SLEEP AND MEMORY I:
THE INFLUENCE OF DIFFERENT SLEEP STAGES ON MEMORY
Vadim S. Rotenberg vadir@post.tau.ac.il
NEUROSCIENCE and BIOBEHAVIORAL REVIEWS, 16(4) 497-502, 1992.
A new approach to the sleep stages role in memory is discussed in the context of the two opposite patterns of behavior-search activity and renunciation of search. Search activity is activity designed to change the situation (or the subjects attitudes to it) in the absence of a definite forecast of the results of such activity, but with the constant consideration of these results at all stages of activity. Search activity increases general adaptability and body resistance while renunciation of search decreases adaptability and requires REM sleep for its compensation.Unprepared learning, which is often accompanied by failures on the first steps of learning, is suggested to produce renunciation of search, which decreases learning ability, suppress retention, and increases REM sleep requirement. A prolonged REM sleep deprivation before training causes learned helplessness and disturbs the learning process, while short REM sleep deprivation causes the "rebound" of the compensatory search activity that interferes with passive avoidance. REM sleep deprivation performed after a training session can increase distress caused by a training procedure, with the subsequent negative outcome on retention.
Key words: Memory, REM sleep, Search activity, Renunciation of search, Learned helplessness, REM sleep deprivation , Prepared vs. unprepared learning
The results of numerous investigations performed on man and animals and directed toward the elucidation of the influence of sleep on memory functions contain many contradictions, and there are some reasons to suggest that these contradictions are due to an underestimation of the peculiarities of the scientific methods used. Therefore, before analyzing the results per se, it is necessary to discuss some methodological aspects of these investigations.
There are three main approaches to the examination of the problem of sleep and memory:
1. The investigation of sleep structure during different stages of the assimilation of new information and at different intervals after instruction.
2. The investigation of the influence of sleep deprivation on the retention and on recall of new information; deprivation may be performed before the beginning of the learning process, during this process, and after this process is finished.
3. The investigation of sleep structure in patients suffering from memory dysfunction caused by organic brain disease.
The first two approaches are used most often, with a variety of methods of education and of sleep stage deprivation. Some contradictions in the data may be partly due to these differences in methods.
First of all, the effect of education and the peculiarities of the interaction between sleep and memory systems depend on the nature of the material that is being assimilated. We will employ the classification of learning created by (30), i.e., pre pared and unprepared learning. Prepared learning involve: tasks for which the organism is already "hard wired," i.e. either it is instinctual or overlearned. Unprepared learning on the other hand, is involved with tasks for which a reorientation of expectations or behavior is required. Prepared learning includes simple active and passive avoidance during aversive stimulation, and also the formation of conditioned reflexes by reinforcement. An analogue of prepared learning in humans may be the assimilation of simple, emotionally unimportant and irrelevant information, for instance the assimilation of a list of words or paired word associates. Unprepared learning includes the two-way avoidance task and the assimilation of complicated behavior required to achieve food reinforcement. An analogue of this type of learning in humans would be the assimilation of a complicated text with a distorted semantic. In the same category some authors (9) include the assimilation of emotionally meaningful and relevant information, for instance, information related to difficult relationships with significant persons, or the activation of unrealized wishes.
It is necessary to stress that in discussing the problem of sleep function in these two forms of learning, the authors do not take into account that unprepared learning is usually accompanied by much more pronounced emotional stress than prepared learning. Therefore, some stages of the assimilation of complicated and new information, being difficult and hard for the subject, may produce renunciation of search (28) because it may be impossible to overcome the serious obstacle in a short time. Search activity means activity designed to change the situation (or the subject's attitude to it) in the absence of a definite confidence in the successful outcome of results of such activity, but with constant consideration of these results at all stages of activity. Search activity, manifesting itself in different concrete forms of behavior (fight, flight, self-stimulation, creativity) increases body resistance; renunciation of search, is displayed in freezing, panicky behavior, depression and neurotic anxiety; and decreases body resistance (25). The renunciation of search, however, not only decreases body resistance, but also puts additional obstacles in the way of the assimilation of the new information.
