Intracellular proteins that condense into droplets via liquid-liquid phase separation (LLPS) have been suspected of favoring the formation of amyloid fibrils, and thereby causing neurodegenerative diseases. For example, the protein alpha-synuclein (ɑSyn) forms both liquid droplets and amyloid fibrils. The question, then, is whether ɑSyn’s droplet formation and amyloid formation are related. According to researchers at the Paul Scherrer Institute (PSI), the answer is, not necessarily. In fact, protein droplets may help dissolve aggregate protein.
Detailed findings appeared in Advanced Science, in an article titled, “Phase Separation and Aggregation of α-Synuclein Diverge at Different Salt Conditions.”
“Previous work shows that factors promoting or inhibiting aggregation have similar effects on LLPS,” the article’s authors wrote. “By contrast, this study finds a monotonic salt dependence of LLPS due to intermolecular interactions.
“Furthermore, it observes time evolution of the two distinct assembly states, with macroscopic fibrillar-like bundles initially forming at medium salt concentration but subsequently converting into droplets after prolonged incubation. The droplet state is therefore capable of inhibiting aggregation or even dissolving aggregates through heterotypic interactions.”
About 15 years ago, researchers discovered that protein molecules could condense into droplets, isolated from the cell’s cytoplasm without an external membrane. And researchers soon realized that LLPS can be highly functional. It can allow cells to compartmentalize molecules and regulate biochemical reactions. LLPS helps organize DNA, protects and regulates mRNA, and—at the ends of microtubules—acts as a molecular glue to position the nucleus for cell division.
As well as their functional significance, protein droplets have been implicated in disease. For example, it has been thought that they can concentrate certain proteins, pushing them to the point of aggregation and causing diseases such as Alzheimer’s and Parkinson’s. This idea has been backed up by observations that certain conditions such as salt concentration or pH simultaneously promote protein aggregation and condensation. Yet whether—or, indeed how—the two processes are linked remains unproven.
The relationship between LLPS and aggregation was revisited by PSI scientists led by Jinghui Luo, PhD, a tenured scientist at PSI. To pin down the true link between aggregation and droplet formation, the researchers methodically investigated the behavior of ɑSyn proteins under a wide variety of conditions: protein concentration, salt concentration, and the presence of various concentrations of crowding agents that mimic the complex molecular environment of the cytoplasm. Each of these were studied at different pH values.
In total, the researchers studied more than 500 different conditions. For each condition, they followed the progression of droplet formation or aggregation for up to four months, taking regular images using light microscopy.
To study so many conditions, the researchers used the robotic crystallization facility at the Swiss Light Source SLS. The technique is typically used to prepare protein crystals for X-ray crystallography experiments.
“Being at the large-scale facilities and working alongside beamline scientists enabled us to approach this problem from a different angle,” explained PSI postdoctoral researcher Rebecca Sternke-Hoffmann, PhD, first author of the study.
“Interestingly, crystallographers have known that proteins could form droplets for a very long time,” Luo added. “It was just another thing they observed in their search for the perfect crystal.”
To complement this macroscopic story, the researchers used small angle X-ray scattering (SAXS) measurements at the Swiss Light Source SLS and conducted simulations to understand the microscopic picture.
The meticulous experiments revealed that the conditions that give rise to stable droplets or protein aggregation are not the same. Contrary to conventional theory that aggregates initiate from droplets, the researchers showed independent formation of droplets and aggregates in αSyn across diverse protein, salt, and crowding conditions.
Through their long timeframe of investigation, the researchers could see whether droplets really did evolve into aggregates. The answer: they didn’t, even after 120 days. Indeed, far from promoting aggregation into fibrils, remarkably droplets appeared to have the opposite effect. During extended incubations, previously considered irreversible fibrils transformed into droplets.
“This observation,” Luo stated, “points to a functional role of liquid droplets in preventing the formation of solid LLPS in the cell identifies it as a highly evolved feature associated with functionality, whereas aggregation, particularly in the case of αSyn, is associated with disease. From this perspective, it would be somewhat surprising if protein droplets were precursors of protein aggregation.”
Through SAXS measurements together with simulations and sequence analysis, the researchers could understand the differences they observed: aggregation occurs mostly due to interactions between the tails of individual protein molecules, while LLPS occurs due to interactions between different protein molecules.
The deeper understanding of the complex interplay between protein aggregation and liquid-liquid phase separation is relevant not only for Parkinson’s, but also other neuro-degenerative diseases characterized by protein aggregation, including Alzheimer’s, Huntingdon’s, and Creutzfeldt-Jakob disease. In turn, this could lead to new treatments.
