Under a transmission electron microscope, infectious prions appear as dense, tangled clusters of fibrils, thin threadlike structures roughly 10 to 20 nanometers wide that bundle together and resist the chemical treatments used to break down normal proteins. These fibrils were first spotted in the early 1980s in brain tissue from scrapie-infected animals and were so distinctive that researchers named them “scrapie-associated fibrils,” or SAF. Since then, the tools used to image prions have advanced dramatically, and the picture has sharpened from blurry fibril outlines to near-atomic-resolution maps that reveal how each misfolded protein molecule stacks against the next.
The First Glimpse Through Electron Microscopy
The earliest views of prions came from a technique called negative-stain transmission electron microscopy, where a heavy-metal stain is applied around biological structures so they stand out against a dark background. In the early 1980s, researchers examining brain extracts from scrapie-infected sheep and hamsters noticed fibrillar structures that did not appear in healthy brain tissue. These abnormal fibrils were dubbed scrapie-associated fibrils and became one of the first physical signatures of prion disease visible under a microscope.1PubMed. Abnormal fibrils from scrapie-infected brain
At this stage, the images were useful but coarse. Negative staining gives you a silhouette of a structure rather than a detailed portrait. Researchers could tell that the fibrils were rod-shaped and that they accumulated in infected brains, but they could not determine how individual prion protein molecules were arranged within them. A later study added an important detail: when brain tissue was treated with a detergent (sodium dodecyl sulfate), larger tube-shaped structures called tubulofilaments appeared to fragment, peeling away an outer coat to reveal thinner fibrils inside. Immunogold labeling confirmed these inner fibrils were SAF, suggesting that the fibrils sit inside a larger structure in living tissue and are not just artifacts created during sample preparation.2PubMed Central. Tubulofilaments in negatively stained scrapie-infected brains: relationship to scrapie-associated fibrils
That last point was contentious at the time. When you grind up brain tissue, dissolve membranes with chemicals, and spin the mixture at high speed to isolate a fraction, you can end up with structures that did not exist in the original tissue. The question of whether these fibrils were real biological features or lab artifacts dogged the field for years. Establishing that the tubulofilaments enclosed SAF within intact tissue helped settle that debate, but it also hinted at a level of structural complexity that would take decades to fully resolve.
The Normal Protein and Why Shape Matters
To appreciate what makes prion fibrils so unusual, it helps to know what the normal prion protein looks like. Every healthy mammalian brain contains millions of copies of the cellular prion protein, known as PrPC. In the early 2000s, researchers used nuclear magnetic resonance spectroscopy to map its three-dimensional shape in solution. They found a compact, globular domain made up of three helical segments and a short stretch of flat sheet-like structure, with a floppy, disordered tail dangling off the front end.3PubMed. NMR solution structure of the human prion protein This normal form is soluble, meaning it floats freely and does its job without clumping.
In prion disease, the protein refolds into a shape dominated by flat, stacked sheets rather than helices. This misfolded version is stickier. Molecules latch onto each other and build up into the fibrils visible under an electron microscope. The change is not a small tweak; it is a wholesale architectural rearrangement. And because the misfolded form can template its shape onto normal copies of the protein, forcing them to refold in the same way, the fibrils grow and spread through brain tissue over time.
Solid-state NMR played a crucial role in filling the gap between what early electron microscopy could show and what researchers needed to know about the internal arrangement of prion fibrils. Unlike solution NMR, which works on freely tumbling molecules, solid-state NMR can probe structures that are locked into rigid assemblies like fibrils. It provided the first atomic-level constraints on how individual protein molecules sit within an amyloid fibril, enabling researchers to build molecular models well before high-resolution imaging caught up.4PubMed Central. Solid-state NMR studies of amyloid fibril structure Comparative studies of different prion proteins in fibrillar form using solid-state NMR helped establish a general picture of how prion molecules stack and interact.5PubMed Central. Prions: En route from structural models to structures
How Cryo-EM Changed Everything
The real breakthrough in seeing what prions look like came with cryo-electron microscopy, a technique that flash-freezes biological samples in a thin layer of ice and images them with an electron beam. Unlike negative staining, cryo-EM preserves the sample in a near-native state, avoiding the distortions caused by drying and heavy-metal coating. And unlike NMR, which gives you distance constraints between atoms that must be assembled into a model computationally, cryo-EM produces a direct three-dimensional density map that shows where the protein is.
In 2016, a cryo-EM study of infectious prion fibrils from hamster brains revealed a fibril composed of two distinct protofilaments wound together. The repeating unit along the fibril axis measured about 19 angstroms, and each prion molecule sat within a volume of roughly 19,000 cubic angstroms, giving each molecule a height of about 17.7 angstroms. The overall architecture pointed to a four-rung structure called a beta-solenoid, where each protein molecule contributes multiple flat rungs that stack on top of one another like a spiral staircase.6PubMed Central. The Structural Architecture of an Infectious Mammalian Prion Using Electron Cryomicroscopy
By 2021, resolution had improved enough to build actual atomic models. A study of the 263K hamster prion strain achieved a resolution of 3.1 angstroms, sharp enough to trace the path of the protein backbone through each rung of the fibril. The images showed highly uniform rungs spaced 4.9 angstroms apart along the fibril axis, consistent with an architecture known as parallel in-register intermolecular beta-sheet, or PIRIBS, where each molecule contributes one rung and every rung is essentially identical.7Molecular Cell. Structure of an Infectious Mammalian Prion Picture a tall stack of identical flat discs, each one a single prion molecule lying flat and bonded to the disc above and below. That is roughly what a prion fibril looks like at the molecular level.
