Hsp104, a yeast protein-remodeling factor, can disaggregate misfolded proteins implicated in neurodegeneration. Although many potentiated Hsp104 variants have been generated, suboptimal properties have limited their application in mammalian systems. Here, we present the development of a high-throughput screening approach for identifying enhanced Hsp104 variants. To screen a large library of variants in parallel and with a quantitative output, we coupled a live-or-die yeast-based selection with next-generation sequencing. The identified Hsp104 variants solubilize preformed α-synuclein and TDP-43 aggregates, inhibit seeding of preformed α-synuclein fibrils in mammalian biosensor cells, restore TDP-43 splicing of native targets, and have diminished off-target toxicity in mammalian cells. Certain variants show distinct changes in ATP hydrolysis, which we suggest is the key driver of these improved properties. We anticipate that our approach is broadly applicable to a range of protein engineering targets to allow coupling of a phenotypic readout to high-throughput quantitative analysis of variants in parallel. • Potentiated Hsp104 variants can be identified via selection and sequencing approach • Next-generation sequencing allows quantitative comparisons of variants in parallel • Potentiated Hsp104 variants suppress misfolding and clear preformed aggregates • Hsp104 variants restore TDP-43 splicing of native targets Ryan et al. have developed a high-throughput approach for identifying variants of the amyloid disaggregase Hsp104. Their approach facilitates analysis of large Hsp104 variant libraries in parallel and with quantitative output. Using this approach, they identified disaggregase variants that restore function to proteins that misfold in ALS and Parkinson’s disease.
Ryan et al. (2026) studied this question.