Mitochondrial Electron Overflow Drives the Biological Need to Sleep

Sleep pressure—the biological drive that intensifies with prolonged waking—has long lacked a precise molecular identity. A new study published in Nature provides compelling mechanistic evidence that the urge to sleep originates in mitochondria, specifically in a small population of sleep-control neurons in the Drosophila brain.

The research team performed single-cell RNA sequencing on over 13,000 neurons from rested and sleep-deprived flies, using a fluorescent marker to enrich for dorsal fan-shaped body neurons (dFBNs), the primary sleep-inducing cells in the fly brain. Among all cell types examined—including antennal lobe projection neurons and mushroom body Kenyon cells—only dFBNs showed a robust transcriptional response to sleep loss. That response was strikingly specific: nearly all upregulated transcripts encoded proteins involved in mitochondrial respiration, electron transport complexes I–V, ATP synthase, the ATP-ADP carrier sesB, and TCA cycle enzymes. Simultaneously, genes related to synaptic vesicle release and presynaptic plasticity were downregulated.

The mechanistic logic centers on an electron surplus. During waking, dopamine release inhibits dFBNs, suppressing their electrical activity and ATP consumption. Because caloric intake continues, mitochondria keep producing ATP, filling reserves and overfilling the coenzyme Q (CoQ) electron pool. This triggers premature, non-enzymatic electron leakage to oxygen, generating reactive oxygen species (ROS). The authors confirmed that ATP concentrations in dFBNs rise approximately 1.2-fold after sleep deprivation, and increase acutely when the neurons are inhibited by arousing stimuli, while optogenetic activation dissipates ATP. Structural correlates were also documented: sleep deprivation caused mitochondrial fragmentation, increased mitophagy, and more mitochondria-endoplasmic reticulum contacts—changes that reversed after recovery sleep.

Critically, these morphological and functional changes could be causally manipulated. Installing the alternative oxidase (AOX) from Ciona intestinalis in dFBN mitochondria—providing an electron overflow valve—blunted the structural changes and relieved sleep pressure. Inducing mitochondrial fission genetically reduced neuronal excitability and sleep, while forcing hyperfusion increased both. Expressing an uncoupling protein (Ucp4) to discharge the proton gradient reduced sleep, whereas driving ATP synthesis with a light-activated mitochondrial proton pump (mito-dR) exacerbated the electron-ATP mismatch and rapidly precipitated sleep.

Taken together, the findings position mitochondrial electron overflow—a byproduct of aerobic metabolism during periods of low neuronal activity—as the molecular substrate of sleep pressure. This connects sleep homeostasis directly to a fundamental feature of oxidative metabolism, suggesting that sleep, like aging, may be an inescapable biological cost of life on oxygen.