- Email: [email protected]
Asleep but aware?
Brain and Cognition 87 (2014) 7–15
Contents lists available at ScienceDirect
Brain and Cognition journal homepage: www.elsevier.com/locate/b&c
Asleep but aware? Stéphanie Mazza a,⇑, Caroline Perchet b,c, Maud Frot b,c, George A. Michael a, Michel Magnin b,c, Luis Garcia-Larrea b,c, Hélène Bastuji b,c,d a
Laboratoire d’Etude des Mécanismes Cognitifs, Université Lumière, Lyon 2, 5 Avenue Pierre Mendes France, Bron F-69676, France INSERM, U1028, CNRS, UMR 5292, Centre de Recherche en Neurosciences de Lyon, Laboratoire «intégration centrale de la douleur», 59 bd Pinel, Bron Cedex F-69677, France c Université Claude Bernard, 43 Boulevard du 11 Novembre 1918, Lyon 1 F-69003, France d Hospices Civils de Lyon, Unité d’Hypnologie, Hôpital Neurologique, 59 Bd Pinel, Bron F-69677, France b
a r t i c l e
i n f o
Article history: Accepted 16 February 2014 Available online 13 March 2014 Keywords: Paradoxical sleep Perception Consciousness Nociception Intracerebral EEG
a b s t r a c t Despite sleep-induced drastic decrease of self-awareness, human sleep allows some cognitive processing of external stimuli. Here we report the fortuitous observation in a patient who, while being recorded with intra-cerebral electrodes, was able, during paradoxical sleep, to reproduce a motor behaviour previously performed at wake to consciously indicate her perception of nociceptive stimulation. Noxious stimuli induced behavioural responses only if they reached the cortex during periods when mid-frontal networks (pre-SMA, pre-motor cortex) were pre-activated. Sensory responses in the opercular cortex and insula were identical whether the noxious stimulus was to evoke or not a motor behaviour; conversely, the responses in mid-anterior cingulate were specifically enhanced for stimuli yielding motor responses. Neuronal networks implicated in the voluntary preparation of movements may be reactivated during paradoxical sleep, but only if behavioural-relevant stimuli reach the cortex during specific periods of ‘‘motor awareness’’. These local activation appeared without any global sleep stage change. This observation opens the way to further studies on the currently unknown capacity of the sleeping brain to interact meaningfully with its environment. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Sleep has frequently been assimilated to a little death, to such an extent that the Greek mythology establishes a fraternal link between Hypnos, the God of Sleep and Thanatos the God of Death. The loss of behavioural control and the fading of consciousness are the distinctive features of sleep that underlie this phenomenon. The accompanying lack of responsiveness towards nocturnal stimulation led to hypothesize a functional disconnection between the cerebral cortex and the external world (Steriade, 1994). This disconnection is especially intriguing in the period of paradoxical sleep (PS), during which the electroencephalographic pattern is similar to that of waking state. In order to explain this paradox, Foulkes (1966) and later Llinas and Ribary (1993) suggested that PS ‘‘is a state of hyperattentiveness during which sensory input cannot address the machinery that generates conscious experience’’. They hypothesized that the oniric activity integrally focuses the attention of the sleeper, excluding the possibility for external sources of stimulation to be
⇑ Corresponding author. Fax: +33 4 78 77 43 51. E-mail address: [email protected] (S. Mazza). http://dx.doi.org/10.1016/j.bandc.2014.02.007 0278-2626/Ó 2014 Elsevier Inc. All rights reserved.
processed. However, this view is challenged by the well-known fact that external events can interfere with, or be incorporated into, the ongoing oniric activity (Dement & Wolpert, 1958). Modern behavioural and neurophysiological studies indicate that the sleeper keeps the possibility to perceive and process external information. Accordingly brain responses related to cognitive activity (e.g. P300) have been recorded during PS in response to frequency- or intensity-deviant stimuli (Bastuji, Garcia-Larrea, Franc, & Mauguière, 1995; Cote & Campbell, 1999; Takahara, Nittono, & Hori, 2006; Macdonald, Jamshidi, & Campbell, 2008), and likewise to more complex stimuli such as the subject’s own name (Perrin, Garcia-Larrea, Mauguiere, & Bastuji, 1999; Pratt, Berlad, & Lavie, 1999). The persistence of a differential brain response (‘‘N400 effect’’) to incongruous words during PS further indicates that the sleeper’s brain keeps some capacity to analyse semantic contents (Brualla, Romero, Serrano, & Valdizan, 1998), even if linguistic absurdity appears to be accepted to a greater extent than during waking (Perrin, Bastuji, & Garcia Larrea, 2002). These results confirm that the sleeping brain remains able to detect and categorize some particular traits of external stimulus significance, even if the processes engaged are not identical to those implicated during wakefulness.
8
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
Notwithstanding the importance of the above work, the question of consciousness of external events during sleep remains unanswered. An empirical criterion used to assess the awareness of information is its ‘‘reportability’’ (Dehaene & Naccache, 2001; Gazzaniga, LeDoux, & Wilson, 1977; Weiskrantz, 1997) defined as ‘‘all voluntary communicative acts that are used to report conscious content’’ (Baars, 1997). In the context of sleep, Laberge (2000) has shown that ‘‘lucid dreamers’’ can voluntarily produce eye movements, to indicate that they are aware of dreaming, during ongoing PS. It has also been demonstrated that operant responses learned during wakefulness may be reactivated by stimuli presented during sleep (Bonnet, 1982; Burton, Harsh, & Badia, 1988; Granda & Hammack, 1961; Lauerma, Kaartinen, Polo, Sallinen, & Lyytinen, 1994). Evans, Gustafson, O’Connell, Orne, and Shor (1970) succeeded to observe complex behavioural responses to verbal suggestion during PS (i.e. ‘‘Whenever I say the word ‘‘itch,’’ your nose will feel itchy until you scratch your nose’’) without sleep interruption. These behaviours remain rare and inconsistent suggesting that stimulus awareness during PS may be fluctuating. In parallel to this, a critical and unsolved question concerns, the brain regions or activity supporting the production of these apparently ‘intentional’ responses while EEG recordings indicate unequivocal PS sleep, characterised by atonia and fading consciousness. Using brain imaging, Dresler and colleagues recently showed that dreamed motor actions performed during PS elicited neuronal activation in the sensorimotor cortex and the SMA, comparable to those observed during movements really acted during wakefulness (Dresler et al., 2011). Here we describe the results of intracerebral EEG investigation that fortuitously captured a spontaneous reactivation of an actual motor behaviour previously used during wakefulness as an indicator of conscious perception of nociceptive stimulation in a patient affected by drug-resistant epilepsy recorded during presurgical assessment. Intra-cerebral recordings demonstrated that preactivation of mid-frontal networks was determinant to allow this adapted behaviour to an external input, whereas activity in sensory cortices appeared irrelevant and global sleep stage unchanged. 2. Materials and methods The subject was a 37-year-old women involved in a study protocol evaluating nociceptive processes during sleep using laser stimulations (Bastuji et al., 2011). She presented with refractory epilepsy and was being investigated using stereotactically implanted intracerebral electrodes before functional surgery. Depth EEG recording electrodes were implanted according to the stereotactic technique of Talairach and Bancaud (1973). The cortical targets were identified on the patient’s MRI. The implantation procedure has been described in detail elsewhere (Frot & Mauguière, 1999; Frot, Rambaud, Guénot, & Mauguière, 1999; Guenot et al., 2001). The patient was fully informed about the fact that Laser EvokedPotential (LEP) recordings during sleep were not a part of the diagnostic procedure but were performed with research purposes, and gave her written informed consent. The laser stimulation paradigm was submitted to, and approved by, the local Ethics Committee (CCPPRB Léon Bérard-Lyon) and the study was promoted by the French National Agency for Medical Research (INSERM).
the right hand (radial territory contralateral to the hemisphere of electrode implantation). The pain threshold was determined at wake. The patient was instructed to lift her left index finger to indicate that she had perceived the stimulation and score its intensity using a Likert-type scale where 0 was ‘‘no sensation’’, 8 = unbearable pain, and 4 was defined as ‘‘pricking, moderately painful’’). An energy level of 80 mJ/mm2, yielded for this patient the expected painful sensation defined as ‘‘pricking, moderately painful’’. After that, ten stimulations, at this pain threshold, were delivered to obtain wakefulness LEP responses. During LEP recordings, the patient was lying, immobile. Then, the patient was allowed to sleep at her own time. During the night, series of up to 30 nociceptive laser pulses were delivered when definitive stages of sleep were reached, as determined online by an experienced sleep researcher. The staging was later confirmed off-line. The laser intensity was kept stable during the whole night (80 mJ/mm2). Laser pulses were transmitted through the optic fibre from the laser stimulator. This 10-m optical fibre connecting the laser generator with the stimulating probe allowed stimulating conveniently the hand dorsum despite subject position during the night. Both the sleeping subject and the stimulating investigator wore eye protections. After preliminary work showing that delivering stimuli at short (<6 s) and constant intervals increased the probability of awakening, inter-stimulus interval (ISI) was pseudo-randomly adjusted on-line and varied between 10 and 20 s. The laser beam was slightly moved over the skin surface between two successive stimuli to avoid habituation and especially peripheral nociceptor fatigue (Schwarz, Greffrath, Buèsselberg, & Treede, 2000). Due to environmental bad conditions (noisy storm), this night session ended prematurely. A second night of stimulation was planned one week later, during which, the emergence of behavioural responses was observed. This night recording was conducted 14 days after electrodes implantation; at that time, anticonvulsant drug intake was drastically reduced for at least one week. EEG recordings were performed with sampling frequency of 256 Hz and a 0.03–100 Hz bandpass (MicromedÒ, Mâcon, France) in referential mode (the reference electrode being an implanted contact located in the skull). In order to limit the lack of comfort induced by the monitoring device, no submental EMG electrodes were placed, but only EKG and 2 EOG electrodes (supero- and infero-lateral right canthus). EEG, EKG and EOG were recorded continuously during the night and stored for off-line analysis. In order to stimulate at wellidentified sleep stages, the different states of vigilance were visually identified on-line, according to the criteria of Rechtschaffen and Kales (1968) adapted to intracranial recordings (see Magnin, Bastuji, Garcia-Larrea, & Mauguiere, 2004; Rey et al., 2007). 2.2. Incorporation of stimuli to dreams In the morning that followed the recording night, a four-level scale adapted from Zimmerman (1970) was used to assess the degree of stimulus incorporation to dreams. The four-levels were (1) no incorporation, the subject had no recall of somatosensory or pain stimulation; (2) possible incorporation, the subject had some somatosensory or pain recall, but unrelated to the stimulations delivered; (3) obvious incorporation, the subject had some somatosensory or pain recall related to the stimulations delivered; (4) awareness of stimulation effectively delivered.
2.1. Laser stimulation procedure and parameters
2.3. Anatomical localization of recording sites
Nociceptive heat pulses of 5 ms were delivered with a Nd:YAPlaser (Yttrium Aluminium Perovskite; wavelength 1.34 lm; El.En.Ò). Series of laser stimuli were delivered on the dorsum of
The final position of each implanted contact with respect to the targeted anatomical structures was verified via frontal and sagittal X-rays at scale 1, using an X-ray source at 4.85 m from the patient’s
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
head thus eliminating the linear enlargement due to X-ray divergence. The T1-weighted MRI midsagittal slice, also at scale 1, was superimposed on the sagittal X-ray using the limits of the cranial bone as a common reference. This allowed to define the position of the anterior commissures (AC) and posterior commissures (PC) and to calculate the anteroposterior and dorsoventral coordinates of each electrode contact. The mediolateral coordinate of each contact was determined on the frontal X-ray with respect to the midsagittal plane. Each contact was then localized in the Talairach space (Talairach & Tournoux, 1988) according to the following 3 dimensional (3-D) coordinates in millimetres: x = distance from the interhemispheric sagittal vertical plane; y = distance from the coronal plane perpendicular to the horizontal AC–PC plane passing through AC; z = distance from horizontal AC–PC plane. The origin of this 3D reference system corresponds to the AC location as seen on the midsagittal plane. The 3-D position of all contacts within the patient’s cerebral structures was checked by plotting their coordinates in her MRI volume (T1-weighted images, 1-mm based voxel), reconstructed in the 3-D space using MRIcro Software (Rorden & Brett, 2000). This patient had 8 implanted electrodes (90 contacts in the grey matter) (Fig. 1).
