The two‐process model of sleep regulation: Beginnings and outlook^†

Abstract

Summary The two‐process model serves as a major conceptual framework in sleep science. Although dating back more than four decades, it has not lost its relevance for research today. Retracing its origins, I describe how animal experiments aimed at exploring the oscillators driving the circadian sleep–wake rhythm led to the recognition of gradients of sleep states within the daily sleep period. Advances in signal analysis revealed that the level of slow‐wave activity in non‐rapid eye movement sleep electroencephalogram is high at the beginning of the 12‐light period and then declines. After sleep deprivation, the level of slow‐wave activity is enhanced. By scheduling recovery sleep to the animal's activity period, the conflict between the sleep–wake‐dependent and the circadian influence resulted in a two‐stage recovery pattern. These experiments provided the basis for the first version of the two‐process model. Sleep deprivation experiments in humans showed that the decline of slow‐wave activity during sleep is exponential. The two‐process model posits that a sleep–wake‐dependent homeostatic process (Process S) interacts with a process controlled by the circadian pacemaker (Process C). At present, homeostatic and circadian facets of sleep regulation are being investigated at the synaptic level as well as in the transcriptome and proteome domains. The notion of sleep has been extended from a global phenomenon to local representations, while the master circadian pacemaker has been supplemented by multiple peripheral oscillators. The original interpretation that the emergence of sleep may be viewed as an escape from the rigid control imposed by the circadian pacemaker is still upheld.

Dynamics of sleep, feeding, and metabolic homeostasis inDrosophilaensheathing glia, astrocytes, and neurons

“…Vice versa , the experience during wakefulness, in particular the duration of time spent awake, but also for example learning or injury, impact subsequent sleep across species 4, 15 . However, how neural circuits in the brain integrate the time spent awake as well as specific behaviors for homeostatic sleep regulation, is little understood 3–6 .…”

“…In the brain, a homeostatic process or signal has been proposed to represent sleep need, steadily increasing during wakefulness and gradually decreasing during sleep 1,2 . However, such dynamics of a sleep homeostat, capturing the changing need to sleep in real time depending on behavior, has so far not been observed in identified cells 3–6 . Here, using a system that we developed for monitoring calcium activity over multiple days in head-fixed, walking fruit flies 7,8 , we find that a class of glia in the fly brain, called ensheathing glia 9–11 , shows dynamics expected of a sleep homeostat. Calcium levels in these cells – monitored in a central brain area important for memory, navigation and sleep – integrate activity during wakefulness, reset during sleep, and saturate under sleep deprivation.…”

“…In the brain, a homeostatic process or signal has been proposed to represent sleep need, steadily increasing during wakefulness and gradually decreasing during sleep 1,2 . However, such dynamics of a sleep homeostat, capturing the changing need to sleep in real time depending on behavior, has so far not been observed in identified cells 3–6 .…”

A two-process model ofDrosophilasleep reveals an inter-dependence between circadian clock speed and the rate of sleep pressure decay

“…Though myriad factors impact sleep [10], the regulation of typical sleep/wake cycles can be explained by just these two processes acting in concert but operating independently [6,7]. This model has been a major conceptual framework in the field for nearly four decades owing to its ability to predict the timing and amount of sleep in mammals, both during normal sleep/wake cycles and in the aftermath of sleep deprivation [5][6][7]. The model has also been used over the years to inform sleep clinicians in the treatment of sleep disorders associated with major depression, seasonal affective disorders, irregular sleep patterns, and the application of bright light therapy [7].…”

“…They are particularly useful by virtue of their provision of testable hypotheses [2][3][4][5][6][7][8][9]. Some of the most successful models of sleep regulation have been variants of Borbély's twoprocess model [2,[4][5][6][7], which posits the presence of a sleep pressure that builds during wakefulness (Fig. 1A).…”