You are here:
Computational Modeling of the Neurovascular Unit to Predict Microglia Mediated Effects on Blood-Brain Barrier Formation
Citation:
Zurlinden, T., K. Saili, R. Spencer, N. Baker, AND T. Knudsen. Computational Modeling of the Neurovascular Unit to Predict Microglia Mediated Effects on Blood-Brain Barrier Formation. Presented at Teratology Society, Clearwater, Florida, June 23 - 27, 2018. https://doi.org/10.23645/epacomptox.7098863
Impact/Purpose:
We utilized several in vitro platforms with human endothelial and neural cells (neuroprogenitor, neural crest, and neural network) to screen for potential angiogenic and/or neurogenic disruption in the ToxCast dataset and integrated these data embryologically using an in silico agent based model (ABM) specified with knowledge of multicellular pathways driving BBB development
Description:
Blood-brain barrier (BBB) development is a complex process invoked by heterotypic cellularization of the embryonic neuroepithelium: vascular invasion from the perineural vascular plexus; neuroprogenitor proliferation, migration and differentiation in the subventricular zone; and cytokine/growth factor signaling interactions with microglia. These interactions are mediated by cell-cell signaling pathways that may be perturbed by genetic and environmental factors, leading to developmental neurovascular toxicity. We utilized several in vitro platforms with human endothelial and neural cells (neuroprogenitor, neural crest, and neural network) to screen for potential angiogenic and/or neurogenic disruption in the ToxCast dataset. These assays generated 18 features for behaviors such as proliferation, migration, tubulogenesis, and neural network formation. Potency metrics for the assay endpoints were calculated for 38 ToxCast chemicals and clustered to determine groups of chemicals with similar bioactivity profiles. To integrate these data embryologically we built an in silico agent based model (ABM) specified with knowledge of multicellular pathways driving BBB development. Executing the ABM simulated microglial-endothelial signaling (Notch/dll4, CSF1, VEGFA, VEGFC, sFLT1). To demonstrate the predictive capability of the integrated in silico and in vitro system, a simulated dose-escalation study was conducted for Oxytetracycline dihydrate and Mancozeb, which demonstrated an incremental concentration response for three critical signaling pathways mediating microglial-endothelial interplay. In the case of Oxytetracycline dihydrate, the computational model predicted decreased branching and vessel length at low concentrations due to altered VEGFC signaling, commensurate with the results from endothelial-specific in vitro assays. Studies are underway to implement neurogenic signaling into the ABM. The integration of in vitro data with in silico models provides a path forward to mechanistic prediction of developmental neurovascular toxicity and demonstrates a methodology for screening chemicals in silico and validating these biological predictions using cell-based in vitro assays. This abstract may not reflect US EPA policy.
URLs/Downloads:
DOI: Computational Modeling of the Neurovascular Unit to Predict Microglia Mediated Effects on Blood-Brain Barrier Formation
ZURLINDEN_A MECHANISTIC ABM FOR BRAIN ANGIOGENESIS_M.PDF (PDF, NA pp, 3554.528 KB, about PDF)