Coherent Neocortical 40-Hz Oscillations Are Not Present During

European Journal of Neuroscience, pp. 1–10, 2013 doi:10.1111/ejn.12143

1 2 3 4 5 6 7 Coherent neocortical 40-Hz oscillations are not present 8 9 during REM sleep 10 11

12 1 1 2,3 1 13 Santiago Castro, Atilio Falconi, Michael H. Chase and Pablo Torterolo 1 14 Department of Physiology, School of Medicine, Universidad de la Republica, General Flores 2125, 11800, Montevideo, 15 Uruguay 2 16 WebSciences International, Los Angeles, CA, USA 3 17 1 UCLA School of Medicine, Los Angeles, CA, USA 18 19 Keywords: cat, consciousness, cortex, EEG, synchronization 20 21 22 Abstract 23 During cognitive processes there are extensive interactions between various regions of the cerebral cortex. Oscillations in the 24 gamma frequency band (40 Hz) of the elecroencephalogram (EEG) are involved in the binding of spatially separated but tempo- 25 rally correlated neural events, which results in a unified perceptual experience. The extent of these interactions can be examined 26 by means of a mathematical algorithm called ‘coherence’, which reflects the ‘strength’ of functional interactions between cortical 27 areas. The present study was conducted to analyse EEG coherence in the gamma frequency band of the cat during alert wake- 28 fulness (AW), quiet wakefulness (QW), non-rapid eye movement (NREM) sleep and REM sleep. Cats were implanted with elec- 29 trodes in the frontal, parietal and occipital cortices to monitor EEG activity. Coherence values within the gamma frequency (30– 30 100 Hz) from pairs of EEG recordings were analysed. A large increase in coherence occurred between all cortical regions in the 31 30–45 Hz frequency band during AW compared with the other behavioral states. As the animal transitioned from AW to QW and 32 from QW to NREM sleep, coherence decreased to a moderate level. Remarkably, there was practically no EEG coherence in 33 the entire gamma band spectrum (30–100 Hz) during REM sleep. We conclude that functional interactions between cortical 34 areas are radically different during sleep compared with wakefulness. The virtual absence of gamma frequency coherence during 35 REM sleep may underlie the unique cognitive processing that occurs during dreams, which is principally a REM sleep-related 36 phenomenon. 37 38 39 Introduction 40 The brain integrates fragmentary neural events that occur at different has been observed during attentive wakefulness not only in humans, 41 times and locations into a unified perceptual experience. Understand- but also in animals (Bouyer et al., 1981; Llinas & Ribary, 1993; 42 ing the mechanisms that are responsible for this integration, ‘the Tiitinen et al., 1993; Steriade et al., 1996; Maloney et al., 1997). 43 binding problem’, is one of the most important challenges that cog- Gamma-band rhythmogenesis is inextricably tied to perisomatic 44 nitive neuroscience has to solve (von der Malsburg, 1995; Velik, inhibition wherein the key ingredient is GABAA-receptor-mediated 45 2009). inhibition (Buzsaki & Wang, 2012). 46 Electroencephalographic (EEG) oscillations in the gamma fre- The degree of EEG coherence between two cortical regions is 47 quency band ( 40 Hz) are involved in the integration or binding believed to reflect the strength of the functional interconnections 48 of spatially separated, but temporally correlated, neural events; (re-entries) that occur between them (Edelman & Tononi, 2000; 49 recent studies based on magnetoencephalography and electrocorti- Bullock et al., 2003). Recently, Siegel et al. (2012) have proposed 50 cography have also reported that gamma band oscillations between that frequency-specific correlated oscillations in distributed cortical 51 60 and 200 Hz may also play a role in this process (Uhlhaas et al., networks provide indices, or ‘fingerprints’, of the network interac- 52 2011). tions that underlie cognitive processes; measures of the association 53 An increase in gamma power typically appears during states/ between brain areas based on such frequency-specific signals are 54 behaviors that are characterized by the active cognitive processing likely to provide more detailed information than corresponding 55 of external percepts or internally generated thoughts and images measures based on broadband electrophysiological signals or blood 56 (Uhlhaas et al., 2009, 2011; Rieder et al., 2010). Gamma activity oxygen level-dependent functional magnetic resonance imaging 57 (BOLD-fMRI) signals (Siegel et al., 2012). 58 Coherent EEG activity in the gamma frequency band increases 59 during different behaviors and cognitive functions in both animals 60 Correspondence: Dr P. Torterolo, as above. and humans (Bouyer et al., 1981; Bressler et al., 1993; Harle et al., 61 E-mail: [email protected] 2004). In addition, gamma activity and gamma coherence between 62 Received 12 March 2012, revised 19 December 2012, accepted 29 December 2012 different brain areas have been viewed as a possible neural correlate

1 of consciousness (Llinas et al., 1998). In this regard, coherence in 2 the gamma frequency band is lost during narcosis (unconsciousness) 3 induced by anesthesia (John, 2002; Mashour, 2006). 4 Cognitive activities not only occur during wakefulness. Dreams, 5 which occur more prominently during rapid eye movement (REM) 6 sleep, are considered a special kind of cognitive activity or proto- 7 consciousness (Hobson, 2009). REM sleep dreams are characterized 8 by their vividness, single-mindedness, bizarreness and loss of volun- 9 tary control over the plot. Attention is unstable and rigidly focused, Fig. 1. Position of recording electrodes. The figure depicts the position of 10 facts and reality are not checked, violation of physical laws and the recording electrodes on the surface of the primary sensory, association 11 sensory and prefrontal cerebral cortices. Recordings from these electrodes bizarreness are passively accepted, contextual congruence is dis- were referred to a common inactive electrode, which was located over the 12 torted and time is altered (Rechtschaffen, 1978; Hobson, 2009; Nir frontal sinus. The electrodes were located in accordance with previous 13 & Tononi, 2010). Interestingly, some authors have suggested that reports (Thompson et al., 1963; Markowitsch & Pritzel, 1977; Berman & 14 cognition during REM sleep mimics frontal lobe dysfunction (Corsi- Jones, 1982; Scannell et al., 1995). C1–C4, individual animals. A1, auditory 15 primary cortex; Pf, prefrontal cortex; Pp, posterior-parietal cortex; S1, so- Cabrera et al., 2003). In contrast, during deep non-REM (NREM) mato-sensory primary cortex; V1, visual primary cortex. 16 sleep there is an absence of, or at least a strong reduction in, cogni- 17 tive functions (Hobson, 2009). The electrodes were connected to a Winchester plug, which 18 What occurs with respect to coherence in the gamma frequency together with two plastic tubes (which were used to maintain the 19 band during sleep? In a pioneering study utilizing magnetoencepha- animal’s head fixed without pain or pressure) were bonded to the 20 lographic recordings, Llinas & Ribary (1993) found that during skull with acrylic cement. 21 REM sleep in humans, as in wakefulness, there is large widespread At the end of these surgical procedures an analgesic (Buprenexâ, 22 40-Hz coherent activity that is charactherized by a fronto-occipital 0.01 mg/kg, i.m.) was administered. Incision margins were kept 23 phase shift over the brain (Llinas & Ribary, 1993); however, recent clean and a topical antibiotic was administered on a daily basis. 24 studies have challenged these results (see Discussion). In the present After the animals had recovered from the preceding surgical proce- 25 report, we answered the preceding question by studing gamma band dures, they were adapted to the recording environment for a period 26 coherence among different cortical areas during natural sleep and of at least 2 weeks. 27 wakefulness in the cat. Preliminary data have been presented in 28 abstract form (Castro et al., 2010, 2011). 29 Experimental sessions 30 Experimental sessions of 4 h in duration were conducted between 31 Material and methods 11:00 and 15:00 in a temperature-controlled environment (21– 32 23 °C). All animals had free access to water and food until 1 h prior Experimental animals 33 to the beginning of each recording session. During these sessions 34 Four adult cats were used in this study. The animals were obtained (as well as during adaptation sessions), the animal’s head was held 35 from, and determined to be in good health by, the Institutional in a stereotaxic position by four steel bars that were placed into the 36 Animal Care Facility. All of the experimental procedures were plastic tubes while the body rested in a sleeping bag (semirestricted 37 conducted in accord with the Guide for the Care and Use of Lab- condition). 38 oratory Animals (8th edition, National Academy Press, Washington The activity of three cortical areas, from the same cerebral hemi- 39 DC, 2010) and approved by the Institutional Animal Care Com- sphere, was recorded simultaneously with monopolar electrodes. A 40 mission (protocol no. 71140-1235-09, School of Medicine, Uni- common electrode reference montage was employed. The reference 41 versidad de la Republica). All efforts were made to use the electrode was placed in the left frontal sinus; this location is excel- 42 minimal number of animals necessary to produce reliable scientific lent for a common inactive reference electrode, which is critical for 43 data. the analysis of coherence (Bullock et al., 1995a; Bullock, 1997; Nu- 44 nez et al., 1997; Cantero et al., 2000). The electromyogram (EMG, 45 of the nuchal muscle which was recorded by means of acutely Surgical procedures 46 placed bipolar electrodes) was also monitored. The EMG of the gas- 47 Surgical procedures that were employed were similar to those used trocnemious and temporal muscles, as well as electrocardiogram 48 in previous studies (Torterolo et al., 2002, 2009, 2011b). Accord- (EKG) and electrooculogram (EOG, via acutely placed electrodes on 49 ingly, the animals were chronically implanted with electrodes to the sides of the eyes’ orbits) were also recorded in control experi- 50 monitor the states of sleep and wakefulness. Prior to being anesthe- ments. Each cat was recorded daily for a period of approximately 51 tized, each cat was premedicated with xylazine (2.2 mg/kg, i.m.), 30 days in order to obtain complete data sets. 52 atropine (0.04 mg/kg, i.m.) and antibiotics (Tribrissenâ, 30 mg/kg, Bioelectric signals were amplified (9 1000), filtered (0.1– 53 i.m.). Anesthesia, which was initially induced with ketamine 100 Hz), sampled (512 Hz, 216 bits) and stored in a PC using Spike 54 (15 mg/kg, i.m.), was maintained with a gas mixture of isoflourane 2 software (Cambridge Electronic Design, Cambridge, UK). To dis- 55 in oxygen (1–3%). The head was positioned in a stereotaxic frame card artifacts in coherence analysis due to digitization of the raw 56 and the skull was exposed. In order to record the EEG, stainless recordings, control experiments were also performed with different 57 steel screw electrodes (1.4 mm diameter) were placed on the surface combinations of filters (500 Hz highpass filters, with and without a 58 (above the dura matter) of prefrontal and sensory cortical regions 50-Hz notch filter) as well as with sampling rates of 1024 Hz. No 59 (Fig. 1). To reduce volume conduction, the inter-electrode distance differences were observed in the data obtained in these experiments 60 was at least 6 mm (between posterior-parietal and auditory primary compared with the data presented in the Results. 61 cortices); the greatest was 22 mm (between prefrontal and visual Data were obtained during spontaneously occurring quiet wake- 62 primary cortices). fulness (QW), REM sleep and NREM sleep. The presence of a