9
EEG data were segmented into epochs ranging from 100 ms before the stimulation to 1000 ms after. Each EEG epoch was baseline corrected using the 100 ms pre-stimulus interval as reference. Trials contaminated by epileptic transient activities were removed. The LEP components on the insular, opercular and cingulate corteces were identified as the peaks with the highest amplitude within a 160–420 ms latency window encompassing the corresponding waveform (Frot, Mauguière, Magnin, & Garcia-Larrea, 2008). The components were analysed when their amplitude value during wakefulness exceeded the mean prestimulus baseline by at least 3 SD. They were labelled C1 and C2 according to their latency (Bastuji et al., 2011). Mean voltages and latencies are given ±1 SD. LEPs obtained during wakefulness were compared to those obtained during PS, separately whether they were or not followed by an overt finger movement (PS with behaviour vs PS without behaviour). Latencies and amplitudes of LEPs were tested using one-way analyses of variance (ANOVA) with ‘‘vigilance state’’ as factor (waking vs PS with behaviour vs PS without behaviour). Tukey-Kramer tests were used for post hoc comparisons.
2.4.2. EEG analysis in the time-frequency domain: continuous wavelet transform
2.4. Data analysis Sleep stages were visually identified using more than 20 cerebral contacts including the dorso-lateral extent of the frontal, parietal and temporal lobes, and the thalamus, with both bipolar and referential traces plus EOG. Sleep scoring was done according to the criteria of Rechtschaffen and Kales (1968) and thalamic activity (see Magnin et al., 2004; Rey et al., 2007), so as to derive hypnograms based on 30 s epochs and determine the vigilance state during which laser stimuli were delivered. Data reported in this study concern exclusively stimulations delivered during the waking session and the last PS at 06:20. Epochs of 10 s prior and 5 s after stimuli delivered during PS were classified as occurring into a tonic (without rapid eye movement) or a phasic period (with rapid eye movements) (Sallinen, Kaartinen, & Lyytinen, 1996). The occurrence of EEG arousing reactions or full awakenings was controlled online considering raw cortical EEG activities. In absence of EMG recording, arousals were defined as an abrupt shift of EEG activities towards higher frequencies (alpha, theta and/or >16 Hz) that lasted at least 3 s and less than 15 s only (AASM: Iber, Ancoli-Israel, Chesson, & Quan, 2007). These frequency modifications were considered as stimulus-related if they occurred within 15 s after stimulus onset. A full awakening was scored when this cortical reaction lasted more than 15 s. The delay separating each finger lift from the stimulus was calculated using the video recording synchronised with the EEG. 2.4.1. Intracranial laser-evoked potentials (LEPs) LEPs analyses were performed using ASAÒ software (Advanced Neuro Technology (ANT), Eschende, The Netherlands). Continuous
– In order to determine whether a specific EEG activation pattern was associated to the production of a behavioural response, a time-frequency (TF) analysis of the EEG preceding each stimulus was performed, for all contact localisations (Fig. 1) (Letswave software; Mouraux & Iannetti, 2008). Wavelet transforms are particularly suitable for the analysis of nonstationary signals in the time–frequency domain. Thus, the continuous wavelet TF transform was applied to each individual epoch in each brain structure explored, using a time window analysis of 2 s before the stimulus (stimulus-referenced). The resulting TF amplitude maps were then averaged for the condition with behaviour and without behaviour. Results were displayed as an increase or decrease of oscillation amplitude after baseline correction and were analysed separately for each standard EEG frequency bands: delta (1.5–4.5 Hz), theta (4.75–7.75 Hz), alpha (8–12 Hz), sigma (12.25–15 Hz), beta (15.25–30 Hz) and gamma (30.25–45 Hz). TF amplitude values of epochs preceding or not a behaviour were compared with the Wilcoxon test for paired data. Pairing was achieved through the ordering procedure proposed by Michael, Boucart, Degreef, and Godefroy (2001), already used in behavioural case studies, which avoids masking of effects due to the random and unpredictable post-stimulus appearance of the behavioural response. In each condition (behavioural response or no behavioural response), amplitude values were ordered from the smallest to the largest ones for each EEG frequency band. Then, the smallest value in one condition was paired with the smallest one in the other condition. The same procedure was repeated sequentially for the two series of increasing values.
Fig. 1. Electrode implantation. This figure illustrates the electrode implantation sites on the lateral view of the patient’s 3D MRI, the anatomical structures and the corresponding Brodmann areas explored by each electrode.
10
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
– TF analyses were also performed using a time window locked on the onset of the behavioural response (response-referenced), EEG epochs corresponding to the 2 s preceding the beginning of each motor response. Continuous wavelet transform was applied to each epoch after a baseline correction and then averaged. Within average TF map obtained, the highest amplitude value (Vmax) was identified. A region of interest (ROI) was centred on the location of the Vmax. Its limits circumscribed a zone in which amplitude values were at least 60% of the Vmax and thus considered as significantly increased. Then, a ROI of same surface was applied to each TF single map composing the TF average map at the same coordinates in terms of frequency and time. To verify if activations defined by ROIs were specific to behavioural responses, identical ROIs were applied to TF maps obtained when behavioural responses were absent. To take into account the latency jitter observed between the laser stimulation and the behavioural response the ROIs were placed in each epoch without behavioural responses at variable times with respect to the laser stimulation, reproducing thus the delays observed when the behavioural response was present (see Fig. 2). As mentioned above a Wilcoxon test for paired data was performed on average amplitude value within the ROIs to determine regions responding differently between stimulations generating or not behavioural response. The Average amplitude values within the ROIs are given ±1 SEM. The level of significance was set at 0.05 for all statistical analyses.
3. Results 3.1. Behavioural responses At the end of a whole night of laser stimulation, the patient unexpectedly lifted her left index finger in response to 11 of the
20 laser stimuli delivered during PS at 06:20. These 11 behavioural responses were not obtained after consecutive stimulations, but dispersed within the stimulation series (Fig. 3B, see movie). All the stimuli were delivered during tonic PS, attested by EOG, cortical activity and the presence of a concomitant delta activity at the thalamic level. These stimulations never induced any global sleep stage shift. When visually compared using video recording, the latencies of finger movements observed during PS were not different from those of the motor responses observed during wakefulness (mean latency after laser stimulation = 840 ms (range: 460–1670 ms) vs 1008 ms (range 470–1750) respectively; p > 0.05). However, while the amplitudes of movements appeared similar, movements were performed at slower velocity during the PS. Among these 11 stimulations followed by a movement, 8 were followed by an obvious micro-arousal identified through changes in EEG cerebral activities appearing at a mean latency of 1400 ms (range: 980–1810 ms). The micro-arousal always lasted less than 15 s and never gave rise to further sleep stage shifts. One microarousal only could be observed among the 9 stimulations not followed by a motor response. There was neither incorporation of nociceptive stimuli into dreams content nor recollection of pain stimulation on the following morning (questionnaire score = 1).
3.2. Laser evoked potentials (LEPs) Among the 30 laser stimulations analysed (20 delivered during PS, 10 during wakefulness), 3 were discarded due to epileptic activity. Thus, the LEPs results presented are based on 27 cortical responses, including 9 stimulations delivered during PS and followed by a behavioural response (7 preceding a micro-arousal), 9 during PS without behavioural response and 9 during wakefulness. In the insular cortex, ANOVA showed a significant effect of state of vigilance on LEP latencies (component C1: F(2, 24) = 4.74;
Fig. 2. Definition of region of interest (ROI) during the period of time preceding the apparition of a behavioural response. Within the average TF map locked on the onset of the behavioural response, a ROI (black scare) is centred on the highest amplitude value (Vmax = red cross) and circumscribes a zone in which amplitude values are at least 60% of the Vmax. ROI of same surface is applied to each TF single map composing the TF average map: same frequency coordinates (Fmin and Fmax) and time (Tmin and Tmax). For each TF single map the delay between the laser stimulation and the Tmin is calculated (in the example: x1, x2, x3, . . .). ROIs are placed in each epoch without behavioural responses reproducing the delays observed in the condition with behavioural response (in the example: x1, x2, x3, . . .). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
11
Fig. 3. Laser evoked potentials. (A) Averaged laser evoked potentials record in the ‘pain matrix’, in a bipolar recording mode from two adjacent contacts (red crosses). Negative potentials at the intracortical recording site are represented upward. For each structure, the Talairach coordinates of the contacts are indicated and illustrated on the coronal patient’s MRI. (B) Infra-red video images showing the patient during PS after two laser nociceptive stimulations on the right hand: on the left panel the stimulation gives rise to a left finger lift (red arrow) while, on the right panel no motor response follows the stimulation. Middle panel: enlargement of the C1 component recording during waking, PS with and PS without a behavioural response, in the anterior cingulate gyrus. Asterisk indicates the statistically significant difference between PS without motor response as compared with the others. (C) The five sets of raw intracranial recordings illustrate the frontal (a), temporal (b) and parietal (c and d) cortical and concomitant thalamic (e) PS activities. Laser stimulations were deliverated during delta oscillations in the thalamus. Asterisk: micro-arousal. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
p = 0.02 and C2: F(2, 24) = 17.75; p = 0.0001) and amplitudes for component C2 (component C1: F(2, 24) = 0.3; p = n.s. and component C2: (F(2, 24) = 4.35; p = 0.02). Post-hoc analysis specified that, as compared to waking, LEPs obtained during PS had delayed latencies and decreased amplitudes. Further statistical analysis did not reveal any differences between the two PS conditions, i.e., with or without behavioural response. In the opercular cortex, similar results were obtained with a significant effect of state of vigilance for latencies (components C1: F(2, 24) = 10.28; p = 0.0006) and C2: F(2, 24) = 15.96; p = 0.0003) and amplitudes (component C1: F(2, 24) = 4.74; p = 0.02 and C2: F(2, 24) = 3.82; p = 0.036). During the 2 PS conditions the LEPs were delayed and attenuated as compared to waking and no differences were observed between LEPs during PS with or without motor responses. In the anterior cingulate gyrus, ANOVA showed a significant effect of state of vigilance on latencies for C2 component (component C1: F(2, 24) = 0.54; p = ns and C2: F(2, 24) = 25.56; p < 0.001) and amplitudes (component C1: F(2, 24) = 5.77; p = 0.009 and C2: F(2, 24) = 8.55; p = 0.002). The LEP latencies were different with the following gradient: wake = PS without behaviour < PS with behaviour. The amplitudes of C1 obtained during PS with motor behaviour were significantly enhanced relative to those in periods with no overt response, making them similar to waking LEPs (Fig. 3B). Component C2 had greater amplitude during the waking state as compared to either PS condition.
To summarise, opercular and insular LEPs remained similar during PS whether the stimulus evoked a motor response or not. In contrast, nociceptive stimuli giving rise to a finger movement entailed a significant increase of cingulate LEPs, which became similar to those obtained during wakefulness (see Fig. 3 and Table 1). 3.3. EEG activity preceding laser stimuli During the pre-stimulus interval ( 2 to 0 s), time-frequency analyses did not reveal any significant statistical difference between the two PS conditions, with or without behavioural response, except for the pre-supplementary motor area (pre-SMA) and the premotor area (BA6) (Fig. 4A and B). In the pre-SMA, low frequencies were reduced by 1.55 lV for delta (t(8), p < 0.02), 4.87 lV for theta (t(8), p < 0.004), and 1.20 lV for alpha (t(8), p < 0.006), when followed by a behavioural response. In contrast, pre-stimulation high frequencies were enhanced by 0.90 lV for beta (t(8), p < 0.004). In BA6, pre-stimulus EEG activity was significantly reduced in theta and alpha frequency-bands ( 3.40 lV (t(8), p < 0.004, and 0.99 lV (t(8), p < 0.001) and increased in beta and gamma frequency-bands (0.6 lV (t(8), p < 0.004, and 0.4 lV (t(8), p < 0.006) when a finger lift occurred. In the cingulate, no significant difference in the 1.5–45 frequency band could be observed (Fig. 4C), despite the LEPs modification described.
