Investigating the functional role of long-duration sleep spindles in memory consolidation via non-invasive brain stimulation

Neuronal activity underlying spontaneously generated brain rhythms during sleep are crucial for many forms of memory consolidation. Thalamocortical sleep spindles, i.e., waxing and waning oscillations at around 12 Hz lasting from 0.5 to 3 s present a central endogenous rhythm. In particular, the fine temporal coupling between sleep spindles and both the cortical slow oscillation (SO, ~ 1 Hz) and much faster (~ 150 Hz) hippocampal ripples is suggested to reflect neural reactivation and memory consolidation. However, the contribution of spindles in this multi-regional communication remains poorly understood. Several studies have revealed the importance of spindle and ripple durations for memory processing. Furthermore, spindle duration decreases with age and in multiple neuropsychiatric disorders. Direct evidence for a causal role of spindle duration on memory consolidation is missing, however. The application of non-invasive brain stimulation methods such as electric or auditory stimulation can boost sleep rhythms and improve memory providing a tool to investigate the causal role of sleep rhythms for memory consolidation. Moreover, others and we have suggested trait-like features in humans, especially cognitive ability metrics and baseline sleep parameters, to be relevant markers for the susceptibility of an individual to brain stimulation. Such inter-individual variances may well explain variabilities in stimulation efficacy and poor reproducibility of memory benefits of stimulation in some studies. In this project, we firstly aim to directly target the duration of ongoing sleep spindles to explore the causal role of long-duration spindles on inter-regional temporal interactions and memory performance both experimentally and computationally. We propose a novel Closed-Loop Amplitude-Modulated auditory Stimulation design with amplitudes oscillating at subjects’ individual spindle frequency. Complimentary to experiments we will further develop our hippocampal thalamo-cortical neural mass model that can spontaneously generate SOs, spindles and ripples. We will computationally explore effects of stimulation parameters on sleep rhythms especially the spindle duration as well as spindle-ripple coupling, which can not be investigated by non-invasive methods in healthy humans experimentally. To obtain reliable information on stimulation efficacy at the individual subject level, we secondly aim to characterize interactions between cognitive ability, non-learning baseline EEG and stimulation efficacy on memory consolidation and EEG measures. In the computational model, the impact of SO and spindle properties controlled by relevant model parameters, at the time of stimulation, are exemplary parameters for investigation. Our long-term goal is to develop individualized stimulation methods based on our experimental and computational modelling results for future applications, e.g., in precision medicine and for treatments of patients with memory problems.