0 1992 Gordon and Breach Science Publishers S.A. Printed in the United States of America
Intern. J . Neuroscience, 1992, Vol. 6 3 , p. 105-114 Reprints available directly from the publisher Photocopying permitted by license only
MELATONIN AND MATURATION OF REM SLEEP REUVEN SANDYK
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
Democrition University of Thrace, Department of Medical Physics and Polytechnic School, Alexandroupolis and Xanthi, Greece (Received November 15, 1991)
The discovery in 1953 of rapid eye movement (REM) sleep and the appreciation that sleep is a heterogenous physiological state stimulated major research into sleep disorders. Electroencephalographic studies have shown that the amount of REM sleep changes with age. While newborns spend almost 50% of their sleep time in REM, the percentage of REM sleep decreases to 30% by the age of 3 months and to 20% by the age of 6 months. In addition, newborns enter REM sleep soon after the initiation of sleep, but by the age of 4 months entry into sleep assumes the adult pattern in which a significant period of non-REM sleep precedes the onset of REM sleep. Since reduction in the amount of REM sleep is associated with cerebral maturation and since the pineal gland has been implicated both in cerebral development and in the organization of REM sleep, the pineal gland may be involved in the maturation of the adult REM sleep pattern. Prior to the age of 3 months melatonin plasma levels are low and the characteristic circadian rhythms of melatonin are absent. Thereafter, melatonin secretion increases and circadian rhythmicity of melatonin becomes apparent. Thus, the abundance of REM sleep during the first 3 months of infancy is associated with deficient pineal melatonin functions, while the decline in the percentage of REM sleep coincides with the emergence of melatonin secretion coincident with the maturation of the pineal gland. I propose, therefore, that a state of low melatonin secretion is permissive for REM sleep and that maturation of the pineal gland retards REM sleep. This hypothesis is supported by the findings that melatonin suppresses REM sleep in cats and that in rats and humans pinealectomy induces a narcolepticlike pattern of REM sleep which strikingly resembles that of the newborn and which is reversed by the administration of melatonin. A further hypothesis is advanced to explain the pathophysiology of narcolepsy in terms of a maturational defect of the pineal gland in infancy. Keywords: REM sleep, Pineat gland. Melatonin. Infancy. Cerebral maruralion, Narcolepsy.
The discovery by Aserinsky and Kleitman (1953) of Rapid Eye Movement (REM) sleep and its association with dreaming stimulated great interest in sleep research. Polygraphic studies have revealed that sleep is a complex physiological process in which over the course of the night, there is an alternating pattern of non-REM (NREM) sleep and REM sleep with cycles lasting about 90 minutes each (Stock, 1982). A normal adult, upon falling asleep, exhibits a typical succession of electroencephalographic (EEG) changes which have been designated as “stages.” These EEG changes are all considered phases of the NREM sleep (Roffwarg et al., 1966). After fragmentation and obliteration of the alpha rhythm, the waves diminish slightly in frequency as their amplitude increases (Stage 1). This pattern is then replaced by highvoltage, notched slow-waves (K-complexes) (Stage 2 ) . Tall “delta” waves of 1-2 cycles per second progressively fill the record (Stage 3) and finally dominate it in virtually unbroken sequence (Stage 4). ~
Reprints requests to: Professor Reuven Sandyk M.D. ,M.Sc., Department of Medical Physics, Democrition University of Thrace, Alexandroupolis, Greece. 105
106
R. SANDYK
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
Adult Pattern of REM Sleep
Approximately one hour after the onset of sleep, the initial REM period of the night commences. REM periods recur every 80 to 90 minutes and comprise 20-2570 of the conventional sleep time of the adult. REM periods are shorter in the initial stages of sleep and become longer towards the morning with each REM period averaging about 20 minutes duration (Roffwarg et al., 1966). REM sleep is a period characterized by considerable cerebral excitation. This notion is supported by REM sleep EEG and physiological studies which reveal a remarkable similarity between REM sleep and a state of wakefulness (Roffwarg et al., 1966; Llinas and Pare, 1991). Animal studies indicate that REM sleep is associated with active cerebral processes, which include increase in cerebral blood flow to the cortex, rise in brain temperature, and increase in the frequency of spontaneous neuronal firing in the mesencephalic reticular formation, medial and descending vestibular nuclei, pyramidal tract, and occipital cortex (Roffwarg et al., 1966). During REM sleep excitability in the motor cortex is higher than during NREM sleep while in the sensory cortex it is as high as during NREM sleep. In both regions, however, excitability is greater than it is in the waking state. Moreover, REM sleep is associated also with increased responsiveness of thalamic and cortical neurons as compared to the waking state (Roffwarg et al., 1966). Early studies by Jouvet et al. (1962; 1965) have suggested that the generation of REM sleep is dependent upon an intact rostra1 pons (nucleus pontis caudalis). Following ablation of this area in cats, REM sleep disappears and the animal shows only two states, NREM and wakefulness which may gradually progress to insomnia and death. A decorticated cat, on the other hand, shows n o evidence of NREM sleep. Subsequent studies in decorticate and decerebrate humans have supported the analogues dependence of REM sleep upon brainstem mechanisms and of NREM sleep upon cortical mechanisms (cf. Roffwarg et a\., 1966). Based on the assumption that REM sleep is a brainstem phenomenon and NREM sleep a cortical dependent process, Jouvet ( 1 962; 1965) considered REM sleep to be a phylogenetically archaic state. The findings that in the newborn REM sleep matures earlier then NREM or waking and the observations that sustained periods of REM sleep appear directly after arousal without intermediary NREM sleep (Roffwarg et al., 1966) have been used in support of the “primitive” quality of the REM state. Conversely, NREM sleep, being dependent on the functioning of the neocortex, has been thought to be acquired in the process of telencephalization and was considered a “neo-sleep” (Jouvet, 1962; 1965). However, more recent studies suggest that the mechanisms of REM sleep generation are more complex, possibly involving the interaction between several CNS structures (Steriade and Hobson, 1976; Stock, 1982; Drucker-Colin and Valverde, 1982). The Evolutiori of REM Sleep
It has been proposed that REM and NREM sleep evolved at different times during the phylogeny of mammals. Allison and van Twyver (1970) postulated that NREM sleep evolved 180 million years ago in animals which are considered true mammals. REM sleep, on the other hand, was considered to evolve 50 million years later. Moruzzi (1972) viewed the adult human sleep-wake cycle as the result of inborn factors, namely, the development and maturation of the cerebral cortex from infancy (Kleitman, 1963) and its relationship with the brainstem as well as learned factors under the influence of day and night and social influences. Meddis (1977) hypoth-
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
107
esized that REM sleep is evolutionally the more ancient type of sleep and that it is less effective, especially since it is not appropriate for homeothennic species such as birds and mammals. This hypothesis is supported by the inappropriateness of REM sleep as a regulator of body temperature in mammals (Parmeggiani, 1977). Based on the findings that REM sleep is prominent in immature mammals and in the human newborn, it has been hypothesized to accompany and probably be causally related to cerebral maturation (Roffwarg et al., 1966). Extensive investigations on the evolution of REM sleep in humans have demonstrated an association between the percentage of REM/NREM sleep and cerebral maturation (Roffwarg et al., 1966; Williams et al., 1974). Observations in the newborn and immature animals revealed that REM sleep consumed a high proportion of total sleep time in the first days of life and that its amount diminished as maturation proceeded (Roffwarg et al., 1966). Specifically, in infancy when the proportion of time awake is smaller than in any other period of life, there is a large amount of REM activity. Characteristically, these REM periods appear soon after the onset of sleep and are of random duration and occurrence during the night. Moreover, during infancy there is neither inhibition of motor activity (atonia) that is characteristic in REM sleep, nor are phasic events restricted to REM periods (Ellman and Weinstein, 1991). Subsequently, when the developing infant spends longer intervals awake, the total amount as well as the proportion of REM sleep decreases (Roffwarg et al., 1966). In addition, the latency of REM onset, which is significantly longer in children (3-4 hours in the 4-7 year old as compared to one hour in the adult), continues to shorten with maturation reaching the adult pattern in midadolescence (Roffwarg et al., 1966). In the human fetus, REM sleep has been recorded as early as 6 months postconception, prior to any detection of NREM sleep. Later in fetal development, REM sleep occupies almost 90% of the overall sleep time. In premature infants of 32 to 36 weeks’ gestation, REM sleep constitutes 75% of total sleep. This drops sharply to about 50% of the overall sleep time at birth, and to about 30% at the age of 3 months and to 20% at the age of 6 months (Williams et al., 1974; Ellman and Weinstein, 1991). REM Sleep and Cerebral Maturation The high percentage of REM sleep time in the fetus and newborn as well as the progressive decline in the percentage of REM sleep time throughout childhood strongly suggest that REM sleep may be fulfilling a vital function in early brain development. Ephron and Carrington (1966) considered REM sleep a mechanism aimed at restoring cortical homeostasis, while Roffwarg et al. (1966) proposed that REM sleep affords intense stimulation to the central nervous system. It has been suggested that ascending impulses originating in the pons during REM sleep act to enhance neuronal differentiation, maturation, and myelinization in cortical areas (Roffwarg et al., 1966). This hypothesis receives support froni studies which have shown that administration of clonidine, which induces REM deprivation, attenuated cortical development induced by environmental stimulation in young rats (Mirmiran et al., 1983 a; b; Lewin and Singer, 1991). In accord with this hypothesis, several investigators have provided evidence to indicate a role for REM sleep in learning and consolidation of memory (Ellman and Weinstein, 1991). The main impetus for this hypothesis comes from animal data revealing high levels of REM sleep after learning various tasks and the observation that REM sleep deprivation interferes with learning processes (Ellman et al., 1991). In humans, increase in REM sleep time has been noted to occur
I08
R. SANDYK
several days after intense learning, mental stress, and particularly demanding events (Greenberg, 1981). Moreover, memory impairment due to subcortical dementia has been linked to diminished REM sleep (Parkes, 1985). According to Llinas and Pare (1991) “REM sleep can be considered as a modified attentive state in which attention is turned away from the sensory input, toward memories” (p. 525).
