Citation:

Knudsen, T., T. Zurlinden, K. Saili, AND Nancy C. Baker. Computational Simulation: agent-based modeling of neurovascular development. 5th International Conference on Developmental Neurotoxicity Testing (DNT5), Konstanz, GERMANY, April 05 - 08, 2020.

Impact/Purpose:

Abstract submitted to the 5th International Conference on Developmental Neurotoxicity Testing (DNT5) scheduled for April 2020 in Konstanz Germany. This invitational travel is to present at the DNT5 conference. This series of conferences has started in 2006 with the aim to offer a platform to the most relevant experts of the field to discuss the state-of-art and the challenges which the DNT field is facing. To do so, as in the past editions, the program will include sessions on the application of hPSC-based assays, organoids, zebrafish models, and in silico approaches both for DN disease modeling and regulatory decision-making processes. Notably, this year the conference has also been chosen to host the annual meeting of the OECD DNT expert group. The attached abstract reflects an invitational talk to be presented in a session on “Novel Approaches to Test Methods”.

Description:

For predictive toxicology, in vitro data and biological information must be organized and assimilated into computer models that bridge different levels of biological organization. This concept has been demonstrated with several virtual tissue model prototypes that predict adverse developmental phenotypes with ToxCast data. For example, a cellular agent-based model of the neurovascular unit provides a means to simulate chemical disruption of neural tube vascularization at critical stages of blood-brain barrier (BBB) formation. The ‘computational neurovascular unit’ (cNVU) can be used to translate concentration-response data from ToxCast assays into integrative in silico models of neurodevelopmental toxicity. With a cNVU, interactions between neuroprogenitor cells, microglia, and endothelial cells can be reconstructed in silico, cell-by-cell and interaction-by-interaction, leading to emergent phenomena that can be mapped quantitatively as the system develops. This enables a unique way to generate (or test) how particular biomolecular lesion(s) might interact with signaling pathways driving early neurodevelopmental growth and differentiation. Items for discussion include how smart these computer models must be to support decision-making in the animal-free (3Rs) zone, and practical considerations for technology development, application, and training for predictive toxicology. Disclaimer: does not reflect EPA policy.

Record Details: