Memory-behavior correlation in pre-onset Alzheimer's reverses following beta amyloid accumulation

doi:10.21203/rs.3.rs-4913269/v1

Download PDF

Research Article

Memory-behavior correlation in pre-onset Alzheimer's reverses following beta amyloid accumulation

https://doi.org/10.21203/rs.3.rs-4913269/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Alzheimer’s disease (AD) causes cognitive decline with aging, hypothetically due to the accumulation of beta-amyloid (Aβ) plaques. Animal models are critical in the study of AD, and the 3xTg-AD mouse model is increasingly used due to its initial absence of significant physical or behavioral impairments in youth and progressive Aβ plaque development with age. This mouse model thus provides an opportunity to find early biomarkers for AD through two stages of study. However, while altered structural and functional networks occur across the whole brain in human AD, such whole-brain networks have never been studied changes in 3xTg-AD mice. Using wild-type (WT) and 3xTg-AD mice, aged 22 and 40 weeks (before and after Aβ plaque development), we measured resting state functional magnetic resonance imaging to examine functional connectivity (FC) between brain regions and diffusion tensor imaging to evaluate the structural connectivity (SC) and axonal integrity of brain white matter fiber bundles. At 22 weeks, 3xTg-AD mice unexpectedly had higher SC and FC, and there was positive correlation between behavioral performance and functional connectivity density. By 40 weeks, SC and FC was lower in AD mice (similar to human AD patients), but the behavior-functional correlation was negative. Thus, our novel methods identified a shift in 3xTg-AD mice between two abnormal states, with the latter state resembling human AD patients. Such a shift could be an early biomarker in human patients, or, if it is not present, that the 3xTg-AD mouse model only becomes relevant after this shift occurs.

Alzheimer’s disease (AD) is a type of dementia related to age, which substantially reduces the quality of life for millions of people worldwide. AD reduces gray matter volume in the neocortex ¹, reduces the functional connections of brain networks ², and causes behavioral problems, notably anxiety in early stages and memory loss in late stages ^{3, 4}. Several studies have also have shown hyperactive brain networks in the early stages of both AD patients and an AD mouse model ^{5, 6}. Understanding the integration of structural, behavioral, and functional problems in AD is critical to developing techniques to diagnose or treat the disease. At present, there are no effective therapies for advanced AD, thus diagnosing AD early can delay the progression of AD and reduce the pain of patients. However, AD is difficult to diagnose, and subclinical problems may emerge very early ^{7, 8}, making studying its progression difficult in actual patients. Therefore, animal models are critical for studying structural-functional-behavioral developments in AD, especially to follow developments across the entire brain.

The 3xTg-AD mouse model is increasingly used to study AD due to its initial absence of significant physical or behavioral impairments in youth, while progressively developing Aβ plaques as it ages ⁹. The progressive increase in Aβ is detectable in some brain regions as early as 3 months, with synaptic transmission and potentiation being measurably impaired by 6 months ¹⁰. However, how these progressive changes affect the structural and functional brain networks in these mice has not been systematically studied at a whole-brain level. Therefore, a study between two stages of networks across the whole brain is needed, especially as human AD patients have greatly altered brain networks ^{11, 12}, including the general loss of central network hubs in the brain.

Magnetic resonance imaging (MRI) can safely and non-invasively image internal structures in human patients, and functional MRI (fMRI) uses the blood-oxygen-level-dependent (BOLD) signal to generate high temporal and spatial resolution images of functional activation in the brain ¹³. In “resting state” fMRI, no overt stimuli or task that targets a particular brain region is presented, and the whole brain’s activity is thus available for investigation ^14–17. The spontaneous BOLD signal can be regarded as the marker of neuronal signal transduction ¹⁸ and thus reflects the brain’s function ¹⁵. Diffusion tensor imaging (DTI) measures the restricted random movement of water molecules in the brain to visualize the microstructure of white matter (WM) ^19–21. DTI can also be performed on ex vivo brains to allow longer imaging sessions, and thus the derived fiber tracks can be used to trace the physical link between brain regions ^{22, 23}, whereas in vivo DTI can detect the integrity of white matter ^{24, 25}. By combining fMRI and DTI, researchers can study the relationship between functional and structural connections to determine if and how structural networks constrain, maintain, and regulate functional networks, while behavior can reflect how different brain regions work together. MRI can thus reveal the relationship between structural and functional connectivity in AD, which assists with diagnosis and drug discovery.

Here, we combined multi-modal MRI (in vivo fMRI and DTI, and ex vivo DTI) and behavioral testing to elucidate the differences in brain structure and brain function between the AD and WT mice in a comparison between 22 and 40 weeks. Previous studies have determined that focus on these two age points is ideal as it allows study prior to and after the sharp increase in Aβ^{10, 26}. We used resting-state fMRI to study their whole brain function connectivity (FC) and global functional connectivity density (FCD) and used in vivo and ex vivo DTI to study the fiber integrity and brain structural connectivity (SC). We also used two behavior tests: the open field test (OFT) to study anxiety and the novel object recognition (NOR) test to study cognition and memory. Comparisons were then made between the FC and SC, FC and in vivo DTI, and FCD with behavior results in order to understand how changes to whole-brain functional and structural networks are linked. This was designed to help with both validating the utility the 3xTg AD mouse model and locating potential biomarkers of AD for testing human patients.

AD mouse model

The mice were purchased from Beijing Vitalstar Biotechnology Co., Ltd. The 3xTg-AD mice were maintained on the C57BL6/129SvJ hybrid background. The study employed female 3xTg-AD mice and age-matched control female C57BL/6N mice, aged 22 weeks and 40 weeks, with ten mice per group (AD₂₂, WT₂₂, AD₄₀, WT₄₀). Only two time points were used to maximize statistical power and apply the reduction principle ²⁷; these specific time points were chosen based on previous work that indicates that they are sufficient to examine the state prior to, and succeeding, the large increase in Aβ plaques²⁸. Mice were cage-housed under a 12:12-h light/dark cycle in a constant 25℃ and 60–70% humidity environment where mice could access food and water freely.