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
Regulation of REM Sleep
The findings that REM sleep occupies a great proportion of the infant’s sleep time stimulated research into the mechanisms responsible for the profusion of REM sleep in the immature organism or, for that matter, for the regulation of the proportion of REM sleep at any age. Ellman and Weinstein (1991) proposed that REM restriction, which is initiated at 3 months of age, is an important factor in infant development. One of the more attractive theories discussed by Roffwarg et al. (1966) was that the amount of REM sleep is a consequence of deficient restraining influences on the brainhtem REM center. This hypothesis maintains that abundance of REM sleep in the newborn reflects the presence of insufficient cortical inhibitory influences on the REM center. If this hypothesis is correct, then the amount of REM sleep should increase after a mature organism had been subjected to decortication. However, observations in decorticated cats failed to support this hypothesis in part because these animals were found to have less REM sleep than intact animals (cf. Roffwarg et al., 1966). In experimental animals, pinealectomy is associated with delayed cerebral maturation (Relkin et al., 1973; Kamback et al., 1982). Moreover, pinealectomy both in rats (Kawakami et al., 1972; Mouret et al., 1974) and man (Mouret et al., 1974) induces a narcoleptic-like pattern of REM sleep which is reversed by administration of melatonin (Kawakami et al., 1972). Therefore, since the pineal gland is involved in the organization of circadian sleep/wake cycle including REM sleep (Anton-Tay et al., 1971; Mouret et al., 1974; Cramer et al., 1976) and since it is implicated in psychosexual and pubertal development in humans (Silman et al., 1979; Waldhauser et al., 1984; Lissoni et al., 1985), the pineal gland may be involved also in the maturation of the adult REM/NREM sleep cycle. In the present communication, I propose that the abundance of REM sleep in infancy is maintained by a state of deficient melatonin functions. Furthermore, maturation of the pineal gland associated with the development of circadian rhythms of melatonin secretion exerts a restraining influence on REM sleep. Ontogeny of Pineal Chronobiology
Since the discovery of melatonin as the principal hormone of the pineal gland, it was found that the secretion of the hormone varies in relation to age and level of sexual maturation (Silman et al., 1979; Waldhauser et al., 1984). Lemaitre et al. (1981) assayed melatonin in 58 subjects aged one day to 30 years and noted that urinary elimination progressively decreased from birth to adulthood. In a subsequent study, plasma and urinary melatonin estimations were undertaken between 0500 hrs and 0600 hrs in 26 normal newborn males ranging in age from 15 days to one year (Hartmann et al., 1982). Plasma melatonin levels were found to be low during the first 3 months of life and subsequently increased progressively and significantly between 3 and 12 months. Similar findings have been obtained by Attanasio et al. (1986). These observations reveal that during the first three months of life melatonin
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
109
levels are low as a result of either a relatively inactive synthesis or a lack of utilization of melatonin due to the presence of immature receptors. Studies on the ontogeny of pineal chronobiology in humans reveal that melatonin plasma levels fail to show any circadian rhythmicity during the first month of life (Gupta, 1986). However, the third month of life is associated with the emergence of a highly significant circadian rhythm for melatonin (Gupta, 1986). It is noteworthy that in species such as the rat in which the organization of the adult type of REM sleep does not occur at birth (Verley and Garma, 1975), melatonin also is not present at birth in the pineal gland (Klein and Lines, 1969) and its appearance (Klein and Lines, 1969) coincides with the organization of the adult type of REM sleep (Verley and Garma, 1975). However, in species such as the sheep, in which the organization of the adult type of REM sleep does occur at birth (Verley and Garma, 1975), melatonin is present in the fetal pineal gland (Kennaway and Seamark, 1978). In humans, the period during which the circadian rhythmicity of melatonin emerges in infants is also a crucial phase of neurologic development. Around the age of 3 months the neurological patterns characteristic of the newborn period disappear, replaced by new motor and behavioral patterns (Gupta, 1986). It has been suggested that the development of the overt circadian rhythmicity for melatonin may be involved in cerebral maturation (Gupta, 1986).
The Pineal Gland and REM Sleep The period associated with the maturation of the circadian rhythms of melatonin and the phase during which plasma melatonin levels begin to rise (i.e., 3 months) appear also to coincide with important milestones in the ontogeny of REM sleep. By age 3 months, the percentage of REM sleeptime has dropped from 50% in the newborn to 30% of the total sleep time. In addition, this period is associated with major changes in the organization of REM sleep, namely the elimination of undifferentiated REM states (waking, fussing, and drowsy REM) and the coalescing of random eye movements into burst patterns (Ellman and Weinstein, 1991). Furthermore, by 4 months of age the onset of sleep has been shown to shift from the newborn pattern of entry through REM sleep to the adult mode of entry through NREM sleep (Parkes, 1985). Collectively, these findings demonstrate that abundance of REM sleep and initiation of sleep through REM prior to the age of 3 months occur during a period which is associated with low melatonin secretion coupled with deficient circadian melatonin rhythms. A significant fall in the percentage of REM sleep and the emergence of an adult pattern of entry into sleep through NREM sleep occurs after age 3 months, a period in which concurrently there is a rise in melatonin secretion and maturation of the melatonin circadian rhythms. Melatonin receptors have been demonstrated in various nuclei of the brainstem reticular system (Sallanon et al., 1982), which are implicated in the regulation of the sleep-wake cycle (Parkes, 1985). Melatonin has been shown to modify the spontaneous electrical activity of single cells in the mesencephalic reticular formation of the rat (Pazo, 1979) and pinealectomy has been reported to disrupt the circadian REM sleep in rats and humans (Mouret et al., 1974). Moreover, the pineal gland has been shown to be involved in the regulation of attentional, behavioral, and sleepwake mechanisms (Marczynski et al., 1964; Anton-Tay, 1974; Romijn, 1978; Datta and King, 1980; Birkeland, 1982; Dugovic et al., 1989). Thus, it is unlikely that the association between melatonin secretion and the maturation of REM sleep in infancy reflects an epiphenomenon.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
110
R. SANDYK
I propose, therefore that melatonin deficiency is permissive for REM sleep and that maturation of the pineal gland may exert a restraining effect on REM sleep while simultaneously enhancing the development of REM/NREM adult sleep pattern. This hypothesis is supported by the findings that melatonin suppresses REM sleep in cats (cf. Pave1 et al., 1980) and that pinealectomy, which removes the main source of circulating melatonin (Wurtman et al., 1968), induces both in rats and man (Kawakami et al., 1972; Mouret et al., 1974) a narcoleptic-like pattern of REM sleep which strikingly resembles that of the newborn (Passouant et al., 1969). The hypothesis that decreased melatonin plasma levels is permissive for REM sleep is supported also by the findings in adults that REM periods are associated with melatonin nadirs (Birkeland, 1982). Since melatonin plasma levels rise at the onset of NREM sleep (Birkeland, 1982), it is possible that activation of the pineal gland during this period of sleep may function to inhibit REM sleep. Consequently, the transitions between REM and NREM sleep cycles may be influenced by fluctuations in the secretory activity of the pineal gland, which may thus function to maintain the homeostasis of the sleep cycle. Maturation of REM Sleep and Temperature Regulation
