INVESTIGATION

[PSI+] Prion Transmission Barriers ProtectSaccharomyces cerevisiae from Infection:

Intraspecies ’Species Barriers’David A. Bateman and Reed B. Wickner1

Laboratory of Biochemistry and Genetics, National Institute of Diabetes and Digestive and Kidney Diseases,National Institutes of Health, Bethesda, Maryland 20892-0830

ABSTRACT [PSI+] is a prion of Sup35p, an essential translation termination and mRNA turnover factor. The existence of lethal [PSI+]variants, the absence of [PSI+] in wild strains, the mRNA turnover function of the Sup35p prion domain, and the stress reaction to prioninfection suggest that [PSI+] is a disease. Nonetheless, others have proposed that [PSI+] and other yeast prions benefit their hosts. Wefind that wild Saccharomyces cerevisiae strains are polymorphic for the sequence of the prion domain and particularly in the adjacentM domain. Here we establish that these variations within the species produce barriers to prion transmission. The barriers are partiallyasymmetric in some cases, and evidence for variant specificity in barriers is presented. We propose that, as the PrP 129M/V poly-morphism protects people from Creutzfeldt–Jakob disease, the Sup35p polymorphisms were selected to protect yeast cells from prioninfection. In one prion incompatibility group, the barrier is due to N109S in the Sup35 prion domain and several changes in the middle(M) domain, with either the single N109S mutation or the group of M changes (without the N109S) producing a barrier. In another, thebarrier is due to a large deletion in the repeat domain. All are outside the region previously believed to determine transmissioncompatibility. [SWI+], a prion of the chromatin remodeling factor Swi1p, was also proposed to benefit its host. We find that none of 70wild strains carry this prion, suggesting that it is not beneficial.

PRIONS are infectious proteins, with no nucleic acid re-quired for transmission. Most prions are self-propagating

amyloids, b-sheet-rich filamentous polymers of a singleprotein. The mammalian prion causes a uniformly fatal trans-missible spongiform encephalopathy (TSE) based on amyloidof PrP, normally a nonessential cell surface GPI-anchoredprotein that becomes toxic in amyloid form (reviewed inCaughey et al. 2009). Interspecies transmission of TSEs oc-curs with difficulty, or not at all, as a result of sequence differ-ences between the prion proteins of donor and recipient,a phenomenon called the “species barrier” (Prusiner et al.1990; Bruce 2003).

Human TSEs include the spontaneous form (Creutzfeldt–Jakob disease, CJD) and an infectious form, most famouslyan epidemic (called Kuru) among the Fore people of New

Guinea due to a ritual mortuary cannibal custom (reviewedin Collinge and Alpers 2008). The human gene encoding PrPhas a polymorphism with about half of alleles encoding Metat residue 129 and half Val at that site. Remarkably, onlyrarely do patients heterozygous at this site develop any formof CJD, although both M/M and V/V people are susceptible.It has been suggested that this protective polymorphism wasselected to avoid the devastating effects of cannibalismwhen that phenomenon was more common (Mead et al.2003). Indeed, the Kuru epidemic selected a new resistantallele at residue 127 (Mead et al. 2009b).

The [PSI+] prion (infectious protein) is a self-propagatingamyloid form of Sup35p, a subunit of the translation termi-nation factor (Wickner 1994; King and Diaz-Avalos 2004;Tanaka et al. 2004), reviewed in Wickner et al. (2010).[URE3] is a prion of Ure2p, a regulator of nitrogen catabolism(Wickner 1994), likewise based on a self-propagating amy-loid (Brachmann et al. 2005), and [SWI+] is a prion of thechromatin remodeling factor Swi1p (Du et al. 2008). Each ofthese prions produces decreased function of the respectiveprotein. Sup35p consists of an N-terminal Q/N-rich domain(N; residues 1–123), a highly charged middle (M; residues

Copyright © 2012 by the Genetics Society of Americadoi: 10.1534/genetics.111.136655Manuscript received October 24, 2011; accepted for publication November 11, 2011Supporting information is available online at http://www.genetics.org/content/suppl/2011/11/18/genetics.111.136655.DC1.1Corresponding author: National Institutes of Health, Bldg. 8, Room 225, 8 Center Dr.MSC 0830, Bethesda, MD 20892-0830. E-mail: wickner@helix.nih.gov

Genetics, Vol. 190, 569–579 February 2012 569

124–253) domain, and a C-terminal domain (C; residues254–685). Sup35NM functions normally in the generalmRNA degradation system, linking translation terminationto mRNA lifetime (Hoshino et al. 1999; Hosoda et al. 2003;Funakoshi et al. 2007), and is necessary and largely sufficientfor prion generation and propagation (Teravanesyan et al.1994; King 2001; King and Diaz-Avalos 2004). However,some variant-specific information requires sequence to resi-due 137 (Bradley and Liebman 2004). Alterations in Sup35Mcan also affect prion propagation in subtle ways, but deletionof Sup35M is reported not to interfere with [PSI+] propaga-tion (Liu et al. 2002). Sup35C carries out the essential trans-lation termination function of the protein (Teravanesyanet al. 1994; Stansfield et al. 1995; Zhouravleva et al. 1995).

A single prion protein sequence can stably propagate anyof several amyloid conformations, resulting in different“prion variants” or “prion strains.” Thus, different isolatesof sheep scrapie (a TSE) propagated in mice can show dra-matically different incubation periods and different distribu-tion of pathology in the brain, even though the sequence ofthe infecting prion protein and that of the infected animalare identical in each prion strain. In yeast, prion variantsmay differ in the intensity of the prion phenotype, stabilityof prion propagation, response to overproduction or defi-ciency of chaperones, or differing ability to overcome a spe-cies barrier (Derkatch et al. 1996; Chernoff et al. 1999;Kushnirov et al. 2000b; Edskes et al. 2009). Recently, prionvariants of [PSI+] and [URE3] that are severely toxic oreven lethal have also been found (Mcglinchey et al. 2011).The sequence of the prion protein is the same in differentprion variants, but the conformation of the protein in theamyloid is different (Bessen and Marsh 1994; Tanaka et al.2006), and this conformation is propagated faithfully as thefilaments elongate, are severed to make new filaments, andare distributed to daughter cells.

The mechanism by which a single protein can faithfullytransmit its conformation to another molecule of the same issuggested by the architecture of yeast prions. Infectiousamyloid of the prion domains of Ure2p, Sup35p, and Rnq1p(the latter the basis of the [PIN+] prion) each have an in-register parallel b-sheet architecture with multiple folds inthe b-sheet along the long axis of the filaments (Shewmakeret al. 2006; Baxa et al. 2007; Wickner et al. 2008a). Thistype of amyloid, also typical of most pathological humanamyloids (Tycko 2011), is characterized by lines of identicalamino acid side chains running the length of the filament.The in-register structure is enforced by the favorable inter-actions between these aligned identical side chains that arepossible only if the molecules are in-register. Hydrogenbonds between the side-chain amides of glutamines or as-paragines produce a great stabilization of the structure, andserine or threonine residues can likewise form a chain ofH-bonds with their side chain2OH group. Hydrophobic sidechains can also have interactions that promote in-registeralignment. Of course, a line of charged residues would beunfavorable for in-register structure and charged residues

are few in these prion domains. Prion variants are proposedto differ in the location of the folds of the sheet (i.e., theturns of the chain). The same favorable interactions of iden-tical side chains that enforce the in-register architecturewould force new molecules to assume the same conforma-tion as molecules already in the chain, thus explaining thefaithful propagation of prion variants (Wickner et al. 2007,2008b, 2010).

