[URE3] is a prion of the yeast Ure2 protein. Hsp40 is a cochaperone that regulates Hsp70 chaperone activity. When overexpressed, the Hsp40 Ydj1p cures yeast of [URE3], but the Hsp40 Sis1p does not. On the basis of biochemical data Ydj1p has been proposed to cure [URE3] by binding soluble Ure2p and preventing it from joining prion aggregates. Here, we mutagenized Ydj1p and find that disrupting substrate binding, dimerization, membrane association, or ability to transfer substrate to Hsp70 had little or no effect on curing. J-domain point mutations that disrupt functional interactions of Ydj1p with Hsp70 abolished curing, and the J domain alone cured [URE3]. Consistent with heterologous J domains possessing similar Hsp70 regulatory activity, the Sis1p J domain also cured [URE3]. We further show that Ydj1p is not essential for [URE3] propagation and that depletion of Ure2p is lethal in cells lacking Ydj1p. Our data imply that curing of [URE3] by overproduced Ydj1p does not involve direct interaction of Ydj1p with Ure2p but rather works through regulation of Hsp70 through a specific J-protein/Hsp70 interaction.

---

PRION diseases are amyloidoses with the distinction of being transmissible between organisms. Amyloid is a highly ordered, fibrous aggregate of protein in which each protein monomer has a similar misfolded conformation. Although it is not clear how amyloid arises, once present it continues to propagate by recruiting the soluble form of the protein and converting it into the same misfolded form as it joins the aggregate. Replication of prions is thought to occur when fibers break into more numerous pieces, either spontaneously or through the action of protein chaperones. As in mammals, certain proteins of the yeast Saccharomyces cerevisiae can misfold and propagate as prions that possess amyloid properties and are efficiently transmitted between strains and to daughter cells. Since yeast cells propagating prions are viable, they provide a valuable system to study various components of cellular machinery involved in prion propagation (WICKNER et al. 2007).

By binding exposed hydrophobic surfaces of proteins, the universally conserved Hsp40 and Hsp70 family molecular chaperones help other proteins attain native conformations, prevent partially unfolded proteins from aggregating, and help disaggregate and refold damaged proteins (BUKAU et al. 2006). Hsp40's can act alone as chaperones to inhibit protein aggregation, but their primary roles are as cochaperones that assist Hsp70 and regulate its activity. Ydj1p is a major cytosolic Hsp40 of yeast that regulates Hsp70 function in many important cellular processes, such as transport of proteins across membranes, targeted degradation of misfolded proteins, and protection of nascent chains against degradation (CAPLAN et al. 1992a; LEE et al. 1996; MANDAL et al. 2008).

Altered expression of Hsp40 and Hsp70 strongly influences the dynamics of yeast prion propagation. The yeast prions [URE3], [PSI⁺], and [PIN⁺] are self-replicating misfolded forms of the Ure2, Sup35, and Rnq1 proteins, respectively. Overexpressing Ydj1p cures cells of [URE3] but another Hsp40, Sis1p, has no effect (MORIYAMA et al. 2000; KRYNDUSHKIN and WICKNER 2007; LIAN et al. 2007). Conversely, Sis1p is essential for [PIN⁺] propagation but Ydj1p is dispensable (SONDHEIMER et al. 2001; LOPEZ et al. 2003; ARON et al. 2007). It was suggested that the difference in effect of Ydj1p and Sis1p on [PIN⁺] is related to specific Hsp40/Hsp70 interactions (LOPEZ et al. 2003). With regard to [PSI⁺], Ydj1p is dispensable and was shown to eliminate weak variants of this prion, but only when the cytosolic Hsp70 Ssa1p is also overexpressed (KUSHNIROV et al. 2000; JONES and MASISON 2003). Overexpressing Ssa1p, but not the nearly identical Ssa2p, antagonizes [URE3], and cells lacking both Ssa1p and Ssa2p cannot propagate [URE3] (SCHWIMMER and MASISON 2002; SHARMA and MASISON 2008). In vitro, both Ydj1p and Ssa1p inhibit Ure2p amyloid formation (LIAN et al. 2007; SAVISTCHENKO et al. 2008). These latter observations suggest that Hsp70 and Hsp40 each interact with Ure2p in vivo, and it has been proposed that Ydj1p and Ssa1p cure [URE3] by binding Ure2p and inhibiting its ability to form amyloid in cells (LIAN et al. 2007; SAVISTCHENKO et al. 2008). Such a mechanism has not been confirmed experimentally.

Here we dissect the role of Ydj1p in [URE3] curing by studying the effect of Ydj1p mutations spanning the entire protein. Unexpectedly, mutations that disrupt substrate binding had little effect on ability of Ydj1p to cure [URE3]. Similarly, mutations expected to disrupt function of all regions outside the J domain had little or no effect on [URE3] curing. Mutants unable to cure [URE3] were obtained only in the J domain, and the J domain alone was capable of curing [URE3], indicating that Ydj1p interaction with Hsp70 plays a critical role in [URE3] curing. Our data broaden understanding of how Ydj1p effects [URE3] and provide insight into the mechanism of Ydj1p functions.

Strains, growth conditions and plasmids:

Yeast strain 1075 (MAT, kar1-1, P_DAL5::ADE2, his3202, leu21, trp163, ura3-52 [URE3]) (SHARMA and MASISON 2008) was used for routine curing experiments. Isogenic diploid strains 1229 (YDJ1/ydj1::KanMX) and 1252 (YDJ1/ydj1::KanMX and URE2/ure2::URA3) were constructed by crosses and standard gene replacement (ROTHSTEIN 1991). Strains were grown at 30° unless indicated. Media are as described (SHERMAN 1991) with slight modifications. To enhance red pigmentation of strains lacking [URE3], solid rich medium (1/2 YPD) contains 0.5% yeast extract instead of the usual 1%. YPAD liquid medium is similar but lacks agar and contains 1% yeast extract and 400 mg/liter adenine. Synthetic dextrose (SD) media contain 7 g/liter yeast nitrogen base, 2% dextrose, required nutritional supplements, and limiting (9 mg/liter) adenine.