Behavior test

Mice were acclimated to the test room for 1 hour before trials and testing. Open field tests (OFT) were carried out when the mice were awake to test their anxiety ²⁹. Mice were put in the testing arena (70 × 70 × 70 cm) where they could move freely for 10 min. We analyzed the trace of each mouse’s movement in MATLAB R2020b (MathWorks, America). Novel object recognition (NOR) tests were carried out to test the mice's cognition and memory ³⁰. Each mouse was first habituated to the testing arena (70 × 70 × 70cm) for 10 min on two consecutive days. Next, they were allowed to explore two identical objects for 10 min. After 3 hours, they were put in the area again, with a novel object with a different shape and material in place of one of the previous objects, and they were allowed to explore the two differing objects for another 10 min. The time that the mice explored the novel object was recorded ³¹.

MRI acquisition

The mice were anesthetized using intraperitoneal bolus injections of 25% urethane (Sigma-Aldrich, U2500) dissolved in saline, administered in two separate doses totaling 7 µl/g ^32–34. The breath rate of mice under urethane was approximately 150–240, similar to previous studies ³⁴, however it was significantly higher in WT₂₂ than WT₄₀ (p = 0.0021, two-sample t-test) but there was no significant difference between AD₂₂ and AD₄₀ mice, AD₂₂ and WT₂₂ mice, nor AD₄₀ and WT₄₀ mice.

Functional and diffusion tensor images were acquired on the 9.4 Tesla MRI scanner (BioSpec 94/30 USR; Bruker Biospin MRI, Ettilingen, Germany), with a 4-element surface coil and an 86-mm volume transmit coil. T2 Turbo RARE (Rapid Acquisition with Relaxation Enhancement) sequence was used for 2D T2-weighted brain anatomical imaging. The acquisition parameters were as follows: field of view (FOV) = 24.2×9.6 mm², echo time/repetition time (TE/TR) = 35/2003.56 ms, number of slices = 23, slice thickness = 0.3 mm, resolution = 0.1× 0.1 mm², matrix size = 242×96, total scan time = 2 min 36 sec.

Echo-planar imaging (EPI) sequence was used to acquire T2*-weighted free induction decay (FID) images for 2D functional imaging. The acquisition parameters were as follows: FOV = 24.2×9.6 mm², TE/TR = 13/1000 ms, matrix size = 121×48, number of slices = 13, slice thickness = 0.6 mm, resolution = 0.2 × 0.2 mm², repetition = 480, total scan time = 8 min.

To characterize the microstructure of axon tracts, in vivo Diffusion Tensor Imaging (DTI) was performed using spin-echo EPI. The acquisition parameters were as follows: FOV = 16×8 mm², TE/TR = 19/2500 ms, matrix size = 80×40, number of slices = 60, slice thickness = 0.20 mm without gap, voxel size = 0.20×0.20×0.20 mm³, b-value = 1000 s/mm², diffusion gradient duration/separation time (∆/δ) = 3.5/10.0 ms, with a total of 32 diffusion directions and five b = 0 s/mm² images acquired. No. of segments = 4, averages = 4, total scan time = 24 min 40 sec.

For tractography analysis, ex vivo DTI was acquired using spin-echo echo-planar imaging (EPI) with the following parameters: a 3D acquisition, FOV = 12×7.5×16 mm2, TE/TR = 32.5/250 ms, matrix = 120×75×160, voxel size = 0.10×0.10×0.10 mm3, b = 3000 s/mm², ∆/δ = 3.0/21.0 ms, 60 diffusion directions, 5 b = 0 s/mm² images, No. of segments = 4, average = 4, total scan time = 11 h 33 min 20 sec. Due to animal death (see “Preparation of ex vivo samples” section), limitations in scanner availability, and the extremely long scan time, only 5 mice per group for AD₂₂, WT₂₂, and AD_40, and 6 mice for WT₄₀ were scanned using this protocol.

Preparation of ex vivo samples

Following the MR scan, mice were intracardially perfused with 5 mM Gadopentetic acid (Gd-DTPA) (Adamas 86050-77-3). The brain sample was stored in a 15 ml tube with 4% PFA solution at 4℃ overnight for fixation. After this, the sample was moved to a 15 ml tube with 5mM Gd-DTPA saline and stored at 4℃ for at least 2 weeks for rehydration. On the day of ex vivo imaging, the brain samples were transferred into a 5 ml tube filled with Galden® perfluoropolyether (Solvay Specialty Polymers). Air bubbles were removed by vacuum before scanning. Due to the long scan time under urethane anesthesia and stress of perfusion setup, 2 AD₂₂, 2 WT₂₂, 2 AD₄₀, and 4 WT₄₀ mice did not survive long enough to be able to successfully perfuse and thus were unable to be used for ex vivo DTI.

fMRI processing

The fMRI and DTI data processing are shown in “Supplementary methods.” Our methods comparing DTI, fMRI, and behavior are in part based on a previous study by Kaneko and colleagues ³⁵

For resting-state fMRI (rs-fMRI) analysis, functional images were pre-processed by brain extraction, slice-timing correction, and motion correction in MATLAB R2020b (https://uk.mathworks.com/products/matlab) with SPM12 (http://www.fil.ion.ucl.ac.uk/spm/software/spm12/) to adjust as if all slices were acquired simultaneously, eliminating motion artifacts. Functional images were then aligned with anatomical images within the same subject by rigid transformation and then anatomical images were co-registered to the Allen mouse brain atlas ³⁶ in MATLAB R2020b and BioImage Suite Web (https://bioimagesuiteweb.github.io/webapp/). All images were smoothed with FWHM (full-width-half-maximum) of 5 to allow for better comparison between mice and to facilitate the group-based, ROI (region of interest)-based analysis in this study.

Group independent component analysis (group ICA) took place in MATLAB R2020b using Group ICA of the fMRI Toolbox (GIFT) (https://trendscenter.org/software/gift/), and the components related to cerebral spinal fluid (CSF) noise were removed ³⁷.

Preprocessed data were divided into 5 epochs, with each epoch including 96 repetitions. We removed epochs which had a Pearson’s correlation coefficient (pcc) of < 0.1 between the left and right brain regions of lower forelimb primary somatosensory cortex (S1FL). This was undertaken to improve signal quality, as the S1FL region normally has high correlation if the mouse is in a good physiological state ^{32, 33}. This resulted in the complete removal of 3 AD₂₂, 2 WT₂₂, 1 AD₄₀, and 2 WT₄₀ mice. (Note that behavioral data from these mice were not removed for calculation of average behavioral parameters.)

