Infectious Protein Strains No Longer Strain the Truth?

The heretical idea that prions--naked protein particles without a stitch of genetic material--can cause transmissible disorders such as mad cow disease and Creutzfeldt-Jakob disease (CJD) in people has just received a big boost. In tomorrow's issue of Science, a team led by Glenn Telling and Stanley Prusiner of the University of California, San Francisco, provides compelling new evidence that may help explain how prions, which seem to exist in multiple strains, each with somewhat different characteristics, can reproducibly transmit those strain characteristics--a seeming impossibility for an infectious agent that lacks any genetic material.

Previous work by the Prusiner team and others had shown that prions consist of a cellular protein that has misfolded into an abnormal form. This can happen, for example, if the normal prion protein (Prp) comes into contact with an abnormal version, which then forces it to misfold, too. In the past couple of years, other researchers, studying a prion disease called mink transmissible encephalopathy, had provided evidence that different prion strains consist simply of differently misfolded forms of the prion protein, each of which can transmit its own shape, or conformation, to normal PrP. Now Telling, Prusiner, and their colleagues have extended the work to human prions in a mouse model.

They injected a special breed of transgenic mice with extracts from the brains of patients who died either from CJD or another prion disease, fatal familial insomnia. The prions causing the two diseases were known to fold differently because they yield different size fragments when cut by a protein-splitting enzyme--an indication that different amino acids are exposed on their surfaces. The Prusiner team found that prions produced in the mice yield the same size fragments as the parent molecules used for the injections. The findings are "a very powerful piece of support for the idea that an abnormal protein can confer its abnormal configuration onto the host prion protein," says neurologist Dennis Choi of Washington University in St. Louis.

Not everyone is convinced by the result, however, mainly because of the difficulty in eliminating the possibility that a more conventional disease-causing agent, such as a virus, might have been lurking in the injected material. "The group has taken a relatively crude [brain] extract," says Michael Harrington, who studies nervous system diseases at the Calfornia Institute of Technology in Pasadena. Therefore he concludes that the work "doesn't confirm" that protein-folding differences amount to different strains. But to Prusiner and his colleagues, the failure of years of research to finger a virus in prion diseases is proof that none is involved. "There's nothing to say to [people who think viruses cause the diseases] except that the evidence is overwhelming," Prusiner maintains.