PS1 mouse model for Alzheimer’s disease: Applying MRS, DTI, and ASL

Accepted Manuscript Research report Multifaceted Assessment of the APP/PS1 Mouse Model for Alzheimer’s disease: Applying MRS, DTI, and ASL Zhiwei Shen, Jianfeng Lei, Xueyuan Li, Zhanjing Wang, Xinjie Bao, Renzhi Wang PII: DOI: Reference:

S0006-8993(18)30415-3 https://doi.org/10.1016/j.brainres.2018.08.001 BRES 45898

To appear in:

Brain Research

Received Date: Revised Date: Accepted Date:

4 June 2018 28 July 2018 1 August 2018

Please cite this article as: Z. Shen, J. Lei, X. Li, Z. Wang, X. Bao, R. Wang, Multifaceted Assessment of the APP/ PS1 Mouse Model for Alzheimer’s disease: Applying MRS, DTI, and ASL, Brain Research (2018), doi: https:// doi.org/10.1016/j.brainres.2018.08.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Multifaceted Assessment of the APP/PS1 Mouse Model for Alzheimer’s disease: Applying MRS, DTI, and ASL

Zhiwei Shena, Jianfeng Leib, Xueyuan Lia, Zhanjing Wangb, Xinjie Baoa,*, Renzhi Wanga,*

aDepartment

of Neurosurgery, Peking Union Medical College Hospital, Chinese

Academy of Medical Sciences & Peking Union Medical College, Beijing, China bCenter

for Medical Experiments and Testing, Capital Medical University,

Beijing, China

*Correspondence

Xinjie Bao, Department of Neurosurgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China, 100730, E-mail: [email protected]; Renzhi Wang, Department of Neurosurgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China, 100730, E-mail: [email protected];

Abstract Transgenic animal models of Alzheimer’s disease (AD) can mimic pathological and behavioral changes occurring in AD patients, and are usually viewed as the first choice for testing novel therapeutics. Validated biomarkers, particularly non-invasive ones, are urgently needed for AD diagnosis or evaluation of treatment results. However, there are few studies that systematically characterize pathological changes in AD animal models. Here, we investigated the brain of 8-month-old amyloid precursor protein/presenilin 1 (APP/PS1) transgenic and wild-type (WT) mice, employing 7.0-T magnetic resonance imaging (MRI). Magnetic resonance spectroscopy (MRS), diffusion tensor imaging (DTI), and arterial spin labeling (ASL) were obtained through micro-MRI scanning. After MRI examination in both transgenic (n = 12) and WT (n = 12) mice, immunohistochemical staining and ultrastructural analysis were subsequently performed. Cerebral blood flow (CBF) was significantly decreased in the left hippocampus, left thalamus, and right cortex of AD mice (P < 0.05). Moreover, MRS showed significantly changed NAA/Cr, Glu/Cr, and mI/Cr ratios in the hippocampus of transgenic mice. While only NAA/Cr and mI/Cr ratios varied significantly in the cortex of transgenic mice. Regarding DTI imaging, however, the values of FA, MD, DA and DR were not significantly different between transgenic and WT mice. Finally, it is worth noting that pathological damage of metabolism, CBF, and white matter was more distinct between transgenic and WT mice by pathological examination. Altogether, our results suggest that intravital imaging evaluation of 8-month-old APP/PS1 transgenic mice by MRS and ASL is an alternative tool for AD research.

Keywords: Alzheimer’s disease, MRI, MRS, ASL, DTI, APP/PS1 transgenic mice

1. Introduction Alzheimer’s disease (AD) is the most common form of dementia, associated with multifaceted dysfunction involving brain metabolism, cerebral blood flow (CBF), and white matter. With the increase of AD research, comprehensive and non-invasive assessment of pathological changes in the brain of AD patients is an urgent requirement. In this regard, rapid development of imageological techniques, especially magnetic resonance imaging (MRI), currently offers a realistic possibility. In AD research, different types of MRI sequences have been used for detection of AD pathological changes (Spencer et al., 2013; Sun et al., 2005; Verclytte et al., 2016; Westman et al., 2010), establishing a foundation for systematic assessment of AD in the future. Commonly, diffusion tensor imaging (DTI) is used to estimate white matter connectivity through multiple diffusion tensor measures, such as fractional anisotropy (FA) and mean diffusivity (MD) (Alexander et al., 2007). Magnetic resonance spectroscopy (MRS) is an analytical technique that uses 1H signals to determine the relative shift of targeted metabolites, including N-acetyl aspartate (NAA), choline (Cho), creatine (Cr), and lactate (Gujar et al., 2005). Further, arterial spin labeling (ASL) is a tissue perfusion imaging technique that employs magnetically labeled arterial blood 1H protons as an intrinsic tracer (Petcharunpaisan et al., 2010). Thus far, few studies have reported on the possibility of combinational assessment of different MRI methods (and in particular, MRS, ASL, and DTI together) in evaluation of AD mice. Currently, the amyloid precursor protein/presenilin 1 (APP/PS1) transgenic mouse model has been widely used to explore underlying AD pathophysiology and test novel therapies (Li et al., 2016b). This AD animal model replicates formation of amyloid-beta (Aβ) plaques in the mouse brain, with obvious learning and memory dysfunction at 8 months-old (Li et al., 2016d). Consequently, this model can be regarded as a valuable tool for validation and clarification of potential neuroimaging techniques. Here, we examined the brain of 8-month-old APP/PS1 mice and age-matched wildtype (WT) littermates by applying MRS, DTI, and ASL. Afterwards, we correlated

