Comment on 'Lack of evidence for associative learning in pea plants'.

In 2016 we reported evidence for associative learning in plants (Gagliano et al., 2016). In view of the far-reaching implications of this finding we welcome the attempt made by Markel to replicate our study (Markel, 2020). However, as we discuss here, the protocol employed by Markel was unsuitable for testing for associative learning.

Three decades of continuous wrist-activity recording: analysis of sleep duration.

Motor activity recording by a wrist-worn device is a common method to monitor the rest-activity cycle. The first author wore an actimeter continuously for more than three decades, starting in 1982 at the age of 43.5 years. Until November 2006 analysis was performed on a 15-min time base, and subsequently on a 2-min time base. The timing of night-time sleep was determined from the cessation and re-occurrence of daytime-level activity. Sleep duration declined from an initial 6.8 to 6 h in 2004. The declining trend was reversed upon retirement, whereas the variance of sleep duration declined throughout the recording period. Before retirement, a dominant 7-day rhythm of sleep duration as well as an annual periodicity was revealed by spectral analysis. These variations were attenuated or vanished during the years after retirement. We demonstrate the feasibility of continuous long-term motor activity recordings to study age-related variations of the rest-activity cycle. Here we show that the embeddedness in a professional environment imparts a temporal structure to sleep duration.

Learning by Association in Plants.

In complex and ever-changing environments, resources such as food are often scarce and unevenly distributed in space and time. Therefore, utilizing external cues to locate and remember high-quality sources allows more efficient foraging, thus increasing chances for survival. Associations between environmental cues and food are readily formed because of the tangible benefits they confer. While examples of the key role they play in shaping foraging behaviours are widespread in the animal world, the possibility that plants are also able to acquire learned associations to guide their foraging behaviour has never been demonstrated. Here we show that this type of learning occurs in the garden pea, Pisum sativum. By using a Y-maze task, we show that the position of a neutral cue, predicting the location of a light source, affected the direction of plant growth. This learned behaviour prevailed over innate phototropism. Notably, learning was successful only when it occurred during the subjective day, suggesting that behavioural performance is regulated by metabolic demands. Our results show that associative learning is an essential component of plant behaviour. We conclude that associative learning represents a universal adaptive mechanism shared by both animals and plants.

The two-process model of sleep regulation: a reappraisal.

In the last three decades the two-process model of sleep regulation has served as a major conceptual framework in sleep research. It has been applied widely in studies on fatigue and performance and to dissect individual differences in sleep regulation. The model posits that a homeostatic process (Process S) interacts with a process controlled by the circadian pacemaker (Process C), with time-courses derived from physiological and behavioural variables. The model simulates successfully the timing and intensity of sleep in diverse experimental protocols. Electrophysiological recordings from the suprachiasmatic nuclei (SCN) suggest that S and C interact continuously. Oscillators outside the SCN that are linked to energy metabolism are evident in SCN-lesioned arrhythmic animals subjected to restricted feeding or methamphetamine administration, as well as in human subjects during internal desynchronization. In intact animals these peripheral oscillators may dissociate from the central pacemaker rhythm. A sleep/fast and wake/feed phase segregate antagonistic anabolic and catabolic metabolic processes in peripheral tissues. A deficiency of Process S was proposed to account for both depressive sleep disturbances and the antidepressant effect of sleep deprivation. The model supported the development of novel non-pharmacological treatment paradigms in psychiatry, based on manipulating circadian phase, sleep and light exposure. In conclusion, the model remains conceptually useful for promoting the integration of sleep and circadian rhythm research. Sleep appears to have not only a short-term, use-dependent function; it also serves to enforce rest and fasting, thereby supporting the optimization of metabolic processes at the appropriate phase of the 24-h cycle.

Metabolic response of the cerebral cortex following gentle sleep deprivation and modafinil administration.

