COMMENTARYMemory, sleep and the evolution of mechanisms of synaptic efficacy maintenance
Introduction
Roffwarg et al.[243]conjectured that spontaneous, repetitive activations of circuitry in the CNS during rapid-eye-movement (REM) sleep in the human embryo facilitate circuit development and maintenance. They suggested that such activations during REM sleep maintain circuitry throughout life. This concept has been highly fruitful, with further contributions and elaborations by many investigators, and with increasing experimental support.
The concept provided the basis for a paradigm of dynamic stabilization (DS) of neural circuitry (reviewed in [145]). Circuits storing inherited information, frequently referred to as hard-wired, e.g., circuits for autonomic functions, are referred to as phylogenetic memories, while those storing experiential information, such as learned responses, are referred to as ontogenetic memories. According to the paradigm of DS, synaptic efficacy in circuits storing ontogenetic and phylogenetic memories is enhanced, and the enhancements are maintained at length, by both frequent functional use and activations induced by self-generated, spontaneous oscillatory brain activity. Since the spontaneous activations maintain enhanced synaptic efficacy, but usually do not trigger circuit functions (often inhibited by temporarily increased activation thresholds), they are referred to as being non-utilitarian. Unless functional DS is referred to specifically, non-utilitarian DS is implied.
Pursuing the logical implications and consequences of long-term maintenance of synaptic efficacy by DS has led to the formulation of a chain of causal evolutionary neural links, from the development of activity-dependent plasticity of synapses in simple metazoans to the neural adaptations represented by REM sleep in mammals and birds. In the course of this pursuit, causal factors for some of the enigmatic phasic events of REM sleep, and a basis for the genesis and properties of dreams, may have been identified. These formulations are elaborated here, taking into account the pertinent implications of the existence of unihemispheric sleep in birds and cetaceans, the dominating influences of visual input on brain activity, and the evolution of parcellation of the processing of visual input between the retina and central brain regions.
It has long been suspected that self-generated, spontaneous brain oscillations play a fundamental role in brain activity.[170]The major function of many of these oscillations during sleep may be the DS of synapses in infrequently used circuitry. Examples of spontaneous oscillatory activity that may accomplish DS are: the theta rhythm of REM sleep, which may enhance synaptic efficacy in the hippocampus;145, 219, 222, 244field irregular sharp spikes (SPWs) of non-rapid-eye-movement (NREM) sleep, which may potentiate synaptic efficacy in targets of the hippocampus;31, 42and continuous synaptic bombardment from the intrinsic cortical networks that generate the slow sleep rhythm during NREM sleep, which may reinforce synapses in cortical association neurons.294, 300
Spontaneous brain oscillations also may maintain memories over the short term. In the proposal of Lisman and Idiart,[169]they would provide timing signals that control serial processing of multiple short-term memories in neural networks, each memory stored in a different high-frequency (∼40 Hz) nested subcycle of a network's low-frequency oscillations. Firing of network neurons could be sustained by a neuromodulator-induced transient increase in membrane excitability (short-term activity-dependent efficacy enhancement) that is refreshed on each cycle of network oscillations (dynamically stabilized).
Oscillating and resonating neurons and networks are able to project their rhythms and generate synchronous firing in large neuronal populations.[170]Neurons in neocortical layer 5, in particular, can generate highly synchronous activity and impose it on other neurons in the cortex.[45]Discharges can be synchronized on a millisecond time-scale even when the neurons are widely distributed.[273]Reinforcing activations of DS apparently have their origin in these oscillations. Oscillatory neurons are widely distributed, being found in the neocortex, entorhinal cortex, thalamus, hypothalamus, inferior olivary nucleus, olfactory system and brainstem. The major part of thalamocortical connectivity is devoted to self-generated oscillatory activity (largely re-entrant[64]or reverberating) that produces global oscillatory brain states.[171]
One source of selective pressure for the evolutionary origin of neurons with oscillatory firing capacities, or for their development from primitive non-spiking oscillator or pacemaker cells, may have been the need for spontaneous activations to enhance the efficacy of synapses in infrequently used circuits. In support of this conjecture, single neurons with endogenous oscillatory properties, with multiple conductances contributing to their rhythmic bursts, and with capacities for modulated switching between a variety of activity patterns, are of wide taxonomic occurrence.83, 98, 298
Oscillator neurons and pacemakers occur even in organisms with the simplest nervous systems, for example, sea anemones.[287]That a need for self-generated neural activations is probably of very ancient origin is suggested by the finding that “coordinated spontaneous activity…is a fundamental feature of the most primitive nervous systems”.[28]One expects the primordial basis for long-term memory to exist in some simple nervous systems, and this coordinated spontaneous activity appears to be the underlying mechanism.