Regions of interest (ROI) were created for 25 regions using a labeled atlas of the template brain ³⁸. Abbreviations are used as per the Allen atlas. Functional connectivity matrices were calculated using the pcc of the mean BOLD signals between different brain regions. Student’s t-tests were calculated using MATLAB and resulting p values were corrected for multiple comparisons by sequential goodness of fit metatest (SgoF) for family-wise error rate (FWER) of 0.05 ³⁹.

Global functional connectivity density (gFCD) was calculated by creating a correlation matrix for all voxels in the brain, then counting each voxel’s number of connections with other voxels as the number of voxels it correlated with above a threshold of 0.1 as per ⁴⁰. Thus, gFCD reflected the level at which each voxel in the brain was a functional connectivity hub. In addition, gFCD was used to perform outlier removal. Mice whose data included greater than 3 standard deviations from whole brain mean gFCD were removed from behavior-gFCD analysis, this was only one AD₂₂ mouse. (Thus, the total number removed including prior removals was 4 AD₂₂ 2 WT₂₂, 1 AD₄₀, and 2 WT₄₀ mice.)

DTI processing

Pre-processing: both in vivo and ex vivo DTI datasets were initially converted into 4D NIFTI images using DSI studio (2022, Jan, 3) as described by Yeh ⁴¹. The dataset underwent denoising using nlsam (https://nlsam.readthedocs.io/en/latest/autoapi/nlsam/index.html). Summed DWI was utilized for brain extraction mask delineation in BrainSuite (Version 21a) (https://brainsuite.org/). Motion and Eddy current correction procedures were conducted.

The in vivo DTI metrics, including Fractional Anisotropy (FA) as an indicator of axon integrity; Mean Diffusivity (MD) and Axial Diffusivity (AD) as an indicator of axon injury; and Radial Diffusivity (RD) as an indicator of myelination, were computed in DSI studio (Version 2022 Jul) using the denoised DTI dataset. Q-Space Diffeomorphic Reconstruction (QSDR) was performed based on the Allen Mouse Brain Common Coordinate Framework version 3 (ABA). Subsequently, the average parametric values for each 3D brain region were extracted using MATLAB scripts (Mathworks, R2021b). Prior to further analysis, mice were excluded if visual inspection of their in vivo DTI images could not clearly differentiate white matter from gray matter, which resulted in 3 AD₂₂, 3 WT₂₂, 1 AD₄₀, and 3 WT₄₀ mice being removed from in vivo DTI results.

The ex vivo DTI connectome analysis employed QSDR reconstruction with the ABA mouse brain template. The tracking threshold was set at qa = 0.01, angular threshold at 45°, and minimum/maximum length at 1.0/40.0mm, with terminal seeds of 1,000,000. Connectivity analysis utilized 15 ROIs derived from rsfMRI analysis. For each mouse brain, a connectivity matrix was calculated using these 15 brain regions. Group comparison was conducted through a two-sample t-test across ages and genotypes. To visualize the connectogram displaying significant group differences, the averaged connectivity matrix with significant differences was plotted on the website of circos (http://mkweb.bcgsc.ca/tableviewer/visualize/).

Statistical analysis

Statistical analysis regarding behavior and also regarding the relationship between FC and both ex vivo and in vivo DTI were calculated in GraphPad Prism (version 8, GraphPad Software, San Diego, CA, USA). For behavior, unpaired two-tailed Student’s t-tests were used to compare the significance between different groups. The relationship between FC and both ex vivo and in vivo DTI was calculated by Pearson correlation. (Only mice whose data were neither excluded for fMRI nor ex vivo DTI were used for the FC versus in vivo comparison, and only mice whose data were not excluded for fMRI and who survived long enough to perform brain perfusion were used for the FC versus ex vivo comparison.) For each group, the mean slope from all linear regressions within the group was calculated. Unpaired two-tailed Student’s t-tests were used to compare groups with the relationship between FC and in vivo DTI. No multiple comparison corrections were performed for the behavior test or for the relationship between FC and in vivo DTI. (Note that when matching data from the mice in the correlation between FC and in vivo DTI, 5 AD₂₂, 5 WT₂₂, 2 AD₄₀, and 3 WT₄₀ mice had data removed due to a combination of exclusion criteria described previously.)

In vivo rs-fMRI functional connectivity

For brain function, a comparison of the FC maps between AD and WT mice at 22 and 40 weeks, revealed a number of important observations. For all tested regions, in AD₂₂ and AD₄₀ mice, the gustatory areas (GU) showed a negative correlation with other regions, but in WT₂₂ and WT₄₀ mice, the GU showed a positive correlation with all tested regions (Fig. 1a-d). Next, we compared the changes of FC between different groups. First, we found the FC of the FRP, MO, and GU (see brain region abbreviations in Table S1) with other regions were significantly lower in AD₂₂ mice compared to WT₂₂ mice, while the FC of the other cortical regions and other subcortical regions with other brain regions were significantly higher in AD₂₂ mice (Fig. 1e). The FC between most pairs of brain regions were significantly lower in AD₄₀ mice than in WT₄₀ mice (Fig. 1f). While FC of the MO and GU with other regions was found to be lower in AD₂₂ mice than AD₄₀ mice, the FC of the ILA, ORB and OLA with other regions were significantly higher in AD₂₂ mice (Fig. 1g). In WT mice, the FC was significantly lower at 22 weeks (Fig. 1h). These results revealed that in normal aging, FC increased from 22 to 40 weeks, but with AD this increase only existed in MO and GU regions. In addition, compared with WT mice, the FC was higher in AD at the younger age.

Ex vivo DTI structural connectivity

Brain regions were chosen for ex vivo DTI structural connectivity based on their significance in the in vivo fMRI functional connectivity results. Thus, the brain regions chosen were the AI, ECT, FRP, GU, ILA, ORB, PERI, TEa, and VISC cortical regions and the AMY (amygdala), HF, HY, OLF, PAL and TH subcortical regions. We calculated the structural connectivity (SC) between all of these regions and compared the results. We found that the SC between ORB with HY and PAL in AD₂₂ mice were significantly lower compared to WT₂₂ mice, while the SC between FRP, GU, TEa, AI, and ILA were significantly higher (Fig. 2a). The SC between cortical regions of AD₄₀ mice were significantly lower than in WT₄₀ mice, whereas the SC between subcortical regions were significantly higher (Fig. 2b). The SC was found to be significantly lower in AD₂₂ than AD₄₀ mice while SC between GU and AMY was only detectable in AD₂₂ mice, not AD₄₀ mice (Fig. 2c). In WT mice, the SC of the TH, HY, and PAL between other regions were significantly higher for WT₂₂, while the SC of the cortical regions with other regions were significantly lower (Fig. 2d). From these results, we concluded that the SC increases in AD mice, but otherwise the SC between cortical regions increases with age and SC between subcortical regions decreases with age in normal mice.

