Séminaire Vision algorithmique et biologique
< précédent |
|Modeling neuron–glial interactions: from Hodgkin–Huxley to neurometabolic coupling and back|
Renaud Jolivet (EPFL)
28 novembre 2006
In recent years, accumulating evidences have shown that astrocytes play a critical role at synapses and in providing energy substrates for neurons. Despite the ongoing revolution regarding their role in the nervous system, astrocytes have attracted only very little attention from the computational neuroscience community. In this talk, we will present a summary of the most important features of astrocytes known today and show how to extend the Hodgkin–Huxley framework to include neurometabolic coupling between neurons, astrocytes, extracellular space and cerebral blood flow.
We first extend the Hodgkin–Huxley model to include the Na,K–ATPase electrogenic pump that maintains electrochemical gradients. Our analysis shows that the pump significantly affects the spiking of the neuron and the pattern of spike–frequency adaptation. Then, in order to study how typical neuronal and astrocytic timescales relate, we extend our model to the metabolic neuron–astrocyte interaction using recent work by Aubert and Costalat connecting cerebral blood flow to principal metabolic pathways. The resulting model bridges the gap between astrocytes and the classic Hodgkin–Huxley framework. This allows us to focus on relations between neuronal activity and metabolism. The final model is fitted on experimental data so that it quantitatively reproduces recent results of NADH fluorescence dynamics. Interestingly, the only acceptable parameter set resulting from this optimization procedure yields a model that strongly supports the so–called astrocyte–neuron lactate shuttle hypothesis. Namely, the astrocyte releases lactate that is consumed as an energy substrate by the neuron placing the astrocyte at the center of brain energy metabolism. Our results bring support for an active and central role of the astrocytes and as such, point to them as a valid and key topic for computational neuroscience.
We will conclude by discussing perspectives and possible extensions of this work.