Shortly after, two independent teams published cryo-EM structures of mouse-adapted prion strains (called RML) at resolutions of 2.7 and 3.0 angstroms, confirming the PIRIBS architecture in a different host species.8PubMed Central. 2.7 Å cryo-EM structure of ex vivo RML prion fibrils9Nature Communications. Cryo-EM structure of anchorless RML prion reveals variations in shared motifs between distinct strains A review of all sufficiently resolved prion fibril structures to date found that every single one shares the same fundamental layout: a parallel in-register beta-sheet stack.10PubMed Central. Structural biology of ex vivo mammalian prions The basic visual motif of prion fibrils, at every resolution from the blurry SAF images of the 1980s to the atom-traced maps of the 2020s, is the same: long, straight or gently curved rods built from flat molecules stacked one atop another.
Different Strains, Different Shapes
One of the most striking findings from high-resolution imaging is that different prion strains, which cause different disease patterns in animals, produce fibrils with visibly distinct shapes. Prion strains are not defined by mutations in the gene; the same protein sequence can fold into multiple misfolded conformations, and each conformation breeds true as it templates new copies. Under the microscope, these conformational differences translate into structural differences you can see.
Comparing the hamster 263K strain to the mouse-adapted RML strain illustrates this well. A high-resolution study noted major differences in fibril shape between the two, including an asymmetric fibril cross-section in 263K with no paired protofilaments, a feature that differed from what had been assumed from earlier, lower-resolution work.11PubMed. High-resolution structure and strain comparison of infectious mammalian prions The differences are not subtle. Side-chain orientations of specific amino acids can flip between strains, loops in the protein backbone bend in opposite directions, and the overall flatness of each rung varies. A detailed comparison of two mouse prion strains (called aRML and a22L) showed that their internal arches bend in opposite directions, that the interface between two lobes of the protein is less staggered in one strain than the other, and that multiple pairs of amino acid side chains flip their orientations relative to the backbone.12PLOS Pathogens. Cryo-EM of prion strains from the same genotype of host identifies conformational determinants
These structural differences matter because they help explain why prion strains behave differently in living animals: they target different brain regions, cause different symptoms, and have different incubation periods. The protein sequence and the particular misfolded conformation of the seed are the primary factors that determine whether a prion can cross between species and how aggressively it replicates once it does.13PubMed Central. Contributions of the Prion Protein Sequence, Strain, and Environment to the Species Barrier In other words, what a prion looks like under the microscope is not just a curiosity; it is a direct readout of the information that governs its biological behavior.
Why Preparation Methods Can Fool You
A persistent headache in prion imaging is that sample preparation can create structures that look like disease-associated fibrils but are not. When researchers isolate prion protein from brain tissue, they typically use detergents to strip away cell membranes and other debris. But some of those detergents can force even normal, healthy prion protein into aggregated forms that look worryingly fibrillar under an electron microscope.
One study demonstrated this directly. When normal cellular prion protein was exposed to low concentrations of the detergent sarkosyl, which is routinely used in prion purification, it formed both amorphous clumps and fibrillar aggregates. Spectroscopy confirmed that these aggregates had a high proportion of beta-sheet structure, mimicking the signature of disease-associated prions. Yet they were not protease-resistant in the same way as true infectious prions, they had a different infrared spectrum, and they were not infectious.14PubMed. Conformational change, aggregation and fibril formation induced by detergent treatments of cellular prion protein The takeaway is stark: not every fibril you see under a microscope is a prion. High beta-sheet content and a fibrillar shape are necessary features of infectious prions, but they are not sufficient. The field learned the hard way that looking like a prion does not mean being one.
This is partly why so many complementary techniques are needed. Electron microscopy can show you the shape, but you need immunogold labeling to confirm the protein identity, protease digestion assays to check for the characteristic resistance of infectious prions, and increasingly, cryo-EM at high resolution to verify that the internal molecular architecture matches known prion structures rather than some other aggregated state.
Yeast Prions Under the Microscope
Prions are not exclusive to mammals. Yeast cells harbor their own prion-like proteins, the best studied of which is Sup35. When Sup35 adopts its prion form, it aggregates into structures that are visible under a fluorescence microscope as bright puncta, small bright dots inside the cell. But what do those dots look like at higher magnification?