12
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
Table 1 Laser evoked potential (LEP) values recorded in the 3 cortical areas. Latencies (mean ± SD) and amplitudes (mean ± SD) values of the two LEP components in the 3 structures of the pain matrix. Post-hoc analysis following statistically significant ANOVA are presented in the three conditions: wake (1), PS with behaviour (2), PS without behaviour (3). Component 1
Component 2
Latency (ms)
Post hoc
Amplitude (lV)
Post hoc
Latency (ms)
Post hoc
Insula Wake (1) PS with behaviour (2) PS without behaviour (3)
Amplitude (lV)
178 ± 11 198 ± 21 196 ± 13
1<2=3
109 ± 22 101 ± 46 95 ± 40
NS
273 ± 6 345 ± 27 339 ± 13
1<2=3
107 ± 26 76 ± 26 68 ± 35
1>2=3
Operculum Wake (1) PS with behaviour (2) PS without behaviour (3)
179 ± 11 206 ± 8 211 ± 9
1<2=3
91 ± 15 30 ± 15 27 ± 13
1>2=3
264 ± 17 371 ± 24 354 ± 14
1<2=3
57 ± 33 31 ± 15 38 ± 13
1>2=3
Anterior cingulate Wake (1) PS with behaviour (2) PS without behaviour (3)
250 ± 12 248 ± 8 254 ± 19
NS
81 ± 35 63 ± 32 30 ± 13
1=2>3
308 ± 12 365 ± 26 314 ± 13
1=3<2
64 ± 25 26 ± 9 24 ± 31
1>2=3
Post hoc
NS: non significant. Values resulting from mean LEP responses in the pain matrix.
Fig. 4. Time-frequency maps preceding the laser stimulation. Average TF amplitude maps stimulus-referenced in the pre-SMA, premotor area, and anterior cingulate gyrus. Left panel illustrates the location and Talairach coordinates of the contacts on the 3 planes of the patient’s MRI for each structure. For each region the Talairach coordinates of the two contacts ensuring bipolar recording are indicated. Average TF amplitude maps (middle panel) display increase and decrease oscillations amplitude 2s before the laser stimulation, for the condition with (top) and without (bottom) behavioural response, in the pre-SMA, the premotor area and the anterior cingulate gyrus. Graphs on the right illustrate the amplitude in each frequency band for the condition with (with squares) and without (grey circles) behavioural response in the 3 structures. Asterisk indicates the statistically significant differences.
To summarise, stimuli giving rise to motor behaviour were preceded by a significant amplitude reduction in low frequencies and increase in high-frequency activities in the pre-supplementary motor area (pre-SMA) and the premotor area (BA6). 3.4. EEG activity preceding behaviours Average time frequency maps presented in Fig. 5 illustrate cortical activities during the interval of time between the
nociceptive stimulus and the occurrence or the absence of finger lift. When a behavioural response occurred, a specific and significant enhancement of cortical activities in the 7–11 Hz ROI appeared 790 ms before finger movements in the parietal cortex (BA40) (23.07 ± 6.09 lV with behaviour vs 0.7 ± 3.26 lV without behaviour, t(8) p < 0.004). This activation was followed by a similar activity increase in the pre-SMA with a delay of 450 ms, in a 5–8 Hz ROI, (12.69 ± 2.92 lV vs 1.25 ± 4.02 lV, t(8) p < 0.004).
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
13
Fig. 5. Time-frequency maps preceding the apparition of a behavioural. Average TF maps illustrate activity levels in cortical regions showing a significant difference during the time interval between the nociceptive stimulus and the occurrence (top) or the absence (bottom) of a behavioural response. For each region the Talairach coordinates of the two contacts ensuring bipolar recording are indicated. Squares illustrate regions of interest (ROIs) in which activity levels are significantly different after the stimulations leading or not to a behavioural response. Values in TF maps correspond to the mean frequency amplitude ((lV) ± SEM) defined within the ROIs. Asterisks indicate the statistically significant differences between PS with and without motor responses.
In the frontal area (BA11) an increased activity starting 400 ms before the motor response in a 2–7 Hz ROI which endured during finger lift execution (4.97 ± 0.97 lV vs 0.89 ± 1.42 lV, t(8) p < 0.004), while in the hippocampus the peak of ROI activity (7– 15 Hz) was limited and concomitant to the motor execution (17.95 ± 6.27 lV vs 1.64 ± 2.89 lV, t(8) p < 0.004). 4. Discussion The patient, presented herein, spontaneously lifted her index finger in response to 11 of the 20 nociceptive stimulations delivered during the last PS period of the night. This motor behaviour suggests that, during PS, this patient was not only able to detect nociceptive stimuli but also to initiate a motor sequence previously used at wake to indicate her perception. In parallel, intracerebral recordings uncovered activations of brain structures brought into play during this sensory-motor processing. Among the above electrophysiological results, two of them seem specific to the occurrence of a behavioural manifestation. First, LEP cingulate responses associated to a meaningful motor behaviour were selectively enhanced and became similar to those obtained during wakefulness. Second, the level of activation in motor and pre-motor networks was also increased before the stimulations giving rise to a motor behaviour. As these data have been obtained in an epileptic patient, one can wonder if they are more or less related to this pathophysiology. However, comparing cortical activities recorded during our experiment with those observed during the patient’s seizures and even during interictal periods leaves no doubt as to the physiological validity of the phenomena we observed. The possibility to generate a movement during the muscle atonia characterising PS remains puzzling. However, several data converge to consider that muscle atonia can be removed in various circumstances. Voluntary movements, such as eye movements, deep breath or complex motor movements, can be obtained from subjects in response to external stimulations during this sleep stage (Burton et al., 1988; Evans, Gustafson, O’Connell, Orne, & Shor, 1970; Laberge, 1990; Lauerma et al., 1994). PS dreams involving body movements can be associated with an increase in EMG activity of the corresponding muscle (Erlacher & Schredl, 2008). Furthermore, it has been demonstrated that, despite the general muscle blockade, brief bursts of phasic EMG activity can be recorded during PS, more frequently in distal than in proximal muscles of the limbs (Gardner & Grossman, 1975; Rye, 2003). This has been recently quantified, indicating that the strength of atonia is weaker in distal territories (Frauscher et al., 2008). All these elements demonstrate that motor inhibition during PS can be alleviated, allowing movements triggered by external or internal
stimuli, although their dynamics are partly impeded, as observed in our patient’s finger lifts. Intracerebral recordings have shown that while the lateral operculo-insular system supporting intensity coding and stimulus localization remains similarly active during PS and waking, the anterior cingulate cortex (ACC) involved in motor control, orienting and attention for action (Dum, Levinthal, & Strick, 2009; Frot et al., 2008; Vogt, 2005) is selectively depressed (Bastuji et al., 2011). In the present study, cingulate responses during PS were specifically restored before each finger lift, reaching amplitudes comparable to those of the waking state. This result confirms the direct access of spinothalamic input to cingulate motor networks, bypassing the lateral operculo-insular system. Besides, our results revealed that noxious stimuli induced behavioural responses only if they reached the cortex during PS sleep periods when motor and pre-motor networks (pre-SMA, and BA 6), were activated before stimulus arrival. These transitory increase in high frequency bands and decrease of low-frequency rhythms, which could be view as topographically circumscribed microarousals, may permit an appropriate motor response to be generated. In support to this hypothesis, intracranial EEG recording techniques have provided new and interesting information pointing to a local regulation of sleep. We recently showed that during sleep onset, the thalamus undergoes a deactivation process that precedes by several minutes the one occurring within the cortex. Moreover, we confirmed that the cortical deactivation also showed marked topographical differences (Magnin et al. 2010). Nobili and coworkers (2011) reported the coexistence of wake-like activity in the motor cortex and sleep-like EEG patterns in other cortical areas during the SWS. By capturing a confusional arousal episode in a subject recorded with intracerebral EEG, the same team documented the presence of a local arousal of the motor and cingulate cortices beginning before the confusional episode, while the associative cortices remained asleep (Terzaghi et al., 2009). Studying lucid dreams during PS, Voss, Holzmann, Tuin, and Hobson (2009) showed that lucidity, a period during which sleepers became able to voluntary interact with their environment, occurs in a state with features of both PS and waking, constituting an hybrid state of consciousness. Thus, we could speculate that local activity modification or local ‘‘awakening’’ during PS, could explain the possibility of sensory processing and behavioural responses, without any modification of the global sleep state. The behaviours we observed were preceded by the activation of cerebral regions involved in the voluntary preparation of movement, namely the pre-frontal and parietal cortices, and the caudal part of the pre-SMA (Deiber et al., 1991; Jenkins, Jahanshahi, Jueptner, Passingham, & Brooks, 2000). The latter is known to be specifically involved in self-initiated actions (Lau, Rogers, Ramnani, &
14
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
Passingham, 2004) and externally-driven movements (Cunnington, Windischberger, Deecke, & Moser, 2002). This pattern of activation may suggest that a sleeper is able to engage cerebral networks implicated in intentional movements. Dehaene and Naccache (2001) postulated that, during wakefulness, consciousness depends on the connectivity of a brain-scale workspace comprising the prefrontal cortex, the anterior cingulate, and the areas that connect to them. Despite a general state of frontal hypo-activity during PS (Maquet, 2000), Massimini et al. (2005) proposed that sleep is characterised by a dynamic variation of the effective connectivity among cortical areas. This connectivity, which is drastically reduced during SWS, is partially restored during PS, suggesting a potential resurgence of cortical integration and, accordingly, a resurgence of consciousness during this state (Massimini et al., 2010). In this context the local modulation in cerebral activities we observed before laser stimulation and the behaviour occurrence may participate to a kind of restored awareness during PS. Nevertheless, since cerebral activities reported in this case study were restricted to electrode implantation needed to define the epileptogenic area for surgical purposes, we do not know if other cerebral structures apart from implantation sites may contribute to this phenomenon. Other factors influencing responsiveness during PS may have participated to these behaviours. Lammers and colleagues described that responsiveness to external stimulations increases from the first to the second half of the night, reaching its maximum at early morning (05:30). This phenomenon, related to both homeostatic and circadian processes (Borbély, 1982), may have favoured the late occurrence of behavioural responses in the present case (at 06:20). Moreover, a mechanism of memory consolidation may also have contributed to these behaviours, since they were observed in the only patient who was tested twice at a one-week interval. Jouvet (1979), Nielsen, Kuiken, Alain, Stenstrom, and Powell (2004) have reported the reappearance in dreams of features from events occurring about a week prior, suggesting that processes with circaseptan morphology underlie dream incorporation. This delay may correspond to the sleepdependent redistribution of hippocampus-related memories to neocortical sites supporting memory consolidation (Nielsen & Stenstrom, 2005). In our study, the patient did not report neither incorporation into dreams of nociceptive stimuli, nor even recollection of pain stimulation on the morning questionnaire. The fact that cortical activations remain temporally and topographically limited may explain the lack of insight, a function requiring a much larger cortical network (Dehaene & Naccache, 2001). Nevertheless, this lack of recollection does not exclude some kind of temporally mnesic reactivation during PS, attested by the observed hippocampal activation, but remaining too limited to persist once awake. 5. Conclusions The present study suggests that during PS, not only the perception of sensory inputs but also the capacity for the sleeper to indicate his perception can, under particular circumstances, be preserved. A selective enhancement in motor and pre-motor networks associated with a re-activation of the anterior cingulate may allow an appropriate motor response to be generated. This case report opens the way to further studies on the currently unknown capacity of the sleeping brain to interact meaningfully with its environment. Acknowledgments The work was supported by the French National Agency for Medical Research (INSERM), the university of Lyon 2 (dispositif
