Author(s): Ben D. Fulcher 1,*, Andrew J. K. Phillips 2, Svetlana Postnova 1,3,4, Peter A. Robinson 1,3,4,5
Since the discovery of the orexin A and orexin B neurotransmitters (also termed hypocretin 1 and 2) by Sakurai et al.  and de Lecea et al.  in 1998, the orexinergic neurons of the lateral hypothalamic area (Orx) have been implicated in a wide range of neurological processes, including a key role in the regulation of sleep and wake , . The orexins have also been shown to have a role in feeding, emotion, reward function, and motivation , -. The neurodegenerative disorder narcolepsy  is characterized by a loss of approximately 90% of Orx , and is a condition that affects approximately 0.05% of the population . It is thought that some process-perhaps an autoimmune attack -selectively destroys these orexinergic cells . Narcolepsy is characterized by awakenings during sleep, unintentional naps during wake, drowsiness, and difficulty in waking from sleep . The condition is often accompanied by cataplexy, the sudden loss of muscle tone triggered by strong emotions , although over one quarter of all narcoleptics do not have cataplexy , perhaps due to less severe loss of Orx , .
Although the link between the loss of Orx and narcolepsy has been established and the key neurological pathways of Orx are known, the mechanisms through which loss of Orx causes narcoleptic symptoms remain unclear . For example, it is commonly thought that Orx excites the wake-promoting monoaminergic neurons (MA) during wake and thereby acts to stabilize the sleep-wake switch , , but it is not clear how Orx also stabilizes sleep, the destabilization of which is a hallmark of narcolepsy . Homeostatic control of sleep in narcoleptics is thought to be normal, since they exhibit normal recovery from sleep deprivation and have a normal total daily sleep duration . The underlying circadian dynamics in both orexin knockout mice and narcoleptic humans also appears to be normal , . Thus, despite apparently normal homeostatic and circadian processes, a reduction in Orx somehow produces 'behavioral state instability', with low thresholds to transition between sleep and wake , . In this work, we present a detailed, physiologically justified explanation of this phenomenon and explain how the loss of Orx gives rise to these characteristically low thresholds for behavioral state transitions in narcolepsy.
Phenomenological models of sleep-wake dynamics that have built upon Borbély's two-process model  have been successful in predicting a range of sleep-wake behaviors , including subjective fatigue during sleep deprivation, internal desynchronization, fragmented sleep during continuous bedrest, and the sleep durations of shift workers , . However, incorporating Orx into such models is problematic because they lack a physiological framework. In contrast, physiologically based models of sleep represent the neuronal populations and their interactions explicitly, allowing new physiological information to be incorporated straightforwardly. Following advances in the understanding of key sleep-regulatory nuclei in the brainstem and hypothalamus , , a range of physiologically based sleep models have been developed -. In this...