
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0020086Research ArticleBiotechnologyCell BiologyInfectious DiseasesMolecular Biology/Structural BiologySaccharomycesDissection and Design of Yeast Prions Dissection and Design of Yeast PrionsOsherovich Lev Z lxoshe@itsa.ucsf.edu
1

2
Cox Brian S 
2
Tuite Mick F 
2
Weissman Jonathan S 
1
1Department of Cellular and Molecular Pharmacology and Howard Hughes Medical Institute, University of CaliforniaSan Francisco, CaliforniaUnited States of America2Department of Biosciences, University of KentCanterburyUnited Kingdom4 2004 23 3 2004 23 3 2004 2 4 e8611 10 2003 21 1 2004 Copyright: © 2004 Osherovich et al.2004This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Artificial Prions Created from Portable Control Elements 

Yeast Prions: Protein Aggregation Is Not Enough 
Many proteins can misfold into β-sheet-rich, self-seeding polymers (amyloids). Prions are exceptional among such aggregates in that they are also infectious. In fungi, prions are not pathogenic but rather act as epigenetic regulators of cell physiology, providing a powerful model for studying the mechanism of prion replication. We used prion-forming domains from two budding yeast proteins (Sup35p and New1p) to examine the requirements for prion formation and inheritance. In both proteins, a glutamine/asparagine-rich (Q/N-rich) tract mediates sequence-specific aggregation, while an adjacent motif, the oligopeptide repeat, is required for the replication and stable inheritance of these aggregates. Our findings help to explain why although Q/N-rich proteins are relatively common, few form heritable aggregates: prion inheritance requires both an aggregation sequence responsible for self-seeded growth and an element that permits chaperone-dependent replication of the aggregate. Using this knowledge, we have designed novel artificial prions by fusing the replication element of Sup35p to aggregation-prone sequences from other proteins, including pathogenically expanded polyglutamine.

Artificial prions - infectious, misfolded proteins - can be created by fusing the replication element of one prion to aggregation sequences from another
==== Body
Introduction
The aggregation of misfolded proteins underlies a diverse range of human diseases, including sporadic amyloidoses such as Alzheimer's disease and hereditary neuropathies such as Huntington's disease (Dobson 1999). Prions are a special class of protein aggregates that replicate their conformation and spread infectiously (Prusiner 1998). After the discovery that prion aggregates are responsible for the mammalian transmissible spongiform encephalopathies, several epigenetically heritable traits in fungi were also found to depend on a prion mechanism (Wickner 1994; Uptain and Lindquist 2002; Osherovich and Weissman 2004). In Saccharomyces cerevisiae and Podospora anserina, prions are transmitted from cell to cell through mating and cell division, resulting in readily assayed phenotypes with a non-Mendelian pattern of inheritance (Liebman and Derkatch 1999).

The yeast non-Mendelian factors [PSI+] (Cox 1965) and [URE3] (Lacroute 1971), which are prion forms of the translation termination factor Sup35p and the transcriptional activator Ure2p, respectively, have served as useful models for the formation and replication of heritable protein aggregates. Prion forms of Sup35p and Ure2p lead to defects in their respective biochemical activities (translation termination and nitrogen catabolite repression). Mutational analysis has shown the glutamine/asparagine-rich (Q/N-rich) amino-terminal (N) domains of these proteins to be critical for prion behavior (Ter-Avanesyan et al. 1993; Masison and Wickner 1995; Patino et al. 1996; Paushkin et al. 1996; DePace et al. 1998). In vitro, these Q/N-rich domains form self-seeding, β-sheet-rich amyloid fibrils similar to those associated with Alzheimer's and Huntington's diseases (Glover et al. 1997; King et al. 1997; Taylor et al. 1999). The autocatalytic aggregation of yeast prion proteins often shows a high specificity for like molecules; for example, Sup35p N domains from different yeast species form prion aggregates that preferentially interact with molecules of their own kind (Santoso et al. 2000; Chernoff et al. 2000; Kushnirov et al. 2000; Zadorskii et al. 2000; Nakayashiki et al. 2001). [PSI+] and [URE3] can be eliminated by transient growth in the presence of guanidine hydrochloride (GuHCl), which “cures” cells of prions by inhibiting Hsp104p, a molecular chaperone needed for prion replication (Chernoff et al. 1995; Jung et al. 2002; Ness et al. 2002).

A surprisingly large number of proteins in S. cerevisiae and other eukaryotes have lengthy Q/N-rich tracts ostensibly similar to those found in the prion-forming domains of Sup35p and Ure2p (Michelitsch and Weissman 2000). From among these, we and another group identified two novel proteins, New1p and Rnq1p, with prion-forming domains resembling those of Sup35p and Ure2p (Santoso et al. 2000; Sondheimer and Lindquist 2000). When these Q/N-rich domains were fused to green fluorescent protein (GFP) and overexpressed, they formed visible aggregates resembling those of GFP-labeled Sup35p in [PSI+] cells. Fusion proteins in which these domains were introduced in place of the Sup35p prion domain could support distinct, self-specific prion states that recapitulated the translation termination defect associated with [PSI+]. Rnq1p was later shown to underlie a naturally occurring prion called [PIN+], which promotes the aggregation of Q/N-rich proteins such as Sup35p; overexpressed Sup35p forms aggregates and stimulates the appearance of [PSI+] only in [PIN+] strains (Derkatch et al. 1997; Derkatch et al. 2001). Aggregates of the New1p prion domain, whether resulting from overexpression or from a constitutive prion form (termed [NU+]), also promoted the aggregation of other Q/N-rich proteins, emulating the effect of [PIN+] (Osherovich and Weissman 2001).

Many sequences with Q/N content as high as that of Sup35p and Ure2p, including human polyglutamine expansion disease proteins, form visible aggregates when overexpressed in yeast as GFP fusions (Krobitsch and Lindquist 2000; Osherovich and Weissman 2001; Meriin et al. 2002). However, only a limited number of Q/N-rich sequences are bone fide prion domains capable of propagating these aggregates over multiple cell generations even when expressed at low levels (J. Hood and J.S.W, unpublished data). To understand what distinguishes generic Q/N-rich aggregates from heritable prions, we conducted a detailed dissection of the prion-forming regions of Sup35p and New1p. We found that the prion properties of Sup35p and New1p require the presence of two independent and portable sequence elements within their prion domains. One element mediates the growth of prion aggregates by incorporation of soluble monomers. The second promotes the inheritance of aggregates, generating new heritable “seeds” which can be partitioned between mother and daughter cells during cell division.

Results
Distinct Regions of the New1p Prion Domain Mediate Prion Growth and Division
Sup35p can alternate between a biochemically active, soluble form ([psi–]) and an aggregated prion state ([PSI+]) with diminished translation termination activity, which can be monitored by nonsense suppression of the mutant ade1–14 allele (Liebman and Derkatch 1999). Whereas [psi–] strains form red colonies on yeast extract-peptone-dextrose (YEPD) medium and cannot grow in the absence of adenine, [PSI+] strains suppress the premature stop codon in ade1-14, and thus appear pink or white on YEPD medium and grow on adenine-free medium (a phenotype termed adenine prototrophy, Ade+). The N or prion domain of Sup35p (residues 1-112) is required for [PSI+] formation but is dispensable for the translation termination activity of the carboxy-terminal C domain (Ter-Avanesyan et al. 1993). The charged middle domain (M) is not required for prion behavior, but modulates the efficiency of chaperone-dependent prion transmission (Liu et al. 2002; L.Z.O., unpublished data) (Figure 1). Two distinct regions in the N domain have previously been implicated in Sup35p aggregation: a Q/N-rich tract (residues 1–39) (DePace et al. 1998) and an oligopeptide repeat (residues 40–112) that consists of five and a half degenerate repeats of the consensus sequence P/QQGGYQQ/SYN (Liu and Lindquist 1999; Parham et al. 2001; Crist et al. 2003).

Figure 1 Schematic Diagram of Sup35p and New1p
Prion domains of both proteins are enlarged in the center, highlighting the Q/N-rich tract of Sup35p (blue), the NYN tripeptide repeat of New1p (purple), and the oligopeptide repeat sequences (orange) found in both proteins. The sequence of the NEW1 oligopetide repeat (residues 50–70) is QQQRNWKQGGNYQQGGYQSYN, while that of the adjacent tripeptide repeat region (residues 71–100) is SNYNNYNNYNNYNNYNNYNNYNKYNGQGYQ.

We had earlier identified New1p as an uncharacterized protein with a Sup35p-like N-terminal domain; when fused to the M and C domains of Sup35p, the first 153 residues of New1p (New11–153) supported a [PSI+]-like prion state termed [NU+] (Santoso et al. 2000). Sup35p and New1p have regions of clear similarity beyond their high Q/N content (Figure 1). The prion domains of both have Q/N-rich tracts and oligopeptide repeat regions, although their order is reversed. The C-terminal domains of New1p and Sup35p also appear to be related, based on modest homology and the similarity of the translation termination defects in sup35 (Song and Liebman 1985) and new1 mutants (L.Z.O., unpublished data).

To understand the sequence requirements for the prion behavior of New1p, we constructed a series of truncated prion domains (Figure 2A) and examined their participation in several critical steps of the prion replication cycle. We first asked whether these truncated prion domains could form visible foci when fused to GFP (aggregation). Next, we examined whether such aggregates could convert New11–153 into a [NU+] prion state (induction). Finally, we fused these constructs to the M and C domains of Sup35p (–M-C), introduced them in place of endogenous SUP35, and assessed whether these proteins could adopt stable prion states (maintenance).

Figure 2 Dissection of the New1p Prion Domain Reveals Distinct Regions Responsible for Aggregation and Prion Inheritance
(A) Indicated fragments of New1p (left) were expressed as GFP fusions (inducers) in a [nu–] [pin–] strain, examined by microscopy for GFP aggregation, then plated on SD-ade medium to assess induction of [NU+]. The symbol “+” indicates induction frequencies of at least 5%; the symbol “–” indicates no induction. Maintenance was assessed by the ability of an episomal maintainer version of the indicated fragment to support an Ade+ state after overexpression of New11–153-GFP (see Materials and Methods). The aggregation of New1-GFP fusions (second column) has been previously reported (Osherovich and Weissman 2001).

(B) The NYN repeat of New1p induces [NU+] and [NU+]mini. New170–100-GFP was overexpressed in [nu–] and [nu–]mini strains ([pin–] and [PIN+] derivatives of each), along with vector only or New11–153-GFP controls. Averages of three independent trials, representing 600–2000 colonies, are shown for most induction experiments; inductions using New170–100-GFP were conducted twice. Error bars show minimal and maximal observed induction efficiencies.

(C) Reversibility of [NU+]mini. The [pin–] Ade+ convertants obtained in (B) were colony purified on SD-ade medium and confirmed to have lost the inducer plasmid. A stable [NU+]mini isolate is shown before and after induction, as well as after GuHCl treatment, along with [nu–] and [NU+] reference strains.

We found that distinct regions within the New1p prion domain are necessary for the induction and maintenance of [NU+] (Figure 2A). The asparagine-tyrosine-asparagine (NYN) repeat (residues 70–100), which we had earlier shown to be sufficient for aggregation (Osherovich and Weissman 2001), also proved sufficient for induction of [NU+]. As with the full-length New1p prion domain, overexpression of the NYN repeat efficiently stimulated the appearance of Ade+ in [nu–] cells (Figure 2B, left). However, stable prion maintenance required both the NYN repeat and the adjacent oligopeptide repeat. In a strain with this minimized New1p prion domain (residues 50–100), overexpression of the full prion domain or of the NYN repeat alone promoted the appearance of Ade+ colonies (Figure 2B, right). The resulting convertants remained Ade+ after loss of the inducer plasmid but reverted to Ade- after transient GuHCl treatment (Figure 2C). We conclude that the oligopeptide repeat and the NYN repeat of New1p together are sufficient to support a prion state, termed [NU+]mini, which recapitulates the characteristics of [NU+].

Dissection of the Sup35p Prion Domain
In light of the similarity between New1p and Sup35p prion domains, we asked whether separate regions of Sup35p were involved in the induction and maintenance of [PSI+] aggregates (Figure 3). We constructed a series of truncated Sup35p N domains and analyzed their behavior in the aggregation, induction, and maintenance assays described above for [NU+]. Additionally, we examined the ability of truncated N domains to decorate preexisting Sup35p aggregates in [PSI+] strains.

Figure 3 Dissection of the Sup35p Prion Domain
At top are schematic diagrams of these experiments; positive outcomes are shown below the arrows. In some cases, similar experiments have been reported by Parham et al. (2001) (indicated by “a”) and are repeated here as controls.

Aggregation: Plasmid-borne M-GFP fusions of the indicated Sup35p N domain fragments (green) were overexpressed in a [psi–] [PIN+] strain and examined for fluorescent focus formation. The symbol “+” indicates that 10% or more of cells displayed aggregates. Sup351–57-M-GFP displayed a lower frequency of aggregation (approximately 1%).

Induction: Strains from the aggregation experiment were plated onto SD-ade medium and scored for growth to test whether aggregates of truncated protein (green) convert chromosomally encoded protein (blue) to [PSI+]. The symbol “+” indicates approximately 5–10% conversion frequency. Consistent with the aggregation experiment, Sup351–57-M-GFP displayed a lower frequency of [PSI+] induction (approximately 1%).

Decoration: Indicated proteins were expressed as –M-GFP fusions in [PSI+] [PIN+] cells, which were examined to determine whether GFP-labeled truncations (green) decorate preexisting aggregates of full-length Sup35p (blue). Curiously, Sup351–49-M-GFP in [PSI+] cells formed abnormally large “ribbon” aggregates of the kind typically observed during de novo [PSI+] induction; furthermore, approximately 10% of the cells reverted to [psi–] (indicated by “*”). Thus, this truncation was a potent dominant PNM mutant.