MRI results with histological staining and ultrastructural observations to determine whether in vivo imaging reflects pathological changes in APP/PS1 mice, including metabolism, CBF, and white matter structure. Accordingly, this study may be valuable for choosing appropriate tools for non-invasive evaluation of AD pathologies or novel therapies in the APP/PS1 mouse model.

2. Results 2.1 Decreased regional CBF in APP/PS1 mice detected by ASL For each animal investigated, a slice of 2 mm thickness was examined, in which the hippocampus was identified (Figure 1A). From both the hippocampus and brain, ROIs were transferred from the corresponding anatomical images to CBF maps. Average CBF values (in mL min-1 100 g tissue-1) were collected from these ROIs. As shown in Figure 1B, 8-month-old APP/PS1 transgenic mice demonstrated relatively decreased regional CBF in most brain areas (right hippocampus, right cortex, right thalamus, left hemisphere, left amygdala, left hippocampus, left cortex, and left thalamus) compared with age-matched WT mice. Further, statistically significant hypoperfusion was observed in the left hippocampus, left thalamus, and right cortex (P < 0.05, FDR corrected), suggesting that these regions might be the earliest affected brain areas in AD.

Hypoperfusion

in

AD

mouse

brain

was

also

demonstrated

by

immunohistochemistry. In the hippocampus of AD mice, average MVD determined by CD31 staining was significantly lower compared with WT mice (26.06 ± 4.87 vs. 44.27 ± 5.67, P = 0.03). A similar result was observed in the cortex (13.03 ± 2.90 vs. 28.87 ± 3.51, P = 0.005) (Figure 1C). Overall, these results suggest that ASL is a useful tool for evaluating reduced CBF in AD mice.

Figure 1. Arterial spin labeling (ASL) magnetic resonance (MR) images and vascular immunohistochemical staining of APP/PS1 (n=8) and WT (n=8) mice. (A) Left, regions of interest (cortex, hippocampus, thalamus, and amygdala) in echoplanar imaging–fluid-attenuated inversion recovery (EPI–FLAIR) maps were selected for cerebral blood flow (CBF) measurement. Middle and right, representative CBF images of APP/PS1 and wild-type (WT) groups. High CBF is indicated in red. (B) Analysis of CBF in specific brain areas by ASL MR imaging; based upon FDRcorrected t-statistics (P=0.05). (C) Corresponding immunohistochemical staining of CD31. Scale bar: 20 μm. Values represent mean ± SEM. *P < 0.05.

2.2 Slight differences between APP/PS1 and WT mice detected by DTI Next, we analyzed quantitative measurements (including FA, MD, DA and DR) in ROIs from both white and gray matter structures. As shown in Figure 2A and B, APP/PS1 mice showed no significant differences compared with WT mice. In comparison, mean FA values were slightly higher in the olfactory bulb, left hippocampus, and left internal capsule of APP/PS1 mice. Other brain regions (including the amygdala, right hippocampus, cortex, right internal capsule, cingulate

gyrus, and corpus callosum) showed relatively lower FA values in APP/PS1 mice. For MD parametric statistics, almost all ROIs (including the olfactory bulb, amygdala, corpus callosum, hippocampus, cortex, internal capsule, and cingulate gyrus) exhibited relatively lower MD values in APP/PS1 mice. However, none of them showed a significant difference in MD value between APP/PS1 and WT mice (P >0.05, FDR corrected). Similarly, all ROIs (including the amygdala, corpus callosum, hippocampus and cortex) exhibited relatively lower DA values in APP/PS1 mice (P>0.05, FDR corrected). However, the DR values in corpus callosum and cortex of APP/PS1 mice were slightly increased compared to the controls (P>0.05, FDR corrected). Nevertheless, MBP staining showed a more abundant number of MBP-positive fibers in the cortex and corpus callosum of the APP/PS1 group compared with the control group. Moreover, mean density (IOD/area) was significantly lower in APP/PS1 mice than WT mice (P < 0.05) (Figure 2C).

