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

By introducing proline residues to disrupt $\beta$-sheet formation in the FUS prion-like domain, researchers demonstrated that preventing aggregation and liquid-to-solid transitions can eliminate neurotoxicity in disease models while preserving the protein's essential phase separation and biological functions.

The Gist

The Big Picture: The "Sticky" Protein Problem

Imagine your body is a bustling city, and inside every cell, there are millions of tiny workers called proteins. One of these workers is named FUS.

In a healthy city, FUS is a very social worker. It loves to gather with other FUS workers and RNA (the instruction manuals) to form temporary, liquid-like "town squares" (called condensates). In these town squares, they chat, organize files, and get work done. This process is called phase separation. It's like a drop of oil mixing with vinegar to form a distinct blob, but in a way that allows things to flow freely inside.

The Problem:
Sometimes, due to stress or genetic mutations, these FUS workers get too excited. Instead of staying in a fluid, flowing town square, they start sticking together too hard. They turn from a liquid drop into a solid, rock-hard lump. This is called aggregation. In diseases like ALS (Lou Gehrig's disease) and Frontotemporal Dementia, these rock-hard lumps pile up in the brain, killing the cells.

Scientists have long debated two big questions:

1. The Structure Question: Does the "rock-hard" part (a specific structure called a beta-sheet) need to exist for the workers to form the town square in the first place? Or can they form the town square without turning into rocks?
2. The Toxicity Question: Is the toxicity (the killing of cells) caused because the workers are in the wrong place (the cytoplasm instead of the nucleus), or is it caused specifically because they turned into those hard, rock-like lumps?

The Experiment: The "Proline" Trick

The researchers in this paper decided to test these questions by playing a game of "molecular Lego."

They knew that FUS turns into a rock because of a specific way its backbone folds, creating a rigid structure called a beta-sheet. They wanted to see if they could break that structure without stopping the protein from doing its job.

The Analogy:
Imagine FUS is a long, flexible rope. To turn into a rock, the rope has to fold into a very specific, tight knot (the beta-sheet).
The scientists took this rope and inserted Proline residues. Think of Proline as a rigid, un-bendable hinge or a kink in the rope. You can't tie a tight knot if the rope has a stiff hinge in the middle of it.

They created a new version of the FUS protein (called FUS-12P) that had 12 of these "kinks" inserted into it.

What They Found: A Miracle Decoupling

Here is what happened when they tested this new "kinked" FUS:

1. The Town Squares Still Formed (Phase Separation is Safe)
Even with the kinks, the FUS-12P workers still gathered into liquid-like town squares. They flowed, merged, and moved just like the normal FUS.

• The Takeaway: You don't need the "rock-hard" beta-sheet structure to form the liquid town square. The workers can do their social gathering without turning into rocks.

2. The Rocks Disappeared (Aggregation is Stopped)
When they tried to force the FUS-12P to turn into rocks (by shaking them or waiting a long time), they simply couldn't. The kinks prevented the tight knots from forming. The protein stayed liquid and healthy.

• The Takeaway: The "kink" successfully stopped the protein from turning into toxic, solid aggregates.

3. The Work Still Got Done (Function is Preserved)
The scientists checked if the FUS-12P could still do its actual job: regulating genes, repairing DNA, and responding to stress.

• The Takeaway: Yes! The FUS-12P worked exactly as well as the normal FUS. It went to the right places in the cell and did the right things. The "kink" didn't break the worker; it just broke the ability to turn into a rock.

4. The Flies Were Saved (Toxicity is Gone)
This is the most exciting part. The scientists put these proteins into fruit flies (a common model for studying human disease).

• Normal FUS (or disease-mutant FUS): The flies got sick. Their eyes degenerated, they couldn't climb, and they died young.
• FUS-12P (The "Kinked" version): The flies were completely healthy. Even if the protein was stuck in the wrong part of the cell (which usually causes trouble), the flies didn't get sick.
• The Takeaway: The toxicity wasn't caused by the protein being in the wrong place; it was caused specifically by the protein turning into those solid, beta-sheet rocks. If you stop the rocks from forming, the protein is harmless, even if it's misplaced.

