ABSTRACT
Saccharomyces cerevisiae contains several prion elements, which are epigenetically transmitted as self-perpetuating protein conformations. One such prion is [SWI+], whose protein determinant is Swi1, a subunit of the SWI/SNF chromatin-remodeling complex. We previously reported that [SWI+] formation results in a partial loss-of-function phenotype of poor growth in nonglucose medium and abolishment of multicellular features. We also showed that the first 38 amino acids of Swi1 propagated [SWI+]. We show here that a region as small as the first 32 amino acids of Swi1 (Swi11–32) can decorate [SWI+] aggregation and stably maintain and transmit [SWI+] independently of full-length Swi1. Regions smaller than Swi11–32 are either incapable of aggregation or unstably propagate [SWI+]. When fused to Sup35MC, the [PSI+] determinant lacking its PrD, Swi11–31 and Swi11–32 can act as transferable prion domains (PrDs). The resulting fusions give rise to a novel chimeric prion, [SPS+], exhibiting [PSI+]-like nonsense suppression. Thus, an NH2-terminal region of ∼30 amino acids of Swi1 contains all the necessary information for in vivo prion formation, maintenance, and transmission. This PrD is unique in size and composition: glutamine free, asparagine rich, and the smallest defined to date. Our findings broaden our understanding of what features allow a protein region to serve as a PrD.
INTRODUCTION
Self-perpetuating conformational conversion of the prion protein, PrP, underlies the pathogenesis of a group of mammalian neurodegenerative disorders known as transmissible spongiform encephalopathies (TSEs), which are characterized by the formation of protease-resistant amyloid and neuronal loss (1, 2). Recent evidence suggests that the prion concept can be extended to other neurodegenerative diseases, such as Alzheimer's disease (AD), Parkinson's disease (PD), Huntington's disease (HD), and amyotrophic lateral sclerosis (ALS) (3–6). Interestingly, the budding yeast Saccharomyces cerevisiae also contains several protein-based epigenetic elements, known as yeast prions, which propagate as altered conformations resulting in the transmission of heritable phenotypes (7, 8). These proteins with diverse functions, including Sup35, Ure2, Rnq1, Swi1, Cyc8, Mot3, Mod5, Nup100, and Lsb2, are the protein determinants of the yeast prions [PSI+] (9), [URE3] (10, 11), [PIN+] (12, 13), [SWI+] (14), [OCT+] (15), [MOT3+] (16), [MOD+] (17), [NUP100+] (18), and [LSB+] (19), respectively. Yeast prion proteins contain regions termed prion domains (PrDs), which are usually dispensable for normal protein function but essential and sufficient for prion formation and propagation. Most PrDs have been found to be biased in their amino acid compositions, as they are enriched in glutamine and asparagine residues (20). However, some proteins that are not enriched in glutamine or asparagine can also switch to altered conformations and confer prion-like patterns of inheritance (17, 21, 22). It has been shown that extracellular environments can modulate the frequency of de novo formation and loss of yeast prions, demonstrating the important role of prion-mediated inheritance in adaptation (23–26). Therefore, elucidating the determinants of PrDs will shed light on our understanding of how prions are formed and propagated in vivo in response to changes in environmental conditions.
The prion protein Swi1 is a subunit of the SWI/SNF chromatin remodeling complex, which regulates approximately 6% of gene expression in yeast (27). Swi1 forms aggregates in [SWI+] cells, and cells harboring [SWI+] exhibit a partial loss-of-function phenotype of poor growth in nonglucose (e.g., raffinose) medium (14) and a complete loss of multicellular features (28). Recently, Nizhnikov et al. demonstrated that the interaction of [SWI+] and [PIN+] causes inactivation of the SUP45 gene that leads to transmission of heritable phenotypes of nonsense suppression in strains bearing a deleted or modified Sup35 N-terminal domain (29).
Importantly, loss of Swi1 production causes [SWI+] elimination, confirming Swi1 as the protein determinant of [SWI+] (14). Furthermore, [SWI+] was shown to be “infectious,” as cytoduction, cytoplasmic mixing in the absence of nuclear fusion, results in transmission of [SWI+] to nonprion cells (14).
Investigation into Swi1's PrD has focused on the NH2-terminal region of Swi1 due to its high glutamine/asparagine content. It has been shown that deletion of the first 327 amino acids (Swi11–327 or Swi1N) at the chromosomal SWI1 locus resulted in [SWI+] loss (30). Additionally, transformation of fibrils made from recombinant Swi1N protein resulted in de novo formation of [SWI+] in naive nonprion cells, demonstrating that the Swi1 PrD lies within the Swi1N region and amyloid is the structural basis of [SWI+] (30). Further deletion analysis of Swi1N showed that a region consisting of the first 38 amino acids of Swi1, which is rich in asparagine but lacking glutamine, could propagate the prion fold of [SWI+] in the absence of full-length Swi1 (31). Still to be determined is whether smaller regions of Swi1 can propagate [SWI+] and promote de novo prion formation and thus constitute the minimal Swi1 PrD.
In this report, we have examined the ability of small truncation mutants ranging from Swi11–26 to Swi11–37 to propagate [SWI+] and found that the minimum length of Swi1 NH2-terminal regions required for [SWI+] propagation differs at different Swi1 expression levels. Under our examined conditions, a region as small as the first 32 amino acid residues (Swi11–32) is sufficient to stably support [SWI+] propagation. While Swi11–31 is not able to stably propagate [SWI+], it can confer [PSI+]-like prion phenotypes when fused to Sup35's MC region, establishing that this small region can serve as a transferable PrD.