Figure 2. Diffusion tensor imaging (DTI) scanning and myelin basic protein (MBP) staining in APP/PS1 (n=8) and control (n=8) mice. (A) Representative color-encoded fractional anisotropy (FA) directional map of continuous scanning in

APP/PS1 and wild-type (WT) groups. (B) Comparison of FA, mean diffusivity (MD), radial diffusivity (DR) and axial diffusivity (DA) values in selected brain areas; based upon FDR-corrected t-statistics (P=0.05). (C) MBP staining in the cortex and corpus callosum. Values represent mean ± SEM. *P < 0.05. Scale bar: 20 μm.

2.3 Distinct neurochemical changes in the hippocampus and cortex of APP/PS1 transgenic mice To examine brain metabolic differences between APP/PS1 and WT mice, 1H-MRS was obtained from a corresponding coronal T2-weighted image of both the hippocampus and cortex (Figure 3A, red box). Compared with WT mice, decreased NAA and Glu peaks and increased mI peak were found in the cortex and hippocampus of APP/PS1 mice. Quantification analysis showed significantly reduced NAA/Cr (1.53±0.03 vs. 1.69±0.03, P=0.032, FDR corrected) and Glu/Cr (0.49±0.03 vs. 0.61±0.03, P=0.04, FDR corrected) ratios in the hippocampus of APP/PS1 mice, while mI/Cr (0.61±0.04 vs. 0.42±0.02, P=0.016, FDR corrected) ratio was significantly higher compared with WT mice (Figure 3A). In the cortex, only decreased NAA/Cr (1.55±0.08 vs. 1.84±0.03, P=0.036, FDR corrected) ratio and elevated mI/Cr (0.89±0.09 vs. 0.56±0.05, P=0.036, FDR corrected) ratio were significantly different between APP/PS1 and WT mice (Figure 3A). No significant difference was observed in Cho/Cr ratio, regardless of brain region (P > 0.05, FDR corrected). Furthermore, Nissl staining was performed to assess neuronal activity in the brain of APP/PS1 mice (Figure 3B). Compared with WT mice, APP/PS1 mice showed neuronal loss and disappearance of Nissl bodies in the cortex and hippocampus (Figure 3B). Quantification analysis showed significantly reduced neuronal number in APP/PS1 mice compared with WT mice (P < 0.01) (Figure 3B). Altogether, these results suggest that NAA/Cr and mI/Cr ratios may be potential biomarkers for discriminating between APP/PS1 and WT mice.

Figure 3. Neuronal metabolic activity detected by magnetic resonance spectroscopy (MRS) in APP/PS1 (n=8) and WT (n=8) mice. (A) Localization and representative MRS spectra from the hippocampus and frontal cortex (red box) of T2weighted scans of APP/PS1 and wild-type (WT) mice. 1H MRS exhibits a creatine (Cr) peak at 3.0 ppm, N-acetyl aspartate (NAA) peak at 2.0 ppm, glutamate (Glu) peak at 2.2 ppm, and myo-inositol (mI) peak at 3.6 ppm. Quantification of the relative ratio of each metabolite to Cr in the cortex and hippocampus; based upon FDRcorrected t-statistics (P=0.05). (B) Nissl-stained images and comparison of neuronal counts in the frontal cortex and hippocampus of APP/PS1 and WT mice. Scale bar: 20 μm. Values represent mean ± SEM. *P < 0.05; **P < 0.01.

2.4 Aβ deposition and pathological changes validated by TEM in the brain of APP/PS1 mice Ultrastructural analysis showed the presence of Aβ plaques and degenerated changes in the neocortex and hippocampus of APP/PS1 mice. Dense and mature forms of senile plaques (which are typical of neuritic plaques) were observed in the neocortex of APP/PS1 mice (Figure 4A). These plaques were surrounded by abundant

phagolysosomes (Figure 4A). Moreover, high resolution TEM images showed that Aβ plaques were filled with tight Aβ fibers (Figure 4D). We also observed formation of huge lysosomes containing multivesicular bodies, indicating dysfunction of neuronal lysosomes in APP/PS1 mouse brain (Figure 4B). Additionally, focal demyelination, dilated perivascular space, and neural apoptosis were found in the hippocampus of APP/PS1 mice (Figure 4C, E, and F).

Figure 4. Transmission electron microscopy findings in the neocortex and hippocampus of APP/PS1 mice. (A) Low-magnification view of robust amyloid plaque deposition (stars) surrounded by abundant phagolysosomes (white arrows). (B) Ultrastructural images of lysosomes and multivesicular bodies. (C) Perivascular space dilation and collapse of capillaries were present. (D) High-magnification of amyloid plaques composed of compact fibrillary amyloid. (E) Degenerated myelin sheath. (F) Apoptotic neurons: misshapen nuclei with condensed chromatin masses attached to the nuclear membrane.

