Introduction

The Model Organism Development and Evaluation for Late-Onset Alzheimer’s Disease (MODEL-AD) Consortium was established to develop the next generation of Alzheimer’s disease (AD) models based on human genomic and imaging data. Recently, MODEL-AD has generated a new human Aβ Knock-In (hAβKI) mouse that substitutes 3-amino acids in the mouse Aβ locus with that of the human amyloid precursor protein (App) gene (Baglietto-Vargas et al., 2021). This substitution leads to age-dependent cognitive and synaptic impairments as these mice progressively age. Neuroimaging of AD is a critical component of human diagnoses and preclinical magnetic resonance imaging (MRI) can investigate phenotypic alterations in these new mice models. Specifically, we used diffusion MRI (dMRI) to investigate brain-wide structural alterations in conjunction with connectomic analyses in the hAβKI mouse across its lifespan.

Methods

We examined hippocampal synaptic plasticity in slices from male and female hAβKI mice at 4, 12 and 18mo of age (n=5/group/sex/genotype). For imaging a total of 105 male and female hAβKI mice at the same time points (WT n=51, hAβKI n=54; ~n=8/sex/time point). All imaging was performed on a 9.4T (Bruker Biospin, PV5.1) MR scanner. The following sequences were utilized: 1) diffusion tensor imaging (DTI): 5 B0, 30 directions b=0, 3000 mm2/sec, 25 slices 0.5mm thick for full brain coverage, 2) multiecho T2 for quantitative relaxation, 3) 3D T2 RARE for volumetric assessments and 4) susceptibility weighted imaging for iron deposition. Our automated pipeline for regional analyses utilized the Australian Mouse Brain Mapping Consortium (AMBMC) atlas which we split for bilateral measurements. DTI metrics (axial (AxD), radial (RD), mean (MD) diffusivities and fractional anisotropic (FA) maps) data were extracted from 42 bilateral regions encompassing cortical, subcortical and white matter regions. Tractography and connectomics were derived using DSI studio.

Results

Impairments in hippocampal long-term (LTP) were observed in male hAβKI mice at 4 months of age relative to WT with no significant change observed in females. In female hAβKI mice, impairments in LTP were observed at 12 and 18 months of age relative to WT with no significant change observed in males (Fig. 1A). Brain volumes were significantly decreased in male hAβKI compared to WT mice, but no changes were found in female hAβKI mice. Volumetric increases in select brain regions, such as the anterior cingulate cortex in male hAβKI mice compared to WT were observed but was not related to tissue relaxation (T2) (Fig. 1). dMRI heatmaps of MD highlighted regional differences between males and females across time and genotype, with MD progressively decreasing with time (Fig. 2). Statistical analyses for regional changes showed that male hAβKI mice had progressive changes within the brain with advancing age whereas the largest dMRI changes in females was at 12mo of age (Fig. 3). For example, AxD in 18mo male hAβKI mice was increased in 4 of 7 white matter regions but not in female hAβKI mice compared to WT. This would suggest regional and temporal sensitivity to diffusion metrics in the hAβKI mouse brain. As these animals exhibit impaired hippocampal long-term potentiation, we probed the CA1 region for diffusion abnormalities. Compared to male WT mice, we found decreased RD and MD but no change in AxD whereas FA was increased within the CA1 region of the hippocampus of male hAβKI mice (no change in female mice). Ongoing multishell dMRI analyses will report neurite and cellular density measures from NODDI modeling.

Discussion

Human AD is a complex, progressive and multi-faceted disease process in late-onset AD (LOAD) patients. At the present time there are no animal models that recapitulate all the known features in LOAD patients. The MODEL-AD consortium has been tasked to develop better LOAD mouse models. One such recently developed model is the hAβKI mouse. In male hAβKI mice dMRI metrics, particularly RD and MD (along with FA) appeared to be sensitive to identify ongoing histopathological perturbations. The hippocampus of male hAβKI mice but not females, showed reduced diffusion (RD, MD) consistent with altered tissue structure. White matter was also similarly affected but to an even larger extent. These findings complement our observations of impaired hippocampus-dependent memory and LTP in hAβKI mice.

Conclusion

Preclinical MRI can be used to phenotype new mouse models of AD. Using a newly developed mouse model of AD, the hAβKI mouse, we identified progressive altered regional tissue metrics. The hippocampal CA1 region in the hAβKI mouse appeared to be particularly vulnerable, findings consistent with observed impairments in confirmed LTP and behavioral findings. MRI imaging of new mouse models is a powerful tool to investigate tissue level modifications to brain structure and function. Comparison of MRI from mouse models of AD and those from human MRI acquisitions provides the opportunity to compare and contrast species for evaluation of potential imaging biomarkers.

Acknowledgements

MODEL-AD was established with funding from The National Institute on Aging (U54 AG054345-01, U54 AG054349-01).

References

Baglietto-Vargas, D., Forner, S., Cai, L. et al. Generation of a humanized Aβ expressing mouse demonstrating aspects of Alzheimer’s disease-like pathology. Nat Commun 12, 2421 (2021). https://doi.org/10.1038/s41467-021-22624-z