The prion proteins each have a discrete region needed forprion propagation that corresponds roughly to the part ofthe protein that actually forms amyloid (Teravanesyan et al.1994; Masison and Wickner 1995; King et al. 1997; Tayloret al. 1999). The prion domain of Sup35p was first defined(Teravanesyan et al. 1994) as the N-terminal residues 1–114necessary to propagate the original [PSI+] variant (Cox1965). Later studies showed that residues 1–137 are suffi-cient to propagate any of several [PSI+] variants (Bradleyand Liebman 2004), although several variants were faith-fully propagated by 1–64 (Shkundina et al. 2006) or 1–61(King and Diaz-Avalos 2004). The prion domains of Sup35pof several other Saccharomyces species can form [PSI+]in cerevisiae, but because of sequence differences with S.cerevisiae Sup35p and with each other, there are species bar-riers to transmission of [PSI+] (Chen et al. 2007; Afanasievaet al. 2011), although which differences are responsible forthis barrier have not yet been defined. Mutational studies ofthe S. cerevisiae prion domain have shown that transmissionbarriers can be produced by mutations in Sup35p residues1–33, a very Q/N-rich part of the protein (Depace et al.1998). Solid-state NMR studies of infectious amyloid ofSup35NM showed in-register parallel b-sheet structure ofessentially all of N (1–123) and suggested that this structureextends partially into the M domain (residues 124–253)(Shewmaker et al. 2006, 2009).

The N and M domains of Sup35p are known to be farmore variable between yeast species than the C domains(Kushnirov et al. 1990, 2000a; Santoso et al. 2000). Previ-ous studies of SUP35 sequences from wild S. cerevisiaestrains have revealed several variable sites, including codingchanges Q30R, A34T, A42G, Y55H, Q90H, N109S, V147F,G152C, G162D, D169E, T206K, and H255D and a 19-amino-acid deletion in the oligopeptide repeats (Jensenet al. 2001; Resende et al. 2003; Supporting Information,Table S4). This deletion was reported to alter the expressionof [PSI+] without causing its loss (Resende et al. 2003), butother polymorphisms have not been tested for their effectson prion transmission or prion-forming ability.

We have examined the sequence, prion-forming ability,and susceptibility to prion infection of SUP35 in wild iso-lates of S. cerevisiae. We observe considerable polymor-phism, with variations particularly in the N and Mdomains. While all alleles tested can form prions, we showthat there are strong transmission barriers between differ-ent Sup35p’s. Surprisingly, residues in the C-terminal partof Sup35N and in M contribute to these transmission bar-riers. We also find that [SWI+] is not found in our

570 D. A. Bateman and R. B. Wickner

collection of wild strains, suggesting that it is not adaptivefor yeast.

Materials and Methods

Nomenclature

We refer to the standard laboratory yeast sequence (Kikuchiet al. 1988; Kushnirov et al. 1988; Wilson and Culbertson1988) as the “reference sequence,” and that of differing wildisolates are called “polymorphs.” A prion originating withthe Sup35p sequence of strain G2, for example, but beingpropagated in a strain expressing only the reference se-quence will be designated [PSI+G2]ref, in analogy with sim-ilar nomenclature for [URE3] (Edskes et al. 2009).

Scoring the [PSI+] prion

Sup35p is a subunit of the translation termination complex,and the incorporation of a large proportion of Sup35p intothe prion amyloid filaments makes it inactive, resulting inincreased read-through of termination codons. This is mea-sured by read-through of ade2-1, with an ochre terminationcodon in the middle of the ADE2 gene. In addition to ade2-1,strains carry the SUQ5 weak suppressor mutation, whichleaves cells Ade2 unless the [PSI+] prion is also present.

Strains and media

The strains used are listed in Table 1 (laboratory strains)and Table S1 (wild strains). All yeast media and plates con-tained 20 mM copper sulfate unless noted. Rich and minimalmedia (YPAD and SD) are as described (Sherman 1991).Required nutrients were added to minimal plates.

Sequencing the SUP35 of wild strains

Genomic yeast DNA was isolated using the QAIprepspin miniprep kit. Dilutions of 1:10 and 1:100 were madeand PCR was performed using Platinum PCR SuperMix(Invitrogen) with primer DB082 with either DB051 orDB089 (Table S2) for S. cerevisiae strains; primers DB093and DB094 for S. paradoxus; and primers DB095 and DB119with S. bayanus or S. pastorianus strains. PCR products werecleaned using the QIAquick PCR purification kit (Qiagen)and sequenced (University of Maryland Sequencing facility,UMBI). Sequences were assembled and aligned, and fastaformat output files were generated for resulting DNA andtranslated protein sequences using Codon Code Alignerv. 3.7.1 (CodonCode Corp.). Translated protein sequenceswere aligned to reference sequences using ClustalW (Net-

work Protein Sequence Analysis), to determine codingpolymorphisms.

Wild-strain SUP35 plasmid constructions

Using primers DB082 and DB085, the full SUP35 open readingframe was amplified from the wild strains indicated (in paren-theses) and cloned between the Bam1H and XhoI sites of pDB03to generate pDB73 (strain F9, D19) and pDB86 (strain F8), orinto pDB66 replacing Sup35C–GFP to generate pDB88 (strainA9), pDB89 (strain E9), pDB91 (strain F7), and pDB93 (strainG2). A list of all plasmids is included in Table 2. Plasmids weretransformed into strain 4828, p1215 loss was selected bygrowth on 5-fluoroorotic acid (5-FOA) media, and clonesthat were Ade2 and red on 1/2 YPD were picked. Prionswere selected by plating cells at high density on media lack-ing adenine or by induction with overnight growth with 160mM copper in liquid YPAD media to overproduce Sup35pand plating on media lacking adenine. Colonies that wereAde+ were streaked on 1/2 YPD and then mated witha lawn of strain 4830 to determine dominance. Isolates thatproved to be dominant were streaked on 1/2 YPD contain-ing 5 mM guanidine hydrochloride to isolate cured strains.

Generation of 109 mutants

Using PCR with primers DB082, DB129, DB130, and DB133on plasmid pDB89 (strain E9) and pDB93 (strain G2), weconverted the serine 109 residue back to the referencesequence asparagine, generating plasmids pDB102 (strainE9) and pDB103 (strain G2). Using primers DB082, DB131,DB132, and DB133 with plasmid pDB08 (reference Sup35sequence), the point mutation asparagine to serine at posi-tion 109 was made, generating plasmid pDB101.

Cytoduction

Cytoplasm may be transformed from one strain to anotherutilizing the kar1-1 mutation (Conde and Fink 1976), de-fective for nuclear fusion. Cells fuse, but the nuclei do notfuse, and nuclei separate at the next cell division. However,cytoplasmic mixing has occurred, and so a genetic element(prion or mitochondrial DNA) present in one strain (identi-fied by its nuclear genotype) will be transferred to the other.We use transfer of mitochondrial DNA as a marker of cyto-plasmic transfer, and score prion transfer. Wild-strain se-quence plasmids were transformed into strain 4830, lossof p1215 was selected by growth on 5-FOA media, andAde2 transformants were made rhoo by growth on YPADcontaining 1 mg/ml ethidium bromide. Donor and recipient

Table 1 Laboratory strains of S. cerevisiae

Strain no. Genotype

4828 MATa ade 2-1 SuQ5 trp1 kar1-1 his3 leu2 ura3 sup35::kanMX [PIN+] [psi2] p1215: CEN URA3 SUP35MC4830 MATa ade 2-1 SUQ5 trp1 kar1-1 lys2 leu2 ura3 sup35::kanMX [pin2] [psi2] p1215: CEN URA3 SUP35MC779-6a MATa ade 2-1 SUQ5 trp1 kar1-1 his3 leu2 ura3 [PSI+] (Jung and Masison 2001)74D-694 MATa ade 1-14 trp1-289 his3-200 leu2-3 ura3-52 [swi2]74D-694 SWI+ MATa ade 1-14 trp1-289 his3-200 leu2-3 ura3-52 [SWI+]

The entire ORF of SUP35 is deleted in strains 4828 and 4830, and substituted with the kanMX gene conferring resistance to G418.