The mechanisms by which the pineal gland may restrain REM sleep in infancy are unknown. One possible mechanism may involve the interactions among the pineal gland, thermoregulation, and REM sleep. The onset of sleep is usually preceded by a fall in body temperature (Geschickter et al., 1966). During the initial stages of sleep body temperature continues to fall. This is associated with increased heat loss, a relative high percentage of stage 3-4 sleep, and a low percentage of REM sleep. During the second half of sleep, body temperature rises. This is associated with decreased heat loss, a low percentage of stage 3-4 sleep, and a high percentage of REM sleep (Avery, 1987). REM sleep is most likely to occur at the nadir of the body temperature cycle (Czeisler et al., 1980). These findings thus indicate a strong coupling of the circadian rhythm of the REM sleep events and changes in body temperature. The amount of REM sleep and the ratio of REM to NREM sleep has been shown also to be influenced by the ambient temperature. Both low and extremely high ambient temperatures have been shown to decrease the amount of REM sleep in animals (Parmeggiani and Rabini, 1970) and man (Beck et al., 1976). Increasing the body temperature by inducing fever reduces REM sleep in humans (Karacan et al., 1968). There is evidence that the hypothalamus mediates the effects of the ambient temperature on REM sleep. It has been shown that the occurrence of REM sleep is related to a narrow range of hypothalamic temperature, a temperature “gate” (Parmeggiani et al., 1975). For instance, once an animal is in stage REM sleep, warming the hypothalamus increases the duration of the REM period (Parmeggiani et al., 1974). The occurrence of REM sleep may thus be increased by experimental manipulations of the hypothalamus and ambient temperatures (Avery, 1987). There is now abundant evidence to implicate the pineal gland in circadian thermoregulation (Ralph, 1984). In fact, it has been suggested that the pineal gland is an integral, if not a key feature in the translation of photoperiodic stimuli into thermoregulator~responses (Ralph, 1984). In a comprehensive review on thermoregulation, Ralph et al. (1979) proposed that the pineal gland acts to prevent excessive or debilitating elevations of body temperature. Such a role becomes obvious given the pivotal position of the pineal gland in the adaptation of the organism to the
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
111
environment. It is presently thought that thermoregulatory functions of the pineal gland involve the mediation of the hypothalamus (Ralph, 1987). In laboratory animals administration of melatonin has been shown to affect body temperature and thermoregulation. In rats, injection of 60 micrograms of melatonin into the cerebral ventricles produces a rise in body temperature for 2 hours (Fioretti and Martini, 1968). Similarly, intraperitoneal injections of melatonin (4 or 8 microgramslg body weight) induced marked hyperthermia in rats (Fioretti et al., 1974), while pinealectomy induced hypothermia (Fioretti et al., 1974; Spencer et al., 1976). Moreover, hypothermia induced by barbiturates, neuroleptics, or morphine has been shown to be reversed by melatonin, suggesting a central action of melatonin on temperature regulation (Ralph, 1984). As indicated above, REM sleep, which is associated with a melatonin nadir (Birkeland, 1982) is also accompanied by temperature nadirs (Czeisler et al., 1980). On the other hand, increased body temperature reduces REM sleep (Karacan et al., 1968). Since administration of melatonin is associated with elevation of body temperature, which in turn decreases REM sleep, melatonin may thus impact on the maturation of REM sleep. Specifically, since the infant in the first 3 months of life is functionally “pinealectomized” and thus has a tendency to become hypothemic, the presence of low body temperature would enhance the profusion of REM sleep. However, with the rise in melatonin secretion and the establishment of the melatonin circadian rhythms after the 3rd month of life, body temperature would tend to rise. This in turn could lead to inhibition of REM sleep. In summary, the pineal gland, through its action on the hypothalamic thermoregulatory systems, may induce or inhibit REM sleep by altering body temperature. Maturation of the Pineal Gland and Narcolepsy
The hypothesis that the pineal gland is involved in the maturation of REM sleep may be relevant to the pathophysiology of several sleep disorders including Kleine-Levin syndrome and narcolepsy. Narcolepsy is a syndrome of excessive sleepiness associated with disruption of circadian rhythms of the sleep-wake cycle (Rechtschaffen and Dement, 1969; Hishikawa et al., 1976). Polygraphic investigations of the sleep in narcoleptic patients show the so-called Sleep Onset REM Periods (SOREMPs) to be a fundamental characteristic of the disorder (Rechtschaffen et al., 1963), occurring at the beginning of the nocturnal sleep period as well as during daytime naps (Dement et al., 1966). SOREMPs occur also in patients with sleep apnea, sleep-wake schedule disturbances, major depression, and in normal infants (Spielman and Herrera, 1991). The occurrence of SOREMPs in infancy and in narcolepsy suggest that in both SOREMPs may be related to a common mechanism related to a failure of the pineal gland to restrain REM sleep in early life. Consequently, narcolepsy may be viewed as a maturational defect of REM organization which is likely to result from dysfunction of the pineal gland in infancy. It is therefore not surprising that in the majority of patients the disease manifests itself in late adolescence (Sours, 1963), a period which is associated with dramatic changes in pineal functions (Silman et al., 1979; Waldhauser et al., 1984). The findings that the circadian rhythms of melatonin are absent in narcolepsy (Birau et al., 1982) and that melatonin fails to induce REM sleep at sleep onset or to increase the amount of REM sleep in narcoleptic patients (Pave1 et al., 1980) are supportive of the hypothesis that the disease may be causally related to abnormal pineal melatonin functions. Finally, since it is now recognized that the pineal gland plays a
112
R . SANDYK
fundamental role in immunomodulation (Maestroni et al., 1987 a; b) , it is possible that the abnormalities of the immune system found in the disease (Langdon et al., 1986) are linked to disruption of pineal melatonin functions.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REFERENCES Allison, T. & van Twyver, H. (1970). The evolution of sleep. (pp. 56-65). Vol. 79, New York: Natural History. Anton-Tay, F. (1974). Melatonin: effects on brain function. Advances in Biochemical Psychopharmacology, 11, 315-324. Anton-Tay, F., Diaz, J . L. & Femandez-Guardiola, A. (1971). On the effect of melatonin upon human brain. Its possible therapeutic implications. Life Sciences, 10, 841-850. Aserinsky, E. & Kleitman, N. (1953). Regularly occumng periods of eye motility and concomitant phenomena during sleep. Science, 118, 273-274. Attanasio, A,, Rager, K. & Gupta, D. (1986). Ontogeny of circadian rhythmicity for melatonin, serotonin and N-acetylserotonin in humans. Journal of Pineal Research, 3. 25 1-258. Avery. D. H. (1987). REM sleep and temperature regulatiun in affective disorder. In A. Halaris (Ed.), Chronobiology and psychiufric disorders (pp. 75-10]), New York: Elsevier. Beck, U., Reinhard, H., Kendel, K. & Schmidt-Kessen, W. (1976). Temperature and endocrine activity during sleep in man. Archives der Psychiatrie und Nervenkrankheiten, 232, 245-256. Birau, N., Mayer, C., Matsubayashi, K. & Meier-Ewert, K. H. (1982). Melatonin serum concentration during the nocturnal sleep of narcoleptics. IRCS Journal of Medical Science, 1 0 , 814. Birkeland, A . J . (19x2). Plasma melatonin levels and nocturnal transitions between sleep and wakefulness. Neuroendocrinology, 34, 126-131. Cramer, H., Bohme, W., Kendel, K. & Donnadieu, M. (1976). Freisetzung von Wachtumshormon und von Melanozyten stimulierenden Hormone in durch Melatonin gebahnten Schlaf beim Menschen. Arzneimittel-ForschunR, 2 6 , 1076- 1078. Czeisler, C. A., Weitzman, E. D., Moore-Ede, M. C., Zimmerman, J. C. & Kronauer, R. E. (19x0). Human sleep: its duration and organization depend on its circadian phase. Science. 210, 12641267. Datta, P. C. & King, M. G . (1980). Melatonin: effects on brain and behavior. Neuroscience & Biobehavioral Reviews. 4 , 451-458. Dement, W. C . , Rechtschaffen, A. & Gulevich, G. (1966). The nature of narcoleptic sleep attack. Neurology. 16, 18-33. Drucker-Colin, R. & Valverde, C. (1982). Endocrine and peptide functions in the sleep-waking cycle. In G. Ganten & D. Pfaff (Eds.), Sleep. clinical and experimental aspects (pp. 37-81). Berlin: Springer Verlag. Dugovic, C., Leysen, J. E. & Wauquier, A. (1989). Melatonin modulates the sensitivity of 5-hydroxytryptamine-2 receptor-mediated sleep-wakefulness regulation in the rat. Neuroscience Letters. 104, 320-325. Ellman, S. J. & Weinstein, L. N. (1991). REM sleep and dream formation: A theoretical integration. In S. J. Ellman & J. S. Antrobus (Eds.), The mind in sleep. psychology andpsychophysiology (pp. 466-488). New York: Wiley. Ellman, S. J., Luck, D., Steiner, S. S. & Halperin, R. (1991). REM deprivation: a review. In S. J . Ellman & J. S. Antrohns (Eds.), The mind in sleep. psychology and ps,ychophysiology (pp. 329376). New York: Wiley. Ephron, H. S. & Carrington, P. (1966). Rapid eye movement sleep and cortical homeostasis. Psychological Review, 75, 500-526. Fioretti, M. C. & Martini, L. (1968). Effetti della malatonina sulla temperatura corporea. Atti Accudemia Anatomica Chirurgia, 6 0 , 3-8. Fioretti, M. C . Riccardi, C., Menconi, E. & Martini, L. (1974). Control of the circadian rhythm of the body temperature in the rat. Life Sciences, 14, 211-219. Geschickter, E. H., Andrews, P. A. & Bullard, R. W. (1966). Nocturnal body temperature regulation in man: A rationale for sweating in sleep. Journul of Applied Physiology, 2 1 , 623-630. Greenherg, R. (1981). Dreams and REM sleep-an integrative approach. In W. Fishbein (Ed.), Sleep, dreams and memory (pp. 125-135). Lancaster: MTP Press. Gupta, D. (1986). The pineal gland in relation to growth and development in children. Journal of Neural Transmission, 2I (snppl), 2 17-232. Hartmann, L., Roger, M., Lemaitre, B. J . , Massias, J . F. & Chaussain, J. L. (1982). Plasma and urinary melatonin in male infants during the first 12 months of life. Clinica Chimicu Acta, 121, 37-42.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