3. Discussion

In vivo 7T MRI, high-field MRI with high resolution, is regarded as a novel tool for evaluation of pathological changes in AD. Previous MRI studies in AD mice have mainly focused on brain volume changes (Hayes et al., 2014; Yin et al., 2014) and Aβ burden (Kim et al., 2014; Li et al., 2016a). In this study, we assessed pathological changes of APP/PS1 transgenic mice and WT mice by applying a series of MRI sequences (specifically ASL, DTI, and MRS) to identify comprehensive AD biomarkers. After examination, the MRI results were verified by histological examination and ultrastructural analysis. Although Wells et al. performed a similar multi-parametric imaging study (including DTI and ASL) in a tau-pathology AD mouse model(Wells et al., 2015), its evidence of blood perfusion and microstructure from pathological studies was deficient. Our data show that CBF and brain metabolism were markedly changed in AD mice while white matter structure was barely significant compared with WT mice. Nevertheless, the pathological results of CBF, brain metabolism and white matter structure in light microscopy and electron microscopy were significantly different between AD and WT groups. Currently, ASL is gaining growing popularity in clinical evaluation of AD patients (Zhang, 2016), but studies on its use in AD mice are still limited. Recently, several studies using MRI showed cortical hypoperfusion in an AD transgenic mouse model (Decker et al., 2018; Hebert et al., 2013; Massaad et al., 2010), but few studies have examined other brain regions apart from the cortex, with the connection between histology and MRI rarely investigated. Here, we investigated bilateral brain areas of the hippocampus, cortex, amygdala, and thalamus. Consequently, we found significantly decreased CBF in the left hippocampus, left thalamus, and right cortex of AD mice. This is in agreement with the findings of a previous study (Kobayashi et al., 2008), which found markedly lower CBF in the left hippocampal and right temporal regions in patients with mild AD. Furthermore, decreased blood flow was validated by our CD31 staining results showing decreased vascular number. Therefore, ASL offers the potential to visualize brain vascular dysfunction. Further, the hippocampus and cortex might be influenced in the early AD stage. As the disease progresses, diffuse decreased blood flow would occur throughout the brain.

In recent years, MRS has been widely used to detect neurochemical alterations in specific brain regions of AD mice. However, most MRS studies in AD mice have focused on the hippocampus (Chen et al., 2012; Kuhla et al., 2017; Liang et al., 2017). In the present study, we obtained MRS data from the cortex and hippocampus of control and APP/PS1 transgenic mice. In the hippocampus, NAA/Cr and Glu/Cr ratios were significantly decreased in AD mice, while mI/Cr ratio, a marker for the glial cell proliferation, was significantly increased compared with WT mice, consistent with a report by Chen et al., (Chen et al., 2012). However, in the cortex, only reduced NAA/Cr ratio and increased mI/Cr ratio showed significant differences in AD mouse brain. This may be because metabolic dysfunction of neuron and glia in the hippocampus initially contributes to AD progression before cortical involvement. Furthermore, decreased NAA/Cr ratio and increased mI/Cr ratio might be more sensitive than other biomarkers for detection or evaluation of AD, at least in 8-monthold APP/PS1 mice. Due to the clear directionality of water diffusion, DTI has been widely used to assess white matter structure of the brain. However, DTI findings in AD mice are inconsistent. For example, one study reported that APP/PS1 mice had higher FA values in the hippocampus (Shu et al., 2013), while another study in 3 × Tg mice showed the opposite result (Snow et al., 2017). In our study, FA value was not significantly different between APP/PS1 and control groups. Still, FA value in AD mice was slightly higher in the left hippocampus, olfactory bulb, and left internal capsule. This might be due to the left hippocampus being the early region infiltrated in AD mice, consistent with lower CBF in the left hippocampus, as detected by ASL sequence. In contrast to previous studies (Qin et al., 2013; Shu et al., 2013), MD value in our study showed no significant difference. However, average MD value was slightly decreased in most brain regions in the APP/PS1 group, which is consistent with another study (Sun et al., 2005). Furthermore, DR and DA value were also compared in our study. We found a slight increase of DR in the corpus callosum and cortex of APP/PS1 mice, which may reflect demyelination in these brain regions. Overall, in our study, DTI imaging results are not significantly different. This might

