Contagious Shape: Prions, the Protein-Only Hypothesis, and the Disease Without Genes

The Hook: The Agent You Could Not Kill

In the 1960s, the radiobiologist Tikvah Alper at London's Hammersmith Hospital faced a puzzle that refused to fit the textbook. She was studying the agent of scrapie, a sheep disease known since the eighteenth century that slowly turns the animals' brains into a spongy ruin. Alper did what one routinely did with infectious agents: she bombarded infected tissue with ultraviolet and ionizing radiation. The logic was molecular-biology bedrock. Every known agent—virus, bacterium, or fungus—carries its genetic material as DNA or RNA, and nucleic acids are exquisitely sensitive to radiation. Destroy the genome and the agent loses its ability to reproduce. It is dead.

The scrapie agent, however, would not die. Doses of UV and X-rays that would reliably neutralize any virus simply bounced off it. It stayed infectious. It also survived formaldehyde, boiling water, and treatments that shred nucleic acids. In 1967, Alper drew the only logical but outrageous conclusion: if the agent was indifferent to everything that attacks nucleic acids, then perhaps it had none.

This was heresy. The central dogma of molecular biology, formulated by Francis Crick just years earlier, held that information flows from nucleic acid to protein—never the reverse, and never without a nucleic acid as its store. An agent that reproduces without possessing a gene would be something like a clockwork that ticks without a spring. That same year the mathematician John Stanley Griffith ventured an even bolder guess: perhaps the agent was a pure protein that multiplied by forcing a normal protein to adopt its own shape.

It took fifteen more years for that idea to acquire a name and a stubborn champion. In 1982, the American neurologist Stanley B. Prusiner coined the term prion, contracted from "proteinaceous infectious particle"—an infectious particle made of protein. He claimed this particle was "entirely devoid of nucleic acids," free of both DNA and RNA. The field responded with ridicule and open hostility. Fifteen years later, in 1997, Prusiner received the Nobel Prize in Physiology or Medicine for exactly that idea. In between lies one of the most instructive stories in modern biology: the story of how a shape can become contagious.

The Core Concept: Same Sequence, Different Shape

To grasp what a prion is, you first have to shed a deeply rooted intuition. We are used to the idea that a protein's properties are determined by its amino-acid sequence—the order of the building blocks, which is in turn encoded in a gene. Two proteins with different sequences are different proteins. That is correct. But prions force us to take a second, subtler layer seriously: one and the same sequence can exist in more than one stable three-dimensional shape—and those shapes can have utterly different properties.

At the center stands a perfectly ordinary, native protein: the cellular prion protein, abbreviated PrP^C (the C is for "cellular"). It is encoded by the gene PRNP, tethered by a fatty anchor to the outside of our cell membranes, and found in abundance on nerve cells. Its healthy form is built mostly from α-helices—those corkscrew spirals that are among the most common folding motifs of proteins. In this shape, PrP^C is soluble, mobile, and readily broken down and recycled by cellular enzymes such as protease K. It is, in short, an unremarkable citizen of the cell surface.

The disease-causing form carries the same amino-acid sequence—identical letter for letter—but is folded differently. It is called PrP^Sc (the Sc is for "scrapie," the sheep disease that gave it its name). In it, large portions of the α-helices have flipped over into β-sheets, the flat, extended strands that stack easily into larger assemblies. This refolding changes everything. PrP^Sc is no longer soluble; it clumps into insoluble aggregates and fibrous amyloids. It resists protease K, which digests its healthy twin with ease. And—this is the decisive point—it is infectious in the most literal biochemical sense.

The mechanism postulated by the protein-only hypothesis is disturbingly simple. When a molecule of PrP^Sc meets a normal PrP^C, it acts like a mold. It binds to the healthy protein and forces it to adopt its own β-sheet-rich shape. One PrP^Sc becomes two; two become four. The new form is itself a mold. It is a chain reaction of refolding, a self-amplifying avalanche that requires not a single letter of genetic material. The information passed on does not reside in a sequence of bases but in a fold alone. You are not copying a message—you are copying a deformation.

This is the conceptual core over which everything ignited back then. Inheritance without genes, multiplication without copying a nucleic acid, information stored in geometry rather than chemistry. Anyone who wants to understand why a sequence can adopt more than one shape at all, and why that normally does not happen, would do well to look at the fundamental problem of protein folding (see The Shape of Life: Levinthal's Paradox, Chaperones, and How Proteins Find Their Form).