Section snippets
Methods of maintaining synaptic efficacy
Possessing only non-plastic synapses, an animal merely would respond passively to external stimuli. Responses might vary with physiological state, but behavioral patterns would be limited essentially to successions of fixed reflexive actions. The animal would contribute nothing adaptive to the behavior, with no evidence of an active drive to achieve a goal.[220]The first evolutionary step beyond fixed reflexive behavior doubtless involved the development of synapses with activity-dependent
Neural activity in sleeping and waking brains
Before embarking on considerations of the selective pressures for sleep and multiple sleep states, and their roles in synaptic efficacy enhancement and maintenance, it is desirable to review briefly some aspects of the ontogeny of sleep. Because DS appears to be associated with spontaneous, endogenous brain waves, it also is desirable to review briefly the origins and properties of the principal waves.
Origin of the primitive sleep state
As non-sleeping animals acquired increasingly complex brains, behavioral repertoires and visual competencies, together with ever enlarging stores of ontogenetic and phylogenetic memories, increasing amounts of DS of memory circuits were required during their periods of restful waking. Like sleep, restful waking (or inactivity other than sleep) in present-day vertebrates is not an exclusive function of external circadian rhythms, but is determined by additional internal regulatory mechanisms.
Evolution of non-rapid-eye-movement and rapid-eye-movement sleep
Most lines of evidence suggest that NREM sleep is, or most resembles, the primitive sleep state.100, 127, 191, 347Sleeping ectotherms, such as reptiles, do not engage in REM sleep,43, 100which exists only in birds and mammals. REM sleep correlates with a highly developed forebrain in both groups, reflecting highly developed dorsal and dorsolateral ventricular ridges and wulst in birds, and a highly developed cortex in mammals.108, 127, 142
Both NREM and REM sleep exist in all studied marsupial
Sleep states and circuit consolidation and reinforcement
As noted above, fetal evidence indicates that processes involved in circuit development, maturation, fine-tuning and prenatal maintenance in mammals and birds occur largely during REM sleep. A predominant role of REM sleep in motor circuit reinforcement is indicated by various aspects of brain activity and sleep pathology, the fetal evidence, concordance of the theta rhythm with movement-related activities and other correlations.
We also have noted studies suggesting that the theta rhythm of REM
Inconsequential phasic events during adult rapid-eye-movement sleep
Numerous very minor muscle movements occur during REM sleep in adult mammals and birds. These include REMs (in unison), fine digit movements, and movements of the tongue, small facial and middle ear muscles, and muscles of the nape of the neck,86, 175, 228resulting from “strikingly potent motor excitation drives” that phasically overcome motoneuronal inhibition: [41]“…the underlying motor control landscape is actually ravished by storms of inhibition and brief whirlwinds of excitation directed
Criteria
In view of these considerations, emphasizing the roles of perception of the environment and processing of complex visual information, it can be suggested that the most reliable criteria for distinguishing sleep from restful waking in ectothermic vertebrates are elevated thresholds for sensory stimuli, occlusion of the pupillary apertures, behavioral quiescence, characteristic postures and rapid arousability. The presence of muscular hypotony would be indicative but not necessarily determinate,
Hibernation and deep torpor
It was noted earlier that an absence of EEG activity reflects death of the brain, except when it is under the influence of extreme anesthesia or severe hypothermia.[229]The state of deep torpor of the Arctic ground squirrel (Spermophilus parryii) involves severe hypothermia, with an absence of EEG activity for many days. Since, from the present perspectives, EEG components play roles in the DS of memories, parameters of hibernation and torpor for this animal have the potential to provide unique
Dream genesis and significance
Within the context of the preceding analyses, an effort can be made to elucidate the origin, content, significance and evanescent nature of dreams, the state of consciousness (or altered consciousness)[283]in which contact with reality is minimal.[76]The view championed by Hobson and McCarley,[111]and also espoused by Greenberg[91]and Antrobus,8, 9is that dreaming is a byproduct of, and tightly linked to, the mental activities that normally occur during REM sleep. The essential psychological
Acknowledgements
This research was supported by grants from the UCLA Faculty Research Committee. I thank Marisa G. Kavanau for assistance with the manuscript.
References (347)
- et al.
Energy conservation and sleep
Behav. Brain Res.