Plasmid pRS423 and pRS426 are multicopy HIS3- and URA3-based vectors, respectively (SIKORSKI and HIETER 1989). Plasmid p423YDJ is pRS423 with YDJ1 and 500 bp of 5' and 3' flanking DNA inserted at the BamHI site. Plasmid p423YDJ-AA was engineered by site-directed mutagenesis to have an AatII site immediately preceding the initiator ATG codon and AgeI site immediately following the TAG termination codon. Plasmid p426YDJ-AA is pRS426 with the BamHI fragment from plasmid p423YDJ-AA. Plasmids pDCM12 (MASISON et al. 1997) and pDCM16 (MASISON and WICKNER 1995) are single-copy HIS3-based plasmids with URE22-65 (lacking codons 2–65) and intact URE2, respectively.

Mutagenesis and screening:

YDJ1 was amplified with a 5' primer containing an AatII site and a 3' primer containing an AgeI site using the Stratagene GeneMorph II Random Mutagenesis kit. We used 3.5 micrograms of target DNA (in a 50-µl reaction) and 22 amplification cycles. The reaction products were purified using the QIAGEN MinElute kit, digested with AatII and AgeI, purified from agarose gels after electrophoresis, and then ligated to p423YDJ-AA digested by the same enzymes. The ligated DNA was used to transform Escherichia coli DH5- and plasmids were isolated from a pool of the transformant colonies. This plasmid library was used to transform [URE3] strain 1075. Primary selection was for transformants remaining white on limiting adenine, indicating [URE3] was not affected by the mutant YDJ1 allele. The presence of [URE3] was confirmed by its dominant phenotype and ability to be eliminated by growth on medium containing 3 mM guanidine-hydrochloride, which cures prions by inactivating Hsp104 in vivo (JUNG and MASISON 2001).

Comparison of curing efficiency:

Measurements were done by pooling cells from 3–6 of the whitest colonies from the primary transformation plates, suspending them at very low density in YPAD, growing them 15–20 hr, and then spreading onto 1/2 YPD plates to a cell density of 300–500 per plate. Entirely red colonies were scored as arising from cells having lost [URE3] and the presence of [URE3] was confirmed by replica-plating colonies onto SD plates lacking adenine and 1/2 YPD plates containing 3 mM guanidine-hydrochloride. Data are averages of at least two experiments for each allele tested. We note that this method, although clearly defined and highly reproducible, was used to quantify differences rather than to determine absolute values of curing. Many other ways of quantifying frequency or rate of [URE3] loss are possible. Importantly, our method gives a value less than but close to 100% cured cells for wild-type Ydj1p, and any detectable curing in this assay is at least three orders of magnitude higher than the spontaneous loss of [URE3] we observed from the combined empty vector control measurements.

Western analysis:

Strain 1075 transformants carrying pRS423 (empty vector), p423YDJ-AA (wild type), and versions of p423YDJ-AA containing mutant YDJ1 alleles were grown to mid-log phase in medium selecting for plasmid maintenance. Cells were collected by centrifugation, suspended in an equal volume of buffer, and broken by agitation with glass beads. Whole lysates were transferred to fresh tubes and clarified by centrifugation at top speed in a microfuge for 10 min. Ten micrograms of protein from the lysates were separated on a denaturing 12% polyacrylamide gel and then transferred to a nylon membrane. The membrane was probed with antibody raised in rabbit against a synthesized peptide spanning Ydj1p amino acids 354–372, followed by horseradish peroxidase-conjugated goat anti-rabbit antibody [BioRad catalog (cat) no. 170-6515] and chemiluminescence (ThermoScientific cat no. 34080).

J-domain mutations prevented curing of [URE3]:

Figure 1A is a diagram of Ydj1p and its structural domains. Ydj1p has an amino-terminal J domain followed by a glycine/phenylalanine-rich region (GF), a zinc finger-like region (ZFLR), a C-terminal peptide-binding domain (CTD), a dimerization motif (not indicated), and a farnesylation signal for membrane localization (F). The J domain is necessary for Ydj1p to stimulate Hsp70 ATPase activity, the GF region is thought to influence substrate specificity, and the ZFLR aids transfer of substrate to Hsp70 (TSAI and DOUGLAS 1996; LOPEZ et al. 2003; FAN et al. 2005).

View larger version (76K): In this window In a new window Download PPT slide   FIGURE 1.— Ydj1p J-domain mutations disrupt [URE3] curing activity. (A) Diagram of Ydj1p coding region with defined domains. Numbers indicate amino acid residues. (B) Prion phenotype and curing effect of overexpressed Ydj1p. All plates have limiting adenine. Transformed by an empty plasmid (vector) or plasmid expressing Ydj1p, 1075 cells were selected on –His plates (indicated). Cells from colonies on –His were grown on rich medium and then spread onto 1/2 YPD as described in MATERIALS AND METHODS. Vector control exemplifies phenotype of cells with ([URE3]) and without ([ure-o]) prions. [URE3] cells grow slower and have white colony color on –His and 1/2 YPD plates, whereas [ure-o] cells and transformants overexpressing Ydj1p have a red phenotype. For this and similar experiments, only 5–25% of colonies on the 1/2 YPD plates contain cells that still have the plasmid, so plasmid is not required to maintain [ure-o] phenotype. (C) As in B for cells transformed by plasmids with indicated Ydj1p J-domain substitutions. (D) As in B for cells expressing J-domain triple mutant allele (left) as well as double and single mutant alleles (as indicated) derived from it.

