The thalamus, once believed to be a simple relay station between the body periphery and the neocortex, has started to be recognized as a key player in higher-order functions, such as attention. It participates in the transition between brain states, such as sleep and wakefulness, and thalamocortical oscillations. Many of these functions involve the reticular nucleus (Rt), an inhibitory brain structure that surrounds the thalamus and where thalamocortical and corticothalamic axons establish synaptic contacts. Recent technical advances have made possible the study of thalamic activity in vivo and its contributions to cortical processing and behavior. However, a unified understanding of how network dynamics in the thalamus and the thalamocortical system is shaped by its neurons and synapses is still lacking.
To address these challenges, we developed biophysically-detailed computational models of thalamic and reticular neurons. We then incorporated them in a large-scale model of thalamo-reticular microcircuitry, whose connectivity was directly constrained by three-dimensional reconstructions of neuronal morphologies. We included intrathalamic connections through chemical synapses, gap junctions and synapses from the sensory periphery and the neocortex. A large part of the data used to build and validate the model was extracted from the literature and leveraged a framework for its systematic and collaborative curation. This framework allowed us to keep track of parameter values used in the model, along with metadata describing the species, brain regions and experimental conditions.
As a first validation of the model at the network level, we studied the generation of spindle-like oscillations. We found that external inputs are necessary to initiate this rhythm, while its termination can be sustained by the synaptic interplay alone, with a key role of mutual inhibition between reticular neurons. We found that that waxing-and-waning oscillations have a clear spatial component, that reflected the connectivity through chemical synapses as well as gap junctions. Finally, we investigated how differential depolarization in reticular and thalamocortical neurons influence the properties of spindle-like oscillations and predicted conditions where they are less easy to evoke.
Taken together this thesis demonstrates how a bottom-up modelling approach can be successfully applied to reconstruct and simulate thalamic microcircuitry, which shows emergent network behavior compatible with experimental findings, and paves the way for understanding the role of the thalamus in thalamocortical functions.
EPFL2020