The "Aha!" Moment

The paper solves a major mystery in neurodegenerative disease:

• Old Theory: Maybe the liquid-to-solid transition is necessary for the protein to work, and we just have to live with the risk of it going wrong.
• New Finding: No! The protein can do its job perfectly fine as a liquid. The "solid rock" part is purely a disease mechanism. It's like a car engine that works fine, but if you let the oil turn into concrete, the car explodes. You don't need the concrete to make the car run; you just need to stop the oil from turning into concrete.

Why This Matters for the Future

This study suggests a brand-new way to treat diseases like ALS.
Instead of trying to kill the protein (which might stop it from doing its good work), we could potentially give patients a "super-version" of the protein (like the FUS-12P) that is immune to turning into rocks.

It's like giving the city a new type of worker who is just as good at their job but has a "force field" that prevents them from ever getting stuck in a traffic jam (aggregation). This could allow us to treat the disease without losing the essential functions of the protein.

In short: The researchers found a way to make the "bad" part of the protein (the rock formation) disappear while keeping the "good" part (the liquid function) alive. This proves that the toxicity of these diseases comes from the solid clumps, not just the protein being in the wrong place.

Technical Summary

1. Problem Statement

The RNA-binding protein Fused in Sarcoma (FUS) is essential for RNA processing and forms biomolecular condensates via liquid-liquid phase separation (LLPS). However, in neurodegenerative diseases like Amyotrophic Lateral Sclerosis (ALS) and Frontotemporal Lobar Degeneration (FTLD), FUS mislocalizes to the cytoplasm and undergoes a liquid-to-solid transition (LST), forming toxic, amyloid-like aggregates rich in cross-β-sheet structures.

Two critical questions remain unresolved:

1. Structural Dependency: Are the transient or labile β-sheet interactions (potentially forming structures like LARKS or steric zippers) required for the physiological function and phase separation of FUS, or are they solely pathological features driving aggregation?
2. Mechanism of Toxicity: Is FUS toxicity driven by the loss of nuclear function (mislocalization) alone, or is the formation of solid aggregates the primary driver of cell death?

Current methods to study this (e.g., domain deletion or tyrosine-to-serine mutations) often disrupt both phase separation and aggregation simultaneously, making it impossible to decouple these processes.

2. Methodology

The authors employed a structure-breaking design strategy to selectively disrupt β-sheet formation without altering the primary sequence grammar or disrupting phase separation drivers.

• Protein Engineering: They introduced proline residues (known to break backbone hydrogen bonding required for β-sheets) into the low-complexity (LC) prion-like domain of FUS.
  • Variants Created:
    • FUS 4P: Targeted a specific amyloid core.
    • FUS 12P: Targeted multiple predicted β-sheet regions throughout the LC domain.
    • FUS 16P: Combined approaches.
  • Contexts: These mutations were tested in the isolated LC domain, the LC-RGG1 fragment, and full-length FUS.
• Biophysical Characterization:
  • Phase Separation: Measured saturation concentrations ($C_{sat}$) under varying salt, temperature, and crowding conditions.
  • Aggregation Assays: Subjected samples to orbital shaking (1200 rpm) and monitored for Thioflavin T (ThT) fluorescence and morphological changes.
  • NMR Spectroscopy: Used $^{1}$H-$^{15}$N HSQC, relaxation measurements ($R_1, R_2$, NOE), and DOSY to assess secondary structure, backbone dynamics, and hydrodynamic radius in both dispersed and condensed phases.
  • Microrheology: Used video particle-tracking to measure the mechanical properties (viscoelasticity) and aging of condensates over 72 hours.
• Cellular Assays:
  • Localization: Genome-edited U2OS cells expressing eGFP-FUS (WT vs. 12P) were used to assess recruitment to paraspeckles and sites of DNA damage (laser microirradiation).
  • Function: RT-qPCR and Western blots were used to measure FUS auto-regulation and cross-regulation of EWS.
  • DNA Damage Response: Assessed recruitment of FUS and markers ($\gamma$H2AX, 53BP1) following etoposide or calicheamicin treatment.
• In Vivo Models:
  • Drosophila: Expressed WT, P525L (disease mutant), 12P, and P525L/12P variants in motor neurons and eyes.
  • Readouts: Eye degeneration, wing morphology, pupal eclosion rates, climbing assays (motor function), lifespan, and TUNEL staining for neuronal death.