 
The colony color assay we used to monitor [URE3] is based on the P_DAL5::ADE2 reporter, in which the ADE2 gene is regulated by the DAL5 promoter (SCHLUMPBERGER et al. 2001; BRACHMANN et al. 2005) (see Figure 1B). The DAL5 promoter is repressed by Ure2p when a good source of nitrogen is available, so on standard growth media containing ammonium Ure2p represses ADE2 expression and cells require adenine to grow. On media where adenine is limiting, cells failing to express Ade2p are red due to accumulation of a metabolite of the adenine biosynthesis pathway. Depletion of Ure2p into prion aggregates when [URE3] is present relieves repression of the DAL5 promoter so ADE2 is expressed and cells grow without adenine and are white.

Figure 1B also shows the effect of Ydj1p overexpression on [URE3]. We transformed strain 1075 with a multicopy plasmid (p423YDJ-AA) encoding Ydj1p controlled by its own promoter or the empty vector (pRS423). When selected on medium containing limiting adenine, all pRS423 transformants remained white while all p423YDJ-AA transformants were red. When cells from the pRS423 transformant colonies were grown on YPAD liquid medium and then spread onto 1/2 YPD plates all of several hundred colonies were white, indicating that all the cells from the transformant colonies that were tested had [URE3]. In contrast, 94% of the p423YDJ-AA progeny were red (Table 1). Less than 10% of these red colonies on 1/2 YPD contained cells that still had the plasmid, indicating that the red phenotype was due to loss of [URE3] and was not dependent on continued high expression of Ydj1p. Thus, [URE3] was highly sensitive to being cured by Ydj1p when overexpressed from its own promoter on a high-copy plasmid.

View this table: In this window In a new window   TABLE 1 Effects of Ydj1p J-domain mutations on [URE3] curing efficiency

 
The functions of conserved regions of Hsp40's have been extensively studied but the importance of the different Ydj1p functions for eliminating [URE3] are not known. We therefore randomly mutagenized YDJ1 to identify structural regions important for [URE3] curing. Roughly one-fifth of 12,000 transformants of a library of mutant alleles were not cured of [URE3]. Among these we selected 110 colonies that had growth rates typical of [URE3] cells. YDJ1 alleles from 100 of them had a nonsense mutation, deletion, or insertion causing a premature stop codon. The remaining alleles encode full-length open reading frames with missense mutations. For all alleles with a single substitution the alteration was in the J domain. Additionally, one mutant with three substitutions had all three mutations in the J domain. As this domain is critical for Hsp40/Hsp70 interaction, these data point to Hsp40 regulation of Hsp70 as an important factor in the curing mechanism.

Notably, although the remaining alleles with multiple substitutions had mutations outside the J domain, all had at least one or more substitutions within the J domain. To test if the J-domain substitutions in these alleles were responsible for the curing defect, we constructed alleles containing only these J-domain mutations. For all of the alleles tested, a single J-domain substitution was enough to block curing (see Figure 1C and Table 1). Among these, the mutations of conserved residues L11P and A30T are novel but F47S was previously shown to be important for Hsp70 interaction (GENEVAUX et al. 2002), again pointing to a role of Hsp70 in Ydj1p-mediated curing of [URE3]. Thus, altered J-domain function was responsible for the curing defect in all of our randomly isolated Ydj1 mutants.

The triple J-domain mutant defective for curing contains substitutions in the three conserved residues D9A, Y26H, and D67A. We constructed alleles with all possible combinations of double and single substitutions of these residues and tested them for [URE3] curing ability (Figure 1D). The singly mutated proteins cured [URE3] almost as well as wild type. Double mutants D9A/Y26H and D9A/D67A also cured [URE3] efficiently, but the Y26H/D67A mutant had only partial curing activity (Table 1). These results show that combining mutations at positions 26 and 67 decreased ability of Ydj1p to cure [URE3], but all three substitutions were required to more fully inhibit the curing mechanism.

The J domain spans amino acids 1–69 and contains four -helices (Figure 2). Helices II and III are important for Hsp70 interaction and helices I and IV provide structural stabilization to these helices. A universally conserved HPD tripeptide motif essential for Hsp40–Hsp70 interaction and stimulation of Hsp70 ATPase resides in the loop joining helices II and III. Half of the noncuring J-domain mutants had substitutions in this motif (H34P, H34Q, P35L, P35R, and D36A). As noted above, when present alone, each of the HPD substitutions blocked curing activity (see Figure 1C and Table 1). These data suggest that [URE3] curing by Ydj1p overexpression requires Ydj1p–Hsp70 interaction and that no other Ydj1p activity can fulfill this function.

View larger version (31K): In this window In a new window Download PPT slide   FIGURE 2.— Structure of Ydj1p J domain and location of mutations that inhibit [URE3] curing. Backbone structure is green. Helices I–IV are indicated. Residues that were identified as mutated in our screen are numbered and shown with wild-type side chain, each allele represented by a different color. The conserved HPD motif composes residues 34–36.

 

Mutations outside the J domain did not prevent [URE3] curing:

Although the mutagenesis was random and HPD alterations were isolated several times, we did not find mutants without J-domain alterations, which suggested regions outside the J domain are not involved in the curing. To test this hypothesis, we constructed mutations known to disrupt substrate binding, ZFLR function, dimerization or C-terminal prenylation (CAPLAN et al. 1992b; LI et al. 2003; FAN et al. 2005; LI and SHA 2005).