Natural Barriers to [PSI+] Transmission 571

strains were mixed in water at high density and 0.1 ml wasspotted onto a YPAD plate. After 18 hr at room temperature,cells were streaked for single colonies on media selectiveagainst the donor strains. Clones are shown to be cytoduc-tants by their growth on glycerol and failure to grow onmedia selective for diploids. As further tests of a sampleconfirm, Ade+ cytoductants are judged to have receivedand propagated [PSI+].

Analysis of [SWI+] in wild strains

SWI1 was amplified from p416TEFSWI1NQ-YFP (Du et al.2008) by PCR using primers DB087 and DB097 and clonedbetween the BamHI and XhoI sites of pDB64 (Nakayashikiet al. 2005) resulting in pDB67 (CEN KanMX SWI1–GFP).This plasmid was sequenced and transformed into wildstrains using 0.3 mg/ml G418 (Mediatech Inc., Herndon,VA) for plasmid selection and maintenance. Cells were im-aged using a Nikon Eclipse TE2000-U spinning disc confocalmicroscope with 100· NA 1.4 Nikon oil lens with 1.5· mag-nifier and captured with a Hamamatsu EM-CCD ImagEMdigital camera with IPLab version 4.08. The [SWI+] pheno-type was assessed by streaking each wild strain on YPAD,YPARaf (1% yeast extract, 2% peptone, 0.04% adenine sul-fate, 2% raffinose), YPG (1% yeast extract, 2% peptone, 2%glycerol), and synthetic medium with 2% raffinose (SRaf).

Results

Sup35 sequences in wild S. cerevisiae strains

Among the SUP35 genes of 55 wild S. cerevisiae strains (Ta-ble S1) (Nakayashiki et al. 2005) sequenced in this study, 13

infrequent mutations were found in the C domain (residues254–685), each observed in 5 or fewer strains (Table 3 andTable S3). The M domain (residues 124–253) had a similarnumber of mutations per residue (8/129 �13/431), but ata much higher frequency in the wild strains (Figure 1). Allbut two strains had the G162D mutation, and other commonmutations within the M domain include the D169E mutationfound at 22%, P186A mutation at 13%, T206K at 45%, andH225D found at 27% of the S. cerevisiae population sample.Very few mutations within the N domain (1–123) were ob-served (Table 3 and Table S3); however, the N109S muta-tion was observed in 31% of the population.

Only the N and M domains of the non-cerevisiae SUP35genes were sequenced. The SUP35 genes of four wild S.bayanus strains had only one common N domain (1–116)mutation P4S, and no M domain mutations compared to thereference sequence (Cliften et al. 2003) (Table 3 and TableS3). Five sequenced SUP35 genes from S. pastorianus wildstrains were sequenced and compared to the referenceS. pastorianus sequence (Nakao et al. 2009)). Of thesesequences the N5I and N8I mutations were found with high-est frequency among the five N domain (1–116) mutationsand two M domain (117–242) mutations (Table 3 and TableS3). For S. paradoxus, only two mutations were observedcompared to the reference sequence (Kellis et al. 2003),Y29S and L84P within the N domain (Table 3 and Table S3).

Three major transmission compatibility groups

We found the common N109S, G162D, D169E, P186A,T206K, and H225D mutations most interesting within ourwild S. cerevisiae strains (Table 3 and Table S3), since all

Table 2 Plasmids used in this study

Plasmid Name Description Origin

pDB08 pH 953 CEN SUP35 LEU2 SUP35 promoter H. E. EdskespDB09 p1215 CEN SUP35MC URA3 SUP35 promoter This studypDB03 p1147 CEN LEU2 ADH1 promoter This studypDB22 pH 610 2m TRP1 GAL1 promoter H. E. EdskespDB11 pH 952 2m SUP35NM TRP1 GAL1 promoter H. E. EdskespDB61 p416TEFSWI1NQ–YFP CEN SWI1(1–536)–YFP URA3 TEF promoter Du et al. (2008)pDB64 pKanMXURE2N–GFP (#210) CEN URE2–GFP KanMX ADH promoter Nakayashiki et al. (2005)pDB65 pCUP1SUP35NM–GFP (#211) CEN SUP35NM-GFP KanMX CUP1 promoter Nakayashiki et al. (2005)pDB66 Cup10 CEN SUP35C–GFP LEU2 CUP1 promoter This studypDB67 Swi-E CEN SWI1(1–536)–GFP KanMX ADH promoter This studypDB73 D19-1147 CEN SUP35 (from strain F9) LEU2 ADH promoter This studypDB81 A8-210 CEN SUP35NM (strain A8) KanMX ADH promoter This studypDB82 F9-210 CEN SUP35NM (strain F9) KanMX ADH promoter This studypDB86 F8-1147 CEN SUP35 (strain F8) LEU2 ADH promoter This studypDB88 A9-Cup10 CEN SUP35 (strain A9) LEU2 CUP1 promoter This studypDB89 E9-Cup10 CEN SUP35 (strain E9) LEU2 CUP1 promoter This studypDB91 F7-Cup10 CEN SUP35 (strain F7) LEU2 CUP1 promoter This studypDB93 G2-Cup10 CEN SUP35 (strain G2) LEU2 CUP1 promoter This studypDB97 E9-610 2m SUP35NM (strain E9) TRP1 GAL1 promoter This studypDB99 F9-610 2m SUP35NM (strain F9) TRP1 GAL1 promoter This studypDB100 G2-610 2m SUP35NM (strain G2) TRP1 GAL1 promoter This studypDB101 Ref-N109S CEN SUP35 N109S LEU2 SUP35 promoter This studypDB102 E9-S109N CEN SUP35 (strain E9) S109N LEU2 SUP35 promoter This studypDB103 G2-S109N CEN SUP35 (strain G2) S109N LEU2 SUP35 promoter This study

All plasmids carry AmpR.

572 D. A. Bateman and R. B. Wickner

these mutations lie well outside the region (residues 1–33)commonly proposed to determine the transmission barrierbetween Sup35 molecules with different sequences (Depaceet al. 1998). We transformed laboratory strain 4828[sup35D p(URA3 SUP35C)] with plasmids generated fromthese wild S. cerevisiae SUP35 genes, selected cells that hadlost the URA3 plasmid on FOA, and named these laboratorystrains with the corresponding wild-strain code. This gener-ated strains A9 and F7 (both G162D, T206K), E9 and A8(both N109S, G162D, D169E, P186A, T206K, H225D), G2(N109S, G162D, D169E, T206K, H225D), and F8 (N109S,G162D, D169E, P186A, T206K, E216D, H225D, Y467H), aslisted in Table 3 and Table S3. The D19 allele, found in

strains B11 and F9, lacked residues 59–77 and had theG162D mutation (Table 3 and Table S3). These strains weretested to determine if [PSI+] could be transmitted into thesemutants from a strain with the reference sequence. Cytoduc-tion, a transfer of cytoplasm between strains without transferof DNA plasmids or nuclei, was used in this test as the closestparallel to mammalian transmission events. Because the donorand recipient Sup35 molecules are only transiently coex-pressed, this type of experiment gives the best opportunity todetermine transmission barriers, with the least likelihood ofconfusion resulting from prion variant changes being selected.