Relationship between FC and SC

We compared the relationship between in vivo rs-fMRI-based FC and ex vivo DTI-based SC using nine brain regions for which they showed connectivity significance in both FC and SC. We found that FC increased as SC increased in all nine regions (Fig. 3)

The correlation in FRP was weak in AD₂₂ mice, but higher in AD₄₀ mice. The FC-SC correlation in GU was weak in WT₄₀ mice. The FC-SC correlation in PAL was weak in AD mice as compared to WT mice at 22 weeks and 40 weeks. The correlation in TH was weak at 22 weeks versus 40 weeks. The correlations in HY were weak in AD₂₂ and WT₄₀ mice (Table S2). Overall, these results showed correlated FC and SC within cortical regions, and also that higher SC supported higher FC.

Relationship between functional connectivity and microstructure

We used in vivo DTI to measure the microstructure of the brain and to study the relationship between FC and microstructure of brain parenchyma. Fractional anisotropy (FA) as the parameter of in vivo DTI was used to measure the integrity of white matter. The FRP, GU, ILA, and AI cortical regions were chosen to study the relationship between FC and the microstructure of brain parenchyma because these regions showed both FC and SC significance in prior analyses (Fig. 1e, f, 2a-d). The GU region showed a positive correlation between FC and FA in both AD₂₂ and WT₂₂ mice (Fig. 4a, b), but a negative correlation at 40 weeks (Fig. 4c, d). The FRP, ILA, and AI regions showed a negative correlation (Fig. 4, Table S3) where an FC increase accompanied an FA decrease.

Mouse behavior and brain function

We used the novel object recognition (NOR) test to study mouse memory and the relationship between memory and brain function. Here, the greater the proportion of time a mouse explored the novel object, the better their memory. Using this approach, we found that AD₂₂ mice had significantly worse memory during the task compared to AD₄₀ mice (p = 0.003). Comparing AD and WT mice at both time periods, AD mice showed significantly worse memory than WT mice (p < 0.0001, p = 0.0180) (Fig. 5a). AD₂₂ and WT₄₀ mice groups showed a positive correlation between NOR and gFCD, while the WT₂₂ and AD₄₀ mice had a negative correlation (Fig. 5b-i). These results revealed a correlation between better memory and more brain connectivity in the AD₂₂ and WT₄₀ mice, whereas in AD₄₀ mice, better memory was correlated with less brain connectivity.

We used the open field test (OFT) to study the anxiety level of mice and the relationship between anxiety and brain function. Here, the longer the proportion of time the mouse stayed in the center of the arena, the less anxiety the mouse showed. Using this method, we found that AD₂₂ mice had significantly more anxiety during the task than AD₄₀ mice (p = 0.0046) (Fig. 6a). The AD₂₂ and WT₄₀ mice showed a negative correlation between OFT and gFCD, while WT₂₂ and AD₄₀ mice had a positive correlation between OFT and gFCD (Fig. 6b-i). Our results reveal that more anxiety is correlated with more brain connectivity in AD₂₂ and WT₄₀ mice, whereas in AD₄₀ mice, more anxiety correlated with less brain connectivity.

We used the open field test (OFT) to study the anxiety level of mice and the relationship between anxiety and brain function. Here, the longer the proportion of time the mouse stayed in the center of the arena, the less anxiety the mouse showed. Using this method, we found that AD₂₂ mice had significantly more anxiety during the task than AD₄₀ mice (p = 0.0046) (Fig. 6a). The AD₂₂ and WT₄₀ mice showed a negative correlation between OFT and gFCD, while WT₂₂ and AD₄₀ mice had a positive correlation between NOR and gFCD (Fig. 6b-i). Our results reveal that more anxiety is correlated with more brain connectivity in AD₂₂ and WT₄₀ mice, whereas in AD₄₀ mice, more anxiety correlated with less brain connectivity.

We found that the 3xTg-AD mouse model already differed from WT by 22 weeks in some ways that were the opposite of what was expected; however, these differences inverted to more expected results by 40 weeks. Key features of this inversion were the overall strength of FC and SC, which was generally higher in AD₄₀ mice and lower in AD₂₂ mice, and also behavior-functional correlation, which was similar in AD₂₂ and WT₄₀ mice and similar in AD₄₀ and WT₂₂ mice. Comparisons between AD₄₀ and WT₄₀ better resembled comparisons between human AD patients and age-matched healthy controls, notably the lower FC in AD₄₀. Thus, rather than seeing the emergence of an abnormal state at 40 weeks, we instead observed a shift in AD mice between two abnormal states.

The deficit-compensation hypothesis

Our results are consistent with the hypothesis that the increase in FC of subcortical regions is caused by a compensatory effect at the early stage due to the slight accumulation of Aβ plaques and tau protein aggregation ^42–44. As this injury would happen to synapses in the neocortical region, the FC between it and other regions in AD₂₂ mice was lower than in WT₂₂ mice (Fig. 1e). Later on, however, significant differences of FC between brain regions in AD₄₀ and WT₄₀ mice became pronounced and FC were all lower in AD₄₀ (Fig. 1f) suggesting that AD mice were unable to fully compensate for the injury. From 22 to 40 weeks in both groups, the whole brain FC increased as age increased (Fig. 1) due to normal development ⁴⁵. Severe injury in AD₄₀ mice due to amyloid and tau deposition decreased the FC more than any compensatory increases from normal development, resulting in no significant difference between the different ages of AD mice in FC (Fig. 1g). We further hypothesize that there is a compensatory effect in the subcortex that results in the cortex being more vulnerable than the subcortical regions. AD₂₂ mice had lower SC in subcortical regions but higher SC in cortical regions than WT₂₂ mice, while conversely AD₄₀ mice had higher SC in subcortical regions and lower SC in cortical regions than WT40 mice (Fig. 2a, b). From these results, we concluded that the SC increases as the FC declines due to the effects of Aβ-based injury on brain function.

In terms of SC, AD₂₂ mice showed an increase in cortical regions while AD₄₀ mice had increased SC in subcortical regions. One possible explanation for this observation is that selective axon loss led to an increase in FA in the cross-fiber region, which tended towards an overestimation of SC ^{43, 46}. Another possibility is that, at an early stage, the brain developed myelin and affected the reorganization of higher structure connections in the cortex of AD mice, and, as age increased, the myelin in the cortex became severely injured, reducing SC. However, in normal mice, in the absence of injury, they developed higher SC. Conversely, in subcortical regions, AD₄₀ mice developed higher SC than WT₄₀ mice likely because in the former, there exist both compensatory effects and the effects of damage, and at that time, compensatory effects were higher, while in the latter there was more effects from damage ^{47, 48}. AD₄₀ mice also had higher SC than AD₂₂ mice (Fig. 2), likely due to a combination of WM atrophy, brain volume reduction, and other geometric variability. In some cases, this can result in multiple ROIs being partially merged together into a smaller space, which causes a tendency towards elevated SC measurements ^{43, 49}.

The relationship between FC and SC

Our results demonstrated that FC has a positive correlation with SC. Previous studies have also shown that SC increases are usually correlated with FC increases in human subjects ^50–52, and our results support their conclusions. The weak correlation between FC and SC may indicate that there exists indirect SC of these two region, thus the brain functional connectivity is not generated via a direct structural connection, and the regions with greater correlation between FC and SC appear to be more severely injured in disease ^{53, 54}. Our results showed FA had a negative correlation with FC in the FRP, AI, and ILA regions, and had a positive correlation with FC in the GU region at 22 weeks. The higher FA brain regions had higher FC. Previous studies found that FA had a positive correlation with FC during the resting state, but other studies have found that the relationship between FA and FC was not significant ⁵⁵. In the mouse model of Alzheimer’s disease, the brain is known to develop tau protein aggregation and Aβ plaques. This causes neuroinflammation by microglia, and such neuroinflammation will cause synapse loss ^{48, 56}. FA may increase due to compensation; however, the FC may not increase because the connection pathways have been damaged ^{57, 58}.

Relationship between behavior and brain function

Our behavioral tests results showed that AD mice had worse memory than WT mice (Fig. 5a), and that AD₂₂ mice had more anxiety than both WT₂₂ and AD₄₀ mice (Fig. 6a). While the AD₂₂ and WT₄₀ mice showed a positive correlation between memory and gFCD, they showed a negative correlation between anxiety and gFCD. Conversely, the AD₄₀ mice had a negative correlation between memory and gFCD, and a positive correlation between anxiety and gFCD. These results demonstrated that AD₂₂ mice and WT₄₀ mice had the same behavior-functional modal, which was influenced by age. This modal change also exists in healthy humans, according to King et. al, they found the functional connectivity–memory relationship changes during aging. ⁵⁹. Based on our results regarding the relationship between gFCD and behavior, we hypothesize that the Aβ plaques and tau protein accumulation is changing the function of these brain regions, which are involved in memory, cognition and anxiety ⁶⁰.

In humans, FCD correlates with brain glucose use spatially across different brain regions ⁶¹, and with changing states from eyes open to eyes closed ⁶². In mice, a decrease in brain oxygen use from awake to under anesthesia ⁶³ correlated with a decrease in FCD in the same mice ⁶⁴. In our study, gFCD in WT₂₂ mice was lower on average than in WT₄₀ mice (Fig. 5–6). Thus, a higher cerebral metabolic rate of glucose would be expected at 40 weeks versus 22 weeks, which is indeed what two previous studies of glucose uptake in similar mouse models reported ^{65, 66}. However, this does not hold for the AD mouse models: in those studies, was no difference at 26–30 weeks, but WT glucose uptake was higher than AD glucose uptake at 39–40 weeks. This discrepancy indicates that the FCD-metabolic relationship is disrupted in AD mice, and we suggest that future studies could combine metabolic and functional measurements.

Potential early biomarkers for human AD

The progression of AD can be delayed if diagnosed early ⁶⁷. Many interventional clinical trials are currently shifting their focus to cognitively healthy individuals at risk of developing AD as the best strategy to reduce the incidence and prevalence of AD. Non-invasive neuroimaging plays an important role in diagnosing AD ^{68, 69}, thus our results, collected prior to and succeeding the large increase in Aβ plaques in the 3xTg-AD mouse model ¹⁰, provide several potential early biomarkers which merit study in human populations at risk of developing AD.

We found that the FC of the cortical regions especially GU of AD₂₂ (~ 5 months old) mice had already decreased and that SC had already increased, suggesting that these early correlated changes may be an important biomarker of disease. At 40 weeks (~ 9 months) the AD mice showed decreased whole-brain FC and increased SC as compared with WT but showed no significant changes in FC as compared with the AD₂₂ mice. When comparing AD₂₂ to AD₄₀ mice, the behavior-functional modal changes from positive to negative in memory tests and from negative to positive in anxiety tests. Behavioral testing in human AD patients, compared against both young and age-matched healthy controls, can determine if this change is a biomarker that can assess the progression of AD in humans.

Previous studies have used the AD mouse model as young as 17 weeks, and most studies have used from 26 to 35 weeks and older ⁷⁰. Flurkey and Currer suggested that a 22 week-old mouse is equivalent to a 29 to 30-year old human, and a 40 week-old mouse is equivalent to a 37 to 38-year old human ⁷¹. Conversely, studies of behavior in wild mice have suggested that ‘vigor’ begins decreasing as early as 13 weeks old ⁷² with 50% mortality already reached at this point ⁷³. Compare this to the 3xTg-AD mouse model that already has detectable Aβ at 13 weeks according to the Jackson Laboratory and Belfiore, et al. ^{10, 74}. Therefore, changes in this mouse model likely occur at older equivalent ages in human patients than suggested by Dutta and Sengupta, and we suggest that the observations at 22 weeks may be homologous to AD risk and at 40 weeks may be homologous to active AD.

