🧵15/ Huge thanks to our amazing team and coauthors!

Endless thanks to @jwvogel.bsky.social for guiding and supporting this work from day one. To our amazing team DeMON lab, especially @anlijuncn.bsky.social for enormous support.

🧵9/ (Cont.) We identified two (of the four subtypes) with especially interesting mechanisms:

🔷Posterior subtype: Linked strongly to acetylcholine—already a treatment target!
🔷MTL-sparing subtype: Spread better explained by specific *functional* brain networks.

🧵7/ What drives 🔴Tau Load? — Determined by connectivity and local biological factors

Connectivity isn’t enough—local vulnerability is key. Factors such as regional amyloid-beta levels, MAPT expression, cerebral blood flow, and neurotransmission markers profiles all play a role.

🧵6/ Surprisingly, the balance of neurotransmission signals 🎯 (excitatory vs. inhibitory) and 5-HT receptors strongly overlap with tau presence. Regions with more ionotropic excitatory signaling are particularly vulnerable—supporting work focusing on E:I balance in AD.

🧵5/ What drives ⚪️Tau Presence? — Primarily brain connectivity

As expected, tau follows structural connectivity. But unexpectedly, we also found 🎯neurotransmitter marker distribution does an even better job explaining the pattern. We found this fascinating and explored further.

🧵4/ We next conducted an exploratory analysis using eight types of normative connectomes and 49 regional biological factors (thanks to @misicbata.bsky.social, his team, and the Allen Human Brain Atlas).

🧵3/ We first tested well-known factors that influence tau in AD. Found:

⚪️ Tau presence is largely explained by synaptic spread, measured via structural brain connectivity.
🔴 Tau load is shaped by both synaptic spread and regional MAPT gene expression / beta-amyloid deposition

🧵2/ Why? We adapted a mechanistic computational model simulating tau accumulation in the human brain (SIR, journals.plos.org/plosbiology/...). Besides connectome-based spread, the model simulates regional tau synthesis, misfolding and clearance, allowing empirical regional data to affect.

🧵1/ Using tau-PET imaging data from
@biofinder.bsky.social, we distinguish between:

⚪️Tau Presence: Whether and when tau appears in a brain region?
🔴Tau Load: How much tau builds up there?

We show that ⚪️tau presence is highly consistent across people, but 🔴tau load varies widely.