Using the genetically tractable yeast Saccharomyces cerevisiae, we are studying prions. Prions are infectious proteins. They can transmit a disease or a trait without an accompanying nucleic acid. The idea that proteins can form infectious agents is derived from studies of the transmissible spongiform encephalopathies (TSEs) in mammals. TSEs are fatal neurodegenerative diseases. Examples are scrapie of sheep, mad cow disease, chronic wasting disease of deer and elk, and kuru and Creutzfeldt-Jakob disease in humans. At least 10 proteins in S. cerevisiae are capable of forming a prion. Most of these proteins acquire their infectious nature through the formation of a type of ordered linear protein aggregate called amyloid. However, not all yeast prions are based on protein aggregation. Amyloid deposits are also found in people suffering from Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, and late-onset diabetes.
Ure2 protein and the [URE3] prion
We have focused our attention mainly on the [URE3] prion (the brackets indicate its nonchromosomal inheritance). The prion arises spontaneously in a population at a frequency of approximately one to 10 per million cells. Overexpression of Ure2 protein dramatically increases the frequency with which [URE3] is found in a population. The normal function of Ure2 protein is to turn off the uptake and utilization systems of nitrogen sources that don’t support optimal growth when good nitrogen sources are available. This nitrogen-response system, often referred to as nitrogen catabolite repression, is found in many fungi and bacteria. The C-terminal domain of the protein resembles glutathione S-transferases and like glutathione S-transferases the Ure2 protein is a dimer. The asparagine-rich N-terminal domain of 94 amino acids is natively unstructured and is needed for the normal stability of the full-length protein. It forms amyloid in vitro and is sufficient for [URE3] propagation. As such, this domain is called the prion domain. Introduction of Ure2 amyloid into yeast results in cells with [URE3], thus this prion is an infectious amyloidosis of the Ure2 protein.
Amyloid and [URE3] variants
One of the remarkable features of [URE3] is that many stably propagating variants exist. These variants can be identified through differences in propagation efficiency, effect on engineered reporters and effect on yeast growth. [URE3] variants can also dramatically differ in infection ability. We learned this when we replaced URE2 with homologues from ascomycete yeasts in S. cerevisiae. A [URE3] variant from one Ure2 protein could only infect S. cerevisiae expressing the same protein. However, another variant from the same Ure2 protein could infect S. cerevisiae expressing Ure2 proteins from any of several yeasts. Thus, infection barriers exist that depend on the protein sequence and the prion variant. Similar infection barriers have been reported for TSEs. For instance, a variety of mammals, including humans, have been infected with mad cow disease, but there are no reports of people having been infected with sheep scrapie.
How can one protein form several prion variants? In amyloid filaments the N-terminal domain of Ure2 protein forms parallel in-register beta-sheets with the beta-strands perpendicular to the long axis of the filaments. Thus in the sheets the same amino acid is lined up in adjacent molecules. These sheets are folded multiple times along the long axis of the filament. Variants arise from different folds of the sheets as well as from the sheets being at different positions along the peptide chain. A monomer newly joining the end of the filament will assume the same fold as the molecules already in the fiber.
Ure2 proteins of yeasts
Ure2 protein homologues from ascomycete yeasts all have a C-terminal glutathione S-transferase like domain. When these protein sequences are aligned the N-terminal domains show much greater variation than the C-terminal domains. Even within one species, amino acid substitutions are more prevalent in the prion domain of Ure2 protein. This variation restricts the spread of [URE3] as it affects optimal alignment of a newly joining molecule with the existing Ure2 proteins in the amyloid filament. We showed that the Ure2 proteins of Naumovozyma castellii and the human pathogenic yeast Candida glabrata, despite the presence of a prion-like domain, are not capable of forming a prion in S. cerevisiae. The Candida glabrata protein also does not form amyloid in vitro. Both proteins fulfill the nitrogen regulatory function when they replace Ure2 protein in S. cerevisiae. Together these results indicate that the N-terminal domains of Ure2 proteins are not primarily evolved to form prions.
[URE3] a prion disease in yeast
Newly de novo formed [URE3] is sometimes poorly transferred to the daughter cell. However, once established, [URE3] is reliably propagated. If [URE3] were a benefit to cells carrying it, we would expect to find the prion in wild populations. A survey of 70 wild Saccharomyces isolates did not identify any that contained [URE3]. The same survey identified wild isolates that contained viruses and plasmids not known to provide any advantage to the cells. In addition, the [PIN+] prion was found with the same frequency as viruses and plasmids. Thus [URE3] does not provide the cells a net benefit.
Cells containing newly de novo formed [URE3] often form small colonies. Upon propagation, colony size becomes more comparable to that of uninfected cells.This suggests that initially, [URE3] negatively affected the cells but then changed to a variant with a more neutral effect on the cells. The absence of [URE3] in wild Saccharomyces isolates and the presence of unstable toxic variants points to this prion as a disease of yeast.
Although the prion concept has been firmly established in yeast, there are still many unresolved issues. The mechanisms of prion formation, replication, and transmission are only vaguely understood. Likewise, many questions remain about the interactions between prions and their host cells. We are further exploring prion formation of Ure2 and other proteins.