1 desynchronized EEG and a high level of EMG activity were used to et al., 2011). Nevertheless, we also examined 60–100 Hz oscilla- 2 identify wakefulness. NREM sleep was determined by the occur- tions with the aim of determining if some of the observed phenom- 3 rence of high-amplitude slow waves (0.5–4 Hz) and sleep ‘spindles’ ena extend to higher frequencies. 4 (9–14 Hz) in the EEG, while REM sleep was identified by a desyn- To normalize the data and evaluate them by means of parametric 5 chronized EEG with ponto-geniculo-occipital (PGO) waves and ato- statistical tests, we applied the Fisher z′ transform to the gamma 6 nia (Ursin & Sterman, 1981). coherence values. Thereafter, the profile of the z′-coherence of the 7 Alert wakefulness (AW) was induced for a period of 300 s by a gamma band in 100-s epochs for each pair of EEG recordings, as 8 sound stimulus, which was introduced approximately 30 min after well as the average of 12 epochs, was analysed; the results were 9 the beginning of the recording. The sound stimulus consisted of presented in graphic form (see Figs 6–8). The z′-coherence of the 10 clicks (0.1 ms in duration) of 60–100 dB SPL in intensity with a gamma band for each pair of EEG channels, which was also aver- 11 variable frequency of presentation (1–500 Hz, modified at random aged across behavioral states independently for each cat, was 12 by the operator) in order to avoid habituation (Torterolo et al., expressed as the mean Æ standard error. The significance of the dif- 13 2003, 2011a). In selected experiments, AW was also induced either ferences among behavioral states was evaluated with ANOVA and 14 by the presentation of a moving light or by placing a mirror in front Tamhane post-hoc tests. The criterion used to reject the null hypoth- 15 of the cat. Because the state of alertness always outlasted the dura- eses was P < 0.05. 16 tion of the stimulus, a waking state was considered as QW when Selected recordings were filtered (band pass 35–45 Hz) using 17 there was a relatively low EMG and EOG activity several minutes Spike 2 digital finite impulse response filters. The amplitude of 18 after stimulation. simultaneously recorded pairs of filtered EEG signals was also anal- 19 ysed by means of the Pearson correlation. Autocorrelations and 20 cross-correlation functions were also performed. 21 Data analysis 22 Sleep and waking states were determined for 10-s epochs. To ana- 23 lyse coherence between pairs of EEG channels, 12 artifact-free peri- Results 24 ods of 100 s were examined during each behavioral state (1200 s Raw and filtered (30–45 Hz) EEG recordings during AW and 25 for each behavioral state). For each pair of recordings, data were REM sleep 26 obtained during four recording sessions. 27 For each 100-s period, the magnitude squared coherence was The EEG fluctuates from a desynchronized pattern of activity during 2 28 analysed as follows – cohab (f) = [∑csdab (f)] /[∑psda (f) ∑psdb (f)], wakefulness and REM sleep to synchronized activity during NREM 29 where psd is the power spectral density, and a and b are the waves sleep. Although EEG activity during wakefulness and REM sleep is 30 that were analyzed. csd, which is the cross spectra density or the similar, there are subtle but striking differences. Representative EEG 31 Fourier analysis of the cross covariance function, provides a reflec- recordings (prefrontal and posterior parietal leads) are shown in 32 tion of how common activity between two processes is distributed Figs 2 and 3. ‘Bursts’ of 35–40 Hz oscillations can be readily 33 across frequencies. Coherence between two waveforms is a function observed in raw recordings during AW (Fig. 2); in contrast, these 34 of frequency and ranges from 0 for totally incoherent waveforms to electrographic events are difficult to find during REM sleep (Fig. 3). 35 1 for maximal coherence. In order for two waveforms to be com- These oscillations were unmasked after digital filtering of the 36 pletely coherent at a particular frequency over a given time range, recordings to include only 30–45 Hz activity (Figs 2 and 3). During 37 the phase shift between the waveforms must be constant and the AW, the ‘bursts’ of gamma oscillations exhibited spindle morphol- 38 amplitudes of the waves must have a constant ratio. This mathemati- ogy with an amplitude and duration of approximately 25 lV and 39 cal approach was described in detail by Bullock & McClune (1989) 200–500 ms, respectively. These gamma oscillations during AW, 40 and has been validated by several investigators (Bouyer et al., 1981; which are clearly observed in the autocorrelation histogram, are 41 Bullock & McClune, 1989; Bullock et al., 1995b; Achermann & accompanied by high values of gamma band power (Fig. 4). In con- 42 Borbely, 1998a,b; Cantero et al., 2004). trast, during REM sleep the amplitude and duration of these events 43 Magnitude squared coherence was measured using the Spike 2 are drastically reduced (Figs 3 and 4). 44 script COHER 1S (Cambridge Electronic Design). With this algo- A strong coupling of EEG signals recorded in different cortical 45 rithm we analysed the coherence between two EEG channels that sites was present during AW, but not during REM sleep (Figs 2–4). 46 were recorded simultaneously during 100-s periods. This period of This high level of coupling between EEG oscillations can be 47 analysis was divided into 100 time-blocks for a sample rate of observed in filtered recordings. Coupling is highlighted when the 48 512 Hz, with a bin size of 1024 samples and a resolution of 0.5 Hz. amplitudes of the signals between pairs of simultaneous EEG record- 49 In pilot recordings and analyses, we determined that the level of ings are correlated. During AW, there was a high correlation between 50 coherence between cortical leads and EMG recordings (neck, gastro- both EEG signals, which was absent during REM sleep (Figs 2–4). 51 cnemious or temporal), the EKG or the EOG, was approximately A visual form to represent the increase in gamma power as well 52 0.1 (Fig. S1). as the coupling among different cortices during AW is shown in 53 In this report, we concentrated on examining the coherence of the Fig. 5. Compared with AW, during REM sleep, gamma power was 54 EEG in the gamma frequency band (30–45 Hz), as originally reduced and coupling was not present. However, note that gamma 55 described by Jasper & Andrews (1938); this narrow band corre- band power during REM sleep is similar to QW and larger than 56 sponds to the 40-Hz cognitive rhythm introduced by Das & Gastaut NREM sleep. 57 (1955). We chose this narrow band of the gamma spectrum for anal- 58 ysis because in preliminary studies we observed that when an ani- Coherent 30–45 Hz activity is present during wakefulness but 59 mal is alerted by different means, there is a noticeable peak in not during REM sleep 60 coherence in this frequency (see Fig. 6). Although some data sug- 61 gest a role in cognitive functions of higher gamma (60–100 Hz) In spite of the fact that EEG coupling was not present during REM 62 oscillations, the mechanisms and functions are not clear (Uhlhaas sleep when analysed by filtered recordings and correlation methods,

