Back

Programmable self-assembling system to model intracellular protein aggregation and decode neurodegenerative diseases

Fan, Y.; Hoang, J.; Steele, C.; Benchetrit, Y.; Dutta, S.; Coughlin, G.; Appling, N.; Ding, X.; Lin, C.; Qu, Z.; Mackey, E. D.; Gradinaru, V.

2025-10-06 neuroscience
10.1101/2025.03.06.641009 bioRxiv
Show abstract

Pathological protein aggregation is a conserved feature of neurodegenerative diseases. However, slow progression, aggregates heterogeneity and selective vulnerability make it challenging to model diseases and dissect mechanisms. Here, we present a genetically-encoded, modular platform of self-assembling protein that enables inducible, tunable and cell-type-specific formation of intracellular protein aggregates. We engineered self-assembling variants of -synuclein (SAS), Tau, and TDP-43, recapitulating hallmarks of Parkinsons disease, Alzheimers disease, and ALS, respectively. SAS formed Lewy body-like inclusions, nucleated endogenous -synuclein, triggered neuroinflammation and neurite degeneration; Self-assembling-Tau induced tangles formation, extracellular A{beta} and severe neurodegeneration; Self-assembling-TDP caused TDP-43 mislocalization and cytoplasmic inclusions with mild degeneration. Notably, the system revealed disease-specific aggregate-organelle interactions. In vivo, systemic or dopamine neuron-targeted delivery of SAS induced progressive motor deficits, dopaminergic neuron loss, and microglial activation. Our self-assembling system robustly mirrors histological, transcriptional, and behavioral features of human disease, providing a powerful platform to dissect mechanisms and accelerate therapeutic development. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=189 HEIGHT=200 SRC="FIGDIR/small/641009v2_ufig1.gif" ALT="Figure 1"> View larger version (71K): org.highwire.dtl.DTLVardef@1906bb2org.highwire.dtl.DTLVardef@8e8141org.highwire.dtl.DTLVardef@1c7e62dorg.highwire.dtl.DTLVardef@d8c05a_HPS_FORMAT_FIGEXP M_FIG C_FIG

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.