Human bone marrow-derived mesenchymal stem cells play a role as a vascular pericyte in the reconstruction of human BBB on the angiogenesis microfluidic chip

Abstract

A blood-brain barrier (BBB) on a chip similar to the in vivo BBB is important for evaluating the efficacy of reparative cell therapeutics for ischemic stroke in vitro. In this study, we established human BBB-like microvasculature on an angiogenesis microfluidic chip and analyzed the role of human pericytes (hPCs) and human astrocytes (hACs) on the architecture of human brain microvascular endothelial cells (hBMEC)-derived microvasculature on a chip. We found that human bone marrow mesenchymal stem cells (hBM-MSCs) play a role as perivascular pericytes in tight BBB reformation with a better vessel-constrictive capacity than that of hPCs, providing evidence of reparative stem cells on BBB repair rather than a paracrine effect. We also demonstrated that pericytes play an important role in vessel constriction, and astrocytes may induce the maturation of a capillary network. Higher expression of VEGF, SDF-1α, PDGFRβ, N-cadherin, and α-SMA in hBM-MSCs than in hPCs and their subsequent downregulation with hBMEC co-culture suggest that hBM-MSCs may be better recruited and engaged in the BBB-microvasculature than hPCs. Collectively, the human BBB on a chip may be adopted as an alternative to evaluate in vitro cellular behavior and the engagement of cell therapeutics in BBB regeneration and may also be used for studying stroke.

Introduction

The neurovascular disease stroke is the third leading cause of death worldwide and leads to long-term disability and dependency [1]. Approximately, 80% of strokes are caused by ischemia and the remaining 20% by hemorrhage [2]. Currently, stroke therapeutics are limited owing to the absence of successful tissue regenerative treatments, even though many previous studies have focused on cell therapy [3]. A variety of stem cell therapeutics, such as mesenchymal stem cells (MSCs), neuronal stem cells, and induced pluripotent stem cells (iPSCs) have shown efficacy in animal stroke models in preclinical studies [[4], [5], [6]]. However, several limitations of the animal stroke model make it difficult to evaluate and prove the efficacy of stroke cell therapeutics. The animal stroke model is technically challenging; it is permitted only for skilled experts and requires numerous expensive medical tools specifically designed for the animal. Furthermore, efficacy evaluation mainly relies on behavioral tests, endpoint infarct size, and tissue staining, which require a long-time analysis [7]. Therefore, it is not possible to estimate the functional role and kinetics of how individual stem cells are engaged in the tissue repair and regeneration process as reparative tissue-resident cells in stroke animals. Currently, many cell therapeutics showing good efficacy in animal stroke models have failed to reproduce their effectiveness in clinical studies [8]. Thus, an in vitro system to reconstruct the blood-brain barrier (BBB) similar to the in vivo brain microvasculature and its validation standard may be essential to develop therapeutic stem cells or drugs that can repair brain disease and damage.

