|
Models of thalamic oscillationsIn collaboration with Agnessa Babloyantz (University of Brussels, Belgium), as well as with Terrence Sejnowski (Salk Institute, USA), models of single-cell electrophysiological behavior and of synaptic receptors were designed to investigate mechanisms for oscillations in thalamic circuits. We first investigated the oscillatory behavior in the isolated RE nucleus, based on experiments realized on cats in vivo in Mircea Steriade’s laboratory. We investigated if the biophysical properties of RE cells and the kinetics of their synaptic interactions are sufficient elements to generate a 10 Hz network oscillation similar to in vivo observations [1,4]. We found and predicted that a 10 Hz oscillation with waxing-and-waning field potentials can be sustained by networks of RE cells connected with GABAAreceptors [4]. A similar conclusion was also reached by the group of J. Rinzel (New York University, USA). The presence of GABAA receptors in the RE nucleus was subsequently demonstrated experimentally in several laboratories.Second, models were designed to investigate an alternative mechanism for spindle rhythmicity based on interactions between TC and RE cells. This model was based on in vitro experiments realized in ferret thalamic slices in the laboratory of David McCormick (Yale University, USA). Models suggested that the known intrinsic voltage- and calcium-dependent currents present in TC and RE neurons, combined with their interconnectivity patterns with AMPA and GABAAsynapses, can generate sustained rhythmicity in the spindle frequency range and its waxing-and-waning [3,6]. A series of predictions were generated, and among them, the waxing-and-waning behavior was due to upregulation of Ih by Ca2+ [2,3,6], a property that was later demonstrated by experiments performed in David McCormick’s laboratory. The models above predicted both TC-RE spindling activity and spindle rhythmicity in the isolated RE nucleus, but an explanation was still needed to explain why the isolated RE nucleus shows spontaneous oscillations in vivo but not in vitro. With Mircea Steriade and Diego Contreras (University of Pennsylvania, USA), we proposed a possible explanation for these discrepancies based on the effect of neuromodulators on RE cells [5]. We simulated the depolarizing action of serotonin and noradrenaline on RE cells and observed that, if RE cells have a hyperpolarized resting membrane potential, oscillations require to take into account the presence of neuromodulation, at a reduced level [5], which may well be present in vivo but not in vitro. This model predicts that application of weak concentrations of noradrenaline or serotonin should restore oscillatory behavior in slices of the RE nucleus. We next investigated a network model where all the above properties are present together [6]. In addition, this model accounted for various network properties found in thalamic slices, such as the refractoriness of the network and the propagating properties of the oscillations. We also investigated the transformation of spindle to a slower and more synchronized oscillation at around 3 Hz following block of GABAA receptors. The genesis of these epileptic-like discharges is examined in more detail in Section 3.4. More recently, thalamic and thalamocortical networks models were studied based on adaptive exponential integrate-and-fire models [7]. These models were shown to generate various network states, such as spindle oscillations, slow-wave oscillations (Up/Down states) and desynchronized network states (AI states). Alain Destexhe |