The Pick fold in tau filaments from human MAPT mutants

This study utilizes cryo-EM to reveal that specific MAPT mutations associated with Pick's disease and FTDP-17T generate distinct structural variants of the Pick fold in tau filaments, demonstrating that missense mutations within the filament core can modify this architecture and providing a basis for structure-based diagnostics and therapeutics.

The Gist

The Big Picture: The "Lego" Protein That Goes Wrong

Imagine your brain is a bustling city. Inside the neurons (the city's workers), there are tiny proteins called Tau. Under normal circumstances, Tau acts like construction scaffolding. It holds the city's roads (microtubules) steady so trucks can drive on them without crashing.

However, in certain diseases like Pick's disease and a type of dementia called FTDP-17T, these Tau proteins get confused. Instead of building roads, they stop working and start clumping together into hard, twisted knots called filaments. These knots clog up the brain cells, causing them to die.

This paper is like a team of high-tech detectives using a super-powerful microscope (called Cryo-EM) to take 3D snapshots of these knots. They wanted to see exactly how the knots are built when caused by specific genetic "typos" (mutations) in the Tau instructions.

The Main Discovery: The "Pick" Fold and Its Variants

The researchers looked at brain tissue from four different people who had specific genetic mutations in their Tau gene. They found that the resulting knots all looked very similar, but with some interesting twists.

Think of the standard knot as a two-story house with a specific blueprint. The scientists call this the "Pick Fold."

Here is what they found for each mutation:

1. The "Perfect Copy" Mutations (D252V and DG389-I392):

   • The Analogy: Imagine you have a Lego house blueprint. Two of the people had typos in the instructions, but when they built the house, it looked exactly like the standard Pick Fold blueprint.
   • What it means: Even though the instructions were slightly different, the final structure was identical to the classic "Pick's disease" knot. The mutation didn't change the shape of the house; it just added a tiny decoration or removed a non-essential piece.

2. The "Renovated" Mutations (G272V and S320F):

   • The Analogy: These two mutations were like a contractor who decided to rotate a whole wing of the house. The foundation and the main walls were still there, but one part of the structure was twisted about 20–25 degrees.
   • What it means: These mutations caused the "Pick Fold" to become a "Variant." It's still the same type of house, but the layout is slightly different. The researchers call this an "open" version of the fold. It's like a door that was supposed to be closed is now slightly ajar, changing how the whole building sits.

The "Seeded" Experiment: Can We Build It in a Lab?

One of the biggest challenges in studying these diseases is that you can't just grow these knots in a test tube easily. They are stubborn.

To solve this, the scientists tried a method called "Seeded Assembly."

• The Analogy: Imagine you have a pile of loose Lego bricks (recombinant Tau protein). They are just a mess. But if you drop in one tiny, pre-built piece of the knot (a "seed" taken from a patient's brain), the loose bricks will suddenly snap onto it and start building the exact same structure.
• The Result: They successfully used seeds from the patients to build these knots in a lab.
  • For the "Perfect Copy" mutations, they built the standard house.
  • For the "Renovated" mutations, they built the twisted version.
  • The Surprise: In one case (S320F), the lab-built version didn't twist like the patient's brain version did. It built the standard house instead. This tells the scientists that in the human brain, there might be other invisible helpers (like cofactors or chemical tags) that force the knot to twist. In the clean lab environment, those helpers were missing.

Why Does This Matter?

This research is like creating a detailed architectural catalog of brain diseases.

1. Better Diagnosis: Just as a detective can identify a criminal by their unique fingerprint, doctors might one day use these specific 3D shapes to diagnose exactly which type of dementia a patient has. If the "knot" looks like the "Renovated House," they know it's the G272V mutation.
2. New Treatments: Now that we know the exact shape of these knots, scientists can design "keys" (drugs) that fit perfectly into the cracks of the specific knot to break it apart or stop it from forming. You can't fix a lock if you don't know what the keyhole looks like.
3. Understanding the Rules: By seeing how small changes in the DNA (the typos) lead to big changes in the 3D shape, we learn the rules of how these proteins fold. This helps us understand why some diseases are more aggressive than others.

Summary

In short, this paper is a map of the "crime scenes" in the brain. It shows that while different genetic mutations cause the Tau protein to clump, they mostly build the same type of "Pick Fold" knot. However, some mutations act like a wrench in the gears, twisting the knot slightly. By understanding these tiny structural differences, we are one step closer to unlocking the secrets of dementia and finding a cure.

Technical Summary

1. Problem Statement

Tauopathies are neurodegenerative diseases characterized by the accumulation of misfolded tau protein into amyloid filaments. While cryo-electron microscopy (cryo-EM) has successfully determined the structures of tau filaments in sporadic Alzheimer's disease (AD), Chronic Traumatic Encephalopathy (CTE), and some inherited forms of frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17T), the structural basis of Pick's disease and specific MAPT mutations causing Pick-like pathology remained incompletely understood.

Specifically, it was unclear how different missense mutations within the MAPT gene (which encodes tau) influence the specific folding of tau filaments. While some mutations (e.g., P301L) produce unique folds, others (e.g., D252V, G272V, S320F, ΔG389-I392) are clinically associated with Pick's disease, but their atomic structures and whether they adopt the canonical "Pick fold" or distinct variants were unknown. Furthermore, there was a lack of robust in vitro models to reconstitute these specific disease-relevant folds for mechanistic study and therapeutic development.

