A direct test of competitive versus cooperative episodic–procedural network dynamics in human memory. (1st February 2022)

• Record Type: Journal Article
• Authors: Freedberg, Michael V; Reeves, Jack A; Fioriti, Cynthia M; Murillo, Jorge; Voss, Joel L; Wassermann, Eric M
• Is Part Of: Cerebral cortex. Volume 32:Number 21(2022)
• Page Start: 4715
• Page End: 4732
• Publication Date: 2022-02-01
• Subjects: caudate -- episodic memory -- hippocampus -- procedural memory -- rTMS; Cerebral cortex -- Periodicals; Brain -- Periodicals; 612.825
• DOI: 10.1093/cercor/bhab512
• Languages: English
• ISSNs: 1047-3211

Abstract

Abstract: Classical lesion studies led to a consensus that episodic and procedural memory arises from segregated networks identified with the hippocampus and the caudate nucleus, respectively. Neuroimaging studies, however, show that competitive and cooperative interactions occur between networks during memory tasks. Furthermore, causal experiments to manipulate connectivity between these networks have not been performed in humans. Although nodes common to both networks, such as the precuneus and ventrolateral thalamus, may mediate their interaction, there is no experimental evidence for this. We tested how network-targeted noninvasive brain stimulation affects episodic–procedural network interactions and how these network manipulations affect episodic and procedural memory in healthy young adults. Compared to control (vertex) stimulation, hippocampal network-targeted stimulation increased within-network functional connectivity and hippocampal connectivity with the caudate. It also increased episodic, relative to procedural, memory, and this persisted one week later. The differential effect on episodic versus procedural memory was associated with increased functional connectivity between the caudate, precuneus, and ventrolateral thalamus. These findings provide direct evidence of episodic–procedural network competition, mediated by regions common to both networks. Enhanced hippocampal network connectivity may boost episodic, but decrease procedural, memory by co-opting resources shared between networks.