1 A A 2 3 4 5 6 7 8 9 10 11 B B 12 C 13 14 15 16 17

18 Colour online, B&W in print 19 20 C Colour online, B&W in print 21 22 Fig. 2. Gamma oscillations during alert wakefulness. (A) Simultaneous raw fi 23 and ltered (35–40 Hz) cortical recordings from the prefrontal (Pf) and posterior-parietal cortices (Pp) during alert wakefulness. Gamma oscillations, 24 which are readily observed in the raw recordings, are highlighted after filter- 25 ing. Scale bars – 1 s and 200 lV for raw recordings and 20 lV for filtered 26 recordings. (B) Gamma oscillations enclosed in A are presented with a longer calibration time. Scale bars – 0.2 s, voltage calibration as in A. (C) 27 fi Regression and correlation coef cient between the amplitudes of Pf and Pp Fig. 4. 28 were analysed from representative filtered recordings during 6 s of alert Gamma oscillations during alert wakefulness (AW) and REM sleep. 29 wakefulness. Regression line equation – y = 0.8x + 2À5. (A) Gamma band power during AW (left) and REM sleep (right), of the 30 EEG recorded from the prefrontal cortex (100-s window). (B) Autocorrela- tion function (ACF) from filtered (35–40 Hz) 300-s EEG recordings from the 31 prefrontal cortex during AW (left) and REM sleep (right). (C) Cross-correla- 32 tion function (CCF) from filtered (35–40 Hz) 300 s of simultaneous EEG 33 recordings from the prefrontal cortex and posterior parietal cortex during 34 A AW (left) and REM sleep (right). 35 36 we applied the ‘coherence’ algorithm for an in-depth analysis of dif- 37 ferent pairs of EEG signals that were simultaneously recorded dur- 38 ing wakefulness and sleep. 39 In pilot studies we analysed the EEG coherence from 0.5 up to 40 100 Hz (Fig. 6). It is readily observed that when an animal is 41 alerted by different means, there is a clear increase in coherence 42 between 30 and 45 Hz, with a narrow peak between 35 and 40 Hz; 43 hence, this band was selected for further analysis. 44 B z′-Coherence profiles for representative pairs of EEG leads are 45 C shown in Fig. 7; the profile of 12 100-s periods and their average 46 for each behavioral state are shown. It is readily observed that z′- 47 coherence profiles for all of the 100-s periods, as well as the aver- 48 age, were larger during AW, with a restricted peak at 35–40 Hz. 49 The z′-coherence decreased during QW and NREM sleep, and was 50 virtually absent during REM sleep. 51 Examples of the averaged gamma (30–45 Hz) band z′-coherence

52 Colour online, B&W in print profiles for different combination of EEG recordings are shown in 53 Fig. 8. For these cortical combinations, the z′-coherence was larger 54 during AW and was drastically reduced during REM sleep. Note 55 Fig. 3. Gamma oscillations during REM sleep. (A) Simultaneous raw and fi that in Fig. 8A, the results of three different procedures for the 56 ltered (35–40 Hz) cortical recordings from the prefrontal (Pf) and posterior- parietal cortices (Pp) during REM sleep. The amplitude and duration of induction of AW (sound, moving light or exposure to a mirror) were 57 gamma oscillations decreased compared with alert wakefulness (see Fig. 2). included; the results of these procedures were similar. During AW, 58 Scale bars – 1 s and 200 lV for raw recordings and 20 lV for filtered the z′-coherence peak reached values up to 2.1, which corresponds 59 recordings. (B) Gamma oscillations enclosed in A are shown with a longer to a coherence of 0.97. 60 calibration time. Scale bars – 0.2 s, voltage calibration as in A. (C) Regres- The averaged z′-coherence in the 35–40 Hz band across behav- sion and correlation coefficient between the amplitudes of Pf and Pp were 61 analysed from representative filtered recordings during 6 s of REM sleep. ioral states for all combinations of cortical recordings is presented in 62 Regression line equation – y = 0.28x + 2À6. Table 1. In these 15 independent analyses, the results were consis-

1 A A 2 3 4 5 6 7 8 9 10 11 12 13

14 Colour online, B&W in print

15 Colour online, B&W in print 16 17 B 18 19 B 20 21 22 23 24 25 26 27 Fig. 6. fi 28 Pro le of EEG coherence during wakefulness. (A) z′-Coherence spectra (0–45 Hz) between the EEG recorded from prefrontal and posterior 29 parietal cortices (1000 s in 100-s epochs). The thin arrow indicates the 30 Fig. 5. Gamma band power during alert wakefulness and REM sleep. (A) moment when the experimenter entered the recording room; the thick dou- 31 Hypnogram and simultaneous gamma power spectrogram from EEG record- ble-headed arrow indicates the time when sound stimulation was applied. 32 ings of the primary auditory (A1), posterior parietal (Pp) and prefrontal cor- When the animal was alerted by a novel stimulus, there was a clear peak in tices (Pf). During alert wakefulness, gamma power values are larger and gamma (30–45 Hz) coherence. The increase in gamma coherence outlasted 33 more coupled between the recorded cortices than during other behavioral the sound stimulation. (B) Twelve z′-coherence profiles (thin lines), from a 34 states. A bar under the hypnogram indicates the time of sound stimulation. representative pair of recordings (prefrontal and posterior-parietal cortices) 35 (B) Twenty-second windows of alert wakefulness and REM sleep. The during alert wakefulness. The average of these profiles (thick line) is also 36 dynamic of gamma power during alert wakefulness was very similar in the shown. The profiles were constructed from periods of 100 s, during sound cortices that were simultaneously recorded; this fact was not evident during 37 stimulation, with a resolution of 0.5 Hz. The coherence for the gamma band REM sleep. (30–45 Hz) is highlighted to the right. 38 39 40 tent. During AW, there was a significant increase in z′-coherence in 41 most of the combinations, while during REM sleep there was a sig- 42 nificant decrease in this parameter. During QW and NREM sleep, z 43 ′-coherence values were intermediate, and in several combinations 44 they were larger during QW than during NREM sleep. Differences 45 in the extent of coherence between different recording pairs during 46 the same behavioral state could be related to either the distance 47 between the electrodes and/or the degree of reciprocal connections 48 between them (Cantero et al., 2000). 49 The z′-coherence, normalized to AW values, highlights the differ- 50 ences between wakefulness and REM sleep (Fig. 9). Depending on 51 the pair of cortical areas analysed, z′-coherence during REM sleep

52 ranged from 60 to 2.6% of the AW values. Colour online, B&W in print 53 Fig. 7. Profile of EEG gamma (30–45 Hz) coherence during wakefulness fi 54 and sleep. Twelve pro les of z′-coherences (thin lines) of a representative Coherence in high (60–100 Hz) gamma band during sleep and pair of recordings (prefrontal and posterior-parietal cortices) as well as the 55 fi wakefulness averages of these 12 pro les (thick lines) are shown for alert and quiet wake- 56 fulness, NREM and REM sleep. The profiles were constructed from periods 57 Figure 6B shows EEG coherence from 0.5 to 100 Hz during AW. of 100 s with a resolution of 0.5 Hz. Gamma coherence was maximal during 58 A coherence peak in the 30–45 Hz band occurs during AW; this alert wakefulness and practically absent during REM sleep. 59 phenomenon is not observed at higher frequencies. 60 The dynamic evolution of EEG coherence in high gamma during z′-coherence was drastically reduced during REM sleep (Fig. 10). 61 wakefulness and sleep is shown in Fig. 10. While changes in As shown in Table 2, a significant decrease in z′-coherence during 62 z′-coherence during AW, QW and NREM sleep were subtle, REM sleep occurred in all cortical combinations that were analysed.