Substrate binding region:

We altered residues known to be important for substrate binding to decrease hydrophobicity of the peptide binding pocket. Figure 3 illustrates effects of these mutations on [URE3] curing and Table 2 indicates the degree to which efficiency of curing is altered. The I116A substitution, which abolishes binding of a model substrate (LI and SHA 2005), did not affect curing significantly. Further decreasing hydrophobicity by combining I116A and L137A also had little effect. An additional mutant with the four substitutions I116A, L137A, V248D, and F249A (designated 4xSBD), which introduces charge as well as reducing hydrophobicity, was only slightly lower in curing efficiency compared with wild-type Ydj1p. We also tested three other alterations in the ZFLR and CTD (R131G, R213G, and G315D) that are known to disrupt substrate binding and ability of Ydj1 to act as a chaperone. None of these substitutions had a considerable effect on curing. Together these data confirm that substrate binding function is not important for Ydj1p curing of [URE3].

View larger version (102K): In this window In a new window Download PPT slide   FIGURE 3.— Ydj1p mutations in residues important for substrate-binding, dimerization, farnesylation or zinc finger-like region (ZFLR) function do not block [URE3] curing mechanism. Transformation selection plates (–His) and outgrowth plates (1/2 YPD) are as described in Figure 1. Ydj1p amino acid substitutions disrupt substrate binding (L116A and 4XSBD), dimerization (F335D), farnesylation (C406S), or ZFLR function (C201S, C162S, and C143S). The 4XSBD mutant contains substitutions I116A, L137A, V248D, and F249A. The –His plates show [URE3] phenotype of cells overexpressing alleles with indicated substitutions from plasmids. The 1/2 YPD plates show [URE3] phenotypes of cells after allowing loss of plasmids during growth on rich medium. As in Figure 1, in this experiment [URE3] was completely stable in cells transformed by empty plasmid and was almost entirely cured from cells transformed by plasmids expressing wild-type Ydj1p (not shown).

 

View this table: In this window In a new window   TABLE 2 Effects of Ydj1p functional domain mutations on efficiency of [URE3] curing

 

Zinc finger motif:

We constructed and tested previously described zinc finger-like region (ZFLR) mutants having substitutions C143S or C201S in the first zinc-binding domain (ZBDI) and C162S in the second (ZBDII). Although all three Ydj1p mutants bind denatured luciferase and stimulate ATPase activity of Hsp70, C143S, and C201S partially inhibit cooperation of Ydj1p with Hsp70 to refold luciferase while C162S completely inhibits this function, indicating that the ZFLR helps transfer substrate to Hsp70 (LINKE et al. 2003; FAN et al. 2005). The C201S mutant was only slightly defective in curing [URE3] and the C143S and C162S mutants cured [URE3] like wild-type Ydj1p (see Figure 3 and Table 2). These results suggest that curing does not depend on ability of Ydj1p to transfer nonnative polypeptides to Hsp70.

Dimerization and prenylation:

Replacing phenylalanine residue 335 of Ydj1p with aspartate disrupts Ydj1p dimerization (LI et al. 2003). Ydj1(F335D) cured [URE3] relatively efficiently (see Figure 3 and Table 2), showing that dimerization of Ydj1p is not critical for prion curing function.

Ydj1p has a C-terminal farnesylation signal (CASQ). About 20% of Ydj1p is membrane-associated and the Ydj1(C406S) mutant is neither farnesylated nor associated with membranes (CAPLAN and DOUGLAS 1991; CAPLAN et al. 1992b). This mutant cured ony slightly less efficiently than wild-type Ydj1p (see Figure 3 and Table 2), indicating that membrane localization is not critical for [URE3] curing.

J domain alone cured [URE3]:

Together our findings suggested the J domain of Ydj1p is sufficient for [URE3] curing. While transformants with the control plasmid never had sectors, 1% of transformants overexpressing this domain alone showed red sectors, which reflect progeny of cells that lost [URE3] during growth of the colony, whereas the remaining colonies appeared entirely white. To confirm [URE3] was being lost, cells from five apparently nonsectoring white colonies were pooled, grown overnight in rich medium, and spread onto 1/2 YPD. Ten percent of resulting colonies were found to be [ure-o] (see Figure 4 and Table 3). Thus, although the curing was less efficient than with intact Ydj1p, the J domain alone was enough to cure [URE3].

View larger version (76K): In this window In a new window Download PPT slide   FIGURE 4.— Ydj1p and Sis1p J domains cure [URE3]. [URE3] stability in cells overexpressing truncated versions of Ydj1p and Sis1p that include amino-terminal residues indicated in parentheses was monitored as described in Figure 1 except that –Ura medium was used for plasmid selection. As in Figure 1, in this experiment [URE3] was completely stable in cells transformed by empty plasmid and was almost entirely cured from cells transformed by plasmids expressing wild-type Ydj1p (not shown).

 

View this table: In this window In a new window   TABLE 3 Efficiency of [URE3] curing by C-terminally truncated Ydj1 proteins

 
To stimulate ATPase activity of Hsp70 the J domain requires close association of the GF region, which is thought to help by providing structural stabilization (HUANG et al. 1999). We therefore constructed and tested a gene encoding Ydj1 residues 1–104, which includes part of the GF region to add possible structural stability. Ydj1(1–104) overexpression similarly resulted in white colonies with and without red sectors, but transformant colonies contained 2–3 times as many [ure-o] cells when tested as above for the J domain alone (Figure 4 and Table 3). Thus, [URE3] curing was significantly increased by appending part of the GF region to the J domain.

Sis1p J domain cured [URE3]:

Although Sis1p and Ydj1p have some overlapping functions in vivo, overexpression of Sis1p does not cure [URE3] (KRYNDUSHKIN and WICKNER 2007; LIAN et al. 2007). Since the most highly conserved region among J proteins is the J domain, we anticipated if J-domain function alone was enough to cure [URE3], then the Sis1p J domain would cure [URE3]. We tested this prediction by expressing the Sis1p J domain from the YDJ1 promoter on the same high-copy plasmid. While we did not detect loss of [URE3] among 3000 colonies examined from transformants expressing intact Sis1p, the J domain of Sis1p reproducibly cured [URE3]. Although it cured less efficiently than the J domain of Ydj1p (see Figure 4 and Table 3), which might reflect a slightly lower ability of Sis1p to stimulate Ssa1p ATPase (LU and CYR 1998), the J domain of Sis1p eliminated [URE3] at least 80 times more frequently than the intact protein. Thus, the inability of intact Sis1p to cure [URE3] was not due to a functionally distinct J domain.

