Most of what we know about prions tells us what they are not. They're not bacteria, viruses or parasites. They're not transferred by sneezing, coughing, using dirty needles, or having sex. Prion diseases cannot be cured by conventional treatments such as antibiotics, radiation, or surgery.

All of these "nots" add up to a lot of weirdness. So what exactly are prions, and what do they do, and how do they do it?

The first question, thanks to recent scientific breakthroughs, can be answered easily. What are prions? Prions are proteins. The second question is fairly simple, too. What do prions do? Prions kill you. The answer to the third question -- the "how" -- is a little more complicated.

Back to the Future

Nothing comes from nothing. For decades, the rule of biological reproduction has stated that in order for something to replicate itself, genetic material is required.

Prions, it would seem, forgot to read that particular rule.

Cells, bacteria, viruses, and yeast all contain chromosomes, which are made up of genes. The genes are made up of deoxyribonucleic acids (or, in the case of some viruses, ribonucleic acids) -- commonly referred to as DNA. This is the genetic material. It defines what the organism is and what it does.

When a cell or a bacterium replicates, the DNA first gets duplicated. The cell or bacterium then splits in two, and each half gets one complete copy of the DNA. Viruses do much the same thing, except that they often commandeer the genetic material of the cell they've infected and make it do their work for them.

Each gene in the DNA codes for an individual protein. Each protein is synthesized as it is needed, and each protein exists to perform a particular task or set of tasks. Together, the proteins assemble to form the physical structure of the cell or bacterium, with the genetic material acting as the cell's "blueprint" or "instruction manual."

Infection, by the classic definition, means that a bacterium or virus or other small organism has entered our body and reproduced itself many times over, causing damage in the process. In response to the damage, we sneeze, vomit, cough or bleed. Even a futuristic definition of infection, which includes the possibility of "gene therapy gone wrong" causing a kind of "infection by DNA," still requires the presence of genetic material.

Proteins contain no genetic material. They're composed of amino acids, synthesized from the blueprint provided by a cell's genes. Proteins are made because the cell needs them to perform a particular job, such as providing structural support or activating a cellular function. Proteins can't create new copies of themselves. Therefore, according to the "rules," they cannot cause an infection.

And prions are just proteins. No genetic material. So, logically, prions cannot cause infection. But prions have their own form of logic.

The New Rules

PrP is a perfectly normal protein produced by nerve cells. When the cell needs more PrP protein, the gene that codes for PrP gets "turned on" and a new PrP molecule is synthesized. Inside the cell, PrP gets folded into its normal three-dimensional configuration. It's then excreted and anchored to the outside of the nerve cell, where it is believed to play a role in synaptic function.

If you happen to eat some PrP, it will be denatured by the cooking heat, then chopped up by enzymes in your digestive system. Tomorrow, its leftover bits will be flushed away -- literally. No harm done.

Now: don't change the gene that codes for PrP. Don't even change the sequence of amino acids that make up the protein. Just take that perfectly normal PrP protein and tweak the 3-D structure a bit -- uncoil some of the helices and re-fold them into flat sheets. Chemically and genetically, this new PrP is identical to the old one. It's simply shaped differently, like a person who has changed from a sitting to a standing position. It's a new "isoform": different in physical appearance but not in genetic or chemical makeup.

Don't eat this one.

The two isoforms of the PrP protein are technically designated PrPC and PrPSc. The "C" stands for cellular; the "Sc" stands for scrapie. PrPSc is believed to be the main, and probably the only, component of the prion (a term derived from "proteinaceous infectious particle").

Prion diseases include scrapie in sheep, bovine spongiform encephalopathy (BSE) in cows, and Creutzfeld-Jakob disease (CJD) in humans. They're all caused by the malformed PrPSc protein. They can all be passed on by ingesting the PrPSc. And they're all fatal.

"Resistance is Futile. You Will Be Assimilated."

If you ingest a prion, it may not be "flushed away" the next morning. It's capable of crossing some species barriers. It's resistant to food-preparation treatments such as high heat and ultraviolet irradiation. It's stubbornly insoluble. And it knows how to find its way from your digestive system into your nervous system.

The "how" is not yet clearly defined. Recent studies in animal models suggest that the prion's first step would be to move from the digestive system to the immune system. The presence of specialized lymph nodes called "Peyer's patches" on the outside of the gut gives the prion a direct route from the digestive system to the immune system. From the Peyer's patches, the prion's second step would be to travel through the immune system to other immune organs such as the spleen.

At this point the prion may "hibernate" for years in the immune system, or it may move swiftly. The spleen contains nerve cells. Taking advantage of this, the prion's third step would be to move from the immune system to the nervous system. And the nerve cells in the spleen express the normal PrP protein.