113
Hishikawa, Y., Wakamatsu, H., Furuya, R., Sugita, Y., Masaoka, S., Kaneda, H., Sato, M., Nanno, H. & Zaneko, Z. (1976). Sleep satiation in narcoleptic patients. Electroencephalography and Clinical Neurophysiology, 4 1 , 1-18. Jouvet, M. (1962). Recherches sur les structures nerveuses et les mechanismes responsables des differentes phases du sommeil physiologique. Archives Italiennes de Biologie, 100, 125-206. Jouvet, M. (1965). Paradoxical sleep. A study of its nature and mechanisms. Progress in Brain Research, 18, 20-57. Kamback, D. O., Rich, R. A. & Relkin, R . (1982). Effect of pinealectomy on fatty acid composition of rat brain myelin. Endocrinology, 110, 907-909. Karacan, I., Wolff, S. M., Webb, W. B. & Williams, R. L. (1968). The effects of fever on sleep patterns. Psychophysiology, 5 , 255. Kawakami, M., Yamaoka, S. & Yamaguchi, T. (1972). Influence of light and hormones upon circadian rhythm of EEG slow-wave and paradoxial sleep. In S . Itoh, K. Ogata & H. Yoshimura (Eds.), Advanes in climatic physiology (pp. 349-366). Tokyo: Shoin. Kennaway, D. J. & Seamark, R. F. (1978). The occurrence and synthesis of melatonin in the fetal sheep pineal gland. Journal of Neural Transmission, 13, 371-372. Klein, D. C. & Lines, S . V. (1969). Pineal hydroxyindole-0-methyltransferaseactivity in the growing rat. Endocrinology, 8 4 , 1523-1525. Kleitman, N. (1963). Sleep and wakefulness (p. 552). Chicago: University of Chicago Press. Langdon, N., Lock, C., Welsh, K., Vergani, D., Dorow, R. Wachtel, H., Palenschaft, D. & Parkes, D. (1986). Immune factors in narcolepsy. Sleep, 9, 143-148. Lemaitre, B. J., Bouillie, J . & Hartmann, L. (1981). Variations of urinary melatonin excretion in humans during the first 30 days of life. Clinica Chimica Acfa. 110, 77-82. Lewin, 1. & Singer, J. L. (1991). Psychological effects of REM ("dream") deprivation upon waking mentation. In S. J . Ellman & J. S. Antrobus (Eds.), The mind in sleep. psychology and psychophysiology (pp. 398-412). New York: John Wiley. Lissoni, P., Mauri, R., Resentini, M., Morabito, F., Djemal, S. & Franshini, F. (1985). Melatonin rhythms in some idiopathic hypothalamic-hypophyseal diseases. In G. M. Brown & S. D. Wainwright (Eds.), The pineal gland; endocrine aspects (pp. 307-3 12). Oxford: Pergamon Press. Llinas, R. R. & Pare, D. (1991). Commentary of dreaming and wakefulness. Neuroscience, 4 4 , 521535. Maestroni, G. J. M., Conti, A. & Pierpaoli, W. (1987~).The pineal gland and the circadian opiatergic, immunoregulatory role of melatonin. Aanals of the New York Academy of Science, 496, 67-77. Maestroni, G. J. M., Conti, A. & Pierpaoli, W. (1987b). Melatonin counteracts the effect of acute stress on the immune system via an opiatergic system. In G. P. Trentini & C. de Gaetani (Eds.), Fundamentals and clinics in pineal research (pp. 277-279). New York: Raven Press. Marczynsky, T. J., Yamaguchi, N., Ling, G. M. & Grodziska, L. (1964). Sleep induced by the administration of melatonin (5-methoxy-N-acetyltryptamine)to the hypothalamus in unrestrained cats. Experientia, 2 0 , 435-436. Meddis, R. (1977). The sleep instinct. London: Routlege & Kegan Paul. Mirmiran, M., Scholtens, J., van de Poll, N . E.,Uylings, H. B., van der Gugten, J. & Boer, G. J. (1983~).Effects of experimental suppression of active (REM) sleep during early development upon adult brain and behavior in the rat. Brain Research, 283, 277-286. Mirmiran, M., Uylings, H. B. & Comer, M. A. (1983b). Pharmacological suppression of REM sleep prior to weaning counteracts the effectiveness of subsequent environmental enrichment on cortcal growth in rats. Brain Research, 283, 102-105. Moruzzi, G. (1972). The sleep-waking cycle. Ergebnisse der Physiologie, 6 4 , 1-165. Mouret, J., Coindet, J . & Chouvet, G. (1974). Effet de la pinealectomie sur les etats et rythmes de sommeil du rat male. Brain Research. 8 1 , 97-105. Parkes, J. D. (1985). Sleep and its disorders (pp. 216-221). London: W. B. Saunders. Parmeggiani, P. L. (1977). Interaction between sleep and thermoregulation. Waking B Sleeping, I , 123132. Parmeggiani, P. L. & Rabini, C. (1970). Sleep and environmental temperature. Archives of Italian Biology, 108, 369-387. Parmeggiani, P. L., Zamboni, G. & Cianci, T. (1974). Influence of anterior hypothalamic heating on the duration of fast wave sleep episode. Electroencephalography and Clinical Neurophysiology, 36, 465-470. Parmeggiani, P. L., Agnati, L. F., Zamboni, G. & Cianci, T. (1975). Hypothalamic temperature during the sleep cycle at different ambient temperatures. Electroencephalography and Clinical Neurophysiology, 3 8 , 589-596. Passouant, P., Halberg, F., Genicot, L., Popoviciu, L. & Baldy-Moulinier, M. (1969). La periodicite des acces narcoleptiques et le rytme ultradien du sommeil rapide. Revieu Neurologique, 121, 155164.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
114
R. SANDYK
Pavel, S., Goldstein, R. & Petrescu, M . (1980). Vasotocin, melatonin and narcolepsy: possible involvement of the pineal gland in its patho-physiological mechanisms. Peptzdes. I, 281-284. Pazo, J. H. (1979). Effects of melatonin on spontaneous and evoked neuronal activity in the mesencephalic reticular formation. Brain Research Bulletin, 4 , 725-730. Ralph, C. L. (1978). Pineal bodies and thermoregulation. In R. J. Reiter (Ed.), The pineal gland (pp. 193-219). New York: Raven Press. Ralph, C. L., Firth, B. T., Gern, W. A. & Owens, D. W. (1979). The pinel complex and thermoregulation. Biological Reviews, 5 4 , 41-72. Rechtschaffen, A. & Dement, E. C. (1969). Narcolepsy and hypersomnia. In A. Kales (Ed.), Sleep: physiology and parhology (pp. 1 19- 130). Philadelphia: Lippincott. Rechtschaffen, A , , Wolpert, E. A , , Dement, E. C., Mitchell, S. A. & Fisher, C. (1963). Nocturnal sleep of narcoleptics. Electroencephalography and Clinical Neurophysiology, 15, 599-609. Relkin, R., Fok, W. Y. & Schneck, L. (1973). Pinealectomy and brain myelination. Endocrinology. 4’2, 1426-1432. Roffwarg, H. P., Muzio, J. N. & Dement, W. C. (1966). Ontogenetic development of the human sleepdream cycle. Science, 152, 604-619. Romijn, H. J. (1978). The pineal, a tranquillizing organ? Life Sciences, 23, 2257-2274. Sallanon, M . , Claustrat, B. & Touret, M. (1982). Presence of melatonin in various cat brainstem nuclei determined by radioimmunoassay. Acta Endocrinologica, 101, 161-165. Silman, R. E., Leone, R. M. & Hooper, R. J. L. (1979). Melatonin, the pineal gland and human puberty. Nature, 282, 301-303. Sours, J. A. (1963). Narcolepsy and other disturbances in the sleep-waking rhythm: a study of I 1 5 cases with review of the literature. Journal of Nervous and Mental Disease, 137, 525-542. Spencer, F., Schirer, H. W. & Yochim, J. M. (1976). Core temperature in the female rat: effect of pinealectomy or altered lighting. American Journal of Physiology, 231, 355-360. Spielman, A. & Herrera, C. (1991). Sleep disorders. In S. J. Ellman & J. S . Antrobus (Eds.), The mind in sleep. psychology and psychophysiology (pp. 25-80). New York: Wiley. Steriade, M. & Hobson, J. A. (1976). Neuronal activity during the sleep-waking cycle. Progress in Neurobiology, 6 , 155-376. Stock, G. (1982). Neurobiology of REM sleep. A possible role for dopamine. In D. Ganten & D. Pfaff (Eds.), Sleep. clinical and experimental aspects (pp. 1-36). Berlin: Springer Verlag. Waldhauser, F., Weiszenbacher, G., Frisch, H., Zeitlhuber, U., Waldhauser, M. & Wurtman, R. J. (1984). Fall in nocturnal serum melatonin during prepuberty and pubescence. Lancet, I , 362-365. Williams, R. L., Karacan, I. & Hursch, C. J. (1974). Electroencephalography (EEG) of human sleep: clinical implications. New York: Wiley. Wurtman, R. J., Axelrod, J. & Kelly, D. E. (1968). The pineal. New York: Academic Press. Verley, R. & Garma, L. (1975). The criteria of sleep during ontogeny in different animal species. In G. C. Lairy & P. Salzarulo (Eds.), Experimental study of human sleep: methodological problems (pp. 111-125). Amsterdam: Elsevier.