As mentioned in our previous publications, REM sleep is a compensatory mechanism that helps to restore search activity (25,27,28), and this REM sleep function must be taken into account during the examination of sleep structure alterations in the process of prepared and unprepared learning.
REM sleep deprivation causes even more complicated methodological problems. First of all, there are different methods of REM sleep deprivation in animals and these have different effects on behavior and health. The classical method of the wooden platform, surrounded by water (11) is very effective, but is undoubtedly combined with severe stress due to relative immobilization, periodic falls into cold water and a pronounced restriction of spontaneous behavior, including the frustration of search behavior (25). It is impossible to control all these stress conditions in an appropriate manner. First of all, on the wooden platform all these conditions work together, and it is not the same as the reproduction of each of them in the isolation from each other.
Secondly, most investigators ignore the frustration of the need for search activity, which is extremely important. And, finally, the conditions on the little wooden platform are exceptional because all the above-mentioned factors are combined with the chronic REM sleep deprivation. I suggest that the essence of this method lies in the combination of conditions that are able to produce renunciation of search (learned helplessness) and at the same time deprive REM sleep, even as the requirement for the latter has been sharply aroused due to the frustration of search during wakefulness.
Such a combination must differ from REM sleep deprivation and stressful conditions used separately. For instance, if REM sleep deprivation is performed by stimulation of the brain stem reticular formation in animals placed in comfortable conditions (13,29), the experimental animals do not display any signs of stress. Moreover, the activity of the hypo-thalamus-pituitary-adrenal system is even reduced, and there are no visible changes in behavior that are so obvious after REM deprivation on the wooden platform. (For example: self-stimulation immediately after REM deprivation is not increased; there are no signs of agitation in the open field; no signs of aggression, hyperphagia or hypersexuality). If the animal is deprived of REM sleep by awakenings with noxious stimulation and demonstrates active defensive or orienting behavior (19) the total REM sleep deprivation also does not lead to the above-mentioned behavior disturbances, and REM sleep rebound after this procedure is restricted. These studies help to explain what happens on the wooden platform. Conditions on the platform surrounded by water frustrate search behavior. A long-term experience of uncontrollable events produces learned helplessness. The animal is not able to change its situation which, in the end, inevitably provokes a state of renunciation of search. The emergence of somatic disturbances, hemorrages in the gastric epithelium, and gastric ulcers furnish relevant evidence. There most likely is a parallel increase in REM requirement that is constantly suppressed.
When the animal is subsequently removed from the platform and placed in conditions where nothing prevents behavioral (search) activity, the latter takes a variety of forms ranging from intensified self-stimulation and an increase number of runs in the open field, to hypersexuality. Something like rebound of frustrated search activity takes place. This rebound is heightened by the preceding high emotional tension. Such a rebound shows that the brain still has sufficient functional possibilities, and that the REM sleep deprivation has not yet led to their exhaustion. The change of situation for a more favourable and sufficiently simple one fosters their activation. But, if the REM sleep deprivation on platform lasts a long time, the body's reserves can become depleted and renunciation of search will prevail (25). I wish to stress that exposure to uncontrollable events performed during the classical "learned helplessness" training, as well as the procedure of sleep deprivation on the platform, has a biphasic effect as a function of the aversive experience duration: it first produces increased behavior activation, which then gives way to its opposite (38).
Compared to the classical REM sleep deprivation procedure according to Jouvet, Rechtschaffen's procedure (24), at first glance, permits a comparison of control and experimental groups by the deprivation conditions, while avoiding excessive stress for both groups. The control and the test animal forced to make efforts to avoid getting into the water an equal number of times a day; but the test animals do so whenever they fall asleep or enter into the REM sleep phase, whereas in the control group such plunges are not connected with sleep, and they have a chance to make up for lost sleep while experimental animals are awake.