Thin-section electron microscopy of yeast cells carrying the prion state revealed that the fluorescent aggregates contain bundled fibrillar structures, much like the fibrils seen when Sup35 is assembled into amyloid in a test tube.15PubMed Central. In vivo evidence for the fibrillar structures of Sup35 prions in yeast cells This was significant because it showed that the fibrils formed in the lab are not just an artifact of artificial conditions; the same general structures exist inside living cells. Researchers used rapid-freeze electron microscopy combined with immunogold labeling to confirm that the fibrils were genuinely made of Sup35, not some other cellular component caught up in the aggregate.
Yeast prion fibrils also show strain-dependent morphological variation, just like mammalian prions. When the prion domain of Sup35 (called NM) is assembled into fibrils under different conditions, the resulting fibrils confer either a “strong” or “weak” prion phenotype when introduced into yeast cells. Electron microscopy reveals that strong and weak fibrils are both mixtures of thin and thick fibrils, but in different ratios: strong prion fibrils tend to have more thin fibrils, while weak prion fibrils have more thick ones. The thickness differences come from how the M domain of the protein packs along the fibril surface.16PubMed Central. Amyloid fibrils embodying distinctive yeast prion phenotypes exhibit diverse morphologies So even in yeast, the way a prion looks under the microscope is tied to how it behaves biologically.
Following the Early Stages of Aggregation
Most prion imaging captures the end product: mature fibrils pulled from infected brains or grown in a dish. But researchers also want to understand the earliest stages of aggregation, when individual protein molecules first begin to cluster together into small assemblies called oligomers before elongating into full fibrils. These early intermediates are thought to be particularly toxic in many protein-misfolding diseases, and catching them under a microscope is technically challenging because they are small and transient.
One approach combines transmission electron microscopy with light-scattering measurements and fluorescence assays to track aggregate size and structure in real time. A study of the yeast prion protein Sup35NM used this combination to follow the initial aggregation process, finding that mechanical shocks could alter how the protein first clumped together and shifted the balance between different early aggregate forms.17PubMed. The effect of mechanical shocks on the initial aggregation behavior of yeast prion protein Sup35NM Under the electron microscope, these early-stage aggregates look like small amorphous blobs or short, immature fibril fragments rather than the long, well-ordered fibrils seen at later stages. Watching the transition from blob to fibril is one of the frontiers of prion imaging.
Seeing Prions in Context With Emerging Techniques
A limitation of all the methods discussed so far is that they look at prion fibrils in isolation, pulled out of the brain and purified. What researchers really want is to see prions as they exist inside cells, surrounded by membranes, organelles, and all the other machinery of a living neuron. Correlative light and electron microscopy, which overlays a fluorescence image and an electron microscopy image of the same sample, has begun to make this possible for protein deposits in brain tissue.
An even newer approach is cryo-electron tomography, or cryo-ET, which takes a series of cryo-EM images at different tilt angles and reconstructs them into a three-dimensional volume, like a CT scan for cells. Cryo-ET has recently been applied to prion-infected neurons, allowing researchers to visualize the nanoscale structural features associated with infection in three dimensions and in a near-native frozen state, without the chemical fixation and staining that can distort structures.18Proceedings of the Prion 2025 – Advancing the understanding and treatment of prion diseases. Cryo-ET reveals subcellular pathology in prion-infected neurons This is still early-stage work, but it promises to answer questions that purified fibril studies cannot: where exactly in the cell do prion fibrils form? How do they interact with membranes? What does the surrounding cellular landscape look like as infection progresses?
The gap between seeing an isolated fibril and understanding how that fibril wreaks havoc inside a neuron is enormous. Cryo-ET and similar in-situ techniques are the tools that may finally close it, offering not just pictures of what prions look like, but pictures of what prion disease looks like at the molecular scale as it unfolds in real tissue.
How Prion Fibrils Compare to Other Disease-Related Amyloids
If you put a prion fibril next to a tau fibril from Alzheimer’s disease or an alpha-synuclein fibril from Parkinson’s disease, a non-specialist peering through the microscope would struggle to tell them apart. All of them are long, unbranched filaments a few nanometers wide. At low resolution, amyloid fibrils from unrelated proteins look remarkably alike, which is one reason the field took so long to determine prion-specific structural features.
At atomic resolution, the picture diverges. Cryo-EM structures from patient-derived tau and alpha-synuclein fibrils have been published alongside structures from infectious prion fibrils, and the comparison reveals both shared principles and meaningful differences.19PubMed Central. The structural line between prion and “prion-like”: Insights from prion protein and tau All of them use stacked beta-sheet arrangements. But the fold of each protein within its fibril is unique, shaped by the amino acid sequence and the particular misfolding pathway it followed. Prion fibrils from infectious material also tend to be more structurally uniform along their length than many synthetic amyloid preparations, which can be a mix of structural variants. That uniformity is part of what makes true prions so efficient at replicating their shape: each new molecule that joins the fibril is forced into the same precise fold.
The practical difference is infectivity. Tau and alpha-synuclein aggregates can spread between cells and even between brain regions, which is why they are sometimes called “prion-like.” But only mammalian prion protein fibrils have been shown to be transmissible between organisms in the classical sense, causing fatal disease through environmental or dietary exposure. Whether the structural features visible under cryo-EM can explain that difference in transmissibility is one of the most active questions in the field.