Quantitative Live Imaging of Human Embryonic Stem Cell Derived Neural Rosettes Reveals Structure-Function Dynamics Coupled to Cortical Development

Neural stem cells (NSCs) are progenitor cells for brain development, where cellular spatial composition (cytoarchitecture) and dynamics are hypothesized to be linked to critical NSC capabilities. However, understanding cytoarchitectural dynamics of this process has been limited by the difficulty to quantitatively image brain development in vivo. Here, we study NSC dynamics within Neural Rosettes—highly organized multicellular structures derived from human pluripotent stem cells. Neural rosettes contain NSCs with strong epithelial polarity and are expected to perform apical-basal interkinetic nuclear migration (INM)—a hallmark of cortical radial glial cell development. We developed a quantitative live imaging framework to characterize INM dynamics within rosettes. We first show that the tendency of cells to follow the INM orientation—a phenomenon we referred to as radial organization, is associated with rosette size, presumably via mechanical constraints of the confining structure. Second, early forming rosettes, which are abundant with founder NSCs and correspond to the early proliferative developing cortex, show fast motions and enhanced radial organization. In contrast, later derived rosettes, which are characterized by reduced NSC capacity and elevated numbers of differentiated neurons, and thus correspond to neurogenesis mode in the developing cortex, exhibit slower motions and decreased radial organization. Third, later derived rosettes are characterized by temporal instability in INM measures, in agreement with progressive loss in rosette integrity at later developmental stages. Finally, molecular perturbations of INM by inhibition of ACTIN or NON-MUSCLE MYOSIN-II (NMII) reduced INM measures. Our framework enables quantification of cytoarchitecture NSC dynamics and may have implications in functional molecular studies, drug screening, and iPS cell-based platforms for disease modeling.