Part 1: The Historical Trail—Sheep, Cannibals, and a Heretic

The diseases were known long before their explanation. Scrapie in sheep has been documented since at least the eighteenth century; the name comes from the fact that afflicted animals scrape themselves so violently against fences that they rub off their wool. In the 1920s, the German neurologists Hans Gerhard Creutzfeldt and Alfons Maria Jakob independently described a rare, fatal brain disease of humans that would later carry their names: Creutzfeldt-Jakob disease (CJD). No one at the time suspected any link between the sheep plague and the human affliction.

The bridge was built by an event at the other end of the world. In the 1950s, the American physician Daniel Carleton Gajdusek studied a mysterious disease among the Fore people of the New Guinea highlands. They called it kuru, which in their language means "trembling." Sufferers lost control of their movements, endured uncontrollable laughter, wasted away, and died within months. Strikingly, it was mainly women and children who fell ill, men only rarely. The solution lay in a custom: the Fore honored their dead by consuming their corpses in mourning rituals. The women and children preferentially ate the brain—the tissue with the highest load of the agent. Kuru was a transmissible disease, passed on through the consumption of infected nervous tissue.

The decisive clue came from the veterinary pathologist William Hadlow. He had seen pictures of kuru-affected brains at a London exhibition and instantly recognized their startling resemblance to what he knew from scrapie: the same spongy perforation of the tissue, the same destruction without the usual signs of inflammation. In a 1959 letter to The Lancet, he proposed that kuru, like scrapie, might be caused by a transmissible agent—and that one should try to transmit it to animals. Gajdusek took up the suggestion and began transmission experiments in chimpanzees in 1960. After long incubation periods, the animals did indeed fall ill. The proof was in: kuru was infectious. Gajdusek received the Nobel Prize for it in 1976. People spoke of "slow viruses," because an agent without a nucleic acid was still unimaginable.

Into this landscape burst Stanley Prusiner. As a young physician, he had watched a patient die of CJD and became fascinated by an agent that seemingly defied all the rules. Over years, he enriched infectious hamster tissue to pin the agent down. What he isolated was not a virus but a protein. When he published his results in 1982 and coined the term "prion," a storm broke. Colleagues accused him of violating the central dogma; they suspected a hidden, not-yet-found nucleic acid; some considered the whole idea a career ploy. Prusiner himself later described how isolated he felt in that period. But the evidence accumulated: the PRNP gene encoding the normal protein was found; it was shown that animals lacking this gene could not develop scrapie at all (because there was no substrate to refold); the characteristic refolding was demonstrated. In 1997, the Nobel committee awarded Prusiner the prize "for his discovery of prions—a new biological principle of infection."

It is worth noting that this Nobel story has its shadows, too. Gajdusek was convicted of child abuse years after his prize. And the history of science still debates whether the groundwork of others—Alper, Griffith, Hadlow—was adequately honored in the awards. Science is a human enterprise, and its hero stories are rarely as clean as they appear in hindsight.

Part 2: The Molecule—Two Shapes of One Gene

Let us return to the biochemistry, for this is where the real beauty of the matter lies. The gene PRNP sits on chromosome 20 in humans and encodes a short protein of roughly 250 amino acids. After its manufacture, it is transported to the outside of the cell membrane and held there by a so-called GPI anchor (a glycolipid attachment). It is especially abundant on nerve cells. What PrP^C actually does in its healthy state is still not fully settled—proposed roles include copper binding, protection against oxidative stress, and signaling between cells. I am of the opinion that this open question about the normal function is too often underemphasized: we understand the disease better than the healthy protein it betrays.

The healthy fold of PrP^C is well mapped. Its structured part consists of three α-helices and a short, two-stranded β-sheet; a long front section remains mobile and disordered. This protein is soluble and is continually built up and broken down in normal cellular metabolism.