be due to pathological changes of white matter fibers in 8-month-old APP/PS1 mice, which are not significant to be detected by DTI sequence, as age-dependent Aβ accumulation is linked with white matter abnormalities (Song et al., 2004). For example, studies have shown significant differences by DTI imaging with AD transgenic mice older than 12 months old (Qin et al., 2013; Shu et al., 2013; Sun et al., 2005; Zerbi et al., 2013). Although we evaluated multifaceted brain changes from both MRI imaging and pathological examination, there was one main limitation to our study: we only tested APP/PS1 mice at 8 months old and did not assess longitudinal changes. Moreover, this AD mouse model doesn’t include all pathological changes in AD patients, translation of the MRI findings in this study to AD patients should be treated with caution. Thus, the observed results imply this combinational MRI sequences is merely appropriate for evaluation of 8-month-old APP/PS1 mice or elder ones. In conclusion, this is the first study to apply ASL, DTI, and MRS simultaneously for evaluation of AD transgenic mice to detect pathological changes in vivo. Although differences in DTI results between AD mice and WT mice were not as significant as MRS and ASL findings, our findings suggest that ASL, MRS, and DTI are associated in some way with CBF, brain metabolism, and white matter abnormalities. Moreover, compared with age-matched WT mice, we found the hippocampus (particularly the left hippocampus) to be more damaged than other brain regions, providing new insight into understanding of AD pathophysiology. Finally, further studies require AD patients and more animals to fully validate the meaning of our findings.

4. Experimental procedures 4.1 Animal model Experimental protocols were performed in strict accordance with the Institutional Animal Care and Use Committee of the Institute of Laboratory Animal Science of Peking Union Medical College. Double-transgenic APP/PS1 male mice (n=12; 8 months of age) were used. APP/PS1 mice were produced by co-injecting APPswe and

PS1∆E9 vector, and confirmed by PCR genotyping using appropriate oligonucleotide primers, as previously described (Wang et al., 2010). Non-transgenic C57BL/6J male littermates (n=12) were used as controls. All mice were housed under a 12-hour light/12-hour dark schedule and had free access to food and water.

4.2 MRI scanning MRI experiments were performed on a 7.0T MRI scanner (Bruker Pharmascan 70/16; Bruker, Germany), equipped with a 23-mm surface coil to receive signals and a 12cm diameter self-shielded gradient system. The system was interfaced to a Linux PC running Topspin 2.0 and Paravision 5.1 software (Bruker BioSpin; Bruker). Before MRI examination, mice were anesthetized with 5% isoflurane/95% O2 mixture for induction in the interior of a plexiglass chamber, and then administered 2% isoflurane/98% O2 mixture for maintenance while in the center of the magnetic field. Animal temperature was carefully maintained at normal and respiratory rate closely monitored using an animal physiological guarding system. T2-weighted images were obtained using a rapid acquisition method with relaxation enhancement (RARE) and the following parameters: repetition time (TR) = 3,500 ms, echo time (TE) = 33 ms, RARE factor = 4, field of view (FOV) = 21 × 21 mm, acquisition matrix = 256 × 256, and slice thickness = 0.6 mm. Acquired images were used for subsequent location of MRS. 1H-MRS data were acquired for sizing two regions of interest (ROI) using a point-resolved water suppression pulse sequence (PRESS) with the following parameters: TR/TE = 2500/20 ms, ejection fraction (EF) = 1000, hippocampal voxel = 1.5 × 1.5 × 1.5 mm, cortical voxel = 1 × 1 × 1.5 mm, and time (T) = 40 min. 1H-MRS image data were analyzed by Bruker processing software 2D WIN-NMR (BrukerFranzen Analytik, Bruker). For each ROI, metabolites were evaluated including NAA, Cho, glutamate (Glu), myo-inositol (mI), and Cr. Creatine was used as an internal reference to calculate relative levels of other metabolites as it is not affected in various diseases. DTI was acquired using a multi-slice and echo-planar imaging sequence with TR/TE = 6300/21 ms, 30 diffusion encoding directions, slice thickness of 0.6 mm, two b values = 0 and 1000 s/mm2, and FOV = 2.1 × 2.1 cm. FA images

were reconstructed with Paravision version 5.1 software (Bruker, Pharmascan). The parameters FA, MD, DA and DR were measured using standard methods(Shu et al., 2013) in the following regions: olfactory bulb, thalamus, corpus callosum, cingulate gyrus, bilateral cortical areas, amygdala, hippocampus, and the internal capsule. ASL images were acquired from echo-planar imaging–fluid-attenuated inversion recovery (EPI–FLAIR) sequences and reconstructed with Paravision version 5.1 software (PharmaScan). Acquisition parameters were TR/TE = 18000/16 ms, FOV = 1.8 × 1.8 cm, slice thickness = 2 mm, matrix size = 128 × 128, and number of segments = 2.

