The essential "prion-determining" portion of a prion protein can be removed and fused to a wholly separate protein, creating a new prion with the essential prion qualities of state changing and heritability, report Susan Lindquist and Liming Li from the University of Chicago in tomorrow's issue of Science. The researchers fused the NH terminus and the middle portion (which provides solubilizing and/or stability functions) of the yeast prion protein Sup35 to the rat glucocorticoid receptor. The fusion produced a protein that could still, in its normal state, perform its usual function of binding hormones and regulating transcription. In addition, however, the fused protein could also switch into a prion form (in which it could not perform its usual functions), could switch between normal and prion states spontaneously or through experimental manipulation, and could be inherited by daughter yeast cells.
The prion-determining portion of the protein is thus modular. "The astonishing thing is that the prion property can transferred to a totally different protein," says Adriano Aguzzi or the University of Zurich, quoted in a News article (pp. 562-3) in the same issue. The evidence adds weight to the idea that a prion can be a pathogenic agent in its own right, without a trigger such as a virus, though dissenters still argue that yeast prions may not be relevant for mammalian disorders. They point out that yeast prions do not lead to pathogenic states, though they do change the phenotype.
"If they can put in a pure or recombinant PrP protein, made in a virus-free cell, and get something that replicates infectivity in mammals, then I would be convinced [that the protein-only hypothesis] is correct," said Laura Manuelidis of Yale University, in the same News article. "I've been waiting 20 years to see that experiment."-Hakon Heimer.
References: Li L, Lindquist S. Creating a protein-based element of inheritance. Science 2000 January 28;287:661-4. Abstract