Digestive system, immune system, nervous system. One, two, three simple steps. Now the prion is snuggled up right next to a PrP protein. And there are PrP proteins all through the nervous system -- up the axons, along the spinal cord, and in the brain. According to the "old rules," none of this should be a cause for concern. But the prion has other ideas.

As isoforms of each other, the prion and PrP have a lot in common. The prion presents itself to the normal PrP proteins as a template. The prion apparently fascinates its PrP counterparts, for the PrP proteins take to the prion like social zeroes to a cult leader. And, like slavish cult members, the PrP proteins emulate their new leader, converting their helical structures into the flat beta-sheet conformation. This shape-changing is the equivalent of infection.

Why the shape change occurs is not clear. Each prion may carry within itself enough kinetic energy to catalyze the conversion of the next PrP molecule. Alternatively, the physical proximity of a nearby prion molecule may stabilize the otherwise thermodynamically-unfavorable physical conformation of the first prion. Or there may even exist small "chaperone" or "helper" proteins that assist in the conformational change from the PrP isoform to the prion isoform. But, whatever the reason, the domino effect continues. Up the spinal cord and into the brain, normal PrP proteins become prions.

It's a No-brainer

The prion isoform is resistant to being degraded by cellular enzymes, so the newly-formed prions accumulate in the central nervous system. They then form aggregates, and these clumps of prions continue to increase in number. The aggregates in the brain tissue are called "amyloid plaques." The plaques display consistent characteristics such as positive staining with Congo red dye, and they are a classic feature of human and animal prion diseases.

Prion infection is characterized by a loss of neurons in the brain. Much of prion biology remains speculative, including how prions, or prions aggregated into amyloid plaques, could damage the neurons or the brain tissue. Clumps of prions outside the cells may poke, prod, squash or pinch the neurons around them, causing injury to the neurons that leads to their eventual death. Alternatively, the conversion to the prion shape may occur while PrP proteins are still inside the neuron. A pile-up of prions inside the cell would eventually cause the neuron to burst like an overfilled water balloon, releasing prions that could then "infect" the neighboring cells.

Virtually all prion-infected brains contain tiny holes or "vacuoles" that are visible under a microscope and, in many cases, even to the naked eye. These vacuoles are found in the "grey matter," the material between the cells. The grey matter is where much of the brain's synaptic activity occurs. The individual vacuoles in the gray matter are usually about a tenth of a millimeter in diameter. They eventually become so prevalent that they merge together and form large vacuoles, which distort the cellular architecture of the brain tissue and give it the appearance of a sponge. This has given the prion diseases their common name of "spongiform encephalopathies."

Since the prion-infected brain is literally riddled with holes, it is unsurprising that hosts exhibit a number of behavioral changes. A loss of coordination may cause an animal to rub against fence posts to prop itself up, so that it eventually scrapes its hair off: "scrapie" in sheep. The lack of balance and bodily control may make it extraordinarily apprehensive: "mad cow disease" in bovines. The lack of coordination can become so severe that the afflicted animal cannot stand and eat, causing it to waste away: "chronic wasting disease" in mule deer and elk. Irritability can become dementia: "CJD" in humans.

Sporadic prion disease in humans, or classic CJD, is rare. Overall, the occurrence of spontaneous prion disease in humans is estimated to be about one in a million. It is generally found in people around the age of 60. In classic CJD, the initial prion isn't ingested. Instead, by that one-in-a-million chance, the prion spontaneously forms in the brain or central nervous tissue.

The mutation of a single amino acid in the PrP protein has been documented in several cases of CJD. This mutation lowers the amount of energy needed to convert the PrP structure to that of a prion. Because it is caused by a mutation that occurs in the DNA of an adult person's nerve cell, it will not necessarily be passed on to the person's offspring; however, a genetic tendency or susceptibility for this mutation to occur may run in families. There are also rare cases of CJD contracted by patients who are given organ grafts or hormone treatments that were inadvertently taken from donors with CJD.

Dozens of prion "strains" in humans and animals have been identified, each with their own unique "fingerprint" identified in part by the protein fragments resulting from breaking the prion down with specific enzymes. An individual prion strain will faithfully reproduce itself through a series of hosts, causing a distinct pathology in the brain and a specific set of behavioral changes. Other rare, inherited prion diseases from non-CJD strains include Gerstmann-Straussler-Scheinker (GSS) syndrome and Fatal Familial Insomnia.

You Are What You Eat

Prions preferentially interact with PrP proteins of like chemical composition. This creates a species barrier, so that one species tends to be resistant to infection by prions generated in another species. The more differences in prion gene sequence between species, the greater the species barrier. Ingesting prion-containing tissue from the same species as yourself is therefore the easiest way to contract a prion disease.

Kuru, or the "laughing death," was a prion disease described in a tribe of highlanders in Papua New Guinea in the late 1950s. Like CJD, the disease caused dementia, and it was invariably fatal. This prion strain likely arose in one individual by a chance mutation, but it was then perpetuated through ritual cannibalism. The tribe's practice of honoring the dead by eating their brains has since stopped, and the disease has disappeared.

