Mapping vulnerability or resilience in Alzheimer's through imaging and genetics

It's been recognized for some time that Alzheimer's disease affects brain regions differently and that tau - a protein known to misbehave - plays an important role in the disease. Normally, tau helps stabilize neurons, but in Alzheimer's disease, it begins to misfold and tangle inside neurons. It spreads across the brain forming toxic clumps that impair neuronal function and ultimately lead to cell death.

Brain areas like the entorhinal cortex and hippocampus succumb early to tau tangles, while other areas, like the primary sensory cortices, remain resilient to the disease. In the quest to better understand this selective vulnerability (SV) or resilience (SR) to Alzheimer's disease, researchers have looked to gene association and transgenic studies to identify Alzheimer's risk genes. But past research has not shown a clear link between the location of genetic risk factors and associated tau pathology.

Now, a new study by UC San Francisco researchers has made a leap toward answering that question - by combining brain imaging, genetics, and advanced mathematical modeling into a powerful new lens. The study, published July 9 in Brain, shows multiple distinct pathways by which risk genes confer vulnerability or resilience in Alzheimer's disease.

The study introduced a model of disease spread called the extended Network Diffusion Model (eNDM). The researchers applied this model on brain scans from 196 individuals at various stages of Alzheimer's. They subtracted what the model predicted from what they saw in the scans. The leftovers, called "residual tau," pointed to areas where something else besides brain connections influence the buildup of tau - in this case, genes.

Using brain gene expression maps from the Allen Human Brain Atlas, the researchers tested the degree to which Alzheimer's risk genes explain the patterns of both actual and residual tau. This allowed them to tease apart genetic effects that act with or independently of the brain's wiring.

We think of our model as Google Maps for tau. It predicts where the protein will likely go next, using real-world brain connection data from healthy people."

Ashish Raj, PhD, senior study author, UCSF professor of Radiology and Biomedical Imaging

This upends traditional view of how tau moves in the brain

The study team uncovered four distinct gene types based on how much and in what manner they were predictive of tau: Network-Aligned Vulnerability (SV-NA), which are genes that boost tau spread along the brain's wiring; Network-Independent Vulnerability (SV-NI), which are genes that promote tau buildup in ways unrelated to connectivity; Network-Aligned Resilience (SR-NA), which are genes that help protect regions that are otherwise tau hotspots; and Network-Independent Resilience (SR-NI), which are genes that offer protection outside of the network's usual path - like hidden shields in unlikely spots.

"Vulnerability-aligned genes dealt with stress, metabolism, and cell death; resilience-related ones were involved in immune response and the cleanup of amyloid-beta - another Alzheimer's culprit," said study first author Chaitali Anand, PhD, a UCSF post-doctoral researcher. "In essence, the genes that make parts of the brain more or less likely to be affected by Alzheimer's are working through different jobs - some controlling how tau moves, others dealing with internal defenses or cleanup systems."

This research built on another recent UCSF study in mice, published May 21 in Alzheimer's & Dementia, which demonstrated that tau does not travel randomly or diffuse passively; instead, it follows the brain's wiring pathways with a distinct directional preference. Using a system of differential equations called the Network Diffusion Model (NDM), the research team was able to show the dynamics of tau spread between connected brain regions, challenging the traditional view that tau spreads simply by diffusing through extracellular space or leaking from dying neurons.

"Our research showed that tau propagates trans-synaptically, traveling along axonal projections driven by active transport processes rather than passive diffusion, and exploiting active neural pathways in the preferred retrograde direction," said Justin Torok, PhD, a post-doctoral researcher working in the Raj lab.

In the current study, network-based analyses complemented the existing approaches for validating and identifying gene-based determinants of selective vulnerability and resilience. Genes that respond independently of the network having different biological functions than those genes that respond in concert with the network.

"This study offers a hopeful map forward: one that blends biology and brain maps into a smarter strategy for understanding and eventually stopping Alzheimer's disease," said Raj. "Our findings offer new insights into vulnerability signatures in Alzheimer's disease and may prove helpful in identifying potential intervention targets."