Neurometabolic and structural alterations of medial septum and hippocampal CA1 in a model of post-operative sleep fragmentation in aged mice: a study combining 1H-MRS and DTI

Abstract

Post-operative sleep disturbance is a common feature of elderly surgical patients, and sleep fragmentation (SF) is closely related to post-operative cognitive dysfunction (POCD). SF is characterized by sleep interruption, increased number of awakenings and sleep structure destruction, similar to obstructive sleep apnea (OSA). Research shows that sleep interruption can change neurotransmitter metabolism and structural connectivity in sleep and cognitive brain regions, of which the medial septum and hippocampal CA1 are key brain regions connecting sleep and cognitive processes. Proton magnetic resonance spectroscopy (1H-MRS) is a non-invasive method for the evaluation of neurometabolic abnormalities. Diffusion tensor imaging (DTI) realizes the observation of structural integrity and connectivity of brain regions of interest in vivo. However, it is unclear whether post-operative SF induces harmful changes in neurotransmitters and structures of the key brain regions and their contribution to POCD. In this study, we evaluated the effects of post-operative SF on neurotransmitter metabolism and structural integrity of medial septum and hippocampal CA1 in aged C57BL/6J male mice. The animals received a 24-h SF procedure after isoflurane anesthesia and right carotid artery exposure surgery. 1H-MRS results showed after post-operative SF, the glutamate (Glu)/creatine (Cr) and glutamate + glutamine (Glx)/Cr ratios increased in the medial septum and hippocampal CA1, while the NAA/Cr ratio decreased in the hippocampal CA1. DTI results showed post-operative SF decreased the fractional anisotropy (FA) of white matter fibers in the hippocampal CA1, while the medial septum was not affected. Moreover, post-operative SF aggravated subsequent Y-maze and novel object recognition performances accompanied by abnormal enhancement of glutamatergic metabolism signal. This study suggests that 24-h SF induces hyperglutamate metabolism level and microstructural connectivity damage in sleep and cognitive brain regions in aged mice, which may be involved in the pathophysiological process of POCD.

Keywords: 1H-MRS; CA1; DTI; aging; hippocampus; medial septum; post-operative sleep disturbance; sleep fragmentation.

FIGURE 1

Flow chart of the procedures and magnetic resonance imaging (MRI) processing in the study. (A) Experimental protocol. Aged C57BL/6J mice were divided into three groups: Control, isoflurane and surgery (I/S), and I/S + sleep fragmentation (I/S + SF) groups. Group I/S + SF received 24-h sleep fragmentation following isoflurane anesthesia and right carotid artery exposure surgery, group I/S received isoflurane anesthesia and right carotid artery exposure surgery, and group Control received no intervention. MRI study of diffusion tensor imaging (DTI) and proton magnetic resonance spectroscopy (1H-MRS), and behavior tests of Y maze, 1 h and 24 h novel object recognition (NOR) were performed. (B) 1H-MRS data processing and analysis. The volumes of interest were set at medial septum (MS) and CA1. Typical spectra were recorded to analyze the changes of neurometabolic signals. Glu, glutamate; Gln, glutamine; Glx, glutamate/glutamine; GABA, gamma-aminobutyric acid; tCho, total choline; mI, myo-Inositol; Tau, taurine; NAA, N-acetylaspartate; Cr, creatine; Lac, lactate; Lip, lipid. (C) DTI data processing and analysis (reprinted from DSI studio). MS and CA1 were identified as regions of interest. DTI signal intensity of the regions of interest were subjected to structural integrity analysis.

FIGURE 2

Comparison of 1H-MRS data in MS and CA1 regions in the three groups. Column charts of glutamate (Glu) (A), glutamine (Gln) (B), glutamate/glutamine (Glx) (C), gamma-aminobutyric acid (GABA) (D), total choline (tCho) (E), myo-Inositol (mI) (F), taurine (Tau) (G), and N-acetylaspartate (NAA) (H) show alterations of the concentration ratio of the neuro-metabolites to creatine (Cr). Data are expressed as mean ± SME (n = 10/group). Symbols representing significant differences in the ANOVA of the Control, I/S and I/S + SF groups. ^nsP > 0.05; *P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

FIGURE 3

Comparison of DTI data in MS and CA1 regions in the three groups. (A) Representative images of DTI in the MS and CA1 regions of mice in the Control, I/S and I/S + SF groups. (B) Quantitative analysis of fractional anisotropy (FA) in the Control, I/S and I/S + SF groups. (C) Quantitative analysis of mean diffusivity (MD) in the Control, I/S and I/S + SF groups. (D) Quantitative analysis of axial diffusivity (AD) in the Control, I/S, and I/S + SF groups. (E) Quantitative analysis of radial diffusivity (RD) in the Control, I/S and I/S + SF groups. Data are expressed as mean ± SME (n = 10/group). Symbols representing significant differences in the ANOVA of the Control, I/S and I/S + SF groups. ^nsP > 0.05; *P < 0.05; ^**P < 0.01; ^***P < 0.001.

FIGURE 4

Hippocampal volume. (A) Representative T2-weighted images of the left (blue) and right (red) hippocampus in the Control, I/S and I/S + SF groups. (B) Quantitative analysis of the volume of left hippocampus in in the Control, I/S and I/S + SF groups. (C) Quantitative analysis of the volume of left hippocampus in in the Control, I/S and I/S + SF groups. Data are expressed as mean ± SME (n = 10/group). ns representing no significant difference with P > 0.05 in the ANOVA of the Control, I/S and I/S + SF groups.

FIGURE 5

Short-term spatial memory in the Y maze test. (A) Representative trajectories of mice in the Y maze in the Control, I/S and I/S + SF groups. (B–D) Quantitative analysis of the percentage of time, distance and entries in the novel arm in the Control, I/S and I/S + SF groups. (E) Quantitative analysis of total distance in the Control, I/S and I/S + SF groups. (F) Quantitative analysis of average speed in the Control, I/S and I/S + SF groups. Data are expressed as mean ± SME (n = 10/group). Symbols representing significant differences in the ANOVA of the Control, I/S and I/S + SF groups. ^nsP > 0.05; *P < 0.05; ^**P < 0.01.

FIGURE 6

Short- and long-term recognition memory in the NOR test. (A) Representative trajectories of mice in the NOR in the Control, I/S and I/S + SF groups. (B) Quantitative analysis of discrimination index in 1-h NOR in the Control, I/S and I/S + SF groups. (C) Quantitative analysis of discrimination index in 24-h NOR in the Control, I/S and I/S + SF groups. Data are expressed as mean ± SME (n = 10/group). Symbols representing significant differences in the ANOVA of the Control, I/S and I/S + SF groups. ^nsP > 0.05; *P < 0.05; ^**P < 0.01.