The pathological form PrP^Sc arises from precisely the same molecule through nothing more than a change of fold. The proportion of β-sheets rises dramatically, the proportion of α-helices falls. This seemingly small geometric shift has drastic consequences, best captured in a side-by-side comparison:

The last row is the linchpin. PrP^Sc does not "multiply" by building fresh copies of itself—a gene would do that. It multiplies by seizing existing healthy PrP^C and forcing it into its own shape. Two models are commonly used to describe this. In the template-directed refolding model, a single PrP^Sc molecule acts like a stencil and imprints its shape onto a single PrP^C. In the seeded (nucleation-dependent) polymerization model, the true mold is a small crystallization seed made of several molecules, onto which further monomers attach and refold; when the growing aggregate breaks into pieces, new seeds arise and the spread accelerates exponentially. Today's evidence suggests that real prion propagation bears features of both models.

As elegant as this sounds, the refolding also explains why these diseases are so cruel. The resulting amyloid aggregates can no longer be broken down by the cell. They accumulate, disturb and kill nerve cells, and leave behind those spongy holes that gave the whole family of diseases its name: transmissible spongiform encephalopathies, transmissible spongy brain diseases.

Part 3: The Proof and Its Resistance

A thesis this radical demands rigorous proof, and the field was right to demand it. The sharpest objection ran: perhaps the prion preparations do contain a tiny, well-hidden nucleic acid after all—a virus one simply has not found yet. As long as even the theoretical possibility remained, the protein-only hypothesis was not proven, only suggested.

The gold standard of proof was therefore called synthetic prions. If you assemble a prion protein in the test tube from purely recombinant building blocks made in bacteria—blocks guaranteed to contain no animal nucleic acid—drive it into the pathological fold, and thereby infect a healthy animal, then there is no room left for a hidden virus. Precisely this demonstration succeeded in the 2000s. Fibrils made from recombinant prion protein triggered genuine, transmissible prion disease in mice. In parallel, Claudio Soto and others developed Protein Misfolding Cyclic Amplification (PMCA), a method that mimics prion propagation in the test tube: you add a tiny amount of PrP^Sc to an excess of healthy PrP^C, let it refold, break up the aggregates with ultrasound into new seeds, and repeat the cycle. The result is a chain reaction, conceptually akin to PCR—except that here it is not DNA but a fold that is amplified.

With that, the basic thesis was secured—but the story became more complicated, not simpler. It turned out that pure recombinant protein alone often struggles to form fully infectious prions. In many systems, cofactors are needed: certain lipids, negatively charged molecules such as glycosaminoglycans, or even small amounts of nucleic acid that act as a scaffold to ease the correct fold. This is not a refutation of the protein-only hypothesis—the information about the disease type still resides in the protein, not in these helpers—but it is a refinement. I am of the opinion that the most honest present-day formulation is this: the protein carries and transmits the disease-causing information, yet in the living cell it rarely folds entirely on its own.

The most impressive support for the protein-only idea comes from a phenomenon that would be simply inexplicable without it: prion strains. From one and the same PRNP gene, in one and the same animal, there exist different prion variants that reliably behave differently—different incubation times, different affected brain regions, different biochemical signatures. In classical genetics, one explains such strain differences through sequence differences in the genome. Here, though, the sequence is identical. The only place the strain information can reside is in the fold itself: different strains are different, each stably propagated, three-dimensional conformations of the same molecule. A shape that replicates itself and thereby produces variants with stable inheritance—that is almost a caricature of life, built from a single protein.

Part 4: The Human Diseases—Four Roads into the Same Doom

In humans, prion disease appears in several forms that can be organized by their origin. The most common is Creutzfeldt-Jakob disease, and one distinguishes four roads by which it arises.

Sporadic is by far the most frequent form, at around 85 percent of cases. It occurs without any recognizable external cause and without family history, usually around the age of 60. It is thought that in rare cases a single PrP^C molecule spontaneously misfolds and triggers the avalanche—a biochemical accident that becomes more likely the longer a brain has been producing proteins. Sporadic CJD is rare (about one to two cases per million people per year) but evenly distributed worldwide.

Familial or genetically caused forms account for roughly 10 to 15 percent of cases. Here the affected person carries a mutation in the PRNP gene that makes the prion protein less stable from birth and renders spontaneous misfolding more likely. This group includes, besides familial CJD, Gerstmann-Sträussler-Scheinker syndrome (GSS) and Fatal Familial Insomnia (FFI)—a particularly eerie disease in which sufferers progressively lose the ability to sleep and die of it within months to years. That a single point mutation in the genome should decide the folding fate of a protein connects the prion world directly to classical genetics (on rewriting genes, see The Programmable Scissors: CRISPR and the Rewriting of Life).

