Breaking β-sheets in FUS prion-like domain preserves phase separation and function but prevents aggregation and toxicity

The RNA-binding protein Fused in Sarcoma (FUS) undergoes phase separation associated with RNA processing. However, the prion-like low complexity (LC) domain of FUS forms solid-like aggregates in neurodegenerative diseases. Whether the formation of β-sheet structure associated with pathology is also physiologically/functionally relevant is debated. Similarly, if mislocalization alone or concomitant aggregation is responsible for FUS gain-of-function toxicity remains to be probed. Here, we introduce β-sheet breaking proline residues into FUS LC with the goal of preventing cross-β-driven aggregation without disrupting essential functions and phase separation. β-sheet-deficient FUS variants maintain native-like global motions, disorder, and phase separation, but no longer show a liquid-to-solid transition (LST). Biochemical partitioning, cellular localization, and auto- and cross-regulatory functions of FUS all remain essentially unchanged. Conversely, FUS-induced neurodegeneration in several Drosophila models is drastically reduced. These findings suggest a strategy for mitigating disease-related toxicity through backbone structure modulation to prevent prion-like domain protein aggregation. GRAPHICAL ABSTRACT SUMMARY The RNA-binding protein Fused in Sarcoma (FUS) undergoes phase separation as part of its physiological function but can aberrantly aggregate into solid-like assemblies in amyotrophic lateral sclerosis and frontotemporal dementia. To dissect the role of β-sheets in both function and pathological transition, we engineered β-sheet-preventing FUS variants via targeted proline residue insertions in the prion-like disordered region. These variants retained native structure, motions, and phase behavior yet showed dramatically reduced aggregation, both as an isolated prion-like domain and in full-length FUS. Crucially, these variants maintained a panel of FUS cellular functions that depend on FUS condensation but prevented FUS toxicity in fly models of neurodegeneration. Our findings implicate β-sheets as key drivers of FUS condensate maturation and neuronal toxicity, highlighting β-sheet modulation as a therapeutic strategy against FUS-related neurodegeneration. HIGHLIGHTS Targeted proline additions disrupt β-sheet formation in FUS without altering native conformations, dynamics, or phase separation behavior β-sheet-deficient FUS variants prevent aggregation and liquid-to-solid transitions while retaining key biological functions In vivo models reveal attenuated toxicity of β-sheet-deficient FUS in Drosophila β-sheets are identified as central drivers of condensate maturation and neuronal death, offering a therapeutic entry point for modulating prion-like domain pathology