GENETICALLY DIVERSE MOUSE MODELS OF ALZHEIMER’S DISEASE EXHIBIT DIFFERENTIAL MYELOID CELL RESPONSE AND NEURODEGENERATION

Current mouse models of familial Alzheimer's Disease (FAD) have contributed to a deeper understanding of amyloidosis. However, they have not proven to be ideal to study amyloid-induced neurodegeneration or as models for preclinical testing. The majority of FAD mouse models have been generated on the C57BL/6 (B6) genetic background, which may actually confer resilience, limiting the emergence of key, human-relevant AD phenotypes. To address this, we backcrossed transgenes carrying two FAD mutations (amyloid precursor protein, APPswe and presenilin 1, PS1de9; referred to as APP/PS1) from B6 to three inbred wild-derived strains – CAST/EiJ, PWK/PhJ and WSB/EiJ. These strains harbor a similar degree of variation to the human population in terms of genetic variation and significant differences have been previously reported in associated AD-risk behaviors such as cardiovascular health, metabolic disease risk, gut microbiota, telomere length and activity patterns such as circadian rhythm. This is the first time that these wild-derived strains have been utilized to study AD. APP/PS1 and WT mice from each strain were run through an extensive battery to assess metabolic and cognitive function, and AD-relevant neuropathology. Our results show that 8 month wild derived strains carrying APP/PS1 show significant modification of AD-relevant phenotypes compared to B6.APP/PS1 including cognitive decline, neurodegeneration, and neuroinflammation. WSB.APP/PS1 and CAST.APP/PS1 mice failed to demonstrate preference in a novel spatial recognition task, and exhibited neuronal cell loss in memory-related brain regions. Significant increases in Aβ42 was accompanied by the presence of cerebral amyloid angiopathy in both CAST.APP/PS1, and more prominently, WSB.APP/PS1. Lastly, baseline myeloid cell numbers varied across the strains, and differential plaque associated response was observed in APP/PS1 mice. These findings challenge the currently accepted reductionist approach of examining a human mutation in a single mouse strain in order to model multifactorial diseases such as AD. These genetically diverse AD mouse models represent a unique resource to understand the biological underpinnings of AD, and lay the foundation for identifying novel therapeutic targets.