(a) Prions of S. cerevisiae cause heritable changes in phenotype. In this particular genetic background, the prion [PSI⁺] can be observed by white coloration and adenine prototrophy due to translational readthrough of a nonsense mutation in the ADE1 gene. However, the cryptic genetic variation that can be revealed by [PSI⁺] is inherently polymorphic resulting in a wide variety of strain-specific [PSI⁺] phenotypes [3].
(b) Prion phenotypes are generally caused by a reduction of the prion protein’s normal cellular activity. In vivo, the aggregation and partial loss-of-function of the prion protein, can be observed by the presence of Sup35-GFP foci in [PSI⁺] cells. These foci are composed of self-templating prion aggregates that are cytoplasmically transmitted during cell division.
(c) Nucleated aggregation of a prion protein. Purified prion protein populates a soluble state for an extended period of time, then polymerizes exponentially after the appearance of amyloid nuclei (blue trace). The lag phase can be eliminated by the addition of small quantities of preformed aggregates (red trace), demonstrating the biochemical property underlying the self-propagating prion state [11].
(d) The self-propagating prion conformation is amyloid-like, as seen by the highly ordered, fibrillar appearance of prion domain aggregates visualized by transmission electron microscopy. Amyloid is a one-dimensional protein polymer. Its free ends template a protein folding reaction that incorporate new subunits while regenerating the active template with each addition.

(a) Prions create multiple stable phenotypic states, or “strains”. [PSI⁺] strains differ by their levels of nonsense suppression, with stronger strains having less functional Sup35 available to fulfill its role in translation termination, giving rise to a whiter coloration in a particular genetic background (top). At the molecular level, strains are determined by amyloid conformational variants (bottom) that arise during nucleation but then stably propagate themselves.
(b) Along with the conformational diversity apparent in the end products of amyloid formation, multiple conformational variants are also transiently populated during the early stages of amyloid assembly, and may constitute integral on-pathway species [47]. These oligomeric intermediates likely have limited self-templating capacity, but nevertheless may contribute to the weak phenotypes associated with incipient prion states.
(c) Incipient prion states acquire progressively stronger phenotypes and stabilities, possibly via mass-action population dynamics of prion particles. A number of elegant studies have correlated the phenotypic strength of the prion state with the intracellular number of prion particles [75, 76]. Upon de novo nucleation within a prion-free cell, prion polymerization onto limiting fiber ends proceeds during the “maturation” phase under pre-steady state conditions. Upon each cell division, prion particles are distributed passively and asymmetrically to daughter cells [22]. Progeny that inherit more particles will have faster total prion polymerization rates and correspondingly stronger phenotypes, and will tend to accumulate more prion particles that will in turn strengthen the prion phenotype in subsequent generations (light pink and white cells). Conversely, cells that inherit fewer particles will have slower polymerization rates and weaker phenotypes (red and pink prion-containing cells), and themselves will tend to accumulate fewer particles to pass on to their progeny. Such noise in prion distribution may allow prions to stratify protein functionality along a continuum of semi-stable phenotypes (e.g. red cells, pink cells, and white cells) within a small number of cell generations.