Brain Amide Proton Transfer Imaging of Rat With Alzheimer's Disease Using Saturation With Frequency Alternating RF Irradiation Method

Abstract

Amyloid-β (Aβ) deposits and some proteins play essential roles in the pathogenesis of Alzheimer's disease (AD). Amide proton transfer (APT) imaging, as an imaging modality to detect tissue protein, has shown promising features for the diagnosis of AD disease. In this study, we chose 10 AD model rats as the experimental group and 10 sham-operated rats as the control group. All the rats underwent a Y-maze test before APT image acquisition, using saturation with frequency alternating RF irradiation (APT_SAFARI) method on a 7.0 T animal MRI scanner. Compared with the control group, APT (3.5 ppm) values of brain were significantly reduced in AD models (p < 0.002). The APT_SAFARI imaging is more significant than APT imaging (p < 0.0001). AD model mice showed spatial learning and memory loss in the Y-maze experiment. In addition, there was significant neuronal loss in the hippocampal CA1 region and cortex compared with sham-operated rats. In conclusion, we demonstrated that APT imaging could potentially provide molecular biomarkers for the non-invasive diagnosis of AD. APT_SAFARI MRI could be used as an effective tool to improve the accuracy of diagnosis of AD compared with conventional APT imaging.

Keywords: Alzheimer’s disease; amide proton transfer; chemical exchange saturation transfer; magnetic resonance imaging; saturation with alternating frequency RF irradiation.

FIGURE 1

Learning and memory assessment of AD model group and sham operated control group measured by Y-maze. (A) Alternation behavior: AD model group had significantly decreased the spontaneous alternation behavior compared with the sham operated control group. (B) Time spent in the new arm: The time spent in the new arm of AD model group decreased significantly compared with the sham operated control group. (C) Total distance and the total new arm distance: AD model group had significant decrease of the total distance and the total new arm distance compared with the sham operated control group. (D) There was no significant difference between two groups on the numbers of arm entries. ^∗p < 0.05.

FIGURE 2

(A) APT imaging of sham operated control, (B) APT_SAFARI imaging of sham operated control, (C) APT imaging of AD model, (D) APT_SAFARI imaging of AD model.

FIGURE 3

(A) The T2w image demonstrates the ROIs in the coronal brain slices for the cortex (CX), hippocampus (HI) and thalamus (TH). The red dotted lines indicate the ROI tissues. (B) ROIs for the APT image were manually drawn on the EPI-based image. The blue dotted lines indicate the ROI tissues.

FIGURE 4

(A) The Z-spectra and MTR_asym curve between AD model and sham controls in the whole brain, (B) The Z-spectra and MTR_asym curve between AD model and sham controls in hippocampus regions, (C) The plot for the different region APT effect in AD model and sham groups, (D) The plot for the different region APT_SAFARI effect in AD model and sham groups. ^∗p < 0.05, ^∗∗p < 0.01.

FIGURE 5

(A) T1 map of sham operated control, (B) T1 map of AD model, (C) The plot for the different region T1 map values in AD model and sham groups, (D) T2 map of sham operated control, (E) T2 map of AD model, (F) The plot for the different region T2 map values in AD model and sham groups.

FIGURE 6

HE staining and double-labeling immunofluorescence and confocal microscopy of GFAP (green) of Hippocampal CA1 and cortex. HE staining (A) CA1 region of sham operated control, (B) Cortex of sham operated control, (D) CA1 region of AD model, (E) Cortex of AD model (above all original magnification × 40). AD model group showed significant neuronal loss in the hippocampus CA1 region and cortex (red arrows); Double-labeling immunofluorescence (C) Hippocampal CA1 region of sham operated control, (F) Hippocampal CA1 region of AD model; GFAP staining is strongly enhanced in reactive astrocytes identified by double-labeling immunofluorescence (red arrows) in AD model, Scale bars = 20 μm.

FIGURE 7

(A) HE staining, AD model group showed significant neuronal loss in the hippocampus CA1 region and cortex compared sham operated control, (B) Bar graphs of mean densities of GFAP-positive reactive astrocytes of AD model and sham group. ^∗∗p < 0.01.

FIGURE 8

(A) The linear regression analysis of the correlation between alternation behavior (%) and the APT (%) of the hippocampus in AD model rats (R² = 0.9453, p < 0.0001 and R² = 0.8077, p = 0.0004 for the APT_SAFARI and APT values, respectively). (B) The linear correlation between the APT (%) and GFAP-positive astrocytes/field of the hippocampus in AD model rats (R² = 0.9410, p < 0.0001 and R² = 0.7598, p = 0.0010 for the APT_SAFARI and APT values, respectively).