Generation of an hiPSC-Derived Co-Culture System to Assess the Effects of Neuroinflammation on Blood-Brain Barrier Integrity

Abstract

The blood-brain barrier (BBB) regulates the interaction between the highly vulnerable central nervous system (CNS) and the peripheral parts of the body. Disruption of the BBB has been associated with multiple neurological disorders, in which immune pathways in microglia are suggested to play a key role. Currently, many in vitro BBB model systems lack a physiologically relevant microglia component in order to address questions related to the mechanism of BBB integrity or the transport of molecules between the periphery and the CNS. To bridge this gap, we redefined a serum-free medium in order to allow for the successful co-culturing of human inducible pluripotent stem cell (hiPSC)-derived microglia and hiPSC-derived brain microvascular endothelial-like cells (BMECs) without influencing barrier properties as assessed by electrical resistance. We demonstrate that hiPSC-derived microglia exposed to lipopolysaccharide (LPS) weaken the barrier integrity, which is associated with the secretion of several cytokines relevant in neuroinflammation. Consequently, here we provide a simplistic humanised BBB model of neuroinflammation that can be further extended (e.g., by addition of other cell types in a more complex 3D architecture) and applied for mechanistic studies and therapeutic compound profiling.

Keywords: blood–brain barrier (BBB) integrity; brain microvascular endothelial cells (BMECs); cytokine secretion; hiPSC co-culture; microglia; neuroinflammation; neurovascular unit (NVU); transendothelial electrical resistance (TEER); transwell.

Figure 1

(a) Timeline for the differentiation of the human inducible pluripotent stem cell (hiPSC)-derived brain microvascular endothelial cells (BMECs) (pink, in the transwell insert) and hiPSC-derived microglia (green, at the bottom of the plate) cultures, as well as when to combine the cultures to generate the transwell co-culture system. Images were created with BioRender.com. (b) Immunofluorescence image of hiPSC-derived BMECs with zonula occludens-1 (ZO-1, green) and 4′,6-Diamidino-2-Phenylindole (DAPI, blue) staining. (c) Immunofluorescence image of hiPSC-derived BMECs with occludin (green) and DAPI (blue) staining. (d) Bright field image of the hiPSC-derived microglia cells during differentiation. (e) Immunofluorescence image of hiPSC-derived microglia with ionized calcium-binding adaptor molecule 1 (IBA1, green) and DAPI (blue) staining.

Figure 2

Recombinant human transforming growth factor-beta 1 (rhTGF-β1) strongly diminishes hiPSC-derived BMEC barrier properties. (a) On day 13 of the differentiation process, basal medium supplemented with ±100 ng/mL recombinant human interleukin-34 (rhIL-34), ±25 ng/mL recombinant human macrophage-colony stimulating factor (rhM-CSF) and ±50 ng/mL rhTGF-β1 was added to the BMECs with the TEER measured 24 h later. The percentage change in trans-endothelial electrical resistance (TEER) was normalised to the basal medium experimental condition (One-way ANOVA with Dunnett’s multiple comparisons test, p values displayed on the graph, n = 3). (b) On day 13, the medium on the hiPSC-derived BMECs was replaced with either human endothelial medium or co-culture medium (Table S5), and the TEER was measured 24 h later. The percentage change in TEER was normalised to the human endothelial medium experimental condition (two-tailed unpaired t-test, p values displayed on the graph, n = 3). Each graphical symbol shape (square, triangle and circle) represents each set of experimental replicates.

Figure 3

hiPSC-derived microglia exposed to lipopolysaccharides (LPS) disrupts hiPSC-derived BMEC barrier integrity. (a) On day 14, 0, 1, 10 or 100 ng/mL of LPS was added to the basolateral compartment of hiPSC-derived BMECs (monoculture), and the percentage change in TEER between 0 h and 9 h calculated (0 h TEER baseline: 2395–3741 Ω·cm²). Statistical comparisons were performed against the 0 ng/mL LPS monoculture experimental condition (One-way ANOVA with Dunnett’s multiple comparisons test, p values displayed on the graph. (b) On day 13 of the differentiation process, both hiPSC-derived cultures were combined to form the co-culture system. Then, 24 h later, 0, 1, 10 or 100 ng/mL of LPS was added to the basolateral compartment and the percentage change in TEER between 0 h and 9 h calculated (0 h TEER baseline: 2134–3656 Ω·cm²). Statistical comparisons were performed against the 0 ng/mL LPS co-culture experimental condition (One-way ANOVA with Dunnett’s multiple comparisons test, p values displayed on the graph. (c) The time-course of the percentage change in TEER observed for the hiPSC-derived BMECs (monoculture) upon LPS (0, 1, 10 or 100 ng/mL) exposure. (d) The time-course of the percentage change in TEER observed for the hiPSC-derived co-culture system upon LPS (0, 1, 10 or 100 ng/mL) exposure. In (a,b), we performed the experiment three times (three biological replicates), and in each experiment, we have included three technical replicates per condition. Each symbol in a condition (square, triangle and circle) indicates a separate experiment, and the bar graph illustrates the mean ± standard deviation of the three experiments. In (c,d) we performed three time-course experiments with four different conditions, and we included at least three technical replicates per condition. At each timepoint, the mean ± standard deviation of the three biological replicates is shown for each condition.

Figure 4

hiPSC-derived microglia exposed to LPS secrete elevated levels of cytokines. After 0 and 9 h of LPS (0, 1, 10, 100 ng/mL) exposure in the co-culture system, cell culture supernatant was removed and cytokine concentration quantified with a Luminex^® multiplex assay. Data displays the mean change in cytokine concentrations between 0 h and 9 h of LPS exposure. Each graphical symbol shape (square, triangle and circle) represents each set of biological replicates with at least three technical replicates. Mean values ± standard deviation of cytokine concentrations for the 0 h and 9 h timepoints are provided in Table S6.