RESULTS
Minimal region of Swi1 required for [SWI+] propagation is dependent on Swi1 levels.We previously showed that a small region consisting of the first 38 amino acids of Swi1 fused with yellow fluorescent protein (YFP) (Swi11–38YFP) is able to decorate [SWI+] aggregates (31). Swi11–38YFP was also shown to maintain an aggregated prion conformation in the absence of Swi1 and to transmit the prion conformation to endogenous Swi1 (31). However, truncation to the first 32 amino acids resulted in a significant loss of aggregation (31). To examine whether regions between 32 and 38 amino acids were able to form Swi1 prion aggregates in [SWI+] cells and were capable of prion transmission, corresponding truncation mutants were constructed and fused to a C-terminal YFP reporter (Fig. 1A). These truncation mutants (Swi1TruncYFP) were expressed in [SWI+] and [swi−] cells of either wild-type (WT) BY4741 or swi1Δ BY4741 cells expressing full-length Swi1 from a plasmid, p416TEFSwi1 (pSwi1) (Fig. 1B). All proteins were expressed at the expected sizes (Fig. 1F). We found a size-dependent decrease in the aggregation frequency in both WT and swi1Δ/pSwi1 [SWI+] cells: as Swi1 was increasingly truncated, it became less aggregated (Fig. 1B and D). Interestingly, the minimum region required for aggregation differed depending on the full-length Swi1 protein levels. In WT [SWI+] cells, which express Swi1 from the endogenous chromosomal Swi1 promoter, we found that only 5 to 10% of [SWI+] cells expressing Swi11–32YFP contained aggregation. In contrast, in swi1Δ/pSwi1 [SWI+] cells, which express Swi1 at higher levels from the TEF1 promoter (Fig. 1E) (30), ∼40% of cells contained Swi11–32YFP aggregates (Fig. 1B and D). This result encouraged us to construct a set of additional truncation mutants ranging from Swi11–26 to Swi11–31 (Fig. 1A). Similar experiments were carried out, and we found that although further deletion to Swi11–31YFP resulted in a complete loss of aggregation in WT [SWI+] cells, significant aggregation of Swi11–31YFP was observed in swi1Δ/pSwi1 [SWI+] cells and detectable aggregation could be seen in cells expressing Swi11–28YFP (Fig. 1C and D). All truncation mutants exhibited diffuse fluorescence in [swi−] cells, suggesting that overexpression of Swi1 truncation mutants per se did not result in aggregation (Fig. 1B and C, bottom). Taken together, these results show that the increase of Swi1 protein levels promotes prion-like aggregation of smaller NH2-terminal regions of Swi1.
Minimal region of Swi1 required for [SWI+] propagation is dependent on Swi1 levels. (A) Diagram illustrating Swi1 and its truncation mutants used in this study. The amino acid sequences of the truncation mutants are shown. Asparagine residues are highlighted in blue. (B) Fluorescence microscopy using YFP fusions of Swi11–38 to Swi11–32. Swi1N and Swi11–38, which were previously shown to propagate [SWI+], were used as positive controls, and YFP was used as a negative control. Truncation mutants were expressed in WT and swi1Δ/pSwi1 cells in both [SWI+] (top) and [swi−] (bottom) backgrounds. Three individual transformants were imaged for each truncation mutant, and representative images are shown. (C) Additional fluorescence microscopy using YFP fusions of Swi11–31 to Swi11–26. Experiments were done as described for panel B. (D) Percentage of cells containing aggregates in WT [SWI+] and swi1Δ/pSwi1 [SWI+] cells imaged in panels B and C. Aggregation was averaged for three individual transformants. Bars indicate standard errors. (E) Western blot showing YFP-tagged Swi1-NYFP expressed from the TEF1 promoter and SWI1 promoter in WT BY4741 [swi−] cells. Blots were probed with an anti-GFP antibody and antiactin antibody for a loading control. Values to the left of blots (E and F) are in kilodaltons. (F) Western blot showing YFP-tagged Swi1 truncation mutants expressed at the expected sizes in WT BY4741 [SWI+] (left) and [swi−] cells (right). The blots were then stripped and reprobed with an antiactin antibody for a loading control (bottom).
Swi11–32 stably maintains [SWI+] in the absence of full-length Swi1.We next asked if the aggregatable small regions of Swi1 could maintain an aggregated prion conformation in the absence of full-length Swi1. To this end, swi1Δ/pSwi1/pSwi1TruncYFP [SWI+] cells were grown in medium supplemented with 5-fluoroorotic acid (5-FOA) to counterselect against pSwi1, eliminating the source of full-length Swi1 (Fig. 2A). If Swi1 truncation mutants maintained [SWI+], we would expect to see fluorescent foci. In contrast, if Swi1 truncation mutants did not maintain [SWI+], we would expect to see diffuse YFP fluorescence. In both cases, if full-length Swi1 was eliminated, we expected a transition from a phenotype of reduced growth on medium using raffinose, a nonglucose sugar, as the sole carbon source (Raf±) to a phenotype of no growth on raffinose medium (Raf−) due to the loss of Swi1 function. In agreement with our previous report (31), we found that cells of swi1Δ/Swi11–38YFP and swi1Δ/Swi1N-YFP backgrounds maintained the [SWI+] conformation in the absence of Swi1 (Fig. 2B). Moreover, regions as small as Swi11–30 maintained the [SWI+] conformation in the absence of pSwi1. We did, however, observe changes in the stability of [SWI+] in cells expressing smaller truncations in the absence of Swi1. We found that only 4% of the 5-FOA-treated (−pSwi1) Swi11–30YFP colonies examined maintained aggregation and that [SWI+] was frequently lost upon passaging (data not shown). 5-FOA-treated Swi11–31YFP and Swi11–32YFP cells maintained aggregation in 14% and 80% of colonies examined, respectively. Interestingly, in the absence of full-length Swi1, we observed a change in the morphology of the aggregates upon shortening of the expressed fragment length. While cells containing Swi11–38YFP to Swi11–32YFP had aggregates whose morphology was exclusively dot-shaped in nature, cells containing Swi11–31YFP and Swi11–30YFP contained not only the dot-like aggregates but also ring-like aggregates in [SWI+] cells lacking Swi1 (Fig. 2B). These results demonstrate that while smaller regions of Swi1 can maintain [SWI+] in the absence of the full-length Swi1, Swi11–32 is the smallest domain able to do so stably under our examined conditions.
Swi11–32 stably maintains [SWI+] in the absence of full-length Swi1. (A) Schematic of the experimental design. swi1Δ/pSwi1/pSwi1TruncYFP [SWI+] cells were treated with 5-FOA to counterselect against pSwi1. After loss of full-length Swi1 (pSwi1), swi1Δ/pSwi1TruncYFP cells that maintain [SWI+] are expected to have Swi1 aggregates (middle), while cells that have not maintained [SWI+] are expected to lack aggregation (right). Both cell types will have a Raf− phenotype if full-length Swi1 is eliminated. (B) Fluorescence microscopy using YFP fusion proteins of Swi1 truncation mutants. Negative (YFP) and positive (Swi1N and Swi11–38) controls for Swi1 aggregation are also shown. A minimum of six individual colonies from each truncation mutant were imaged, and representative images are shown. (C) Loss of full-length Swi1 (pSwi1) in swi1Δ/pSwi1TruncYFP cells treated with 5 mM GdnHCl was confirmed by assessing the growth phenotype on medium containing raffinose as the sole carbon source. Log-phase cells were serially diluted and spotted onto rich medium (YPD) or raffinose medium (raffinose). WT, [SWI+], and swi1Δ cells are included as controls. (D) Diagram showing experimental design of RT-PCR to detect the presence or absence of the SWI1 gene in the examined strains (left). RT-PCR was performed with Swi1-specific primers. swi1Δ/pSwi1/pSwi1TruncYFP [SWI+] cells (+) contain SWI1 transcript, while 5-FOA-treated cells do not contain SWI1 transcript (−). [swi−] and swi1Δ controls are included (right). Values to the left of gels are in base pairs.
To confirm that the treatment with 5-FOA had indeed resulted in the loss of pSwi1, we assessed the raffinose phenotype of cells before and after treatment with 5-FOA. In agreement with our published results, we found that WT [SWI+] cells, which express Swi1 endogenously, displayed a Raf± phenotype, whereas swi1Δ cells, which do not express Swi1, had no growth on raffinose medium (Raf−) (Fig. 2C) (14). We found that swi1Δ/pSwi1TruncYFP [SWI+] cells expressing pSwi1 (+pSwi1) exhibited a Raf± phenotype; however, the 5-FOA-treated cells (−pSwi1) displayed a Raf− phenotype, suggesting that the pSwi1 plasmid was indeed lost as a result of 5-FOA treatment (data not shown). Next, we treated cells with 5 mM guanidine hydrochloride (GdnHCl), a chemical known to inhibit prion propagation by inactivating the molecular chaperone Hsp104, a disaggregase that is required for [SWI+] propagation (14). After treatment with GdnHCl, the growth of swi1Δ/pSwi1/pSwi1TruncYFP cells on raffinose medium was restored to levels comparable to those of [swi−] cells (Fig. 2C). In contrast, after GdnHCl treatment, swi1Δ/pSwi1TruncYFP cells still maintained the Raf− phenotype, confirming that these cells had in fact lost full-length Swi1. The loss of pSwi1 for the 5-FOA-treated cells was further confirmed by a reverse transcription PCR (RT-PCR) experiment using primers specific for the C-terminal region of Swi1. An ∼300-bp band representative of the targeted region of Swi1 was observed in the swi1Δ/pSwi1/pSwi1TruncYFP samples, while this band was absent in the 5-FOA-treated cells, confirming that pSwi1 was in fact lost (Fig. 2D).
Insoluble protein aggregates are formed by truncated Swi1.To complement our microscopic analysis, we studied the solubility of Swi11–32YFP and Swi11–38YFP by a centrifugation assay (Fig. 3A). Lysates from [SWI+] and [swi−] cells of a swi1Δ/pSwi11–32YFP or swi1Δ/pSwi11–38YFP background but lacking endogenous Swi1 were separated by centrifugation at 20,000 × g and resolved by SDS-PAGE. Western blot analysis revealed that Swi11–38YFP protein was found mostly in the pelleted, insoluble fraction in [SWI+] cells. Analysis also revealed that a fraction of Swi11–32YFP protein was found in the pelleted insoluble fraction of [SWI+] cells. In contrast, Swi11–32YFP or Swi11–38YFP proteins were found entirely in the soluble, supernatant fraction in [swi−] cells. As a control, actin was found only in the total and supernatant fractions (Fig. 3A). This result demonstrates that Swi11–32YFP and Swi11–38YFP can maintain the [SWI+] fold in the form of insoluble aggregates in [SWI+] cells without full-length Swi1.
Prion-like features of Swi1 truncation mutants. (A) The insoluble fractions of Swi11–38 and Swi11–32 were seen in [SWI+] samples but were not observed in [swi−] samples. [SWI+] and [swi−] cells of a swi1Δ/pSwi11–38YFP or swi1Δ/pSwi11–32YFP background lacking Swi1 were lysed, cleared, and separated by high-speed centrifugation (20,000 × g) as described in Materials and Methods. Cleared whole-cell lysate (T), supernatant (S), and pellet (P) fractions were resolved by SDS-PAGE and probed with anti-GFP antibody by Western blotting. The blot was stripped and reprobed with an antiactin antibody for a loading control. (B) The function of Hsp104 is required for [SWI+]Trunc propagation. [SWI+] cells of swi1Δ/pSwi1N-YFP, swi1Δ/pSwi11–38YFP, or swi1Δ/pSwi11–32YFP background were transformed with an empty vector of pRS316 or pKT218,620 (Hsp104DN). Aggregation of truncation mutants was assessed by fluorescence microscopy. Aggregation was averaged for six individual colonies grown in liquid medium for Swi1N and Swi11–32 and three colonies for Swi11–38. Bars indicate standard errors.
Functional Hsp104 is required for maintaining the prion conformation of truncated Swi1.Most yeast prions examined to date, including [SWI+], require the function of Hsp104 for their propagation. A treatment with millimolar levels of GdnHCl, which inactivates Hsp104, or expression of a dominant negative mutant of Hsp104 eliminates these Hsp104-dependent prions (12, 14, 15, 32–35). To test if the truncated Swi1 prions also require Hsp104 function for propagation, we transformed [SWI+] cells of a swi1Δ/pSwi1N-YFP, swi1Δ/pSwi11–38YFP, or swi1Δ/pSwi11–32YFP background with a dominant negative variant of Hsp104, pKT218,620 (Hsp104DN), which has been shown to cure other amyloid yeast prions (15, 33, 35–37). We found that the YFP fusion aggregation decreased dramatically for Swi1N and Swi11–38 after overproduction of Hsp104DN in comparison to vector controls (Fig. 3B). We also found that Swi11–32 aggregation was completely gone in cells transformed with Hsp104DN as opposed to the persistent aggregation in cells transformed with a vector control. Similar results were obtained after serially passaging cells on medium containing 5 mM GdnHCl (data not shown), demonstrating that the aggregation of Swi1N-YFP, Swi11–38YFP, and Swi11–32YFP is curable by Hsp104 deficiency.
[SWI+] maintained by small NH2-terminal fragments of Swi1 can transmit the prion fold to full-length Swi1.Next, we examined if small regions of Swi1 could propagate [SWI+] by transmitting a prion fold back to full-length Swi1. As illustrated in Fig. 4A, [SWI+] cells of a swi1Δ/pSwi1TruncYFP background were transformed with mCherry-tagged Swi1, pSwi1mCherry, and an aggregation assay was carried out to determine if the small Swi1Trunc prion was able to transmit the prion fold to the Swi1mCherry. Consistent with our previous report, we found that cells of swi1Δ/pSwi1mCherry/pSwi11–38YFP or swi1Δ/pSwi1mCherry/pSwi1N-YFP background harbored punctate fluorescent foci when mCherry fluorescence was examined (Fig. 4B), indicating that these two regions are able to transmit the prion fold back to Swi1 (31). We found that truncation mutants ranging from Swi11–37YFP to Swi11–31YFP were also able to transmit the prion fold back to full-length Swi1, as these cells exhibited punctate foci formed by Swi1mCherry (Fig. 4B). Additionally, our aforementioned results showed that Swi11–30 is able to form prion aggregates and maintain [SWI+] in the absence of full-length Swi1, but [SWI+] was unstable. We found that such a prion fold could be transmitted back to full-length Swi1, as Swi1mCherry foci were observed upon transformation (Fig. 4B). Remarkably, most ring-shaped aggregates formed by Swi11–31YFP and Swi11–30YFP transition back into dot-like aggregates when cells are retransformed with full-length Swi1 (Fig. 4B and C and data not shown). Interestingly, Swi1TruncYFP aggregates were found to be colocalized with full-length Swi1mCherry in a portion of the cells that coexpress both proteins, suggesting that small domains of Swi1 may cross-seed or decorate full-length Swi1 (Fig. 4C).
[SWI+] maintained by small NH2-terminal fragments of Swi1 can transmit the prion fold to full-length Swi1. (A) Schematic of the experimental design. Individual swi1Δ/pSwi1TruncYFP colonies were transformed with full-length Swi1mCherry (pSwi1mCherry). If a Swi1 truncation mutant can transmit the prion fold to Swi1, we expect the appearance of Swi1mCherry foci, but otherwise a diffuse mCherry signal is expected. (B) Fluorescence microscopy using mCherry-tagged Swi1. Six individual colonies from each truncation mutant were imaged, and representative images are shown. (C) Colocalization of YFP-tagged truncation mutants and mCherry-tagged full-length Swi1 is shown by fluorescence microscopy.
A small NH2-terminal region of Swi1 is a transferable PrD that supports prion de novo formation.We next examined whether Swi11–38, Swi11–32, and Swi11–31 could be transferable and could support de novo prion formation using a well-established Sup35 reporter assay (12, 16, 38).
Sup35, the protein determinant of the [PSI+] prion, is a translational termination factor containing three domains—an amino-terminal prion domain that is glutamine/asparagine rich and essential for [PSI+] formation and propagation (N), a highly charged middle domain (M), and a C-terminal functional domain necessary for translational termination (C) (39). The prion state of Sup35 can be easily assessed in a strain containing the ade1-14 allele with a premature stop codon in the open reading frame (ORF) of ADE1 (33). In nonprion cells, functional Sup35 efficiently terminates translation at the premature stop codon, resulting in a lack of growth on medium lacking adenine (−Ade) and red colonies on yeast extract-peptone-dextrose (YPD) medium due to the blockage of the adenine biosynthesis pathway leading to pigment accumulation. In contrast, Sup35 is aggregated in [PSI+] cells, and its function of translation termination is compromised, resulting in the growth on −Ade medium and pink/white colonies on YPD due to nonsense suppression. We replaced the N region of Sup35 with Swi11–38, Swi11–32, or Swi11–31, resulting in constructs expressing chimeric proteins Swi11–38MC, Swi11–32MC, and Swi11–31MC (Swi1TruncMC), respectively, under the control of TEF1 promoter (Fig. 5A). Next, we introduced these Swi1TruncMC constructs into a W303 strain containing a chromosomal deletion of SUP35 and expressing full-length Sup35 from a plasmid, p316Sup35FL (40). Given that the Sup35FL-expressing plasmid was under uracil selection, we were able to counterselect to remove this plasmid by growing cells in 5-FOA medium. The resulting strains, which carried Swi11–38MC, Swi11–32MC, or Swi11–31MC as the only source of Sup35 function, produced red colonies on YPD medium (Fig. 5B, left, and data not shown). As Sup35 function is required for cell survival and the red color of cell colonies is an indication of a functional Sup35, our results suggest that Swi1TruncMC proteins are functional fusion proteins that can provide cells with the essential function of Sup35.
Small NH2-terminal regions of Swi1 are transferable and support prion de novo formation. (A) Diagram illustrating cloning scheme for Swi11–38MC, Swi11–32MC, and Swi11–31MC (Swi1TruncMC) fusions. Swi1Trunc was inserted in place of the NH2-terminal (amino acids 1 to 123) PrD of Sup35, resulting in Swi1TruncMC fusions that were expressed under the control of the TEF1 promoter. An asterisk denotes Swi1Trunc. (B) Experimental design. W303 sup35Δ/pSwi1TruncMC [psi−] strains with an Ade− phenotype and red color on YPD were transformed with the corresponding pSwi1TruncYFP. Transformants were grown and plated on −Ade plates to select for the prion state (white on YPD, Ade+, and aggregated Swi1TruncYFP). Prion candidate isolates were treated with 5 mM GdnHCl. If this GdnHCl treatment resulted in red, Ade− isolates with diffuse Swi1TruncYFP, the corresponding isolates were scored as a chimeric prion termed [SPS+]. (C) Representative [SPS+] isolates and their corresponding GdnHCl-treated [sps−] cells were streaked onto YPD medium and SC−Ade plates to assess the color phenotype and growth ability. sup35Δ/pSup35FL [PSI+] and [psi−] strains were streaked simultaneously as controls. (D) Western blot showing the presence of either full-length Sup35 in sup35Δ/p316Sup35FL cells or Swi1TruncMC fusion proteins in sup35Δ/pSwi1TruncMC/pSwi1TruncYFP [SPS+] cells. Sup35FL and Swi1TruncMC are expressed at different sizes when probed with C domain-specific anti-Sup35 antibody. The blot was stripped and reprobed with antiactin antibody. Values to the left of blots are in kilodaltons. (E) Microscopy assay showing aggregation status of cells described in panel C. (F) [SPS+]1–38, [SPS+]1–32, or [SPS+]1–31 and [sps−]1–38, [sps−]1–32, or [sps−]1–31 cells of a sup35Δ/pSwi1TruncMC/pSwi1TruncYFP background were subjected to a centrifugation assay as described in Materials and Methods. The cleared whole-cell lysate (T), supernatant (S), and pellet (P) fractions were resolved by SDS-PAGE, and the presence of Swi1TruncMC was detected by Western blotting with anti-Sup35 antibody.
Next, we introduced a second plasmid to overexpress Swi1TruncYFP that corresponded to the Swi1TruncMC in the cell, as this may increase the induction of the prion state of Swi1TruncMC in nonprion cells (Fig. 5B, middle). We confirmed that Swi1TruncMC fusion proteins were expressed at the expected size (Fig. 5D). Given the low rate of prion appearance in the absence of positive selection, we plated cells on SC medium lacking adenine to select for Ade+ colonies. After Swi1TruncYFP overexpression, cells that grew on SC−Ade and were white on YPD were considered putative prion candidates and were assayed further via examination of the aggregation state of Swi1TruncYFP through fluorescence microscopy assays. As shown in Fig. 5E, these candidate colonies contain punctate fluorescence foci and such aggregation could be stably inherited. Combined, our results demonstrated that Swi11–38MC, Swi11–32MC, and Swi11–31MC support de novo prion formation that can be promoted by overproduction of the corresponding Swi1 truncation mutant. We termed the emerged chimeric prions [SPS+]1–38, [SPS+]1–32, and [SPS+]1–31 (stands for Swi1 conferred [PSI+]).
We then tested the curability of [SPS+]1–38, [SPS+]1–32, and [SPS+]1–31 by 5 mM GdnHCl. We found that colonies became red in color on YPD after GdnHCl treatment, as would be expected if Swi11–38MC, Swi11–32MC, and Swi11–31MC were soluble and functional (Fig. 5B, right, and C). Further examination showed that GdnHCl-treated cells lacked growth on −Ade medium and exhibited diffuse fluorescence of Swi1TruncYFP, confirming that [SPS+]1–38, [SPS+]1–32, and [SPS+]1–31 are curable by GdnHCl, resulting in a nonprion state, [sps−] (Fig. 5C and E). While cured [sps−] cells stably maintained their nonprion conformation after GdnHCl treatment, [SPS+] cells did reappear at a rate of 1 to 2% for all three truncation mutants upon selection on medium lacking adenine.
Next, we examined the curability of [SPS+]1–32 and [SPS+]1–31 upon Hsp104 overexpression. While overexpression of Hsp104 has been shown to result in curing of [PSI+], such curing is not observed for [SWI+]. Therefore, we transformed [SPS+]1–38, [SPS+]1–32, and [SPS+]1–31 with either p2HG, a vector control, or p2HG-Hsp104, which expresses Hsp104 from a 2μ plasmid driven by a GPD promoter. Individual transformants were grown in selective medium and spread on YPD. While [SPS+]1–32 cells generated 5.5% red colonies when transformed with p2HG, overexpression of Hsp104 resulted in an appearance of red colonies of about 7.6%. Results were similar when the rate of curing was examined for [SPS+]1–31. While [SPS+]1–31 cells generated 10.9% red colonies when transformed with p2HG, overexpression of Hsp104 resulted in an appearance of red colonies of 19.2%. Therefore, our results demonstrate that like [SWI+], [SPS+]1–32 and [SPS+]1–31 are not significantly sensitive to Hsp104 overexpression, which differs from [PSI+].
We next assessed the solubility of the Swi1TruncMC fusion protein in both [SPS+] and [sps−] cells by a centrifugation assay and found that while the Swi11–38MC, Swi11–32MC, and Swi11–31MC fusion proteins were mostly in the soluble supernatant fraction in [sps−] cells, they were mostly found in the insoluble protein fraction in [SPS+] cells, suggesting a prion-mediated change in the solubility of the fusion proteins (Fig. 5F). Thus, our results demonstrate that Swi11–38MC, Swi11–32MC, and Swi11–31MC can exist in two heritable conformational states associated with distinct phenotypes.
DISCUSSION
A prion domain is a limited portion of a prion protein that is necessary and sufficient for prion formation and propagation. In this study, we found that a small region containing the first 32 amino acids of Swi1 is able to decorate [SWI+] aggregates and maintain the prion fold in the absence of full-length Swi1 (Fig. 1 and 2). In addition, when a region consisting of the first 31 amino acids of Swi1 was fused to Sup35MC, the resulting fusion was able to form a [PSI+]-like nonsense suppression prion ([SPS+]) de novo (Fig. 5). The Swi1 PrD is unique in some respects. First, its size, ∼30 amino acids, is unique. To our knowledge, the Swi1 PrD is the smallest PrD currently identified that supports de novo prion formation and propagation in vivo. Second, the composition of this small Swi1 PrD is also unique; being highly rich in asparagine and threonine, it is distinguishable from most other PrDs, which are also enriched in glutamine.
In general, the prevalence of glutamine/asparagine-rich regions in the proteomic sequences of different organisms may be an evolutionary selection for eukaryotic proteins as a means to regulate protein-protein interactions (42). The enrichment of glutamine and asparagine residues in PrDs, however, is not an absolute requirement, as some prion proteins, namely, PrP in mammals, Mod5 and several newly identified prions in S. cerevisiae (17, 21), and HET-s in Podospora anserine (43), lack an enrichment in these residues. Swi1 has a PrD that is uniquely glutamine free and asparagine rich. Research into the contributions of both glutamine and asparagine residues to prion formation and propagation have yielded interesting results that may give us some insight into the ability of these small regions of Swi1 to propagate [SWI+]. While the two amino acids differ only subtly in their chemical compositions, Halfmann et al. found that they have opposing effects on prion and amyloid formation (44). Examination of glutamine and asparagine replacement variants (all glutamines in the Sup35 PrD were replaced with asparagines [Sup35N] and vice versa [Sup35Q]) showed that while the Sup35N variant promoted self-templating amyloidogenesis and prion formation, the Sup35Q variant promoted the formation of nonamyloid aggregates and thus inhibited prionogenesis. Asparagine and glutamine substitution variants were also made for other prion proteins, and the prion-promoting effects of asparagine and the prion-inhibiting effect of glutamine were also observed, suggesting that prion promotion and inhibition are not protein specific but rather general properties of these amino acids (44). Additionally, examination of the thermodynamic properties of regions composed of 30 asparagines (N30) or 30 glutamines (Q30) found that the polyasparagine molecule had a higher propensity to form β-sheets and a lower disorderedness than the polyglutamine molecule (44). Combined, these studies demonstrate that glutamine and asparagine residues have dramatic and disparate effects on prion formation. It is proposed that the smaller side chains found in asparagine allow for stronger hydrogen binding than in glutamine, which in turn allows for the formation of β-sheets and a decreased amount of nonspecific interactions, thus preventing formation of nonamyloid pathway aggregates (off-pathway) (44). The unusually high asparagine content of the extreme N terminus of Swi1 is likely an important factor contributing to its ability to form and propagate [SWI+]. However, it was reported that while polyasparagine molecules 21 residues in length were highly disordered and had a high amyloidogenic propensity compared to polyglutamine sequences, these residues alone did not endorse prionogenicity (45), suggesting that being asparagine rich is not the sole factor determining PrD properties and that other residues are likely required for prionogenesis. Moreover, a longer polyasparagine peptide consisting of 104 asparagines did not demonstrate prion-like heritability, suggesting that simply increasing the size of this moderately amyloidogenic residue is not enough to cause prion formation (46).
Our study illustrates the importance of protein levels on the differing abilities of small regions of Swi1 to form prion aggregates and to maintain and propagate [SWI+]. While Swi11–31 and Swi11–30 are not able to form aggregates in WT [SWI+] cells, they do so when Swi1 protein levels are higher. Swi11–30 was capable not only of adopting the prion fold in swi1Δ/pSwi1 [SWI+] cells, which express Swi1 at a higher level than in WT cells, but also of propagating the acquired prion fold in the absence of full-length Swi1. Previous research has shown that overexpression of a prion protein or its PrD results in a significant increase in de novo prion formation (7). Our results show that the length requirement for a region that constitutes the Swi1 PrD is strongly influenced by the expression level of full-length Swi1. Increased levels of Swi1 may enable previously nonprionogenic fragments that are on the edge of aggregation to form prion aggregates. Increased levels of Swi1 in swi1Δ/pSwi1 [SWI+] cells may also explain the increased level of aggregation of Swi1TruncYFP in general that we observed in these cells compared to WT cells. Increased levels of Swi1 in the cells may result in an increased amount of prion seeds that are available for cross-seeding or decoration, this resulting in the increased levels of aggregation that we observed upon overexpression of our Swi1 truncation mutants.
Another interesting finding in this study was the change of aggregation morphology of Swi11–30 and Swi11–31 in swi1Δ/Swi1TruncYFP [SWI+] cells upon removal of the full-length Swi1. The presence of ring/rod-shaped (here referred to as ring-shaped) aggregates is usually observed in premature [PSI+], [PIN+], and [SWI+] cells (47, 48). Although combined results from various studies have shown that these premature aggregates are processed into dot-shaped aggregates, and this processing is often linked to the transition of premature prion aggregation to mature, stable prion conformations (48–52), a recent study suggests that the de novo formation of [PSI+] can involve multiple pathways and both ring- and dot-shaped Sup35 aggregations are prionogenic (53). Our results show that in the presence of full-length Swi1, Swi11–30 and Swi11–31 formed punctate aggregates, some of which transitioned into ring-shaped aggregates when the full-length protein was lost. Remarkably, most of these ring-shaped aggregates transition back into punctate dots when cells are retransformed with full-length Swi1, where the truncation mutant and full-length protein colocalize (Fig. 4C). It is possible that full-length Swi1 provides stability to the [SWI+] formed by Swi11–30 or Swi1–31 through interaction of the full-length protein and Swi1 truncation mutants. In the absence of full-length Swi1, the morphological changes in the aggregates are perhaps due to interaction with the insoluble protein deposit (IPOD), an ancient quality control compartment, or with the actin cytoskeleton, as shown for Sup35 ring structures (51, 54). More research is required, however, to determine why these ring-shaped aggregates appear.
Taken together, our results show that a small NH2-terminal region of ∼30 amino acids contains all necessary information for in vivo prion formation, maintenance, and transmission and for de novo prion formation. Our research thus sheds light on the determinants that contribute to protein misfolding, aggregation, amyloid formation, and prionization and their significance in biology. Its small size and transferability may also allow us to tag a protein of interest to subject it to prion-mediated functional modulation.
MATERIALS AND METHODS
Yeast strains and media.All yeast strains used in this study were grown and maintained according to methods outlined previously (31) and in the Cold Spring Harbor manual (56). Strains were propagated in rich (yeast extract, peptone, dextrose [YPD]) or synthetic complete (SC) medium. Medium was supplemented with 5 mM guanidine hydrochloride (GdnHCl) or 5-fluoroorotic acid (5-FOA) when indicated. Glucose was used as the carbon source unless otherwise indicated. For the raffinose phenotype assay, glucose was replaced with raffinose and supplemented with 0.5 μg/ml antimycin (Sigma-Aldrich, St. Louis, MO). Plates were incubated at 30°C for 3 days unless otherwise indicated. Agar plates were made as outlined previously (30). Lysogeny broth (LB) supplemented with 100 μg/ml ampicillin was used to select for plasmids, with corresponding selection markers in Escherichia coli.
BY4741 [SWI+] [pin−] and BY4741 swi1Δ/p416TEFSwi1 [SWI+] were described previously (31). The W303 sup35Δ/SUP35::TRP1/p316Sup35FL [PSI+] MATα strain was obtained from the Weissman laboratory (University of California at San Francisco).
Plasmids and oligonucleotides. Table 1 lists the primers used for this study, and Table 2 shows the plasmids used in this study. The pRS316 vector plasmid and pKT218,620 expressing dominant negative Hsp104 (Hsp104DN) were gifts from the Chernoff laboratory (Georgia Institute of Technology). p2HG vector control and p2HG-Hsp104 were previously described (14). Plasmids p416TEFSwi11–38YFP, p415TEFSwi11–38YFP, p416TEFYFP, p415TEFYFP, p416TEFSwiNYFP, p415TEFSwiNYFP, and p416TEFSwi1mCherry were previously described (31). To construct plasmid p416SWI1-NYFP, the SWI1 promoter was cut from p416SWI1-NQYFP (30) through SacI/SpeI and used to replace the TEF1 promoter of p416TEFSwiNYFP (30). Our previous study investigated the aggregation of Swi11–37YFP and Swi11–31YFP. Subsequent sequencing determined that Swi11–37YFP is actually Swi11–38YFP and that Swi11–31YFP is Swi11–32YFP. Plasmid p416TEFNQYFP was previously described (30). All YFP-tagged Swi1 truncation mutants in the p416TEF vector were constructed by amplifying the Swi1 fragment from genomic DNA using a common forward primer (p415Swi1ForwardNEW) and unique reverse primers (Age1Swi1–37Rev, Age1Swi1–36Rev, Age1Swi1–35Rev, Age1Swi1–34Rev, Age1Swi1–33Rev, Age1Swi1–31Rev, Age1Swi1–30Rev, Age1Swi1–29Rev, Age1Swi1–28Rev, Age1Swi1–27Rev, Age1Swi1–26Rev) that contained an AgeI site. Swi11–37 to Swi11–33 were amplified from the p415TEFSwi11–38YFP template. Swi11–31, Swi11–27, and Swi11–26 were amplified from the p415TEFSwi11–33YFP template. Swi11–30YFP, Swi11–29YFP, and Swi11–28YFP were amplified from the p415TEFSwi11–31YFP template. Amplified Swi11–38 to Swi11–33 fragments were digested with SpeI and AgeI, purified, and ligated via SpeI and AgeI sites into p416TEFSwi11–38YFP, replacing Swi11–38 with the smaller truncated version. Amplified Swi11–31 to Swi11–26 fragments were digested with SpeI and AgeI, purified, and ligated via SpeI and AgeI sites into p416TEFSwi11–37YFP, replacing Swi11–37 with the smaller truncated version.
Primers used in this study
Plasmids used in this study
To make YFP-tagged Swi1 truncation mutants in the p415TEF vector, the ∼850-bp fragments containing YFP-tagged Swi1 truncation mutants were digested from p416TEFSwi1TruncYFP described above using SpeI and XhoI and ligated into p415TEFSwiNYFP via SpeI and XhoI sites, replacing YFP-tagged SwiN with the YFP-tagged truncation mutant. Plasmids p415TEFSwi11–38MC, p415TEFSwi11–32MC, and p415TEFSwi11–31MC were constructed using a multistep PCR. First, Swi11–38, Swi11–32, and Swi11–31 were amplified using a forward primer (p415TEFSwiFor 4X HT) that binds upstream of the Swi1 ORF in p415TEFSwi11–38YFP, p415TEFSwi11–32YFP, and p415TEFSwi11–31YFP and a reverse primer containing a sequence encoding a GGPGGG linker (Swi1–38+Linker Rev, Swi1–32+Linker Rev, or Swi1–31+Linker Rev). Next, Sup35MC was amplified from p306Sup35SwiNMC using a forward primer containing a sequence encoding a GGPGGG linker (Linker+Sup35MC Forward) and a reverse primer containing an XhoI site (XhoI+Sup35MC). Products of the Swi1Trunc+Linker and Linker+Sup35MC PCRs were then used as the templates for a third PCR using a forward primer (p415TEFSwiFor 4X HT) and a reverse primer (XhoI+Sup35MC). The amplified products were digested using SpeI and XhoI, purified, and ligated into p415TEFSwi11–38YFP through SpeI and XhoI sites, replacing Swi11–38YFP with Swi11–38MC, Swi11–32MC, or Swi11–31MC. The correct insertion and sequence for all constructs were verified by sequencing.
Yeast transformations.Yeast cells were transformed according to a protocol adapted from reference 31. Alternatively, colony transformations were also adapted from a protocol provided with the S.c. Easy Comp transformation kit (Thermo Fisher Scientific, Waltham, MA). Briefly, colonies were resuspended in 100 μl of solution I. Cells were then collected by centrifugation at 2,500 rpm for 3 min. The cell pellet was then resuspended in 10 μl of solution II. The transformation system was completed by the addition of 2 μl of plasmid DNA and 140 μl of solution III to the 10-μl cell-solution II mixture. The transformation system was then briefly vortexed and incubated at 42°C for 30 min. The transformation system was then placed on ice for a minimum of 3 min; this was followed by spreading 50 μl onto selective medium. Additionally, a high-efficiency transformation was used as described in reference 55 with the following changes: the transformation mixture was set up by resuspending the cell pellet in 240 μl of 50% polyethylene glycol (PEG), followed by the addition of 36 μl 1 M lithium acetate (LiAc), 20 μl single-stranded salmon sperm DNA (10 mg/ml), 4 μl plasmid, and 60 μl water.
Microscopy.Microscope images were taken with a Zeiss Axiovert 200 epifluorescence microscope. Samples were viewed with a 100× objective and filters specific for differential interference contrast (DIC), YFP, mCherry, cyan fluorescent protein (CFP), or 4′,6-diamidino-2-phenylindole (DAPI). Images were captured using Axiovision AC (Zeiss, Oberkochen, Germany).
Immunoblotting.Samples for immunoblots were prepared using a protocol adapted from reference 31. Briefly, cells were grown overnight culture in selective medium. Swi1TruncYFP samples were normalized to 7.5 × 107 cells/ml, and Swi1TruncMC samples (for Fig. 5D) and NYFP samples (for Fig. 1E) were normalized to 2.5 × 107 cells/ml in water. Cells were treated with 0.1 M NaOH and incubated for 5 min at room temperature. Cell lysates were resuspended in 100 to 200 μl 2× Laemmli buffer (2% SDS for Swi1TruncMC and NYFP and 4% SDS for Swi1TruncYFP), boiled for 5 to 10 min, and sonicated in 10 1-s pulses. The lysates were spun down at 13,000 rpm to pellet debris. Five-microliter volumes of lysates were loaded for all samples. Swi1TruncMC cell lysates were loaded onto a 7.5% Tris-glycine gel, Swi1TruncYFP cell lysates were loaded onto a 4-to-20% gradient Tris-glycine gel (Bio-Rad, Hercules, CA), and NYFP samples were loaded onto a 12% Tris-glycine gel. The gels were transferred to a polyvinylidene difluoride (PVDF) membrane and blotted with either 1:2,500 anti-GFP antibody (JL-8 antibody; Clontech, Mountain View, CA), 1:2,500 anti-Sup35 antibody (gift from the Liebman laboratory, University of Nevada, Reno, NV), or 1:5,000 antiactin antibody (Chemicon, Temecula, CA) and 1:2,500 horseradish peroxidase-conjugated rat anti-mouse secondary antibody. The resulting chemiluminescence was detected using ECL Western blotting reagents (Bio-Rad). Blots were imaged using a Bio-Rad Chemidoc imaging system.
Plasmid shuffle.A swi1Δ/p416TEFSwi1 [SWI+] strain was transformed with YFP-tagged truncation mutants using p415TEFSwi1TruncYFP. Three individual colonies were grown in selective medium overnight at 30°C. The three swi1Δ/p416TEFSwi1/p415TEFSwi1TruncYFP [SWI+] colonies were spread onto SC medium lacking leucine (SC−Leu) supplemented with 5-FOA and incubated for 3 days at 30°C. Colonies from the 5-FOA plate were then examined for p416TEFSwi1 loss. Subsequently, two individual swi1Δ/p415TEFSwi1TruncYFP colonies were transformed with p416TEFSwi1mCherry. At least three individual colonies were imaged at each step of the plasmid shuffle. For smaller fragments, Swi11–32 through Swi11–26, the experiment was repeated with an additional selection step. Aggregation was assessed on SC−Leu plus 5-FOA plates, and cells were subsequently streaked on SC−Leu plates.
Raffinose assay.Cells from colonies were resuspended in 1 ml of water and counted. All samples were equalized to 1 × 106 cells/ml. Two hundred microliters of cell mixture was pipetted into the first well of a 96-well plate. One hundred forty microliters of water was pipetted into rows 2 to 6 of the plate. Forty microliters of cells from the first well was pipetted into the second well and serially diluted. Cells were then spotted onto YPD and raffinose-plus-0.5 μg/ml antimycin plates and incubated for 3 to 5 days at 30°C.
RT-PCR.Total RNA was isolated using an RNeasy minikit (Qiagen, Hilden, Germany). The Superscript III First Strand DNA synthesis kit (Invitrogen, Cambridge, MA) was used to synthesize cDNA according to the manufacturer's protocol. One microliter of cDNA was used as the template for amplification with primers specific for the C-terminal region of Swi1, Swi1 RT For, and Swi1 RT Rev. The PCR products were run on a 1.8% agarose gel with a 100-bp DNA ladder (New England Biolabs, Ipswich, MA).
Centrifugation assay.Isogenic BY4741 [SWI+] and [swi−] cells of a swi1Δ/p415TEFSwi11–38YFP or swi1Δ/p415TEFSwi11–32YFP background and isogenic [SPS+]1–38, [SPS+]1–32, or [SPS+]1–31 and [sps−]1–38, [sps−]1–32, or [sps−]1–31sup35Δ/p415TEFSwi1TruncMC/p416TEFSwi1TruncYFP cells were incubated in selective medium overnight at 30°C. Overnight cultures were diluted and grown to log phase at 30°C. Cells were resuspended in 0.1 M Tris buffer (pH 8) plus 100 mM EDTA and 0.5% β-mercaptoethanol and incubated at 30°C for 45 min. Cells were subsequently washed with ST buffer (10 mM Tris [pH 7.5], 1 M sorbitol) and then resuspended in ST buffer supplemented with 10 μl lyticase (Sigma-Aldrich; 10 U/μl) and incubated 30°C for 30 min. After centrifugation at 2,000 rpm for 5 min, cell pellets were resuspended in 50 mM Tris (pH 7.5) plus 50 mM NaCl supplemented with protease inhibitors (Roche Complete Mini tablet [Roche, Mannheim, Germany] plus leupeptin [20 μg/ml], pepstatin [20 μg/ml], and phenylmethylsulfonyl fluoride [PMSF; 10 mM]). Cells were sonicated for five 1-s pulses, and lysates were cleared of debris with centrifugation at 100 × g for 5 min. The cleared lysate was transferred to a clean tube, one-third of cleared lysate was saved as total (T) samples, and the remaining samples were spun at 20,000 × g for 20 min at 4°C. After centrifugation, the supernatant fraction (S) was transferred to a new tube and the pellet fraction (P) was resuspended into a volume equal to that of the pellet fraction of 50 mM Tris (pH 7.5) plus 50 mM NaCl supplemented with protease inhibitors. Samples were mixed with 2× Laemmli sample buffer (4% SDS) and boiled for 10 min. Samples resolved by SDS-PAGE on a 4-to-20% Tris-glycine gradient gel for YFP-tagged Swi1 samples and a 10% Tris-glycine gel for Swi11–38MC, Swi11–32MC, and Swi11–31MC. The gels were transferred onto a PVDF membrane and blotted with 1:2,500 anti-GFP antibody (JL-8 antibody; Clontech) or 1:2,500 anti-Sup35 antibody (for [SPS+] and [sps−] cells) and 1:2,500 horseradish peroxidase conjugated rat anti-mouse secondary antibody. The resulting chemiluminescence was detected using ECL Western blotting reagents (Bio-Rad). Blots were imaged using a Bio-Rad Chemidoc imaging system.
Prion curability with Hsp104DN.[SWI+] cells of swi1Δ/p415TEFSwiNYFP, swi1Δ/p415TEFSwi11–38YFP, swi1Δ/p415TEFSwi11–32YFP backgrounds were transformed independently with pRS316 or pKT218,620 using a high-efficiency yeast transformation protocol. Plates were incubated for 3 days at 30°C. Three colonies from each plate were inoculated into selective medium and grown at 30°C for 2 days. The cells were then imaged, and aggregation was quantified.
Prion curability by Hsp104 overexpression.[SPS+]1–38, [SPS+]1–32, or [SPS+]1–31 cells were transformed with either p2HG or p2HG-Hsp104 and plated on selective medium. Individual transformants were grown on selective medium for 24 h and subsequently spread on YPD. Plates were incubated at 30°C for 3 days, followed by 3 days of incubation at 4°C to allow for color change. The number of completely cured cells (red) or noncured colonies was quantified.
Transferability of PrD.W303 sup35Δ/p316Sup35FL [PSI+] strain was transformed with p415TEFSwi11–38MC, p415TEFSwi11–32MC, or p415TEFSwi11–31MC. Cells were plated on selective medium. Transformants were then replica plated onto YPD to assess the color change. Individual red colonies, whose pigmentation is caused by the efficient termination of a premature nonsense stop codon in the ade1-14 mutant allele (33), from the YPD plate were spread on −Leu plates supplemented with 5-FOA to select for colonies that had lost p316Sup35FL. Colonies were then grown on complete medium (YPD) and SC plates lacking adenine (−Ade). sup35Δ/p415TEFSwi11–38MC, sup35Δ/p415TEFSwi11–32MC, or sup35Δ/p415TEFSwi11–31MC [psi−] [sps−] cells with the Ade− growth phenotype and red color on YPD were then transformed with p416TEFSwi11–38YFP, p416TEFSwi11–32YFP, or p416TEFSwi11–31YFP, respectively, and plated on selective medium. Either transformants were replica plated into −Ade medium or colonies were spread onto −Ade medium to select for Ade+ cells. Ade+ cells were then grown on YPD, YPD plus 5 mM GdnHCl, and SC−Ade plates. Colonies that were white on YPD, red on YPD after GdnHCl treatment, and showed a curable Ade+ growth phenotype were considered [SPS+] candidates. The aggregation status of Swi1TruncYFP was assessed by fluorescence microscopy to further confirm the presence or absence of [SPS+]. Incubation of plates was done at 30°C for 3 days for all steps. For color development, plates were incubated at 30°C for 3 days, followed by incubation at 4°C for 3 days. In order to determine the rate of reappearance of Ade+ colonies of GdnHCl-cured cells, [sps−] cells were grown in selective medium for 24 h and an equal number of cells was subsequently plated on SC−Ura−Leu medium as well as SC−Ade and grown for 3 to 5 days. The number of Ade+ colonies and the total number of cells plated were quantified.
ACKNOWLEDGMENTS
We thank Y. O. Chernoff (Georgia Institute of Technology). J. S. Weissman (University of California, San Francisco), and Susan Liebman (University of Nevada, Reno) for plasmids, yeast strains, and antibodies, as well as D. K. Goncharoff for manuscript editing and critical comments.
This work was supported by grants from the National Institutes of Health (grant 1 R01 GM110045) to L.L., the National Science Foundation (grant 1122135) to L.L., and the National Institute of Aging (grant T32 AG220506) to S.V.
The funders had no role in the study design, data collection and interpretation, or decision to submit the work for publication.
FOOTNOTES
- Received 24 April 2017.
- Returned for modification 15 May 2017.
- Accepted 10 July 2017.
- Accepted manuscript posted online 17 July 2017.
- Copyright © 2017 American Society for Microbiology.