Data in Table 4 show that there is no transmission barrierbetween the reference sequence and recipients A9 or F7,

Table 3 Polymorphs of Saccharomyces Sup35 proteins

Strains of S. cerevisiae Mutations

B2, C4 NoneA5, A7, A10, A11, B1, B3, B5, B6, B7, B10, C1, C3,C5, E2, E4, E6, E11, F1

G162D

B11, F9 D59-77, G162DE12 Q14H, G162DE8 Q10H, G162DF11 N109S, G162DF8 N109S, G162D, D169E, P186A, T206K, E216D, H255D, Y467HA1, A8, B12, E9, F10, F12, N109S, G162D, D169E, P186A, T206K, H225DG2 N109S, G162D, D169E, T206K, H225DE1 N109S, G162D, D169E, Q201H, T206K, H225DE3 N109S, G162D, D169E, Q201H, T206K, H225D, N426YA4 N109S, G162D, D169E, T206K, H225DF4 N109S, G162D, D169E, T206K, H225D, S428R, K439MF3 N109S, G162D, D169E, T206K, H225D, A676S, K679EC2 N109S, G162D, P186A, T206K, H225D, K324PE7 N109S, G162D, T206K, H225DE10 N109S, G162D, T206K, H225D, G678DA9, B9, F7 G162D, T206KA6 G162D, T206K, V402A, S428R, G449AA2, F5 G162D, T206K, V402AB8, F2 G162D, T206K, V402A, N480TA3 G162D, T206K, E560GA12 G162D, T206K, S210LB4 G162D, A676S, G678DF6 G162D, N426YE5 G162D, K324T, K679EG1 G162D, K679E

Strains of S. bayanus MutationsD4 P4SD8 P4SD9 P4SD10 NA

Strains of S. pastorianus MutationsD6 N5I, N8ID7 N5I, N8I, Q41H, T161I, D225ND12 P4S, N5I, N8I, A54SD5 N5I, N8ID11 N8I

Strains of S. paradoxus MutationsC7, NoneC6, C8, C9, C10, 11C, D1, D3 L84PD2 Y29S, L84PC12 Y29S

Natural Barriers to [PSI+] Transmission 573

indicating that the nearly universal mutation G162D doesnot produce a transmission barrier. However, there is a sub-stantial barrier to transmission to strains that have eitherthe D19 mutation or strains with the N109S and multipleM-domain mutations (Table 4, recipients A8, E9, F8, G2,D19). The cytoductants were isolated and used as donorsto [psi2] cells with the reference Sup35 sequence, to con-firm that Ade2 was due to loss of [PSI+] and not, for ex-ample, simply a decrease in the amount of amyloid such thatthere was no prion phenotype. The D19 Ade2 donors didnot transmit [PSI+] back to a strain with the reference se-quence confirming that [PSI+ref] was indeed lost in theD19 host (Table 5). The same result was obtained withthe F8 and G2 Ade2 donors, the latter differing from thereference sequence only by N109S and M-domain changes.The E9 donors, differing in sequence slightly from both G2and F8, showed some retention of infectivity for the refer-ence sequence, indicating that it was losing, but had notcompletely lost [PSI+].

De novo [PSI+] derivatives of strains expressing the poly-morphic sequences were generated for both the D19 and theG2 strains by transiently overexpressing the same polymor-phic Sup35p and selecting Ade+ clones. These Ade+ cloneswere shown to be [PSI+] by their being dominant, by beingefficiently cured by growth on rich medium containing 5mM guanidine, and by the appearance of fluorescent fociwhen transformed with plasmids expressing the correspond-ing Sup35NM fused to GFP (Figure 2). This conclusion wasconfirmed by the efficient transmission of [PSI+] by cyto-duction to cells carrying the same SUP35 allele (Table 6).Cytoductions between these new [PSI+] isolates and cellsexpressing other Sup35p polymorphs, along with data inTable 4, indicate that there are at least three main [PSI+]incompatibility groups (Table 6). Cytoduction from [PSI+]arising in any polymorph into a [psi2] cell expressing the samepolymorph was efficient, but cytoduction into a [psi2] recipi-ent expressing a different polymorph was quite inefficient—sometimes completely blocked. [PSI+D19] propagated poorlyin cells expressing another polymorph, most likely due toa structural rearrangement of the NM domain due to sucha large loss of amino acids within this domain. Similarly,[PSI+G2] propagated best with other N109S containing mu-tant strains, although some transmission (�30%) was ob-served to the A9, F7, reference, and D19 sequences.

We transformed the [PSI+ref]ref, [PSI+G2]G2, and[PSI+D19]D19 strains with the various polymorph NM–GFP

vectors. The refNM–GFP produced dots in each of the hosts(Figure 2), indicating that there is some interaction of theprion domains of the reference Sup35 with each polymorphamyloid, even though those interactions did not generally leadto prion propagation. At the other extreme, the G2NM–GFPshowed dots only in the homologous host, showing that theseinteractions are asymmetric. The D19NM–GFP showed dotswith the homologous amyloid, as well as with the referenceamyloid (Figure 2). These results indicate some sequencespecificity in association with the NM–GFP fusions with amy-loids, with a preference for the same sequence.

Asymmetry and prion variant specificityof transmission barriers

Although [PSI+G2]G2 was transmitted to 5 of 25 cytoduc-tants expressing the reference sequence, [PSI+ref]ref wasnot transmitted to any of 86 cytoductants expressing theG2 polymorph (Tables 4 and 6). Although [PSI+G2]G2

was successfully transmitted in 8 of 22 cytoduction eventsinto a strain expressing D19, [PSI+D19]D19 was not trans-mitted in any of 10 cytoduction events to a G2 strain (Table6). Thus transmission barriers can be asymmetric for [PSI+]as was previously found for [URE3] (Edskes et al. 2009).

Figure 1 Distribution of mutations in Sup35pof wild strains of S. cerevisiae. The width of thevertical lines marking the location of mutationsis in rough proportion to their frequency (TableS3). The box on the left is a region in whichmutation can block [PSI+] transmission (Depaceet al. 1998). Boldface, underlined text indicatesthose mutations observed most frequently. Thelocation of the octapeptide repeats is shown as......

Table 4 Cytoductions of [PSI+] from the reference sequenceto polymorphic Sup35s

Donor RecipientAde+

cytoductantsTotal

cytoductants % Ade+

Reference 80 80 100A9 64 66 97

779-6a F7 37 37 100Ref Seq A8 0 30 0PSI+ E9 16 86 19

G2 0 86 0F8 0 86 0D19 13 101 13

Reference 0 30 0Ref Seq A9 0 30 0[psi2] F7 0 28 0

A8 0 30 0E9 0 30 0G2 0 21 0F8 0 25 0D19 0 42 0

The [PSI+] donor (top) was strain 779-6a and the [psi2] donor (bottom) was 4828.The [psi2] recipient was 4830 in all cases.

574 D. A. Bateman and R. B. Wickner

Although [PSI+D19]D19 was transmitted to only 1 of 10cytoductants expressing the reference sequence (Table 6),[PSI+ref]D19 was transmitted to 97 of 110 cytoductantsexpressing the reference Sup35p (Table 5). This result sug-gests a variant-dependent transmission barrier. The variantthat originated in cells expressing the reference sequencetransmitted poorly to a D19 recipient (Table 4), but“remembers” its origins and returns more easily to the ref-erence sequence Sup35p than does a prion originating ina D19 strain (Table 5).

