Liquid-liquid phase separation in neurodegenerative diseases: Mechanisms, pathological transitions, and therapeutic...

Liquid-liquid phase separation (LLPS) provides a biophysical framework for understanding biomolecular organization, yet its pathological relevance in neurodegeneration remains incompletely resolved. In this narrative review, we critically synthesize current evidence on LLPS in Huntington's disease, Parkinson's disease, Alzheimer's disease, and amyotrophic lateral sclerosis, focusing on mutant huntingtin, α-synuclein, amyloid-β, tau, TDP-43, and FUS. We first outline the molecular principles-including intrinsic disorder, multivalent interactions, and environmental sensitivity-that govern their phase behavior. We then examine how dynamic condensates may, under specific cellular contexts, transition toward less reversible oligomeric, gel-like, or solid assemblies, integrating evidence from in vitro, cellular, and in vivo studies. Importantly, current evidence does not support LLPS as a universal or obligatory precursor of pathological aggregation; rather, its biological consequences appear to depend on condensate composition, material properties, subcellular localization, and proteostasis capacity, while direct demonstration of disease-relevant condensates in vivo remains limited. We further discuss whether aberrant LLPS represents a cause, consequence, or context-dependent component of neurodegeneration and highlight emerging strategies that modulate condensate formation, maturation, or clearance while preserving physiological functions. Overall, this review defines both the mechanistic potential and current limitations of the LLPS framework and identifies key questions that must be addressed before pathological phase-state modulation can become a clinically actionable therapeutic strategy.