Development and Evaluation of Rosette Array Technology for HumanNeurodevelopmental Toxicity Screening

Project Summary Pre-clinical and pre-field toxicology testing of new drugs and chemicals does not routinely include direct assessment of human developmental neurotoxicity (DNT). Current standards for DNT testing require use of animal models with limited throughput and significant differences from human central nervous system (CNS) development. As such, many chemical products have required post-approval (EPA/FDA) restrictions or cancellations due to human DNT. Such limitations of the current regulatory DNT testing paradigm have prompted increased interest in quantitative high-throughput screening (qHTS) using human pluripotent stem cell (hPSC)-based approaches. Here, we propose to use Rosette Array (RA) technology to develop an hPSC- derived qHTS platform (i.e., qHTS-RAs) for effective and efficient DNT testing. Rosette Array technology standardizes in vitro derivation of human neural rosettes tissues that are mimetic of transverse slices of the human neural tube, the anlage of all CNS tissue. Although neural rosettes are not an exact recapitulation of in vivo primary neurulation, they exhibit the same cell phenotypes, tissue cytoarchitecture, and are derived using morphogenetic signaling pathways endogenous to the in vivo neural tube formation process. Rosette arrays are the first in vitro technology to enable spatial and temporal control of neural rosette emergence in a microarray format. This enables rapid assessment via microscope image analysis. This high-yield, standardized generation of in vitro neural tube analogs enables the repeatability necessary to feasibly incorporate hPSC-based CNS morphogenic readouts into quantitative high-throughput toxicology screening. Thus, we hypothesize that the qHTS-RA platform could increase throughput and accuracy of human DNT risk assessment, allowing consolidation and scale-up of commercial DNT screening. Phase 1 Aims will validate the broad applicability of the qHTS-RA platform, confirming compatibility with human induced pluripotent stem cell (hiPSC) lines and establishing methods for automated image acquisition and batch analysis. Phase 2 Aims to validate assay fitness with a 100 compound DNT reference library screen, in which automated AI image analysis is utilized to further increase accuracy and maximize throughput. If successful, the resulting qHTS-RA platform could replace multiple current DNT assays and increase confidence in toxicological readouts relevant to outcomes unique to human physiology. Thus, the work proposed here could have a transformative effect on DNT research, regulatory efforts to prevent DNT exposure, and future translatability of hPSC-based organoid models for broad study of human development.