Maintenance: A SUP35-deleted [PSI+] [PIN+] bearing wild-type SUP35 maintainer (blue) was transformed with maintainer plasmids containing the indicated truncation (purple). The wild-type maintainer was lost by counterselection, and the resulting strain was tested for [PSI+] by color and growth on SD-ade medium. The Sup351–93 mutant displayed an intermediate pink color and grew poorly on SD-ade medium, as previously reported (Parham et al. 2001).

Note: King (2001) reports that Sup351–61-GFP fusion could decorate [PSI+] aggregates in certain strains and could induce [PSI+] de novo when overexpressed.

We found that the Q/N-rich tract and a small portion of the adjacent oligopeptide repeat are responsible for Sup35p aggregation and de novo [PSI+] induction. Deletions within the Q/N-rich tract or of oligopeptide repeat 1 abolished these properties, whereas a construct containing only the Q/N-rich region and the first two oligopeptide repeats (residues 1–64) aggregated and induced [PSI+] at levels comparable to the full prion domain, in agreement with King (2001). A construct (residues 1–57) with a partial deletion of oligopeptide repeat 2 could still aggregate and induce [PSI+], albeit with decreased efficiency. Although a construct lacking oligopeptide repeat 2 entirely (residues 1–49) did not induce [PSI+] de novo, this GFP fusion could nonetheless decorate preexisting Sup35p aggregates. Thus, while oligopeptide repeat 2 contributes to the aggregation of Sup35p, the primary determinants of prion induction reside in the amino-terminal Q/N-rich region and oligopeptide repeat 1.

In contrast, the rest of the oligopeptide repeat region is needed for stable inheritance of [PSI+] aggregates. Constructs that did not form fluorescent foci could not retain [PSI+], suggesting that aggregation is a prerequisite for prion maintenance. However, aggregation is not sufficient for prion inheritance, as Sup35p constructs with deletions spanning oligopeptide repeats 3–5 could not support a prion state despite their ability to form aggregates and efficiently induce [PSI+]. Only the sixth (incomplete) oligopeptide repeat proved dispensable for [PSI+] maintenance, consistent with an earlier report (Parham et al. 2001).

The PNM2-1 Mutation in Oligopeptide Repeat 2 Specifically Compromises the Inheritance of [PSI+]
Our deletion analysis suggested that oligopeptide repeat 2 participated in both the formation and inheritance of Sup35p aggregates. We made use of a point mutation within oligopeptide repeat 2 known as PNM2-1 (G58D) to distinguish between these two functions. PNM2-1 (PSI No More) shows strong interference with [PSI+] in certain strain backgrounds through a poorly understood mechanism (McCready et al. 1977; Doel et al. 1994; Kochneva-Pervukhova et al. 1998; Derkatch et al. 1999).

Using both in vivo and in vitro assays, we established that PNM2-1 does not have a defect in aggregation or [PSI+] induction. Earlier work indicated that PNM2-1 is capable of seeding [PSI+] in vivo (Kochneva-Pervukhova et al. 1998; Derkatch et al. 1999; King 2001). Consistent with these reports, we found that overexpression of a PNM2-1-GFP fusion in [psi–] [PIN+] cells with a wild-type SUP35 locus led to both focus formation and [PSI+] induction (Figure 4A). A previous study of Sup35p polymerization in extracts had suggested that PNM2-1 might interfere with [PSI+] through a defect in seeding (Kochneva-Pervukhova et al. 1998). We tested this by examining the rate of seeded polymerization of recombinant PNM2-1 protein. Like wild-type Sup35p, purified PNM2-1 spontaneously formed amyloid fibrils in vitro; this was accelerated by the addition of preformed Sup35p polymer seeds (data not shown). We measured the initial rates of polymerization of wild-type and PNM2-1 protein seeded by preformed wild-type polymers (Figure 4B) and by PNM2-1 polymers (Figure 4C) using a thioflavin-T–binding assay. We observed that wild-type and PNM2-1 monomers were seeded by wild-type polymers with similar kinetics; likewise, PNM2-1 polymers seeded both wild-type and PNM2-1 monomers equivalently. Thus, the PNM2-1 mutation does not affect polymerization or seeding.

Figure 4 PNM2–1 (G58D) Prevents Inheritance But Not Aggregation of Sup35p Prions
(A) PNM2-1 protein can seed [PSI+]. A Sup35p inducer containing the PNM2-1 (G58D) mutation was overexpressed in [psi–] [PIN+] cells; shown are cells (inset) with representative fluorescent foci, which were the same in frequency and appearance as cells with a wild-type inducer. Cells overexpressing inducer versions of wild-type Sup35p (SUP), an aggregation-defective N-terminal truncation (Δ1–38), and PNM2-1 were plated and scored for Ade+. Approximately 1000 colonies were counted.

(B) PNM2-1 protein polymerization is similar to that of wild-type protein.

(C) Preformed PNM2-1 polymers seed wild-type and PNM2-1 monomers with comparable efficiency. Endpoint PNM2-1 polymers were used to seed fresh reactions.

(D) PNM2-1 displays a partially dominant, incompletely penetrant defect in [PSI+] maintenance. [psi–] (1) and [PSI+] (2) SUP35::TRP1 pSUP35 controls are shown. [PSI+] [PIN+] SUP35::TRP1 pSUP35 was transformed with a second maintainer expressing PNM2-1 (3). The wild-type maintainer (pSUP35) was then lost through counterselection (4). Red sectors from (4) were isolated, retransformed with the wild-type maintainer, and allowed to lose the PNM2-1 maintainer (5).

(E) Mitotic instability of [PSI+] in the PNM2-1 strain. A pink (Ade+) [PSI+] [PIN+] PNM2-1 isolate was grown to log phase in SD-ade liquid then shifted into nonselective (YEPD) medium. At indicated time points, aliquots were plated onto SD-ade and YEPD media to determine the fraction of [PSI+] cells (minimum of 200 colonies counted per time point). Whereas a wild-type control remained [PSI+] through the experiment, the PNM2-1 strain rapidly lost [PSI+] during logarithmic growth; during stationary phase (18 h and beyond), the percentage of [PSI+] PNM2-1 strains remained unchanged (approximately 5%).

(F) Propagon count of PNM2-1 vs. wild-type [PSI+] strains. The majority of PNM2-1 cells had no [PSI+] propagons (i.e., were [psi–]). In both strains, a small number of “jackpot” cells contained over 200 propagons; see Cox et al. (2003).

Instead, the PNM2-1 strain shows a marked defect in the inheritance of [PSI+]. When the wild-type SUP35 gene of a [PSI+] strain was replaced with PNM2-1, the strain retained the prion on synthetic defined (SD) yeast medium that selected for [PSI+] (SD-ade medium) but reverted to [psi–] at a high frequency in nonselective YEPD medium, resulting in sectored colonies (Figure 4D). We measured the rate of [PSI+] loss in a PNM2-1 strain by growing it in YEPD medium and, at various time points, plating aliquots of the culture onto SD-ade medium to determine the fraction of cells that had retained [PSI+] (Figure 4E). A wild-type strain retained [PSI+] in all of the cells throughout the experiment. By contrast, in the PNM2-1 strain the fraction of [PSI+] cells decreased rapidly while the cells grew logarithmically, but remained at a constant level when the cells entered stationary phase. These findings indicate that PNM2-1 acts to eliminate [PSI+] in dividing cells, consistent with a defect in prion replication.

