Rhythmic modulation of subthalamo-pallidal interactions depends on synaptic rewiring through inhibitory plasticity
Rhythmic stimulation offers a paradigm to modulate brain oscillations and, therefore, influence brain function. A growing body of evidence indicates that reciprocal interactions between the neurons of the subthalamic nucleus (STN) and globus pallidus externus (GPe) play a central role in the emergence of abnormal synchronous beta (15-30 Hz) oscillations in Parkinson's disease (PD). The proliferation of inhibitory GPe-to-STN synapses following dopamine loss exacerbates this pathological activity. Rhythmic modulation of the STN and/or GPe, for example, by deep brain stimulation (DBS), can restore physiological patterns of activity and connectivity. Here, we tested whether dual targeting of STN-GPe by rhythmic stimulation can modulate pathologically strong GPe-to-STN synapses through inhibitory spike-timing-dependent plasticity (iSTDP). More specifically, we examined how time-shifted paired stimuli delivered to the STN and GPe can lead to inter-population synaptic rewiring. To that end, we first theoretically analysed the optimal range of stimulation time shift and frequency for effective synaptic rewiring. Then, as a minimal model for generating subthalamo-pallidal oscillations in healthy and PD conditions, we considered a biologically inspired STN-GPe loop comprised of conductance-based spiking neurons. Consistent with the theoretical predictions, rhythmic stimulation with appropriate time shift and frequency modified GPe-to-STN interactions through iSTDP, i.e., by long-lasting rewiring of pathologically strong synaptic connectivity. This ultimately caused desynchronising after-effects within each population such that excessively synchronous beta activity in the PD state was suppressed, resulting in a decoupling of the STN-GPe network and restoration of healthy dynamics in the model. Decoupling effects of the dual STN-GPe stimulation can be realised by time-shifted continuous and intermittent stimuli, as well as monopolar and bipolar simulation waveforms. Our findings demonstrate the critical role of neuroplasticity in shaping long-lasting stimulation effects and may contribute to the optimisation of a variety of multi-site stimulation paradigms aimed at reshaping dysfunctional brain networks by targeting plasticity.