STUDY OBJECTIVES: The main energy reserve of the brain is glycogen, which is almost exclusively localized in astrocytes. We previously reported that cerebral expression of certain genes related to glycogen metabolism changed following instrumental sleep deprivation in mice. Here, we extended our investigations to another set of genes related to glycogen and glucose metabolism. We also compared the effect of instrumentally and pharmacologically induced prolonged wakefulness, followed (or not) by 3 hours of sleep recovery, on the expression of genes related to brain energy metabolism. DESIGN: Sleep deprivation for 6-7 hours. SETTING: Animal sleep research laboratory. PARTICIPANTS: Adults OF1 mice. INTERVENTIONS: Wakefulness was maintained by "gentle sleep deprivation" method (GSD) or by administration of the wakefulness-promoting drug modafinil (MOD) (200 mg/kg i.p.). MEASUREMENTS AND RESULTS: Levels of mRNAs encoding proteins related to energy metabolism were measured by quantitative real-time PCR in the cerebral cortex. The mRNAs encoding protein targeting to glycogen (PTG) and the glial glucose transporter were significantly increased following both procedures used to prolong wakefulness. Glycogenin mRNA levels were increased only after GSD, while neuronal glucose transporter mRNA only after MOD. These effects were reversed after sleep recovery. A significant enhancement of glycogen synthase activity without any changes in glycogen levels was observed in both conditions. CONCLUSIONS: These results indicate the existence of a metabolic adaptation of astrocytes aimed at maintaining brain energy homeostasis during the sleep-wake cycle.

Functional EEG topography in sleep and waking: state-dependent and state-independent features.

Power spectra in the non-rapid eye movement sleep (NREMS) electroencephalogram (EEG) have been shown to exhibit frequency-specific topographic features that may point to functional differences in brain regions. Here, we extend the analysis to rapid eye movement sleep (REMS) and waking (W) to determine the extent to which EEG topography is determined by state under two different levels of sleep pressure. Multichannel EEG recordings were obtained from young men during a baseline night, a 40-h waking period, and a recovery night. Sleep deprivation enhanced EEG power in the low-frequency range (1-8 Hz) in all three vigilance states. In NREMS, the effect was largest in the delta band, in W, in the theta band, while in REMS, there was a peak in both the delta and the theta band. The response of REMS to prolonged waking and its pattern of EEG topography was intermediate between NREMS and W. Cluster analysis revealed a major topographic segregation into three frequency bands (1-8 Hz, 9-15 Hz, 16-24 Hz), which was largely independent of state and sleep pressure. To assess individual topographic traits within each state, the differences between pairs of power maps were compared within (i.e., for baseline and recovery) and between individuals (i.e., separately for baseline and recovery). A high degree of intraindividual correspondence of the power maps was observed. The frequency-specific clustering of power maps suggests that distinct generators underlie EEG frequency bands. Although EEG power is modulated by state and sleep pressure, basic topographic features appear to be state-independent.

Non-rapid eye movement sleep with low muscle tone as a marker of rapid eye movement sleep regulation.

BACKGROUND: It was recently reported that epochs of non-REM sleep (NREMS) with low muscle tone represent a partial correlate of REM sleep (REMS). To further investigate this phenomenon, episodes of restricted night-time sleep (23:00-03.00 h) and subsequent morning sleep (10:00-13:00 h) were analysed. RESULTS: Epochs of NREMS with low muscle tone (NLMT) were identified. Their frequency was higher in morning sleep than in night sleep. At night, the latency to the first occurrence of NLMT showed a bimodal distribution with modes at sleep onset and close to REMS onset. In morning sleep, the distribution was unimodal with the mode at sleep onset. An episode of NLMT at sleep onset occurred in 35.5% of the night sleep episodes and in 60.9% of the morning sleep episodes without sleep onset REMS (SOREMS). Also SOREMS occurred predominantly in morning sleep. REMS episodes were longer and NREMS episodes shorter in morning sleep than in night sleep, whereas cycle duration did not differ. Simulating the time course of slow-wave activity revealed a close correspondence between empirical and computed values for night sleep, and some discrepancies for morning sleep. CONCLUSION: The results provide further evidence that NREMS with low muscle tone is a marker of REMS regulation. NLMT at sleep onset may represent an early manifestation of REMS.

Sleep inertia: performance changes after sleep, rest and active waking.

Napping benefits and sustains subsequent performance. Prophylactic naps have been recommended as a means to maintain performance during extended wakefulness, as required during shiftwork. However, napping may cause short-term performance impairments, because awakening from sleep is followed by sleep inertia, a period of hypovigilance and impaired cognitive and behavioral performance. We investigated sleep inertia after an afternoon nap. Healthy 18-28 year-olds (n=50, not sleep deprived) were assigned to sleep, active wake or rest groups for a 2-h experimental phase with polysomnography starting either at 14:00 or 16:00 for half of each group. Before (baseline, 12:30 or 14:30) and in five sessions during the hour after the experimental phase (16:00-17:00 or 18:00-19:00), subjects completed an addition task, an auditory reaction time task, and the Stanford Sleepiness Scale. In session one, addition speed in the sleep group was reduced compared with baseline and with active wake controls, whereas calculation accuracy did not change. Addition speed in the sleep and rest groups increased substantially from session one to session two and reached a level similar to that of the active wake group by the fifth session. In the first session, auditory reaction speed of the sleep group was reduced compared with baseline and with rest controls but did not differ from the active wake group. The slowest reaction times showed significant recovery after 20 min. The groups reported similar increases in subjective sleepiness after the experimental period. These findings provide evidence for performance slowing and recovery during the hour following a 2-h nap opportunity. They highlight the importance of employing multiple control groups and various objective and subjective measures to assess sleep inertia.