Failure to cure was not due to reduced Ydj1p expression:

To determine if the noncuring Ydj1p mutant proteins were stably expressed at elevated levels, we analyzed Ydj1p abundance by Western blot (Figure 5). Overexpressed proteins were present at widely variable levels, but steady state abundance of the noncuring J-domain mutant proteins was significantly higher than that of endogenous Ydj1p, and all were at least as abundant as the F335D protein, which cured [URE3] efficiently. Moreover, although the A30T and H34Q proteins were very much more abundant than wild-type Ydj1p, they had little or no effect on [URE3] stability. Combined with the fact that any detectable curing would be at least 100-fold more frequent than spontaneous loss of [URE3] (see MATERIALS AND METHODS), these results indicate that the lack of curing by these proteins was not due to deficiency in protein expression but rather to altered protein function.

View larger version (62K): In this window In a new window Download PPT slide   FIGURE 5.— Abundance of overexpressed wild-type and mutant Ydj1p proteins. (Top) Western analysis of lysates of yeast cells transformed by the empty vector control (e) and the same plasmid carrying wild-type YDJ1 (WT) or YDJ1 mutant alleles (amino acid substitutions indicated, 4xSBD contains I116A, L137A, V248D, and F249A. Arrow points to endogenous Ydj1p in lane e. (Bottom) (load) A portion of the blotted membrane stained with amindo black as a loading and transfer control.

 

Ure2p depletion is lethal to cells lacking Ydj1p, and Ydj1p is not essential for [URE3] propagation:

To test if Ydj1p is required for [URE3] propagation, we attempted to transfer [URE3] to a ydj1 strain by cytoduction, a type of abortive mating during which cytoplasm is transferred between mating partners. Although we recovered >50 ydj1 cells with cytoplasm from [ure-o] mating partners, we were unable to recover any cells with cytoplasm from isogenic [URE3] partners. These results suggested that [URE3] is lethal to cells lacking Ydj1p. To test this hypothesis we sporulated a [URE3] diploid heterozygous for ydj1 (strain 1229) and dissected 22 tetrads. Among an expected 44 ydj1 spores, only 1 was viable and it was [ure-o]. Inspection of a dozen ydj1 spore clones revealed that they all germinated but did not divide more than five times (see Figure 6), suggesting the viable ydj1 clone had lost [URE3] shortly after germinating. In contrast, all of 44 ydj1 spores among 22 tetrads of the isogenic [ure-o] diploid were viable. Thus, [URE3] was killing cells that lack Ydj1p.

View larger version (72K): In this window In a new window Download PPT slide   FIGURE 6.— [URE3] kills cells lacking Yjd1p. [URE3] strain 1229 was sporulated and tetrads (numbered) were dissected on 1/2 YPD plates and grown for 4 days. (Top) Spore colonies from dissected tetrads. All viable spores are wild-type YDJ1. (Bottom) Colonies from YDJ1 (left, dark area is edge of colony) and ydj1 (right) spores at the same magnification.

 
To test if the lethality was due to depletion of functional Ure2p by [URE3], rather than the presence of the prion per se, we dissected 40 tetrads from a [ure-o] diploid heterozygous for both ydj1 and ure2 (strain 1252). None of the ydj1 ure2 spores grew, indicating that Ure2p function is essential for growth of cells lacking Ydj1p and the toxicity of [URE3] was likely caused by sequestering soluble Ure2p into prion aggregates. To determine if the prion itself was lethal in cells lacking Ydj1p, we transformed the [URE3] diploid strain 1229 with plasmids carrying wild-type URE2 (pDCM16) or a form of Ure2p lacking its prion-determining region (pDCM12), which provides Ure2p function but cannot be depleted into [URE3] aggregates. As expected, among 20 tetrads from diploids carrying pDCM16 or the empty control vector no ydj1 [URE3] clones were recovered. In contrast, among 16 tetrads from diploids with pDCM12 we recovered 20 ydj1 spores, all of which had pDCM12 and 16 that were confirmed to have [URE3]. These results show that the [URE3] prion itself is not lethal to cells lacking Ydj1p and that Ydj1p is not essential for [URE3] propagation.

Our results are consistent with data from a recent report (HIGURASHI et al. 2008) showing that Ydj1p is dispensable for [URE3], but conflict with the finding in the same study that ydj1 cells supported [URE3] propagation. A plausible explanation for the different findings is that the variant of the [URE3] prion in the other study is weaker than normal and depletes less Ure2p from the cytosol. Alternatively, differences in genetic backgrounds of the strains used by our two groups could affect sensitivity of cells lacking Ydj1p to Ure2p depletion.

All randomly generated Ydj1p mutants we isolated as defective in curing [URE3] had J-domain substitutions, showing that this domain plays a critical role in [URE3] curing. Most of these mutations were in the residues of the HPD tripeptide motif, which defines the J domain, or other conserved residues known to be important for Hsp40 to interact with Hsp70 and regulate its activity (LU and CYR 1998; SUH et al. 1999; GENEVAUX et al. 2002). These results alone suggest the primary mechanism of curing [URE3] by overproduced Ydj1p is through altered regulation of Hsp70 rather than the proposed direct interaction between Ydj1p and Ure2p (LIAN et al. 2007).