N109S and M-domain mutations each affect [PSI+]transmission barrier

To determine the effect of the N109S mutation, pointmutations were created in the reference sequence to generatethis specific point mutation, and corrective mutations wereintroduced into the E9 and G2 sequences (S109N mutation).The resulting strains were used to determine the transmissionof [PSI+]. Table 7 indicates that the [PSI+ref]ref meets someresistance to infection of the N109S point mutant. In addition,correcting the N109S in E9 and G2 strains back to the refer-ence sequence allows more ready infection by [PSI+ref]ref. Inaddition, this corrective mutation in E9 and G2 hinderedthe transmission of [PSI+G2]G2, indicating that the single-res-idue difference results in a transmission barrier in the contextof G2 or E9, as it does in the context of the reference sequence.

The same data also indicate that the M-domain mutationsfound in many polymorphs can produce a transmission barrierwithout differences in the N or C domains (Table 7). Although

E9 S109N and G2 S109N are better recipients for [PSI+ref]ref

than the uncorrected polymorphs, each still shows a significantbarrier. The combination of N109S and the M-domain muta-tions produces a much greater barrier than would be expectedif they acted independently; the barriers are synergistic.

We note also that while the identical E9 and G2appeared significantly different as recipients of [PSI+ref]in the data in Table 4, the difference is much smaller in theexperiment in Table 7, done with the same donor and recip-ients, suggesting that this is simply experimental variation.

[SWI+] is absent from wild strains: Like [PSI+], it hasbeen suggested that [SWI+] benefits yeast (Du et al. 2008).If [SWI+] is indeed a benefit, it should be found in wildstrains (Nakayashiki et al. 2005). All 76 wild strains werestreaked in patches on media reported to prevent growth inthe presence of the [SWI+] prion (Du et al. 2008). StrainsC3, C8, and D7 were unable to grow on YPG media (FigureS1); however, these YPG-negative strains could not be curedwith guanidine. These strains also lacked the presence ofGFP dots with plasmid pDB67, expressing Swi1(1–536)–GFP, as did all the other wild strains tested (Figure S2).The control [SWI+] strain seemed to be unstable, losingits inhibited growth property on raffinose and YPG andneeded to be counterselected for control tests. It was alsodifficult to observe GFP dots with the control strain overprolonged growth on YPD. We concluded that [SWI+] prionis not found in our wild strains and that in our hands the[SWI+] prion is somewhat unstable.

Table 5 Cytoduction from Ade+ and Ade2 cytoductants (in strain4830) from Table 4 back to strain 4828 expressing the referencesequence and lacking the prion

DonorCytoductant Recipient

Ade+cytoductants

Totalcytoductants % Ade+

Ref Ade2 Ref Seq 0 25 0Ref Ade+ A Ref Seq 40 40 100Ref Ade+ B Ref Seq 54 54 100Ref Ade+ C Ref Seq 32 32 100Ref Ade+ D Ref Seq 45 45 100D19 Ade2 A Ref Seq 0 30 0D19 Ade2 B Ref Seq 0 45 0D19 Ade+ A Ref Seq 20 23 87D19 Ade+ B Ref Seq 38 42 90D19 Ade+ C Ref Seq 14 20 70D19 Ade+ D Ref Seq 25 25 100A9 Ade+ A Ref Seq 10 19 53A9 Ade+ B Ref Seq 55 65 85E9 Ade2 A Ref Seq 12 36 33E9 Ade2 B Ref Seq 12 47 26E9 Ade+ A Ref Seq 22 33 67E9 Ade+ B Ref Seq 12 20 60F8 Ade2 A Ref Seq 0 50 0F8 Ade2 B Ref Seq 0 23 0F8 Ade2 C Ref Seq 0 22 0G2 Ade2 A Ref Seq 0 16 0G2 Ade2 B Ref Seq 0 35 0

The Ade2 D19, G2 and F8 donors produced only Ade2 cytoductants suggestingthat the donors did not carry an unexpressed [PSI+]. The Ade2 E9 donors produceda proportion of Ade+ cytoductants indicating that [PSI+] was being lost, but had notbeen completely lost in the E9 strain.

Figure 2 Interaction of Sup35p polymorphs with each other. In [PSI+],strains of one polymorph is expressed Sup35NM–GFP of another (or thesame) polymorph, and the appearance of the GFP fluorescence as dotsindicates association of the indicated NM–GFP with the Sup35 prion aggre-gates. Strain 4828 with each prion (rows) was transformed with plasmidsexpressing the polymorph NM–GFP (columns) fusion proteins from the ADH1promoter and cells were examined as described in methods. GFP expressedalone (and not fused to Sup35) is evenly distributed in the cytoplasm. A8NM–

GFP does differ by one residue (P186A) from the G2 sequence.

Natural Barriers to [PSI+] Transmission 575

Discussion

Following the description of an apparently beneficial prionof Podospora anserina (Coustou et al. 1997), it was reportedthat the [PSI+] prion protected cells from elevated temper-ature or high concentrations of ethanol (Eaglestone et al.1999). Subsequent work showed that there was no con-sistent protection against heat, ethanol, or any of a largenumber of conditions tested and that for most strains undermost conditions [psi2] was preferable to [PSI+] (True andLindquist 2000). However, some strains were reported to bemodestly benefited by [PSI+], and it was suggested thatthese effects could help yeast evolve by allowing them tosurvive stress (True and Lindquist 2000). However, al-though using the same strains, others were unable to repro-duce the reported [PSI+] advantages (Namy et al. 2008),casting doubt on the [PSI+]-as-benefit proposal. Efforts toshow induction of [PSI+] formation by stress conditionswere unsuccessful with the wild-type Sup35p sequence(Tyedmers et al. 2008). Using an artificial Sup35p with highspontaneous [PSI+] formation, a small increase in [PSI+]formation was observed under several stress conditions, butin four of six such conditions, [PSI+] made cell growthworse (Tyedmers et al. 2008), suggesting that this is notan adaptive response.

Since assessment of marginal and variable phenotypedifferences cannot answer the question of prion benefit ordetriment to yeast, several other approaches have been used.We noted that since severely pathogenic viruses and prions(e.g., chronic wasting disease of elk and deer) are readilyfound in nature, a beneficial infectious element will be foundin nearly all wild strains (Nakayashiki et al. 2005). The ab-sence of [PSI+] and [URE3] from the 70 wild strains exam-ined indicates that they are quite detrimental (Nakayashikiet al. 2005). Here, we show that the same strains lack the[SWI+] prion, indicating that it too is a rare disease, not anadaptive change.

The barrier to transmission between mammalian species(the “species barrier”) is a consequence of sequence differ-ences between the PrP proteins. Sheep and elk/deer developprion diseases in the wild and some sheep have PrP allelesthat make them resistant to infection (Hunter et al. 1996).The M/V129 PrP polymorphism in the human populationhas the effect of protecting against prion disease, with het-erozygotes being resistant, and it is argued that it is just thisprotection that has produced the “balancing selection” forthe polymorphism (Mead et al. 2003). The Kuru epidemicapparently selected a nearby mutation that confers resis-tance to transmissible spongiform encephalopathies (Meadet al. 2009a). Here we show that the polymorphism of theSup35p sequence in wild yeast likewise produces barriers totransmission within S. cerevisiae. This clearly protects thecells from the spread of the prion, and we suggest that itis this protection that resulted in the selection of polymor-phism for these sequence variants. The mutations producingthese barriers are in the N and M domains of Sup35, outsidethe region involved in translation termination (Teravanesyanet al. 1993). Whether these mutations also affect the mRNAturnover function of the NM region of Sup35 (Hoshino et al.1999; Hosoda et al. 2003; Funakoshi et al. 2007) is as yetunclear because the sequence requirements for that functionhave not yet been explored. A change that benefited the yeastby improving one of these nonprion functions of Sup35pmight be expected to take over the population, not be present

Table 6 Cytoductions from polymorphic [PSI+] donors

Donor RecipientAde+

cytoductantsTotal

cytoductants%

Ade+

G2 Red [psi2] G2 0 35 0E9 0 35 0F8 0 23 0Reference 0 36 0A9 0 35 0F7 0 46 0D19 0 18 0

G2 Ade+[PSI+G2]G2

G2 34 36 94E9 25 28 89F8 32 35 91Reference 5 20 25A9 10 32 31F7 8 35 23D19 8 22 36

D19 Red [psi2] D19 0 35 0G2 0 45 0E9 0 30 0F8 0 32 0A9 0 32 0F7 0 35 0Reference 0 36 0

D19 Ade+[PSI+D19]D19

D19 7 8 88G2 0 10 0E9 0 10 0F8 0 10 0A9 4 8 50F7 0 10 0Reference 1 10 10

The donors were strain 4828 expressing polymorph Sup35 and [PSI+] originating inthat polymorph. [psi2] recipients were strain 4830 expressing the indicated Sup35ppolymorphs.

