Abstract
It has been postulated that Alzheimer's disease (AD) pathology interacts and resides within system-level circuits of the human brain, long before the onset of cognitive symptoms. However, a side-by-side comparison of tissue loss, amyloid-β, and Tau deposition in early stages of the disease has been precluded until the recent advent of Tau tracer-based neuroimaging. In this study, we used Tau positron emission tomography and network analyses to disentangle these pathological relationships. We found that Tau and amyloid-β deposits are associated with distinctive spatial patterns of brain tissue loss. Moreover, we uncovered the network interdigitations of Tau and amyloid-β in the cortical mantle. These findings contribute significantly to our understanding of how two main hallmarks of AD pathology propagate across the elderly human brain.
Introduction
For decades, the "amyloid hypothesis" has been central to Alzheimer’s disease (AD) research. However, the precise relationship between amyloid-β (Aβ) plaques, Tau tangles, and actual brain tissue loss (atrophy) has remained elusive, particularly in the preclinical stages of the disease. While Aβ can be visualized using PET imaging for years before symptoms appear, Tau has historically been much more difficult to image in living patients. This study addresses the need for a side-by-side, in vivo comparison of Aβ and Tau to understand how these pathologies interact and contribute to the network-level degradation of the aging brain in individuals who are still cognitively normal.
Methods
The researchers analyzed data from 88 cognitively normal elderly participants (mean age 76.2) from the Harvard Aging Brain Study. All participants had a Clinical Dementia Rating (CDR) of 0 and high MMSE scores, representing a "normal aging" state.
A multimodal imaging approach was employed:
- Tau PET ($^{18}$F-AV-1451): To map Tau deposition (measured by SUVR score).
- Amyloid PET ($^{11}$C-PiB): To map Aβ deposition (measured by DVR score).
- Structural MRI: To measure Gray Matter (GM) intensity as a proxy for tissue loss.
The team utilized voxel-wise graph-theoretical analyses to calculate both local-to-local associations (pathology and atrophy in the same region) and local-to-distributed associations (pathology in one region affecting distant circuits). Network topology was further explored using weighted degree connectivity and the Kamada–Kawai algorithm to identify pathological "hubs" and central connectors.
Results
Spatial Distribution of Pathology
The study found distinct spatial signatures for the two proteins. Tau deposition was primarily localized to the ventral and lateral temporal lobes. In contrast, Aβ was widespread, affecting the lateral and midline frontal and parietal cortices.
Pathological Impact on Brain Structure
Tau and Aβ relate to neurodegeneration through different mechanisms:
- Tau as a Local Driver: Tau deposits showed significant negative correlations with local GM intensity, particularly in the lateral inferior temporal lobe ($pR = -0.277$).
- Aβ as a Global Influencer: Aβ deposits showed widespread negative correlations with GM, especially in the lateral parietal and frontal cortices ($pR = -0.340$). Unlike Tau, Aβ was associated with distributed atrophy, suggesting its presence in one region may drive degeneration in distant, connected neural circuits.
The Network Core
Graph analysis identified a "core subnetwork" where Tau and Aβ interactions are most intense. The inferior-lateral temporal lobe emerged as a critical "gateway" or interdigitation zone, showing the strongest positive correlations between the two pathologies. Nodes in this interaction zone exhibited the highest betweenness centrality, indicating they are pivotal in the pathological communication across the brain.
Discussion
These findings suggest that while Tau and Aβ have unique signatures of destruction, they converge in a critical temporal lobe network that may facilitate the transition from normal aging to AD. The observation that Tau acts locally while Aβ influences distributed networks provides a new framework for understanding disease propagation.
Limitations and Future Directions
As a cross-sectional study, these results represent a snapshot in time; longitudinal data is necessary to confirm the causal direction of pathology-driven atrophy. Additionally, potential off-target binding of the Tau tracer in the choroid plexus and the presence of Primary Age-Related Tauopathy (PART) are noted limitations. Future research should focus on the identified "core" network regions as early biomarkers for clinical trials, as they likely represent the tipping point for significant neurodegeneration.
Summary updated