A biochemical signal inside brain cells continuously tracks both how long you have been asleep and whether that sleep was fragmented, according to a new preprint from researchers at Washington University in St. Louis. The signal, a phosphorylation event downstream of the enzyme protein kinase A, declines exponentially during every sleep bout, integrates interruptions, and predicts the moment-to-moment probability of waking up.
The finding, posted July 8 on bioRxiv, identifies the first molecular signal known to encode sleep history within individual sleep bouts, bridging a gap between fast-acting arousal circuits and the slow homeostatic processes that govern sleep need over hours.
The gap in sleep neuroscience
Sleep unfolds across dramatically different timescales. Neural circuits and neuromodulators drive sleep-wake transitions in seconds. Individual sleep bouts last minutes to hours. And the homeostatic sleep drive, the pressure to sleep that builds during wakefulness and dissipates during rest, is tracked over many hours, as reflected in slow-wave activity measured by EEG.
But no signal was previously known to operate at the intermediate timescale: within individual sleep bouts, where the brain must continuously monitor how much sleep has accumulated and how likely waking is at any given moment.
The Washington University team hypothesized that biochemical signals downstream of sleep- and wake-associated neuromodulators might have slower dynamics than the neuromodulators themselves, making them natural candidates for encoding within-bout sleep history.
What the team found
Elizabeth Tilden, Antonio Fontenele, and colleagues used a real-time fluorescent biosensor in freely behaving mice to measure protein kinase A substrate phosphorylation (PKA-SP) at the cell membrane during spontaneous sleep-wake cycles.
Three results stood out:
First, membrane PKA-SP decreased exponentially during each sleep bout, with remarkably consistent kinetics across bouts. The decay followed the same mathematical pattern whether a bout lasted minutes or longer.
Second, the signal integrated both sleep duration and sleep interruption. Because the exponential decline continued from where it left off across fragmented bouts, the PKA-SP level at any point reflected the cumulative history of prior sleep, including how many times sleep had been broken up.
Third, the PKA-SP level continuously forecast the probability of waking. Lower levels predicted a higher likelihood of a transition to wakefulness, suggesting the signal serves as a dynamic readout of when the brain is ready to emerge from sleep.
After six hours of sleep deprivation, end-of-bout PKA-SP levels reached even lower values, consistent with the increased dissipation of sleep need that occurs when an animal finally gets to rest after extended wakefulness.
Implications
The findings position PKA-SP as a molecular integrator that operates at the timescale of individual sleep bouts, linking fast neuromodulator dynamics to the slow accumulation and discharge of sleep pressure. “These findings identify a molecular signal encoding within-bout sleep history, revealing how biochemical dynamics bridge fast arousal circuits and the slow timescale of classical sleep homeostasis,” the authors write.
The work suggests that sleep homeostasis is not purely an electrical or network-level phenomenon but has a biochemical substrate that tracks sleep in real time at the membrane level. Whether analogous signals operate in humans, and whether disruptions in this molecular integrator contribute to sleep disorders characterized by fragmentation, remain open questions.
The study is a preprint and has not yet undergone peer review.
Source
Tilden EI, Fontenele AJ, Goggans KM, Ma S, Gorecki D, Berriman-Rozen ZD, Oldenborg A, Shew WL, Chen Y. “A molecular integrator of sleep duration and interruption.” bioRxiv [Preprint]. 2026 Jul 8:2026.07.03.736427. PMID: 42465383.