We next used a recently described assay to measure the number of heritable prion seeds (propagons) in a PNM2-1 strain. Here, prion replication is inhibited by GuHCl treatment. As the cells divide, preexisting propagons are diluted but not destroyed. The number of propagons present in a colony arising from a single cell is then evaluated by removing the GuHCl prion replication block after a large number (10 or more) of cell divisions and counting the total number of [PSI+] cells in that colony (Cox et al. 2003). Whereas a wild-type strain had a median of 92 (n = 24) propagons per cell, the PNM2-1 strain had dramatically fewer: 41 of 50 cells had no [PSI+] propagons at all (i.e., were [psi–]), and among the remaining nine [PSI+] cells, the median propagon number was six (Figure 4F). Thus, although a PNM2-1 strain can harbor [PSI+] prions, a defect in propagon replication causes mitotic instability, demonstrating the importance of oligopeptide repeat 2 in prion replication or segregation.

Design of Novel Prion Domains
Our data suggested that the formation and inheritance of prions involve distinct regions of Sup35p and New1p prion domains. To assess the interchangeability of these prion domain components, we constructed a chimeric prion domain, termed F, in which the aggregation-determining NYN repeat of New1p was fused to the oligopeptide repeats of Sup35p (Figure 5A). While initially soluble and active, a fusion of F and the Sup35p M and C domains (F-M-C) could be converted into an aggregated state, termed [F+], after transient overexpression of F-M-GFP. As with [NU+], [F+] induction did not require [PIN+] (data not shown). [F+] could be eliminated by GuHCl treatment (Figure 5B) and was inherited in a dominant, non-Mendelian manner (Figure 5C). As with Sup35p in a [PSI+] strain, F-M-C protein in [F+] but not in [f –] extracts sedimented entirely to the pellet fraction following high-speed centrifugation (Figure 5D). Thus, [F+] results from a prion state of F-M-C.

Figure 5 F, A New1p–Sup35p Chimera, Shows Prion Characteristics of New1p
(A) Schematic diagram illustrating the construction of chimera F.

(B) Chimera F forms a prion, [F+]. The SUP35 gene in a [psi–] [pin–] strain was replaced with the F-M-C fusion; after transient overexpression of F-M-GFP, approximately 10% of these cells converted from an Ade- ([f –]) to an Ade+ ([F+]) state. Shown are examples of[f –] and [F+] strains, before and after GuHCl treatment, along with [psi–] and [PSI+] controls.

(C) Non-Mendelian inheritance of [F+]. A diploid made by mating a [F+] MATa strain against an [f –] MATα displayed a [F+] phenotype and, when sporulated, produced four [F+] meiotic progeny. All 11 tetrads examined showed this 4:0 pattern of inheritance.

(D) Sedimentation analysis of F-M-C. Extracts of [f –] and [F+] strains, along with [psi–] and [PSI+] controls, were subjected to 50K × g ultracentrifugation for 15 min. Total, supernatant, and pellet fractions were separated by SDS-PAGE, transferred to nitrocellulose, and probed with anti-SUP35NM serum. As with Sup35p, the prion form of F-M-C sediments primarily to the pellet but remains in the supernatant in [f –].

(E) F-M-GFP overexpression induces [NU+] but not [PSI+]. Indicated inducers and maintainers were used in an induction experiment. The symbol “+” indicates approximately 5–10% conversion to Ade+. F induced [NU+] at a comparable efficiency to New11–153; although New11–153 overexpression promoted the appearance of Ade+ colonies in the F-M-C strain, these were fewer in number (less than 5%) and reverted to Ade- after restreaking.

(F) [F+] and [NU+] prion proteins interact with each other but not with [PSI+]. Episomal “second maintainers” were introduced into the indicated strains, along with an empty vector control. Antisuppression (red) indicates that the second maintainer is soluble, while white/pink indicates coaggregation of the endogenous and episomal maintainers.

We next explored the specificity of [F+] prion seeding. Overexpression of the Sup35p prion domain did not induce [F+]; conversely, F-M-GFP overexpression did not induce [PSI+] (Figure 5E). However, F-M-GFP readily induced [NU+], indicating that mismatched sequences outside of the aggregating region did not prevent cross-interactions between heterologous proteins. Interestingly, overexpression of New11–53-GFP induced Ade+ colonies in the [f –] strain, but this adenine prototrophy proved unstable. We also examined the ability of preexisting prion aggregates to recruit different prion-forming proteins using an antisuppression assay (Santoso et al. 2000) (Figure 5F). [PSI+], [F+], and [NU+] strains were transformed with Sup35p–, F-M-C– or New11–153-M-C–encoding plasmids; the color of the resulting colonies indicates whether the second maintainer protein is soluble (red) or aggregates as a result of the resident prion (pink/white). Consistent with the induction data, F-M-C and New11–153-M-C were not incorporated into [PSI+] aggregates; likewise, Sup35p did not interact with [F+] or [NU+] aggregates. However, [F+] prions recruited New11–153-M-C and, to a lesser extent, [NU+] recruited F-M-C. Thus, F and New1p prion domains can cross-interact during de novo induction and at normal levels of expression, indicating that the NYN repeat is sufficient to specify homotypic interaction between two otherwise distinct prion domains.

Can a simple aggregation-prone sequence such as polyglu-tamine (Zoghbi and Orr 2000) be turned into a heritable prion by fusion to an oligopeptide repeat? We designed artificial prion domains containing short (Q22) and pathogenically expanded (Q62) polyglutamine tracts, either alone or adjacent to the Sup35p oligopeptide repeat (Figure 6A). These domains were fused to –M-GFP and –M-C to create polyglutamine inducers and maintainers, respectively. Q22 inducers did not aggregate upon overexpression, but Q62 inducers (with and without oligopeptide repeats) formed visible foci in [psi–] [PIN+] cells (Figure 6B). Transient overexpression of Q62 inducers had no effect on the Q22 maintainer with the oligopeptide repeat or on the Q62 maintainer lacking the oligopeptide repeat. However, the Q62 maintainer with an oligopeptide repeat supported prion inheritance, converting to a stable Ade+ state following overexpression of the cognate inducer (Figure 6C). Several tests confirmed the prion nature of this state, termed [Q+]. Like [PSI+], [Q+] did not require the presence of the inducer plasmid after transient overexpression, was sensitive to GuHCl treatment (Figure 6D), and displayed a dominant, non-Mendelian pattern of inheritance (Figure 6E). We further tested the specificity of the [Q+] state by introducing a plasmid encoding a noncognate second maintainer into a [Q+] strain (Figure 6F). The Q62 maintainer failed to be incorporated into [PSI+] aggregates, causing antisuppression (red); conversely, Sup35p did not enter [Q+] aggregates.