1 ABCFigure 10C presents 60–100 Hz power spectrum profiles from 2 representative 100-s time epochs. In comparison to AW or QW, 3 high gamma power is reduced during NREM sleep. Note that even 4 though high gamma coherence is reduced during REM sleep 5 (Fig. 10A), high gamma power is similar than waking levels. 6 7 8 Discussion 9 In the present study, we demonstrated that the EEG intra-hemi- 10 spheric coherence in the gamma (30–45 Hz) frequency band is lar- 11 ger during AW than during QW. In addition, gamma coherence 12 decreased to a lower level during NREM sleep, and was absent dur- 13 ing REM sleep; the lack of coherence during REM sleep also 14 extended to high gamma (60–100 Hz) band frequencies. Therefore, 15 DEFduring REM sleep, the coupling of high-frequency neuronal activity 16 among different cortical regions is practically non-existent. 17 18 19 20 21 Colour online, B&W in print 22 23 24 25 26 27 28 29 Fig. 8. Gamma band z′-coherence profiles from representative prefrontal- 30 perceptual and perceptual regions during sleep and wakefulness. Graphics at 31 the top (A–C) show z′-coherence between prefrontal and perceptual cortices. At the bottom (D–F), z′-coherences profiles between perceptual cortices are 32 shown. In A, three types of stimuli produced alert wakefulness – sound Fig. 9. Gamma band (35–40 Hz) z′-coherence for all analysed cortical com- 33 (AWs), mirror (AWm) and moving light (AWml). In B–F, alertness was pro- binations. The data from Table 1 were normalized to the values obtained Colour online, B&W in print 34 duced only by sound stimulation. For all combinations of recordings, z′- during alert wakefulness. C1–C4, individual animals. A1, auditory cortex; Pf, 35 coherence in the gamma band was larger during AW than during REM sleep. prefrontal cortex; Pp, posterior-parietal cortex; S1, somato-sensory primary 36 QW, quiet wakefulness; NREM, non-REM sleep; REM, REM sleep; A1, cortex; V1, visual primary cortex. In the derivatives indicated by an asterisk, auditory cortex; Pf, prefrontal cortex; Pp, posterior-parietal cortex; V1, visual AW was produced by light stimulation. AW, alert wakefulness; QW, quiet 37 primary cortex. wakefulness; NREM, non-REM sleep; REM, REM sleep. 38 39 40 41 Table 1. Gamma (35–40 Hz) z’-coherence values during sleep and wakefulness 42 43 Animal Derivates AW QW NREM REM Statistical significance F 44 ‡ † 45 C1 S1-Pp 1.09 Æ 0.05 0.62 Æ 0.06 0.53 Æ 0.04 0.32 Æ 0.01 AW vs. all ; REM vs. QW and NREM 54 Æ Æ Æ Æ ‡ † 46 Pp-V1 1.31 0.06 0.89 0.06 0.92 0.06 0.75 0.06 AW vs. all ; REM vs. QW and NREM 46 S1-V1 0.70 Æ 0.05 0.15 Æ 0.01 0.29 Æ 0.04 0.11 Æ 0.01 AW vs. all‡; REM vs. NREM†; NREM vs. QW† 54 47 ‡ † † 48 C2 Pf-Pp 0.67 Æ 0.02 0.30 Æ 0.04 0.16 Æ 0.01 0.10 Æ 0.01 AW vs. all ; REM vs. QW and NREM ; NREM vs. QW 114 Æ Æ Æ Æ ‡ † 49 Pf-A1 0.87 0.05 0.49 0.03 0.36 0.01 0.27 0.01 AW vs. all ; REM vs. QW and NREM 86 Pp-A1 1.27 Æ 0.04 0.89 Æ 0.05 0.81 Æ 0.01 0.76 Æ 0.01 AW vs. all‡; REM vs. QW and NREM† 29 50 † † † 51 C3 Pf-V1 0.14 Æ 0.01 0.13 Æ 0.02 0.06 Æ 0.01 0.05 Æ 0.01 AW vs. REM and NREM ; REM vs. QW ; NREM vs. QW 18 Æ Æ Æ Æ ‡ † † 52 S1-A1 0.99 0.02 0.70 0.03 0.48 0.01 0.48 0.01 AW vs. all ; REM vs. QW ; NREM vs. QW 137 V1-A1 0.76 Æ 0.02 0.46 Æ 0.03 0.28 Æ 0.01 0.24 Æ 0.03 AW vs. all‡; REM vs. QW†; NREM vs. QW‡ 102 53 ‡ ‡ 54 C4 Pf-Pp 0.80 Æ 0.05 0.27 Æ 0.04 0.14 Æ 0.01 0.05 Æ 0.01 AW vs. all ; REM vs. QW and NREM 118 Æ Æ Æ Æ ‡ ‡ ‡ 55 Pf-Pp* 0.94 0.03 0.31 0.02 0.13 0.01 0.04 0.00 AW vs. all ; REM vs. all ; NREM vs. AW and REM 121 Pf-A1 0.79 Æ 0.04 0.29 Æ 0.04 0.16 Æ 0.01 0.07 Æ 0.01 AW vs. all‡; REM vs. QW and NREM‡; NREM vs. QW† 173 56 Pp-A1 1.57 Æ 0.03 1.20 Æ 0.03 0.99 Æ 0.02 0.88 Æ 0.01 AW vs. all‡; REM vs. QW and NREM‡; NREM vs. QW‡ 515 57 Pf-V1* 0.86 Æ 0.02 0.26 Æ 0.02 0.10 Æ 0.00 0.02 Æ 0.00 AW vs. all‡; REM vs. QW‡; NREM vs. QW‡ 616 58 Pp-V1* 1.70 Æ 0.02 1.28 Æ 0.03 0.99 Æ 0.02 0.96 Æ 0.02 AW vs. all‡; REM vs. QW and NREM‡; NREM vs. QW‡ 235 59 60 ANOVA with Tamhane tests. The degrees of freedom were 3 (between groups) and 44 (within groups) for all the derivates that were analysed. A1, auditory cortex; Pf, prefrontal cortex; Pp, posterior-parietal cortex; S1, somato-sensory primary cortex; V1, visual primary cortex. AW, alert wakefulness; QW, quiet wake- 61 fulness; NREM, non-REM sleep; REM, REM sleep. *In these cortical pairs AW was induced by light stimulation. The values represent mean Æ standard error. 62 †P < 0.05 and ‡P < 0.0001.