Table 7 N109S is responsible for part of the species barrierin wild strains

Donor RecipientAde+

cytoductantsTotal

cytoductants%

Ade+

779-6a [PSI+ref]ref Reference 49 50 98Ref N109S 93 123 76E9 8 102 8E9 S109N 112 130 86G2 1 67 1G2 S109N 22 33 67

[PSI+G2]G2 Reference 5 50 10Ref N109S 18 85 21E9 35 45 78E9 S109N 14 95 15G2 36 40 90G2 S109N 9 85 11

The [psi2] recipients were strain 4828 expressing the indicated Sup35p.

576 D. A. Bateman and R. B. Wickner

as a polymorphism. However, if any of these polymorphs tookover the population, there would be no protection from prioninfection. The fact of this polymorphism argues that its pres-ence is selected to protect against prion infection, as in thecase of the PrP 129M/V polymorphism (Mead et al. 2003).

The part of the Sup35p prion domain that determinesbarriers to transmission between different sequences hasbeen assigned to residues 2–33 on the basis of mutagen-esis studies of the laboratory S. cerevisiae allele (Depaceet al. 1998). We find that in wild S. cerevisiae strains, thebarrier is provided by either a deletion in the repeat re-gion (residues 59–77) or by a combination of N109S anda group of mutations in the M domain between residues162 and 225, each well outside the region determined bymutation. Previous studies showed that making a [PSI+ref]ref strain express both Sup35ref and Sup35D19 elim-inated the phenotype of [PSI+], but did not eliminate theprion (Resende et al. 2003). When [PSI+ref]ref cells weremade to express only Sup35D19, the [PSI+] phenotype(Ade+) was lost, but it was not determined whether thiswas eliminating the prion or simply interfering with thephenotype (Resende et al. 2003). We find that [PSI+ref]introduced from a cell expressing only the referenceSup35p is lost if the recipient expresses only D19. Thisconstitutes a transmission barrier and is consistent withthe results of Resende et al. (2003).

Deletions or substitutions in the M domain have beenshown to affect [PSI+] in subtle ways, but no prion variantsfailed to be transmitted to SUP35DM strains (Liu et al.2002). A T341A mutation in the C domain appears to block[PSI+] propagation (Kabani et al. 2011), but this was notfound in wild strains. We find that the specific changesfound in the E9 polymorph (and often in other wildstrains)—G162, D169E, T206K H225—are sufficient to sig-nificantly block transmission of [PSI+] from the referencesequence. More dramatically, the single N109S mutationproduces a substantial barrier to transmission of [PSI+ref]ref. Since many point mutations in the N and M partsof Sup35 do not produce barriers to [PSI+] transmission(Depace et al. 1998; Liu et al. 2002), it is likely that thesevariants were selected to produce the transmission barrierthat we observe, suggesting that acquisition of [PSI+] isundesirable.

The finding that mutations in the M region from G162to H225 can affect prion transmission efficiency isconsistent with our previous evidence that there is somein-register b-sheet structure in M, although we were un-able to further localize this structure (Shewmaker et al.2006, 2009).

The interpretation of [PSI+] as a disease is consistentwith an array of other accumulating evidence. The pres-ence of [PSI+] or [URE3] induces the synthesis of the“stress proteins,” Hsp104 and Hsp70, suggesting that cellsview these prions as a stress (Jung et al. 2000; Schwimmerand Masison 2002). Ability of Ure2p to form a prion is notconserved in S. castellii, Candida glabrata, or Kluyveromyces

lactis (Edskes et al. 2009, 2011; Safadi et al. 2011), al-though their putative prion domains are quite similar insequence. If prion formation were advantageous, it wouldbe widely conserved. The fact that the prion domains ofSup35p and Ure2p have other well-established functions,unrelated to prion formation, suggests again that thesedomains are not present for the purpose of the very rareprion formation (Hoshino et al. 1999; Hosoda et al. 2003;Funakoshi et al. 2007; Shewmaker et al. 2007). The findingthat [PSI+] can kill cells by inactivating Sup35p and that[URE3] can be extremely toxic (by an as-yet unknownmechanism) again suggests that these prions are not adap-tive (Mcglinchey et al. 2011). Certain mutants in theHsp40-encoding SIS1 gene make even the usual “mild[PSI+]” become lethal, further indicating the potential tox-icity of this prion (Kirkland et al. 2011). Further work willbe needed to explore the mechanisms of prion pathogene-sis in yeast.

Acknowledgments

We thank Donna MacCallum for kindly sending strainsJ041047 and J940610 and Herman Edskes for severalplasmids. We are grateful to Zhiqiang Du and Liming Lifor kindly providing p416TEFSWI1NQ-YFP and strains 7D-694 and 7D-694SWI+. This work was supported by theIntramural Program of the National Institute of Diabetes andDigestive and Kidney Diseases.

Literature Cited

Afanasieva, E. G., V. V. Kushnirov, M. F. Tuite, and M. D. Ter-Avanesyan, 2011 Molecular basis for transmission barrierand interference between closely related prion proteins in yeast.J. Biol. Chem. 286: 15773–15780.

Baxa, U., R. B. Wickner, A. C. Steven, D. Anderson, L. Marekovet al., 2007 Characterization of b-sheet structure in Ure2p1–89 yeast prion fibrils by solid state nuclear magnetic resonance.Biochemistry 46: 13149–13162.

Bessen, R. A., and R. F. Marsh, 1994 Distinct PrP properties sug-gest the molecular basis of strain variation in transmissible minkencephalopathy. J. Virol. 68: 7859–7868.

Brachmann, A., U. Baxa, and R. B. Wickner, 2005 Prion genera-tion in vitro: amyloid of Ure2p is infectious. EMBO J. 24: 3082–3092.

Bradley, M. E., and S. W. Liebman, 2004 The Sup35 domainsrequired for maintenance of weak, strong or undifferentiatedyeast [PSI+] prions. Mol. Microbiol. 51: 1649–1659.

Bruce, M. E., 2003 TSE strain variation: an investigation intoprion disease diversity. Br. Med. Bull. 66: 99–108.

Caughey, B., G. S. Baron, B. Chesebro, and M. Jeffrey,2009 Getting a grip on prions: oligomers, amyloids, and path-ological membrane interactions. Annu. Rev. Biochem. 78: 177–204.

Chen, B., G. P. Newnam, and Y. O. Chernoff, 2007 Prion speciesbarrier between the closely related yeast proteins is detecteddespite coaggregation. Proc. Natl. Acad. Sci. USA 104: 2791–2796.

Chernoff, Y. O., G. P. Newnam, J. Kumar, K. Allen, and A. D. Zink,1999 Evidence for a protein mutator in yeast: role of the

Natural Barriers to [PSI+] Transmission 577

Hsp70-related chaperone Ssb in formation, stability and toxicityof the [PSI+] prion. Mol. Cell. Biol. 19: 8103–8112.