Human central auditory plasticity associated with tone sequence learning.

We investigated learning-related changes in amplitude, scalp topography, and source localization of the mismatch negativity (MMN), a neurophysiological response correlated with auditory discrimination ability. Participants (n = 32) underwent two EEG recordings while they watched silent films and ignored auditory stimuli. Stimuli were a standard (probability = 85%) and two deviant (probability = 7.5% each for high [HD] and low [LD]) eight-tone sequences that differed in the frequency of one tone. Between recordings, subjects practiced discriminating the HD or LD from the standard for 6 min. The amplitude of the LD MMN increased significantly across recordings in both groups, whereas the amplitude of the HD MMN did not. The LD was easier to discriminate than was the HD. Thus, practicing either discrimination increased the MMN for the easier discrimination. Learning and changes in the LD MMN amplitude were highly correlated. Source localizations of event-related potentials (ERPs) to all stimuli revealed bilateral sources in superior temporal regions. Compared with the standard ERP, the LD ERP revealed a stronger source in the left superior temporal region in both recordings, whereas the right-sided source became stronger after learning. Consistent with prior studies of auditory plasticity in animals and humans, tone sequence learning induced rapid neurophysiological plasticity in the human central auditory system. The results also suggest that there is asymmetric hemispheric involvement in tone sequence discrimination learning and that discrimination difficulty influences the time course of learning-related neurophysiological changes.

Development of the nocturnal sleep electroencephalogram in human infants.

The development of nocturnal sleep and the sleep electroencephalogram (EEG) was investigated in a longitudinal study during infancy. All-night polysomnographic recordings were obtained at home at 2 wk and at 2, 4, 6, and 9 mo after birth (analysis of 7 infants). Total sleep time and the percentage of quiet sleep or non-rapid eye movement sleep (QS/NREMS) increased with age, whereas the percentage of active sleep or rapid eye movement sleep (AS/REMS) decreased. Spectral power of the sleep EEG was higher in QS/NREMS than in AS/REMS over a large part of the 0.75- to 25-Hz frequency range. In both QS/NREMS and AS/REMS, EEG power increased with age in the frequency range <10 Hz and >17 Hz. The largest rise occurred between 2 and 6 mo. A salient feature of the QS/NREMS spectrum was the emergence of a peak in the sigma band (12-14 Hz) at 2 mo that corresponded to the appearance of sleep spindles. Between 2 and 9 mo, low-frequency delta activity (0.75-1.75 Hz) showed an alternating pattern with a high level occurring in every other QS/NREMS episode. At 6 mo, sigma activity showed a similar pattern. In contrast, theta activity (6.5-9 Hz) exhibited a monotonic decline over consecutive QS/NREMS episodes, a trend that at 9 mo could be closely approximated by an exponential function. The results suggest that 1) EEG markers of sleep homeostasis appear in the first postnatal months, and 2) sleep homeostasis goes through a period of maturation. Theta activity and not delta activity seems to reflect the dissipation of sleep propensity during infancy.

Radio frequency electromagnetic field exposure in humans: Estimation of SAR distribution in the brain, effects on sleep and heart rate.