1 A Animal model 2 fi 3 We used the cat as an animal model because it has well-de ned, 4 consolidated sleep and waking states. In addition, there are data that 5 suggest that complex, dream-like, cognitive processes are present 6 during REM sleep in this species (Sastre & Jouvet, 1979; Morrison, 7 1983). Experimental lesions targeted to the dorsal pontine reticular 8 formation elicit REM sleep without atonia; in this condition, dream- 9 like behaviors with complex motor activities, such as jumps and 10 predation-like movements, are released (Sastre & Jouvet, 1979; 11 Morrison, 1983). Furthermore, cortical development is greater in the 12 cat than in rodents, especially in associative cortices; consequently, the cat is an special attractive species in which to study cognitive 13 B processes (Thompson et al., 1963; Markowitsch & Pritzel, 1977; 14 Colour online, B&W in print 15 Fuster, 1989; Scannell et al., 1995). 16 Recordings were carried out in semirestricted condition. The 17 advantage is that the differences among states are the states per se; fl 18 changes in postures or movements did not in uence the recordings, 19 reducing the possibility of artifacts. C 20 21 Extra-cerebral potentials did not explain gamma band 22 coherence 23 24 There are two main arguments against the possibility that extra-cere- 25 bral potentials generated by muscles, heart or eye movements could ‘ ’ 26 produce the large coherence observed during AW. First, gamma 27 coherence between the EEG and the activity of different muscles EMGs, EKG or EOG recorded by means of either bipolar or monopo- 28 Fig. 10. Dynamic evolution of EEG coherence during sleep and wakeful- lar electrodes was negligible (a portion of these results are shown in 29 ness. (A) Three-dimensional spectrogram of z′-coherence of the EEG (60– 30 100 Hz) from simultaneous recordings from prefrontal and posterior parietal Fig. S1). Second, there was a variable latency between most of the 31 cortices (3300 s). Time and frequency are shown on the horizontal and verti- peaks of the gamma waves between the different cortices that were 32 cal (depth) axes, respectively, while z′-coherence is represented in a color recorded simultaneously. This latency was proportional to the dis- code. z′-Coherence is drastically reduced during REM sleep. (B) Hypnogram 33 tance between the recorded cortices (not shown); it would not be of recordings analysed in A. The bar under the hypnogram indicates the time present if the potentials were produced from an extra-cerebral source. 34 of sound stimulation. (C) Gamma band power spectrum profiles from repre- 35 sentative 100-s time epochs enclosed in B during AW, QW, NREM sleep This fact also allows one to reject the volume conduction as an 36 and REM sleep. important component of gamma band coherence in the present study. 37 38 39 40 Table 2. ’ 41 Gamma (60–100 Hz) z -coherence values during sleep and wakefulness 42 Animal Derivates AW QW NREM REM Statistical significance F 43 44 C1 S1-Pp 0.63 Æ 0.02 0.53 Æ 0.01 0.40 Æ 0.02 0.19 Æ 0.01 REM vs. all‡; AW vs. QW and NREM† 149 45 Pp-V1 0.83 Æ 0.01 0.76 Æ 0.01 0.61 Æ 0.02 0.48 Æ 0.01 REM vs. all‡; AW vs. NREM‡ 122 46 S1-V1 0.40 Æ 0.01 0.35 Æ 0.01 0.21 Æ 0.01 0.04 Æ 0.01 REM vs. all†; AW vs. QW and NREM† 175 47 C2 Pf-Pp 0.43 Æ 0.01 0.34 Æ 0.03 0.31 Æ 0.01 0.12 Æ 0.01 REM vs. all‡; AW vs. QW and NREM† 58 48 Pf-A1 0.51 Æ 0.01 0.45 Æ 0.02 0.45 Æ 0.01 0.25 Æ 0.01 REM vs. all‡; AW vs. NREM† 70 49 Pp-A1 0.98 Æ 0.04 0.86 Æ 0.05 0.88 Æ 0.02 0.71 Æ 0.01 REM vs. all† 10 50 Æ Æ Æ Æ † † 51 C3 Pf-V1 0.08 0.00 0.08 0.00 0.05 0.01 0.02 0.00 REM vs. all ; NREM vs. AW and QW 48 S1-A1 0.61 Æ 0.01 0.43 Æ 0.01 0.36 Æ 0.02 0.30 Æ 0.01 REM vs. all‡; AW vs. QW and NREM†; QW vs. NREM† 46 52 V1-A1 0.39 Æ 0.02 0.36 Æ 0.01 0.18 Æ 0.02 0.12 Æ 0.02 REM vs. AW and QW‡; NREM vs. AW and QW† 93 53 ‡ 54 C4 Pf-Pp 0.35 Æ 0.02 0.31 Æ 0.01 0.31 Æ 0.03 0.07 Æ 0.01 REM vs. all 53 Æ Æ Æ Æ ‡ 55 Pf-Pp* 0.37 0.02 0.30 0.02 0.26 0.02 0.03 0.00 REM vs. all 110 Pf-A1 0.31 Æ 0.01 0.27 Æ 0.01 0.28 Æ 0.03 0.07 Æ 0.01 REM vs. all† 49 56 Pp-A1 0.92 Æ 0.02 0.89 Æ 0.01 0.85 Æ 0.04 0.64 Æ 0.05 REM vs. all‡; AW vs. QW and NREM† 16 57 Pf-V1* 1.07 Æ 0.02 1.11 Æ 0.02 0.93 Æ 0.02 0.70 Æ 0.01 REM vs. all‡; QW vs. NREM† 100 58 Pp-V1* 0.45 Æ 0.02 0.37 Æ 0.02 0.31 Æ 0.00 0.06 Æ 0.02 REM vs. all‡; AW vs. NREM† 113 59 60 ANOVA with Tamhane tests. The degrees of freedom were 3 (between groups) and 44 (within groups) for all the derivates that were analysed. A1, auditory cortex; Pf, prefrontal cortex; Pp, posterior-parietal cortex; S1, somato-sensory primary cortex; V1, visual primary cortex. AW, alert wakefulness; QW, quiet wake- 61 fulness; NREM, non-REM sleep; REM, REM sleep. *In these cortical pairs AW was induced by light stimulation. The values represent mean Æ standard error. 62 †P < 0.05 and ‡P < 0.0001.

1 Data in humans indicate that saccadic eye movements can be a However, the coupling of gamma waves, per se, was not the focus 2 source of gamma wave oscillations in frontal derivations (Yuval- of this study. In rats, Gervasoni et al. (2004) analysed pooled 3 Greenberg et al., 2008). In addition to the lack of coherence gamma (30–55 Hz) coherence between neocortical, hippocampal as 4 between the EEG and the EOG, the absence of gamma coherence well as subcortical loci (caudate-putamen and thalamus). They found 5 during REM sleep, when saccadic eye movements predominate, that REM sleep was the state of maximum pooled coherence in the 6 indicates that these movements did not play a role in the generation gamma range; however, gamma coherence among different neocorti- 7 of gamma coherence in the cat. cal sites was not analysed. 8 The results of the present study led to a different conclusion 9 compared with the study of Llinas & Ribary (1993), who reported Coherence analysis in a 100-s window 10 that during REM sleep magneto-EEG 40-Hz oscillations were simi- 11 To determine general trends in coherence analyses, we employed lar in distribution, phase and amplitude to those observed during 12 time windows of 100 s, although the timescale for functionally sig- wakefulness; in contrast, gamma coherent activity was reduced dur- 13 nificant events and fluctuations in the brain is much shorter. In rela- ing NREM sleep. They also found that gamma oscillations become 14 tion to cognitive functions, Libet et al. (1991) demonstrated that robustly coherent in short time windows after sensory stimulation 15 appropriate neuronal activation of approximately 500 ms in duration during waking; in other words, sensory stimuli ‘reset’ gamma oscil- 16 is required to elicit a conscious sensory experience (Libet et al., lations. This effect was not observed during REM sleep. The pre- 17 1991). Therefore, from an electrophysiological point of view, in ceding authors considered that dreams during REM sleep represent 18 100-s time-windows one can average the results of different EEG a state of hyper-attentiveness in which sensory inputs cannot 19 events that occur during wakefulness (i.e. sensory or motor-evoked address the processes that generate conscious experience (Llinas & 20 events), during NREM sleep (i.e. spindles and slow waves) or REM Ribary, 1993). The discrepancies with our data may arise from the 21 sleep (i.e. PGO waves and tonic EEG desychronization). Neverthe- different analytical approaches; Llinas & Ribary (1993) analysed 22 less, an increase in gamma coupling was also evident during shorter coherence by superimposing 37 traces (filtered at 35–45 Hz) 23 periods of analysis. Good correlation (R2 = 0.81) was evident in fil- recorded during 0.6–3-s epochs. It is likely that small time periods 24 tered recordings during 6-s epochs of AW. Furthermore, coupling of high gamma coherence may appear in humans during phasic 25 was readily observed in the 200–500 ms gamma ‘bursts’ that were periods of REM sleep; in fact an increase in gamma activity was 26 present in the raw and filtered recordings during AW; these synchro- found during these periods (Gross & Gotman, 1999). However, we 27 nized gamma ‘bursts’ are also identified by their power in Fig. 5 hypothesize that this increase in gamma coherence would not be 28 (see also the cross correlation function in Fig. 4). In contrast, maintained if larger periods of REM sleep (that include tonic REM 29 gamma band coupling disappeared during REM sleep (Figs 3–5). sleep) were analysed. 30 In contrast to the references cited in the preceding paragraphs, 31 several reports are in agreement with our results. Perez-Garci et al. Gamma coherence during wakefulness 32 (2001) reported that there is a decrease in correlation spectra in 2-s 33 It is well known that gamma power and gamma coherence increase epochs of fast (27–48 Hz) frequencies restricted to intra-hemispheric 34 during wakefulness in animals and humans (Llinas & Ribary, 1993; frontal-perceptual cortical regions during REM sleep in humans 35 Maloney et al., 1997; Rieder et al., 2010). In the cat, gamma band (Perez-Garci et al., 2001). However, among perceptual regions, cor- 36 coherence increases during AW between electrodes located in cortical relation values for REM sleep and wakefulness were similar. Cante- 37 and thalamic sites (Bouyer et al., 1981). In accordance with these ro et al. (2004) employed human intracranial EEG recordings for 38 results, we demonstrated that during AW, coherence increases coherence analyses during sleep, which allow a much finer spatial 39 between different cortical areas; this fact was clearly observed at fre- scale than do scalp-recorded signals. They found that local (within 40 quencies of 40 Hz, but not at higher gamma bands. This increase neocortical regions) and long-range (between intra-hemispheric neo- 41 in coherence occurs during alert behaviors that were induced either cortical regions) gamma coherence was significantly greater during 42 by sound, moving light stimulation or when a mirror was located in wakefulness than during sleep (Cantero et al., 2004). However, no 43 front of the animal (or immediately after these procedures, as shown differences in coherence in the 35–58 Hz frequency band were 44 in Fig. 6). Coherence also increased when the level of alertness was found between NREM and REM sleep. Furthermore, functional 45 high at the beginning of the recording or when an individual entered gamma-range coupling between the neocortex and hippocampus was 46 into the recording room (Fig. 6). Therefore, it is the state of alertness, observed during wakefulness, but not during REM sleep. A recent 47 per se, and not the stimuli that increase gamma band coherence. study by Voss et al. (2009) demonstrated that coherence for all 48 EEG spectra decreases during REM sleep compared with wakeful- 49 ness (Voss et al., 2009); interestingly, during lucid dreaming, coher- Gamma coherence during sleep 50 ence values are intermediate between wakefulness and REM sleep. 51 The intercortical dialogue during wakefulness and sleep has been In the present paper, we have demonstrated in the cat that gamma 52 analysed with standard EEG recordings in humans (Achermann & intra-hemispheric coherence is practically absent during REM sleep. 53 Borbely, 1998a,b), and an increase in gamma band (up to 50 Hz) This was true for coherence measured between prefrontal and per- 54 coherence during REM sleep in comparison with NREM sleep was ceptual regions of the cortex, and among perceptual cortices, includ- 55 observed (Achermann & Borbely, 1998a,b). This effect was subtle ing primary as well as associative sensory areas. Interestingly, 56 and mainly present between anterior inter-hemispheric homologous gamma band uncoupling during REM sleep was readily observed by 57 leads. However, the preceding authors were cautious in interpreting simple inspection of filtered recordings. Future studies are needed to 58 their findings because of the low amplitude of high-frequency sig- analyse the degree of gamma band coupling of cortical and subcorti- 59 nals of the standard EEG recordings. cal areas that increase their activity during REM sleep, such as the 60 Nir et al. (2008) studied inter-hemispheric correlations of enve- limbic cortices and amygdala (Schwartz & Maquet, 2002). 61 lopes (0.1 Hz) of gamma oscillations during REM sleep and Stage 2 Our demonstration that there is a radical reduction in gamma 62 NREM sleep in comparison with wakefulness (Nir et al., 2008). coherence during REM sleep between different and distant cortical