4.3 Histological examination After MRI examination, mice were anesthetized and subjected to cardiac perfusion with saline solution followed by 4% paraformaldehyde (pH 7.4, 4C). Next, the entire brain was dissected and post-fixed in the same fixative for at least 24 h. Nissl staining was used to determine neuronal number. As described previously (Li et al., 2016c), selected sections were sequentially hydrated, incubated in 1% cresyl violet acetate, washed with phosphate-buffered saline (PBS), decolorized with 95% ethanol, and dehydrated. Expression levels of CD31 and myelin basic protein (MBP) were correlated with pathological changes in capillary endothelial cells and white matter, respectively (Giannoni et al., 2016; Kam et al., 2002). Immunohistochemical staining was performed as previously described for CD31 (Drachman et al., 2017) and MBP (Spencer et al., 2013). Briefly, brain sections were incubated at 4C for 40 h with anti-MBP (1:400, servicebio, Wuhan, China) and anti-CD31 (1:300, servicebio, Wuhan, China) antibodies. Subsequently, sections were incubated at 37C for 60 min with biotin-labeled secondary antibody. Counterstaining was performed with hematoxylin. Finally, all sections were coverslipped and visualized with a light microscope (40× objective) (Nikon Eclipse C1; Nikon, Tokyo, Japan). Labeled cells were quantitatively analyzed by a blinded observer. Expression levels of MBP were determined from integrated optical density (IOD) values. Microvessel density (MVD) was measured as total blood vessels per mm2, based on endothelial CD31 staining. Number of Nissl-stained cells was also calculated by neuronal count per mm2. Six

random fields of bilateral hippocampus or cortex was quantified from each slice. Ultrastructural analysis For tissue processing, mice were transcardially perfused with 2.5% glutaraldehyde and 1% paraformaldehyde in 0.1 mol/L phosphate buffer (pH 7.4, 4C). Subsequently, the brains were rapidly removed and cut into 1 mm sagittal slices on ice. The hippocampus and frontal lobe were dissected into blocks of approximately 1 mm3, and post-fixed in 2.5% glutaraldehyde at 4C. Until processing, the tissue was dehydrated in gradient ethanol and propylene oxide, embedded in epoxy resin, cut into ultra-thin sections, stained with uranyl acetate, and examined by transmission electron microscopy (TEM) (JEM-1400plus; JEOL Ltd, Tokyo, Japan).

4.4 Statistical analysis Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) version 22.0 (IBM Corporation, Armonk, NY, USA) and GraphPad Prism 6.0 (GraphPad Software, Inc., La Jolla, CA, USA). Data are expressed as mean ± standard error of the mean (SEM). Levene’s test was performed to assure homogeneity of variance between APP/PS1 mice and WT mice. Differences between two groups were compared using an independent samples t-test (two sided). The results were corrected after multiple comparisons by applying a threshold of P<0.05 with the false discovery ratio (FDR).

Declarations of interest None.

Author contributions RW and XB designed the research. ZS, JL, XL, and ZW performed the research. ZS and XB analyzed the data. ZS and RW wrote the article.

Acknowledgements

This work was supported by the National High Technology Research and Development Program of China (2013AA020106 and 2014AA020513), National Basic Research Program of China (2014CB541603), National Natural Science Foundation of China (81671272), Beijing Natural Science Foundation (7182134), CAMS initiative for Innovative Medicine (CAMS-I2M) (2016-I2M-1-017), and Beijing Nova Program (Z181100006218003).

References Alexander, A.L., Lee, J.E., Lazar, M., Field, A.S., 2007. Diffusion tensor imaging of the brain. Neurotherapeutics. 4, 316-29. Chen, S.Q., Cai, Q., Shen, Y.Y., Wang, P.J., Teng, G.J., Zhang, W., Zang, F.C., 2012. Age-related changes in brain metabolites and cognitive function in APP/PS1 transgenic mice. Behav Brain Res. 235, 1-6. Decker, Y., Muller, A., Nemeth, E., Schulz-Schaeffer, W.J., Fatar, M., Menger, M.D., Liu, Y., Fassbender, K., 2018. Analysis of the vasculature by immunohistochemistry in paraffin-embedded brains. Brain Struct Funct. 223, 1001-1015. Drachman, D.A., Smith, T.W., Alkamachi, B., Kane, K., 2017. Microvascular changes in Down syndrome with Alzheimer's-type pathology: Insights into a potential vascular mechanism for Down syndrome and Alzheimer's disease. Alzheimers Dement. 13, 1389-1396.