The fact that intact proteins with single point mutations in the J domain were unable to cure [URE3] indicates that no other region of Ydj1p can provide this function. Confirming this conclusion we show that all non-J-domain activities are dispensable for curing and that the J domain alone was able to eliminate the prion. In addition to showing that J-domain function is necessary and sufficient for curing, these results demonstrate that substrate binding function of Ydj1p is not required for [URE3] curing. Our data do not exclude the formal possibility that the Ydj1p J domain binds to Ure2p; however, unless such an unexpected interaction occurs, then the mechanism of prion curing does not involve direct binding of Ydj1p to Ure2p, which again points to regulation of Hsp70 activity as the key to Ydj1p curing of [URE3]. Among other things, this conclusion points out the importance of in vivo analysis to explain effects caused by complex chaperone interactions.

Curing of [URE3] by the Sis1p J domain shows that the effect is a general one with regard to J-domain function, which attests to the functional similarity among J domains in regulating Hsp70 activity. J domains mediate physical interaction with Hsp70 and are critical for stimulating Hsp70 ATPase activity. The ability of different J domains to cure [URE3] is consistent with earlier data showing that increased stimulation of Ssa1p ATPase by a variety of factors can have an antiprion effect (JONES et al. 2004), and suggests that J domains of other J proteins would have a similar effect. In fact, further support of a general function was found recently in that the J domain of Jjj1p, another Ssa1p cochaperone, could cure [URE3], although intact Jjj1p could not (HIGURASHI et al. 2008). Overexpressing the J domain of Sis1p or Jjj1p restores normal growth to cells lacking Ydj1p, showing they function like Ydj1p (SAHI and CRAIG 2007). Whether there is a strict correlation between complementation of Ydj1p function and [URE3] curing among other J proteins remains to be tested. With regard to [URE] curing the authors also interpret their results as indicating an interaction of Ydj1p with Hsp70, and suggest that normal recruitment of J proteins to specific locations in the cell, such as Jjj1p to ribosomes, precludes their ability to affect prion propagation.

Since the J domain of Sis1p cures [URE3], the inability of intact Sis1p to cure is not because its J domain functions differently. Sis1p is more widespread in the cytosol and we presume that specific functions of its other structural regions prevent it from interacting with other components of the chaperone machinery, Hsp70 in particular, in a manner that inhibits prion propagation. Both structural and functional differences of type I and type II Hsp40's are mediated by their central regions (LOPEZ et al. 2003; FAN et al. 2005; RAMOS et al. 2008) and it is likely that these regions influence interactions with other chaperones. Although GF regions of various Hsp40's are structurally similar, this region also confers some functional specificity that could contribute to differences in antiprion effects.

Since regions of Ydj1p outside the J domain are dispensable for curing, substrate specificity of curing effects for different prions must be mediated by other chaperones. Members of Hsp100, Hsp70, and Hsp40 chaperone families cooperate to resolubilize and refold aggregated proteins (GLOVER and LINDQUIST 1998; GURLEY 2000; ZIETKIEWICZ et al. 2004). Although the Hsp70/Hsp40 system can promote refolding of proteins from smaller aggregates, solubilization of larger aggregates requires the addition of Hsp104. When assayed in vitro, Hsp104 shows refolding activity only in the presence of both Hsp70 and Hsp40. So, while a direct interaction of Sis1p with Rnq1p appears to be important for propagation of the [PIN⁺] prion (DOUGLAS et al. 2008), Hsp70 or Hsp104 could be contributing significantly to recognition of prions as substrates. Hsp70/40 are believed to interact first with larger aggregates to make them more accessible to the action of Hsp100 chaperones (ZIETKIEWICZ et al. 2004), so Hsp70 could be a primary factor for specifying curing effects of chaperone machinery with different prions. The ability of Hsp40 to interact with and regulate specific Hsp70 isoforms could determine which prion will be affected by altered expression of a particular Hsp40.

Hsp40, Hsp70, and Hsp104 are similar in that they all are important for normal prion propagation and yet have antiprion effects when overexpressed. The inhibition of prion propagation by mutants of any of these chaperone families shows that prion propagation in yeast depends on their normal activity to prevent protein aggregation and resolubilize proteins from aggregates. Nevertheless, overexpressing Hsp40 and Hsp70 antagonizes [URE3] but has little or no effect on other prions, while overexpressing Hsp104 cures [PSI⁺] but not other prions. Although overexpression of a component of the chaperone machinery is usually envisioned to disrupt prion propagation by altering a direct functional interaction of the machinery with the prion, we showed earlier that a structurally altered Hsp104 that was fully functional in both protein disaggregation and prion propagation was incapable of curing [PSI⁺] when overexpressed, indicating that [PSI⁺] curing by overexpressed Hsp104 is not due to disruption or enhancement of a direct interaction of the Hsp40/70/104 machinery with [PSI⁺] prions (HUNG and MASISON 2006). Thus it is likely, at least in some instances, that the curing effect of overproduced chaperones is due to alteration of an unidentified cellular process in a way that interferes with prion propagation rather than a direct effect on chaperone machinery interaction with prion protein. Such a process could involve active transmission of prions from mother to daughter or sequestration or degradation of prion aggregates.

Our data also identify residues important for Hsp40 function. Among the J-domain mutations, L11P and A30T are novel, which suggests these conserved residues are important for Hsp40/Hsp70 interaction. A30 is in close spatial proximity to F47 (see Figure 2) and these particular residues may be important to provide proper spacing between helices II and III for structural integrity of the J domain or proper positioning of the HPD motif. L11 is near D9 and directly across from Y26, which is analogous to residues Y25 of E. coli DnaJ and Y24 of human Hdj1p that were shown to be important for Hsp40 function (GENEVAUX et al. 2002). These data again suggest these residues might be important for J-domain structure or transmission of signals to Hsp70. Our rearranging of substitutions also showed that other J-domain residues were not important for Ydj1p curing activity. The combination of substitutions K48R, P60L, and D67G had only a modest effect on curing (see Table 1), indicating that these residues likely are not critical for structural integrity of the J domain, interaction of Ydj1p with Hsp70, or stimulation of Hsp70 ATPase.