incitation à la recherche), Fondation pour la Recherche Médicale (FRM), APICIL Foundation and NRJ Foundation of the Institut de France and by a Region Rhone-Alpes/France CIBLE 2010 grant. We are indebted to Drs J. Isnard and H. Catenoix for the opportunity to study their patient and to Dr M. Guénot for stereotactic electrode implantation. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bandc.2014. 02.007. References Baars, B. J. (1997). Some essential differences between consciousness and attention, perception and working memory. Consciousness and Cognition, 6, 363–371. Bastuji, H., Garcia-Larrea, L., Franc, C., & Mauguière, F. (1995). Brain processing of stimulus deviance during slow wave sleep and paradoxical sleep: A study of human auditory evoked responses using the oddball paradigm. Journal of Clinical Neurophysiology, 12, 155–167. Bastuji, H., Mazza, S., Perchet, C., Frot, M., Mauguière, F., Magnin, M., et al. (2011). Filtering the reality: Functional dissociation of lateral and medial pain systems during sleep in humans. Human Brain Mapping, 16. http://dx.doi.org/10.1002/ hbm.21390. Bonnet, M. H. (1982). Performance during sleep. In W. Webb (Ed.), Biological rhythms, sleep and performance (pp. 205–237). Chichester: John Wiley & Sons Ltd. Borbély, A. A. (1982). A two process model of sleep regulation. Human Neurobiology, 1, 195–204. Brualla, J., Romero, M. F., Serrano, M., & Valdizan, J. R. (1998). Auditory event-related potentials to semantic priming during sleep. Electroencephalography and Clinical Neurophysiology, 108, 283–290. Burton, S. A., Harsh, J. R., & Badia, P. (1988). Cognitive activity in sleep and responsiveness to external stimuli. Sleep, 11, 61–69. Cote, K. A., & Campbell, K. B. (1999). P300 to high intensity stimuli during REM sleep. Clinical Neurophysiology, 110, 1345–1350. Cunnington, R., Windischberger, C., Deecke, L., & Moser, E. (2002). The preparation and execution of self-initiated and externally-triggered movement: A study of event-related fMRI. Neuroimage, 15, 373–385. Dehaene, S., & Naccache, L. (2001). Towards a cognitive neuroscience of consciousness: Basic evidence and a workspace framework. Cognition, 79, 1–37. Deiber, M. P., Passingham, R. E., Colebatch, J. G., Friston, K. J., Nixon, P. D., Frackowiak, R., et al. (1991). Cortical areas and the selection of movement: a study with positron emission tomography. Experimental Brain Research, 84, 393–402. Dement, W., & Wolpert, E. A. (1958). The relation of eye movements, body motility, and external stimuli to dream content. Journal of Experimental Psychology, 55, 543–553. Dresler, M., Koch, S. P., Wehrle, R., Spoormaker, V. I., Holsboer, F., Steiger, A., et al. (2011). Dreamed movement elicits activation in the sensorimotor cortex. Current Biology, 21, 1833–1837. Dum, R. P., Levinthal, D. J., & Strick, P. L. (2009). The spinothalamic system targets motor and sensory areas in the cerebral cortex of monkeys. The Journal of Neuroscience, 29, 14223–14235. Erlacher, D., & Schredl, M. (2008). Do REM (lucid) dreamed and executed actions share the same neural substrate? International journal of dream research, 1, 7–14. Evans, F. J., Gustafson, L. A., O’Connell, D. N., Orne, M. T., & Shor, R. E. (1970). Verbally induced behavioural responses during sleep. Journal of Nervous and Mental Disease, 150, 171–218. Foulkes, D. (1966). The psychology of sleep. New York: Scribner. Frauscher, B., Iranzo, A., Högl, B., Casanova-Molla, J., Salamero, M., Gschliesser, V., et al. (2008). Quantification of electromyographic activity during REM sleep in multiple muscles in REM sleep behaviour disorder. Sleep, 31(5), 724–731. Frot, M., & Mauguière, F. (1999). Timing and spatial distribution of somato-sensory responses recorded in the upper bank of the sylvian fissure (SII area) in humans. Cerebral Cortex, 9, 854–863. Frot, M., Mauguière, F., Magnin, M., & Garcia-Larrea, L. (2008). Parallel processing of nociceptive A-delta inputs in SII and midcingulate cortex in humans. Journal of Neuroscience, 28, 944–952. Frot, M., Rambaud, L., Guénot, M., & Mauguière, F. (1999). Intracortical recordings of early pain-related CO2 laser evoked potentials in the human second somatosensory (SII) area. Clinical Neurophysiology, 110, 133–145. Gardner, R., Jr., & Grossman, W. I. (1975). Normal patterns in sleep in man. Advances in Sleep Research, 2, 67–107. Gazzaniga, M. S., LeDoux, J. E., & Wilson, D. H. (1977). Language, praxis, and the right hemisphere: Clues to some mechanisms of consciousness. Neurology, 27, 1144–1147. Granda, A. M., & Hammack, J. T. (1961). Operant behaviour during sleep. Science, 133, 1485–1486.
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15 Guenot, M., Isnard, J., Ryvlin, P., Fischer, C., Ostrovsky, K., Mauguiere, F., et al. (2001). Neurophysiological monitoring for epilepsy surgery: The Talairach SEEG method. Stereo Electro Encephalo Graphy. Indications, results, complications and therapeutic applications in a series of 100 consecutive cases. Stereotactic and Functional Neurosurgery, 77, 29–32. Iber, C., Ancoli-Israel, S., Chesson, A., & Quan, S. F. (2007). For the American academy of sleep medicine. In The AASM manual for the scoring of sleep and associated events: Rules, terminology and technical specifications. Westchester: American Academy of Sleep Medicine. Jenkins, I. H., Jahanshahi, M., Jueptner, M., Passingham, R. E., & Brooks, D. J. (2000). Self-initiated versus externally triggered movements. II. The effect of movement predictability on regional cerebral blood flow. Brain, 123, 1216–1228. Jouvet, M. (1979). Memories and ‘‘split brain’’ during dream. 2525 Memories of dream. Revue du Praticien, 29, 27–32. LaBerge, S. (2000). Lucid dreaming: Evidence and methodology. behavioural and Brain Sciences, 23, 962–963. LaBerge, S. (1990). Lucid dreaming: Psychophysiological studies of consciousness during REM sleep. In R. R. Bootzin, J. F. Kihlstrom, & D. L. Schacter (Eds.), Sleep and Cognition. Washington, DC: APA. Lammers, W. J., Badia, P., Hughes, R., & Harsh, J. (1991). Temperature, time-of-night of testing, and responsiveness to stimuli presented while sleeping. Psychophysiology, 28, 463–467. Lau, H. C., Rogers, R. D., Ramnani, N., & Passingham, R. E. (2004). Willed action and attention to the selection of action. Neuroimage, 21, 1407–1415. Lauerma, H., Kaartinen, J., Polo, O., Sallinen, M., & Lyytinen, H. (1994). Asymmetry of instructed motor response to auditory stimuli during sleep. Sleep, 17, 444–448. Llinas, R., & Ribary, U. (1993). Coherent 40-Hz oscillation characterizes dream state in humans. Proceedings of the National Academy of Sciences of the United States of America, 90, 2078–2081. Macdonald, M., Jamshidi, P., & Campbell, K. (2008). Infrequent increases in stimulus intensity may interrupt central executive functioning during Rapid Eye Movement sleep. NeuroReport, 19(3), 309–313. Magnin, M., Bastuji, H., Garcia-Larrea, L., & Mauguiere, F. (2004). Human thalamic medial pulvinar nucleus is not activated during paradoxical sleep. Cerebral Cortex, 14, 858–862. Magnin, M., Rey, M., Bastuji, H., Guillemant, P., Mauguiere, F., & Garcia-Larrea, L. (2010). Thalamic deactivation at sleep onset precedes that of the cerebral cortex in humans. Proceedings of the National Academy of Sciences of the United States of America, 23(107(8)), 3829–3833. Maquet, P. (2000). Functional neuroimaging of normal human sleep by positron emission tomography. Journal of Sleep Research, 9, 207–231. Massimini, M., Ferrarelli, F., Huber, R., Esser, S. K., Singh, H., & Tononi, G. (2005). Breakdown of cortical effective connectivity during sleep. Science, 309, 2228–2232. Massimini, M., Ferrarelli, F., Murphy, M., Huber, R., Riedner, B., Casarotto, S., et al. (2010). Cortical reactivity and effective connectivity during REM sleep in humans. Cognitive Neuroscience, 1, 176–183. Michael, G. A., Boucart, M., Degreef, J. F., & Godefroy, O. (2001). The thalamus interrupts top–down attentional control for permitting exploratory shiftings to sensory signals. NeuroReport, 12, 2041–2048. Mouraux, A., & Iannetti, G. D. (2008). Across-trial averaging of event-related EEG responses and beyond. Magnetic Resonance Imaging, 26, 1041–1054. Nielsen, T. A., Kuiken, D., Alain, G., Stenstrom, P., & Powell, R. A. (2004). Immediate and delayed incorporations of events into dreams: Further replication and implications for dream function. Journal of Sleep Research, 13, 327–336.
15
Nielsen, T. A., & Stenstrom, P. (2005). What are the memory sources of dreaming? Nature, 27, 1286–1289. Nobili, L., Ferrara, M., Moroni, F., De Gennaro, L., Russo, G. L., Campus, C., et al. (2011). Dissociated wake-like and sleep-like electro-cortical activity during sleep. Neuroimage, 15(58(2)), 612–619. Perrin, F., Bastuji, H., & Garcia Larrea, L. (2002). Detection of verbal discordances during sleep. NeuroReport, 13, 1345–1349. Perrin, F., Garcia-Larrea, L., Mauguiere, F., & Bastuji, H. (1999). A differential brain response to the subject’s own name persists during sleep. Clinical Neurophysiology, 110, 2153–2164. Pratt, H., Berlad, I., & Lavie, P. (1999). Oddball event-related potentials and information processing during REM and non-REM sleep. Clinical Neurophysiology, 110, 53–61. Rechtschaffen, A., & Kales, A. (1968). A manual of standardized terminology. In Techniques and Scoring System for Sleep Stages of Human Subjects. Bethesda: US Department of Health Education and Welfare. Rey, M., Bastuji, H., Garcia-Larrea, L., Guillemant, P., Mauguière, F., & Magnin, M. (2007). Human thalamic and cortical activities assessed by dimension of activation and spectral edge frequency during sleep wake cycles. Sleep, 30, 907–912. Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions. Behavioural Neurology, 12, 191–200. Rye, D. B. (2003). Modulation of normal and pathologic motoneuron activity during sleep. In S. Chokroverty, W. A. Hening, & A. S. Walters (Eds.), Sleep and Movement Disorders (pp. 94–119). Philadelphia: Butterworth Heinemann. Sallinen, M., Kaartinen, J., & Lyytinen, H. (1996). Processing of auditory stimuli during tonic and phasic periods of REM sleep as revealed by event-related brain potentials. Journal of Sleep Research, 5, 220–228. Schwarz, S., Greffrath, W., Buèsselberg, D., & Treede, R. D. (2000). Inactivation and tachyphylaxis of heat-evoked inward currents in nociceptive primary sensory neurones of rats. Journal of Physiology (London), 528, 539–549. Steriade, M. (1994). The thalamus and sleep disturbances. In C. Guilleminault, E. Lugaresi, P. Montagna, & P. Gambetti (Eds.), Fatal familial insomnia. Inherited prion diseases, sleep and the thalamus (pp. 177–189). New York: Raven Press. Takahara, M., Nittono, H., & Hori, T. (2006). Effect of voluntary attention on auditory processing during REM. Sleep, 29, 975–982. Talairach, J., & Bancaud, J. (1973). Stereotactic approach to epilepsy: Methodology of anatomo-functional stereotaxic investigations. Progress in Neurolological Surgery, 5, 297–354. Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the human brain. 3Dimensional proportional system an approach to cerebral imaging. Stuttgart, Germany: Georg Thieme Verlag. Terzaghi, M., Sartori, I., Tassi, L., Didato, G., Rustioni, V., LoRusso, G., et al. (2009). Evidence of dissociated arousal states during NREM parasomnia from an intracerebral neurophysiological study. Sleep, 32(3), 409–412. Vogt, B. A. (2005). Pain and emotion interactions in subregions of the cingulate gyrus. Nature Review of Neuroscience, 6, 533–544. Voss, U., Holzmann, R., Tuin, I., & Hobson, J. A. (2009). Lucid dreaming: A state of consciousness with features of both waking and non-lucid dreaming. Sleep, 32(9), 1191–1200. Weiskrantz, L. (1997). Consciousness lost and found: A neuropsychological exploration. New York: Oxford University Press. Zimmerman, W. B. (1970). Sleep mentation and auditory awakening threshold. Psychophysiology, 6, 540–549.