Intern. J . Neuroscience, 1992, Vol. 6 3 , p. 105-114 Reprints available directly from the publisher Photocopying permitted by license only
MELATONIN AND MATURATION OF REM SLEEP REUVEN SANDYK
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
Democrition University of Thrace, Department of Medical Physics and Polytechnic School, Alexandroupolis and Xanthi, Greece (Received November 15, 1991)
The discovery in 1953 of rapid eye movement (REM) sleep and the appreciation that sleep is a heterogenous physiological state stimulated major research into sleep disorders. Electroencephalographic studies have shown that the amount of REM sleep changes with age. While newborns spend almost 50% of their sleep time in REM, the percentage of REM sleep decreases to 30% by the age of 3 months and to 20% by the age of 6 months. In addition, newborns enter REM sleep soon after the initiation of sleep, but by the age of 4 months entry into sleep assumes the adult pattern in which a significant period of non-REM sleep precedes the onset of REM sleep. Since reduction in the amount of REM sleep is associated with cerebral maturation and since the pineal gland has been implicated both in cerebral development and in the organization of REM sleep, the pineal gland may be involved in the maturation of the adult REM sleep pattern. Prior to the age of 3 months melatonin plasma levels are low and the characteristic circadian rhythms of melatonin are absent. Thereafter, melatonin secretion increases and circadian rhythmicity of melatonin becomes apparent. Thus, the abundance of REM sleep during the first 3 months of infancy is associated with deficient pineal melatonin functions, while the decline in the percentage of REM sleep coincides with the emergence of melatonin secretion coincident with the maturation of the pineal gland. I propose, therefore, that a state of low melatonin secretion is permissive for REM sleep and that maturation of the pineal gland retards REM sleep. This hypothesis is supported by the findings that melatonin suppresses REM sleep in cats and that in rats and humans pinealectomy induces a narcolepticlike pattern of REM sleep which strikingly resembles that of the newborn and which is reversed by the administration of melatonin. A further hypothesis is advanced to explain the pathophysiology of narcolepsy in terms of a maturational defect of the pineal gland in infancy. Keywords: REM sleep, Pineat gland. Melatonin. Infancy. Cerebral maruralion, Narcolepsy.
The discovery by Aserinsky and Kleitman (1953) of Rapid Eye Movement (REM) sleep and its association with dreaming stimulated great interest in sleep research. Polygraphic studies have revealed that sleep is a complex physiological process in which over the course of the night, there is an alternating pattern of non-REM (NREM) sleep and REM sleep with cycles lasting about 90 minutes each (Stock, 1982). A normal adult, upon falling asleep, exhibits a typical succession of electroencephalographic (EEG) changes which have been designated as “stages.” These EEG changes are all considered phases of the NREM sleep (Roffwarg et al., 1966). After fragmentation and obliteration of the alpha rhythm, the waves diminish slightly in frequency as their amplitude increases (Stage 1). This pattern is then replaced by highvoltage, notched slow-waves (K-complexes) (Stage 2 ) . Tall “delta” waves of 1-2 cycles per second progressively fill the record (Stage 3) and finally dominate it in virtually unbroken sequence (Stage 4). ~
Reprints requests to: Professor Reuven Sandyk M.D. ,M.Sc., Department of Medical Physics, Democrition University of Thrace, Alexandroupolis, Greece. 105
106
R. SANDYK
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
Adult Pattern of REM Sleep
Approximately one hour after the onset of sleep, the initial REM period of the night commences. REM periods recur every 80 to 90 minutes and comprise 20-2570 of the conventional sleep time of the adult. REM periods are shorter in the initial stages of sleep and become longer towards the morning with each REM period averaging about 20 minutes duration (Roffwarg et al., 1966). REM sleep is a period characterized by considerable cerebral excitation. This notion is supported by REM sleep EEG and physiological studies which reveal a remarkable similarity between REM sleep and a state of wakefulness (Roffwarg et al., 1966; Llinas and Pare, 1991). Animal studies indicate that REM sleep is associated with active cerebral processes, which include increase in cerebral blood flow to the cortex, rise in brain temperature, and increase in the frequency of spontaneous neuronal firing in the mesencephalic reticular formation, medial and descending vestibular nuclei, pyramidal tract, and occipital cortex (Roffwarg et al., 1966). During REM sleep excitability in the motor cortex is higher than during NREM sleep while in the sensory cortex it is as high as during NREM sleep. In both regions, however, excitability is greater than it is in the waking state. Moreover, REM sleep is associated also with increased responsiveness of thalamic and cortical neurons as compared to the waking state (Roffwarg et al., 1966). Early studies by Jouvet et al. (1962; 1965) have suggested that the generation of REM sleep is dependent upon an intact rostra1 pons (nucleus pontis caudalis). Following ablation of this area in cats, REM sleep disappears and the animal shows only two states, NREM and wakefulness which may gradually progress to insomnia and death. A decorticated cat, on the other hand, shows n o evidence of NREM sleep. Subsequent studies in decorticate and decerebrate humans have supported the analogues dependence of REM sleep upon brainstem mechanisms and of NREM sleep upon cortical mechanisms (cf. Roffwarg et a\., 1966). Based on the assumption that REM sleep is a brainstem phenomenon and NREM sleep a cortical dependent process, Jouvet ( 1 962; 1965) considered REM sleep to be a phylogenetically archaic state. The findings that in the newborn REM sleep matures earlier then NREM or waking and the observations that sustained periods of REM sleep appear directly after arousal without intermediary NREM sleep (Roffwarg et al., 1966) have been used in support of the “primitive” quality of the REM state. Conversely, NREM sleep, being dependent on the functioning of the neocortex, has been thought to be acquired in the process of telencephalization and was considered a “neo-sleep” (Jouvet, 1962; 1965). However, more recent studies suggest that the mechanisms of REM sleep generation are more complex, possibly involving the interaction between several CNS structures (Steriade and Hobson, 1976; Stock, 1982; Drucker-Colin and Valverde, 1982). The Evolutiori of REM Sleep
It has been proposed that REM and NREM sleep evolved at different times during the phylogeny of mammals. Allison and van Twyver (1970) postulated that NREM sleep evolved 180 million years ago in animals which are considered true mammals. REM sleep, on the other hand, was considered to evolve 50 million years later. Moruzzi (1972) viewed the adult human sleep-wake cycle as the result of inborn factors, namely, the development and maturation of the cerebral cortex from infancy (Kleitman, 1963) and its relationship with the brainstem as well as learned factors under the influence of day and night and social influences. Meddis (1977) hypoth-
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
107
esized that REM sleep is evolutionally the more ancient type of sleep and that it is less effective, especially since it is not appropriate for homeothennic species such as birds and mammals. This hypothesis is supported by the inappropriateness of REM sleep as a regulator of body temperature in mammals (Parmeggiani, 1977). Based on the findings that REM sleep is prominent in immature mammals and in the human newborn, it has been hypothesized to accompany and probably be causally related to cerebral maturation (Roffwarg et al., 1966). Extensive investigations on the evolution of REM sleep in humans have demonstrated an association between the percentage of REM/NREM sleep and cerebral maturation (Roffwarg et al., 1966; Williams et al., 1974). Observations in the newborn and immature animals revealed that REM sleep consumed a high proportion of total sleep time in the first days of life and that its amount diminished as maturation proceeded (Roffwarg et al., 1966). Specifically, in infancy when the proportion of time awake is smaller than in any other period of life, there is a large amount of REM activity. Characteristically, these REM periods appear soon after the onset of sleep and are of random duration and occurrence during the night. Moreover, during infancy there is neither inhibition of motor activity (atonia) that is characteristic in REM sleep, nor are phasic events restricted to REM periods (Ellman and Weinstein, 1991). Subsequently, when the developing infant spends longer intervals awake, the total amount as well as the proportion of REM sleep decreases (Roffwarg et al., 1966). In addition, the latency of REM onset, which is significantly longer in children (3-4 hours in the 4-7 year old as compared to one hour in the adult), continues to shorten with maturation reaching the adult pattern in midadolescence (Roffwarg et al., 1966). In the human fetus, REM sleep has been recorded as early as 6 months postconception, prior to any detection of NREM sleep. Later in fetal development, REM sleep occupies almost 90% of the overall sleep time. In premature infants of 32 to 36 weeks’ gestation, REM sleep constitutes 75% of total sleep. This drops sharply to about 50% of the overall sleep time at birth, and to about 30% at the age of 3 months and to 20% at the age of 6 months (Williams et al., 1974; Ellman and Weinstein, 1991). REM Sleep and Cerebral Maturation The high percentage of REM sleep time in the fetus and newborn as well as the progressive decline in the percentage of REM sleep time throughout childhood strongly suggest that REM sleep may be fulfilling a vital function in early brain development. Ephron and Carrington (1966) considered REM sleep a mechanism aimed at restoring cortical homeostasis, while Roffwarg et al. (1966) proposed that REM sleep affords intense stimulation to the central nervous system. It has been suggested that ascending impulses originating in the pons during REM sleep act to enhance neuronal differentiation, maturation, and myelinization in cortical areas (Roffwarg et al., 1966). This hypothesis receives support froni studies which have shown that administration of clonidine, which induces REM deprivation, attenuated cortical development induced by environmental stimulation in young rats (Mirmiran et al., 1983 a; b; Lewin and Singer, 1991). In accord with this hypothesis, several investigators have provided evidence to indicate a role for REM sleep in learning and consolidation of memory (Ellman and Weinstein, 1991). The main impetus for this hypothesis comes from animal data revealing high levels of REM sleep after learning various tasks and the observation that REM sleep deprivation interferes with learning processes (Ellman et al., 1991). In humans, increase in REM sleep time has been noted to occur
I08
R. SANDYK
several days after intense learning, mental stress, and particularly demanding events (Greenberg, 1981). Moreover, memory impairment due to subcortical dementia has been linked to diminished REM sleep (Parkes, 1985). According to Llinas and Pare (1991) “REM sleep can be considered as a modified attentive state in which attention is turned away from the sensory input, toward memories” (p. 525).
