Alzheimer’s disease (AD) pathologies emerge in stages, starting with early amyloid beta accumulation in the brain, followed by formation of tau tangles inside neurons, then synapse loss and neuron death. This chain of events, amyloid to tau to neurodegeneration, is now widely accepted by the AD field, but how these hallmark pathologies interact to ultimately cause clinical symptoms remains a black box. A critical unanswered question is how rising amyloid beta levels cause the tau pathology seen in AD brains. Understanding this mechanism could lead to strategies that break the link between amyloid accumulation and tau tangles, thereby preventing AD progression.
A major roadblock to exploring the relationship between amyloid and tau has been that traditional cell and mouse models do not develop tau tangles following amyloid aggregation. This has forced researchers to study their individual effects in separate systems. However, recent breakthroughs have begun to address this limitation. Drs. Kim, Choi, and Tanzi made significant strides for the field by creating the 3D Alzheimer’s in a Dish model, in which human cells in culture actually develop tau pathologies downstream of amyloid beta. Similarly, Dr. De Strooper and his team achieved another major advance when they developed a novel method to surgically transplant human neurons into the brains of amyloid mouse models (xenograft model). While labs are now adopting these xenograft methods, the process remains far more complex than simply breeding mice with genetic variations to model AD. Developing mouse models that recapitulate all AD pathologies—in the correct order—continues to be a high priority for the field.
Drs. Edwards and Hardy are among the researchers working to address this challenge. In a previous funding cycle with Cure Alzheimer’s Fund, they carefully characterized novel mouse models engineered to make human versions of the amyloid precursor protein (APP) or tau genes independently. They identified a particularly promising combination of two such mouse models: one that makes human APP with mutations that cause AD (NLF) and one that makes a single copy of the normal human tau gene (Taum/h). Breeding these mice together results in offspring that carry both genes (called NLFTaum/h mice). Remarkably, this combination leads to tau pathologies at old ages, following rising levels of amyloid beta. These exciting results suggest that this new mouse model will be a valuable resource for the field; however, further work is needed to fully characterize and validate it for use in preclinical studies. In this follow-on, the team is taking advantage of their colonies of young and aging NLFTaum/h mice to characterize plasma biomarkers across pathological stages and compare these data with the results of biomarker measurements in humans.
They proposed four experimental aims. In the first aim, in collaboration with Dr. Henrik Zetterberg at UCL, they are mapping changes in AD-relevant biomarkers—measured in plasma—across the lifespan of NLFTaum/h mice and controls. They are already collecting preliminary data showing that key plasma biomarkers used in humans (pTau-217, NFL) are measurable and increasing in the NLFTaum/h mouse with age. For these experiments, they are collecting blood samples from NLF and NLFTaum/hmice from 8 to 28 months old and are measuring the levels of approximately 120 proteins in these samples using a new technology (NULISA) established in part by Dr. Zetterberg. In the second aim, they are investigating how changes in these blood biomarkers correspond to changes in the brain at the same ages. They are measuring multiple amyloid- and tau-related pathologies, including microglia inflammation, in the brains of mice for which they are also collecting blood samples. In the third (and related) aim, they are focusing specifically on how changes in blood biomarkers relate to the development of tau pathology in the oldest mice, aged 24–28 months. They are looking at the correspondence between blood and brain data at the same age and are also looking back across time to identify patterns in blood biomarkers that predict the severity of tau pathology in the oldest mice. In the fourth aim, which they are running in parallel with the others, they are collecting similar data from other mouse models with and without different AD pathologies to compare with the NLFTaum/h mice (tau only, APP only, or NLFTau x Trem2). Importantly, the inclusion of human tau alone does not result in tau pathology. These comparisons are helping them identify which biomarker changes correspond to the clinical progression of human AD as opposed to reflecting only amyloid- or tau-induced pathologies.
In the first eight months of funding, they are progressing exactly as planned. The first panel of plasma samples across all ages and genotypes is being analyzed, which is generating substantial preliminary biomarker data for Aim 1. The brains from these mice are being sectioned and prepared for immunohistochemistry and molecular analyses (Aim 2) to determine how the biomarker changes seen in Aim 1 correspond to brain changes. Initial data from the older mice (18–24 months) are already available. Meanwhile, they are maintaining the mouse colony for future studies (Aim 4). In addition to everything being on track for this current proposal, they have also successfully completed and analyzed the human spatial transcriptomics data that they collected under a previous funding cycle. The results, which have been submitted for publication, confirm their previous findings from mouse spatial transcriptomic data that the TREM2 microglial gene, and a range of other Alzheimer’s-related genes, are only upregulated when microglia are directly in contact with plaques. Looking ahead to the next year, they will employ a postdoctoral fellow and additional students to conduct detailed analyses of the brains from the first and subsequent panels, initiating Aims 2–4, and continue serial plasma sampling as the mice age. The second panel is expected to be completed by March, and additional panels obtained through collaboration with Zetterberg’s team are allowing them to extend the project to include more time points and additional genotypes, keeping the project fully on track.