Contents lists available at ScienceDirect
Brain and Cognition journal homepage: www.elsevier.com/locate/b&c
Asleep but aware? Stéphanie Mazza a,⇑, Caroline Perchet b,c, Maud Frot b,c, George A. Michael a, Michel Magnin b,c, Luis Garcia-Larrea b,c, Hélène Bastuji b,c,d a
Laboratoire d’Etude des Mécanismes Cognitifs, Université Lumière, Lyon 2, 5 Avenue Pierre Mendes France, Bron F-69676, France INSERM, U1028, CNRS, UMR 5292, Centre de Recherche en Neurosciences de Lyon, Laboratoire «intégration centrale de la douleur», 59 bd Pinel, Bron Cedex F-69677, France c Université Claude Bernard, 43 Boulevard du 11 Novembre 1918, Lyon 1 F-69003, France d Hospices Civils de Lyon, Unité d’Hypnologie, Hôpital Neurologique, 59 Bd Pinel, Bron F-69677, France b
a r t i c l e
i n f o
Article history: Accepted 16 February 2014 Available online 13 March 2014 Keywords: Paradoxical sleep Perception Consciousness Nociception Intracerebral EEG
a b s t r a c t Despite sleep-induced drastic decrease of self-awareness, human sleep allows some cognitive processing of external stimuli. Here we report the fortuitous observation in a patient who, while being recorded with intra-cerebral electrodes, was able, during paradoxical sleep, to reproduce a motor behaviour previously performed at wake to consciously indicate her perception of nociceptive stimulation. Noxious stimuli induced behavioural responses only if they reached the cortex during periods when mid-frontal networks (pre-SMA, pre-motor cortex) were pre-activated. Sensory responses in the opercular cortex and insula were identical whether the noxious stimulus was to evoke or not a motor behaviour; conversely, the responses in mid-anterior cingulate were specifically enhanced for stimuli yielding motor responses. Neuronal networks implicated in the voluntary preparation of movements may be reactivated during paradoxical sleep, but only if behavioural-relevant stimuli reach the cortex during specific periods of ‘‘motor awareness’’. These local activation appeared without any global sleep stage change. This observation opens the way to further studies on the currently unknown capacity of the sleeping brain to interact meaningfully with its environment. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Sleep has frequently been assimilated to a little death, to such an extent that the Greek mythology establishes a fraternal link between Hypnos, the God of Sleep and Thanatos the God of Death. The loss of behavioural control and the fading of consciousness are the distinctive features of sleep that underlie this phenomenon. The accompanying lack of responsiveness towards nocturnal stimulation led to hypothesize a functional disconnection between the cerebral cortex and the external world (Steriade, 1994). This disconnection is especially intriguing in the period of paradoxical sleep (PS), during which the electroencephalographic pattern is similar to that of waking state. In order to explain this paradox, Foulkes (1966) and later Llinas and Ribary (1993) suggested that PS ‘‘is a state of hyperattentiveness during which sensory input cannot address the machinery that generates conscious experience’’. They hypothesized that the oniric activity integrally focuses the attention of the sleeper, excluding the possibility for external sources of stimulation to be
⇑ Corresponding author. Fax: +33 4 78 77 43 51. E-mail address: [email protected] (S. Mazza). http://dx.doi.org/10.1016/j.bandc.2014.02.007 0278-2626/Ó 2014 Elsevier Inc. All rights reserved.
processed. However, this view is challenged by the well-known fact that external events can interfere with, or be incorporated into, the ongoing oniric activity (Dement & Wolpert, 1958). Modern behavioural and neurophysiological studies indicate that the sleeper keeps the possibility to perceive and process external information. Accordingly brain responses related to cognitive activity (e.g. P300) have been recorded during PS in response to frequency- or intensity-deviant stimuli (Bastuji, Garcia-Larrea, Franc, & Mauguière, 1995; Cote & Campbell, 1999; Takahara, Nittono, & Hori, 2006; Macdonald, Jamshidi, & Campbell, 2008), and likewise to more complex stimuli such as the subject’s own name (Perrin, Garcia-Larrea, Mauguiere, & Bastuji, 1999; Pratt, Berlad, & Lavie, 1999). The persistence of a differential brain response (‘‘N400 effect’’) to incongruous words during PS further indicates that the sleeper’s brain keeps some capacity to analyse semantic contents (Brualla, Romero, Serrano, & Valdizan, 1998), even if linguistic absurdity appears to be accepted to a greater extent than during waking (Perrin, Bastuji, & Garcia Larrea, 2002). These results confirm that the sleeping brain remains able to detect and categorize some particular traits of external stimulus significance, even if the processes engaged are not identical to those implicated during wakefulness.
8
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
Notwithstanding the importance of the above work, the question of consciousness of external events during sleep remains unanswered. An empirical criterion used to assess the awareness of information is its ‘‘reportability’’ (Dehaene & Naccache, 2001; Gazzaniga, LeDoux, & Wilson, 1977; Weiskrantz, 1997) defined as ‘‘all voluntary communicative acts that are used to report conscious content’’ (Baars, 1997). In the context of sleep, Laberge (2000) has shown that ‘‘lucid dreamers’’ can voluntarily produce eye movements, to indicate that they are aware of dreaming, during ongoing PS. It has also been demonstrated that operant responses learned during wakefulness may be reactivated by stimuli presented during sleep (Bonnet, 1982; Burton, Harsh, & Badia, 1988; Granda & Hammack, 1961; Lauerma, Kaartinen, Polo, Sallinen, & Lyytinen, 1994). Evans, Gustafson, O’Connell, Orne, and Shor (1970) succeeded to observe complex behavioural responses to verbal suggestion during PS (i.e. ‘‘Whenever I say the word ‘‘itch,’’ your nose will feel itchy until you scratch your nose’’) without sleep interruption. These behaviours remain rare and inconsistent suggesting that stimulus awareness during PS may be fluctuating. In parallel to this, a critical and unsolved question concerns, the brain regions or activity supporting the production of these apparently ‘intentional’ responses while EEG recordings indicate unequivocal PS sleep, characterised by atonia and fading consciousness. Using brain imaging, Dresler and colleagues recently showed that dreamed motor actions performed during PS elicited neuronal activation in the sensorimotor cortex and the SMA, comparable to those observed during movements really acted during wakefulness (Dresler et al., 2011). Here we describe the results of intracerebral EEG investigation that fortuitously captured a spontaneous reactivation of an actual motor behaviour previously used during wakefulness as an indicator of conscious perception of nociceptive stimulation in a patient affected by drug-resistant epilepsy recorded during presurgical assessment. Intra-cerebral recordings demonstrated that preactivation of mid-frontal networks was determinant to allow this adapted behaviour to an external input, whereas activity in sensory cortices appeared irrelevant and global sleep stage unchanged. 2. Materials and methods The subject was a 37-year-old women involved in a study protocol evaluating nociceptive processes during sleep using laser stimulations (Bastuji et al., 2011). She presented with refractory epilepsy and was being investigated using stereotactically implanted intracerebral electrodes before functional surgery. Depth EEG recording electrodes were implanted according to the stereotactic technique of Talairach and Bancaud (1973). The cortical targets were identified on the patient’s MRI. The implantation procedure has been described in detail elsewhere (Frot & Mauguière, 1999; Frot, Rambaud, Guénot, & Mauguière, 1999; Guenot et al., 2001). The patient was fully informed about the fact that Laser EvokedPotential (LEP) recordings during sleep were not a part of the diagnostic procedure but were performed with research purposes, and gave her written informed consent. The laser stimulation paradigm was submitted to, and approved by, the local Ethics Committee (CCPPRB Léon Bérard-Lyon) and the study was promoted by the French National Agency for Medical Research (INSERM).
the right hand (radial territory contralateral to the hemisphere of electrode implantation). The pain threshold was determined at wake. The patient was instructed to lift her left index finger to indicate that she had perceived the stimulation and score its intensity using a Likert-type scale where 0 was ‘‘no sensation’’, 8 = unbearable pain, and 4 was defined as ‘‘pricking, moderately painful’’). An energy level of 80 mJ/mm2, yielded for this patient the expected painful sensation defined as ‘‘pricking, moderately painful’’. After that, ten stimulations, at this pain threshold, were delivered to obtain wakefulness LEP responses. During LEP recordings, the patient was lying, immobile. Then, the patient was allowed to sleep at her own time. During the night, series of up to 30 nociceptive laser pulses were delivered when definitive stages of sleep were reached, as determined online by an experienced sleep researcher. The staging was later confirmed off-line. The laser intensity was kept stable during the whole night (80 mJ/mm2). Laser pulses were transmitted through the optic fibre from the laser stimulator. This 10-m optical fibre connecting the laser generator with the stimulating probe allowed stimulating conveniently the hand dorsum despite subject position during the night. Both the sleeping subject and the stimulating investigator wore eye protections. After preliminary work showing that delivering stimuli at short (<6 s) and constant intervals increased the probability of awakening, inter-stimulus interval (ISI) was pseudo-randomly adjusted on-line and varied between 10 and 20 s. The laser beam was slightly moved over the skin surface between two successive stimuli to avoid habituation and especially peripheral nociceptor fatigue (Schwarz, Greffrath, Buèsselberg, & Treede, 2000). Due to environmental bad conditions (noisy storm), this night session ended prematurely. A second night of stimulation was planned one week later, during which, the emergence of behavioural responses was observed. This night recording was conducted 14 days after electrodes implantation; at that time, anticonvulsant drug intake was drastically reduced for at least one week. EEG recordings were performed with sampling frequency of 256 Hz and a 0.03–100 Hz bandpass (MicromedÒ, Mâcon, France) in referential mode (the reference electrode being an implanted contact located in the skull). In order to limit the lack of comfort induced by the monitoring device, no submental EMG electrodes were placed, but only EKG and 2 EOG electrodes (supero- and infero-lateral right canthus). EEG, EKG and EOG were recorded continuously during the night and stored for off-line analysis. In order to stimulate at wellidentified sleep stages, the different states of vigilance were visually identified on-line, according to the criteria of Rechtschaffen and Kales (1968) adapted to intracranial recordings (see Magnin, Bastuji, Garcia-Larrea, & Mauguiere, 2004; Rey et al., 2007). 2.2. Incorporation of stimuli to dreams In the morning that followed the recording night, a four-level scale adapted from Zimmerman (1970) was used to assess the degree of stimulus incorporation to dreams. The four-levels were (1) no incorporation, the subject had no recall of somatosensory or pain stimulation; (2) possible incorporation, the subject had some somatosensory or pain recall, but unrelated to the stimulations delivered; (3) obvious incorporation, the subject had some somatosensory or pain recall related to the stimulations delivered; (4) awareness of stimulation effectively delivered.
2.1. Laser stimulation procedure and parameters
2.3. Anatomical localization of recording sites
Nociceptive heat pulses of 5 ms were delivered with a Nd:YAPlaser (Yttrium Aluminium Perovskite; wavelength 1.34 lm; El.En.Ò). Series of laser stimuli were delivered on the dorsum of
The final position of each implanted contact with respect to the targeted anatomical structures was verified via frontal and sagittal X-rays at scale 1, using an X-ray source at 4.85 m from the patient’s
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
head thus eliminating the linear enlargement due to X-ray divergence. The T1-weighted MRI midsagittal slice, also at scale 1, was superimposed on the sagittal X-ray using the limits of the cranial bone as a common reference. This allowed to define the position of the anterior commissures (AC) and posterior commissures (PC) and to calculate the anteroposterior and dorsoventral coordinates of each electrode contact. The mediolateral coordinate of each contact was determined on the frontal X-ray with respect to the midsagittal plane. Each contact was then localized in the Talairach space (Talairach & Tournoux, 1988) according to the following 3 dimensional (3-D) coordinates in millimetres: x = distance from the interhemispheric sagittal vertical plane; y = distance from the coronal plane perpendicular to the horizontal AC–PC plane passing through AC; z = distance from horizontal AC–PC plane. The origin of this 3D reference system corresponds to the AC location as seen on the midsagittal plane. The 3-D position of all contacts within the patient’s cerebral structures was checked by plotting their coordinates in her MRI volume (T1-weighted images, 1-mm based voxel), reconstructed in the 3-D space using MRIcro Software (Rorden & Brett, 2000). This patient had 8 implanted electrodes (90 contacts in the grey matter) (Fig. 1).