Thus, the fundamental difference between the groups confined to the fact that the experimental group is deprived of all sleep or of certain stages, and this eventually lead the death of the animals. There is a high direct correlation between the duration of survival and the degree of REM sleep intactness. Before they died, the test animals showed pronounced distress symptoms and a high energy expenditure. The same lethal consequences follow from total sleep deprivation, from selective full sleep deprivation, and from a similar deprivation of high-amplitude NREM sleep (3).
This may give the impression that only sleep deprivation is responsible for all physiological changes in the organism and the death of the test animals. In the present author's opinion, however, this is not the case. In my opinion, the deprivation conditions described by Rechtschaffen and Jouvet have much more in common than may appear at first glance. Indeed, the test animals are awoken by the threat to plunge into the water whenever they need sleep or certain of its stages, and the specific requirement is frustrated. In the course of such frustration, the animal
learns the inexorability of punishment. Naturally, each time it an eventually escape the water, but it cannot prevent the very fact of frustration used as punishment. That which can easily be coped with in waking is more agonizing when there is great sleep requirement. Whereas a control animal does not develop the experience of the inexorability of punishment at every attempt to satisfy its sleep requirement, the experiment animal has exactly such an experience, and this eventually leads it to learned helplessness. In such a case, there is the simultaneous suppression of REM sleep as the compensatory mechanism for the frustration of search activity. In the case of selective deprivation of high amplitude NREM sleep, there is a rivalry between the increased requirement for it due to the
deprivation of NREM and the increased REM sleep requirement due to the conditions of deprivation (the development of learned helplessness). Such a rivalry shortens REM sleep in these animals compared to the control group (with presumably increased requirement of it). Rechtschaffen's observation is of interest here: that among the control animals in the gravest functional state, there was a rat that showed the shortest REM sleep. Thus, in the final analysis, the lethal effect of the deprivation of sleep and its individual stages is due to a combination of conditions that provokes the state of renunciation of search and simultaneously impedes the use of the REM sleep compensatory mechanism. This conclusion is indirectly supported by the results of the study of post-deprivation restorative sleep. To the authors' own surprise, both after selective REM sleep deprivation and after total deprivation of all sleep, the animals showed, above all, a compensatory "rebound" of only REM sleep. Only the rats administered the selective deprivation of the high-amplitude NREM phase in the restorative sleep exhibited its rebound, but in these cases, too, the NREM sleep rebound alternated with an almost equally intensive REM sleep rebound (4).
The present author discussed the above-mentioned suggestion together with Rechtschaffen in his laboratory in Chicago, and professor Rechtschaffen presented the following arguments against the hypothesis that the fatal result of sleep and REM sleep deprivation is due to the combination of renunciation of search and REM sleep deprivation:
1. Animals in Rechtschaffen's experiment display hyperphagia, while in the ordinary state of learned helplessness, appetite is reduced.
2. Animals in Rechtschaffen's experiment displayed ulcers on their extremities and tails, but not in the mucosa of the stomach; the latter characterise learned helplessness.
3. Animals' deaths was preceded by a disturbance of thermo-regulation, which is not described in experiments with learned helplessness.
The present author can explain two of these contradictions: 1. In ordinary learned helplessness training, all punishments take place only during wakefulness, and only sleep is free from punishment. That is why, during wakefulness, when the punishment is unpredictable and the animal can receive it any minute, all biological motivations are inhibited (sexual motivation and appetite). During REM sleep deprivation, with Rechtschaffen's method, the situation is quite the opposite: All punishments take place only during sleep, and wakefulness is free from punishment, thus it is not surprising that during wakefulness appetite is even increased-a common reaction to stress during periods that are free from punishments, especially if other behavioral reactions are blocked.