In two previous studies we demonstrated that radiofrequency electromagnetic fields (RF EMF) similar to those emitted by digital radiotelephone handsets affect brain physiology of healthy young subjects exposed to RF EMF (900 MHz; spatial peak specific absorption rate [SAR] 1 W/kg) either during sleep or during the waking period preceding sleep. In the first experiment, subjects were exposed intermittently during an 8 h nighttime sleep episode and in the second experiment, unilaterally for 30 min prior to a 3 h daytime sleep episode. Here we report an extended analysis of the two studies as well as the detailed dosimetry of the brain areas, including the assessment of the exposure variability and uncertainties. The latter enabled a more in depth analysis and discussion of the findings. Compared to the control condition with sham exposure, spectral power of the non-rapid eye movement sleep electroencephalogram (EEG) was initially increased in the 9-14 Hz range in both experiments. No topographical differences with respect to the effect of RF EMF exposure were observed in the two experiments. Even unilateral exposure during waking induced a similar effect in both hemispheres. Exposure during sleep reduced waking after sleep onset and affected heart rate variability. Exposure prior to sleep reduced heart rate during waking and stage 1 sleep. The lack of asymmetries in the effects on sleep EEG, independent of bi- or unilateral exposure of the cortex, may indicate involvement of subcortical bilateral projections to the cortex in the generation of brain function changes, especially since the exposure of the thalamus was similar in both experiments (approx. 0.1 W/kg).

Mathematical models of sleep regulation.

The level of EEG slow-wave activity (SWA) is determined by the duration of prior sleep and waking. SWA is a marker of nonREM sleep intensity and may serve as an indicator of sleep homeostasis. The two-process model of sleep regulation posits the interaction of the homeostatic Process S and the circadian Process C. Also models of neurobehavioral functions (three-process model; interactive models of alertness and cognitive throughput) are based on the concept of an interaction between homeostatic and circadian factors. Whether the interaction is linear or non-linear is still unresolved. Models may serve as a guiding principle for specifying the relationship between processes occurring at the macroscopic and microscopic level of analysis.

Interhemispheric sleep EEG asymmetry in the rat is enhanced by sleep deprivation.

Vigilance state-related topographic variations of electroencephalographic (EEG) activity have been reported in humans and animals. To investigate their possible functional significance, the cortical EEG of the rat was recorded from frontal and parietal derivations in both hemispheres. Records were obtained for a 24-h baseline day, 6-h sleep deprivation (SD), and subsequent 18-h recovery. During the baseline 12-h light period, the main sleep period of the rat, low-frequency (<7.0 Hz) power in the non-rapid eye-movement (NREM) sleep EEG declined progressively. Left-hemispheric predominance of low-frequency power at the parietal derivations was observed at the beginning of the light period when sleep pressure is high due to preceding spontaneous waking. The left-hemispheric dominance changed to a right-hemispheric dominance in the course of the 12-h rest-phase when sleep pressure dissipated. During recovery from SD, both low-frequency power and parietal left-hemispheric predominance were enhanced. The increase in low-frequency power in NREM sleep observed after SD at the frontal site was larger than at the parietal site. However, frontally no interhemispheric differences were present. In REM sleep, power in the theta band (5.25-8.0 Hz) exhibited a right-hemispheric predominance. In contrast to NREM sleep, the hemispheric asymmetry showed no trend during baseline and was not affected by SD. Use-dependent local changes may underlie the regional differences in the low-frequency NREM sleep EEG within and between hemispheres. The different interhemispheric asymmetries in NREM and REM sleep suggest that the two sleep states may subserve different functions in the brain.

Sleep deprivation modulates brain mRNAs encoding genes of glycogen metabolism.

Replenishment of brain glycogen stores depleted during waking has been suggested to constitute one of the functions of sleep [Benington, J. H. & Heller H. C. (1995) Prog. Neurobiol., 45, 347]. We have tested the hypothesis that the level of expression of enzymes involved in glycogen metabolism could undergo variations throughout the sleep-waking or rest-activity cycle, and after 6 h of 'gentle' total sleep deprivation in mice. Specifically, we determined the variations in mRNAs coding for protein targeting to glycogen (PTG), glycogen synthase and glycogen phosphorylase, all considered as key regulators of glycogen metabolism. Glycogen synthase and glycogen phosphorylase mRNAs exhibited significant variations throughout the light-dark cycle with a maximum at the middle of the light period and a minimum at the middle of the dark period. Following sleep deprivation, a two-fold increase in PTG mRNA and a decrease of mRNAs encoding glycogen synthase and glycogen phosphorylase were observed. These transcriptional events have functional consequences as the activity of glycogen synthase was increased 2.5-fold indicating a stimulating effect of sleep deprivation on glycogen synthesis. These results indicate that (i) expression of genes related to brain glycogen metabolism exhibit variations throughout the sleep-waking or rest-activity cycle and (ii) given the almost selective localization of glycogen to astrocytes, these cells might participate in the regulation of sleep.

Comparison of the effects of modafinil and sleep deprivation on sleep and cortical EEG spectra in mice.

