Biophysically realistic network-level transport model of tau progression with exosome-mediated release and uptake processes

The spatiotemporal progression of tau aggregates in neurodegenerative diseases like Alzheimers follows the brains structural connectome, yet a profound gap exists between the slow macroscopic spread observed over years and the rapid protein kinetics occurring over hours. Current graph-diffusion models fail to reconcile this timescale disparity or incorporate the cellular mechanisms driving transmission. Here, we advance the Network Transport Model (NTM) to bridge these scales by integrating directional active transport along microtubules, continuous toxic tau production, and exosome-mediated trans-neuronal release and uptake. This framework constitutes one of the most mechanistically complete and biologically detailed models of tau spread on the whole brain to date, representing a significant innovation in how multiscale proteinopathies are simulated. To overcome the computational complexity of the underlying partial differential equations, we developed a quasi-static approximation that separates fast axonal transport from slow network-wide exchange. Simulations on the whole-brain mouse connectome demonstrate that this framework emergently replicates empirical tau propagation patterns without the need for case-specific empirical fitting. Our results identify trans-neuronal release and uptake rates as the primary mechanistic "bottleneck" on macroscopic spread, providing a biologically grounded explanation for the diseases slow progression. Furthermore, we find that high aggregation rates can paradoxically sequester tau within neurons, limiting global transmission, while transport polarity (anterograde vs. retrograde) fundamentally dictates spatial patterning. By linking molecular mechanics to system-wide pathology, this model provides a predictive "in-silico" framework to evaluate how cellular-targeted interventions might alter the trajectory of tauopathic dementias.