2. The same conditions may explain the absence of stomach ulcers, since it is well known that ulcers appear predominantly in the empty stomach, and if the animal continues to eat during stress, ulcers do not appear.
It is necessary to consider the possibility that the effect of REM sleep deprivation on memory depends on the method of deprivation, and that deprivation on the platform, in direct relation to its length, can change the functional state of the animal from hypercompensatory search activity to renunciation of search; and that renunciation of search per se may also interfere with the learning process. For example, hyperactivity can interfere with passive avoidance, but promote simple active avoidance. The above-mentioned position helps us also
to change the point of view regarding the pharmacological method of REM sleep deprivation.- There are some observations (22) that REM sleep deprivation caused by imipramine or some other drugs has the same influence on learning as REM sleep deprivation on the wooden platform. The authors come to the conclusion that REM sleep reduction is the common factor responsible for this effect.
My earlier analysis has led us to the conclusion that the behavioral effect of REM sleep deprivation on the platform depends at least in part on the combination of REM sleep deprivation and distress, while REM sleep deprivation performed by Kovalzon and Tsibulski (13) does not have such an influence on behavior. For this reason, it is also possible to suggest that the similarity between the effect of REM sleep deprivation on memory, whether the deprivation is performed on the platform or by imipramine, is only superficial, and that, therefore, the specific mechanisms of the influence of these two methods of REM deprivation on memory may also be different. For instance, the direct influence of drugs on the brain is not excluded. It is thus necessary to take into consideration that the method of REM sleep deprivation may be the main factor of REM sleep deprivation. Now let us turn to the analysis of the concrete experimental data.
THE INFLUENCE OF THE LEARNING PROCESS ON THE SUBSEQUENT SLEEP STRUCTURE
There have been many studies of sleep structure following the process of task solution. The common result of these studies is that REM sleep increases in the first 24, and especially in the first 3 hours after training (33). REM sleep was not increased after very simple task performance, but only after more complicated ones (see above for prepared vs. unprepared learning). Some investigators found a REM sleep increase after failure (7), but other investigators found a correlation between REM percentage and signs of high achievement in the learning process (14,20). There is also evidence suggesting that the greatest increase of REM sleep occurs during the 24 hours before the critical level of success in the learning process (34). Afterwards, the number of mistakes decreases sharply and REM sleep returns to the initial level. The more intense the learning process, the sooner there appears to be an increase in REM sleep percentage. After such a learning process, the first REM sleep period is especially important with regard to retention (33).
It is possible to suggest that REM sleep is especially important for the animal during the critical period when a sharp conversion from the previous stereotypical behavior to the new style of behavior take place. This conversion period may be prolonged, which could account for the increase in REM sleep over several days.
The question of the functional meaning of such a REM sleep increase after learning remains open. Many authors (2,16,21,33,39) suggest that REM sleep per se plays an important role in the organization of memory functions and helps to transfer the information from a short-term memory to long-term memory. REM sleep deprivation performed during the critical period when the REM sleep requirement is especially high, keeps the information in a labile state and makes it sensitive to all additional stimuli, for instance to electrical shocks.
From this point of view, it is evident that learning of new forms of behavior takes place during wakefulness, and REM sleep only helps to retain these new forms of behavior in memory. This point of view is supported by separate reports that REM sleep deprivation sometimes does not influence the learning process, if the learning prior to the deprivation was performed with intensity and success and was concerned with such simple information that the full retention of this information was realized during wakefulness (16). However, if the solution of the task is realized during wakefulness and the new information is adopted before sleep, it is not easy to understand why the retention of unusual information requires REM sleep, while the retention of ordinary information does not. After the unusual task is resolved (during wakefulness) the body of information connected with this task must be transferred from the subjectively unusual to subjectively ordinary.
This is why there is reason to suggest that REM sleep plays only an indirect role in the process of retention by carrying out its main function -compensation of the renunciation of search and restoration of search activity. Renunciation of search decreases body resistance, disturbs adaptive behavior and interferes with all psychic functions, including memory, hindering the integration of new information.
