0057 Functional Network Architecture Supports Memory Consolidation via NREM Sleep Oscillation Expression and Coupling

Miranda G Chappel-Farley,Jenna Adams,Negin Sattari,Destiny Berisha,Novelle Meza,John Janecek,Richard Betzel,Hamid Niknazar,Soyun Kim,Abhishek Dave,Ivy Y Chen,Kitty K Lui,Ruth Benca,Michael Yassa,Bryce Mander
DOI: https://doi.org/10.1093/sleep/zsae067.0057
IF: 6.313
2024-04-20
SLEEP
Abstract:Abstract Introduction Memory consolidation is facilitated by trace reactivation during non-rapid eye movement slow wave sleep (SWS), involving reinstatement of a hippocampal index driving integration of activity patterns stored in cortical modules. Slow oscillation (SO) and sleep spindle (SP) coupling reflects this process. Here, we used a novel graph theory approach to examine whether network modularity (Q) and the relative import of the hippocampus and amygdala in network function, measured by eigenvector centrality (EC) and betweenness centrality (BC), are associated with SWS expression, SO-SP coupling, and emotional memory consolidation. Methods Thirty-six cognitively intact older adults (μ=72.9±5.6 years, 23 Females) completed structural and resting-state functional magnetic resonance imaging (MRI) at 3T, an overnight polysomnography with electroencephalography, and, prior to and following sleep, performed the Emotional Mnemonic Discrimination Task, with the Lure Discrimination Index quantifying memory. MRI data were preprocessed with the CONN toolbox. Network nodes were defined using the Brainnetome atlas. Graph metrics were computed using the Brain Connectivity Toolbox. In a subset, p-welch derived SO power (N=18) and mean vector length computed phase-amplitude SO-SP coupling (N=16) were derived. Results Q was associated with diminished emotional memory retention (B=-1.394, p=0.042) and greater SWS percentage (B=48.476, p=0.023), which was also associated with emotional memory retention (B=0.013, p=0.020). An inconsistent mediation effect of SWS in the Q-emotional memory retention relationship was found (95% CIs: 0.015, 0.345). Hippocampal EC (B=36.845, p=0.008) and amygdala EC (B=47.277, p=0.022) were associated with emotional memory retention. Hippocampal BC was associated with immediate test performance (B=19.859, p=0.027), whereas amygdala BC was associated with delayed test (B=1.823, SE=0.745, p=0.020) and emotional memory retention (B=1.699, p=0.047). Hippocampal EC (B=0.475, p=0.047) and BC (B=0.426, p=0.086) were associated with frontal SO expression, which was associated with delayed test performance (B=0.492, p=0.038). Hippocampal EC (B=0.649, p=0.006) and amygdala BC (B=0.545, p=0.029) were associated with frontal SO-SP coupling. Conclusion These findings suggest that greater network integration supports SWS-dependent emotional memory consolidation during sleep. A more topologically influential hippocampus aids in memory acquisition and retention, whereas a more influential amygdala supports emotional memory consolidation. Support (if any) This research was supported by NIA grants F31AG074703, F32AG074621, R21AG07955, R01AG053555, K01AG058353, and AASM Foundation SRA-1818.
neurosciences,clinical neurology
What problem does this paper attempt to address?