on semi-permeable filter inserts, with astrocytes on the bottom of the filter was established. The cell material was characterized on the basis of specific cell-type properties and (functional expression of) specific BBB properties. Results: BCEC displayed: (1) characteristic endothelial cell morphology; (2) expression of endothelial cell markers (i.e., CD51, CD62P, CD71 and cadherin 5); (3) marginal F-actin localization; (4) tight junction formation between the cells; (5) expression of γ-glutamyl-transpeptidase (γ-GTP); (6) expression of P-glycoprotein (Pgp); (7) functional transendothelial transferrin transport and uptake; (8) restriction of paracellular transport; and (9) high transendothelial electrical resistance (TEER). Astrocytes displayed characteristic astrocyte morphology and expressed glial fibrillary acidic protein (GFAP). Co-culture with astrocytes increased TEER and decreased paracellular transport. In addition, expression of the glucocorticoid receptor (GR) was demonstrated in the endothelial cells of the BBB, while no expression of the mineralocorticoid receptor (MR) was found. Conclusions: A high quality and mass-production in vitro BBB model was established in which experiments with physiological (e.g., regulation of BBB permeability), pharmacological (e.g., pharmacokinetics and pharmacodynamics) and pathophysiological (e.g., disease influence on BBB permeability) objectives can be reproducibly performed.

Introduction

The in vitro blood–brain barrier (BBB) model is an isolated system that enables the selective study of the physiology, pharmacology and pathophysiology of the BBB. Ideally, an in vitro model comes as close to the in vivo situation as possible, without losing the advantages of being an in vitro system. Furthermore, the system should be flexible, reproducible, abundantly available and functionally characterized based on specific cell-type properties and on the functional expression of specific BBB properties.

Recently several of the different procedures which have been applied to prepare in vitro BBB models have been described in detail in the context of a Concerted Action sponsored by the EEC (de Boer and Sutanto, 1997). These systems differ with respect to the used isolation procedures, culture conditions and model preparation. Moreover, primary, subpassaged and immortalized cell cultures have been used from, e.g., brain capillaries, aortic endothelial cells, umbilical vein endothelial cells and also epithelial cell lines, originating from human, primate, bovine, porcine, rodent and murine species. Since astrocytes were known to induce and maintain BBB properties in endothelial cells (Arthur et al., 1987; Janzer and Raff, 1987), these cells have been co-cultured with primary cultured or subpassaged rat astrocytes and rat or human glioma tumor cell lines, while also cell-conditioned media and pericytes have been used (Dehouck et al., 1990; Rubin et al., 1991; Meyer et al., 1990, Meyer et al., 1991). It is obvious that such a diversity in the currently available different in vitro BBB systems leads to systems with different characteristics, which makes the comparison of results difficult. Therefore, we suggested that each in vitro BBB model has to be functionally characterized and optimized to the needs of the research group (de Boer et al., 1999).

In vitro BBB systems express many specific properties (de Boer and Breimer, 1996), which could be employed for their characterization (de Boer et al., 1999). In particular, endothelial cells may be characterized by their characteristic morphology in culture, i.e., cobblestone shape (when growing in a cluster) and spindle shape (when confluent), with a centered oval nucleus, while pericytes may be characterized by their distinct irregular morphology. The endothelial cells may subsequently be characterized using a panel of general endothelial cell-specific properties, e.g., expression of endothelial cell-specific cluster of differentiation (CD) antigens, Factor VIII-related antigen, non-thrombogenicity, low leukocyte adherence, release of vasoactive compounds (nitric oxide, endothelin-1 and prostacyclins), uptake of DiI-labeled-acetylated low density lipoprotein (DiI-Ac-LDL), lectin binding, and expression of angiotensin-converting enzyme, alkaline phosphatase and monoamine oxidase. In addition, typical barrier markers, like the formation of tight junctions, expression of γ-glutamyl-transpeptidase (γ-GTP), P-glycoprotein (Pgp), glucose transporter, transferrin receptor, marginal F-actin localization and high density of mitochondria may be used, although these are not exclusively specific for the BBB endothelial cells. Astrocytes can be identified by the expression of glial fibrillary acidic protein (GFAP) and their specific morphology in culture. Moreover, when the in vitro BBB system is cultured in a two-compartment system (e.g., on filters), BBB functionality may be characterized by restricted paracellular transport of marker substances, transendothelial electrical resistance (TEER) and analysis of the transport of substrates for transporters expressed on the BBB. Finally, expression of molecules which are unrelated to the BBB phenotype, but relevant for the specific line of research (e.g., receptors or enzymes), may be determined on the cell material constituting the in vitro BBB.

In this paper, the isolation procedure of bovine brain capillaries and newborn rat astrocytes was described. In addition, the primary culture and characterization of brain capillary endothelial cells (BCEC) and astrocytes was described. Furthermore, the establishment and functional characterization of our in vitro (co)-culture model of the BBB was described.

Section snippets

Cell culture

Brain capillaries were isolated from bovine (calf) brain, obtained at a slaughterhouse (Molendijk, Nieuwerkerk a/d IJssel, The Netherlands) from freshly killed animals. The brain was transported to the laboratory in ice-cold phosphate-buffered saline (PBS, pH 7.4). Meninges and white matter were removed and gray matter was collected in DMEM, supplemented with 10% fetal calf serum (DMEM+S). Blood vessel fragments were prepared by manual homogenization using a Wheaton homogenizer and subsequently

Cell culture

Bovine brain capillaries and BCEC were isolated and cultured essentially according to procedures described by Rubin et al. (1991) and published by us in great detail before (de Boer et al., 1997). However, the number of experiments from one isolation and timing possibilities for the preparation of the in vitro model of the BBB were improved by freezing and storing the brain capillaries (personal communication with Louise Morgan from the laboratory of Lee Rubin, Eisai London Research). The

Discussion

In this paper, the isolation procedures and primary culture methods for bovine brain capillaries, BCEC and rat astrocytes, along with the specific requirements to establish a high quality in vitro co-culture model of the BBB are presented. The cell culture material has been extensively examined for the expression and maintenance of known BBB properties. In addition, the influence of astrocytes on the restriction of paracellular transport was determined.

To make sure that all (known and unknown)

Acknowledgements

Louise Morgan and Lee Rubin (help with restart of BCEC model), Dhr Bethlehem (Firma Molendijk, coordinating the slaughterhouse affairs), Marc Fluttert and Win Sutanto (help and advice GR/MR binding studies), colleagues from the Department of Toxicology (help with Western blot analysis and fluorescence microscopy), Inez van der Sandt (cell culture of MDR1 transfected and control LLC-PK1 cells), Margret Blom-Roosemalen (general support) and Hari Sharma (EM photography) are acknowledged for their

References (25)

  • A.G de Boer et al.
  • H.E de Vries et al.

    Effect of endotoxin on permeability of bovine cerebral endothelial cell layers in vitro

    J. Pharmacol. Exp. Ther.

    (1996)
  • Cited by (289)

    • Tissue-based in vitro and ex vivo models for blood–brain barrier permeability studies

      2024, Concepts and Models for Drug Permeability Studies: Cell and Tissue based In Vitro Culture Models
    View all citing articles on Scopus
    View full text