A difficult task that requires unprepared learning may cause the renunciation of search more readily than a simple task.' During the first stage of the process of resolution of the difficult task, the mistakes predominate, which is a sign that the task is difficult. But Seligman's investigations (31) show very clearly that the negative experience may cause learned helplessness (renunciation of search), especially when the task is very important for the animal due to aversive stimulation. This explains why an intensification of the learning process, when the animal perceives a great number of failures within a restricted time, must cause a sharp increase in REM sleep. If the process of learning is less intense and takes longer, then the negative experience is interspersed with the ordinary experiences of normal life, which include the experience of success. This would account for a more gradual increase of REM sleep (33).
It is significant that the state of renunciation of search that is aroused by the stimulation of brain structure or by an aversive situation causes the same changes in sleep structure as during the unprepared learning: The latent period of REM sleep decreases and the REM sleep percentage increases (27). If renunciation of search is not repaired by REM sleep, this would prevent both the correct task decision and the fixing of this decision in memory. As a result, the renunciation of search becomes fixated and deep.
The above sequence would explain the increase of REM sleep percentage just prior to the achievement of mastery. The increase of REM sleep percentage would be useless, however, if the REM sleep per se is functionally deficient (26). In this situation, REM sleep may be most increased in animals that are unsuccessful in their efforts to adopt a new strategy of behavior. On the other hand, not all animals are inclined to renunciation of search, even during unprepared learning, which is why REM sleep is not always increased in this condition. Conversely, some animals are overinclined to learned helplessness and they can display the renunciation of search and the increased requirement of REM sleep even in the process of prepared learning. Many contradictions in experimental data may be resolved by taking into consideration the type of behavior during prepared and unprepared learning.
The locus coeruleus, being a key constituent of the noradrenergic system, is involved simultaneously in the organization of REM sleep and in the process of memory consolidation (39). Destruction of the locus coeruleus prevents the transformation from a labile memory trace to a stable memory trace. It was suggested (18) that due to its involvement in the process of regulation of the brain noradrenergic mechanisms, REM
sleep plays an important role in the process of adaptation to an enriched environment. After the destruction of the brain noradrenergic system, animals were unable to resolve difficult tasks and their REM sleep was not increased in the process of learning (14). By taking into consideration both the important role of the brain noradrenergic system in search behavior, and the role of REM sleep in the process of restoration of the activity of this system after renunciation of search (25), it is possible to understand the interconnections between brain noradrenergic system and REM sleep in the organization of memory. They both service memory due to their role in overcoming the renunciation of search.
It is necessary to stress that REM sleep is increased predominantly during unprepared learning; on the other hand exposure to an enriched environment usually is followed by an increase in slow-wave sleep (33).
It seems that the abundance of information is not a real cause of the increase of REM sleep, and the latter is involved not in information processing, but in the process of adaptation to distress.
REM SLEEP DEPRIVATION BEFORE LEARNING
The influence of REM sleep deprivation which precedes learning is not as definite as the influence of learning on subsequent sleep structure. REM sleep deprivation during 120 hours prevents subsequent learning of simple active avoidance (10,36), and a short REM deprivation can even stimulate active avoidance (19).
A long REM deprivation on the wooden platform destroys simple passive avoidance (36), while a short and more gentle REM deprivation does not cause these disturbances (5,15). The influence on memory of long-lasting REM sleep deprivation on the wooden platform is reminiscent of the effect of brain noradrenaline depletion (8).
If the animal receives 1-DOPA or amphetamine before the testing procedure, the effect of REM sleep deprivation on memory is diminished (37). Perhaps long-lasting REM deprivation destroys memory due to brain noradrenaline exhaustion in some brain regions during the distress caused by unprepared learning.