1 regions does not contradict findings from Steriade’s group, which tions that take place during mentation that occur during wakefulness 2 show an increase in local coupling (within a column or among closely and REM sleep (dreams). 3 cortical sites) during activated states (Steriade et al., 1996; Destexhe 4 et al., 1999). In fact, as suggested in Figs 5 and 10, gamma power 5 values (as a reflection of local gamma synchronization) during REM Acknowledgements 2 6 sleep were similar to QW and greater than during NREM sleep. We thank Luciana Benedetto and Matias Cavelli for their technical assis- 7 tance. We are also grateful to Dr Giancarlo Vanini for critical comments 8 regarding the manuscript. We are indebted to several anonymous reviewers – 9 Absence of gamma coherence during REM sleep impact on for comments on the manuscript. The ‘Programa de Desarrollo de Ciencias ’ 10 cognitive functions Basicas (PEDECIBA), Uruguay, partially supported this work. 11 A proper understanding of cognitive functions cannot be achieved 12 without an understanding of consciousness, which involves a Conflict of interest 13 system’s capacity to integrate information (Tononi, 2010). Recently, 14 utilizing transcranial magnetic stimulation (TMS) and high-density None. 15 EEG recordings in humans, Massimini et al. (2010) demonstrated 16 that although during NREM sleep TMS elicited local and stereotypi- References 17 cal cortical activations, during REM sleep TMS triggered a wide- 18 spread and differentiated patterns of cortical activation that were Achermann, P. & Borbely, A.A. (1998a) Coherence analysis of the human 19 similar to those observed during wakefulness. These data indicate sleep electroencephalogram. Neuroscience, 85, 1195–1208. 20 fi Achermann, P. & Borbely, A.A. (1998b) Temporal evolution of coherence that there is a signi cant impairment in intracortical dialogue during and power in the human sleep electroencephalogram. J. Sleep Res., 7(Sup- 21 NREM sleep, although during REM sleep interactions among corti- pl. 1), 36–41. 22 cal sites are operative (Massimini et al., 2010). However, in this Berman, A.L. & Jones, E.G. (1982). The Thalamus and Basal Telenceph- 23 study the frequency bands of the cortical interactions were not anal- alum of the Cat. A Citoarchitectonic Atlas with Stereotaxic Coordinates. 24 University of Wisconsin, Madison. ysed. Our results in the cat and previous results from humans clearly Bouyer, J.J., Montaron, M.F. & Rougeul, A. (1981) Fast fronto-parietal 25 show that, compared with wakefulness, during REM sleep there is a rhythms during combined focused attentive behaviour and immobility in 26 failure in the capacity of integration among different cortices, at cat: cortical and thalamic localizations. Electroencephalogr. Clin. Neuro- 27 least in high frequency ranges (Perez-Garci et al., 2001; Cantero physiol., 51, 244–252. 28 et al., 2004; Voss et al., 2009). This phenomenon may underlie Bressler, S.L., Coppola, R. & Nakamura, R. (1993) Episodic multiregional 29 cortical coherence at multiple frequencies during visual task performance. some of the unique patterns of REM sleep mentation (Hobson, Nature, 366, 153–156. 30 2009; Nir & Tononi, 2010); in fact, the prevalence of thoughts Bullock, T.H. (1997) Signals and signs in the nervous system: the dynamic 31 decreases from wakefulness to NREM sleep and reach a nadir dur- anatomy of electrical activity is probably information-rich. Proc. Natl 32 ing REM sleep (the same as gamma coherence), while hallucinations Acad. Sci. USA, 94,1–6. 33 Bullock, T.H. & McClune, M.C. (1989) Lateral coherence of the electrocorti- and emotions follow an inverse pattern (Fosse et al., 2001). cogram: a new measure of brain synchrony. Electroencephalogr. Clin. 34 Bullock et al. (2003) considered that the following processes Neurophysiol., 73, 479–498. 35 result in an increase in coherence between two structures: (1) both Bullock, T.H., McClune, M.C., Achimowicz, J.Z., Iragui-Madoz, V.J., Duck- 36 structures are driven by the same generator, (2) both structures can row, R.B. & Spencer, S.S. (1995a) EEG coherence has structure in the 37 mutually drive each other and (3) one of the structures drives the millimeter domain: subdural and hippocampal recordings from epileptic 38 patients. Electroencephalogr. Clin. Neurophysiol., 95, 161–177. other one. It is likely that differential activation of regulatory (cho- Bullock, T.H., McClune, M.C., Achimowicz, J.Z., Iragui-Madoz, V.J., Duck- 39 linergic, monoaminergic and peptidergic) systems during REM sleep row, R.B. & Spencer, S.S. (1995b) Temporal fluctuations in coherence of 40 may be the foundation for cortical uncoupling (Watson et al., 2010). brain waves. Proc. Natl Acad. Sci. USA, 92, 11568–11572. 41 Although it is not clear what role is played by these neuromodulato- Bullock, T.H., McClune, M.C. & Enright, J.T. (2003) Are the electroenceph- 42 alograms mainly rhythmic? Assessment of periodicity in wide-band time ry systems with respect to large-scale cortical interactions, it is series. Neuroscience, 121, 233–252. 43 accepted that at least some of these systems alter the frequency Buzsaki, G. & Wang, X.J. (2012) Mechanisms of gamma oscillations. Annu. 44 range and the strength of local cortical oscillations (Munk et al., Rev. Neurosci., 35, 203–225. 45 1996; Rodriguez et al., 2004; Roopun et al., 2010). Recent studies Cantero, J.L., Atienza, M. & Salas, R.M. (2000) Clinical value of EEG 46 have demonstrated the presence of gamma activity in the reticular coherence as electrophysiological index of cortico-cortical connections dur- 47 ing sleep. Rev. Neurol., 31, 442–454. activating system (RAS), and have suggested that this activity helps Cantero, J.L., Atienza, M., Madsen, J.R. & Stickgold, R. (2004) Gamma 48 to stabilize coherence related to arousal (reviewed by Urbano et al., EEG dynamics in neocortex and hippocampus during human wakefulness 49 2012); however, the role of gamma activity in the RAS during and sleep. Neuroimage, 22, 1271–1280. 50 Castro, S., Benedetto, L., Gutierrez, M., Falconi, A. & Torterolo, P. (2010) REM sleep is unknown. Nevertheless, the study of mechanisms that ‘ ’ 51 fi Coherence in the cortical electrical activity during sleep and wakefulness. support the modi cation of gamma coherence during sleep and In 3rd International Congress on Sleep Medicine and 12th Brazilian Con- 52 wakefulness was beyond the scope of the present report. gress on Sleep Medicine. San Pablo. 53 Castro, S., Gutierrez, M., Falconi, A., Chase, M. & Torterolo, P. (2011) 54 Absence of EEG gamma (35-40 Hz) coherence characterizes REM sleep 55 Conclusions and differentiates it from wakefulness. In 41st Annual Meeting of the Soci- 56 ety for Neuroscience. Washington. During REM sleep in the cat, despite an activated (desynchronized) Corsi-Cabrera, M., Miro, E., del-Rio-Portilla, Y., Perez-Garci, E., Villanueva, 57 EEG, there is an absence of cross-talk between different cortical Y. & Guevara, M.A. (2003) Rapid eye movement sleep dreaming is char- 58 regions in the gamma frequency band. Therefore, functional interac- acterized by uncoupled EEG activity between frontal and perceptual corti- 59 cal regions. Brain Cogn., 51, 337–345. tions among different cortical areas (re-entries), which are critical ’ 60 Das, N.N. & Gastaut, H. (1955) Variations de l activite electrique du cer- for cognitive functions, are different during wakefulness and REM veau, du coeur et des muscles squellettiques au cours de la meditation 61 sleep. This uncoupling of gamma frequency activity during REM et de l’extase yoguique. Electroencephalogr. Clin. Neurophysiol., 6, 62 sleep may be involved in the distinctiveness of the cognitive opera- 211.