Figure 6 [Q+], a Prion Form of Pathogenically Expanded Polyglutamine
(A) Schematic illustrating the construction of polyglutamine-derived prion domains. (Op) indicates the presence of the Sup35p oligopeptide repeats (residues 40–124).

(B) Fluorescence micrographs of [psi–] [PIN+] strains expressing indicated polyglutamine inducers.

(C) Polyglutamine-based prion inheritance. Strains with indicated inducers and maintainers were plated onto SD-ade and YEPD media to determine the fraction of Ade+ after a standard induction experiment. Interestingly, Q62 inducer forms aggregates but does not promote Ade+ in the Q62(Op) maintainer strain. Note that Q62(Op) shows a high rate of spontaneous appearance of Ade+.

(D) GuHCl sensitivity of the [Q+] state. An Ade+ convertant obtained in (C) was restreaked to lose the inducer plasmid, then grown on GuHCl. Shown are plates before and after GuHCl treatment, along with [psi–] and [PSI+] controls.

(E) Dominance and non-Mendelian inheritance of [Q+]. See Figure 5C.

(F) [Q+] does not interact with Sup35p and vice versa. [Q+] and [PSI+] strains were transformed with indicated maintainers; mismatches between the maintainer and the chromosomally encoded allele result in antisuppression (red).

Discussion
A number of epigenetic traits in fungi result from the stable inheritance of self-propagating, infectious protein aggregrates (prions) (Uptain and Lindquist 2002). Prion inheritance requires three sequential events that must keep pace with cell division to preserve the number of heritable prion units, or propagons, per cell (Osherovich and Weissman 2004). First, prion aggregates must grow in size by incorporating newly synthesized protein. Next, these enlarged aggregates must be divided into smaller ones through the action of cellular chaperones (Kushnirov and Ter-Avanesyan 1998; Borchsenius et al. 2001; Ness et al. 2002; Kryndushkin et al. 2003). Finally, these regenerated propagons must be distributed to mother and daughter cells (Cox et al. 2003); for small, cytoplasmic aggregates, this distribution may occur passively by diffusion during cytokinesis. In the present study, we have dissected the prion-forming domains of Sup35p and New1p to discover the sequence elements involved in these steps. We have found that these domains consist largely of modular, interchangeable elements that serve distinct functions of prion growth and division or transmission.

Aggregation underlies the growth phase of the prion replication cycle (Figure 7A) and occurs through the templated addition of conformationally compatible monomers onto preexisting seeds. Like other amyloids, yeast prions display a high specificity for homotypic aggregation (Santoso et al. 2000; Chernoff et al. 2000; Kushnirov et al. 2000; Zadorskii et al. 2000; Nakayashiki et al. 2001). This discrimination arises from differences in the amino acid sequence and the conformation (Chien and Weissman 2001) of the aggregation-promoting Q/N-rich elements found in each yeast prion protein. Aggregation and specificity are dictated by the NYN repeat (residues 70–100) of New1p and by the Q/N-rich amino terminal region (residues 1–57) of Sup35p.

Figure 7 Model for Prion Growth and Division
(A) During prion growth, polymers seed the incorporation of monomers through interactions between Q/N-rich aggregation sequences (blue). Proteins with noncognate aggregation sequences (red) are excluded.

(B) The division phase of prion replication requires the oligopeptide repeats (orange), which may facilitate the action of chaperones such as Hsp104p (scimitar) in breaking the polymer into smaller, heritable units.

In contrast, the conserved oligopeptide repeat sequence mediates the division and/or segregation phase of prion replication (Figure 7B). In New1p, the NYN repeat alone can aggregate and induce [NU+] but requires an adjacent oligopeptide repeat to form a minimal heritable New1p prion, [NU+]mini. Similarly, in Sup35p, the Q/N-rich amino terminal region mediates aggregation whereas most of the oligopeptide repeats are needed for the inheritance of [PSI+] propagons. Oligopeptide repeats 1 and 2 appear to contribute to both growth and inheritance, consistent with earlier reports that expansion and deletion within this region modulate in vitro polymerization of Sup35p and the appearance of [PSI+] in vivo (Liu and Lindquist 1999). However, the two functions can be distinguished by a point mutant in oligopeptide repeat 2 (PNM2-1), which displays a specific defect in [PSI+] inheritance despite normal aggregation. Certain [PSI+] variants are resistant to the dominant negative effect of PNM2-1 (Derkatch et al. 1999; King 2001); this suggests that although oligopeptide repeat 2 is critical for the replication of the [PSI+] variant used in our studies, it may be less important for the replication of other Sup35p prion conformations.

Many studies have established that prion inheritance requires the action of cellular chaperones such as Hsp104p and Hsp70 proteins (reviewed in Osherovich and Weissman 2002), although how these proteins contribute is poorly understood. We propose that oligopeptide repeats turn nonheritable aggregates into prions by facilitating chaperone-mediated division. Oligopeptide repeats may allow the division of aggregates by providing direct binding sites for chaperones or by altering the conformation of the amyloid core to allow chaperone access. An earlier study established that deletion of residues 22–69 of Sup35p (which include parts of both the Q/N tract and the oligopeptide repeat) interferes with both [PSI+] induction and chaperone-mediated prion disaggregation (Borchsenius et al. 2001). Unlike the Δ22–69 mutant, the prion replication defect in PNM2-1 could not be corrected by increasing Hsp104p levels (data not shown), arguing that the mitotic instability of PNM2-1 [PSI+] is not due solely to inadequate Hsp104p binding.

Our findings help to explain why, among many Q/N-rich proteins in yeast, only a small subset form heritable prions. While many Q/N-rich proteins can aggregate when overexpressed (Sondheimer and Lindquist 2000; Derkatch et al. 2001; Osherovich and Weissman 2001), prion inheritance of such aggregates requires that they be divided and passed on to the next generation. Although the inheritance of Sup35p and New1p prions is mediated by oligopeptide repeats, other sequences could also serve this purpose. Ure2p lacks an oligopeptide repeat; interestingly, many isolates of [URE3] are mitotically unstable in the absence of selection (Schlumpberger et al. 2001). Rnq1p, which underlies [PIN+], also lacks a strict oligopeptide repeat, but a region (residues 218–405) within its prion domain has an amino acid content reminiscent of the oligopeptide repeat sequence (i.e., numerous Q, N, S, Y, and G residues) (Resende et al. 2003). Only two other yeast proteins, YDR210W and YBR016W, have clearly recognizable oligopeptide repeats; both proteins also have Q/N-rich regions. YBR016W forms aggregates when overexpressed (Sondheimer and Lindquist 2000), but it is not known whether either protein can maintain a heritable aggregated state. Although the mammalian prion protein PrP contains a sequence resembling the oligopeptide repeat that can functionally replace one of the Sup35p repeats (Parham et al. 2001), it is unclear whether this sequence is important in the replication of the PrPSc state.