If REM deprivation is short, then search behavior is frustrated, but brain noradrenaline is not yet exhausted. The change from conditions on the wooden platform to the conditions of learning and testing is perceived by animals as a change from a stressful to a more comfortable situation, and may cause a rebound of search activity. A high level of motor activity and a moderate level of stress can even promote some forms of learning, for instance active avoidance. But, since search behavior after REM sleep deprivation is compensative, it can become chaotic and overwhelming -and that is why it can interfere with some tasks, for instance with a passive avoidance.
Conversely, if REM sleep deprivation is prolonged, a renunciation of search activity appears, perhaps as a result of brain catecholamine exhaustion, and the whole process of learning is destroyed.
Memory traces then become more sensitive to disturbing influences. Our conclusion is that REM sleep deprivation before learning creates a distress that can interfere with memory.
REM SLEEP DEPRIVATION AFTER TRAINING
In contrast to REM sleep deprivation before training, deprivation after training may be performed in definite periods of time and can be more selective, more restricted, and less stressful for animals As an indicator for the time for sleep deprivation, it is possible to use those periods during the day when, after the training procedure, REM sleep usually increases ["windows" for REM, according to (33)] Investigations performed according to this paradigm brought the same results. Retention is destroyed if REM deprivation coincides with REM "windows ". For example, REM deprivation delays the learning process when used in the first 90 minutes after training (14). When REM deprivation is performed following I.5 hours of sleep immediately after the training process, memory is not disturbed. Even a long REM sleep deprivation (during 10 hours) performed outside of the critical periods for REM sleep increase (REM sleep "windows") does not disturb the retention (12). The more intensive the complicated learning, the earlier the "window" opens - especially when the animal gives up during the training process. The first episode of REM sleep following the training process is the most important for memory, and its deprivation is the most detrimental to learning.
Because REM sleep deprivation is effective only if it is performed during the periods that coincide with the possible REM sleep increase, it is not surprising that this procedure is not effective after prepared learning (1,21,22,32) and after the simple tasks with a positive reinforcement (35).
After the first day of REM deprivation, a simple active avoidance may appear even sooner than in the control group, but after the more prolonged deprivation this type of avoidance is destroyed due to the disruption of behavior (19). REM sleep deprivation after training does not influence the results of training if the training process before REM deprivation is very intensive, or if a rest period of more than 24 hours takes place between a short REM deprivation and a testing procedure (16). If electric shock is used during the first hours of this rest period after REM deprivation, the retention is much less (6). It is significant that the rest period between REM deprivation and the test procedure may be free from REM sleep or any sleep at all, nevertheless, this rest period is favourable for subsequent retention (19).
The analysis of all these data gives a basis for the supposition that REM sleep deprivation influences the retention process indirectly. Agitated restless behavior that appears as a compensatory reaction after a short REM sleep deprivation on the wooden platform may facilitate simple active avoidance, but disturbs passive avoidance. Long-lasting REM sleep deprivation that leads to panicky behavior or renunciation or search, interferes with all forms of learning.
The increased REM sleep after unprepared learning must compensate for the unproductive emotional state that corresponds to the renunciation of search that occurs during unprepared learning. Thus, REM sleep deprivation hinders the learning process by stabilizing the emotional tension that has a negative influence on retention, as well as on all other physiological functions.
During prepared learning, REM sleep deprivation has neither a direct or an indirect influence on memory per se. But stress that is apparent through REM sleep deprivation may prevent the reproduction of the previously assimilated information: either exaggerated search activity (after a short deprivation), or renunciation of search (after a long deprivation) may interfere with integrated purposeful behavior That is why the ability to resolve tasks in the proper manner can be restored not only after sleep, but even after rest without sleep.
There are some additional arguments against the idea of the direct participation of REM sleep in the consolidation mechanisms:
1 During the active search behavior in the stressful situation, the consolidation of the new experience is essential for the organism, but REM sleep during this form of behavior does not increase, and even decreases (25).