Previous studies showed the gustatory cortex is involved with gustatory function like food choice and intake ^{75, 76}, and taste preferences ⁷⁷. The decline in FC of the GU region suggests that the AD mice have loss of gustatory function. The relationship between FC and SC in GU was a positive linear relationship in 22 week-old mice, but a negative linear relationship in 40 week-old mice. The WT₄₀ mice had a higher negative slope than AD₄₀ mice while the WT₂₂ mice showed a lower positive slope than AD₂₂ mice (Fig. 4, Table S3). Thus, the gustatory loss is also worth investigating in human AD.

In conclusion, through examining the changes to the structural-functional and functional-behavioral relationships between two ages in 3xTg-AD mice, we have demonstrated that, rather than degrading from a normal state to an abnormal state, they shift from one abnormal state to another, with the latter state better corresponding to human patients with AD. Our results reveal a shifting relationship between FC and SC, and FC and microstructure, providing a new perspective on the relationship between function, structure, and behavior in the brain affected by AD. The shifts we have observed in AD mice show important potential as an early biomarker to test in human patients at risk of AD. Alternately, if these shifts are not present in human AD, it should be considered that the 3xTg-AD mouse model only becomes relevant to the study of human AD after the shift has completed.

Funding sources:

Hui Li has received funding from the National Natural Science Foundation of China Grant (No. 32100555). Garth J. Thompson has received funding from ShanghaiTech University, the Shanghai Municipal Government, and the National Natural Science Foundation of China Grant (No. 81950410637). Qiong Ye has received funding from Collaborative Key Foundation of Hefei Science Center Grant (No. 2022HSC-CIP003).

Author Contribution

Ziyi Wang has made contributions to the design and draft of the work, acquisition, analysis and interpretation of the data and draft the work.Hui Li has contributed to the design of the work and the acquisition and analysis of the data.Bowen Shi has made contributions to the acquisition of the data.Qikai Qin has made contributions to the acquisition of the data.Qiong Ye has made contributions to the design the work, acquisition, analysis and interpretation of the data and draft the work. Garth J. Thompson has made contributions to the design and conception of the work, substantively revised the draft and to have approved the submitted version and to have agreed both to be personally accountable for the author's own contributions and to ensure that questions related to the accuracy or integrity of any part of the work.

Acknowledgments:

Dr. Qikai Qin is currently affiliated with Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, and Department of Radiology, Massachusetts General Hospital and Harvard Medical School.

