Progressive remodeling of global protein interaction networks in a mouse model of tauopathy.

Neurodegenerative disease is marked not just by loss of proteins or cells, but by dynamic rewiring of macromolecular interaction networks that precede and drive pathology. Here, we present a temporally resolved, systems-scale map of multi-protein complex remodeling in a tauopathy model, integrating co-fractionation mass spectrometry, quantitative phosphoproteomics, and machine learning to decode phosphorylation-dependent shifts in protein interactomes across disease progression. This interactomic atlas serves as a resource to investigate functional assemblies, including MAPT-Dpysl2 and Cyfip1-actin, as well as ptk2b-Kars/Dars and ptk2b-Msn-Farsb-complexes, that modulate early disease phenotypes in vivo. Notably, it captures the composition, dynamics, and regulatory state of protein complexes at scale. By revealing how phosphorylation tunes macromolecular complex architecture and function, this work highlights dynamic network instability that develops over the course of tauopathy, and establishes a generalizable framework for mechanistic dissection of functional complexes in neurodegeneration.