The effect of Polybrominated diphenyl ethers at the fetal blood-brain-barrier: evaluation using a microphysiological system.

Abstract

Background: Glutamate dysregulation leading to neuronal excitotoxicity and neuroinflammation are associated with neurological disorders, specifically autism spectrum disorders (ASD) in preterm neonates. The lack of physiologically relevant in vitro models has limited mechanistic understanding of glutamate dysregulation and neuroinflammation during pregnancy. This study evaluated the effect of environmental pollutant and flame retardant, Polybrominated Diphenyl Ethers (PBDE) 99 and 47, on cell viability, glutamate dysregulation, and neuroinflammation using a microphysiologic system (MPS) of human fetal blood-brain-barrier organ on a chip (FB-OOC).

Methods: The FB-OOC is composed of 3-cell culture chambers, connected by microchannels, containing 1) human brain microvessel endothelial cells (HBMEC), 2) human vascular pericytes (HBVP), and 3) a triculture of neurons, astrocytes, and microglia in a 5:2:1 ratio, respectively. To assess the effect of toxicants on glutamate dysregulation and neuroinflammation, control (standard media) endothelial cells were exposed to PBDE 99 and 47 (150 ng/mL). To mimic the passage of PBDE through the placenta, endothelial cells in FB-OOC were exposed to conditioned PDBE media (1:1) derived from a placenta-OOC. In parallel, triculture cells were directly treated in a 96-well plate. Dextran propagation over 72 h confirmed FB barrier function. The activation status of microglia was determined using immunocytochemistry for CD11 b and Iba1, respectively. Cell morphology (microscopy), cell cytotoxicity (Lactate Dehydrogenase and glutamate assays), and cytokines (multiplex assay) were measured.

Results: Physiologic FB-OOCs were characterized by 1) viable cell cultures expressing standard cell morphologies and cell-specific markers, 2) barrier formation confirmed by decreased dextran propagation over 72 h, and 3) baseline glutamate and pro-inflammatory cytokine production. On-chip PBDE and placenta-derived metabolites of PBDE treatment in the endothelial chamber induced cell cytotoxicity and significant upregulation of glutamate in the triculture but did not induce neuroinflammation nor microglia activation compared to the controls. Conversely, 2D triculture experiments showed direct PBDE treatment-induced significant neuroinflammation (TNF-α, GM-CSF, IL-8) compared to PBDE placenta-derived metabolites or controls.

Conclusion: This study established an FB model that recreated intercellular interactions. We report PBDE-induced glutamate dysregulation, often associated with the development of ASD, independent of neuroinflammation.