1 Destexhe, A., Contreras, D. & Steriade, M. (1999) Spatiotemporal analysis Rodriguez, R., Kallenbach, U., Singer, W. & Munk, M.H. (2004) Short- and 2 of local field potentials and unit discharges in cat cerebral cortex during long-term effects of cholinergic modulation on gamma oscillations and 3 natural wake and sleep states. J. Neurosci., 19, 4595–4608. response synchronization in the visual cortex. J. Neurosci., 24, 10369– A Universe of Consciousness 4 Edelman, G.M. & Tononi, G. (2000). . Basic 10378. Books, New York. Roopun, A.K., Lebeau, F.E., Ramell, J., Cunningham, M.O., Traub, R.D. & 5 Fosse, R., Stickgold, R. & Hobson, J.A. (2001) Brain-mind states: reciprocal Whittington, M.A. (2010) Cholinergic neuromodulation controls directed 6 variation in thoughts and hallucinations. Psychol. Sci., 12,30–36. temporal communication in neocortex in vitro. Front. Neural Circuits, 4, 7 Fuster, J.M. (1989). The Prefrontal Cortex. Raven Press, New York. 8. 8 Gervasoni, D., Lin, S.C., Ribeiro, S., Soares, E.S., Pantoja, J. & Nicolelis, Sastre, J.P. & Jouvet, M. (1979) Oneiric behavior in cats. Physiol. Behav., M.A. (2004) Global forebrain dynamics predict rat behavioral states and 22, 979–989. 9 their transitions. J. Neurosci., 24, 11137–11147. Scannell, J.W., Blakemore, C. & Young, M.P. (1995) Analysis of connectiv- 10 Gross, D.W. & Gotman, J. (1999) Correlation of high-frequency oscillations ity in the cat cerebral cortex. J. Neurosci., 15, 1463–1483. 11 with the sleep–wake cycle and cognitive activity in humans. Neuroscience, Schwartz, S. & Maquet, P. (2002) Sleep imaging and the neuro-psychologi- 12 94, 1005–1018. cal assessment of dreams. Trends Cogn. Sci., 6,23–30. fi 13 Harle, M., Rockstroh, B.S., Keil, A., Wienbruch, C. & Elbert, T.R. (2004) Siegel, M., Donner, T.H. & Engel, A.K. (2012) Spectral ngerprints of Mapping the brain’s orchestration during speech comprehension: task-spe- large-scale neuronal interactions. Nat. Rev. Neurosci., 13, 121–134. 14 cific facilitation of regional synchrony in neural networks. BMC Neurosci., Steriade, M., Amzica, F. & Contreras, D. (1996) Synchronization of fast (30 15 5, 40. –40 Hz) spontaneous cortical rhythms during brain activation. J. Neurosci., 16 Hobson, J.A. (2009) REM sleep and dreaming: towards a theory of protocon- 16, 392–417. 17 sciousness. Nat. Rev. Neurosci., 10, 803–813. Thompson, R.F., Johnson, R.H. & Hoopes, J.J. (1963) Organization of audi- Jasper, H.H. & Andrews, H.L. (1938) Brain potentials and voluntary muscle tory, somatic sensory, and visual projection to association fields of cerebral 18 activity in man. J. Neurophysiol., 1,87–100. cortex in the cat. J. Neurophysiol., 26, 343–364. 19 John, E.R. (2002) The neurophysics of consciousness. Brain Res. Brain Res. Tiitinen, H., Sinkkonen, J., Reinikainen, K., Alho, K., Lavikainen, J. & Naat- 20 Rev., 39,1–28. anen, R. (1993) Selective attention enhances the auditory 40-Hz transient 21 Libet, B., Pearl, D.K., Morledge, D.E., Gleason, C.A., Hosobuchi, Y. & Bar- response in humans. Nature, 364,59–60. 22 baro, N.M. (1991) Control of the transition from sensory detection to sen- Tononi, G. (2010) Information integration: its relevance to brain function sory awareness in man by the duration of a thalamic stimulus. The and consciousness. Arch. Ital. Biol., 148, 299–322. 23 cerebral ‘time-on’ factor. Brain, 114, 1731–1757. Torterolo, P., Morales, F.R. & Chase, M.H. (2002) GABAergic mechanisms 24 Llinas, R. & Ribary, U. (1993) Coherent 40-Hz oscillation characterizes in the pedunculopontine tegmental nucleus of the cat promote active 25 dream state in humans. Proc. Natl Acad. Sci. USA, 90, 2078–2081. (REM) sleep. Brain Res., 944,1–9. 26 Llinas, R., Ribary, U., Contreras, D. & Pedroarena, C. (1998) The neuronal Torterolo, P., Yamuy, J., Sampogna, S., Morales, F.R. & Chase, M.H. basis for consciousness. Philos. Trans. R. Soc. Lond. B Biol. Sci., 353, (2003) Hypocretinergic neurons are primarily involved in activation of the 27 1841–1849. somatomotor system. Sleep, 1,25–28. 28 Maloney, K.J., Cape, E.G., Gotman, J. & Jones, B.E. (1997) High-frequency Torterolo, P., Benedetto, L., Lagos, P., Sampogna, S. & Chase, M.H. (2009) 29 gamma electroencephalogram activity in association with sleep-wake states State-dependent pattern of Fos protein expression in regionally-specific 30 and spontaneous behaviors in the rat. Neuroscience, 76, 541–555. sites within the preoptic area of the cat. Brain Res., 1267,44–56. von der Malsburg, C. (1995) Binding in models of perception and brain Torterolo, P., Ramos, O.V., Sampogna, S. & Chase, M.H. (2011a) Hypocret- 31 function. Curr. Opin. Neurobiol., 5, 520–526. inergic neurons are activated in conjunction with goal-oriented survival- 32 Markowitsch, H.J. & Pritzel, M. (1977) A stereotaxic atlas of the prefrontal related motor behaviors. Physiol. Behav., 104, 823–830. 33 cortex of the cat. Acta Neurobiol. Exp. (Wars), 37,63–81. Torterolo, P., Sampogna, S. & Chase, M.H. (2011b) A restricted parabrachial 34 Mashour, G.A. (2006) Integrating the science of consciousness and anesthe- pontine region is active during non-rapid eye movement sleep. Neurosci- 35 sia. Anesth. Analg., 103, 975–982. ence, 190, 184–193. Massimini, M., Ferrarelli, F., Murphy, M., Huber, R., Riedner, B., Casarotto, Uhlhaas, P.J., Pipa, G., Lima, B., Melloni, L., Neuenschwander, S., Nikolic, 36 S. & Tononi, G. (2010) Cortical reactivity and effective connectivity dur- D. & Singer, W. (2009) Neural synchrony in cortical networks: history, 37 ing REM sleep in humans. Cogn. Neurosci., 1, 176–183. concept and current status. Front. Integr. Neurosci., 3, 17. 38 Morrison, A.R. (1983) A window on the sleeping brain. Sci. Am., 248,94–102. Uhlhaas, P.J., Pipa, G., Neuenschwander, S., Wibral, M. & Singer, W. 39 Munk, M.H., Roelfsema, P.R., Konig, P., Engel, A.K. & Singer, W. (1996) (2011) A new look at gamma? High- (>60 Hz) gamma-band activity in Prog. Biophys. 40 Role of reticular activation in the modulation of intracortical synchroniza- cortical networks: function, mechanisms and impairment. tion. Science, 272, 271–274. Mol. Biol., 105,14–28. 41 Nir, Y. & Tononi, G. (2010) Dreaming and the brain: from phenomenology Urbano, F.J., Kezunovic, N., Hyde, J., Simon, C., Beck, P. & Garcia-Rill, E. 42 to neurophysiology. Trends Cogn. Sci., 14,88–100. (2012) Gamma band activity in the reticular activating system. Front. Neu- 43 Nir, Y., Mukamel, R., Dinstein, I., Privman, E., Harel, M., Fisch, L., Gel- rol., 3,6. 44 bard-Sagiv, H., Kipervasser, S., Andelman, F., Neufeld, M.Y., Kramer, U., Ursin, R. & Sterman, M. (1981). Manual for Standardized Scoring of Sleep Arieli, A., Fried, I. & Malach, R. (2008) Interhemispheric correlations of and Waking States in Adult Cats. BIS/BRI, University of California, Los 45 slow spontaneous neuronal fluctuations revealed in human sensory cortex. Angeles. 46 Nat. Neurosci., 11, 1100–1108. Velik, R. (2009) From single neuron-firing to consciousness – towards the 47 Nunez, P.L., Srinivasan, R., Westdorp, A.F., Wijesinghe, R.S., Tucker, true solution of the binding problem. Neurosci. Biobehav. Rev., 34, 993– 48 D.M., Silberstein, R.B. & Cadusch, P.J. (1997) EEG coherency. I: statis- 1001. 49 tics, reference electrode, volume conduction, Laplacians, cortical imaging, Voss, U., Holzmann, R., Tuin, I. & Hobson, J.A. (2009) Lucid dreaming: a and interpretation at multiple scales. Electroencephalogr. Clin. Neurophys- state of consciousness with features of both waking and non-lucid dream- 50 iol., 103, 499–515. ing. Sleep, 32, 1191–1200. 51 Perez-Garci, E., del-Rio-Portilla, Y., Guevara, M.A., Arce, C. & Corsi-Cabre- Watson, C., Baghdoyan, H. & Lydic, R. (2010). A neurochemical perspec- 52 ra, M. (2001) Paradoxical sleep is characterized by uncoupled gamma activ- tive on states of consciousness. In Hudetz, A.G. & Pearce, R. (Eds), Su- 53 ity between frontal and perceptual cortical regions. Sleep, 24, 118–126. pressing the Mind, Contemporary Clinical Neuroscience. Humana Press, Rechtschaffen, A. (1978) The single-mindedness and isolation of dreams. New York, pp. 33–79. 54 Sleep, 1,97–109. Yuval-Greenberg, S., Tomer, O., Keren, A.S., Nelken, I. & Deouell, L.Y. 55 Rieder, M.K., Rahm, B., Williams, J.D. & Kaiser, J. (2010) Human gamma- (2008) Transient induced gamma-band response in EEG as a manifestation 56 band activity and behavior. Int. J. Psychophysiol., 79,39–48. of miniature saccades. Neuron, 58, 429–441. 57 58 59 60 61 62

