Carbachol models of REM sleep: recent developments and new directions.

Since the early '60s, injections of a broad-spectrum muscarinic cholinergic agonist, carbachol, into the medial pontine reticular formation (mPRF) of cats have been extensively used as a tool with which to study the neural mechanisms of rapid eye movement (REM) sleep. During the last decade, new carbachol models of REM sleep were introduced, including chronically instrumented/behaving rats and "reduced" preparations such as decerebrate or anesthetized cats and rats. The combined results from these distinct models show interspecies similarities and differences. The dual nature, both REM sleep-promoting and wakefulness (or arousal)-promoting, of the cholinergic effects exerted within the mPRF is more strongly expressed in rats than in cats. This strengthens the possibility suggested by earlier central neuronal recordings that active wakefulness and REM sleep have extensive common neuronal substrates, and may have evolved from a common behavioral state. Carbachol studies using different intact and reduced models also suggest that powerful REM sleep episode-terminating effects originate in suprapontine structures. In contrast, the timing of REM sleep-like episodes in decerebrate models is determined by a pontomedullary neuronal network responsible for the generation of an ultradian cycle similar to the basic rest-activity cycle of N. Kleitman. Other presumed species differences, such as the more widespread distribution of carbachol-sensitive sites or the relative failure of carbachol to increase the duration of REM sleep episodes in rats when compared to cats, may be of a quantitative or technical nature. While carbachol and many other neurotransmitters and peptides microinjected into the mPRF evoke, enhance or suppress REM sleep, the most sensitive site(s) of their actions have not been fully mapped, and the nature of the cellular and neurochemical interactions taking place at the sites where carbachol triggers the REM sleep-like state remain largely unknown. Similarly, little is known about the pathways between the mPRF and medial medullary reticular formation, but the existing evidence suggests that they are reciprocal and essential for the generation of both natural and carbachol-induced REM sleep. Studies of the mesopontine cholinergic neurons, which are hypothesized to be the main source of endogenous acetylcholine for the mPRF, need to be extended to neurons of the mPRF and cells located functionally downstream from this important site for REM sleep, or both REM sleep and active wakefulness.