The interchangeable nature of prion domain components allowed us to design novel artificial prions. The F chimera, consisting of the aggregation sequence of New1p and the oligopeptide repeat of Sup35p, demonstrates that the growth and specificity of prions is largely determined by the Q/N-rich tract, not by the oligopeptide repeat. Despite a sequence derived primarily from Sup35p, the F chimera behaved like New1p rather than like Sup35p. The [F+] prion cross-interacted with New1p but not Sup35p. Like [NU+], [F+] could be induced in the absence of a prion-promoting (PIN) factor. Finally, [F+] could itself act as a PIN factor, as does [NU+] (data not shown). Notably, the NYN repeat of New1p functions as an aggregation module apparently without regard to its position within a protein; this sequence induced prions when overexpressed by itself or with oligopeptide repeats at its N-terminal (in New11–153 and New150–100) or C-terminal regions (in the F chimera). These results suggest that aggregation sequences are portable and functionally separable from the oligopeptide repeat, perhaps constituting a structurally discrete amyloid core. Indeed, a peptide derived from the amino-terminal region of Sup35p forms a self-seeding amyloid in vitro (Balbirnie et al. 2001). A simple aggregation-prone sequence, pathogenically expanded glutamine, also supports prion inheritance when adjacent to the oligopeptide repeat, suggesting that prion domains can consist of little more than a generic, aggregating core sequence and an inheritance-promoting element.

In addition to illuminating the principles of yeast prion domain architecture, artificial prions with distinct specificity may be useful as controllable epigenetic regulators of protein activity. Such prion “switches” can be turned on and off by transient overexpression and genetic repression; for example, the Q prion domain could be fused to other proteins in order to conditionally and reversibly inactivate them independently of [PSI+]. It may also be possible to design additional artificial yeast prion domains whose aggregation is driven by non-Q/N-rich amyloidogenic proteins such as the Aβ peptide that accumulates in Alzheimer's disease (Koo et al. 1999) or the mammalian prion protein PrP (Cohen and Prusiner 1998). Such artificial prions could serve as models for aggregate–chaperone interactions in metazoans and could provide a genetic system for the high-throughput screening of modulators of human aggregation diseases.

Materials and Methods

Yeast strains and methods
Derivatives of W303 (Osherovich and Weissman 2001), with the initial genotypes ade1-14, his3-11,15, leu2-3, trp1-1, and ura3-1, were used throughout unless otherwise noted; all strains were [PIN+]. Strain numbers, with indicated genotypic differences, are as follows: YJW 584 [psi–] MATa, YJW 508 [PSI+] MATα, YJW 716 [nu–] MATα
sup35::TRP1 pRS315SpNew11–153-M-C, YJW 717 [NU+] MATα
sup35::TRP1 pRS315SpNew11–153-M-C, YJW 844 [f –] MATα
sup35::F-M-C C.g. HIS3, YJW 881 [F+] MATa
sup35::F-M-C C.g. HIS3, YJW 867 [q–] MATα
sup35::Q-M-C C.g. HIS3, YJW 868 [Q+] MATa
sup35::Q-M-C C.g. HIS3. Maintainer plasmids used in Figure 3 (see plasmid and gene replacement construction, below) were introduced by plasmid shuffling into YJW 716 or YJW 753 ([PSI+] MATa
sup35::TRP1 pRS316SpSUP35), followed by loss of the maintainer spontaneously or through 5-FOA counterselection. The PNM2-1 strain in Figure 4 was generated in this manner and was subsequently restreaked on SD-ade to select for [PSI+]. HIS3-marked oligopeptide repeat truncations and PNM2-1 maintainers were from Parham et al. (2001); all other Sup35p and New1p maintainers were marked with LEU2. The [f–] strain was generated by “gamma” chromosomal integration of pRS306 F-M-C into the SUP35 locus of YJW 584; excision of the wild-type gene was confirmed by PCR of Ade- colonies arising from subsequent growth on 5-FOA. The [q–] strain was made by “omega” chromosomal gene replacement (Kitada et al. 1995) of SUP35 with a C.glabrata HIS3-marked –M-C variant (with or without oligopeptide repeats) into the SUP35 locus of a diploid [PSI+] [PIN+] strain. After sporulation, gene replacement was confirmed by PCR and by loss of [PSI+] in half of the haploid progeny. Yeast culture methods were according to standard procedures (Sherman 1991), but YEPD-medium plates contained 1/4 of the standard amount of yeast extract to accentuate color phenotypes. For prion curing, strains were grown on YEPD medium plus 3 mM GuHCl, then restreaked onto YEPD medium.

Plasmid and gene replacement construction
The modular SUP35 cloning system described in previous reports was used throughout (Santoso et al. 2000; Osherovich and Weissman 2001). All plasmids are derived from Sikorski and Hieter (1989); sequence files of all constructs are available as a web supplement (Data S1). Maintainer plasmids are low-copy CEN/ARS (pRS31x series) with the native SUP35 promoter (Sp) driving the expression of the indicated prion domain followed by the M and C domains of Sup35p. Inducer plasmids are high-copy 2μ (pRS42x series) with the inducible CUP1 promoter (Cp) driving the expression of the indicated prion domain followed by the Sup35p M domain and GFP. New1p inducers did not include the Sup35p M domain. For polyglutamine constructs, polyglutamine tracts (22 and 62) were amplified out of the MJDtr constructs used in an earlier study (Osherovich and Weissman 2001). To permit amplification, primers contained sequences homologous to several codons adjacent to the 5′ and 3′ ends of the polyglutamine tracts plus an initiator ATG codon. Thus, the polyglutamine sequences read MAYFEK(Q22/62)DLSG. The resulting PCR fragments were cloned into maintainer and inducer plasmids, which were used as templates for gene replacement PCR (see yeast strains and methods, above).

In vivo prion assays
For aggregation, inducers were overexpressed by growth of cells in selective medium with 50 μM CuSO4 until the culture reached stationary phase; cells were then examined by fluorescent microscopy (Zeiss Axiovert, Zeiss, Oberkochen, Germany; Metamorph imaging software, Universal Imaging Corporation, Downingtown, Pennsylvania, United States). Unless otherwise noted, cultures displaying 10% or more cells with foci were scored as positive. For induction, dilutions of the above cultures were plated onto SD-ade and YEPD media to determine percentage of Ade+. In qualitative assessments, strains were scored as positive if 5% or more of plated cells grew on SD-ade medium after 5 d. In [NU+] maintenance experiments, strains with indicated maintainers were tested for the ability to support an Ade+ state following New11–153-GFP overexpression. In [PSI+] maintenance experiments, strains that began as [PSI+] were tested for Ade+ after plasmid shuffle gene replacement with the indicated maintainer. For decoration, a [PSI+] [PIN+] strain was transformed with the indicated inducers, grown in selective medium with 50 μM CuSO4, and examined by fluorescence microscopy during midlogarithmic phase. Propagon counts were performed as described in Cox et al. (2003). For the antisuppression assay, indicated strains were transformed with a second, differently marked maintainer plasmid, and color phenotypes were assayed on medium selective for both plasmids.

