Hyperglycemia-Induced REM Sleep Suppression: Locus Ceruleus-Noradrenergic Hyperactivity as the Underlying Mechanism

Abstract

Sleep is essential for maintaining metabolic homeostasis. Disruptions in sleep and circadian rhythms adversely affect glucose metabolism and hormonal regulation, thereby promoting the development of metabolic syndrome. In individuals with diabetes mellitus, sleep disturbances are frequently observed and often attributed to secondary complications such as nocturia; however, the common co-occurrence of hyperglycemia and impaired sleep indicates a complex bidirectional relationship. Despite accumulating evidence delineating the interplay between metabolic state and sleep physiology, the mechanistic pathways through which acute hyperglycemia directly influences sleep architecture remain inadequately characterized. Acute hyperglycemia was induced in C57BL/6 mice by 8 h of fasting followed by intraperitoneal diazoxide injection and oral glucose gavage, resulting in approximately 6 h of elevated blood glucose. EEG/EMG recordings were performed during this period, with DMSO- and saline-treated mice serving as controls. For immunohistochemistry, mice （n = 3-4/group） were sacrificed 3 h after treatment, perfused with PBS and 4% paraformaldehyde, cryoprotected, sectioned （40 μm）, and stained for c-Fos and NeuN. c-Fos⁺/NeuN⁺ neurons were quantified in sleep-wake regulatory brain regions. To monitor locus ceruleus （LC） activity in vivo, a TH- Cre mouse received AAV（DJ）-EF1α-DIO-GCaMP6f injection into the right LC and implantation of an optic fiber positioned 0.15 mm above the injection site, followed by simultaneous fiber photometry and EEG/EMG recordings. To assess ex vivo neuronal activity, acute LC-containing （300μm） brain slices were prepared 3 h after treatment and recorded using a high-density microelectrode array （HD-MEA）. Local field potential activity was quantified by peak-to-peak amplitude, signal energy, and root mean square （RMS） measurements. Cognitive performance was evaluated using the Novel Object Preference Recognition （NOPR） test 24 h after the induction of hyperglycemia, with Smart v3.0 software. Acute hyperglycemia markedly suppressed REM sleep, reducing REM percentage, bout length, and duration. EEG absolute power decreased across wake and NREM, indicating dampened cortical activity. Acute hyperglycemia selectively reduced NREM sleep spindle amplitude while leaving spindle density unchanged, suggesting intact initiation but weakened thalamocortical synchrony. LC-TH staining and the LC fiber photometry showed increased noradrenergic LC activity during NREM after diazoxide. The increase in neuronal activity was subsequently confirmed by ex vivo electrophysiological recordings of brain slices performed using a high-density microelectrode array （HD-MEA）. Acute hyperglycemia directly suppresses REM sleep, reduces cortical activation, and weakens thalamocortical synchrony, with LC hyperactivity mediating REM inhibition. This study provides the first experimental evidence that a transient elevation in glucose is sufficient to rapidly disrupt REM sleep through LC activation, revealing a metabolic-neuromodulatory mechanism linking hyperglycemia to impaired sleep regulation.