Iatrogenic—transmitted by medical intervention—is a small but tragic group. Prions are extraordinarily resistant to the usual sterilization procedures, the very property that once astonished Alper. This has led to transmissions via contaminated neurosurgical instruments, cornea and dura mater grafts, and, most consequentially, via growth hormone that used to be extracted from the pituitary glands of the deceased. These cases taught medicine that ordinary autoclaving is not enough against prions.

Variant, finally, is the form that made headlines in the 1990s and opened the fourth road: transmission from another species.

A key to understanding all these forms is a tiny spot in the PRNP gene: codon 129. At this position, the protein can naturally carry either the amino acid methionine (M) or valine (V). Every person inherits two copies, so is MM, MV, or VV. This seemingly harmless letter swap strongly influences how susceptible someone is to particular prion diseases and how they progress. In variant CJD, the finding is downright dramatic: nearly all neuropathologically confirmed cases occurred in people with the MM genotype. The fold of a protein and a single letter in the gene together decide life and death.

Part 5: Mad Cow and the Species Barrier

Variant CJD is inseparable from one of the greatest food crises of the twentieth century: bovine spongiform encephalopathy (BSE), popularly "mad cow disease." In the 1980s, a prion disease spread through British cattle herds. Its cause lay in feeding practice: slaughterhouse waste, including nervous tissue, was processed into meat-and-bone meal and fed to cattle—herbivores unwittingly given infectious tissue of their own kind. The prions survived processing and spread through the feed chain.

The decisive and long-disputed question was whether these cattle prions could also afflict humans. Here a second central concept comes into play: the species barrier. Prions do not transmit arbitrarily between species. The more dissimilar the prion-protein sequences of two species are, the harder it is for the agent protein of one species to refold the host protein of the other. This barrier is not an impassable wall but a resistance—sometimes nearly impenetrable, sometimes merely a delayer. In the case of BSE, the barrier to humans proved low enough to be breached.

From 1996 onward, the first cases of a novel CJD appeared in Britain that differed from the sporadic type: the sufferers were strikingly young, often under 30, and the course of the disease deviated. Painstaking studies showed that variant CJD is caused by the same prion as BSE. Humans had become infected by eating contaminated beef. Overall, the death toll remained—at a little over two hundred worldwide, most of them in Britain—far below the initial catastrophe scenarios; the species barrier and the long incubation periods prevented worse. But the crisis permanently changed food safety: feed bans, removal of at-risk material, surveillance programs. It also drove home, drastically, that the fold of a single protein can shake an entire national economy.

Part 6: The Grand Generalization—Prion-Like Spread Everywhere

Perhaps the most consequential turn of this story is the most recent. For what if the prion-like mechanism—the template-directed refolding and protein-to-protein spread—is not confined to the classical prion diseases? That is exactly what the research of the past two decades suggests.

The major neurodegenerative diseases almost always revolve around misfolded, aggregating proteins: amyloid-β and tau in Alzheimer's disease, α-synuclein in Parkinson's disease, TDP-43 in certain forms of dementia and amyotrophic lateral sclerosis. For years, the aggregation of these proteins was regarded as a kind of passive garbage deposition. Today the evidence mounts that it spreads actively according to a prion-like principle: a misfolded molecule acts as a seed that forces neighboring healthy molecules into the same faulty form; these aggregates travel from nerve cell to nerve cell and carry the misfolding step by step through the brain. This explains why the pathological deposits in Alzheimer's and Parkinson's spread in such a characteristic spatial pattern.

Two important clarifications are needed here, and I am of the opinion that they cannot be stressed often enough. First: "prion-like" does not mean "contagious." There is no solid evidence that Alzheimer's or Parkinson's transmits from person to person like an infectious disease. The mechanism of spread within one brain resembles that of prions; transmissibility between organisms does not. Second, much of this remains active and partly contested research—compelling cell-culture and animal experiments, but less direct proof in humans. Even so, this perspective has changed thinking about neurodegeneration and opened new therapeutic ideas: if one could interrupt the spread of the seeds, one might halt the progression.