In vitro prion assays
Centrifugation was performed as described in Ness et al. (2002). Immunoblots were visualized with MT130 anti-Sup35p N-M domain serum.For the polymerization of PNM2-1, the PNM2-1 N and M domains were cloned as 7-histidine fusions into pAED4 and expressed and purified as described in DePace et al. (1998). Thioflavin-T binding was conducted as in Chien et al. (2003). The slope of early (0–6 min) dye binding was obtained from seeded polymerization reactions conducted in triplicate. To correct for a difference in dye binding between wild-type and PNM2-1 protein, these values were normalized to the end point (90 min) maximum signal for each protein. Monomer concentrations were 2.5μM.

Supporting Information
Data S1 DNA Sequences of Constructs
(30 KB ZIP).

Click here for additional data file.

 Accession Numbers
The GenBank accession numbers for the proteins discussed in this paper are Hsp104p (NP_013074), New1p (NP_015098), Rnq1p (NP_09902), Sup35p (NP_010457), Ure2p (NC_014170), YDR210W (NP_010496), and YBR016W (NP_010319).

We thank Maya Shuldiner, Kim Tipton, Peter Chien, Sean Collins, and other members of the Weissman and Tuite labs for critical comments. Work in the Weissman lab was funded by the Howard Hughes Medical Institute and the Packard Foundation. Work in the Tuite lab was funded by the Wellcome Trust and by the Biotechnology and Biological Sciences Research Council LZO was a Howard Hughes Medical Institute Predoctoral Fellow and thanks Werner Herzog for inspiration.


Conflicts of interest. The authors have declared that no conflicts of interest exist.


Author contributions. LZO, BSC, MFT, and JSW conceived and designed the experiments. LZO performed the experiments. LZO and JSW analyzed the data. BSC, MFT, and JSW contributed reagents/materials/analysis tools. LZO and JSW wrote the paper.

Academic Editor: Greg Petsko, Brandeis University


Note Added in Proof


A recent study of the specificity-determining region of Sup35p prion (Hara et al. 2003) describes complementary results.

Hara H, Nakayashiki T, Crist CG, Nakamura Y (2003) Prion domain interaction responsible for species discrimination in yeast [PSI
+] transmission. Genes Cells 8:925–939.

Abbreviations
Ade+adenine prototrophy

Ccarboxy-terminal

GFPgreen fluorescent protein

GuHClguanidine hydrochloride

Mmiddle

Namino-terminal

NYNasparagine-tyrosine-asparagine

Q/Nglutamine/asparagine

PNMPSI No More

SDsynthetic defined (yeast medium with dextrose)

