Respiratory pauses highlight sleep architecture in mice
Nature Communications. 2026-05-19; :
DOI: 10.1038/s41467-026-73106-z
Casali G(1), Miermon C(1), Terral G(1), Ravassard P(1), Gervois T(1), Dolique T(1), Harrell ER(1), Spitsyn A(1), Lesburguères E(1), Jarriault D(2), Gambino F(1), Chenouard N(3)(4), Roux L(5).
Author information:
(1)Univ. Bordeaux, CNRS, Interdisciplinary Institute for Neuroscience, IINS, UMR
5297, Bordeaux, France.
(2)Univ. of Bordeaux, INRAE, Bordeaux INP, NutriNeuro, UMR 1286, Bordeaux,
France.
(3)Sorbonne Univ., INSERM, CNRS, Paris Brain Institute, Institut du Cerveau,
ICM, U1127, UMR 7225, Paris, France. .
(4)Univ. Bordeaux, INSERM, Neurocentre Magendie, U1215, Bordeaux, France.
.
(5)Univ. Bordeaux, CNRS, Interdisciplinary Institute for Neuroscience, IINS, UMR
5297, Bordeaux, France. .
Brain activity and breathing rate influence each other but it remains unclear
how fine respiratory features vary across and within brain states, and how they
coordinate with the micro-architecture of sleep and its associated network
dynamics. Using simultaneous nasal pressure and hippocampal local field
potential recordings in freely-moving mice, we show here that Wake, Rapid Eye
Movement (REM) sleep, and non-REM sleep (NREM) exhibit unique respiratory
signatures with distinct prominence of post-inhalation pauses. Within NREM, the
emergence of pauses aligns with the infra-slow noradrenaline fluctuations and
demarcate not only NREM packets traditionally defined by microarousal movements,
but also shorter (~30 seconds) packets of elevated hippocampal sigma-band power.
Within packets, respiratory feature changes predict moment-to-moment
fluctuations in the sigma-band peak-power, even in hypoxic conditions where
these infra-slow oscillations are accelerated. Overall, our findings reveal that
respiratory features capture the macro- and micro-architecture of sleep, opening
new windows into brain states and network computations through respiration.
© 2026. The Author(s).
DOI: 10.1038/s41467-026-73106-z
PMID: 42156372
Conflict of interest statement: Competing interests: The authors declare no
competing interests.