No document available.
Abstract :
[en] Introduction:
The locus coeruleus (LC) is the main source of norepinephrine in the brain (Szabadi, 2013). The LC contributes to the transition between sleep and wakefulness, and between slow wave sleep (SWS) and rapid eye movement sleep (REMS) (Cirelli, Tononi 2005). It is tightly linked to the oscillatory modes found during sleep, such as sleep spindles and slow waves in animals (Eschenko et al., 2012). While the LC appears to be an important structure for sleep, only a few imaging studies evaluated whether the LC is related to sleep variability due to the difficulty of imaging such a small size nucleus in vivo (Keren et al., 2009).
We aimed to investigate the link between LC activity during wakefulness assessed by 7-Tesla MRI and electroencephalogram (EEG) features of sleep. To trigger a response of the LC, we used a visual perceptual rivalry task and an auditory attentional task. Subjects saw an ambiguous stimulus which could trigger spontaneous switches between two perceptions of the same image (Blake & Logothetis, 2002). Switches are considered to recruit the LC (Einhauser et al., 2008; Murphy et al., 2014). Participants also did an auditory oddball task, which can activate LC (Murphy et al., 2014).
Methods:
52 healthy volunteers, including 33 young (22.3 ± 3.2 y; 28 women) and 19 late middle-aged (61.05 ± 5.3 y; 14 women) individuals completed the protocol. They first underwent a structural 7T MRI session, which allowed collecting high-resolution whole-brain T1-weighted images as well as magnetization transfer-weighted turbo-flash (MT-TFL) sequence over a 6 cm slab placed around the LC. The latter was used to create an individual LC-mask in the brain space of each participant. All individual LC-masks were gathered into a probabilistic LC-mask in a standardized group brain space.
Following 1 week of regular sleep times, participants completed an fMRI session in the morning, 2h to 3h after wake-up time, during which they were administered the perceptual rivalry task and oddball task (TR= 2.34s; voxel size 1.4x1.4x1.4 mm³). Participant’s habitual and baseline sleep was recorded in-lab under EEG to extract 4 sleep features of interest depicting some of the most canonical characteristics of sleep (i.e., cumulated power of the theta frequency band during REMS, slow wave energy (SWE), sleep onset latency, and REMS percentage).
We computed a general linear model by the Statistical Parametric Mapping 12 package (SPM12) over the entire brain in individual space. Subject level maps were normalized to MNI space for group statistic visualization. We then extracted the activity estimates over individual LC masks and then conducted generalized linear mixed models (GLMMs) to test for associations between the activity of the LC during both tasks and EEG features of sleep, including age, sex, BMI, and total sleep time as covariates.
Results:
Inspection of whole brain group data identified increased activation within the left LC group template associated with perceptual switches (uncorrected p<0.001; t > 3.27) and the detection of the target sound of the oddball task (uncorrected p<0.001; t >3.26) (Figure 1).
This supported the extraction of individual LC activity in the individual space to associate with sleep EEG metrics. GLMMs on the LC activity and sleep EEG metrics revealed a positive association between the left LC activity during perceptual rivalry task and EEG cumulated power of the theta frequency band (4-8Hz) during REMS (p = 0.004, t=3.02), which reflects the intensity of REMS. A negative association was found between the left LC activity during oddball task and this sleep metric in the older (p = 0.024, t = -2.33) but not in the younger group.
Conclusions:
The results suggest that maybe there is an inverted U-shape relationship between LC activity during wakefulness and REMS intensity. These results contribute to the understanding of sleep physiology.
Funding: FNRS Belgium, ULiège, FEDER, Alzheimer Foundation (SAO-FRA), Wallonia-Brussels federation
