Simulation of neuronal activity of human embryonic stem cells recorded with multielectrode arrays

K. Lenk,” Invited Talk presented at 3rd International Conference „Aspects of Neuroscience”, University of Warsaw, Poland, Nov. 2013.

Abstract

The neuronal system can be studied by examining brain circuitry. However, a much more direct way is to culture neuronal cells and networks in an in vitro environment. Here, our aim is to produce in vitro and their in silico counterpart model of neuronal networks. Current state of the art in vitro systems utilize 2D cultured neuronal networks grown on multielectrode arrays (MEAs). In in vitro experiments with MEAs cultivated e.g. networks of rodent’s neurons or neuronal cell networks derived from human pluripotent stem cells (hPSCNN) can be recorded. Those recorded data show a rich variability of activity patterns. We developed and implemented a model called INEX (INhibitory/ EXcitatory) to simulate neuronal activity as observed in MEA experiments with frontal cortex tissue of embryonic mice and hPSCNNs. The model is based on Glauber dynamics and a Poisson process. Each neuron has either an inhibitory (negative synaptic strength) or an excitatory (positive synaptic strength) effect to its neighbors. We examined the balance between excitation and inhibition. Single and multiple mechanisms of action when a neuro-active substance was added to the neuronal network can be simulated. Bursts, cascades of action potentials, play a crucial role in neuronal networks on MEAs. This important feature was examined for a developing hPSCNN among others. For hPSCNN network we observed larger variation on the activity development compared to frontal cortex tissue of embryonic mice. This is natural as the hPSCNN cells are less mature and can develop to various neuronal lineages unlike the matured cortical neurons. Thus, with this simple model profound conclusions can be made about the examined biological neural network.