Brain Neurochemical effects of long-term sleep fragmentation investigated in mice at 14.1T using 1H-MRS

Introduction: Sleep-fragmentation (SF) is a common symptom associated to disturbances such as sleep apneas (1) Previous studies showed that SF induces daytime sleepiness and cognitive impairments associated to metabolic changes (2, 3). In the present study, the neurochemical profile of mice submitted to SF during 14 days was examined using H-MRS at 14.1T both in hippocampus and the frontal cortex. The goal of this study was to investigate changes linking sleep fragmentation to modifications in neurotransmitter concentrations. Materials and Methods: Animal preparation: C57BL6 mice (n=10, BW=25±2g) were individually transferred in rotating Plexiglas cylindrical cages equipped with three internal Plexiglas static walls to be submitted to SF. The rotations consisted in a mild rotation (1.5 rpm) of the floor of the cages lasting 20s every minute. Such a rotation brings the mouse in contact with one of the walls and wakes it up. Animals had free access to food and water. Each mouse was left to adapt to its new environment for a week before the first MRS session and prior to the start of the cage rotation to prevent stress. Mice underwent a first MRS session in their basal state and a second MRS 14days after .Prior to MRS, mice were gently manipulated to avoid stress and anesthetized with 2% isoflurane in a mixture of O2 and N2O. Each of them was positioned in a dedicated mouse holder equipped with ear and teeth bars and anesthetized with 1% isoflurane in O2/N2O through a nose cone. Body temperature was maintained at 36.5±1oC by circulating warm water around the animals.H-MRS All the experiments were performed on a 14.1T/26cm horizontal bore magnet ( Magnex, Varian). A quadrature T/R 14-mm surface coil was used. Localized proton spectroscopy was performed using SPECIAL (4) in a 4±1μl voxel for the hippocampus and in a 12±1μl voxel for the frontal cortex placed using multi-slice fast spin echo (FSEMS) images. Field homogeneities were adjusted using FASTMAP (5) up to water linewidth=20-25Hz. At least 40 blocks of 16 FIDs were acquired for a total acquisition time of 40minutes. Metabolite concentrations were calculated using LCmodel (6). A paired t-test was used to compare pre and post-SF metabolite concentrations. A pvalue<0.05 was considered significant. Results: Fig.1a and 1b present the comparison of mean metabolite concentrations pre and post-SF in the hippocampus and the frontal cortex respectively. In the hippocampus, mean GABA concentrations decreased significantly between basal and SF states (pvalue =0.006) and a significant decrease was also detected for lactate (pvalue<0.05). In the cortex, no significant differences were observed between basal and SF states, however a tendency to decrease was also observed for GABA (pvalue=0.21). In Fig.2, the evolution of GABA, and Lactate concentrations between basal and SF states are shown individually in the hippocampus (Black) and in the cortex (red).