Sleep spindle state-dependent motor cortical plasticity induction by EEG-triggered ripple burst TMS

Background: Sleep spindles are fundamental for plasticity and memory consolidation. Here we sought to target different spindle states in sleeping healthy participants with real-time EEG-burst repetitive transcranial magnetic stimulation (rTMS) at hippocampal ripple frequency, and test the spindle state-dependent induction of corticospinal and sensorimotor cortical plasticity. We hypothesized that the spindle-trough is a particularly critical state for plasticity induction because hippocampal ripples are naturally nested in the spindle-trough, reflecting replay of memory traces and facilitating memory consolidation. Methods: Fourteen participants underwent four experimental nights, in which rTMS was applied either at the spindle-trough, spindle-peak, spindle random phase or during spindle-free epochs. Readouts of plasticity were changes in resting-state EEG (rsEEG) power, motor evoked potential (MEP) amplitude, local mean field amplitude (LMFA, for the N45 and P60 potential components), immediate response slope (IRS) and TMS-related time frequency response (TFR), tested 10 and 30 min after the end of the rTMS interventions upon awakening, and compared to pre-sleep baseline. Results: Spindle-trough rTMS resulted in pre- to post-sleep decreases of rsEEG beta-band power, MEP amplitude, P60-LMFA, IRS and TFR in the alpha-band, and an increase in the N45-LMFA. None of the other spindle state-dependent rTMS interventions resulted in consistent plastic changes. Conclusions: Targeting the spindle-trough with ripple-burst rTMS stands out in consistently leading to long-term depression-like changes across a broad array of corticospinal and sensorimotor cortical excitability readouts. This opens the intriguing opportunity of targeted manipulation of human sleep physiology for improving specific behavioral processes, such as memory consolidation.