You are here:
Computational Modeling of an In Vitro Neural Network – Microelectrode Assay for Neurotoxicity Screening
Citation:
Conolly, R. AND Tim Shafer. Computational Modeling of an In Vitro Neural Network – Microelectrode Assay for Neurotoxicity Screening. SOT 2019 Annual Meeting, Baltimore, Maryland, March 10 - 14, 2019.
Impact/Purpose:
Computational modeling of structural features of an in vitro assay can provide mechanistic insights into assay data. Here we describe computational modeling of an assay where primary rat cortical cells are grown over a microelectrode array for up to 12 days in vitro (DIV).This assay is used to screen for developmental neurotoxicity, as chemicals that interfere with neuron growth, synapse formation, and firing will alter electrical activity. The capability of the model for examination of how structural features of the assay affect its performance can be used to infer potential mechanisms by which putative neurotoxicants interfere with the growth and maturation of the neuronal cultures.
Description:
Computational modeling of structural features of an in vitro assay can provide mechanistic insights into assay data. Here we describe computational modeling of an assay where primary rat cortical cells are grown over a microelectrode array for up to 12 days in vitro (DIV). As the neural network develops, random firing is detected at the electrodes by 5 DIV but, over time, firing becomes synchronized in bursts. The assay is used to screen for developmental neurotoxicity, as chemicals that interfere with neuron growth, synapse formation, and firing will alter electrical activity. While the assay itself provides little mechanistic information about how chemicals alter firing behavior, the computational model, coded in MATLAB®, provides a capability for quantitative analysis of how firing behavior is affected by assay characteristics including total number of neurons, relative numbers and firing rates of excitatory and inhibitory neurons, fraction of all neurons in contact with electrodes, and numbers of synapses formed by excitatory and inhibitory neurons. For example, when configured with 105 neurons, of which 75% are excitatory and 25% inhibitory, with 0.1% of neurons contacting electrodes, basal firing rates of 0.15 and 1.15 Hz for excitatory and inhibitory neurons respectively, and numbers of synapses/neuron ranging between 0 to 700 and 0 to 281 for excitatory and inhibitory neurons, respectively, the model is at a transition point between random firing and highly synchronized, nonrandom bursting. On one side of the transition, raster plots of model simulations are visually like raster plots for random firing of actual cultures, while on the other side, simulations are even more highly synchronized than is seen with bursting of actual cultures. The current implementation of the model does not explicitly recreate the spatial distribution of the electrodes or of the neurons, and these explicit spatial characteristics may be needed for more accurate simulations of bursting. This capability of the model for examination of how structural features of the assay affect its performance can be used to infer potential mechanisms by which putative neurotoxicants interfere with the growth and maturation of the neuronal cultures. This abstract does not necessarily reflect any specific policy of the US EPA.