3. Key Contributions

• Decoupling Phase Separation and Aggregation: The study provides the first direct evidence that the β-sheet forming propensity of the FUS LC domain is dispensable for phase separation but essential for aggregation.
• Structural Insight: It demonstrates that the FUS LC domain remains intrinsically disordered in both dispersed and condensed phases, challenging models that suggest stable β-sheet motifs are required to stabilize liquid condensates.
• Toxicity Mechanism: It establishes that FUS toxicity is driven by the formation of solid, β-sheet-rich aggregates rather than cytoplasmic mislocalization alone.
• Therapeutic Strategy: Proposes a novel therapeutic avenue: expressing an "aggregation-proof" but fully functional FUS variant to replace toxic endogenous protein, potentially avoiding the side effects of complete gene knockdown.

4. Key Results

A. Biophysical Properties

• Phase Separation Preserved: The 12P variant (and others) phase separated with kinetics and saturation concentrations nearly identical to Wild-Type (WT) FUS. The addition of prolines did not disrupt the liquid-like nature of the droplets.
• Aggregation Abolished: While WT FUS rapidly aggregated under shaking (forming ThT-positive fibrils), the 12P variant remained liquid-like for over 24 hours (and up to a week) with no ThT signal or solid morphology.
• Dynamics Unchanged: NMR relaxation and diffusion data showed that the 12P variant retained native-like global disorder and backbone dynamics. The slight reduction in diffusion coefficient was attributed to the increased chain length (12 extra residues), not rigidification.
• No Liquid-to-Solid Transition: Microrheology showed that WT condensates aged (became more viscoelastic/solid-like) over 72 hours, whereas 12P condensates maintained fluid properties, indicating that β-sheet cross-linking drives the LST.

B. Cellular Function

• Localization: The 12P variant recruited to nuclear paraspeckles and DNA damage sites with the same efficiency and dynamics as WT.
• Regulation: FUS 12P successfully regulated its own mRNA levels and those of its paralog EWS, indicating that transcriptional regulation does not require β-sheet formation.
• DNA Damage Response: Recruitment to laser-induced damage sites and subsequent phosphorylation by DNA-PK were unaffected in the 12P variant.

C. In Vivo Toxicity (Drosophila)

• Complete Rescue of Phenotypes: In flies expressing the toxic P525L mutation, the addition of the 12P modification completely rescued all toxic phenotypes:
  • Restored normal eye morphology (no degeneration).
  • Corrected wing development (no "curly-wing" phenotype).
  • Restored pupal eclosion rates to control levels.
  • Significantly improved motor function (climbing assays).
  • Extended lifespan to match non-expressing controls.
• Reduced Aggregation: Filter trap assays confirmed that 12P variants formed significantly fewer solid aggregates in fly brains compared to WT or P525L.
• Neuronal Survival: TUNEL staining revealed no neuronal death in flies expressing the 12P variant, whereas WT and P525L flies showed significant apoptosis.

5. Significance

This study fundamentally shifts the understanding of FUS pathology and phase separation biology:

1. Mechanistic Clarity: It proves that the "liquid" state of FUS condensates is stabilized by multivalent, disordered interactions, not by transient β-sheets. The transition to a toxic "solid" state is a distinct process driven by the accumulation of stable β-sheet cross-links.
2. Therapeutic Implications: The findings suggest that antisense oligonucleotide (ASO) therapies currently in clinical trials (which knock down FUS) might be improved or replaced by gene therapy approaches that express an aggregation-resistant variant (like FUS 12P). This would maintain essential FUS functions (preventing loss-of-function toxicity) while eliminating the gain-of-function toxicity caused by aggregation.
3. General Principle: The "proline insertion" strategy offers a generalizable tool to study other prion-like domain proteins (e.g., TDP-43) where aggregation and function are currently difficult to separate.

In conclusion, the paper demonstrates that breaking the backbone's ability to form β-sheets is a viable strategy to uncouple the essential physiological functions of FUS from its pathological aggregation, offering a promising path toward treating FUS-associated neurodegenerative diseases.