Prions, the maddening, infectious proteins, and the diseases they trigger, such as the fatal neurodegenerative disorder in humans, Creutzfeldt-Jakob disease—as well as its bovine counterpart, mad cow disease—have baffled scientists for decades. Although researchers know what they are (abnormally folded proteins) and the illnesses that they cause, how they form and multiply has remained elusive.

Today, a study appearing in an advance online publication of the journal Nature announces a first step in demystifying the mechanisms governing prion behavior. Studying yeast proteins, a pair of researchers at the Massachusetts Institute of Technology's Whitehead Institute of Biomedical Research discovered that a highly specific section of a prion protein's amino-acid sequence controls its switch to the prion state. In addition, this same segment regulates its ability to cross species barriers.

Biology professor Susan Lindquist and postdoctoral researcher Peter Tessier examined the behavior of prions in baker's yeast (Saccharomyces cerevisiae). (Unlike in humans or cows, these yeast prions do not negatively affect their hosts.) The scientists worked with Sup35, a protein that is normally responsible for stopping the translation of messenger RNA (mRNA) into other proteins, but has the ability to readily shift into a prion state. In its prion configuration the protein does not properly halt mRNA translation. "Therefore, in a genome-wide sense you have proteins with extensions added on to them," Tessier explains, adding that this will certainly alter cellular functions. "You're going to have a large change due to this single protein changing its conformational [folded] state."

When the researchers incubated full strands of Sup35 in its nonprion state with the array of peptides, they found that a certain section of the peptides, encompassing about 10 percent of an entire protein sequence, served as a "recognition element," which bound the Sup35 protein and caused it to begin folding into its prion state. This then initiated a conformational cascade, whereby the other Sup35 strands were encouraged to fold into their prion state. The team saw similar results when they tested Sup35 strands and peptides from a pathogenic, gut-dwelling yeast, Candida albicans.

The researchers then cultured Sup35 peptides from both types of yeast on the same array and incubated them with full strands of both varieties of the prion proteins in their nonprion states. The S. cerevisiae prions formed after binding only to recognition elements within their own species and then recruited only other proteins of their species into the prion state. The C. albicans prions behaved similarly. "What that gives us is a glimpse into the basis of species barriers," Tessier says. "Our hypothesis is that part of the origin of the species barrier is governed by local regions within these prion proteins."

Lindquist and Tessier then employed a chimeric prion, one with segments from both yeast species, and found that it contained the recognition elements of both Sup35 proteins. When incubated with Sup35 peptides for S. cerevisiae, it could activate—just as it could when incubated with C. albicans. If cultured with peptides from both species, the researchers discovered that they could bias which yeast species it could infect by adjusting environmental conditions: At a temperature of four degrees Celsius, it sought out the baker's yeast's recognition element; at 37 degrees C, it was activated by C. albicans peptides. The pair also biased the activity of the prion by disrupting its native recognition elements, replacing amino acids in the signature segments for baker's yeast or the pathogenic variety.

Tessier says that he and Lindquist are now working to find recognition elements within mammalian prion protein (PrP), which in its malformed state causes mad cow and Creutzfeldt-Jakob diseases—the latter, a rare, neurodegenerative illness characterized by dementia, loss of muscle control and memory, and eventually death. The sporadic form of the disease affects about one in a million people per year, according to the National Institutes of Health. The researchers believe that PrP likely has more than one recognition element, which can kick off a conformational cascade.

"One kind of pie in the sky idea is, if you can understand these regions within disease-related prions," says Tessier, "[you can] use this to identify small molecules and peptides that may be able to block how these things interact with each other and which are key for blocking this [prion] aggregation from happening."