Abstracts

Rationale: Absence seizures are thought to result from 3~5 Hz generalized thalamocortical oscillations. Understanding how such oscillations persist could lead to novel treatments for the most common form of pediatric epilepsy. In acute thalamic slices, bicuculline-induced GABAB receptor-dependent oscillations were prolonged upon pharmacological blockade of GABA transporters GAT1 or GAT3 individually, but were surprisingly suppressed upon dual blockade of GAT1 and GAT3. Using pharmacological manipulations, distinct temporal profiles of inhibitory post-synaptic currents (IPSCs) from GABAB receptors of thalamocortical (TC) relay neurons could be recorded for each of 4 pharmacological conditions: (1) control, (2) GAT1 blockade, (3) GAT3 blockade, (4) dual blockade. We propose that the differential effects of GAT blockade on thalamic oscillations could be accounted for by the differential activation of GABAB receptors in TC neurons. Methods: We developed a biophysical thalamic network model that can recapitulate differences in oscillations when GABAB activation profiles were varied. We first established 3-compartment single TC neuron models that could recapitulate differences in GABAB IPSC responses. Using dynamic clamp, voltage responses to each of the 4 GABAB IPSC conductance profiles, at 3 different levels of maximal amplitude, as well as voltage responses to a current impulse, was recorded for 36 TC neurons. Results: In general, TC neurons had an increase in low-threshold spike (LTS) probability with GAT1 or GAT3 blockade but a significant decrease with dual blockade. However, there was also high IPSC response heterogeneity among the 36 neurons. By fitting simulated current impulse responses and simulated IPSC responses for each individual cell, we sought to infer geometric and conductance parameters for all 36 neurons. Finally, we incorporated the heterogeneous model TC neurons into an established thalamic network model. A homogeneous network model was sufficient to recapitulate the exacerbation of oscillations with GAT1 or GAT3 blockade and the abolishment of oscillations with dual blockade. Nevertheless, network heterogeneity seemed to be necessary for modelling seizure termination. Conclusions: In summary, we have developed a biophysically-based computational model that recapitulates most effects of GABA transporter antagonists on thalamic oscillations. These results suggest that modulation of GABA transporters is a potential novel treatment for absence epilepsy. Funding: University of Virginia Medical Scientist Training Program, NIH R01-NS099586 (MPB)