9
EEG data were segmented into epochs ranging from 100 ms before the stimulation to 1000 ms after. Each EEG epoch was baseline corrected using the 100 ms pre-stimulus interval as reference. Trials contaminated by epileptic transient activities were removed. The LEP components on the insular, opercular and cingulate corteces were identified as the peaks with the highest amplitude within a 160–420 ms latency window encompassing the corresponding waveform (Frot, Mauguière, Magnin, & Garcia-Larrea, 2008). The components were analysed when their amplitude value during wakefulness exceeded the mean prestimulus baseline by at least 3 SD. They were labelled C1 and C2 according to their latency (Bastuji et al., 2011). Mean voltages and latencies are given ±1 SD. LEPs obtained during wakefulness were compared to those obtained during PS, separately whether they were or not followed by an overt finger movement (PS with behaviour vs PS without behaviour). Latencies and amplitudes of LEPs were tested using one-way analyses of variance (ANOVA) with ‘‘vigilance state’’ as factor (waking vs PS with behaviour vs PS without behaviour). Tukey-Kramer tests were used for post hoc comparisons.
2.4.2. EEG analysis in the time-frequency domain: continuous wavelet transform
2.4. Data analysis Sleep stages were visually identified using more than 20 cerebral contacts including the dorso-lateral extent of the frontal, parietal and temporal lobes, and the thalamus, with both bipolar and referential traces plus EOG. Sleep scoring was done according to the criteria of Rechtschaffen and Kales (1968) and thalamic activity (see Magnin et al., 2004; Rey et al., 2007), so as to derive hypnograms based on 30 s epochs and determine the vigilance state during which laser stimuli were delivered. Data reported in this study concern exclusively stimulations delivered during the waking session and the last PS at 06:20. Epochs of 10 s prior and 5 s after stimuli delivered during PS were classified as occurring into a tonic (without rapid eye movement) or a phasic period (with rapid eye movements) (Sallinen, Kaartinen, & Lyytinen, 1996). The occurrence of EEG arousing reactions or full awakenings was controlled online considering raw cortical EEG activities. In absence of EMG recording, arousals were defined as an abrupt shift of EEG activities towards higher frequencies (alpha, theta and/or >16 Hz) that lasted at least 3 s and less than 15 s only (AASM: Iber, Ancoli-Israel, Chesson, & Quan, 2007). These frequency modifications were considered as stimulus-related if they occurred within 15 s after stimulus onset. A full awakening was scored when this cortical reaction lasted more than 15 s. The delay separating each finger lift from the stimulus was calculated using the video recording synchronised with the EEG. 2.4.1. Intracranial laser-evoked potentials (LEPs) LEPs analyses were performed using ASAÒ software (Advanced Neuro Technology (ANT), Eschende, The Netherlands). Continuous
– In order to determine whether a specific EEG activation pattern was associated to the production of a behavioural response, a time-frequency (TF) analysis of the EEG preceding each stimulus was performed, for all contact localisations (Fig. 1) (Letswave software; Mouraux & Iannetti, 2008). Wavelet transforms are particularly suitable for the analysis of nonstationary signals in the time–frequency domain. Thus, the continuous wavelet TF transform was applied to each individual epoch in each brain structure explored, using a time window analysis of 2 s before the stimulus (stimulus-referenced). The resulting TF amplitude maps were then averaged for the condition with behaviour and without behaviour. Results were displayed as an increase or decrease of oscillation amplitude after baseline correction and were analysed separately for each standard EEG frequency bands: delta (1.5–4.5 Hz), theta (4.75–7.75 Hz), alpha (8–12 Hz), sigma (12.25–15 Hz), beta (15.25–30 Hz) and gamma (30.25–45 Hz). TF amplitude values of epochs preceding or not a behaviour were compared with the Wilcoxon test for paired data. Pairing was achieved through the ordering procedure proposed by Michael, Boucart, Degreef, and Godefroy (2001), already used in behavioural case studies, which avoids masking of effects due to the random and unpredictable post-stimulus appearance of the behavioural response. In each condition (behavioural response or no behavioural response), amplitude values were ordered from the smallest to the largest ones for each EEG frequency band. Then, the smallest value in one condition was paired with the smallest one in the other condition. The same procedure was repeated sequentially for the two series of increasing values.
Fig. 1. Electrode implantation. This figure illustrates the electrode implantation sites on the lateral view of the patient’s 3D MRI, the anatomical structures and the corresponding Brodmann areas explored by each electrode.
10
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
– TF analyses were also performed using a time window locked on the onset of the behavioural response (response-referenced), EEG epochs corresponding to the 2 s preceding the beginning of each motor response. Continuous wavelet transform was applied to each epoch after a baseline correction and then averaged. Within average TF map obtained, the highest amplitude value (Vmax) was identified. A region of interest (ROI) was centred on the location of the Vmax. Its limits circumscribed a zone in which amplitude values were at least 60% of the Vmax and thus considered as significantly increased. Then, a ROI of same surface was applied to each TF single map composing the TF average map at the same coordinates in terms of frequency and time. To verify if activations defined by ROIs were specific to behavioural responses, identical ROIs were applied to TF maps obtained when behavioural responses were absent. To take into account the latency jitter observed between the laser stimulation and the behavioural response the ROIs were placed in each epoch without behavioural responses at variable times with respect to the laser stimulation, reproducing thus the delays observed when the behavioural response was present (see Fig. 2). As mentioned above a Wilcoxon test for paired data was performed on average amplitude value within the ROIs to determine regions responding differently between stimulations generating or not behavioural response. The Average amplitude values within the ROIs are given ±1 SEM. The level of significance was set at 0.05 for all statistical analyses.
3. Results 3.1. Behavioural responses At the end of a whole night of laser stimulation, the patient unexpectedly lifted her left index finger in response to 11 of the
20 laser stimuli delivered during PS at 06:20. These 11 behavioural responses were not obtained after consecutive stimulations, but dispersed within the stimulation series (Fig. 3B, see movie). All the stimuli were delivered during tonic PS, attested by EOG, cortical activity and the presence of a concomitant delta activity at the thalamic level. These stimulations never induced any global sleep stage shift. When visually compared using video recording, the latencies of finger movements observed during PS were not different from those of the motor responses observed during wakefulness (mean latency after laser stimulation = 840 ms (range: 460–1670 ms) vs 1008 ms (range 470–1750) respectively; p > 0.05). However, while the amplitudes of movements appeared similar, movements were performed at slower velocity during the PS. Among these 11 stimulations followed by a movement, 8 were followed by an obvious micro-arousal identified through changes in EEG cerebral activities appearing at a mean latency of 1400 ms (range: 980–1810 ms). The micro-arousal always lasted less than 15 s and never gave rise to further sleep stage shifts. One microarousal only could be observed among the 9 stimulations not followed by a motor response. There was neither incorporation of nociceptive stimuli into dreams content nor recollection of pain stimulation on the following morning (questionnaire score = 1).
3.2. Laser evoked potentials (LEPs) Among the 30 laser stimulations analysed (20 delivered during PS, 10 during wakefulness), 3 were discarded due to epileptic activity. Thus, the LEPs results presented are based on 27 cortical responses, including 9 stimulations delivered during PS and followed by a behavioural response (7 preceding a micro-arousal), 9 during PS without behavioural response and 9 during wakefulness. In the insular cortex, ANOVA showed a significant effect of state of vigilance on LEP latencies (component C1: F(2, 24) = 4.74;
Fig. 2. Definition of region of interest (ROI) during the period of time preceding the apparition of a behavioural response. Within the average TF map locked on the onset of the behavioural response, a ROI (black scare) is centred on the highest amplitude value (Vmax = red cross) and circumscribes a zone in which amplitude values are at least 60% of the Vmax. ROI of same surface is applied to each TF single map composing the TF average map: same frequency coordinates (Fmin and Fmax) and time (Tmin and Tmax). For each TF single map the delay between the laser stimulation and the Tmin is calculated (in the example: x1, x2, x3, . . .). ROIs are placed in each epoch without behavioural responses reproducing the delays observed in the condition with behavioural response (in the example: x1, x2, x3, . . .). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
11
Fig. 3. Laser evoked potentials. (A) Averaged laser evoked potentials record in the ‘pain matrix’, in a bipolar recording mode from two adjacent contacts (red crosses). Negative potentials at the intracortical recording site are represented upward. For each structure, the Talairach coordinates of the contacts are indicated and illustrated on the coronal patient’s MRI. (B) Infra-red video images showing the patient during PS after two laser nociceptive stimulations on the right hand: on the left panel the stimulation gives rise to a left finger lift (red arrow) while, on the right panel no motor response follows the stimulation. Middle panel: enlargement of the C1 component recording during waking, PS with and PS without a behavioural response, in the anterior cingulate gyrus. Asterisk indicates the statistically significant difference between PS without motor response as compared with the others. (C) The five sets of raw intracranial recordings illustrate the frontal (a), temporal (b) and parietal (c and d) cortical and concomitant thalamic (e) PS activities. Laser stimulations were deliverated during delta oscillations in the thalamus. Asterisk: micro-arousal. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
p = 0.02 and C2: F(2, 24) = 17.75; p = 0.0001) and amplitudes for component C2 (component C1: F(2, 24) = 0.3; p = n.s. and component C2: (F(2, 24) = 4.35; p = 0.02). Post-hoc analysis specified that, as compared to waking, LEPs obtained during PS had delayed latencies and decreased amplitudes. Further statistical analysis did not reveal any differences between the two PS conditions, i.e., with or without behavioural response. In the opercular cortex, similar results were obtained with a significant effect of state of vigilance for latencies (components C1: F(2, 24) = 10.28; p = 0.0006) and C2: F(2, 24) = 15.96; p = 0.0003) and amplitudes (component C1: F(2, 24) = 4.74; p = 0.02 and C2: F(2, 24) = 3.82; p = 0.036). During the 2 PS conditions the LEPs were delayed and attenuated as compared to waking and no differences were observed between LEPs during PS with or without motor responses. In the anterior cingulate gyrus, ANOVA showed a significant effect of state of vigilance on latencies for C2 component (component C1: F(2, 24) = 0.54; p = ns and C2: F(2, 24) = 25.56; p < 0.001) and amplitudes (component C1: F(2, 24) = 5.77; p = 0.009 and C2: F(2, 24) = 8.55; p = 0.002). The LEP latencies were different with the following gradient: wake = PS without behaviour < PS with behaviour. The amplitudes of C1 obtained during PS with motor behaviour were significantly enhanced relative to those in periods with no overt response, making them similar to waking LEPs (Fig. 3B). Component C2 had greater amplitude during the waking state as compared to either PS condition.
To summarise, opercular and insular LEPs remained similar during PS whether the stimulus evoked a motor response or not. In contrast, nociceptive stimuli giving rise to a finger movement entailed a significant increase of cingulate LEPs, which became similar to those obtained during wakefulness (see Fig. 3 and Table 1). 3.3. EEG activity preceding laser stimuli During the pre-stimulus interval ( 2 to 0 s), time-frequency analyses did not reveal any significant statistical difference between the two PS conditions, with or without behavioural response, except for the pre-supplementary motor area (pre-SMA) and the premotor area (BA6) (Fig. 4A and B). In the pre-SMA, low frequencies were reduced by 1.55 lV for delta (t(8), p < 0.02), 4.87 lV for theta (t(8), p < 0.004), and 1.20 lV for alpha (t(8), p < 0.006), when followed by a behavioural response. In contrast, pre-stimulation high frequencies were enhanced by 0.90 lV for beta (t(8), p < 0.004). In BA6, pre-stimulus EEG activity was significantly reduced in theta and alpha frequency-bands ( 3.40 lV (t(8), p < 0.004, and 0.99 lV (t(8), p < 0.001) and increased in beta and gamma frequency-bands (0.6 lV (t(8), p < 0.004, and 0.4 lV (t(8), p < 0.006) when a finger lift occurred. In the cingulate, no significant difference in the 1.5–45 frequency band could be observed (Fig. 4C), despite the LEPs modification described.