Cliften, P., P. Sudarsanam, A. Desikan, L. Fulton, B. Fulton et al.,2003 Finding functional features in Saccharomyces genomesby phylogenetic footprinting. Science 301: 71–76.

Collinge, J., and M. P. Alpers, 2008 The end of kuru: 50 years ofresearch into an extraordinary disease. Philos. Trans. R. Soc.Lond. B Biol. Sci. 363: 3607–3763.

Conde, J., and G. R. Fink, 1976 A mutant of Saccharomyces cer-evisiae defective for nuclear fusion. Proc. Natl. Acad. Sci. USA73: 3651–3655.

Coustou, V., C. Deleu, S. Saupe, and J. Begueret, 1997 The pro-tein product of the het-s heterokaryon incompatibility gene ofthe fungus Podospora anserina behaves as a prion analog. Proc.Natl. Acad. Sci. USA 94: 9773–9778.

Cox, B. S., 1965 PSI, a cytoplasmic suppressor of super-suppressorin yeast. Heredity 20: 505–521.

Depace, A. H., A. Santoso, P. Hillner, and J. S. Weissman, 1998 Acritical role for amino-terminal glutamine/asparagine repeats inthe formation and propagation of a yeast prion. Cell 93: 1241–1252.

Derkatch, I. L., Y. O. Chernoff, V. V. Kushnirov, S. G. Inge-Vechto-mov, and S. W. Liebman, 1996 Genesis and variability of [PSI]prion factors in Saccharomyces cerevisiae. Genetics 144: 1375–1386.

Du, Z., K.-W. Park, H. Yu, Q. Fan, and L. Li, 2008 Newly identifiedprion linked to the chromatin-remodeling factor Swi1 in Saccha-romyces cerevisiae. Nat. Genet. 40: 460–465.

Eaglestone, S. S., B. S. Cox, and M. F. Tuite, 1999 Translationtermination efficiency can be regulated in Saccharomyces cerevi-siae by environmental stress through a prion-mediated mecha-nism. EMBO J. 18: 1974–1981.

Edskes, H. K., L. M. Mccann, A. M. Hebert, and R. B. Wickner,2009 Prion variants and species barriers among Saccharomy-ces Ure2 proteins. Genetics 181: 1159–1167.

Edskes, H. K., A. Engel, L. M. Mccann, A. Brachmann, H.-F. Tsaiet al., 2011 Prion-forming ability of Ure2 of yeasts is not evo-lutionarily conserved. Genetics 188: 81–90.

Funakoshi, Y., Y. Doi, N. Hosoda, N. Uchida, M. Osawa et al.,2007 Mechanism of mRNA deadenylation: evidence for a mo-lecular interplay between translation termination factor eRF3and mRNA deadenylases. Genes Dev. 21: 3135–3148.

Hoshino, S., M. Imai, T. Kobayashi, N. Uchida, and T. Katada,1999 The eukaryotic polypeptide chain releasing factor(eRF3/GSPT) carrying the translation termination signal tothe 39-poly(A) tail of mRNA. J. Biol. Chem. 274: 16677–16680.

Hosoda, N., T. Kobayashii, N. Uchida, Y. Funakoshi, Y. Kikuchiet al., 2003 Translation termination factor eRF3 mediatesmRNA decay through the regulation of deadenylation. J. Biol.Chem. 278: 38287–38291.

Hunter, N., J. D. Foster, W. Goldmann, M. J. Stear, J. Hope et al.,1996 Natural scrapie in a closed flock of Cheviot sheep occursonly in specific PrP genotypes. Arch. Virol. 141: 809–824.

Jensen, M. A., H. L. True, Y. O. Chernoff, and S. Lindquist,2001 Molecular population genetics and evolution of a prion-like protein in Saccharomyces cerevisiae. Genetics 159: 527–535.

Jung, G., and D. C. Masison, 2001 Guanidine hydrochloride in-hibits Hsp104 activity in vivo: a possible explanation for its ef-fect in curing yeast prions. Curr. Microbiol. 43: 7–10.

Jung, G., G. Jones, R. D. Wegrzyn, and D. C. Masison, 2000 A rolefor cytosolic Hsp70 in yeast [PSI+] prion propagation and [PSI+] as a cellular stress. Genetics 156: 559–570.

Kabani, M., B. Cosnier, L. Bousset, J.-P. Rousset, R. Melki et al.,2011 A mutation within the C-terminal domain of Sup35pthat affects [PSI+] prion propagation. Mol. Microbiol. 81:640–658.

Kellis, M., N. Patterson, M. Endrizzi, B. Birren, and E. S. Lander,2003 Sequencing and comparison of yeast species to identifygenes and regulatory elements. Nature 423: 241–254.

Kikuchi, Y., H. Shimatake, and A. Kikuchi, 1988 A yeast gene re-quired for the G1 to S transition encodes a protein containing anA kinase target site and GTPase domain. EMBO J. 7: 1175–1182.

King, C. Y., 2001 Supporting the structural basis of prion strains:induction and identification of [PSI] variants. J. Mol. Biol. 307:1247–1260.

King, C. Y., and R. Diaz-Avalos, 2004 Protein-only transmission ofthree yeast prion strains. Nature 428: 319–323.

King, C.-Y., P. Tittmann, H. Gross, R. Gebert, M. Aebi et al.,1997 Prion-inducing domain 2–114 of yeast Sup35 proteintransforms in vitro into amyloid-like filaments. Proc. Natl. Acad.Sci. USA 94: 6618–6622.

Kirkland, P. A., M. Reidy, and D. C. Masison, 2011 Functions ofyeast Hsp40 chaperone Sis1p dispensable for prion propagationbut important for prion curing and protection from prion toxic-ity. Genetics 188: 565–577.

Kushnirov, V. V., M. D. Teravanesyan, M. V. Telckov, A. P. Surguchov,V. N. Smirnov et al., 1988 Nucleotide sequence of the SUP2(SUP35) gene of Saccharomyces cerevisiae. Gene 66: 45–54.

Kushnirov, V. V., M. D. Ter-Avanesyan, S. A. Didichenko, V. N.Smirnov, Y. O. Chernoff et al., 1990 Divergence and conserva-tion of SUP2 (SUP35) gene of yeasts Pichia pinus and Saccharo-myces cerevisiae. Yeast 6: 461–472.

Kushnirov, V. V., N. V. Kochneva-Pervukhova, M. B. Cechenova,N. S. Frolova, and M. D. Ter-Avanesyan, 2000a Prion proper-ties of the Sup35 protein of yeast Pichia methanolica. EMBO J.19: 324–331.

Kushnirov, V. V., D. Kryndushkin, M. Boguta, V. N. Smirnov, and M.D. Ter-Avanesyan, 2000b Chaperones that cure yeast artificial[PSI+] and their prion-specific effects. Curr. Biol. 10: 1443–1446.

Liu, J.-J., N. Sondheimer, and S. Lindquist, 2002 Changes in themiddle region of Sup35p profoundly alter the nature of epige-netic inheritance for the yeast prion. [PSI+] Proc. Natl. Acad.Sci. USA 99: 16446–16453.

Masison, D. C., and R. B. Wickner, 1995 Prion-inducing domain ofyeast Ure2p and protease resistance of Ure2p in prion-containingcells. Science 270: 93–95.

Mcglinchey, R., D. Kryndushkin, and R. B. Wickner, 2011 Suicidal[PSI+] is a lethal yeast prion. Proc. Natl. Acad. Sci. USA 108:5337–5341.

