Here is a good report on research into the pathogenic mechanism for "Mad Cow disease."  It's important to keep in mind that misfolded proteins are likely to be the effect, and not the cause, of the underlying mechanism which begins with some sort of unusual stressor (or a common stressor that is applied in unusually large amounts).  However, these proteins can be a necessary part of the disease "pathway," the disease being the collection of symptoms that eventually develop.  As the researcer discovers, certain environmental conditions seem to be necessary to generate the "disease."  Moreover, he points out that what he has found is a lab construct ("artifact" is also used sometimes), and may not actually occur in diseased organisms

QUOTE:  Proteins are the cell's workhorses, and they need to fold into complex and precise shapes to do their jobs. Prions are proteins that start out normally, but then at some point misfold--rather like an origami swan that comes out looking and acting instead like a vulture.

But prions have another characteristic that enables them to wreak havoc. They recruit other, properly folded proteins into misforming along with them, a process Lindquist calls a "conformational cascade." In many organisms, this conformational cascade creates long fibers called amyloids. (The brains of animals that have died from prion infections are literally packed with amyloid clumps.)...

...Tessier accessed a synthetic yeast prion, one that another research group had assembled from pieces of both the baker's yeast and the pathogenic fungi prion. Earlier studies had shown that this synthetic prion could cross the species barrier but did not identify the mechanism. Tessier found that this synthetic prion contained two recognition elements, one for baker's yeast and one for pathogenic fungi. When the prion was placed with peptide fragments from baker's yeast, the baker's yeast recognition element was activated, and likewise for the pathogenic fungi.

Even more striking, Tessier could activate different recognition elements by manipulating environmental conditions, such as temperature. For example, when he conducted the experiment at 4 degrees Celsius, the baker's yeast recognition element switched on. At 37 degrees Celsius, the pathogenic fungi element was activated. In other words, temperature alone could dictate which yeast species the prion could infect. Additionally, the prion's behavior could be altered by subtle alterations in the recognition element's amino acid sequence.

While this prion is a laboratory construct not found in nature, these findings provide researchers with a new way to approach old questions, such as why some prion diseases can jump from one species to another but others can't. Tessier and Lindquist say it is likely that natural prions contain more than one recognition element, and recognition elements can slide into a neighboring region. Many external factors can determine which recognition element is activated, in turn influencing the downstream behavior of the prion.

"These findings are remarkable for two reasons," says Lindquist, who is also an investigator for Howard Hughes Medical Institute. "For one thing, this is the first time that these peptide arrays have been used to study protein folding. We've taken this platform to a whole new level. Also, we've seen just one small part of this prion inducing proteins to fold. This is an entirely new concept..."  UNQUOTE.

Source:  http://www.sciencedaily.com/releases/2007/05/070509161210.htm