Summary: A new mathematical model investigates the biological changes that occur during normal brain aging and in neurodegenerative conditions. The findings point to pathways that could guide development of therapies for disorders that impair cognition.
Source: eLife
Researchers report that Alzheimer’s disease shares important biological features with normal aging, according to a new mathematical model published in eLife.
The model integrates data across multiple spatial scales to reveal how molecular changes relate to macroscopic brain alterations in older adults and in people with Alzheimer’s disease. These insights could help identify promising targets for interventions to slow or prevent cognitive decline.
To build the model, the team combined detailed gene-expression information with clinical neuroimaging measurements. The imaging data included molecular PET measures of tau and amyloid, MRI-based indicators of brain structure and tissue integrity, measures of neuronal function, cerebrovascular flow, and metabolism. By linking these diverse data types, the model captures the complex interactions that emerge across molecular and whole-brain levels.
“Most studies look at either microscopic measures, such as gene expression, or macroscopic imaging features, but not the direct causal relationships across these scales,” explains first author Quadri Adewale, a PhD candidate in Neurology and Neurosurgery at McGill University. “We aimed to combine whole-brain transcriptomic maps with longitudinal clinical scans in a personalized framework and to validate this model in both healthy aging and Alzheimer’s disease.”
The analysis used longitudinal multimodal imaging from 460 participants in the Alzheimer’s Disease Neuroimaging Initiative. Each person had at least four different scan types collected at four time points. Participants included 151 clinically healthy controls (HC), 161 with early mild cognitive impairment (EMCI), 113 with late mild cognitive impairment (LMCI), and 35 with probable Alzheimer’s disease (AD).
Gene-expression data came from the Allen Human Brain Atlas, covering 20,267 genes across the whole brain. For analysis, the brain was parcellated into 138 gray-matter regions, enabling the team to align regional transcriptomic patterns with regional imaging measures of pathology, function, and structure.
Using this personalized, spatiotemporal modeling approach, the researchers probed causal links between spatial gene-expression patterns and changes observed in neuroimaging. They then related those linked molecular and imaging dynamics to age-related cognitive changes, including memory and executive function.
Model performance was strongest for predicting cognitive decline in Alzheimer’s disease, with progressively lower predictive accuracy for late and early mild cognitive impairment, and the weakest predictions for healthy controls. This gradient demonstrates the model’s ability to reproduce individualized, multifactorial evolution of toxic-protein accumulation, neuronal dysfunction, and tissue changes documented by the clinical scans.
Focusing first on healthy aging, the authors examined a subgroup of controls who remained clinically stable for nearly eight years to identify genes whose regional expression predicted subtle cognitive changes over time. They identified eight genes whose spatial expression patterns were linked to imaging dynamics and to age-related changes in memory and executive function. Notably, several of these genes influence pathways related to tau and amyloid-β, two proteins central to Alzheimer’s pathology.
Applying the same approach to participants with Alzheimer’s disease, the model identified 111 genes whose regional expression was associated with the imaging measures and with cognitive decline in AD. Functional analysis showed these 111 genes participate in 65 distinct biological processes, many of which are already implicated in neurodegeneration and loss of cognitive function.
“This work offers an unprecedented view of how age-related and Alzheimer’s-associated biological factors interact across scales, and suggests mechanistic roles for the genes we identified,” says senior author Yasser Iturria-Medina, Assistant Professor of Neurology and Neurosurgery at McGill University. “Although Alzheimer’s disease and normal aging share complex biological mechanisms, our results also confirm that AD involves more extensive molecular and macroscopic pathway alterations. A personalized, multiscale model like this can guide the search for genetic and biological targets to extend healthy brain aging and to treat Alzheimer’s progression.”
About this brain aging research news
Source: eLife
Contact: Emily Packer – eLife
Image: The image is in the public domain
Original Research: Open access. “Integrated transcriptomic and neuroimaging brain model decodes biological mechanisms in aging and Alzheimer’s disease” by Quadri Adewale, Ahmed F Khan, Felix Carbonell, Yasser Iturria-Medina, Alzheimer’s Disease Neuroimaging Initiative. eLife
Abstract
Integrated transcriptomic and neuroimaging brain model decodes biological mechanisms in aging and Alzheimer’s disease
Both healthy aging and Alzheimer’s disease involve simultaneous changes across multiple biological factors. Existing generative brain models rarely incorporate measures at both molecular (transcriptomic) and macroscopic (neuroimaging) spatial resolutions simultaneously.
Here, the authors present a personalized, bottom-up spatiotemporal brain model that explicitly links regional expression of hundreds of RNA transcripts with multiple neuroimaging modalities (PET and MRI). In cohorts of elderly individuals and people with AD, the model identifies key genes that modulate tau and amyloid-β accumulation, vascular flow, glucose metabolism, functional activity, and atrophy, all of which contribute to cognitive decline.
The findings reveal that healthy aging and Alzheimer’s disease share particular biological mechanisms, while confirming that AD represents a distinct condition with more pronounced alterations across molecular and macroscopic pathways. Overall, this personalized multiscale model provides new insights into the biology of the aging brain and offers a framework to identify genetic targets for promoting healthy cognitive aging and for treating Alzheimer’s disease progression.