2. Methodology

The study employed a multi-disciplinary approach combining neuropathology, structural biology, and biochemical reconstitution:

• Sample Collection: Post-mortem brain tissue (frontal/temporal cortex and caudate nucleus) was obtained from individuals with FTDP-17T carrying specific MAPT mutations: D252V, G272V, S320F, and ΔG389-I392.
• Filament Extraction: Sarkosyl-insoluble tau filaments were extracted from the brain tissues.
• Cryo-EM Structure Determination: High-resolution cryo-electron microscopy was used to determine the atomic structures of the extracted filaments. Data processing involved 2D/3D classification, helical reconstruction, and refinement using RELION. Resolutions ranged from 2.0 Å to 2.9 Å.
• Biochemical Validation:
  • Western Blotting: To confirm the presence of 3R tau isoforms and specific molecular weight bands.
  • Mass Spectrometry (MS): To verify the co-assembly of wild-type and mutant tau proteins within the filaments.
  • Immunohistochemistry: To characterize the pathological inclusions (Pick bodies) in the brain tissue.
• In Vitro Seeded Assembly: Recombinant 0N3R tau with 12 phosphomimetic mutations (PAD12) was engineered with specific MAPT mutations (D252V, G272V, S320F, ΔG389-I392). These were seeded with brain-derived filaments to test if the specific folds could be reconstituted in vitro.

3. Key Contributions

• Structural Classification of FTDP-17T: The authors established a structure-based classification for FTDP-17T cases, linking specific mutations to distinct tau filament folds.
• Discovery of Pick Fold Variants: They demonstrated that while some mutations preserve the canonical Pick fold, others generate structurally distinct but closely related "open" variants of the Pick fold.
• Mechanistic Insight into Mutation Effects: The study elucidated how specific amino acid substitutions (e.g., G272V, S320F) induce local conformational changes that propagate to alter the global architecture of the filament.
• Reconstitution of Disease Folds: The paper successfully reconstituted the Pick fold and its variants in vitro using recombinant tau, providing a platform for future drug discovery and mechanistic studies.

4. Key Results

A. Structures from Human Brain Tissue

1. D252V and ΔG389-I392 Mutations:

   • Filaments from these cases adopted the canonical Pick fold (two-layered, single protofilament).
   • Structural superposition with the known Pick fold (PDB: 6GX5) showed very low RMSD values (0.695 Å for D252V; 0.389 Å for ΔG389-I392), indicating structural identity.
   • Mass spectrometry confirmed that filaments contain a mixture of wild-type and mutant tau.
   • The ΔG389-I392 deletion lies outside the ordered core, explaining why the fold remains canonical.

2. G272V Mutation:

   • Filaments adopted a more open variant of the Pick fold.
   • The mutation (V272) is located in the 269QPGGG273 motif, which links the "stem" and "long arm" of the fold.
   • The hydrophobic side chain of V272 inserts into a pocket formed by V306, I308, and V339, causing a ~25° rotation of the long arm relative to the stem.
   • Side chains of Q269 and P270 switched their in-out orientations compared to the canonical fold.

3. S320F Mutation:

   • Filaments also adopted an open Pick fold variant, structurally similar to the G272V variant but with distinct local features.
   • The long arm was rotated by ~20° relative to the canonical Pick fold.
   • The bulky phenylalanine side chain at position 320 inserts into a hydrophobic pocket (V318, L325, I328).
   • Interestingly, a non-proteinaceous density was observed near G272 in the S320F structure, suggesting a cofactor might stabilize this conformation.

B. In Vitro Seeded Assembly

• D252V and G272V: Seeding recombinant PAD12 tau with brain seeds successfully generated filaments with the Pick fold (D252V) and the open variant (G272V). However, the in vitro G272V variant showed slight conformational differences in the linking motif compared to the brain-derived structure.
• S320F: Seeding produced four distinct filament types:
  • Two types (singlet and doublet) adopted the canonical Pick fold (surprisingly, the open variant seen in the brain was not the dominant form in vitro).
  • Two other types adopted a novel fold (Type 2) with a core spanning V306-F378, distinct from both Pick and Alzheimer folds.
• ΔG389-I392: Seeding this mutant did not yield Pick filaments; instead, it formed Paired Helical Filaments (PHFs) with the Alzheimer fold, likely due to contamination with AD seeds in the brain extract.

C. Structural Comparisons

• The study utilized an "Amyloid Packing Difference" (APD) metric. The APD between the canonical Pick fold and the G272V variant was 12%, and 5% for the S320F variant, confirming they are closely related structural variants rather than entirely new folds.
• The study confirmed that the Pick fold is diagnostic of Pick's disease (both sporadic and inherited), whereas mutations like P301L/P301T produce unique folds not seen in sporadic cases.

5. Significance

• Diagnostic Precision: The findings provide a structural basis for distinguishing between different forms of FTDP-17T. The presence of the Pick fold (canonical or variant) is a hallmark of Pick's disease pathology, regardless of whether it is sporadic or inherited.
• Therapeutic Development: The ability to reconstitute the Pick fold and its variants in vitro using recombinant tau is a major breakthrough. Previously, only Alzheimer and CTE folds could be reliably reconstituted. This enables the screening of structure-specific binders (antibodies or small molecules) for Pick's disease.
• Mechanistic Understanding: The study reveals that missense mutations can act as "structural switches." Even mutations outside the core (like ΔG389-I392) or within the core (like G272V) can dictate the specific amyloid fold, influencing disease phenotype.
• Future Directions: The identification of cofactors (non-proteinaceous densities) that may stabilize specific fold variants suggests that the cellular environment plays a crucial role in determining tau filament structure, opening new avenues for investigating the etiology of tau aggregation.

In summary, this paper resolves the atomic structures of tau filaments in four distinct MAPT mutation cases, defines the structural landscape of the Pick fold, and establishes a critical in vitro model system for studying and treating Pick's disease.