2 The REM sleep percentage increases due to neuroleptic treatment (17) but does not have a beneficial effect on memory.
3 Activating drugs, like amphetamine, have a beneficial effect on memory, but at the same time have a tendency to decrease REM sleep.
All these additional arguments reinforce our suggestion that REM sleep has only an indirect influence on memory due to restoration of search activity.
ACKNOWLEDGEMENT
My best thanks to Prof R Greenberg and Dr J Fisch for help in manuscript preparation
REFERENCES
1 Dodge A , Beatty, W Sleep deprivation does not affect spatial memory in rats. Bull Psychol Soc 16408-409,1980
2 Drucker- Colin, R P. Neuroproteins, brain excitability and REM sleep. In Fishbein, W , ed. Sleep dreams and memory. Lancaster MTP Press, 1981 73-94
3 Everson, C , Bergmann, B, Fang, V S , Leutch, C A , Obermeyer, W , Refetoff, S , Schoeller, D A , Rechtschaffen, A Physiological and biochemical effects of total sleep deprivation m the rat. In Chase M H , McGuinty D J , G Crane, eds. Sleep res. 15:216, 1986
4 Everson, C A , Kushida, C A , Bergmann, B M , Gilhland, M A , Pilcher, J J , Fang, V S , Rechtschaffen, A Recovery from chronic sleep deprivation in the rat. In: Chase, M H , McGuinty, D J , O'Connor, C , eds. Sleep res. 16 :521, 1987
5 Fishbein, W. Interference with conversion of memory from short term to long term storage by partial sleep deprivation. Commun. Behav. Biol. 5, 171-175, 1970
6 Fishbein, W. Disruptive effects of rapid eye movement sleep deprivation on long term memory. Physiol Behav 6279-282,1971.
7 Fishbein, W , Kastamotis, C , Chattman, I Paradoxical sleep. Prolonged augmentation following learning. Brain Res. 79 61-75, 1974
8 Gorehck, D A , Bozewicz T R , Bndger, W H The role of catecholamines in animal learning and memory. In Fnedhof, A I , ed Catecholamines and behavior, vol 2. New York Plenum Press, 1976 1-30
9 Greenberg, R , Pearlman, C. Cutting the REM nerve. An approach to the adaptive role of REM sleep. Perspect. Biol. Med. 17 513-521, 1974
10 Hartmann, E , Stern, W C Desinchronized sleep deprivation. Learning deficit and its reversal by increased catecholamines. Physiol Behav. 8: 585-587, 1972
11 Jouvet, I , Vimont, P , Lelorme, F , Jouvet, M Etude de la privation selective de la phase paradoxale de sommeil chez le chat. C R Soc Biol 158 756-759, 1964
12 Kitahama, , Valatx, J D , Jouvet, M Apprentissage d'un labyrinthe en 4 chez deux souches de souris. Effects de la privation instrumentale et pharmacologique du sommeil. Brain Res 108 75-86, 1976
13 Kovalzon, V M , Tsbulski, V L REM sleep deprivation, stress and emotional behavior in rats. Behav Brain Res 14235-245,1984
14 Leconte, P , Hennevin, E , Block V Analyse des effects d'un apprentissage et de son niveau d'acquisition sur le sommeil paradoxal consecutif. Brain Res 49367-379,1973
15. Maho, C. Concomitants physiologiques de 1'apprentissage et sommeil paradoxal consecutif chez le chat. Physiol. Behav. 18:431-438; 1977.
16. McGrath, M. J.; Cohen, I. B. REM sleep facilitation of adaptive making behavior: A review of the literature. Psychol. Bull. 85: 24-57; 1978.
17. Mendelson, W. B.; Gillin, J. C.; Wyatt, R. J. Human sleep and its disorders. New York: Plenum Press; 1977.
18. Mirmiran, M.; Wylings, H. B. . The environmental enrichment effect upon cortical growth is neutralized by concomitant pharmacological suppression of active sleep in female rats. Brain Res. 261:331-334; 1983.