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
Regulation of REM Sleep
The findings that REM sleep occupies a great proportion of the infant’s sleep time stimulated research into the mechanisms responsible for the profusion of REM sleep in the immature organism or, for that matter, for the regulation of the proportion of REM sleep at any age. Ellman and Weinstein (1991) proposed that REM restriction, which is initiated at 3 months of age, is an important factor in infant development. One of the more attractive theories discussed by Roffwarg et al. (1966) was that the amount of REM sleep is a consequence of deficient restraining influences on the brainhtem REM center. This hypothesis maintains that abundance of REM sleep in the newborn reflects the presence of insufficient cortical inhibitory influences on the REM center. If this hypothesis is correct, then the amount of REM sleep should increase after a mature organism had been subjected to decortication. However, observations in decorticated cats failed to support this hypothesis in part because these animals were found to have less REM sleep than intact animals (cf. Roffwarg et al., 1966). In experimental animals, pinealectomy is associated with delayed cerebral maturation (Relkin et al., 1973; Kamback et al., 1982). Moreover, pinealectomy both in rats (Kawakami et al., 1972; Mouret et al., 1974) and man (Mouret et al., 1974) induces a narcoleptic-like pattern of REM sleep which is reversed by administration of melatonin (Kawakami et al., 1972). Therefore, since the pineal gland is involved in the organization of circadian sleep/wake cycle including REM sleep (Anton-Tay et al., 1971; Mouret et al., 1974; Cramer et al., 1976) and since it is implicated in psychosexual and pubertal development in humans (Silman et al., 1979; Waldhauser et al., 1984; Lissoni et al., 1985), the pineal gland may be involved also in the maturation of the adult REM/NREM sleep cycle. In the present communication, I propose that the abundance of REM sleep in infancy is maintained by a state of deficient melatonin functions. Furthermore, maturation of the pineal gland associated with the development of circadian rhythms of melatonin secretion exerts a restraining influence on REM sleep. Ontogeny of Pineal Chronobiology
Since the discovery of melatonin as the principal hormone of the pineal gland, it was found that the secretion of the hormone varies in relation to age and level of sexual maturation (Silman et al., 1979; Waldhauser et al., 1984). Lemaitre et al. (1981) assayed melatonin in 58 subjects aged one day to 30 years and noted that urinary elimination progressively decreased from birth to adulthood. In a subsequent study, plasma and urinary melatonin estimations were undertaken between 0500 hrs and 0600 hrs in 26 normal newborn males ranging in age from 15 days to one year (Hartmann et al., 1982). Plasma melatonin levels were found to be low during the first 3 months of life and subsequently increased progressively and significantly between 3 and 12 months. Similar findings have been obtained by Attanasio et al. (1986). These observations reveal that during the first three months of life melatonin
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
109
levels are low as a result of either a relatively inactive synthesis or a lack of utilization of melatonin due to the presence of immature receptors. Studies on the ontogeny of pineal chronobiology in humans reveal that melatonin plasma levels fail to show any circadian rhythmicity during the first month of life (Gupta, 1986). However, the third month of life is associated with the emergence of a highly significant circadian rhythm for melatonin (Gupta, 1986). It is noteworthy that in species such as the rat in which the organization of the adult type of REM sleep does not occur at birth (Verley and Garma, 1975), melatonin also is not present at birth in the pineal gland (Klein and Lines, 1969) and its appearance (Klein and Lines, 1969) coincides with the organization of the adult type of REM sleep (Verley and Garma, 1975). However, in species such as the sheep, in which the organization of the adult type of REM sleep does occur at birth (Verley and Garma, 1975), melatonin is present in the fetal pineal gland (Kennaway and Seamark, 1978). In humans, the period during which the circadian rhythmicity of melatonin emerges in infants is also a crucial phase of neurologic development. Around the age of 3 months the neurological patterns characteristic of the newborn period disappear, replaced by new motor and behavioral patterns (Gupta, 1986). It has been suggested that the development of the overt circadian rhythmicity for melatonin may be involved in cerebral maturation (Gupta, 1986).
The Pineal Gland and REM Sleep The period associated with the maturation of the circadian rhythms of melatonin and the phase during which plasma melatonin levels begin to rise (i.e., 3 months) appear also to coincide with important milestones in the ontogeny of REM sleep. By age 3 months, the percentage of REM sleeptime has dropped from 50% in the newborn to 30% of the total sleep time. In addition, this period is associated with major changes in the organization of REM sleep, namely the elimination of undifferentiated REM states (waking, fussing, and drowsy REM) and the coalescing of random eye movements into burst patterns (Ellman and Weinstein, 1991). Furthermore, by 4 months of age the onset of sleep has been shown to shift from the newborn pattern of entry through REM sleep to the adult mode of entry through NREM sleep (Parkes, 1985). Collectively, these findings demonstrate that abundance of REM sleep and initiation of sleep through REM prior to the age of 3 months occur during a period which is associated with low melatonin secretion coupled with deficient circadian melatonin rhythms. A significant fall in the percentage of REM sleep and the emergence of an adult pattern of entry into sleep through NREM sleep occurs after age 3 months, a period in which concurrently there is a rise in melatonin secretion and maturation of the melatonin circadian rhythms. Melatonin receptors have been demonstrated in various nuclei of the brainstem reticular system (Sallanon et al., 1982), which are implicated in the regulation of the sleep-wake cycle (Parkes, 1985). Melatonin has been shown to modify the spontaneous electrical activity of single cells in the mesencephalic reticular formation of the rat (Pazo, 1979) and pinealectomy has been reported to disrupt the circadian REM sleep in rats and humans (Mouret et al., 1974). Moreover, the pineal gland has been shown to be involved in the regulation of attentional, behavioral, and sleepwake mechanisms (Marczynski et al., 1964; Anton-Tay, 1974; Romijn, 1978; Datta and King, 1980; Birkeland, 1982; Dugovic et al., 1989). Thus, it is unlikely that the association between melatonin secretion and the maturation of REM sleep in infancy reflects an epiphenomenon.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
110
R. SANDYK
I propose, therefore that melatonin deficiency is permissive for REM sleep and that maturation of the pineal gland may exert a restraining effect on REM sleep while simultaneously enhancing the development of REM/NREM adult sleep pattern. This hypothesis is supported by the findings that melatonin suppresses REM sleep in cats (cf. Pave1 et al., 1980) and that pinealectomy, which removes the main source of circulating melatonin (Wurtman et al., 1968), induces both in rats and man (Kawakami et al., 1972; Mouret et al., 1974) a narcoleptic-like pattern of REM sleep which strikingly resembles that of the newborn (Passouant et al., 1969). The hypothesis that decreased melatonin plasma levels is permissive for REM sleep is supported also by the findings in adults that REM periods are associated with melatonin nadirs (Birkeland, 1982). Since melatonin plasma levels rise at the onset of NREM sleep (Birkeland, 1982), it is possible that activation of the pineal gland during this period of sleep may function to inhibit REM sleep. Consequently, the transitions between REM and NREM sleep cycles may be influenced by fluctuations in the secretory activity of the pineal gland, which may thus function to maintain the homeostasis of the sleep cycle. Maturation of REM Sleep and Temperature Regulation