Phillips, J.S., F. Da Re, D.J. Irwin, et al., Longitudinal progression of grey matter atrophy in non-amnestic Alzheimer's disease. Brain, 2019. 142(6):1701-1722.
Mondadori, C.R., A. Buchmann, H. Mustovic, et al., Enhanced brain activity may precede the diagnosis of Alzheimer's disease by 30 years. Brain, 2006. 129(Pt 11):2908-2922.
Donovan, N.J., J.J. Locascio, G.A. Marshall, et al., Longitudinal Association of Amyloid Beta and Anxious-Depressive Symptoms in Cognitively Normal Older Adults. American Journal of Psychiatry, 2018. 175(6):530-537.
Ferretti, L., S.M. McCurry, R. Logsdon, et al., Anxiety and Alzheimer's disease. J Geriatr Psychiatry Neurol, 2001. 14(1):52-58.
Pistono, A., M. Senoussi, L. Guerrier, et al., Language Network Connectivity Increases in Early Alzheimer's Disease. J Alzheimers Dis, 2021. 82(1):447-460.
Shah, D., W. Gsell, J. Wahis, et al., Astrocyte calcium dysfunction causes early network hyperactivity in Alzheimer's disease. Cell Rep, 2022. 40(8):111280.
Sabbagh, M.N., M. Boada, S. Borson, et al., Early Detection of Mild Cognitive Impairment (MCI) in Primary Care. J Prev Alzheimers Dis, 2020. 7(3):165-170.
Porsteinsson, A.P., R.S. Isaacson, S. Knox, et al., Diagnosis of Early Alzheimer's Disease: Clinical Practice in 2021. J Prev Alzheimers Dis, 2021. 8(3):371-386.
Oddo, S., A. Caccamo, J.D. Shepherd, et al., Triple-transgenic model of Alzheimer's disease with plaques and tangles:: Intracellular Aβ and synaptic dysfunction. Neuron, 2003. 39(3):409-421.
Belfiore, R., A. Rodin, E. Ferreira, et al., Temporal and regional progression of Alzheimer's disease-like pathology in 3xTg-AD mice. Aging Cell, 2019. 18(1):e12873.
Greicius, M.D., G. Srivastava, A.L. Reiss, et al., Default-mode network activity distinguishes Alzheimer's disease from healthy aging: Evidence from functional MRI. Proceedings of the National Academy of Sciences, 2004. 101(13):4637-4642.
Rombouts, S.A.R.B., F. Barkhof, R. Goekoop, et al., Altered resting state networks in mild cognitive impairment and mild Alzheimer's disease: An fMRI study. Human Brain Mapping, 2005. 26(4):231-239.
Ogawa, S., T.M. Lee, A.R. Kay, et al., Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proceedings of the National Academy of Sciences, 1990. 87(24):9868-9872.
van den Heuvel, M.P. and H.E. Hulshoff Pol, Exploring the brain network: a review on resting-state fMRI functional connectivity. Eur Neuropsychopharmacol, 2010. 20(8):519-534.
Biswal, B., F. Zerrin Yetkin, V.M. Haughton, et al., Functional connectivity in the motor cortex of resting human brain using echo-planar mri. Magnetic Resonance in Medicine, 1995. 34(4):537-541.
Smith, S.M., P.T. Fox, K.L. Miller, et al., Correspondence of the brain's functional architecture during activation and rest. Proceedings of the National Academy of Sciences, 2009. 106(31):13040-13045.
Lee, M.H., C.D. Smyser, and J.S. Shimony, Resting-State fMRI: A Review of Methods and Clinical Applications. American Journal of Neuroradiology, 2013. 34(10):1866-1872.
Fox, M.D. and M.E. Raichle, Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging. Nature Reviews Neuroscience, 2007. 8(9):700-711.
Zhang, J., M. Aggarwal, and S. Mori, Structural insights into the rodent CNS via diffusion tensor imaging. Trends in Neurosciences, 2012. 35(7):412-421.
Assaf, Y. and O. Pasternak, Diffusion Tensor Imaging (DTI)-based White Matter Mapping in Brain Research: A Review. Journal of Molecular Neuroscience, 2008. 34(1):51-61.
Filley, C.M. and R.D. Fields, White matter and cognition: making the connection. Journal of Neurophysiology, 2016. 116(5):2093-2104.
Zhu, D., T. Zhang, X. Jiang, et al., Fusing DTI and fMRI data: a survey of methods and applications. Neuroimage, 2014. 102 Pt 1:184-191.
Dutta, S. and P. Sengupta, Men and mice: Relating their ages. Life Sciences, 2016. 152:244-248.
Sbardella, E., F. Tona, N. Petsas, et al., DTI Measurements in Multiple Sclerosis: Evaluation of Brain Damage and Clinical Implications. Multiple Sclerosis International, 2013. 2013:1-11.
Mascalchi, M., N. Toschi, M. Giannelli, et al., Progression of Microstructural Damage in Spinocerebellar Ataxia Type 2: A Longitudinal DTI Study. American Journal of Neuroradiology, 2015. 36(6):1096-1101.
Duggan, M.R., Z. Steinberg, T. Peterson, et al., Cognitive trajectories in longitudinally trained 3xTg-AD mice. Physiol Behav, 2024. 275:114435.
Hubrecht, R.C. and E. Carter, The 3Rs and Humane Experimental Technique: Implementing Change. Animals (Basel), 2019. 9(10).
Roda, A.R., G. Esquerda-Canals, J. Martí-Clúa, et al., Cognitive Impairment in the 3xTg-AD Mouse Model of Alzheimer's Disease is Affected by Aβ-ImmunoTherapy and Cognitive Stimulation. Pharmaceutics, 2020. 12(10).
Kraeuter, A.-K., P.C. Guest, and Z. Sarnyai, The Open Field Test for Measuring Locomotor Activity and Anxiety-Like Behavior. 2019, Springer New York. p. 99-103.
Whittaker, D.S., L. Akhmetova, D. Carlin, et al., Circadian modulation by time-restricted feeding rescues brain pathology and improves memory in mouse models of Alzheimer's disease. Cell Metabolism, 2023. 35(10):1704-1721.
Lueptow, L.M., Novel Object Recognition Test for the Investigation of Learning and Memory in Mice. Journal of Visualized Experiments, 2017(126):55718.
Johnson, F.K., J.-C. Delpech, G.J. Thompson, et al., Amygdala hyper-connectivity in a mouse model of unpredictable early life stress. Translational Psychiatry, 2018. 8(1):49.
De Arce, K.P., A. Ribic, D. Chowdhury, et al., Concerted roles of LRRTM1 and SynCAM 1 in organizing prefrontal cortex synapses and cognitive functions. Nature Communications, 2023. 14(1):459.
Li, H., Q. Ye, D. Wang, et al., Resting State Brain Networks under Inverse Agonist versus Complete Knockout of the Cannabinoid Receptor 1. ACS Chem Neurosci, 2024.
Kaneko, G., B.G. Sanganahalli, S.M. Groman, et al., Hypofrontality and Posterior Hyperactivity in Early Schizophrenia: Imaging and Behavior in a Preclinical Model. Biol Psychiatry, 2017. 81(6):503-513.
Lein, E.S., M.J. Hawrylycz, N. Ao, et al., Genome-wide atlas of gene expression in the adult mouse brain. Nature, 2007. 445(7124):168-176.
Han, Z., W.J. Chen, X.F. Chen, et al., Awake and behaving mouse fMRI during Go/No-Go task. Neuroimage, 2019. 188:733-742.
Xu, M.Y., B.S. Bo, M.C. Pei, et al., High-resolution relaxometry-based calibrated fMRI in murine brain: Metabolic differences between awake and anesthetized states. Journal of Cerebral Blood Flow and Metabolism, 2022. 42(5):811-825.
Carvajal-Rodríguez, A., J. De Uña-Alvarez, and E. Rolán-Alvarez, A new multitest correction (SGoF) that increases its statistical power when increasing the number of tests. BMC Bioinformatics, 2009. 10(1):209.
Tomasi, D. and N.D. Volkow, Functional connectivity density mapping. Proc Natl Acad Sci U S A, 2010. 107(21):9885-9890.
Yeh, F.C., I.M. Zaydan, V.R. Suski, et al., Differential tractography as a track-based biomarker for neuronal injury. Neuroimage, 2019. 202:116131.
Gaubert, S., F. Raimondo, M. Houot, et al., EEG evidence of compensatory mechanisms in preclinical Alzheimer's disease. Brain, 2019. 142:2096-2112.