To adopt and apply the in vitro BBB-like microvasculature as a substitute or an alternative test model of the BBB in vivo, quantitative analysis of the structural and functional characteristics of the BBB microvasculature needs to be established. The role of individual cellular components in the architecture of the reconstructed BBB needs to be defined, and their criteria must be standardized. In the human BBB, the vessel diameter and segment length of the fine web-like microvasculature are in the range of 8–10 μm and 50–100 μm, respectively [9]. In addition, structurally, the innermost apical linings of endothelial cell tubules are tightly sealed off with tight junctions and adherens junctions, and the basal linings are covered by a basement membrane along with encircling pericytes and astrocyte end-feet [10]. Recently, several in vitro BBB models have been developed. The Matrigel tubular network is most commonly used to test the tubule-forming capacity of endothelial cells or their precursor cells with accessory cells, such as pericytes and astrocytes; however, the patency of the tubular network has not been clearly demonstrated [11]. Since blood flow is important for vascular remodeling and the acquisition of functional microvasculature, a perfusable capillary network on microfluidic chips or equivalents has been developed [[12], [13], [14]]. Ingber et al. reported an in vitro BBB model using a pre-lumenized collagen gel system with a vessel diameter of 600–800 μm [12]. Kamm et al. established an in vitro BBB microfluidic system with iPSC-induced endothelial cells, human pericytes (hPCs), and human astrocytes (hACs) using a microfluidic vasculogenesis model with a vessel diameter of 25 ± 6 μm and a segment length of 136 ± 24 μm, which were closest to that of the in vivo BBB [13]. Jeon et al. developed an in vitro BBB model with human brain microvascular endothelial cells (hBMECs), hPCs, and hACs using a multi-channel microfluidic angiogenesis model with a vessel diameter of 34.64 μm and confirmed the vessel permeability with a fluorescent dye and the drug transport capacity through p-glycoprotein of the in vitro BBB [14]. The recent development of in vitro BBB-like microvasculature on a microfluidic chip has become much closer to the BBB in vivo, both structurally and functionally. However, quantitative criteria to compare the morphological and functional properties of the microvessel between in vitro reconstructed BBB and in vivo BBB are essential for standardized validation.

BBB regeneration is an essential step in ischemic stroke to recover brain function [15]. The BBB regulates the molecular exchange between the brain parenchyma and the blood, establishes its barrier function, and maintains brain homeostasis [16]. Under ischemic stroke, pericytes dissociate from the BBB architecture during the acute phase, causing BBB leakage and the loss of barrier function [17]. Then, immune cells can penetrate the brain parenchyma, initiating neuroinflammation, which causes neuronal cell death and brain damage [18]. Thus, vascular pericytes critically contribute to the formation and integrity of the blood vessel walls within the microcirculation and maintain vascular stability [19]. More importantly, vascular pericytes contribute to remodeling of the microcirculation through the dynamic interactions between pericytes and endothelial cells during angiogenic sprouting, where activated endothelial cells recruit pericytes through the secretion of PDGF-ΒΒ, which binds to PDGFR-β with high affinity [20]. In particular, bone marrow-derived MSCs (BM-MSCs) have been widely studied for the development of cell therapy for stroke and other chronic central nervous system diseases [4]. Many previous preclinical studies have reported that BM-MSCs regulate the inflammatory environment through the paracrine effect of their released cytokines and growth factors. However, BM-MSCs are also considered to act similarly to vascular pericytes and engage in BBB reconstruction [21,22].

In this study, we reconstructed the BBB-like microvasculature on a microfluidic chip with different combinations of hBMECs, hPCs, and hACs and explored how each cell type plays a role in the reconstruction of capillary network architecture and maturation of BBB-like microvasculature by measuring vessel diameter, segment length, number of nodes, meshes, segments, and branches using the NIH Image J angiogenesis program. Based on quantitative criteria of the structural features of capillary networks, biochemical analysis of tight junction formation and basement membrane, and functional test of vascular permeability barrier, we demonstrated the engagement of hBM-MSCs on the BBB-like microvasculature as the perivascular pericyte and comparatively estimated their efficacy with hPCs in vitro. We found that hBM-MSCs seem to retain a better capacity to encircle and constrict the endothelial vessel as a perivascular pericyte than hPCs, and hACs may also play a more important role in the maturation of the fine web-like capillary network of the BBB-like microvasculature rather than in vascular constriction. To delineate the role of hBM-MSCs as perivascular pericytes, levels of proteins involved in cell migration, endothelial association, vasoconstriction, and basement membrane were compared using Western blot analysis and enzyme-linked immunosorbent assay (ELISA) between hBM-MSCs and hPCs in single and co-culture conditions similar to those in the microfluidic chip. Finally, we discussed that the BBB-like microvasculature on a microfluidic chip may be adopted as a competent alternative to evaluate in vitro cellular behavior and the engagement of cell therapeutic candidates in BBB regeneration and can be further developed to the BBB in the disease state or ischemic stroke in the future.