The mechanisms by which the pineal gland may restrain REM sleep in infancy are unknown. One possible mechanism may involve the interactions among the pineal gland, thermoregulation, and REM sleep. The onset of sleep is usually preceded by a fall in body temperature (Geschickter et al., 1966). During the initial stages of sleep body temperature continues to fall. This is associated with increased heat loss, a relative high percentage of stage 3-4 sleep, and a low percentage of REM sleep. During the second half of sleep, body temperature rises. This is associated with decreased heat loss, a low percentage of stage 3-4 sleep, and a high percentage of REM sleep (Avery, 1987). REM sleep is most likely to occur at the nadir of the body temperature cycle (Czeisler et al., 1980). These findings thus indicate a strong coupling of the circadian rhythm of the REM sleep events and changes in body temperature. The amount of REM sleep and the ratio of REM to NREM sleep has been shown also to be influenced by the ambient temperature. Both low and extremely high ambient temperatures have been shown to decrease the amount of REM sleep in animals (Parmeggiani and Rabini, 1970) and man (Beck et al., 1976). Increasing the body temperature by inducing fever reduces REM sleep in humans (Karacan et al., 1968). There is evidence that the hypothalamus mediates the effects of the ambient temperature on REM sleep. It has been shown that the occurrence of REM sleep is related to a narrow range of hypothalamic temperature, a temperature “gate” (Parmeggiani et al., 1975). For instance, once an animal is in stage REM sleep, warming the hypothalamus increases the duration of the REM period (Parmeggiani et al., 1974). The occurrence of REM sleep may thus be increased by experimental manipulations of the hypothalamus and ambient temperatures (Avery, 1987). There is now abundant evidence to implicate the pineal gland in circadian thermoregulation (Ralph, 1984). In fact, it has been suggested that the pineal gland is an integral, if not a key feature in the translation of photoperiodic stimuli into thermoregulator~responses (Ralph, 1984). In a comprehensive review on thermoregulation, Ralph et al. (1979) proposed that the pineal gland acts to prevent excessive or debilitating elevations of body temperature. Such a role becomes obvious given the pivotal position of the pineal gland in the adaptation of the organism to the
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
111
environment. It is presently thought that thermoregulatory functions of the pineal gland involve the mediation of the hypothalamus (Ralph, 1987). In laboratory animals administration of melatonin has been shown to affect body temperature and thermoregulation. In rats, injection of 60 micrograms of melatonin into the cerebral ventricles produces a rise in body temperature for 2 hours (Fioretti and Martini, 1968). Similarly, intraperitoneal injections of melatonin (4 or 8 microgramslg body weight) induced marked hyperthermia in rats (Fioretti et al., 1974), while pinealectomy induced hypothermia (Fioretti et al., 1974; Spencer et al., 1976). Moreover, hypothermia induced by barbiturates, neuroleptics, or morphine has been shown to be reversed by melatonin, suggesting a central action of melatonin on temperature regulation (Ralph, 1984). As indicated above, REM sleep, which is associated with a melatonin nadir (Birkeland, 1982) is also accompanied by temperature nadirs (Czeisler et al., 1980). On the other hand, increased body temperature reduces REM sleep (Karacan et al., 1968). Since administration of melatonin is associated with elevation of body temperature, which in turn decreases REM sleep, melatonin may thus impact on the maturation of REM sleep. Specifically, since the infant in the first 3 months of life is functionally “pinealectomized” and thus has a tendency to become hypothemic, the presence of low body temperature would enhance the profusion of REM sleep. However, with the rise in melatonin secretion and the establishment of the melatonin circadian rhythms after the 3rd month of life, body temperature would tend to rise. This in turn could lead to inhibition of REM sleep. In summary, the pineal gland, through its action on the hypothalamic thermoregulatory systems, may induce or inhibit REM sleep by altering body temperature. Maturation of the Pineal Gland and Narcolepsy
The hypothesis that the pineal gland is involved in the maturation of REM sleep may be relevant to the pathophysiology of several sleep disorders including Kleine-Levin syndrome and narcolepsy. Narcolepsy is a syndrome of excessive sleepiness associated with disruption of circadian rhythms of the sleep-wake cycle (Rechtschaffen and Dement, 1969; Hishikawa et al., 1976). Polygraphic investigations of the sleep in narcoleptic patients show the so-called Sleep Onset REM Periods (SOREMPs) to be a fundamental characteristic of the disorder (Rechtschaffen et al., 1963), occurring at the beginning of the nocturnal sleep period as well as during daytime naps (Dement et al., 1966). SOREMPs occur also in patients with sleep apnea, sleep-wake schedule disturbances, major depression, and in normal infants (Spielman and Herrera, 1991). The occurrence of SOREMPs in infancy and in narcolepsy suggest that in both SOREMPs may be related to a common mechanism related to a failure of the pineal gland to restrain REM sleep in early life. Consequently, narcolepsy may be viewed as a maturational defect of REM organization which is likely to result from dysfunction of the pineal gland in infancy. It is therefore not surprising that in the majority of patients the disease manifests itself in late adolescence (Sours, 1963), a period which is associated with dramatic changes in pineal functions (Silman et al., 1979; Waldhauser et al., 1984). The findings that the circadian rhythms of melatonin are absent in narcolepsy (Birau et al., 1982) and that melatonin fails to induce REM sleep at sleep onset or to increase the amount of REM sleep in narcoleptic patients (Pave1 et al., 1980) are supportive of the hypothesis that the disease may be causally related to abnormal pineal melatonin functions. Finally, since it is now recognized that the pineal gland plays a
112
R . SANDYK
fundamental role in immunomodulation (Maestroni et al., 1987 a; b) , it is possible that the abnormalities of the immune system found in the disease (Langdon et al., 1986) are linked to disruption of pineal melatonin functions.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REFERENCES Allison, T. & van Twyver, H. (1970). The evolution of sleep. (pp. 56-65). Vol. 79, New York: Natural History. Anton-Tay, F. (1974). Melatonin: effects on brain function. Advances in Biochemical Psychopharmacology, 11, 315-324. Anton-Tay, F., Diaz, J . L. & Femandez-Guardiola, A. (1971). On the effect of melatonin upon human brain. Its possible therapeutic implications. Life Sciences, 10, 841-850. Aserinsky, E. & Kleitman, N. (1953). Regularly occumng periods of eye motility and concomitant phenomena during sleep. Science, 118, 273-274. Attanasio, A,, Rager, K. & Gupta, D. (1986). Ontogeny of circadian rhythmicity for melatonin, serotonin and N-acetylserotonin in humans. Journal of Pineal Research, 3. 25 1-258. Avery. D. H. (1987). REM sleep and temperature regulatiun in affective disorder. In A. Halaris (Ed.), Chronobiology and psychiufric disorders (pp. 75-10]), New York: Elsevier. Beck, U., Reinhard, H., Kendel, K. & Schmidt-Kessen, W. (1976). Temperature and endocrine activity during sleep in man. Archives der Psychiatrie und Nervenkrankheiten, 232, 245-256. Birau, N., Mayer, C., Matsubayashi, K. & Meier-Ewert, K. H. (1982). Melatonin serum concentration during the nocturnal sleep of narcoleptics. IRCS Journal of Medical Science, 1 0 , 814. Birkeland, A . J . (19x2). Plasma melatonin levels and nocturnal transitions between sleep and wakefulness. Neuroendocrinology, 34, 126-131. Cramer, H., Bohme, W., Kendel, K. & Donnadieu, M. (1976). Freisetzung von Wachtumshormon und von Melanozyten stimulierenden Hormone in durch Melatonin gebahnten Schlaf beim Menschen. Arzneimittel-ForschunR, 2 6 , 1076- 1078. Czeisler, C. A., Weitzman, E. D., Moore-Ede, M. C., Zimmerman, J. C. & Kronauer, R. E. (19x0). Human sleep: its duration and organization depend on its circadian phase. Science. 210, 12641267. Datta, P. C. & King, M. G . (1980). Melatonin: effects on brain and behavior. Neuroscience & Biobehavioral Reviews. 4 , 451-458. Dement, W. C . , Rechtschaffen, A. & Gulevich, G. (1966). The nature of narcoleptic sleep attack. Neurology. 16, 18-33. Drucker-Colin, R. & Valverde, C. (1982). Endocrine and peptide functions in the sleep-waking cycle. In G. Ganten & D. Pfaff (Eds.), Sleep. clinical and experimental aspects (pp. 37-81). Berlin: Springer Verlag. Dugovic, C., Leysen, J. E. & Wauquier, A. (1989). Melatonin modulates the sensitivity of 5-hydroxytryptamine-2 receptor-mediated sleep-wakefulness regulation in the rat. Neuroscience Letters. 104, 320-325. Ellman, S. J. & Weinstein, L. N. (1991). REM sleep and dream formation: A theoretical integration. In S. J. Ellman & J. S. Antrobus (Eds.), The mind in sleep. psychology andpsychophysiology (pp. 466-488). New York: Wiley. Ellman, S. J., Luck, D., Steiner, S. S. & Halperin, R. (1991). REM deprivation: a review. In S. J . Ellman & J. S. Antrohns (Eds.), The mind in sleep. psychology and ps,ychophysiology (pp. 329376). New York: Wiley. Ephron, H. S. & Carrington, P. (1966). Rapid eye movement sleep and cortical homeostasis. Psychological Review, 75, 500-526. Fioretti, M. C. & Martini, L. (1968). Effetti della malatonina sulla temperatura corporea. Atti Accudemia Anatomica Chirurgia, 6 0 , 3-8. Fioretti, M. C . Riccardi, C., Menconi, E. & Martini, L. (1974). Control of the circadian rhythm of the body temperature in the rat. Life Sciences, 14, 211-219. Geschickter, E. H., Andrews, P. A. & Bullard, R. W. (1966). Nocturnal body temperature regulation in man: A rationale for sweating in sleep. Journul of Applied Physiology, 2 1 , 623-630. Greenherg, R. (1981). Dreams and REM sleep-an integrative approach. In W. Fishbein (Ed.), Sleep, dreams and memory (pp. 125-135). Lancaster: MTP Press. Gupta, D. (1986). The pineal gland in relation to growth and development in children. Journal of Neural Transmission, 2I (snppl), 2 17-232. Hartmann, L., Roger, M., Lemaitre, B. J . , Massias, J . F. & Chaussain, J. L. (1982). Plasma and urinary melatonin in male infants during the first 12 months of life. Clinica Chimicu Acta, 121, 37-42.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
