Sophie Hertel

Intrinsically disordered proteins (IDPs) often associate into biomolecular condensates. While micron‐sized condensates are observed above a well‐defined saturation concentration, the formation of nanoscale clusters also occurs at subsaturated concentrations. The diversity and physical nature of such nanoscale clusters remain underexplored with respect to the well‐established condensate state. Here we show that the Parkinson's disease‐associated protein α‐synuclein forms highly monodisperse assemblies at low ionic strength, distinct from the larger polydisperse assemblies previously observed at higher ionic strengths. Using a range of biophysical methods, we identify monodisperse clusters of α‐synuclein composed of ∼20 molecules that maintain their size over a broad concentration range. These clusters are stable over time and do not progress to amyloid fibril formation. Their response to mutational perturbations and chemical destabilization, as well as their thermal cycling support a micelle‐like assembly. Increasing ionic strength shifts the system toward the formation of larger, polydisperse nanoclusters and condensates that promote amyloid fibril formation. Our results reveal a hitherto uncharacterized assembly state of α‐synuclein and highlight the complexity of nanoscale assembly processes by IDPs.

Phase separated condensates of α-synuclein (α-Syn) accelerate amyloid fibril formation, a process implicated in Parkinson's disease pathogenesis. Yet, the precise effects of pathologically relevant α-Syn sequence modifications on this process remain unclear. Here, we show that sequence truncations exert the strongest influence on condensate thermodynamics, material properties, and amyloid aggregation, whereas familial point mutations impart minimal effects. Among the tested familial variants (A30P, H50Q, G51D and A53T), only G51D forms condensates that show a reduced propensity for amyloid fibril formation. Truncated variants undergo rapid gelation and form amyloid fibrils almost immediately. Extending our study to multicomponent systems where α-Syn is a client, we show that α-Syn can dissolve DNA-peptide coacervates or assemble into Pickering clusters on condensate surfaces-regulating condensate fusion and nucleic acid partitioning. These functions depend on the acidic C-terminal domain of α-Syn. Together, our results show disease-relevant modifications can modulate α-Syn phase behavior, both in pathological and physiological contexts.

Phase separated condensates can accelerate α-synuclein (α-Syn) amyloid fibril formation implicated in Parkinson’s disease pathogenesis. The effects of pathological modifications, i.e., truncations and familial mutations on the thermodynamics, material properties, and the extent of amyloid aggregation within α-Syn condensates remain elusive. Here, we quantitatively demonstrate that terminal truncations significantly alter α-Syn phase separation, while familial mutations impart minimal effects. Spontaneous sol-gel phase transitions of the truncated α-Syn variants could give rise to amyloid fibrils almost instantly within condensates, suggesting similarities between molecular interactions driving both processes. Extending our study to model coacervate and condensate systems where α-Syn acts as a client, we find α-Syn can dissolve coacervates and form Pickering clusters on condensate surfaces—regulating their size. Additionally, the C-terminal region of α-Syn modulates nucleic acid sequestration within condensates. Together, our findings reveal diverse effects of α-Syn modifications on phase separation, both in pathological and physiological contexts. GRAPHICAL

HIV can infect non-dividing cells because the viral capsid can overcome the selective barrier of the nuclear pore complex and deliver the genome directly into the nucleus^1,2. Remarkably, the intact HIV capsid is more than 1,000 times larger than the size limit prescribed by the diffusion barrier of the nuclear pore³. This barrier in the central channel of the nuclear pore is composed of intrinsically disordered nucleoporin domains enriched in phenylalanine–glycine (FG) dipeptides. Through multivalent FG interactions, cellular karyopherins and their bound cargoes solubilize in this phase to drive nucleocytoplasmic transport⁴. By performing an in vitro dissection of the nuclear pore complex, we show that a pocket on the surface of the HIV capsid similarly interacts with FG motifs from multiple nucleoporins and that this interaction licences capsids to penetrate FG-nucleoporin condensates. This karyopherin mimicry model addresses a key conceptual challenge for the role of the HIV capsid in nuclear entry and offers an explanation as to how an exogenous entity much larger than any known cellular cargo may be able to non-destructively breach the nuclear envelope.

Summary Paragraph: HIV can infect non-dividing cells because the viral capsid can overcome the selective barrier of the nuclear pore complex and deliver the genome directly into the nucleus. Remarkably, the intact HIV capsid is over one thousand times greater than the size-limit prescribed by the nuclear pore’s diffusion barrier. This barrier is a phase-separated condensate in the central channel of the nuclear pore and is comprised of intrinsically-disordered nucleoporin domains enriched in phenylalanine-glycine (FG) dipeptides. Through multivalent FG-interactions, cellular karyopherins and their bound cargoes solubilise in this phase to drive nucleocytoplasmic transport. By performing an in vitro dissection of the nuclear pore complex, we show that a pocket on the surface of the HIV capsid similarly interacts with FG-motifs from multiple nucleoporins and that this interaction licenses capsids to penetrate nucleoporin condensates. This karyopherin mimicry model resolves a key conceptual challenge for the role of the HIV capsid in nuclear entry, and explains how an exogenous entity much larger than any known cellular cargo can non-destructively breach the nuclear envelope.