Cortical motor network modulation: Common mechanisms parallel efficient motor integration in implicit motor learning in healthy subjects and subthalamic neurostimulation in Parkinson’s disease

Abstract

On the one hand, the neuronal circuitry and connectivity of the large-scale motor network play an important role in many human cognitive functions, i. e. in implicit motor learning. On the other hand, alterations in connectivity of the motor network are also a hallmark in the pathophysiology of a variety of psychological and neurological diseases, such as Parkinson’s disease. Here, we set out to study the motor network activity (more exactly the cortical and spinal aspects of it) under two different aspects: in healthy controls during implicit motor learning and in Parkinson’s disease patients in the conditions ‘stimulation off’ and ‘stimulation on’. To this end, 12 healthy controls and 20 Parkinson’s disease patients performed externally paced right finger movements with simultaneous recordings of a 64-channel EEG and EMG of the forearm muscles. The healthy controls performed the serial reaction time task. Parkinson’s disease patients conducted the baseline of this task with only random trials in the two conditions ‘stimulation off’ and ‘stimulation on ‘. Cortical and muscular activity was analyzed by time-frequency movement-related spectral perturbations and by power spectral density and corticospinal synchronization was assessed by time-frequency cross-spectra coherence. Clinically, Parkinson’s disease patients improved significantly with deep brain stimulation, assessed by the Unified Parkinson’s Disease Rating Scale III score, the reaction time and the error ratio. Deep brain stimulation lead to an increased cortical beta-band movement-related desynchronization, which was topographically spread over a wider cortical area. Besides, in ‘stimulation off’ after finger tap we found a premature beta-band rebound of the corticomuscular coherence to the extensor digitorum over the primary sensorimotor cortex, which was suppressed with stimulation on. The healthy controls presented with significantly reduced reaction times in the ‘sequence blocks’ compared to ‘random blocks’. In ‘sequence blocks’, power spectral density increased mainly over the right posterior parietal cortex but also over a larger left-hemispheric cortical area in alpha and low beta band. Alpha and beta band movement-related desynchronization presented most pronounced over the bilateral prefrontal, fronto-central and central channels. The movement-related desynchronization was significantly modulated over the course of implicit motor learning. The present findings reveal the impressive modulation of the motor network activity including cortical activations and corticospinal synchronizations introduced by deep brain stimulation therapy of the subthalamic nucleus in Parkinson’s disease.