Giannoni, P., Arango-Lievano, M., Neves, I.D., Rousset, M.C., Baranger, K., Rivera, S., Jeanneteau, F., Claeysen, S., Marchi, N., 2016. Cerebrovascular pathology during the progression of experimental Alzheimer's disease. Neurobiol Dis. 88, 107-17. Gujar, S.K., Maheshwari, S., Bjorkman-Burtscher, I., Sundgren, P.C., 2005. Magnetic resonance spectroscopy. J Neuroophthalmol. 25, 217-26. Hayes, K., Buist, R., Vincent, T.J., Thiessen, J.D., Zhang, Y., Zhang, H., Wang, J., Summers, A.R., Kong, J., Li, X.M., Martin, M., 2014. Comparison of manual and semi-automated segmentation methods to evaluate hippocampus volume in APP and PS1 transgenic mice obtained via in vivo magnetic resonance imaging. J Neurosci Methods. 221, 103-11. Hebert, F., Grand'maison, M., Ho, M.K., Lerch, J.P., Hamel, E., Bedell, B.J., 2013. Cortical atrophy and hypoperfusion in a transgenic mouse model of Alzheimer's disease. Neurobiol Aging. 34, 1644-52. Kam, L., Shain, W., Turner, J.N., Bizios, R., 2002. Selective adhesion of astrocytes to surfaces modified with immobilized peptides. Biomaterials. 23, 511-5. Kim, J.H., Ha, T.L., Im, G.H., Yang, J., Seo, S.W., Chung, J.J., Chae, S.Y., Lee, I.S., Lee, J.H., 2014. Magnetic resonance imaging for monitoring therapeutic response in a transgenic mouse model of Alzheimer's disease using voxel-based analysis of amyloid plaques. Neuroreport.

25, 211-8. Kobayashi, S., Tateno, M., Utsumi, K., Takahashi, A., Saitoh, M., Morii, H., Fujii, K., Teraoka, M., 2008. Quantitative analysis of brain perfusion SPECT in Alzheimer's disease using a fully automated regional cerebral blood flow quantification software, 3DSRT. J Neurol Sci. 264, 27-33. Kuhla, A., Ruhlmann, C., Lindner, T., Polei, S., Hadlich, S., Krause, B.J., Vollmar, B., Teipel, S.J., 2017. APPswe/PS1dE9 mice with cortical amyloid pathology show a reduced NAA/Cr ratio without apparent brain atrophy: A MRS and MRI study. Neuroimage Clin. 15, 581-586. Li, L., Wang, X.Y., Gao, F.B., Wang, L., Xia, R., Li, Z.X., Xing, W., Tang, B.S., Zeng, Y., Zhou, G.F., Zhou, H.Y., Liao, W.H., 2016a. Magnetic resonance T2 relaxation time at 7 Tesla associated with amyloid beta pathology and age in a double-transgenic mouse model of Alzheimer's disease. Neurosci Lett. 610, 92-7. Li, X., Bao, X., Wang, R., 2016b. Experimental models of Alzheimer's disease for deciphering the pathogenesis and therapeutic screening (Review). Int J Mol Med. 37, 271-83. Li, X., Zhu, H., Sun, X., Zuo, F., Lei, J., Wang, Z., Bao, X., Wang, R., 2016c. Human Neural Stem Cell Transplantation Rescues Cognitive Defects in APP/PS1 Model of Alzheimer's Disease by Enhancing Neuronal Connectivity and Metabolic Activity. Front Aging Neurosci. 8, 282.

Li, X.Y., Men, W.W., Zhu, H., Lei, J.F., Zuo, F.X., Wang, Z.J., Zhu, Z.H., Bao, X.J., Wang, R.Z., 2016d. Age- and Brain Region-Specific Changes of Glucose Metabolic Disorder, Learning, and Memory Dysfunction in Early Alzheimer's Disease Assessed in APP/PS1 Transgenic Mice Using 18F-FDG-PET. Int J Mol Sci. 17. Liang, S., Huang, J., Liu, W., Jin, H., Li, L., Zhang, X., Nie, B., Lin, R., Tao, J., Zhao, S., Shan, B., Chen, L., 2017. Magnetic resonance spectroscopy analysis of neurochemical changes in the atrophic hippocampus of APP/PS1 transgenic mice. Behav Brain Res. 335, 26-31. Massaad, C.A., Amin, S.K., Hu, L., Mei, Y., Klann, E., Pautler, R.G., 2010. Mitochondrial superoxide contributes to blood flow and axonal transport deficits in the Tg2576 mouse model of Alzheimer's disease. PLoS One. 5, e10561. Petcharunpaisan, S., Ramalho, J., Castillo, M., 2010. Arterial spin labeling in neuroimaging. World J Radiol. 2, 384-98. Qin, Y.Y., Li, M.W., Zhang, S., Zhang, Y., Zhao, L.Y., Lei, H., Oishi, K., Zhu, W.Z., 2013. In vivo quantitative whole-brain diffusion tensor imaging analysis of APP/PS1 transgenic mice using voxel-based and atlasbased methods. Neuroradiology. 55, 1027-1038. Shu, X., Qin, Y.Y., Zhang, S., Jiang, J.J., Zhang, Y., Zhao, L.Y., Shan, D., Zhu, W.Z., 2013. Voxel-based diffusion tensor imaging of an APP/PS1 mouse model of Alzheimer's disease. Mol Neurobiol. 48, 78-83.

