The Generative Brain

What is a prion?: research

Plain explanation

A prion is a multichain fibrillar assembly of misfolded host prion protein that propagates by seeded protein polymerisation. Near-atomic cryo-EM structures of infectious ex vivo fibrils from the hamster 263K and mouse RML strains show a comparable parallel in-register intermolecular beta-sheet architecture, which is the current structural basis for explaining strain diversity.

How prion disease develops in the brain
  1. Normal proteinEveryone makes prion protein. It sits on the surface of cells.
  2. A wrong foldSome of the protein folds into the wrong shape.
  3. It spreads its shapeThe misfolded form causes more protein to misfold.
  4. Build-upMisfolded protein gathers and is hard to clear.
  5. Nerve cells are damagedThis causes the symptoms of the illness.

The substrate is a normal host protein. PrP is a glycosylphosphatidylinositol-anchored cell surface glycoprotein with two asparagine-linked glycosylation sites. Prions are infectious multichain fibrillar assemblies of that same protein in a misfolded conformation, so the pathogen carries no nucleic acid genome of its own and its properties have to be held in protein conformation and assembly state (Manka and colleagues, Cell and Tissue Research 2023).

How does replication work?

By seeded protein polymerisation. PrP monomers are recruited to existing fibrillar assemblies, which elongate and then fragment, generating more templates. Fragmentation is part of the mechanism rather than a by-product, because an elongating fibril without breakage would increase mass without increasing the number of seeds.

Bellini and colleagues (Viruses 2026) describe the same process from the biochemical side: misfolded PrPSc acts as a template that converts PrPC in a self-propagating reaction, and the product is rich in beta-sheet structure, protease-resistant and aggregation-prone. They note that propagation without nucleic acids is what gives these conditions a dual identity as infectious diseases and as protein-misfolding diseases.

What have cryo-EM structures settled?

High-resolution definition of ex vivo prion fibrils was long considered unlikely. Advances in cryo-electron microscopy and in computational reconstruction of amyloids changed that, and near-atomic resolution structures have been reported for highly infectious ex vivo fibrils from the hamster 263K and mouse RML strains. Both have a comparable parallel in-register intermolecular beta-sheet architecture. Manka and colleagues present that shared architecture as the structural foundation for understanding prion strain diversity in mammals, meaning that strains are expected to differ as variations within this architecture rather than as separate folds.

Two caveats follow from the material itself. Both solved structures come from rodent-adapted laboratory strains, not from human disease tissue. And the fibrils are assemblies of a glycosylated, anchored protein, so glycoform composition and assembly state are part of what distinguishes strains, alongside the polypeptide conformation.

What remains unresolved?

Manka and colleagues state three gaps directly. The mechanism of prion replication is not precisely delineated. The molecular basis of prion strains and of inter-species transmission barriers lacks a detailed explanation. And the neurotoxic PrP species has no structural definition, which separates the question of what propagates from the question of what kills neurons. That third gap is the subject of neurotoxicity.

Current research programmes are organised around the same divisions. The PRIONS team at the IHAP unit in Toulouse, led by Olivier Andreoletti with Tomas Barrio as co-leader and nine listed members, works on prion strain plasticity, transmission within and between species, and the pathogenesis of transmissible spongiform encephalopathies.

What do amplification assays reveal about conversion?

Seeded conversion assays used for diagnosis double as a readout of templating specificity, because they fail in informative ways. Myskiw and colleagues (Annals of Neurology 2026) tested archived CSF against three formats differing mainly in recombinant PrP substrate. With truncated hamster substrate, 0 of 13 fatal familial insomnia samples and 4 of 10 P102L samples seeded a reaction. With deer mouse substrate, 12 of 13 and 10 of 10 did. One V180I sample behaved in the opposite direction, positive with truncated hamster substrate and negative with both other formats.

Schmitz and colleagues (Brain 2022) found the same pattern in a clinical cohort of 302 genetic prion disease cases across 36 PRNP variants: RT-QuIC sensitivity at standard cut-offs was 93 percent in 112 E200K cases and 87 percent in 47 V210I cases, against 43 percent in 14 P102L cases and 28 percent in 68 fatal familial insomnia cases. In octapeptide repeat insertion cases, every biomarker including RT-QuIC fell as repeat number rose, with Kendall’s tau of minus 0.2810 for RT-QuIC, P = 0.0486.

A mechanistic reading of these results is that the sequence of the substrate and the conformer of the seed have to be compatible for templated conversion to proceed efficiently. Both studies are retrospective and their per-variant groups are small, so the direction of the effect is better supported than its size.

Mechanisms hub, overview of prion diseases, PrP function and the plain-language what is a prion.

Sources for this page