Modafinil is a wakefulness-promoting substance whose profile differs from that of the classical psychostimulants. It is still unknown whether waking induced by modafinil and wakefulness induced by sleep deprivation differ in terms of their effect on subsequent sleep. To investigate this problem sleep was recorded in two groups of OF1 mice. One group received modafinil (200 mg/kg, i.p.) at light onset which induced a period of wakefulness of approx. 5 h, while animals of the subsequent control group were injected with vehicle and kept awake for an equivalent duration. The effect of the two treatments on sleep was similar. REM sleep was initially reduced and slow-wave activity (SWA; EEG power in the 0.75-4.0 Hz range) in nonREM sleep was enhanced for several hours. The SWA increase was more prominent over the frontal cortex than over the occipital cortex after both treatments. A minor difference was seen at the occipital site where the initial rise of power in the low-frequency range was larger after vehicle combined with enforced waking than after modafinil. The study shows that the homeostatic sleep response following the modafinil-induced wakefulness corresponds largely to the response following a non-pharmacologically induced extended waking episode.

Sleep EEG in mice that are deficient in the potassium channel subunit K.v.3.2.

Voltage-gated potassium channels containing the K.v.3.2 subunit are expressed in specific neuronal populations such as thalamocortical neurons and fast spiking GABAergic interneurons of the neocortex and hippocampus. These K(+)-channels play a major role in the regulation of firing properties in these neurons. We investigated whether the K.v.3.2 subunit contributes to the generation of the sleep electroencephalogram (EEG). The EEG of a frontal and occipital derivation of K.v.3.2-deficient mice and littermate controls was recorded during a 24-h baseline, 6-h sleep deprivation (SD) and subsequent 18-h recovery to assess also the effects of the K.v.3.2 subunit deficiency under physiological sleep pressure. The K.v.3.2-deficient mice had lower EEG power density in the frequencies between 3.25 and 6 Hz in nonREM (NREM) sleep and 3.25-5 Hz in REM sleep. These differences were more prominent in the frontal derivation than in the occipital derivation. The waking EEG spectrum was not affected by the deletion. In both genotypes SD induced a prominent increase in slow-wave activity in NREM sleep (mean EEG power density between 0.75 and 4.0 Hz), and a concomitant decrease in sleep fragmentation. The effects of SD did not differ significantly between the genotypes. The results indicate that K.v.3.2 channels may be involved in the generation of EEG oscillations in the high delta and low theta range in sleep. They support the notion that GABA-mediated synchronization of cortical activity contributes to the electroencephalogram.

Selective REM sleep deprivation during daytime I. Time course of interventions and recovery sleep.

Although repeated selective rapid eye movement (REM) sleep deprivation by awakenings during nighttime has shown that the number of sleep interruptions required to prevent REM sleep increases within and across consecutive nights, the underlying regulatory processes remained unspecified. To assess the role of circadian and homeostatic factors in REM sleep regulation, REM sleep was selectively deprived in healthy young adult males during a daytime sleep episode (7-15 h) after a night without sleep. Circadian REM sleep propensity is known to be high in the early morning. The number of interventions required to prevent REM sleep increased from the first to the third 2-h interval by a factor of two and then leveled off. Only a minor REM sleep rebound (11.6%) occurred in the following undisturbed recovery night. It is concluded that the limited rise of interventions during selective daytime REM sleep deprivation may be due to the declining circadian REM sleep propensity, which may partly offset the homeostatic drive and the sleep-dependent disinhibition of REM sleep.

Selective REM sleep deprivation during daytime. II. Muscle atonia in non-REM sleep.

One of the hallmarks of rapid eye movement (REM) sleep is muscle atonia. Here we report extended epochs of muscle atonia in non-REM sleep (MAN). Their extent and time course was studied in a protocol that included a baseline night, a daytime sleep episode with or without selective REM sleep deprivation, and a recovery night. The distribution of the latency to the first occurrence of MAN was bimodal with a first mode shortly after sleep onset and a second mode 40 min later. Within a non-REM sleep episode, MAN showed a U-shaped distribution with the highest values before and after REM sleep. Whereas MAN was at a constant level over consecutive 2-h intervals of nighttime sleep, MAN showed high initial values when sleep began in the morning. Selective daytime REM sleep deprivation caused an initial enhancement of MAN during recovery sleep. It is concluded that episodes of MAN may represent an REM sleep equivalent and that it may be a marker of homeostatic and circadian REM sleep regulating processes. MAN episodes may contribute to the compensation of an REM sleep deficit.