From immaturity to complexity: in vitro generation of embryonic stem cell-derived functional neuronal networks assessed by micro-electrode arrays

Biological neuronal networks can be described as populations of synaptically interconnected neurons capable of generating electrophysiological activity that spreads spatially and temporally. Thus, neuronal networks represent the basic principle for brain activity. The possibility to generate functional neuronal networks in vitro, starting from embryonic stem cells that represent highly immature and pluripotent cells, would open up new ways to manipulate and observe complex neuronal activity.

Although embryonic stem (ES) cells can be guided in vitro towards a mature electrophysiologically active neuronal stage characterized e.g. by the expression of voltage dependent ion channels it was unclear if populations of ES cell-derived neurons are able to generate a functional neuronal network that would depend on the sufficient density and appropriate function of a multiplicity of neurons and their synapses. To clarify this issue we performed extracellular recordings by using the microelectrode array (MEA) technique that allows to observe electrophysiological function of populations of ES-derived neurons over longer time periods in vitro. MEA-recordings revealed first spontaneous spike activities nearly 2 weeks after initiation of differentiation. During the following 2 weeks the spatial and temporal density of spike activities increased and 3 weeks after initiation of differentiation we observed spontaneous bursts of spikes. Four weeks after initiation of differentiation we observed synchronously oscillating bursts of spike activity on spatially separated electrodes. Furthermore, we demonstrated that network activity was sensitive to synaptically acting drugs like gamma-aminobutyric acid or N-methyl-D-aspartic acid and their antagonists bicuculline or APV. We also show that magnesium-free salt solution supported synchronous burst activity and inhibited uncorrelated spike activity. These data indicate that populations of ES-derived neurons are able to generate functional neuronal networks that can be applied to investigate pharmacologically active compounds.