This research was supported by the intramural program of the National Institute of Diabetes, Digestive and Kidney Diseases, National Institutes of Health.

¹ These authors contributed equally to this work.

ARON, R., T. HIGURASHI, C. SAHI and E. A. CRAIG, 2007 J-protein co-chaperone Sis1 required for generation of [RNQ+] seeds necessary for prion propagation. EMBO J. 26: 3794–3803.[CrossRef][Medline]

BRACHMANN, A., U. BAXA and R. B. WICKNER, 2005 Prion generation in vitro: amyloid of Ure2p is infectious. EMBO J. 24: 3082–3092.[CrossRef][Medline]

BUKAU, B., J. WEISSMAN and A. HORWICH, 2006 Molecular chaperones and protein quality control. Cell 125: 443–451.[CrossRef][Medline]

CAPLAN, A. J., and M. G. DOUGLAS, 1991 Characterization of YDJ1: a yeast homologue of the bacterial dnaJ protein. J. Cell Biol. 114: 609–621.[Abstract/Free Full Text]

CAPLAN, A. J., D. M. CYR and M. G. DOUGLAS, 1992a YDJ1p facilitates polypeptide translocation across different intracellular membranes by a conserved mechanism. Cell 71: 1143–1155.[CrossRef][Medline]

CAPLAN, A. J., J. TSAI, P. J. CASEY and M. G. DOUGLAS, 1992b Farnesylation of YDJ1p is required for function at elevated growth temperatures in Saccharomyces cerevisiae. J. Biol. Chem. 267: 18890–18895.[Abstract/Free Full Text]

DOUGLAS, P. M., S. TREUSCH, H. Y. REN, R. HALFMANN, M. L. DUENNWALD et al., 2008 Chaperone-dependent amyloid assembly protects cells from prion toxicity. Proc. Natl. Acad. Sci. USA 105: 7206–7211.[Abstract/Free Full Text]

FAN, C. Y., H. Y. REN, P. LEE, A. J. CAPLAN and D. M. CYR, 2005 The type I Hsp40 zinc finger-like region is required for Hsp70 to capture non-native polypeptides from Ydj1. J. Biol. Chem. 280: 695–702.[Abstract/Free Full Text]

GENEVAUX, P., F. SCHWAGER, C. GEORGOPOULOS and W. L. KELLEY, 2002 Scanning mutagenesis identifies amino acid residues essential for the in vivo activity of the Escherichia coli DnaJ (Hsp40) J-domain. Genetics 162: 1045–1053.[Abstract/Free Full Text]

GLOVER, J. R., and S. LINDQUIST, 1998 Hsp104, Hsp70, and Hsp40: a novel chaperone system that rescues previously aggregated proteins. Cell 94: 73–82.[CrossRef][Medline]

GURLEY, W. B., 2000 HSP101: a key component for the acquisition of thermotolerance in plants. Plant Cell 12: 457–460.[Free Full Text]

HIGURASHI, T., J. K. HINES, C. SAHI, R. ARON and E. A. CRAIG, 2008 Specificity of the J-protein Sis1 in the propagation of 3 yeast prions. Proc. Natl. Acad. Sci. USA 105: 16596–16601.[Abstract/Free Full Text]

HUANG, K., J. M. FLANAGAN and J. H. PRESTEGARD, 1999 The influence of C-terminal extension on the structure of the "J-domain" in E. coli DnaJ. Protein Sci. 8: 203–214.

HUNG, G. C., and D. C. MASISON, 2006 N-terminal domain of yeast Hsp104 chaperone is dispensable for thermotolerance and prion propagation but necessary for curing prions by Hsp104 overexpression. Genetics 173: 611–620.[Abstract/Free Full Text]

JONES, G. W., and D. C. MASISON, 2003 Saccharomyces cerevisiae Hsp70 mutations affect [PSI⁺] prion propagation and cell growth differently and implicate Hsp40 and tetratricopeptide repeat cochaperones in impairment of [PSI⁺]. Genetics 163: 495–506.[Abstract/Free Full Text]

JONES, G. W., Y. SONG, S. CHUNG and D. C. MASISON, 2004 Propagation of yeast [PSI⁺] is prion impaired by factors that regulate Hsp70 substrate binding. Mol. Cell. Biol. 24: 3928–3937.[Abstract/Free Full Text]

JUNG, G., and D. C. MASISON, 2001 Guanidine hydrochloride inhibits Hsp104 activity in vivo: a possible explanation for its effect in curing yeast prions. Curr. Microbiol. 43: 7–10.[CrossRef][Medline]

KRYNDUSHKIN, D., and R. B. WICKNER, 2007 Nucleotide exchange factors for Hsp70s are required for [URE3] prion propagation in Saccharomyces cerevisiae. Mol. Biol. Cell 18: 2149–2154.[Abstract/Free Full Text]

KUSHNIROV, V. V., D. S. KRYNDUSHKIN, M. BOGUTA, V. N. SMIRNOV and M. D. TER-AVANESYAN, 2000 Chaperones that cure yeast artificial [PSI⁺] and their prion-specific effects. Curr. Biol. 10: 1443–1446.[CrossRef][Medline]

LEE, D. H., M. Y. SHERMAN and A. L. GOLDBERG, 1996 Involvement of the molecular chaperone Ydj1 in the ubiquitin-dependent degradation of short-lived and abnormal proteins in Saccharomyces cerevisiae. Mol. Cell. Biol. 16: 4773–4781.[Abstract/Free Full Text]

