The yeast prions [PSI+] and [URE3] are molecular degenerative diseases

The yeast prions URE3] and PSI] are not found in wild strains, suggesting they are not an advantage. Prion-forming ability is not conserved, even within Saccharomyces, suggesting it is a disease. Prion domains have non-prion functions, explaining some conservation of sequence. However, in spite of the sequence being constrained in evolution by these non-prion functions, the prion domains vary more rapidly than the remainder of the molecule, and these changes produce a transmission barrier, suggesting that these changes were selected to block prion infection. Yeast prions PSI] and URE3] induce a cellular stress response (Hsp104 and Hsp70 induction), suggesting the cells are not happy about being infected. Recently, we showed that the array of PSI] and URE3] prions includes a majority of lethal or very toxic variants, a result not expected if either prion were an adaptive cellular response to stress.Key words: URE3], PSI⁺], prion, Sup35p, Ure2pfMammalian prions are uniformly fatal, but a lethal yeast prion would not be detected by the usual procedure, which requires growth of a colony under some selective condition. As a result, the prion variants commonly studied are quite mild in their effects. This circumstance has led to the suggestion that yeast prions actually benefit their host. Sup35p, the translation termination subunit whose amyloid becomes the PSI⁺] prion, is essential for growth and Ure2p, the nitrogen regulation protein whose amyloid constitutes the URE3] prion, is important for growth, with ure2 mutants showing noticeably slowed growth.When yeast prions were discovered,¹ we assumed they were diseases, by analogy with the mammalian diseases and the many non-prion amyloid diseases. Inactivating the essential Sup35p or the desireable Ure2p did not seem like a useful strategy. While control of either protein''s activity might be advantageous, and Ure2p activity control is the key to regulation of nitrogen catabolism, prion formation is a stochastic process, so it makes control of activity of these proteins random instead of appropriate to the circumstances. The Het-s] prion changed that picture.² Here was a prion necessary for a normal function, heterokaryon incompatibility, and we suggested that it was the first beneficial prion.³