19. Oniani, T. N.; Lortkipanidze, N. D. Effect of paradoxical sleep deprivation on the learning and memory. In: Oniani, T. N., ed. Neurophysiology of motivation, memory and sleep-wakefulness cycle, vol. 4. Tbilisi: Metsniereba; 1985:214-234.
20. Pagel, I.; Pegram, V.; Vaughan, S.; Donaldson, P.; Bridge, W. The relationship of REM sleep with learning and memory in mice. Behav. Biol. 9:383-388; 1973.
21. Pearlman, C. REM sleep and information processing: Evidence from animal studies. Neurosci. Biobehav. Rev. 3:57-68; 1979.
22. Pearlman, C.; Becker, M. Brief post-trial REM sleep deprivation impairs discrimination learning in rats. Physiol. Psychol. 1:373-376; 1973.
23. Pearlman, C.; Becker, M. Retroactive impairment of cooperation learning by imipramine and chlordiazepoxide in rats. Psycho-pharmacologia 42:63-66; 1975.
24. Rechtschaffen, A.; Gilliland, M. A.; Bergmann, B. .; Winter, J. B. Physiological correlates of prolonged sleep deprivation in rats. Science 221:182-184; 1983.
25. Rotenberg, V. S. Search activity in the context of psychosomatic disturbances of brain monoamines and REM sleep function. Pavlov. J. Biol. Sci. 19:1-15; 1984.
26. Rotenberg, V. S. Functional deficiency of REM sleep and its role in the pathogenesis of neurotic and psychosomatic disturbances. Pavlov. J. Biol. Sci. 23:1-3; 1988.
27. Rotenberg, V. S.; Arshavsky, V. V. REM sleep, stress and search activity. Wak. Sleep. 3:235-244; 1979.
28. Rotenberg, V. S.; Arshavsky, V. V. Search activity and adaptation. Moscow: Nauka; 1984.
29. Rotenberg, V. S.; Kovalzon, V. V.; Tsibulsky, V. L. Paradoxical sleep-protection from stress? Sci. USSR 2:45-51; 1986.
30. Seligman, M. E. P. On the generality of the laws of learning. Psychol. Rev. 77:406-418; 1970.
31. Seligman, M. E. P. Helplessness. On depression, development and death. San Francisco: Freeman; 1975.
32. Sloan, M. A. The effects of rapid eye movement (REM) sleep on maze learning and aggression in the albino rat. J. Psychiat. Res. 9:101-111; 1972.
33. Smith, C. Sleep states and learning. A review of the animal literature. Neurosci. Biobehav. Rev. 9:157-168; 1985.
34. Smith, C.; Kitahama, K.; Valatx, J. L.; Jouvet, M. Increased paradoxical sleep in mice during acquisition of A shock avoidance task. Brain Res. 77:221-230; 1974.
35. Smith, C.; Lowe, D.; Smith, M. J. Increases in paradoxical and slow wave sleep during the acquisition of an appetitive task in rats. Physiol. Psychol. 5:364-372; 1977.
36. Stern, W. C. Acquisition impairments following rapid eye movements sleep deprivation in rats. Physiol. Behav. 7:345-352; 1971.
37. Stern, W. C. Morgane, P. I. Theoretical view of REM sleep function: Maintenance of catecholamine systems in the central nervous system. Behav. Biol. 11: 1-32; 1974.
38. Ursin, H. The instrumental effects of emotional behavior. In: Bateston, J.; Klopfer, L., eds. Perspectives in ethology. Plenum Press: London; 1985:45-62.
39. Zornetzer, S. F. The nucleus locus coeruleus and its involvement in memory processing. In: Fishbein, M., ed. Sleep, dreams and memory. Lancaster: MTP Press; 1981:47-72.