REM SLEEP AND THE PINEAL GLAND
113
Hishikawa, Y., Wakamatsu, H., Furuya, R., Sugita, Y., Masaoka, S., Kaneda, H., Sato, M., Nanno, H. & Zaneko, Z. (1976). Sleep satiation in narcoleptic patients. Electroencephalography and Clinical Neurophysiology, 4 1 , 1-18. Jouvet, M. (1962). Recherches sur les structures nerveuses et les mechanismes responsables des differentes phases du sommeil physiologique. Archives Italiennes de Biologie, 100, 125-206. Jouvet, M. (1965). Paradoxical sleep. A study of its nature and mechanisms. Progress in Brain Research, 18, 20-57. Kamback, D. O., Rich, R. A. & Relkin, R . (1982). Effect of pinealectomy on fatty acid composition of rat brain myelin. Endocrinology, 110, 907-909. Karacan, I., Wolff, S. M., Webb, W. B. & Williams, R. L. (1968). The effects of fever on sleep patterns. Psychophysiology, 5 , 255. Kawakami, M., Yamaoka, S. & Yamaguchi, T. (1972). Influence of light and hormones upon circadian rhythm of EEG slow-wave and paradoxial sleep. In S . Itoh, K. Ogata & H. Yoshimura (Eds.), Advanes in climatic physiology (pp. 349-366). Tokyo: Shoin. Kennaway, D. J. & Seamark, R. F. (1978). The occurrence and synthesis of melatonin in the fetal sheep pineal gland. Journal of Neural Transmission, 13, 371-372. Klein, D. C. & Lines, S . V. (1969). Pineal hydroxyindole-0-methyltransferaseactivity in the growing rat. Endocrinology, 8 4 , 1523-1525. Kleitman, N. (1963). Sleep and wakefulness (p. 552). Chicago: University of Chicago Press. Langdon, N., Lock, C., Welsh, K., Vergani, D., Dorow, R. Wachtel, H., Palenschaft, D. & Parkes, D. (1986). Immune factors in narcolepsy. Sleep, 9, 143-148. Lemaitre, B. J., Bouillie, J . & Hartmann, L. (1981). Variations of urinary melatonin excretion in humans during the first 30 days of life. Clinica Chimica Acfa. 110, 77-82. Lewin, 1. & Singer, J. L. (1991). Psychological effects of REM ("dream") deprivation upon waking mentation. In S. J . Ellman & J. S. Antrobus (Eds.), The mind in sleep. psychology and psychophysiology (pp. 398-412). New York: John Wiley. Lissoni, P., Mauri, R., Resentini, M., Morabito, F., Djemal, S. & Franshini, F. (1985). Melatonin rhythms in some idiopathic hypothalamic-hypophyseal diseases. In G. M. Brown & S. D. Wainwright (Eds.), The pineal gland; endocrine aspects (pp. 307-3 12). Oxford: Pergamon Press. Llinas, R. R. & Pare, D. (1991). Commentary of dreaming and wakefulness. Neuroscience, 4 4 , 521535. Maestroni, G. J. M., Conti, A. & Pierpaoli, W. (1987~).The pineal gland and the circadian opiatergic, immunoregulatory role of melatonin. Aanals of the New York Academy of Science, 496, 67-77. Maestroni, G. J. M., Conti, A. & Pierpaoli, W. (1987b). Melatonin counteracts the effect of acute stress on the immune system via an opiatergic system. In G. P. Trentini & C. de Gaetani (Eds.), Fundamentals and clinics in pineal research (pp. 277-279). New York: Raven Press. Marczynsky, T. J., Yamaguchi, N., Ling, G. M. & Grodziska, L. (1964). Sleep induced by the administration of melatonin (5-methoxy-N-acetyltryptamine)to the hypothalamus in unrestrained cats. Experientia, 2 0 , 435-436. Meddis, R. (1977). The sleep instinct. London: Routlege & Kegan Paul. Mirmiran, M., Scholtens, J., van de Poll, N . E.,Uylings, H. B., van der Gugten, J. & Boer, G. J. (1983~).Effects of experimental suppression of active (REM) sleep during early development upon adult brain and behavior in the rat. Brain Research, 283, 277-286. Mirmiran, M., Uylings, H. B. & Comer, M. A. (1983b). Pharmacological suppression of REM sleep prior to weaning counteracts the effectiveness of subsequent environmental enrichment on cortcal growth in rats. Brain Research, 283, 102-105. Moruzzi, G. (1972). The sleep-waking cycle. Ergebnisse der Physiologie, 6 4 , 1-165. Mouret, J., Coindet, J . & Chouvet, G. (1974). Effet de la pinealectomie sur les etats et rythmes de sommeil du rat male. Brain Research. 8 1 , 97-105. Parkes, J. D. (1985). Sleep and its disorders (pp. 216-221). London: W. B. Saunders. Parmeggiani, P. L. (1977). Interaction between sleep and thermoregulation. Waking B Sleeping, I , 123132. Parmeggiani, P. L. & Rabini, C. (1970). Sleep and environmental temperature. Archives of Italian Biology, 108, 369-387. Parmeggiani, P. L., Zamboni, G. & Cianci, T. (1974). Influence of anterior hypothalamic heating on the duration of fast wave sleep episode. Electroencephalography and Clinical Neurophysiology, 36, 465-470. Parmeggiani, P. L., Agnati, L. F., Zamboni, G. & Cianci, T. (1975). Hypothalamic temperature during the sleep cycle at different ambient temperatures. Electroencephalography and Clinical Neurophysiology, 3 8 , 589-596. Passouant, P., Halberg, F., Genicot, L., Popoviciu, L. & Baldy-Moulinier, M. (1969). La periodicite des acces narcoleptiques et le rytme ultradien du sommeil rapide. Revieu Neurologique, 121, 155164.
Int J Neurosci Downloaded from informahealthcare.com by Freie Universitaet Berlin on 12/04/14 For personal use only.
114
R. SANDYK
Pavel, S., Goldstein, R. & Petrescu, M . (1980). Vasotocin, melatonin and narcolepsy: possible involvement of the pineal gland in its patho-physiological mechanisms. Peptzdes. I, 281-284. Pazo, J. H. (1979). Effects of melatonin on spontaneous and evoked neuronal activity in the mesencephalic reticular formation. Brain Research Bulletin, 4 , 725-730. Ralph, C. L. (1978). Pineal bodies and thermoregulation. In R. J. Reiter (Ed.), The pineal gland (pp. 193-219). New York: Raven Press. Ralph, C. L., Firth, B. T., Gern, W. A. & Owens, D. W. (1979). The pinel complex and thermoregulation. Biological Reviews, 5 4 , 41-72. Rechtschaffen, A. & Dement, E. C. (1969). Narcolepsy and hypersomnia. In A. Kales (Ed.), Sleep: physiology and parhology (pp. 1 19- 130). Philadelphia: Lippincott. Rechtschaffen, A , , Wolpert, E. A , , Dement, E. C., Mitchell, S. A. & Fisher, C. (1963). Nocturnal sleep of narcoleptics. Electroencephalography and Clinical Neurophysiology, 15, 599-609. Relkin, R., Fok, W. Y. & Schneck, L. (1973). Pinealectomy and brain myelination. Endocrinology. 4’2, 1426-1432. Roffwarg, H. P., Muzio, J. N. & Dement, W. C. (1966). Ontogenetic development of the human sleepdream cycle. Science, 152, 604-619. Romijn, H. J. (1978). The pineal, a tranquillizing organ? Life Sciences, 23, 2257-2274. Sallanon, M . , Claustrat, B. & Touret, M. (1982). Presence of melatonin in various cat brainstem nuclei determined by radioimmunoassay. Acta Endocrinologica, 101, 161-165. Silman, R. E., Leone, R. M. & Hooper, R. J. L. (1979). Melatonin, the pineal gland and human puberty. Nature, 282, 301-303. Sours, J. A. (1963). Narcolepsy and other disturbances in the sleep-waking rhythm: a study of I 1 5 cases with review of the literature. Journal of Nervous and Mental Disease, 137, 525-542. Spencer, F., Schirer, H. W. & Yochim, J. M. (1976). Core temperature in the female rat: effect of pinealectomy or altered lighting. American Journal of Physiology, 231, 355-360. Spielman, A. & Herrera, C. (1991). Sleep disorders. In S. J. Ellman & J. S . Antrobus (Eds.), The mind in sleep. psychology and psychophysiology (pp. 25-80). New York: Wiley. Steriade, M. & Hobson, J. A. (1976). Neuronal activity during the sleep-waking cycle. Progress in Neurobiology, 6 , 155-376. Stock, G. (1982). Neurobiology of REM sleep. A possible role for dopamine. In D. Ganten & D. Pfaff (Eds.), Sleep. clinical and experimental aspects (pp. 1-36). Berlin: Springer Verlag. Waldhauser, F., Weiszenbacher, G., Frisch, H., Zeitlhuber, U., Waldhauser, M. & Wurtman, R. J. (1984). Fall in nocturnal serum melatonin during prepuberty and pubescence. Lancet, I , 362-365. Williams, R. L., Karacan, I. & Hursch, C. J. (1974). Electroencephalography (EEG) of human sleep: clinical implications. New York: Wiley. Wurtman, R. J., Axelrod, J. & Kelly, D. E. (1968). The pineal. New York: Academic Press. Verley, R. & Garma, L. (1975). The criteria of sleep during ontogeny in different animal species. In G. C. Lairy & P. Salzarulo (Eds.), Experimental study of human sleep: methodological problems (pp. 111-125). Amsterdam: Elsevier.