Aganj, I., A. Frau-Pascual, J.E. Iglesias, et al. Compensatory Brain Connection Discovery in Alzheimer's Disease. IEEE.
Mormino, E.C., A. Smiljic, A.O. Hayenga, et al., Relationships between Beta-Amyloid and Functional Connectivity in Different Components of the Default Mode Network in Aging. Cerebral Cortex, 2011. 21(10):2399-2407.
Chen, Y., X. Zhao, X. Zhang, et al., Age-related early/late variations of functional connectivity across the human lifespan. Neuroradiology, 2018. 60(4):403-412.
Ryan, N.S., S. Keihaninejad, T.J. Shakespeare, et al., Magnetic resonance imaging evidence for presymptomatic change in thalamus and caudate in familial Alzheimer’s disease. Brain, 2013. 136(5):1399-1414.
Baik, S.H., S. Kang, W. Lee, et al., A Breakdown in Metabolic Reprogramming Causes Microglia Dysfunction in Alzheimer's Disease. Cell Metabolism, 2019. 30(3):493-507.
Rajendran, L. and R.C. Paolicelli, Microglia-Mediated Synapse Loss in Alzheimer's Disease. Journal of Neuroscience, 2018. 38(12):2911-2919.
Salat, D., D. Greve, J. Pacheco, et al., Regional white matter volume differences in nondemented aging and Alzheimer's disease. NeuroImage, 2009. 44(4):1247-1258.
Honey, C.J., O. Sporns, L. Cammoun, et al., Predicting human resting-state functional connectivity from structural connectivity. Proceedings of the National Academy of Sciences of the United States of America, 2009. 106(6):2035-2040.
Koch, M.A., D.G. Norris, and M. Hund-Georgiadis, An Investigation of Functional and Anatomical Connectivity Using Magnetic Resonance Imaging. NeuroImage, 2002. 16(1):241-250.
Li, K., L. Guo, D. Zhu, et al., Individual Functional ROI Optimization Via Maximization of Group-Wise Consistency of Structural and Functional Profiles. Neuroinformatics, 2012. 10(3):225-242.
Deco, G., A.R. McIntosh, K. Shen, et al., Identification of Optimal Structural Connectivity Using Functional Connectivity and Neural Modeling. Journal of Neuroscience, 2014. 34(23):7910-7916.
Straathof, M., M.R.T. Sinke, T.J.M. Roelofs, et al., Distinct structure-function relationships across cortical regions and connectivity scales in the rat brain. Sci Rep, 2020. 10(1):56.
Liu, J.K., L. Yao, W.J. Zhang, et al., Dissociation of fractional anisotropy and resting-state functional connectivity alterations in antipsychotic-naive first-episode schizophrenia. Schizophrenia Research, 2019. 204:230-237.
Nakamura, T., C.K. Oh, L.J. Liao, et al., Noncanonical transnitrosylation network contributes to synapse loss in Alzheimer's disease. Science, 2021. 371(6526):253.
Cho, D.H., T. Nakamura, J.G. Fang, et al., S-Nitrosylation of Drp1 Mediates β-Amyloid-Related Mitochondrial Fission and Neuronal Injury. Science, 2009. 324(5923):102-105.
Terry, R.D., E. Masliah, D.P. Salmon, et al., PHYSICAL BASIS OF COGNITIVE ALTERATIONS IN ALZHEIMERS-DISEASE - SYNAPSE LOSS IS THE MAJOR CORRELATE OF COGNITIVE IMPAIRMENT. Annals of Neurology, 1991. 30(4):572-580.
King, D.R., M. De Chastelaine, and M.D. Rugg, Recollection-related increases in functional connectivity across the healthy adult lifespan. Neurobiology of Aging, 2018. 62:1-19.
Zhao, S., Y. Li, M. Li, et al., 12-h abstinence-induced functional connectivity density changes and craving in young smokers: a resting-state study. Brain Imaging and Behavior, 2019. 13(4):953-962.
Tomasi, D., G.J. Wang, and N.D. Volkow, Energetic cost of brain functional connectivity. Proc Natl Acad Sci U S A, 2013. 110(33):13642-13647.
Thompson, G.J., V. Riedl, T. Grimmer, et al., The Whole-Brain "Global" Signal from Resting State fMRI as a Potential Biomarker of Quantitative State Changes in Glucose Metabolism. Brain Connect, 2016. 6(6):435-447.
Xu, M., B. Bo, M. Pei, et al., High-resolution relaxometry-based calibrated fMRI in murine brain: Metabolic differences between awake and anesthetized states. J Cereb Blood Flow Metab, 2022. 42(5):811-825.
Wang, D., H. Li, M. Xu, et al., Differential Effect of Global Signal Regression Between Awake and Anesthetized Conditions in Mice. Brain Connect, 2024. 14(1):48-59.
Waldron, A.M., L. Wyffels, J. Verhaeghe, et al., Longitudinal Characterization of [18F]-FDG and [18F]-AV45 Uptake in the Double Transgenic TASTPM Mouse Model. J Alzheimers Dis, 2017. 55(4):1537-1548.
Kuhla, A., L. Meuth, J. Stenzel, et al., Longitudinal [(18)F]FDG-PET/CT analysis of the glucose metabolism in ApoE-deficient mice. EJNMMI Res, 2020. 10(1):119.
Dubois, B., A. Padovani, P. Scheltens, et al., Timely Diagnosis for Alzheimer’s Disease: A Literature Review on Benefits and Challenges. Journal of Alzheimer's Disease, 2015. 49(3):617-631.
Tokuchi, R., N. Hishikawa, K. Sato, et al., Age-dependent cognitive and affective differences in Alzheimer's and Parkinson's diseases in relation to MRI findings. J Neurol Sci, 2016. 365:3-8.
Zhang, X.Y., Z.L. Yang, G.M. Lu, et al., PET/MR Imaging: New Frontier in Alzheimer's Disease and Other Dementias. Front Mol Neurosci, 2017. 10:343.
Kosel, F., J.M.S. Pelley, and T.B. Franklin, Behavioural and psychological symptoms of dementia in mouse models of Alzheimer's disease-related pathology. Neuroscience and Biobehavioral Reviews, 2020. 112:634-647.
Flurkey, K., J. M. Currer, and D.E. Harrison, Chapter 20 - Mouse Models in Aging Research, in The Mouse in Biomedical Research (Second Edition), J.G. Fox, M.T. Davisson, F.W. Quimby, et al., Editors. 2007, Academic Press: Burlington. p. 637-672.
Berry, R.J., M.E. Jakobson, and G.S. Triggs, Survival in wild‐living mice*. Mammal Review, 1973. 3(2):46-57.
Collins, C.R. and R.W. Kays, Patterns of Mortality in a Wild Population of White-Footed Mice. Northeastern Naturalist, 2014. 21(2):323-336, 314.
JAX. 3xTg-AD Strain Details. Available from: https://www.jax.org/strain/004807.
Ataei, A., A. Amini, and A. Ghazizadeh, Gustatory Cortex Is Involved in Evidence Accumulation during Food Choice. Eneuro, 2022. 9(3):6-22.
Dossat, A.M., M.M. Kokoska, J.R. Whitaker-Fornek, et al., Glucagon-Like Peptide-1 Receptors in the Gustatory Cortex Influence Food Intake. The Journal of Neuroscience, 2023. 43(23):4251-4261.
Schiff, H.C., J.F. Kogan, M. Isaac, et al., Experience-dependent plasticity of gustatory insular cortex circuits and taste preferences. Sci Adv, 2023. 9(2):eade6561.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Memory-behavior correlation in pre-onset Alzheimer's reverses following beta amyloid accumulation

Status:

Version 1

Abstract

Figures

Introduction

Methods

AD mouse model

Behavior test

MRI acquisition

fMRI processing

DTI processing

Statistical analysis

Results

Relationship between FC and SC

Relationship between functional connectivity and microstructure

Mouse behavior and brain function

Discussion

The deficit-compensation hypothesis

The relationship between FC and SC

Relationship between behavior and brain function

Potential early biomarkers for human AD

Conclusion

Declarations

Funding sources:

Author Contribution

Acknowledgments:

References

Additional Declarations

Supplementary Files

Status:

Version 1