Interplay between cortical inputs and internal dynamics in generating pathological beta oscillations in Basal Ganglia

The complexity of neuronal circuits dysfunctions underlying Parkinson's Disease (PD) has not been captured by computational models yet. In particular, the exogenous or endogenous nature of the main neural hallmark of PD, i.e. beta range ([12 30] Hz) oscillations in the subthalamic nucleus of the basal ganglia, is still debated. Here we investigate the interplay between endogenous and exogenous oscillations in a spiking neural network model of basal ganglia, with a focus on their progressive synchronization in PD. We identify a nonlinear summation of the two oscillation sources and delineate parameter regions where one dominates over the other in shaping subthalamic neuronal activity. Comparison with simplified models of synchronization highlights that nonlinearities are essential to capture the core characteristics of this interaction. On top of that, a coupled forced oscillators model can account for the complexity of the biologically grounded model. Our results on the selective stimulation of basal ganglia neurons by idealized cortical signals suggest a preferential role of hyperdirect (i.e. cortico-subthalamic) projections in propagating pathological beta in the whole network, in line with recent experimental findings. These results shed a new light on our understanding of the genesis of pathological beta oscillations in basal ganglia and pave the way for in silico applications to real biological systems for Parkinson’s disease treatment.