(1995) Two stage model of memory trace formation: a role for “noisy” brain states
Neuroscience
(1989)The functions of sleep: further analysis
Physiol. Behav.
(1991)An integrative analysis to sleep functions
Behav. Brain Res.
(1995)- et al.
Animal sleep: a review of sleep duration across phylogeny
Neurosci. Biobehav. Rev.
(1984) - et al.
The function of dendritic spines: a review of theoretical issues
Behav. neural. Biol.
(1985) - et al.
Warming up for sleep?—Ground squirrels sleep during arousals from hibernation
Neurosci. Lett.
(1991) - et al.
Sleep, brain activation and cognition
Physiol. Behav.
(1990) Functional consequences of sustained sleep deprivation in the rat
Behav. Brain Res.
(1995)Synaptic plasticity: on the trail of the retrograde messenger
Trends Neurosci.
(1992)
The EEG and behavioral continuum of the crocodilian, Caiman sclerops
Electroenceph. clin. Neurophysiol.
Neuronal polarity: targeting of microtubule components into axons and dendrites
Trends biochem. Sci.
The effects of epileptic cortical activity on the development of callosal projections
Devl Brain Res.
Evolution of the mammalian middle ear
J. Morphol.
Sleep in mammals: ecological and constitutional correlates
Science
Electrophysiological studies of the echidna, Tachyglossus aculeatus. I. Waking and sleep
Archs ital. Biol.
A synthesis of sleep in wild birds
Behaviour
Dreaming: cognitive processes during cortical activation and high afferent thresholds
Psychol. Rev.
Serotonin-mediated endocytosis of apCAM: an early step of learning-related synaptic growth in Aplysia
Science
Anatomic basis of cognitive–emotional interactions in the primate prefrontal cortex
Neurosci. Biobehav. Rev.
Biochemistry of information storage in the nervous system
Science
The tuatara: why is it a lone survivor?
Sci. Monthly
Does trampolining and anaerobic physical fitness affect sleep?
Percept. Motor Skills
Motor learning and ultradian sleep cycle: an electroencephalographic study of trampoliners
Percept. Motor Skills
Retinal projections in the bowfin, Amia calva, cytoarchitectonic and experimental analysis
Brain Behav. Evol.
High-frequency network oscillation in the hippocampus
Science
Selective stabilization of developing synapses as mechanism for the specification of neuronal networks
Nature
The atonia and myoclonia of active (REM) sleep
A. Rev. Psychol.
Selective activation of deep layer (V–VI) retrohippocampal cortical neurons during hippocampal sharp waves in the behaving rat
J. Neurosci.
Cited by (99)
The Role of Sleep in Learning Placebo Effects
2018, International Review of NeurobiologyCitation Excerpt :Sleep has been proposed as serving an energy-saving function (Berger & Phillips, 1995; Webb, 1988), thermoregulation (Rechtschaffen & Bergmann, 1995), the repairing of cell tissue (Oswald, 1980), cardiovascular and metabolic regulation (Knutson, Spiegel, Penev, & Van Cauter, 2007; Van Cauter, Eve, Tasali, & Leproult, 2008), or immune functions (Lange, Dimitrov, & Born, 2010). However, the decreased responsiveness to the external environment during sleep strongly speaks for the notion that sleep is essential for brain functions (Hobson, 2005; Kavanau, 1997). Sleep is important for memory consolidation.
The energy allocation function of sleep: A unifying theory of sleep, torpor, and continuous wakefulness
2014, Neuroscience and Biobehavioral ReviewsSleep and synaptic changes
2013, Current Opinion in NeurobiologyCitation Excerpt :Most of this energy is used to sustain the various components of synaptic activity, from the release of neurotransmitter vesicles and their recycling to the restoration of Na+ and K+ gradients following postsynaptic potentials [4,5]. It is not surprising, therefore, that many hypotheses about the functions of sleep have focused on its effects on synapses, suggesting that sleep may allow the recovery of the synapses that underwent plastic changes during wake [6], the further strengthening of the synapses activated by learning during wake [7], the stimulation of the synapses that were not sufficiently activated during wake [8–10], or the generalized downregulation of synaptic strength after wake-induced synaptic potentiation [11]. Irrespective of the specific hypothesis, the underlying assumption is that sleep, or lack of it, will affect how synapses work, with significant consequences for brain functions.
Molecular neurobiology of sleep
2011, Handbook of Clinical NeurologyAn Overview of Normal Sleep
2009, Sleep Disorders Medicine: Basic Science, Technical Considerations, and Clinical Aspects: Third Edition