Directed Differentiation of Human Pluripotent Stem Cells into Radial Glia and Astrocytes Bypasses Neurogenesis

Derivation of astrocytes from human pluripotent stem cells (hPSCs) is inefficient and cumbersome, impeding their use in biomedical research. Here, we developed a highly efficient chemically defined astrocyte differentiation strategy that overcomes current limitations. This approach largely bypasses neurogenesis, which otherwise precedes astrogliogenesis during brain development and in vitro experiments. hPSCs were first differentiated into radial glial cells (RGCs) exhibiting in vivo-like radial glia signatures. Activation of NOTCH and JAK/STAT pathways in bona fide RGCs resulted in direct astrogliogenesis confirmed by expression of various glial markers (NFIA, NFIB, SOX9, CD44, S100B, GFAP). Transcriptomic and genome-wide epigenetic analyses confirmed RGC-to-astrocyte differentiation and absence of neurogenesis. The morphological and functional identity of hPSC-derived astrocytes was confirmed by using an array of methods (e.g. electron microscopy, calcium imaging, co-culture with neurons, grafting into mouse brains). Lastly, the scalable protocol was adapted to a robotic platform and used to model Alexander disease. In conclusion, our findings uncover remarkable plasticity in neural lineage progression that can be exploited to manufacture large numbers of human hPSC-derived astrocytes for drug development and regenerative medicine.