© 2013 Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience,1–10 Graphical Abstract The contents of this page will be used as part of the graphical abstract of html only. It will not be published as part of main article.

During cognitive processes there are extensive interactions between various regions of the cerebral cortex. Oscillations in the gamma frequency band ( 40 Hz) of the elecroencephalogram (EEG) are involved in the binding of spatially separated but temporally correlated neural events, which results in a uniﬁed perceptual experience. Author Query Form

Journal: EJN

Article: 12143

Dear Author, During the copy-editing of your paper, the following queries arose. Please respond to these by marking up your proofs with the necessary changes/additions. Please write your answers on the query sheet if there is insufﬁcient space on the page proofs. Please write clearly and follow the conventions shown on the attached corrections sheet. If returning the proof by fax do not write too close to the paper's edge. Please remember that illegible mark-ups may delay publication.

Many thanks for your assistance.

Query reference Query Remarks 1 AUTHOR: Please check that authors and their afﬁliations are correct.

2 AUTHOR: Please abbreviaton list if applicable.

EJN 12143 Dispatch: 22.1.13 Journal: EJN CE: Nivedha Journal Name Manuscript No. B Author Received: No. of pages: 12 PE: Indumathi

O n c e y o u h a v e A c r o b a t R e a d e r o p e n o n y o u r c o m p u t e r , c l i c k o n t h e C o m m e n t t a b a t t h e r i g h t o f t h e t o o l b a r :

h i s w i l l o p e n u p a p a n e l d o w n t h e r i g h t s i d e o f t h e d o c u m e n t . h e m a j o r i t o f

T T y

t o o l s o u w i l l u s e f o r a n n o t a t i n g o u r p r o o f w i l l e i n t h e A n n o t a t i o n s s e c t i o n ,

y y

p i c t u r e d o p p o s i t e . W e ’ v e p i c k e d o u t s o m e o f t h e s e t o o l s e l o w :

S t r i k e s a l i n e t h r o u g h t e x t a n d o p e n s u p a t e x t S t r i k e s a r e d l i n e t h r o u g h t e x t t h a t i s t o e

o x w h e r e r e p l a c e m e n t t e x t c a n e e n t e r e d . d e l e t e d .

b b

i g h l i g h t a w o r d o r s e n t e n c e . i g h l i g h t a w o r d o r s e n t e n c e . H

‚ H ‚

R e p l a c e ( I n s ) i c o n i n t h e A n n o t a t i o n s S t r i k e t h r o u g h ( D e l ) i c o n i n t h e l i c k o n t h e l i c k o n t h e C

‚ C ‚

s e c t i o n . A n n o t a t i o n s s e c t i o n .

p e t h e r e p l a c e m e n t t e x t i n t o t h e l u e o x t h a t

b b y

‚ T

a p p e a r s .

i g h l i g h t s t e x t i n e l l o w a n d o p e n s u p a t e x t

a r k s a p o i n t i n t h e p r o o f w h e r e a c o m m e n t

H y

o x w h e r e c o m m e n t s c a n e e n t e r e d .

b b

n e e d s t o e h i g h l i g h t e d .

i g h l i g h t t h e r e l e v a n t s e c t i o n o f t e x t .

l i c k o n t h e A d d s t i c k n o t e i c o n i n t h e

H y

‚ ‚ C

A n n o t a t i o n s s e c t i o n .

l i c k o n t h e A d d n o t e t o t e x t i c o n i n t h e

‚ C

A n n o t a t i o n s s e c t i o n .

l i c k a t t h e p o i n t i n t h e p r o o f w h e r e t h e c o m m e n t

‚ C

s h o u l d e i n s e r t e d .

p e i n s t r u c t i o n o n w h a t s h o u l d e c h a n g e d

b y

‚ T

r e g a r d i n g t h e t e x t i n t o t h e e l l o w o x t h a t

p e t h e c o m m e n t i n t o t h e e l l o w o x t h a t

y y

‚ T

a p p e a r s .

I n s e r t s a n i c o n l i n k i n g t o t h e a t t a c h e d f i l e i n t h e I n s e r t s a s e l e c t e d s t a m p o n t o a n a p p r o p r i a t e

a p p r o p r i a t e p a c e i n t h e t e x t . p l a c e i n t h e p r o o f .

l i c k o n t h e A t t a c h F i l e i c o n i n t h e A n n o t a t i o n s l i c k o n t h e A d d s t a m p i c o n i n t h e A n n o t a t i o n s C

‚ C ‚

s e c t i o n . s e c t i o n .

l i c k o n t h e p r o o f t o w h e r e y o u ’ d l i k e t h e a t t a c h e d e l e c t t h e s t a m p y o u w a n t t o u s e . ( T h e A p p r o v e d S

‚ C ‚

f i l e t o b e l i n k e d . s t a m p i s u s u a l l y a v a i l a b l e d i r e c t l y i n t h e m e n u t h a t

a p p e a r s ) .

e l e c t t h e f i l e t o b e a t t a c h e d f r o m y o u r c o m p u t e r

‚ S

o r n e t w o r k . l i c k o n t h e p r o o f w h e r e y o u ’ d l i k e t h e s t a m p t o

‚ C

a p p e a r . ( W h e r e a p r o o f i s t o b e a p p r o v e d a s i t i s ,

e l e c t t h e c o l o u r a n d t y p e o f i c o n t h a t w i l l a p p e a r

‚ S

t h i s w o u l d n o r m a l l y b e o n t h e f i r s t p a g e ) .

i n t h e p r o o f . l i c k O K .

A l l o w s s h a p e s , l i n e s a n d f r e e f o r m a n n o t a t i o n s t o b e d r a w n o n p r o o f s a n d f o r

c o m m e n t t o b e m a d e o n t h e s e m a r k s . .

l i c k o n o n e o f t h e s h a p e s i n t h e D r a w i n g

‚ C

s e c t i o n . M a r k u p s

l i c k o n t h e p r o o f a t t h e r e l e v a n t p o i n t a n d

‚ C

d r a w t h e s e l e c t e d s h a p e w i t h t h e c u r s o r .

o a d d a c o m m e n t t o t h e d r a w n s h a p e ,

‚ T

m o v e t h e c u r s o r o v e r t h e s h a p e u n t i l a n

a r r o w h e a d a p p e a r s .

o u b l e c l i c k o n t h e s h a p e a n d t y p e a n y

‚ D

t e x t i n t h e r e d b o x t h a t a p p e a r s .