Human iPSC-derived neuronal networks. Development and application for compound evaluation

Abstract: Research on human brain development and function in health and disease has been hampered by limited access to primary human tissue and limited translatability of animal studies. This knowledge gap is encouraging the use of human induced pluripotent stem cell (hiPSC)-derived neural in vitro models. The current hope is that person-specific hiPSC-based in vitro models for human brain development and neuronal network function will increase the success in translating research results from bench to bedside. The aim of this thesis was to characterize and validate a person-specific human iPSC-based neural in vitro model to study the development, properties, and pharmacological modulation of human neuronal networks. In the first article we presented a procedure to generate 3D neural aggregates comprising astrocytes, oligodendrocytes and highly functional neurons that generated synchronous neuronal networks in less than three weeks. Further, by culturing hiPSC-derived 3D neural aggregates in human cerebrospinal fluid (hCSF), we demonstrated in article II that this adult brain-like milieu promotes morphological and functional maturation. Although hCSF is superior to currently used cell culture media, it has very limited availability for routine cell culturing purposes. This motivated the search for soluble factors that can mimic the observed maturational effects. In article III, we identified TGF-β1 as a physiologically relevant factor that can suppress proliferation and enhance neuronal and glial differentiation in a human 3D neural in vitro model. In article IV, we utilized this optimized model to provide insights in how therapeutically effective and overdose concentrations of lithium influence human single neuronal and network function. We showed that epileptiform discharges caused by overdose concentrations of lithium were suppressed by the antiepileptic drug Perampanel. The demonstrated functional impact of clinically relevant pharmacological compounds on human neuronal network function represents a proof-of-concept for the enhanced translational value of the human 3D neural aggregate in vitro model. The work presented in this thesis advances the field with a fast functional isogenic in vitro hiPSC-derived neuronal network model with improved physiological relevance and applicability for drug evaluation. Hopefully, our findings will bring the field of neuroscience closer to more translatable modeling and more successful clinical trials in the future.