While the species barrier does provide protection, it is not insurmountable. Until recently, few people had heard of classic CJD or kuru, but when prions crossed the species barrier, the results were devastating and eminently newsworthy. "Mad cow disease" filled headlines around the world after animal feed containing protein from scrapie-infected sheep caused a BSE outbreak in cattle.

The amino acid sequences of sheep and bovine PrP are very similar; scrapie was therefore able to cross into cattle with relative ease. In contrast, the dose needed to infect a mouse with scrapie from sheep is a thousand times higher than the dose that will infect another sheep. PrP from human and other primate species is also significantly different from that of both rodents and ruminants, and for a time it was hoped that only prions from humans could infect humans or other primates.

Unfortunately, this proved incorrect. CJD from humans can infect laboratory rodents. BSE from cattle has been used to infect macaques. And, in 1996, clinicians in the United Kingdom described new variant CJD (nvCJD), a prion disease in humans that was believed to be caused by eating beef from BSE-infected cattle. Once again, prions became headline news.

The prion strain "fingerprint" in nvCJD is different from that found in classic CJD. It is, however, identical to that found in BSE. Unlike classic CJD, nvCJD is not limited to older adults; of the over two dozen cases identified to date, several have been adolescents and young adults. The duration of the illness is also different: approximately 14 months for nvCJD, as opposed to 4-5 months in classic CJD.

Prion strains differ, but they do have one thing in common: they're incurable. The prion plaques can't be removed from the brain without killing the patient. Prions cannot be destroyed by antibiotics, chemotherapy, or anti-viral medications. So far, no treatment has been shown to cure a prion-infected animal or human.

Pentosan polysulfate, a drug used by veterinarians to treat arthritis in dogs and horses, has recently been shown to block prion formation in cell cultures and to slow the progression of the disease in infected laboratory rodents. It has not yet been tested on human prions. However, the lack of viable treatments for prion diseases recently led the families of two teenaged nvCJD patients to fight in court so that the infected youngsters could receive injections of pentosan polysulfate directly into their brains. Such experimental treatments are, at present, the only option.

The majority of cattle do not have BSE, and the majority of BSE-infected cattle are culled and destroyed rather than processed into saleable beef. Therefore, the odds that a package of beef in your freezer would be prion-contaminated are extremely small.

One can reduce the risk of contracting a prion disease by avoiding high-risk foods, but not completely eliminate it. Even a strict vegan diet would not prevent the occurrence of sporadic CJD through a chance mutation. Until prions are more clearly understood, future cases or epidemics cannot be accurately predicted. Still, the risk of a given individual contracting a prion disease is likely to remain very low.

Curiouser and Curiouser

The prion protein has been identified as the main cause of prion diseases. It has been difficult to prove that additional factors are not required. To date, no "miniature virus" or "helper protein" or "novel genetic material" has been isolated in conjunction with prions. Absence of proof, however, does not negate the possibility, and some scientists continue to search for the "real cause" of prion diseases.

Scientific research is focused on prevention and cure as well as mechanism and diagnosis. Prion research using mice that have been genetically altered has provided scientists with much valuable information. Mice that have had the entire PrP gene deleted from their genetic material cannot contract prion diseases, a finding which conclusively proved that the PrP gene is required for the disease to progress.

Oddly, the PrP-deficient mice appear to be perfectly normal. This suggests that the normal PrP protein may be as unnecessary as the appendix. If PrP is proven to be non-essential, genetically deleting the PrP gene in patients may be a way of halting the disease's progress. Other future treatments may involve designer compounds that physically bind to PrP, stabilizing the protein's structure and preventing shape-flipping.

In part because of their extreme weirdness, prions have a lot of potential. We may learn to exploit the prion's "tricks" and turn them into tools. Designer prions could someday be used to remodel brains, in order to cure mental diseases or forge artificial memories or implant nano-computer chips. Such futuristic tools could, of course, also be turned into weapons. A murderer could use prions to erase a witness's memory of the crime; a murder weapon made of proteins could be altered beyond identification by changing the proteins' structures.

Other potential uses -- or misuses -- for prions are within our current capabilities. A killer who's not on a deadline could simply feed his victim a BSE-contaminated steak. Anthrax can be mailed in an envelope; why not put prions in a town's drinking water? It's all too easy to imagine prions as the ultimate in biological warfare.

Proteins can and do change shape for many different reasons. If PrP can become a disease-causing prion, what about other proteins? So far, none have been identified. But, again, absence of proof is not proof of absence. The future may hold new and even more frightening prion diseases -- or it may not.

Let's hope not.

Fran Wolber is a research scientist who, in her spare time, raises an eclectic mini-herd of (non-infected) cattle in beautiful New Zealand.