YEPDyeast extract-peptone-dextrose
==== Refs
References
Balbirnie M  Grothe R  Eisenberg DS   An amyloid-forming peptide from the yeast prion Sup35 reveals a dehydrated beta-sheet structure for amyloid Proc Natl Acad Sci U S A 2001 98 2375 2380 11226247 
Borchsenius AS  Wegrzyn RD  Newnam GP  Inge-Vechtomov SG  Chernoff YO   Yeast prion protein derivative defective in aggregate shearing and production of new “seeds.” EMBO J 2001 20 6683 6691 11726504 
Chernoff YO  Lindquist SL  Ono B  Inge-Vechtomov SG  Liebman SW   Role of the chaperone protein Hsp104 in propagation of the yeast prion-like factor [psi+ ] Science 1995 268 880 884 7754373 
Chernoff YO  Galkin AP  Lewitin E  Chernova TA  Newnam GP    Evolutionary conservation of prion-forming abilities of the yeast Sup35 protein Mol Microbiol 2000 35 865 876 10692163 
Chien P  Weissman JS   Conformational diversity in a yeast prion dictates its seeding specificity Nature 2001 410 223 227 11242084 
Chien P  DePace AH  Collins SR  Weissman JS   Generation of prion transmission barriers by mutational control of amyloid conformations Nature 2003 424 948 951 12931190 
Cohen FE  Prusiner SB   Pathologic conformations of prion proteins Annu Rev Biochem 1998 67 793 819 9759504 
Cox BS   ψ, a cytoplasmic suppressor of super-suppressor in yeast Heredity 1965 20 505 521 
Cox BS  Ness F  Tuite MF   Analysis of the generation and segregation of propagons: Entities that propagate the [PSI+ ] prion in yeast Genetics 2003 165 23 33 14504215 
Crist CG  Nakayashiki T  Kurahashi H  Nakamura Y   [PHI+ ], a novel Sup35-prion variant propagated with non-Gln/Asn oligopeptide repeats in the absence of the chaperone protein Hsp104 Genes Cells 2003 8 603 618 12839621 
DePace AH  Santoso A  Hillner P  Weissman JS   A critical role for amino-terminal glutamine/asparagine repeats in the formation and propagation of a yeast prion Cell 1998 93 1241 1252 9657156 
Derkatch IL  Bradley ME  Zhou P  Chernoff YO  Liebman SW   Genetic and environmental factors affecting the de novo appearance of the [PSI+ ] prion in Saccharomyces cerevisiae 
 Genetics 1997 147 507 519 9335589 
Derkatch IL  Bradley ME  Zhou P  Liebman SW   The PNM2 mutation in the prion protein domain of SUP35 has distinct effects on different variants of the [PSI+ ] prion in yeast Curr Genet 1999 35 59 67 10079323 
Derkatch IL  Bradley ME  Hong JY  Liebman SW   Prions affect the appearance of other prions: The story of [PIN(+) ] Cell 2001 106 171 182 11511345 
Dobson CM   Protein misfolding, evolution and disease Trends Biochem Sci 1999 24 329 332 10470028 
Doel SM  McCready SJ  Nierras CR  Cox BS   The dominant PNM2-mutation which eliminates the psi factor of Saccharomyces cerevisiae  is the result of a missense mutation in the SUP35 gene Genetics 1994 137 659 670 8088511 
Glover JR  Kowal AS  Schirmer EC  Patino MM  Liu JJ    Self-seeded fibers formed by Sup35, the protein determinant of [PSI+ ], a heritable prion-like factor of S. cerevisiae 
 Cell 1997 89 811 819 9182769 
Jung G  Jones G  Masison DC   Amino acid residue 184 of yeast Hsp104 chaperone is critical for prion-curing by guanidine, prion propagation, and thermotolerance Proc Natl Acad Sci U S A 2002 99 9936 9941 12105276 
King CY   Supporting the structural basis of prion strains: Induction and identification of [PSI ] variants J Mol Biol 2001 307 1247 1260 11292339 
King CY  Tittmann P  Gross H  Gebert R  Aebi M    Prion-inducing domain 2–114 of yeast Sup35 protein transforms in vitro into amyloid-like filaments Proc Natl Acad Sci U S A 1997 94 6618 6622 9192614 
Kitada K  Yamaguchi E  Arisawa M   Cloning of the Candida glabrata  TRP1 and HIS3 genes, and construction of their disruptant strains by sequential integrative transformation Gene 1995 165 203 206 8522176 
Kochneva-Pervukhova NV  Paushkin SV  Kushnirov VV  Cox BS  Tuite MF    Mechanism of inhibition of Psi+ prion determinant propagation by a mutation of the N-terminus of the yeast Sup35 protein EMBO J 1998 17 5805 5810 9755180 
Koo EH  Lansbury PT  Kelly JW   Amyloid diseases: Abnormal protein aggregation in neurodegeneration Proc Natl Acad Sci U S A 1999 96 9989 9990 10468546 
Krobitsch S  Lindquist S   Aggregation of huntingtin in yeast varies with the length of the polyglutamine expansion and the expression of chaperone proteins Proc Natl Acad Sci U S A 2000 97 1589 1594 10677504 
Kryndushkin DS  Alexandrov IM  Ter-Avanesyan MD  Kushnirov VV   Yeast [PSI+ ] prion aggregates are formed by small Sup35 polymers fragmented by Hsp104 J Biol Chem 2003 278 49636 49643 14507919 
Kushnirov VV  Ter-Avanesyan MD   Structure and replication of yeast prions Cell 1998 94 13 16 9674422 
Kushnirov VV  Kochneva-Pervukhova NV  Chechenova MB  Frolova NS  Ter-Avanesyan MD   Prion properties of the Sup35 protein of yeast Pichia methanolica 
 EMBO J 2000 19 324 331 10654931 
Lacroute F   Non-Mendelian mutation allowing ureidosuccinic acid uptake in yeast J Bacteriol 1971 106 519 522 5573734 
Liebman SW  Derkatch IL   The yeast [PSI+ ] prion: Making sense of nonsense J Biol Chem 1999 274 1181 1184 9880481 
Liu JJ  Lindquist S   Oligopeptide-repeat expansions modulate “protein-only” inheritance in yeast Nature 1999 400 573 576 10448860 
Liu JJ  Sondheimer N  Lindquist SL   Changes in the middle region of Sup35 profoundly alter the nature of epigenetic inheritance for the yeast prion [PSI+ ] Proc Natl Acad Sci U S A 99 2002 (Suppl 4) 16446 16453 
Masison DC  Wickner RB   Prion-inducing domain of yeast Ure2p and protease resistance of Ure2p in prion-containing cells Science 1995 270 93 95 7569955 
McCready SJ  Cox BS  McLaughlin CS   The extrachromosomal control of nonsense suppression in yeast: An analysis of the elimination of [psi+ ] in the presence of a nuclear gene PNM Mol Gen Genet 1977 150 265 270 321935 
Meriin AB  Zhang X  He X  Newnam GP  Chernoff YO    Huntington toxicity in yeast model depends on polyglutamine aggregation mediated by a prion-like protein Rnq1 J Cell Biol 2002 157 997 1004 12058016 
Michelitsch MD  Weissman JS   A census of glutamine/asparagine-rich regions: Implications for their conserved function and the prediction of novel prions Proc Natl Acad Sci U S A 2000 97 11910 11915 11050225 
Nakayashiki T  Ebihara K  Bannai H  Nakamura Y   Yeast [PSI+ ] “prions” that are crosstransmissible and susceptible beyond a species barrier through a quasi-prion state Mol Cell 2001 7 1121 1130 11430816 
Ness F  Ferreira P  Cox BS  Tuite MF   Guanidine hydrochloride inhibits the generation of prion “seeds” but not prion protein aggregation in yeast Mol Cell Biol 2002 22 5593 5605 12101251 
Osherovich LZ  Weissman JS   Multiple Gln/Asn-rich prion domains confer susceptibility to induction of the yeast [PSI(+) ] prion Cell 2001 106 183 194 11511346 
Osherovich LZ  Weissman JS   The utility of prions Dev Cell 2002 2 143 151 11832240 
Osherovich LZ  Weissman JS   Buchner J   Insights into the structure of yeast prions Protein folding handbook 2004 New York Wiley and Sons 
Parham SN  Resende CG  Tuite MF   Oligopeptide repeats in the yeast protein Sup35p stabilize intermolecular prion interactions EMBO J 2001 20 2111 2119 11331577 
Patino MM  Liu JJ  Glover JR  Lindquist S   Support for the prion hypothesis for inheritance of a phenotypic trait in yeast Science 1996 273 622 626 8662547 
Paushkin SV  Kushnirov VV  Smirnov VN  Ter-Avanesyan MD   Propagation of the yeast prion-like [psi+ ] determinant is mediated by oligomerization of the SUP35-encoded polypeptide chain release factor EMBO J 1996 15 3127 3134 8670813 
Prusiner SB   Prions Proc Natl Acad Sci U S A 1998 95 13363 13383 9811807 
Resende CG  Outeiro TF  Sands L  Lindquist S  Tuite MF   Prion protein gene polymorphisms in Saccharomyces cerevisiae 
 Mol Microbiol 2003 49 1005 1017 12890024 
Santoso A  Chien P  Osherovich LZ  Weissman JS   Molecular basis of a yeast prion species barrier Cell 2000 100 277 288 10660050 
Schlumpberger M  Prusiner SB  Herskowitz I   Induction of distinct [URE3 ] yeast prion strains Mol Cell Biol 2001 21 7035 7046 11564886 
Sherman F   Getting started with yeast Meth Enzymol 1991 194 3 21 2005794 
Sikorski RS  Hieter P   A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae 
 Genetics 1989 122 19 27 2659436 
Sondheimer N  Lindquist S   Rnq1: An epigenetic modifier of protein function in yeast Mol Cell 2000 5 163 172 10678178 
Song JM  Liebman SW   Interaction of UAG suppressors and omnipotent suppressors in Saccharomyces cerevisiae 
 J Bacteriol 1985 161 778 780 3881411 
Taylor KL  Cheng N  Williams RW  Steven AC  Wickner RB   Prion domain initiation of amyloid formation in vitro from native Ure2p Science 1999 283 1339 1343 10037606 
Ter-Avanesyan MD  Kushnirov VV  Dagkesamanskaya AR  Didichenko SA  Chernoff YO    Deletion analysis of the SUP35 gene of the yeast Saccharomyces cerevisiae  reveals two non-overlapping functional regions in the encoded protein Mol Microbiol 1993 7 683 692 8469113 
Uptain SM  Lindquist S   Prions as protein-based genetic elements Annu Rev Microbiol 2002 56 703 741 12142498 
Wickner RB   [URE3 ] as an altered URE2 protein: Evidence for a prion analog in Saccharomyces cerevisiae 
 Science 1994 264 566 569 7909170 
Zadorskii SP  Sopova IV  Inge-Vechtomov SG   Prionization of the Pichia methanolica  SUP35 gene product in the yeast Saccharomyces cerevisiae 
 Genetika 2000 36 1322 1329 11094743 
Zoghbi HY  Orr HT   Glutamine repeats and neurodegeneration Annu Rev Neurosci 2000 23 217 247 10845064

