The tripartite synapse during metabolic stress

点击翻译按钮获取中文摘要

The Z-project connects state-of-the-art experimental approaches to advanced computational modelling. Based on experimental findings obtained in the RU, we have generated a biophysical model of the tripartite synapse under metabolic stress. The model includes a presynaptic neuron and an astrocyte embedded in a finite extracellular space (ECS). The various transmembrane currents and their energy dependency are expressed as coupled differential equations calibrated with experimental data. The model faithfully reproduces key experimental observations, including ion, volume and electrophysiological dynamics. We found that the response to transient energy deprivation is bi-stable: neurons can either recover from transient ATP depletion and repolarise with the restoration of synaptic transmission or remain in a depolarised state, depending on NKA pump strength and the size of the ECS. In the second Funding Period, we will extend this model by incorporating important new elements, including Na+-dependent acid-base transporters (NHE1 and NBCe1) and dynamic water permeabilities for astrocytes and neurons to explore volume regulation. Endosomal compartments will be added for a more detailed description of intracellular ion distributions and a postsynaptic compartment to simulate synaptic transmission failure during energy deprivation more faithfully. Moreover, we will implement the glutamate-glutamine and include mitochondrial function. To simulate the energy-dependent activity of interacting neurons, we will add inhibitory synapses and subsequently describe the population activity with a neural mass based on our biophysical model. This neural mass model enables us to simulate brain rhythms (EEG) and their changes during metabolic stress. Calibrating to new experimental data and using advanced bifurcation analysis, we will gain insight into the selective vulnerability of inhibitory and excitatory synapses, including the various energy-dependent processes at the tripartite synapse, within a neural network. Our computational model also provides a platform to test interventions in silico, thus generating hypotheses for new experiments. Ultimately, our model and the consortiums' experimental data will advance our understanding of the clinical phenomenology of many neurological diseases where ATP generation is reduced, including stroke, hypoxic-ischemic encephalopathy, seizures and mitochondriopathies.