A causal role for the precuneus in network-wide theta and gamma oscillatory activity during complex memory retrieval

Abstract

Complex memory of personal events is thought to depend on coordinated reinstatement of cortical representations by the medial temporal lobes (MTL). MTL-cortical theta and gamma coupling is believed to mediate such coordination, but which cortical structures are critical for retrieval and how they influence oscillatory coupling is unclear. We used magnetoencephalography (MEG) combined with continuous theta burst stimulation (cTBS) to (i) clarify the roles of theta and gamma oscillations in network-wide communication during naturalistic memory retrieval, and (ii) understand the causal relationship between cortical network nodes and oscillatory communication. Retrieval was associated with MTL-posterior neocortical theta phase coupling and theta-gamma phase-amplitude coupling relative to a rest period. Precuneus cTBS altered MTL-neocortical communication by modulating theta and gamma oscillatory coupling. These findings provide a mechanistic account for MTL-cortical communication and demonstrate that the precuneus is a critical cortical node of oscillatory activity, coordinating cross-regional interactions that drive remembering.

Keywords: episodic memory; human; medial temporal lobe; neural oscillations; neuroscience; precuneus; theta burst stimulation.

Figure 1.. Autobiographical memory paradigm.

Participants were cued with familiar words (locations, people, objects), and told to recall a past event in relation to this word. Construction was terminated when participants indicated they had an event in mind via button press.

Figure 2.. Exploratory between-subjects analysis showing that precuneus stimulation leads to (A) reduced vividness ratings and (B) reduced ease of recall compared to vertex stimulation when considering only the first session for each participant.

Asterisks indicate a significant difference between precuneus and vertex stimulation (p<0.05).

Figure 2—figure supplement 1.. Exploratory between-subjects analysis showing non-significant differences between precuneus and vertex stimulation for (A) vividness and (B) ease of recall, when considering only the second session for each participant.

Figure 3.. Theta activity during autobiographical memory retrieval for vertex stimulation sessions.

(A) Theta power increases during AM retrieval relative to rest. (B) Increased theta phase coupling during AM retrieval using a right MTL seed (blue box). Theta power and phase coupling images displayed at p<0.005, cluster corrected. (C) Comodulogram showing theta-gamma phase-amplitude coupling between right MTL theta phase and left precuneus gamma amplitude. Black dotted lines on comodulogram shows areas of significantly different phase-amplitude coupling between memory and rest. Comodulogram displayed at p<0.05, cluster corrected.

Figure 4.. Effects of cTBS on theta activity during memory elaboration relative to rest.

(A) Precuneus stimulation led to decreased theta phase coupling between a right MTL seed (blue box) and the left occipital lobe. Images displayed at p<0.005, cluster corrected. (B) Comodulograms showing theta-gamma phase-amplitude coupling separately for vertex (left) and precuneus (right) stimulation sessions, between right MTL theta phase and left precuneus gamma amplitude. Black dotted lines show areas of significantly different phase-amplitude coupling between memory and rest. (C) Comparison between comodulograms in (B), showing precuneus compared to vertex stimulation during memory retrieval, between right MTL theta phase and left precuneus gamma amplitude. Black dotted lines show areas of significantly different phase-amplitude coupling between precuneus and vertex stimulation. Comodulograms displayed at p<0.05, cluster corrected.

Figure 4—figure supplement 1.. Phase-amplitude coupling using a broader amplitude frequency range reveals that effects are specific to high gamma.

Comodulograms show PAC for (A) vertex and (B) precuneus stimulation sessions, and (C) comparison between precuneus and vertex stimulation sessions. Comodulograms displayed at p<0.05, cluster corrected.