Mead, S., M. P. Stumpf, J. Whitfield, J. A. Beck, M. Poulter et al.,2003 Balancing selection at the prion protein gene consistentwith prehistoric kurulike epidemics. Science 300: 640–643.

Mead, S., J. Whitfield, M. Poulter, P. Shah, J. Uphill et al.,2009a A novel protective prion protein variant that colocalizeswith Kuru exposure. N. Engl. J. Med. 361: 2056–2065.

Mead, S., J. Whitfield, M. Poulter, P. Shah, J. Uphill et al.,2009b A novel protective prion protein variant that colocalizeswith kuru exposure. N. Engl. J. Med. 361: 2056–2065.

Nakao, Y., T. Kanamori, T. Itoh, Y. Kodama, S. Rainieri et al.,2009 Genome sequence of the lager brewing yeast, an inter-species hybrid. DNA Res. 16: 115–129.

Nakayashiki, T., C. P. Kurtzman, H. K. Edskes, and R. B. Wickner,2005 Yeast prions [URE3] and [PSI+] are diseases. Proc. Natl.Acad. Sci. USA 102: 10575–10580.

Namy, O., A. Galopier, C. Martini, S. Matsufuji, C. Fabret et al.,2008 Epigenetic control of polyamines by the prion. [PSI+]Nat. Cell Biol. 10: 1069–1075.

Prusiner, S. B., M. Scott, D. Foster, K.-M. Pan, D. Groth et al.,1990 Transgenic studies implicate interactions between ho-mologous PrP isoforms in scrapie prion replication. Cell 63:673–686.

578 D. A. Bateman and R. B. Wickner

Resende, C. G., T. F. Outeiro, L. Sands, S. Lindquist, and M. F.Tuite, 2003 Prion protein gene polymorphisms in Saccharomy-ces cerevisiae. Mol. Microbiol. 49: 1005–1017.

Safadi, R. A., N. Talarek, N. Jacques, and M. Aigle, 2011 Yeastprions: Could they be exaptations? The URE2/[URE3] system inKluyveromyces lactis. FEM. Yeast Res. 11: 151–153.

Santoso, A., P. Chien, L. Z. Osherovich, and J. S. Weissman,2000 Molecular basis of a yeast prion species barrier. Cell100: 277–288.

Schwimmer, C., and D. C. Masison, 2002 Antagonistic interac-tions between yeast [PSI+] and [URE3] prions and curing of[URE3] by Hsp70 protein chaperone Ssa1p but not by Ssa2p.Mol. Cell. Biol. 22: 3590–3598.

Sherman, F., 1991 Getting started with yeast, pp. 3–21 in Guide toYeast Genetics and Molecular Biology, edited by C. Guthrie, andG. R. Fink. Academic Press, San Diego.

Shewmaker, F., R. B. Wickner, and R. Tycko, 2006 Amyloid of theprion domain of Sup35p has an in-register parallel b-sheetstructure. Proc. Natl. Acad. Sci. USA 103: 19754–19759.

Shewmaker, F., L. Mull, T. Nakayashiki, D. C. Masison, and R. B.Wickner, 2007 Ure2p function is enhanced by its prion domainin Saccharomyces cerevisiae. Genetics 176: 1557–1565.

Shewmaker, F., D. Kryndushkin, B. Chen, R. Tycko, and R. B. Wickner,2009 Two prion variants of Sup35p have in-register b-sheetstructures, independent of hydration. Biochemistry 48: 5074–5082.

Shkundina, I. S., V. V. Kushnirov, M. F. Tuite, and M. D. Ter-Avanesyan, 2006 The role of the N-terminal oligopeptide re-peats of the yeast Sup35 prion protein in propagation andtransmission of prion variants. Genetics 172: 827–835.

Stansfield, I., K. M. Jones, V. V. Kushnirov, A. R. Dagkesamanskaya,A. I. Poznyakovski et al., 1995 The products of the SUP45(eRF1) and SUP35 genes interact to mediate translation termi-nation in Saccharomyces cerevisiae. EMBO J. 14: 4365–4373.

Tanaka, M., P. Chien, N. Naber, R. Cooke, and J. S. Weissman,2004 Conformational variations in an infectious protein deter-mine prion strain differences. Nature 428: 323–328.

Tanaka, M., S. R. Collins, B. H. Toyama, and J. S. Weissman,2006 The physical basis of how prion conformations deter-mine strain phenotypes. Nature 442: 585–589.

Taylor, K. L., N. Cheng, R. W. Williams, A. C. Steven, and R. B.Wickner, 1999 Prion domain initiation of amyloid formationin vitro from native Ure2p. Science 283: 1339–1343.

Teravanesyan, M. D., V. V. Kushnirov, A. R. Dagkesamanskaya, S. A.Didichenko, Y. O. Chernoff et al., 1993 Deletion analysis of theSUP35 gene of the yeast Saccharomyces cerevisiae reveals twonon-overlapping functional regions in the encoded protein. Mol.Microbiol. 7: 683–692.

Teravanesyan, A., A. R. Dagkesamanskaya, V. V. Kushnirov, andV. N. Smirnov, 1994 The SUP35 omnipotent suppressor geneis involved in the maintenance of the non-Mendelian determinant[psi+] in the yeast Saccharomyces cerevisiae. Genetics 137: 671–676.

True, H. L., and S. L. Lindquist, 2000 A yeast prion providesa mechanism for genetic variation and phenotypic diversity. Na-ture 407: 477–483.

Tycko, R., 2011 Solid-state NMR studies of amyloid fibril struc-ture. Annu. Rev. Phys. Chem. 62: 279–299.

Tyedmers, J., M. L. Madariaga, and S. Lindquist, 2008 Prionswitching in response to environmental stress. PLoS Biol. 6:e294.

Wickner, R. B., 1994 [URE3] as an altered URE2 protein: evidencefor a prion analog in S. cerevisiae. Science 264: 566–569.

Wickner, R. B., H. K. Edskes, F. Shewmaker, and T. Nakayashiki,2007 Prions of fungi: inherited structures and biological roles.Nat. Rev. Microbiol. 5: 611–618.

Wickner, R. B., F. Dyda, and R. Tycko, 2008a Amyloid of Rnq1p,the basis of the [PIN+] prion, has a parallel in-register b-sheetstructure. Proc. Natl. Acad. Sci. USA 105: 2403–2408.

Wickner, R. B., F. Shewmaker, D. Kryndushkin, and H. K. Edskes,2008b Protein inheritance (prions) based on parallel in-registerb-sheet amyloid structures. BioEssays 30: 955–964.

Wickner, R. B., F. Shewmaker, H. Edskes, D. Kryndushkin, J. Nemeceket al., 2010 Prion amyloid structure explains templating: howproteins can be genes. FEM. Yeast Res. 10: 980–991.

Wilson, P. G., and M. R. Culbertson, 1988 SUF12 suppressor pro-tein of yeast: a fusion protein related to the EF-1 family ofelongation factors. J. Mol. Biol. 199: 559–573.

Zhouravleva, G., L. Frolova, X. Legoff, R. Leguellec, S. Inge-Vectomovet al., 1995 Termination of translation in eukaryotes is gov-erned by two interacting polypeptide chain release factors,eRF1 and eRF3. EMBO J. 14: 4065–4072.

Communicating editor: A. P. Mitchell

Natural Barriers to [PSI+] Transmission 579

GENETICSSupporting Information

http://www.genetics.org/content/suppl/2011/11/18/genetics.111.136655.DC1

[PSI+] Prion Transmission Barriers ProtectSaccharomyces cerevisiae from Infection:

Intraspecies ‘Species Barriers’David A. Bateman and Reed B. Wickner

Copyright © 2012 by the Genetics Society of AmericaDOI: 10.1534/genetics.111.136655