Abstracts

Abstract

Narcolepsy is characterized by excessive daytime sleepiness and cataplexy (loss of muscle tone during wakefulness triggered by emotions), along with hypnagogic hallucinations and sleep paralysis. It is accompanied with a short latency to rapid-eye-movements sleep (REM; also called paradoxical sleep) and sleep onset REM periods (SOREMS). Symptoms highlight a deregulation of REM regulation. Cataplexy is very similar to REM muscle atonia suggesting that part of the network generating REM is activated during cataplexy. Extensive work has been done to identify REM atonia network in cats and rats. However, it has not been identified in mice, a model of choice for the study of cataplexy thanks to the availability of genetically modified animals. CFos is an immediate early gene expressed in a wide range of neurons when they experience a sustained activity, and is commonly used as a neuronal activation marker. It has been used in rats to identify the neuronal populations specifically active during REM. As REM amounts are low in basal condition, a protocol of homeostatic challenge (i.e. a REM deprivation followed by REM hypersomnia) was used to trigger CFos expression in REM active neurons. The combined use of CFos immunostaining, pharmacological microinjections and electrophysiological recordings demonstrated in rats that the neurons generating REM are glutamatergic and localized in the pontine sublaterodorsal tegmental nucleus (SLD). In addition, we showed that the SLD send direct projections to glycinergic neurons of the medullary ventral gigantocellular reticular nucleus previously shown to be responsible for motoneurons' hyperpolarization during REM. Using a similar paradigm of specific REM hypersomnia combined with double immunostaining of Cfos with several markers such as tyrosine hydroxylase, acetylcholine transferase, hypocretin-1 or melanin concentrating hormone, we evaluated the similarities and discrepancies of the network generating REM in mice compared to rats. We found a small region just ventral to the periaqueductal gray, corresponding to the rat SLD, which is immunoreactive for CFos after REM hypersomnia. We also found activated neurons in the ventrolateral part of the periaqueductal gray, in the ventral medulla and the tuberal hypothalamus (including the melanin concentrating hormone neurons). Like in rats, the pontine cholinergic neurons were not Cfos positive. Our observations indicate that the neuronal network generating REM is similar in mice to rats with some variations in the precise localization of the structures.