# Electrical Synapses Contribute to Sleep‐Dependent Declarative Memory Retention

**Authors:** Gordon B. Feld, Niels Niethard, Jianfeng Liu, Sandra Gebhardt, Lisa Kleist, Kerstin Brugger, Andreas Fritsche, Jan Born, Hong‐Viet V. Ngo, Manfred Hallschmid

PMC · DOI: 10.1111/ejn.70401 · 2026-01-23

## TL;DR

Blocking electrical synapses in the brain with a drug impaired verbal memory retention during sleep but improved sensorimotor memory.

## Contribution

This study shows that electrical synapses specifically support sleep-dependent verbal memory consolidation in humans.

## Key findings

- Mefloquine impaired retention of word pairs during sleep but not during wakefulness.
- Mefloquine disrupted coupling between sleep spindles and slow oscillations in EEG.
- Mefloquine improved sensorimotor memory retention regardless of sleep.

## Abstract

Sleep supports memory formation by neurophysiological mechanisms that are yet to be fully uncovered. We investigated the contribution of the direct coupling of neurons via electrical synapses (gap junctions). The administration of mefloquine (250 mg p.o. vs. placebo), an antimalarial, which blocks electrical synapses, to healthy young men before nocturnal sleep impaired the retention of word pairs learned before drug administration and disrupted the coupling of sleep spindles to EEG slow oscillations. In control experiments, in which participants received mefloquine before a consolidation interval of nocturnal wakefulness or after rather than before sleep, word‐pair memory retention was not affected by the drug, suggesting that electrical synapses specifically support the sleep‐dependent retention of verbal declarative memory. Irrespective of sleep, mefloquine enhanced the retention of sensorimotor memory assessed with a finger sequence tapping task. In supplemental experiments in rats, mefloquine administered i.p. at escalating doses of 20 and 40 mg/kg did not alter hippocampal sharp‐wave/ripple activity, a prominent mechanism of hippocampal memory replay. While mefloquine effects beyond gap junctions in the present experiments cannot be fully excluded, we conclude that electrical coupling enhances the oscillatory coordination between sleep spindles and slow oscillations and, thereby, supports systems memory consolidation.

Blocking electrical synapses (gap junctions) in healthy humans by mefloquine impaired the sleep‐dependent retention of verbal declarative memory and improved sensorimotor memory consolidation irrespective of sleep. Mefloquine also disrupted the coupling of sleep spindles to EEG slow oscillations in humans but did not affect hippocampal sharp‐wave/ripple activity in rodents. These results indicate a role for electrical synapses in the systems consolidation of memory contents.

## Linked entities

- **Chemicals:** mefloquine (PubChem CID 4046)
- **Species:** Homo sapiens (taxon 9606), Mus musculus (taxon 10090)

## Full-text entities

- **Genes:** GJA8 (gap junction protein alpha 8) [NCBI Gene 2703] {aka CAE, CAE1, CTRCT1, CX50, CZP1, MP70}, GJD2 (gap junction protein delta 2) [NCBI Gene 57369] {aka CX36, GJA9}, ACHE (acetylcholinesterase (Yt blood group)) [NCBI Gene 43] {aka ACEE, ARACHE, N-ACHE, YT}, CA3 (carbonic anhydrase 3) [NCBI Gene 761] {aka CAIII, Car3}, LMNA (lamin A/C) [NCBI Gene 4000] {aka CDCD1, CDDC, CMD1A, CMT2B1, EMD2, FPL}
- **Diseases:** tinnitus (MESH:D014012), schizophrenia (MESH:D012559), Sleepiness (MESH:D000077260), cognitive impairments (MESH:D003072), malaria (MESH:D008288), PVT (MESH:D000405), rapid eye movement (MESH:D020923), sleep (MESH:D012893), impaired motor-coordination learning (MESH:D001259), and declarative memory dysfunction (MESH:D008569), muscle twitches (MESH:D019042), nonrapid eye movement (MESH:D015835), neuropsychiatric effects (MESH:D065606), Alzheimer's disease (MESH:D000544), mental or physical disease (MESH:D008607), slow-wave sleep (MESH:C535500), Mood (MESH:D019964), learning-to-sleep delay (MESH:D007859), vertigo (MESH:D014717), vision and hearing loss (MESH:D054062), Deprivation (MESH:D012892)
- **Chemicals:** fentanyl (MESH:D005283), Mefloquine (MESH:D015767), quinidine (MESH:D011802), calcium (MESH:D002118), N-methyl-D-aspartate (MESH:D016202), midazolam (MESH:D008874), Quinine (MESH:D011803), carbenoxolone (MESH:D002229), Ag-AgCl (-), AMPA (MESH:D018350), caffeine (MESH:D002110), DMSO (MESH:D004121), saline (MESH:D012965), isoflurane (MESH:D007530), alcohol (MESH:D000438)
- **Species:** Solanum lycopersicum (tomato, species) [taxon 4081], Mus musculus (house mouse, species) [taxon 10090], Rattus norvegicus (brown rat, species) [taxon 10116], Rodentia (rodent, order) [taxon 9989], Homo sapiens (human, species) [taxon 9606]

## Figures

4 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12828869/full.md

---
Source: https://tomesphere.com/paper/PMC12828869