12
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
Table 1 Laser evoked potential (LEP) values recorded in the 3 cortical areas. Latencies (mean ± SD) and amplitudes (mean ± SD) values of the two LEP components in the 3 structures of the pain matrix. Post-hoc analysis following statistically significant ANOVA are presented in the three conditions: wake (1), PS with behaviour (2), PS without behaviour (3). Component 1
Component 2
Latency (ms)
Post hoc
Amplitude (lV)
Post hoc
Latency (ms)
Post hoc
Insula Wake (1) PS with behaviour (2) PS without behaviour (3)
Amplitude (lV)
178 ± 11 198 ± 21 196 ± 13
1<2=3
109 ± 22 101 ± 46 95 ± 40
NS
273 ± 6 345 ± 27 339 ± 13
1<2=3
107 ± 26 76 ± 26 68 ± 35
1>2=3
Operculum Wake (1) PS with behaviour (2) PS without behaviour (3)
179 ± 11 206 ± 8 211 ± 9
1<2=3
91 ± 15 30 ± 15 27 ± 13
1>2=3
264 ± 17 371 ± 24 354 ± 14
1<2=3
57 ± 33 31 ± 15 38 ± 13
1>2=3
Anterior cingulate Wake (1) PS with behaviour (2) PS without behaviour (3)
250 ± 12 248 ± 8 254 ± 19
NS
81 ± 35 63 ± 32 30 ± 13
1=2>3
308 ± 12 365 ± 26 314 ± 13
1=3<2
64 ± 25 26 ± 9 24 ± 31
1>2=3
Post hoc
NS: non significant. Values resulting from mean LEP responses in the pain matrix.
Fig. 4. Time-frequency maps preceding the laser stimulation. Average TF amplitude maps stimulus-referenced in the pre-SMA, premotor area, and anterior cingulate gyrus. Left panel illustrates the location and Talairach coordinates of the contacts on the 3 planes of the patient’s MRI for each structure. For each region the Talairach coordinates of the two contacts ensuring bipolar recording are indicated. Average TF amplitude maps (middle panel) display increase and decrease oscillations amplitude 2s before the laser stimulation, for the condition with (top) and without (bottom) behavioural response, in the pre-SMA, the premotor area and the anterior cingulate gyrus. Graphs on the right illustrate the amplitude in each frequency band for the condition with (with squares) and without (grey circles) behavioural response in the 3 structures. Asterisk indicates the statistically significant differences.
To summarise, stimuli giving rise to motor behaviour were preceded by a significant amplitude reduction in low frequencies and increase in high-frequency activities in the pre-supplementary motor area (pre-SMA) and the premotor area (BA6). 3.4. EEG activity preceding behaviours Average time frequency maps presented in Fig. 5 illustrate cortical activities during the interval of time between the
nociceptive stimulus and the occurrence or the absence of finger lift. When a behavioural response occurred, a specific and significant enhancement of cortical activities in the 7–11 Hz ROI appeared 790 ms before finger movements in the parietal cortex (BA40) (23.07 ± 6.09 lV with behaviour vs 0.7 ± 3.26 lV without behaviour, t(8) p < 0.004). This activation was followed by a similar activity increase in the pre-SMA with a delay of 450 ms, in a 5–8 Hz ROI, (12.69 ± 2.92 lV vs 1.25 ± 4.02 lV, t(8) p < 0.004).
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
13
Fig. 5. Time-frequency maps preceding the apparition of a behavioural. Average TF maps illustrate activity levels in cortical regions showing a significant difference during the time interval between the nociceptive stimulus and the occurrence (top) or the absence (bottom) of a behavioural response. For each region the Talairach coordinates of the two contacts ensuring bipolar recording are indicated. Squares illustrate regions of interest (ROIs) in which activity levels are significantly different after the stimulations leading or not to a behavioural response. Values in TF maps correspond to the mean frequency amplitude ((lV) ± SEM) defined within the ROIs. Asterisks indicate the statistically significant differences between PS with and without motor responses.
In the frontal area (BA11) an increased activity starting 400 ms before the motor response in a 2–7 Hz ROI which endured during finger lift execution (4.97 ± 0.97 lV vs 0.89 ± 1.42 lV, t(8) p < 0.004), while in the hippocampus the peak of ROI activity (7– 15 Hz) was limited and concomitant to the motor execution (17.95 ± 6.27 lV vs 1.64 ± 2.89 lV, t(8) p < 0.004). 4. Discussion The patient, presented herein, spontaneously lifted her index finger in response to 11 of the 20 nociceptive stimulations delivered during the last PS period of the night. This motor behaviour suggests that, during PS, this patient was not only able to detect nociceptive stimuli but also to initiate a motor sequence previously used at wake to indicate her perception. In parallel, intracerebral recordings uncovered activations of brain structures brought into play during this sensory-motor processing. Among the above electrophysiological results, two of them seem specific to the occurrence of a behavioural manifestation. First, LEP cingulate responses associated to a meaningful motor behaviour were selectively enhanced and became similar to those obtained during wakefulness. Second, the level of activation in motor and pre-motor networks was also increased before the stimulations giving rise to a motor behaviour. As these data have been obtained in an epileptic patient, one can wonder if they are more or less related to this pathophysiology. However, comparing cortical activities recorded during our experiment with those observed during the patient’s seizures and even during interictal periods leaves no doubt as to the physiological validity of the phenomena we observed. The possibility to generate a movement during the muscle atonia characterising PS remains puzzling. However, several data converge to consider that muscle atonia can be removed in various circumstances. Voluntary movements, such as eye movements, deep breath or complex motor movements, can be obtained from subjects in response to external stimulations during this sleep stage (Burton et al., 1988; Evans, Gustafson, O’Connell, Orne, & Shor, 1970; Laberge, 1990; Lauerma et al., 1994). PS dreams involving body movements can be associated with an increase in EMG activity of the corresponding muscle (Erlacher & Schredl, 2008). Furthermore, it has been demonstrated that, despite the general muscle blockade, brief bursts of phasic EMG activity can be recorded during PS, more frequently in distal than in proximal muscles of the limbs (Gardner & Grossman, 1975; Rye, 2003). This has been recently quantified, indicating that the strength of atonia is weaker in distal territories (Frauscher et al., 2008). All these elements demonstrate that motor inhibition during PS can be alleviated, allowing movements triggered by external or internal
stimuli, although their dynamics are partly impeded, as observed in our patient’s finger lifts. Intracerebral recordings have shown that while the lateral operculo-insular system supporting intensity coding and stimulus localization remains similarly active during PS and waking, the anterior cingulate cortex (ACC) involved in motor control, orienting and attention for action (Dum, Levinthal, & Strick, 2009; Frot et al., 2008; Vogt, 2005) is selectively depressed (Bastuji et al., 2011). In the present study, cingulate responses during PS were specifically restored before each finger lift, reaching amplitudes comparable to those of the waking state. This result confirms the direct access of spinothalamic input to cingulate motor networks, bypassing the lateral operculo-insular system. Besides, our results revealed that noxious stimuli induced behavioural responses only if they reached the cortex during PS sleep periods when motor and pre-motor networks (pre-SMA, and BA 6), were activated before stimulus arrival. These transitory increase in high frequency bands and decrease of low-frequency rhythms, which could be view as topographically circumscribed microarousals, may permit an appropriate motor response to be generated. In support to this hypothesis, intracranial EEG recording techniques have provided new and interesting information pointing to a local regulation of sleep. We recently showed that during sleep onset, the thalamus undergoes a deactivation process that precedes by several minutes the one occurring within the cortex. Moreover, we confirmed that the cortical deactivation also showed marked topographical differences (Magnin et al. 2010). Nobili and coworkers (2011) reported the coexistence of wake-like activity in the motor cortex and sleep-like EEG patterns in other cortical areas during the SWS. By capturing a confusional arousal episode in a subject recorded with intracerebral EEG, the same team documented the presence of a local arousal of the motor and cingulate cortices beginning before the confusional episode, while the associative cortices remained asleep (Terzaghi et al., 2009). Studying lucid dreams during PS, Voss, Holzmann, Tuin, and Hobson (2009) showed that lucidity, a period during which sleepers became able to voluntary interact with their environment, occurs in a state with features of both PS and waking, constituting an hybrid state of consciousness. Thus, we could speculate that local activity modification or local ‘‘awakening’’ during PS, could explain the possibility of sensory processing and behavioural responses, without any modification of the global sleep state. The behaviours we observed were preceded by the activation of cerebral regions involved in the voluntary preparation of movement, namely the pre-frontal and parietal cortices, and the caudal part of the pre-SMA (Deiber et al., 1991; Jenkins, Jahanshahi, Jueptner, Passingham, & Brooks, 2000). The latter is known to be specifically involved in self-initiated actions (Lau, Rogers, Ramnani, &
14
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15
Passingham, 2004) and externally-driven movements (Cunnington, Windischberger, Deecke, & Moser, 2002). This pattern of activation may suggest that a sleeper is able to engage cerebral networks implicated in intentional movements. Dehaene and Naccache (2001) postulated that, during wakefulness, consciousness depends on the connectivity of a brain-scale workspace comprising the prefrontal cortex, the anterior cingulate, and the areas that connect to them. Despite a general state of frontal hypo-activity during PS (Maquet, 2000), Massimini et al. (2005) proposed that sleep is characterised by a dynamic variation of the effective connectivity among cortical areas. This connectivity, which is drastically reduced during SWS, is partially restored during PS, suggesting a potential resurgence of cortical integration and, accordingly, a resurgence of consciousness during this state (Massimini et al., 2010). In this context the local modulation in cerebral activities we observed before laser stimulation and the behaviour occurrence may participate to a kind of restored awareness during PS. Nevertheless, since cerebral activities reported in this case study were restricted to electrode implantation needed to define the epileptogenic area for surgical purposes, we do not know if other cerebral structures apart from implantation sites may contribute to this phenomenon. Other factors influencing responsiveness during PS may have participated to these behaviours. Lammers and colleagues described that responsiveness to external stimulations increases from the first to the second half of the night, reaching its maximum at early morning (05:30). This phenomenon, related to both homeostatic and circadian processes (Borbély, 1982), may have favoured the late occurrence of behavioural responses in the present case (at 06:20). Moreover, a mechanism of memory consolidation may also have contributed to these behaviours, since they were observed in the only patient who was tested twice at a one-week interval. Jouvet (1979), Nielsen, Kuiken, Alain, Stenstrom, and Powell (2004) have reported the reappearance in dreams of features from events occurring about a week prior, suggesting that processes with circaseptan morphology underlie dream incorporation. This delay may correspond to the sleepdependent redistribution of hippocampus-related memories to neocortical sites supporting memory consolidation (Nielsen & Stenstrom, 2005). In our study, the patient did not report neither incorporation into dreams of nociceptive stimuli, nor even recollection of pain stimulation on the morning questionnaire. The fact that cortical activations remain temporally and topographically limited may explain the lack of insight, a function requiring a much larger cortical network (Dehaene & Naccache, 2001). Nevertheless, this lack of recollection does not exclude some kind of temporally mnesic reactivation during PS, attested by the observed hippocampal activation, but remaining too limited to persist once awake. 5. Conclusions The present study suggests that during PS, not only the perception of sensory inputs but also the capacity for the sleeper to indicate his perception can, under particular circumstances, be preserved. A selective enhancement in motor and pre-motor networks associated with a re-activation of the anterior cingulate may allow an appropriate motor response to be generated. This case report opens the way to further studies on the currently unknown capacity of the sleeping brain to interact meaningfully with its environment. Acknowledgments The work was supported by the French National Agency for Medical Research (INSERM), the university of Lyon 2 (dispositif