Snow, W.M., Dale, R., O'Brien-Moran, Z., Buist, R., Peirson, D., Martin, M., Albensi, B.C., 2017. In Vivo Detection of Gray Matter Neuropathology in the 3xTg Mouse Model of Alzheimer's Disease with Diffusion Tensor Imaging. J Alzheimers Dis. 58, 841-853. Song, S.K., Kim, J.H., Lin, S.J., Brendza, R.P., Holtzman, D.M., 2004. Diffusion tensor imaging detects age-dependent white matter changes in a transgenic mouse model with amyloid deposition. Neurobiol Dis. 15, 640-7. Spencer, N.G., Bridges, L.R., Elderfield, K., Amir, K., Austen, B., Howe, F.A., 2013. Quantitative evaluation of MRI and histological characteristics of the 5xFAD Alzheimer mouse brain. Neuroimage. 76, 108-15. Sun, S.W., Song, S.K., Harms, M.P., Lin, S.J., Holtzman, D.M., Merchant, K.M., Kotyk, J.J., 2005. Detection of age-dependent brain injury in a mouse model of brain amyloidosis associated with Alzheimer's disease using magnetic resonance diffusion tensor imaging. Exp Neurol. 191, 77-85. Verclytte, S., Lopes, R., Lenfant, P., Rollin, A., Semah, F., Leclerc, X., Pasquier, F., Delmaire, C., 2016. Cerebral Hypoperfusion and Hypometabolism Detected by Arterial Spin Labeling MRI and FDG-PET in Early-Onset Alzheimer's Disease. J Neuroimaging. 26, 207-12. Wang, H., Liu, J., Zong, Y., Xu, Y., Deng, W., Zhu, H., Liu, Y., Ma, C., Huang, L., Zhang, L., Qin, C., 2010. miR-106b aberrantly expressed in a

double transgenic mouse model for Alzheimer's disease targets TGFbeta type II receptor. Brain Res. 1357, 166-74. Wells, J.A., O'Callaghan, J.M., Holmes, H.E., Powell, N.M., Johnson, R.A., Siow, B., Torrealdea, F., Ismail, O., Walker-Samuel, S., Golay, X., Rega, M., Richardson, S., Modat, M., Cardoso, M.J., Ourselin, S., Schwarz, A.J., Ahmed, Z., Murray, T.K., O'Neill, M.J., Collins, E.C., Colgan, N., Lythgoe, M.F., 2015. In vivo imaging of tau pathology using multi-parametric quantitative MRI. Neuroimage. 111, 369-78. Westman, E., Wahlund, L.O., Foy, C., Poppe, M., Cooper, A., Murphy, D., Spenger, C., Lovestone, S., Simmons, A., 2010. Combining MRI and MRS to distinguish between Alzheimer's disease and healthy controls. J Alzheimers Dis. 22, 171-81. Yin, J., Turner, G.H., Coons, S.W., Maalouf, M., Reiman, E.M., Shi, J., 2014. Association of amyloid burden, brain atrophy and memory deficits in aged apolipoprotein epsilon4 mice. Curr Alzheimer Res. 11, 283-90. Zerbi, V., Kleinnijenhuis, M., Fang, X., Jansen, D., Veltien, A., Van Asten, J., Timmer, N., Dederen, P.J., Kiliaan, A.J., Heerschap, A., 2013. Gray and white matter degeneration revealed by diffusion in an Alzheimer mouse model. Neurobiol Aging. 34, 1440-50. Zhang, J., 2016. How far is arterial spin labeling MRI from a clinical reality? Insights from arterial spin labeling comparative studies in Alzheimer's disease and other neurological disorders. J Magn Reson Imaging. 43,

1020-45.

Highlights

 We acquired ASL, DTI and MRS scans simultaneously from an 8months-old mouse model of Alzheimer’s disease.  After MRI examination, we performed immune-histochemical staining and ultrastructure to validate the corresponding results of MRI.  ASL revealed CBF of AD mice was lower than WT mice, especially in the brain regions of left hippocampus, left thalamus and right cortex.  MRS showed the metabolic changes of APP/PS1 mice in hippocampus were more serious than that in cortex.  The results of DTI were not as significant as ASL and MRS.