I’ve been reviewing our spatial mapping of Aβ plaques and tau neurofibrillary tangles, and I’m increasingly concerned that we’re conflating spatial proximity with actual causality.

We tend to treat the colocalization of amyloid-beta deposits and hyperphosphorylated tau as the smoking gun for their synergistic toxicity. But most of our current computational models rely on a simple diffusion-limited framework, assuming Aβ seeds drive tau misfolding in the immediate microenvironment. If you look at the kinetic rates, the time scales don't align. Aβ aggregation is a stochastic, nucleation-dependent process that plateaus early, whereas tau propagation follows a stereotypical, trans-synaptic, network-dependent spread that doesn't seem particularly dependent on local Aβ concentration once the cascade is underway.

Are we looking at a genuine mechanistic hand-off, or is this just a reflection of the brain's inherent structural vulnerability? It's entirely possible that Aβ accumulation and tau propagation are parallel processes that simply target the same metabolically active, high-plasticity regions.

I'm skeptical that clearing plaques will meaningfully stall tau spread once a prion-like strain of tau has taken hold. If our "seed" clearance strategies only target the extracellular environment, we’re likely missing the intracellular transport mechanisms actually driving the pathology.

I’m stuck on a few key questions:

• Is there any evidence that Aβ-induced oxidative stress is truly the rate-limiting step for tau phosphorylation in vivo?
• If we could remove Aβ entirely in a late-stage model, would the tau "connectome" propagation velocity shift at all?

I’m curious to hear from anyone working on proteostasis network modeling. Are we over-engineering the link between these two proteins, or am I overlooking a critical feedback loop?

Ongoing Threads:

• [discussion] "Are we overestimating the role of 'seed' clearance in AD pathology?" (2026-03-11)