Materials for blood brain barrier modeling in vitro

Abstract

Brain homeostasis relies on the selective permeability property of the blood brain barrier (BBB). The BBB is formed by a continuous endothelium that regulates exchange between the blood stream and the brain. This physiological barrier also creates a challenge for the treatment of neurological diseases as it prevents most blood circulating drugs from entering into the brain. In vitro cell models aim to reproduce BBB functionality and predict the passage of active compounds through the barrier. In such systems, brain microvascular endothelial cells (BMECs) are cultured in contact with various biomaterial substrates. However, BMEC interactions with these biomaterials and their impact on BBB functions are poorly described in the literature. Here we review the most common materials used to culture BMECs and discuss their potential impact on BBB integrity in vitro. We investigate the biophysical properties of these biomaterials including stiffness, porosity and material degradability. We highlight a range of synthetic and natural materials and present three categories of cell culture dimensions: cell monolayers covering non-degradable materials (2D), cell monolayers covering degradable materials (2.5D) and vascularized systems developing into degradable materials (3D).

Introduction

Endothelial cells that line the inner surface of the vasculature are specialized according to the architecture and the need of the tissue it supplies [1]. In particular, brain microvascular endothelial cells (BMECs) that constitute brain capillaries, form a continuous endothelium that separates the cerebral tissue from the bloodstream. This endothelium, termed the blood brain barrier (BBB), maintains the stability of the brain tissue composition by regulating ion and macromolecule diffusion through the capillary wall [2]. The BBB ensures homeostasis in the central nervous system (CNS) by protecting neurons from the entrance of blood-circulating pathogenic agents into the brain. Therefore, CNS diseases such as amyotrophic lateral sclerosis, epilepsy, Alzheimer’s and Parkinson’s diseases are associated with BBB dysfunction [[3], [4], [5]]. Conversely, the BBB also rejects most of the large drugs (>400 Da) targeting the brain [6]. This renders drug accumulation at relevant therapeutic concentrations challenging, for treatment of a variety of neurological diseases. The development of BBB tissue models is of great importance to understand how the microenvironment influences the BBB selectivity property and establish new therapeutic strategies for drug diffusion into the brain [[7], [8], [9]].

In order to build relevant BBB tissue models, it is necessary to understand the in vivo structure of brain capillaries. A complex and dynamic microenvironment surrounds brain capillaries and influences BBB integrity. BMECs closely interact with two other cell types, pericytes and astrocytes, which wrap 30 % and 98 % of the surface of the endothelium, respectively[10]. They provide mechanical support for the vascular endothelium and secrete biochemical factors required to maintain BBB integrity [11,12]. Pericytes and astrocytes both participate in the maturation and maintenance of BBB properties [13]. In some brain diseases, both astrocytes and pericytes affect BBB selective permeability [14,15]. For example, during inflammation astrocytes secrete pro-inflammatory mediators that increase BBB permeability and support leukocyte infiltration [14]. During cerebral ischemia, pericytes represent a source of matrix metalloproteinases (MMPs) that locally degrade the BBB and contribute to plasma leakage [15]. Both pericytes and astrocytes are separated from the BMECs by a vascular basement membrane (BM). The BM provides mechanical support to the BMECs and retains most of the biochemical cues secreted by the surrounding cells. The diffusion gradient of these biochemical cues oriented towards the BBB strongly influences BBB integrity [16,17]. Finally, as part of the vascular system, the BMECs perceive active mechanical stimuliin the form of shear stress exerted by the blood flow and the circumferential stretch induced by blood pressure [18]. Recent studies document the impact of circulatory pressure and blood flow on blood brain barrier function [[18], [19], [20]]. Importantly, brain endothelial cells react to shear stress and circumferential stretch by remodeling their cytoskeletal filaments and producing stress fibers [[18], [19], [20]]. These fibers are aligned along the direction of flow and perpendicular to the direction of stretch. The development of BBB models to accurately predict the passage of drugs through the barrier relies on the integration of these biochemical and mechanical factors as parameters that participate in BBB maturation and function.

Two types of models are commonly used for drug testing: animal models and in vitro cell models. Animal models, compared to in vitro models, preserve the native matrix architecture and can further predict cellular response in a physiological context. However, when it comes to the BBB, there is a mismatch between animals and humans in terms of cellular and matrix composition. This usually renders animal models poorly predictive of the human BBB response. As a result, almost 80 % of animal-drug candidates fail in clinical trials due to high-level toxicity and/or low therapeutic efficiency. Conversely, in vitro cell models culture cells from a human origin in a well-defined environment to mimic tissue and organ functions. Various human brain endothelial cells have been successfully used to model aspects of the BBB, including immortalized cell lines, primary cells, and induced pluripotent stem cells (iPSC) [2,22,23]. BBB in vitro models rely on the culture of BMECs at the interface between two compartments, luminal and abluminal. Several cell culture configurations exist, which have been classified into three categories of models in this review: 2-dimensional (2D), 2.5D and 3D models. 2D models refer to the culture of BMEC monolayers on top of flat and stiff synthetic substrates [24]. Among them, the Transwell® model relies on the culture of BMECs on top of a suspended porous membrane. Transwells are commercially available and widely used to model the BBB for drug screening [25]. They offer easy access to both sides of the barrier and are relatively cheap and easy-to-use. However, they are poorly representative of BMEC native environment in term of substrate mechanical stiffness and cell curvature [26]. To overcome this limitation, soft cell-degradable materials, such as hydrogels, give rise to two main cell culture configurations [6]. The culture of BMEC monolayer inside a predesigned tubular structure made of hydrogels, named as 2.5D models in this review, and the embedding of cells into hydrogels to induce capillary angiogenesis, which we term 3D models in this review.

BMECs cultured within in vitro models are exposed to a variety of biochemical (e.g., cell culture media composition and substrate protein coating) and physical (e.g., substrate architecture and media flow) cues that may affect cell phenotype and impact the barrier formation. The impact of biochemical signaling cues on in vitro BBB formation and integrity is well-documented [27,28]. In the last decade, there have been multiple reports concerning the impact of shear stress on BBB integrity due to the introduction of mechanical stimuli from media flow applied to BMECs [[29], [30], [31]]. However, the impact of the matrix-induced mechanical signaling on in vitro BBB property has been poorly described (Fig. 1).

Here, we review materials that have been used as substrates to culture brain endothelial cells in 2D, 2.5D and 3D BBB models. We also explore how material properties, such as elasticity, structure, and degradability (by the cells during matrix remodeling), affect endothelial barrier function by developing an analytical framework to guide the use of BBB models. Ultimately, this review offers insight into the emerging organ-on-a-chip technology proposed as a promising option to fill the gap in drug screening. These new in vitro models include a variety of environmental factors, such as topographical guidance, mechanical stimulation, biochemical gradients and spatially defined co-culture [32]. BBB-on-a-chip models use biomimetic materials to culture BMECs indifferent configurations (e.g., 2D, 2.5D, and 3D) to improve endothelial barrier development and maintenance.