Protein-only (prion) epigenetic elements confer unique phenotypes by adopting alternate conformations that specify new traits. Given the conformational flexibility of prion proteins, protein-only inheritance requires efficient self-replication of the underlying conformation. To explore the cellular regulation of conformational self-replication and its phenotypic effects, we analyzed genetic interactions between [PSI ؉ ], a prion form of the S. cerevisiae Sup35 protein (Sup35 [PSI ؉ ] ), and the three N ␣ -acetyltransferases, NatA, NatB, and NatC, which collectively modify ϳ50% of yeast proteins. Although prion propagation proceeds normally in the absence of NatB or NatC, the [PSI ؉ ] phenotype is reversed in strains lacking NatA. Despite this change in phenotype, [PSI ؉ ] NatA mutants continue to propagate heritable Sup35 [PSI ؉ ] . This uncoupling of protein state and phenotype does not arise through a decrease in the number or activity of prion templates (propagons) or through an increase in soluble Sup35. Rather, NatA null strains are specifically impaired in establishing the translation termination defect that normally accompanies Sup35 incorporation into prion complexes. The NatA effect cannot be explained by the modification of known components of the [PSI ؉ ] prion cycle including Sup35; thus, novel acetylated cellular factors must act to establish and maintain the tight link between Sup35 [PSI ؉ ] complexes and their phenotypic effects.
INTRODUCTIONThe transmission of phenotypes from one individual to another is a fundamental process in biology. Much of our understanding of these events arises from decades of study on nucleic acid metabolism, but new traits may also be passed between individuals without changes in nucleic acid content through a number of epigenetic mechanisms. One particularly intriguing example of such a process is the prion phenomenon, in which the activity of a protein is altered in a heritable way to transmit a new phenotype. How is such a feat accomplished? In 1967, Griffith proposed that some proteins, now known as prions (Prusiner, 1982), can adopt more than one stable form in vivo (Griffith, 1967). Since a protein's structure determines its function, two cells containing the same protein but in diffrent physical states will have distinct phenotypes. This protein-based process has been linked to a number of previously inexplicable events, including the development and spread of the transmissible spongiform encephalopathies in mammals (Prusiner, 1982) and the non-Mendelian inheritance of some traits in fungi (Wickner, 1994).Protein-based traits can only become transmissible, however, if the inherent structural flexibility of prion proteins can be constrained by regulatory mechanisms to create an epigenetic element. For example, if each newly synthesized prion polypeptide chain independently folded to a unique form, all cells would display the same phenotype, which would reflect the average of the accessible states. The appearance of distinct protein-based phenotypes suggests that ...