incitation à la recherche), Fondation pour la Recherche Médicale (FRM), APICIL Foundation and NRJ Foundation of the Institut de France and by a Region Rhone-Alpes/France CIBLE 2010 grant. We are indebted to Drs J. Isnard and H. Catenoix for the opportunity to study their patient and to Dr M. Guénot for stereotactic electrode implantation. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bandc.2014. 02.007. References Baars, B. J. (1997). Some essential differences between consciousness and attention, perception and working memory. Consciousness and Cognition, 6, 363–371. Bastuji, H., Garcia-Larrea, L., Franc, C., & Mauguière, F. (1995). Brain processing of stimulus deviance during slow wave sleep and paradoxical sleep: A study of human auditory evoked responses using the oddball paradigm. Journal of Clinical Neurophysiology, 12, 155–167. Bastuji, H., Mazza, S., Perchet, C., Frot, M., Mauguière, F., Magnin, M., et al. (2011). Filtering the reality: Functional dissociation of lateral and medial pain systems during sleep in humans. Human Brain Mapping, 16. http://dx.doi.org/10.1002/ hbm.21390. Bonnet, M. H. (1982). Performance during sleep. In W. Webb (Ed.), Biological rhythms, sleep and performance (pp. 205–237). Chichester: John Wiley & Sons Ltd. Borbély, A. A. (1982). A two process model of sleep regulation. Human Neurobiology, 1, 195–204. Brualla, J., Romero, M. F., Serrano, M., & Valdizan, J. R. (1998). Auditory event-related potentials to semantic priming during sleep. Electroencephalography and Clinical Neurophysiology, 108, 283–290. Burton, S. A., Harsh, J. R., & Badia, P. (1988). Cognitive activity in sleep and responsiveness to external stimuli. Sleep, 11, 61–69. Cote, K. A., & Campbell, K. B. (1999). P300 to high intensity stimuli during REM sleep. Clinical Neurophysiology, 110, 1345–1350. Cunnington, R., Windischberger, C., Deecke, L., & Moser, E. (2002). The preparation and execution of self-initiated and externally-triggered movement: A study of event-related fMRI. Neuroimage, 15, 373–385. Dehaene, S., & Naccache, L. (2001). Towards a cognitive neuroscience of consciousness: Basic evidence and a workspace framework. Cognition, 79, 1–37. Deiber, M. P., Passingham, R. E., Colebatch, J. G., Friston, K. J., Nixon, P. D., Frackowiak, R., et al. (1991). Cortical areas and the selection of movement: a study with positron emission tomography. Experimental Brain Research, 84, 393–402. Dement, W., & Wolpert, E. A. (1958). The relation of eye movements, body motility, and external stimuli to dream content. Journal of Experimental Psychology, 55, 543–553. Dresler, M., Koch, S. P., Wehrle, R., Spoormaker, V. I., Holsboer, F., Steiger, A., et al. (2011). Dreamed movement elicits activation in the sensorimotor cortex. Current Biology, 21, 1833–1837. Dum, R. P., Levinthal, D. J., & Strick, P. L. (2009). The spinothalamic system targets motor and sensory areas in the cerebral cortex of monkeys. The Journal of Neuroscience, 29, 14223–14235. Erlacher, D., & Schredl, M. (2008). Do REM (lucid) dreamed and executed actions share the same neural substrate? International journal of dream research, 1, 7–14. Evans, F. J., Gustafson, L. A., O’Connell, D. N., Orne, M. T., & Shor, R. E. (1970). Verbally induced behavioural responses during sleep. Journal of Nervous and Mental Disease, 150, 171–218. Foulkes, D. (1966). The psychology of sleep. New York: Scribner. Frauscher, B., Iranzo, A., Högl, B., Casanova-Molla, J., Salamero, M., Gschliesser, V., et al. (2008). Quantification of electromyographic activity during REM sleep in multiple muscles in REM sleep behaviour disorder. Sleep, 31(5), 724–731. Frot, M., & Mauguière, F. (1999). Timing and spatial distribution of somato-sensory responses recorded in the upper bank of the sylvian fissure (SII area) in humans. Cerebral Cortex, 9, 854–863. Frot, M., Mauguière, F., Magnin, M., & Garcia-Larrea, L. (2008). Parallel processing of nociceptive A-delta inputs in SII and midcingulate cortex in humans. Journal of Neuroscience, 28, 944–952. Frot, M., Rambaud, L., Guénot, M., & Mauguière, F. (1999). Intracortical recordings of early pain-related CO2 laser evoked potentials in the human second somatosensory (SII) area. Clinical Neurophysiology, 110, 133–145. Gardner, R., Jr., & Grossman, W. I. (1975). Normal patterns in sleep in man. Advances in Sleep Research, 2, 67–107. Gazzaniga, M. S., LeDoux, J. E., & Wilson, D. H. (1977). Language, praxis, and the right hemisphere: Clues to some mechanisms of consciousness. Neurology, 27, 1144–1147. Granda, A. M., & Hammack, J. T. (1961). Operant behaviour during sleep. Science, 133, 1485–1486.
S. Mazza et al. / Brain and Cognition 87 (2014) 7–15 Guenot, M., Isnard, J., Ryvlin, P., Fischer, C., Ostrovsky, K., Mauguiere, F., et al. (2001). Neurophysiological monitoring for epilepsy surgery: The Talairach SEEG method. Stereo Electro Encephalo Graphy. Indications, results, complications and therapeutic applications in a series of 100 consecutive cases. Stereotactic and Functional Neurosurgery, 77, 29–32. Iber, C., Ancoli-Israel, S., Chesson, A., & Quan, S. F. (2007). For the American academy of sleep medicine. In The AASM manual for the scoring of sleep and associated events: Rules, terminology and technical specifications. Westchester: American Academy of Sleep Medicine. Jenkins, I. H., Jahanshahi, M., Jueptner, M., Passingham, R. E., & Brooks, D. J. (2000). Self-initiated versus externally triggered movements. II. The effect of movement predictability on regional cerebral blood flow. Brain, 123, 1216–1228. Jouvet, M. (1979). Memories and ‘‘split brain’’ during dream. 2525 Memories of dream. Revue du Praticien, 29, 27–32. LaBerge, S. (2000). Lucid dreaming: Evidence and methodology. behavioural and Brain Sciences, 23, 962–963. LaBerge, S. (1990). Lucid dreaming: Psychophysiological studies of consciousness during REM sleep. In R. R. Bootzin, J. F. Kihlstrom, & D. L. Schacter (Eds.), Sleep and Cognition. Washington, DC: APA. Lammers, W. J., Badia, P., Hughes, R., & Harsh, J. (1991). Temperature, time-of-night of testing, and responsiveness to stimuli presented while sleeping. Psychophysiology, 28, 463–467. Lau, H. C., Rogers, R. D., Ramnani, N., & Passingham, R. E. (2004). Willed action and attention to the selection of action. Neuroimage, 21, 1407–1415. Lauerma, H., Kaartinen, J., Polo, O., Sallinen, M., & Lyytinen, H. (1994). Asymmetry of instructed motor response to auditory stimuli during sleep. Sleep, 17, 444–448. Llinas, R., & Ribary, U. (1993). Coherent 40-Hz oscillation characterizes dream state in humans. Proceedings of the National Academy of Sciences of the United States of America, 90, 2078–2081. Macdonald, M., Jamshidi, P., & Campbell, K. (2008). Infrequent increases in stimulus intensity may interrupt central executive functioning during Rapid Eye Movement sleep. NeuroReport, 19(3), 309–313. Magnin, M., Bastuji, H., Garcia-Larrea, L., & Mauguiere, F. (2004). Human thalamic medial pulvinar nucleus is not activated during paradoxical sleep. Cerebral Cortex, 14, 858–862. Magnin, M., Rey, M., Bastuji, H., Guillemant, P., Mauguiere, F., & Garcia-Larrea, L. (2010). Thalamic deactivation at sleep onset precedes that of the cerebral cortex in humans. Proceedings of the National Academy of Sciences of the United States of America, 23(107(8)), 3829–3833. Maquet, P. (2000). Functional neuroimaging of normal human sleep by positron emission tomography. Journal of Sleep Research, 9, 207–231. Massimini, M., Ferrarelli, F., Huber, R., Esser, S. K., Singh, H., & Tononi, G. (2005). Breakdown of cortical effective connectivity during sleep. Science, 309, 2228–2232. Massimini, M., Ferrarelli, F., Murphy, M., Huber, R., Riedner, B., Casarotto, S., et al. (2010). Cortical reactivity and effective connectivity during REM sleep in humans. Cognitive Neuroscience, 1, 176–183. Michael, G. A., Boucart, M., Degreef, J. F., & Godefroy, O. (2001). The thalamus interrupts top–down attentional control for permitting exploratory shiftings to sensory signals. NeuroReport, 12, 2041–2048. Mouraux, A., & Iannetti, G. D. (2008). Across-trial averaging of event-related EEG responses and beyond. Magnetic Resonance Imaging, 26, 1041–1054. Nielsen, T. A., Kuiken, D., Alain, G., Stenstrom, P., & Powell, R. A. (2004). Immediate and delayed incorporations of events into dreams: Further replication and implications for dream function. Journal of Sleep Research, 13, 327–336.
15
Nielsen, T. A., & Stenstrom, P. (2005). What are the memory sources of dreaming? Nature, 27, 1286–1289. Nobili, L., Ferrara, M., Moroni, F., De Gennaro, L., Russo, G. L., Campus, C., et al. (2011). Dissociated wake-like and sleep-like electro-cortical activity during sleep. Neuroimage, 15(58(2)), 612–619. Perrin, F., Bastuji, H., & Garcia Larrea, L. (2002). Detection of verbal discordances during sleep. NeuroReport, 13, 1345–1349. Perrin, F., Garcia-Larrea, L., Mauguiere, F., & Bastuji, H. (1999). A differential brain response to the subject’s own name persists during sleep. Clinical Neurophysiology, 110, 2153–2164. Pratt, H., Berlad, I., & Lavie, P. (1999). Oddball event-related potentials and information processing during REM and non-REM sleep. Clinical Neurophysiology, 110, 53–61. Rechtschaffen, A., & Kales, A. (1968). A manual of standardized terminology. In Techniques and Scoring System for Sleep Stages of Human Subjects. Bethesda: US Department of Health Education and Welfare. Rey, M., Bastuji, H., Garcia-Larrea, L., Guillemant, P., Mauguière, F., & Magnin, M. (2007). Human thalamic and cortical activities assessed by dimension of activation and spectral edge frequency during sleep wake cycles. Sleep, 30, 907–912. Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions. Behavioural Neurology, 12, 191–200. Rye, D. B. (2003). Modulation of normal and pathologic motoneuron activity during sleep. In S. Chokroverty, W. A. Hening, & A. S. Walters (Eds.), Sleep and Movement Disorders (pp. 94–119). Philadelphia: Butterworth Heinemann. Sallinen, M., Kaartinen, J., & Lyytinen, H. (1996). Processing of auditory stimuli during tonic and phasic periods of REM sleep as revealed by event-related brain potentials. Journal of Sleep Research, 5, 220–228. Schwarz, S., Greffrath, W., Buèsselberg, D., & Treede, R. D. (2000). Inactivation and tachyphylaxis of heat-evoked inward currents in nociceptive primary sensory neurones of rats. Journal of Physiology (London), 528, 539–549. Steriade, M. (1994). The thalamus and sleep disturbances. In C. Guilleminault, E. Lugaresi, P. Montagna, & P. Gambetti (Eds.), Fatal familial insomnia. Inherited prion diseases, sleep and the thalamus (pp. 177–189). New York: Raven Press. Takahara, M., Nittono, H., & Hori, T. (2006). Effect of voluntary attention on auditory processing during REM. Sleep, 29, 975–982. Talairach, J., & Bancaud, J. (1973). Stereotactic approach to epilepsy: Methodology of anatomo-functional stereotaxic investigations. Progress in Neurolological Surgery, 5, 297–354. Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the human brain. 3Dimensional proportional system an approach to cerebral imaging. Stuttgart, Germany: Georg Thieme Verlag. Terzaghi, M., Sartori, I., Tassi, L., Didato, G., Rustioni, V., LoRusso, G., et al. (2009). Evidence of dissociated arousal states during NREM parasomnia from an intracerebral neurophysiological study. Sleep, 32(3), 409–412. Vogt, B. A. (2005). Pain and emotion interactions in subregions of the cingulate gyrus. Nature Review of Neuroscience, 6, 533–544. Voss, U., Holzmann, R., Tuin, I., & Hobson, J. A. (2009). Lucid dreaming: A state of consciousness with features of both waking and non-lucid dreaming. Sleep, 32(9), 1191–1200. Weiskrantz, L. (1997). Consciousness lost and found: A neuropsychological exploration. New York: Oxford University Press. Zimmerman, W. B. (1970). Sleep mentation and auditory awakening threshold. Psychophysiology, 6, 540–549.