LI, J., X. QIAN and B. SHA, 2003 The crystal structure of the yeast Hsp40 Ydj1 complexed with its peptide substrate. Structure 11: 1475–1483.[Medline]

LI, J., and B. SHA, 2005 Structure-based mutagenesis studies of the peptide substrate binding fragment of type I heat-shock protein 40. Biochem. J. 386: 453–460.[CrossRef][Medline]

LIAN, H. Y., H. ZHANG, Z. R. ZHANG, H. M. LOOVERS, G. W. JONES et al., 2007 Hsp40 interacts directly with the native state of the yeast prion protein Ure2 and inhibits formation of amyloid-like fibrils. J. Biol. Chem. 282: 11931–11940.[Abstract/Free Full Text]

LINKE, K., T. WOLFRAM, J. BUSSEMER and U. JAKOB, 2003 The roles of the two zinc binding sites in DnaJ. J. Biol. Chem. 278: 44457–44466.[Abstract/Free Full Text]

LOPEZ, N., R. ARON and E. A. CRAIG, 2003 Specificity of class II Hsp40 Sis1 in maintenance of yeast prion. [RNQ+] Mol. Biol. Cell 14: 1172–1181.

LU, Z., and D. M. CYR, 1998 Protein folding activity of Hsp70 is modified differentially by the hsp40 co-chaperones Sis1 and Ydj1. J. Biol. Chem. 273: 27824–27830.[Abstract/Free Full Text]

MANDAL, A. K., N. B. NILLEGODA, J. A. CHEN and A. J. CAPLAN, 2008 Ydj1 protects nascent protein kinases from degradation and controls the rate of their maturation. Mol. Cell. Biol. 28: 4434–4444.[Abstract/Free Full Text]

MASISON, D. C., M. L. MADDELEIN and R. B. WICKNER, 1997 The prion model for [URE3] of yeast: spontaneous generation and requirements for propagation. Proc. Natl. Acad. Sci. USA 94: 12503–12508.[Abstract/Free Full Text]

MASISON, D. C., and R. B. WICKNER, 1995 Prion-inducing domain of yeast Ure2p and protease resistance of Ure2p in prion-containing cells. Science 270: 93–95.[Abstract/Free Full Text]

MORIYAMA, H., H. K. EDSKES and R. B. WICKNER, 2000 [URE3] Prion propagation in Saccharomyces cerevisiae: requirement for chaperone Hsp104 and curing by overexpressed chaperone Ydj1p. Mol. Cell. Biol. 20: 8916–8922.[Abstract/Free Full Text]

RAMOS, C. H., C. L. OLIVEIRA, C. YANG-FAN, I. L. TORRIANI and D. M. CYR, 2008 Conserved central domains control the quaternary structure of type I and type II Hsp40 molecular chaperones. J. Mol. Biol. 383: 155–166.[CrossRef][Medline]

ROTHSTEIN, R., 1991 Targeting, disruption, replacement, and allele rescue: integrative DNA transformation in yeast. Methods Enzymol. 194: 281–301.[CrossRef][Medline]

SAHI, C., and E. A. CRAIG, 2007 Network of general and specialty J protein chaperones of the yeast cytosol. Proc. Natl. Acad. Sci. USA 104: 7163–7168.[Abstract/Free Full Text]

SAVISTCHENKO, J., J. KRZEWSKA, N. FAY and R. MELKI, 2008 Molecular chaperones and the assembly of the prion Ure2p in vitro. J. Biol. Chem. 283: 15732–15739.[Abstract/Free Full Text]

SCHLUMPBERGER, M., S. B. PRUSINER and I. HERSKOWITZ, 2001 Induction of distinct [URE3] yeast prion strains. Mol. Cell. Biol. 21: 7035–7046.[Abstract/Free Full Text]

SCHWIMMER, C., and D. C. MASISON, 2002 Antagonistic interactions 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.[Abstract/Free Full Text]

SHARMA, D., and D. C. MASISON, 2008 Functionally redundant isoforms of a yeast hsp70 chaperone subfamily have different antiprion effects. Genetics 179: 1301–1311.[Abstract/Free Full Text]

SHERMAN, F., 1991 Getting started with yeast. Methods Enzymol. 194: 3–21.[CrossRef][Medline]

SIKORSKI, R. S., and P. HIETER, 1989 A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122: 19–27.[Abstract/Free Full Text]

SONDHEIMER, N., N. LOPEZ, E. A. CRAIG and S. LINDQUIST, 2001 The role of Sis1 in the maintenance of the [RNQ+] prion. EMBO J. 20: 2435–2442.[CrossRef][Medline]

SUH, W. C., C. Z. LU and C. A. GROSS, 1999 Structural features required for the interaction of the Hsp70 molecular chaperone DnaK with its cochaperone DnaJ. J. Biol. Chem. 274: 30534–30539.[Abstract/Free Full Text]

TSAI, J., and M. G. DOUGLAS, 1996 A conserved HPD sequence of the J-domain is necessary for YDJ1 stimulation of Hsp70 ATPase activity at a site distinct from substrate binding. J. Biol. Chem. 271: 9347–9354.[Abstract/Free Full Text]

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.[CrossRef][Medline]

ZIETKIEWICZ, S., J. KRZEWSKA and K. LIBEREK, 2004 Successive and synergistic action of the Hsp70 and Hsp100 chaperones in protein disaggregation. J. Biol. Chem. 279: 44376–44383.[Abstract/Free Full Text]

Communicating editor: M. D. ROSE

Related articles in Genetics:

ISSUE HIGHLIGHTS

Genetics 2009 181: NP. [Full Text]  

International Access Link

Online ISSN: 1943-2361  Print ISSN: 0016-6731
Copyright 2009 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu