Guanidine hydrochloride inhibits the generation of prion "seeds" but not prion protein aggregation in yeast

Abstract

[PSI(+)] strains of the yeast Saccharomyces cerevisiae replicate and transmit the prion form of the Sup35p protein but can be permanently cured of this property when grown in millimolar concentrations of guanidine hydrochloride (GdnHCl). GdnHCl treatment leads to the inhibition of the replication of the [PSI(+)] seeds necessary for continued [PSI(+)] propagation. Here we demonstrate that the rate of incorporation of newly synthesized Sup35p into the high-molecular-weight aggregates, diagnostic of [PSI(+)] strains, is proportional to the number of seeds in the cell, with seed number declining (and the levels of soluble Sup35p increasing) in the presence of GdnHCl. GdnHCl does not cause breakdown of preexisting Sup35p aggregates in [PSI(+)] cells. Transfer of GdnHCl-treated cells to GdnHCl-free medium reverses GdnHCl inhibition of [PSI(+)] seed replication and allows new prion seeds to be generated exponentially in the absence of ongoing protein synthesis. Following such release the [PSI(+)] seed numbers double every 20 to 22 min. Recent evidence (P. C. Ferreira, F. Ness, S. R. Edwards, B. S. Cox, and M. F. Tuite, Mol. Microbiol. 40:1357-1369, 2001; G. Jung and D. C. Masison, Curr. Microbiol. 43:7-10, 2001), together with data presented here, suggests that curing yeast prions by GdnHCl is a consequence of GdnHCl inhibition of the activity of molecular chaperone Hsp104, which in turn is essential for [PSI(+)] propagation. The kinetics of elimination of [PSI(+)] by coexpression of a dominant, ATPase-negative allele of HSP104 were similar to those observed for GdnHCl-induced elimination. Based on these and other data, we propose a two-cycle model for "prionization" of Sup35p in [PSI(+)] cells: cycle A is the GdnHCl-sensitive (Hsp104-dependent) replication of the prion seeds, while cycle B is a GdnHCl-insensitive (Hsp104-independent) process that converts these seeds to pelletable aggregates.

FIG. 1.

Recovery of heat-inactivated luciferase in the presence of 3 mM GdnHCl and absence of Hsp104. (A) Strain 74-D694 [PSI⁺] grown at 30°C for 1 h in the presence (□) or absence (○) of GdnHCl (3 mM) or grown in the presence of 3 mM GdnHCl for five generations (▵) before assaying for luciferase reactivation as described in Materials and Methods. Luciferase activity was determined before heat treatment and at several subsequent time points and expressed as percentages of the activity before the heat treatment. (B) Cells were grown at 30^oC and then shifted to 37^oC for 1 h before being placed back at 30^oC, when 3 mM GdnHCl was added. Luciferase reactivation was carried out as for panel A. The strains used were 74-D694 [PSI⁺], either untreated (○) or grown in the presence of 3 mM GdnHCl for 1 h (□) and [psi⁻] strain 74-D694 hsp104::LEU2 (▴).

FIG. 2.

Subcellular distribution of Sup35p in [PSI⁺] and [psi⁻] cells grown in 3 mM GdnHCl. [PSI⁺] and [psi⁻] derivatives of strain 74-D694 were grown in YNBD to stationary phase and then transferred to YPD medium. After six generations of growth (6g) in the presence of 3 mM GdnHCl a subcellular fractionation analysis of Sup35p was performed as described in Materials and Methods by using an anti-Sup35p polyclonal antibody. Samples from the 0g time point were similarly studied. [PSI⁺]-to-[psi⁻] conversion after GdnHCl treatment was analyzed by plating cells on 1/4YPD medium and visually determining the percentage of white ([PSI⁺]) colonies at each time point. Approximately 500 colonies were counted for each time point. The translation termination readthrough assays were performed as described in Materials and Methods. The levels of β-galactosidase in the pUKC819 transformants were expressed as percentages of the level seen in the control pUKC815 transformants. The percentages of readthrough shown (with standard deviations) were calculated from the averages of two repetitions for each of two independent transformants. T, total crude extract; S, soluble fraction; P, pellet fraction.

FIG. 3.

The addition of a hexahistidine tag on the C terminus of Sup35p does not alter its prion-like behavior in [PSI⁺] cells. (A) Subcellular distribution of Sup35p and Sup35Hp in the [PSI⁺] and [psi⁻] derivatives of the 74-D694 strain transformed with plasmid pUKC1809. After growth to log phase on YPD medium (A₆₀₀ = 0.5) the cultures were transferred to YPGal medium to induce Sup35Hp synthesis. Cell samples were removed every 30 min over a 6-h incubation period and analyzed as described in the legend to Fig. 1 by using either an anti-Sup35p (data not shown) or an antipentahistidine polyclonal antibody (right) to detect Sup35p or Sup35Hp, respectively. For Sup35Hp identical results were obtained at the different time points and the results shown are those obtained after 2 h of growth in YPGal. The analysis of wild-type Sup35p (left) was undertaken before transferring the cells to YPGal medium. (B) The [PSI⁺]-associated suppression phenotype is not altered by the presence of the C-terminal hexahistidine tag. Plasmids pUKC1512SUP (encoding wild-type Sup35p) and pUKC1602 (encoding Sup35Hp) were separately introduced into [PSI⁺] strain MT700/9d sup35::kanMX in place of plasmid pYK810 (encoding the wild-type Sup35p). After growth on 5-fluoroorotic acid, the [PSI⁺]-related suppression phenotypes of both transformants were analyzed by determining colony color on 1/4YPD (top) and after growth in the presence of 3 mM GdnHCl (bottom). Four independent clones were used for each plasmid: pUKC1512SUP, clones 1 to 4; pUKC1602, clones 5 to 8. T, S, and P are as defined for Fig. 2.

FIG. 4.

GdnHCl does not dissociate preexisting Sup35p aggregates in a [PSI⁺] strain. [PSI⁺] strain 74-D694 was transformed with plasmid pUKC1809 (carrying SUP35H under the control of the GAL1 promoter) and transferred to YPGal for 4 to 5 h to induce synthesis of Sup35Hp. Synthesis was then stopped by transferring the cells to YPD medium for 2 h, after which time the cell cultures were split, one-half being diluted into fresh YPD medium and the other being diluted into identical medium containing GdnHCl (3 mM final concentration). The cells were then incubated further to allow them to go through four generations of growth with cell samples being taken at zero (0g), two (2g), and four generations (4g) for subcellular fractionation analysis as described in the legend to Fig. 1. The preexisting Sup35Hp was detected with antipentahistidine antibodies. To assess the relative levels of Sup35Hp in the pellet (P) fraction after two and four generations, the control (0g) P sample was also diluted either 1:4 or 1:16 prior to SDS-PAGE and Sup35Hp was detected with the same antipentahistidine antibody (right). An additional band detected by the antipentahistidine antibody (∗) was variably detected in samples analyzed but was not reproducibly present in any one particular sample. The anti-Sup35p antibody did not detect this protein (data not shown). T and S are as defined for Fig. 2.

FIG. 5.

The influence of growth in the presence of GdnHCl on the physical state of newly synthesized Sup35p in [PSI⁺] cells. (A) [PSI⁺] strain 74-D694 carrying plasmid pUKC1809 (expressing GAL1-SUP35H) was grown overnight in YNBD to ensure plasmid retention and to repress Sup35Hp synthesis. Cells were then transferred to YPGal medium either with or without GdnHCl (3 mM) and grown for either 1 or 4 h. The subcellular distribution of Sup35Hp at these two time points was then determined as described in the legend to Fig. 2 with an antipentahistidine antibody. T, S, and P are as defined for Fig. 2. (B) The 74-D694 [PSI⁺] strain carrying plasmid pUKC1809 was grown overnight in YNBD medium. Cells were then transferred to YPD medium, either with or without GdnHCl (3 mM), and incubated for a further 3 h. Finally, both cultures were transferred to YPGal medium containing 3 mM GdnHCl. After a further incubation for 3 h the subcellular location of the Sup35Hp was assessed as described for panel A with an antipentahistidine antibody. The proportion of Sup35Hp in the pellet fraction (%P) was determined by densitometry.

FIG. 6.

Reducing the number of [PSI⁺] seeds by GdnHCl reduces the ability of newly synthesized Sup35p to enter high-molecular-weight aggregates. (A) [PSI⁺] strain 74-D694, transformed with plasmid pUKC1809, was grown in YPD medium containing 3 mM GdnHCl for up to six generations (0g to 6g), and then, at various points, the cells were transferred to YPGal medium containing 3 mM GdnHCl for a further 2 h to allow for induction of synthesis of the Sup35Hp. For each of the samples taken, the subcellular distribution of the Sup35Hp was assessed as described above. T, S, and P are as defined for Fig. 2. (B) The cell samples were also analyzed for the degree of aggregation of the endogenous Sup35p prior to the transfer to YPGal by using a polyclonal anti-Sup35p antibody. The proportions of Sup35p and Sup35Hp proteins in the pellet fraction (%P) were determined by densitometry. Cells were also plated on 1/4YPD before the 2-h YPGal incubation to assess the percentage of [PSI⁺] cells in the population being analyzed. The subsequent incubation for 2 h in YPGal did not significantly alter the relative numbers of [PSI⁺] cells (data not shown).

FIG. 7.

The kinetics of Sup35p aggregation in vivo. Quantification of the proportion of Sup35p and Sup35Hp was performed by densitometric scanning of the Western blots shown in Fig. 6. For each data point the sum of the area corresponding to the pellet and the soluble fractions equaled the total fraction with no greater than 10% variation. The seed numbers were deduced from the kinetics of elimination of [PSI⁺] from cells over six generations of growth in YPD containing 3 mM GdnHCl (14). The intermediate seed numbers for each generation were calculated as a simple exponential function to reach the total seed number. (A) Plot of the percentage of Sup35p present in the pellet fraction (i.e., the fraction that corresponded to the aggregated Sup35p) against the estimated seed number. (B) Plot of the inverse of each data point from panel A.

FIG. 8.

Influence of GdnHCl on the aggregation of newly synthesized Sup35p in slowly dividing [PSI⁺] cells growing on glucose-limited medium. [PSI⁺] strain 74-D694 transformed with plasmid pUKC1809 was grown for 2 h in YPGal containing 3 mM GdnHCl following four different preincubation regimens (A to D) as indicated. For cultures C and D this included growth in glucose-depleted YPD containing 0.2% glucose rather than the standard 2% glucose. The fate of the newly synthesized Sup35Hp following the 2-h growth on YPGal was then assessed as described in the legend to Fig. 1 by using an antipentahistidine antibody. T, S, and P are as defined for Fig. 2.

FIG. 9.

Rate of prion seed replication in a [PSI⁺] strain following a release from a GdnHCl-induced block. [PSI⁺] strain 74-D694 was grown on YPD containing 3 mM GdnHCl for either 6 (A) or 4.5 (B) generations to generate a population of cells with only a few seeds per cell (14). The cells were then resuspended in fresh YPD (low-glucose [0.2%]) medium without GdnHCl, and cell samples were taken every 10 min for the first 30 min and then at 60 and 120 min. By the 2-h time point the cells had not gone through one generation. The average number of [PSI⁺] seeds per cell in the cell populations at each of the time points was determined by establishing the rate of [PSI⁺] curing by 3 mM GdnHCl essentially as described by Eaglestone et al. (14).

FIG. 10.

[PSI⁺] seed replication does not require ongoing cell division or protein synthesis. [PSI⁺] strain 74-D694 was grown in YPD with 3 mM GdnHCl for six generations to generate a cell population containing a low average seed number per cell. The cells were then resuspended in fresh low-glucose (0.2%) YPD (YPD0.2) without GdnHCl to allow seeds to begin to replicate (see Fig. 8). In parallel, cells from the same culture were also transferred to fresh YPD0.2 with either 10 μg of cycloheximide (Cyh)/ml in 0.1% ethanol (EtOH) or 0.1% ethanol alone. Cells were removed from these cultures at the beginning (t = 0) and after 60 min of incubation (t = 60), and the average seed number in the cell population was determined essentially as described by Eaglestone et al. (14). The results of two independent experiments (A and B) using cultures with slightly different starting seed numbers are shown.

FIG. 11.

The kinetics of [PSI⁺] elimination by Hsp104 depletion or inactivation. The data of Wegrzyn et al. (46) for the elimination of [PSI⁺] in haploid ascospores containing an hsp104 gene deletion are plotted (○) together with a continuous line indicating the predicted kinetics of [PSI⁺] elimination from a cell containing 16 [PSI⁺] seeds calculated by using the model of Eaglestone et al. (14). Also shown for comparison are [PSI⁺] elimination data obtained by inactivation of endogenous Hsp104 by coexpression of an ATPase-negative mutant HSP104 K218T,K620T (▪).

FIG. 12.

A two-cycle model for prionization of Sup35p in [PSI⁺] cells. Cycle A replicates the [PSI⁺] seed (squares), which can either be a conformationally altered monomer or a multimeric form of Sup35p and which is dependent on Hsp104. Cycle B converts these seeds to aggregates, which themselves do not have seeding activity. Cycle B is independent of Hsp104 as is the de novo conversion of the soluble nonprion form of Sup35p (circles) to the seed form in the absence of a preexisting seed.