Even more surprising is an insight that elevates prions from mere pathogens to a general biological principle. In the 1990s, the biologist Reed Wickner showed that baker's yeast possesses functional prions. The best known, [PSI+], is a self-propagating amyloid form of the protein Sup35, which normally terminates protein production at the correct points. When Sup35 refolds into the prion form, it changes the yeast's gene expression—and this change is inherited via the fold to daughter cells, entirely without any change to the DNA. Here the prion is not a killer but an epigenetic switch, a heritable trait stored in the fold of a protein. There are even hints of physiologically useful prions in animals, for instance a prion-like protein (CPEB) that may play a role in the consolidation of long-term memories.

With that, a circle of thought closes. What began in 1967 as unthinkable heresy—information stored in the shape of a protein and passed on through that shape—turns out to be a basic pattern of biology that sometimes kills, sometimes inherits, and sometimes remembers. The language of life knows not only the four letters of the nucleic acids (whose translation into proteins is described elsewhere, see The Machine Made of RNA: The Ribosome, the Ribozyme, and the Translation of the Genetic Code). It also knows a second, geometric script: the fold.

Frameworks: Two Ordering Grids

For orientation, I summarize the two central classifications. First, the four roads by which Creutzfeldt-Jakob disease arises:

And the chain of core concepts on which the whole argument hangs: one gene (PRNP), two forms (PrP^C and PrP^Sc), one mechanism (template-directed refolding), no nucleic acid, many strains (encoded in the fold), one barrier (between species), and one generalization (prion-like spread in Alzheimer's, Parkinson's, and in yeast).

The Central Takeaway

The real lesson of prions is not about a rare disease but about the nature of information itself. We are trained to equate biological information with sequence—with the order of bases in DNA, the letters in a gene. Prions show that information can also reside in structure: one and the same sequence can adopt more than one stable shape, and that shape can reproduce itself, pass itself on, and differentiate into variants. The fold is a second layer of inheritance, existing alongside the sequence.

The practical prompt I draw from this story is a thinking tool: when a system behaves unexpectedly, it is worth asking whether the problem is not merely bad information but a bad state that perpetuates itself. Prusiner's opponents spent years searching for a hidden gene because they were convinced that any reproduction required a sequence store. They searched in the wrong place because they mistook a basic assumption for a law of nature. Anyone debugging a self-reinforcing error state in a software architecture, observing a spread without a pathogen in biology, or examining a dysfunctional culture in an organization that carries itself from person to person, should keep prions in mind: sometimes it is not the content that is contagious, but the form.

A Question to Reflect On

If a mere fold—without a single gene—can store information, multiply itself, and pass itself on in stable variants, then where exactly does the line run between "alive" and "not alive"? And would we even recognize a self-replicating pattern that possesses neither metabolism nor genome if we met it—or would we, as we once did with the scrapie agent, simply deem it impossible?

Cross-References in the Vault

• The Shape of Life: Levinthal's Paradox, Chaperones, and How Proteins Find Their Form – why a protein finds a shape at all and why misfolding is possible.
• The Machine Made of RNA: The Ribosome, the Ribozyme, and the Translation of the Genetic Code – how the sequence information of genes is translated into proteins in the first place.
• The Programmable Scissors: CRISPR and the Rewriting of Life – how mutations in the genome, such as in the PRNP gene, arise and can be changed.
• The Countdown in the Nucleus: Telomeres, Telomerase, and the Clock of Aging – another molecular view of aging and neurodegenerative processes.

Sources

• Nobel Prize in Physiology or Medicine 1997 – Press release (Stanley B. Prusiner), NobelPrize.org: https://www.nobelprize.org/prizes/medicine/1997/press-release/
• Zabel, M. D. & Reid, C.: A brief history of prions, Pathogens and Disease / PMC (2015): https://pmc.ncbi.nlm.nih.gov/articles/PMC4626585/
• Liberski, P. P. et al.: Kuru, the First Human Prion Disease, Viruses 11(3):232 / PMC (2019): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6466359/
• Conformational conversion of prion protein in prion diseases, Acta Biochimica et Biophysica Sinica 45(6):465 (2013): https://academic.oup.com/abbs/article/45/6/465/1298
• Goedert, M. et al.: Like prions: the propagation of aggregated tau and α-synuclein in neurodegeneration, Brain 140(2):266 (2017): https://academic.oup.com/brain/article/140/2/266/2669395
• Wickner, R. B. et al.: Yeast and Fungal Prions, Cold Spring Harbor Perspectives in Biology 8(9):a023531 (2016): https://cshperspectives.cshlp.org/content/8/9/a023531.full.pdf

← All articles