Import into and Degradation of Cytosolic Proteins by Isolated Yeast Vacuoles

Abstract

In eukaryotic cells, both lysosomal and nonlysosomal pathways are involved in degradation of cytosolic proteins. The physiological condition of the cell often determines the degradation pathway of a specific protein. In this article, we show that cytosolic proteins can be taken up and degraded by isolated Saccharomyces cerevisiae vacuoles. After starvation of the cells, protein uptake increases. Uptake and degradation are temperature dependent and show biphasic kinetics. Vacuolar protein import is dependent on cytosolic heat shock proteins of the hsp70 family and on protease-sensitive component(s) on the outer surface of vacuoles. Degradation of the imported cytosolic proteins depends on a functional vacuolar ATPase. We show that the cytosolic isoform of yeast glyceraldehyde-3-phosphate dehydrogenase is degraded via this pathway. This import and degradation pathway is reminiscent of the protein transport pathway from the cytosol to lysosomes of mammalian cells.

INTRODUCTION

Regulated protein synthesis and degradation control protein turnover in cells. Lysosomal and nonlysosomal pathways are responsible for protein degradation. In eukaryotes, the proteasomes, including the ones that are part of the ATP- and ubiquitin-dependent proteolytic system, constitute the main nonlysosomal protein degradation mechanism (Coux et al., 1996). Proteins enter mammalian lysosomes or the homologous organelles in yeast and plant vacuoles through one of several pathways. Classically, the most prominent of these pathways were called autophagy and heterophagy. More recently, it has become apparent that intracellular proteins can gain access to the lysosomal matrix by additional routes: endocytosis, crinophagy, direct conversion of ER cisternae into lysosomes, macroautophagy, microautophagy, and selective transport across the lysosomal membrane (Dunn, 1994; Blommaart et al., 1997; Bryant and Stevens, 1998; Knecht et al., 1998; Thumm and Wolf, 1998).

In the yeast Saccharomyces cerevisiae, autophagy is the most important pathway from the cytosol into the vacuole. Autophagy is thought to be responsible for bulk turnover of proteins (Seglen and Bohley, 1992; Takeshige et al., 1992; Mizushima et al., 1998) and is therefore responsible for degradation of a large quantity of proteins in the vacuole. As in mammalian cells, microautophagy in yeast is different from macroautophagy. It is possible to isolate yeast mutants that only affect the first but not the second mechanism (Tuttle and Dunn, 1995), and the requirements for both lysosomal mechanisms appear to be different (Yuan et al., 1997). Extracellular as well as plasma membrane proteins are delivered to the vacuole for degradation by endocytosis (Davis et al., 1993; Raths et al., 1993).

Transport of proteins from the cytosol into the vacuole has been described for only a few proteins. These are the two vacuolar hydrolases aminopeptidase 1 (Klionsky et al., 1992) and α-mannosidase (Yoshihisa and Anraku, 1990) and the key gluconeogenic enzyme fructose-1,6-bisphosphatase (Chiang and Schekman, 1991; Huang and Chiang, 1997; Shieh and Chiang, 1998). The exact mechanism of protein transport is not known for any of these proteins (for review, see Scott and Klionsky, 1997; Klionsky, 1998). In the case of aminopeptidase 1 and α-mannosidase, both protein translocation and vesicle-mediated autophagocytosis have been proposed (Seguí-Real et al., 1995; Scott and Klionsky, 1997; Klionsky, 1998). Cytosolic fructose-1,6-bisphosphatase can be selectively transported into the vacuole and degraded under conditions where its enzymatic activity is no longer needed (Chiang and Schekman, 1991; Chiang et al., 1996; Huang and Chiang, 1997); however, the protein can also be degraded independently of the major vacuolar protease proteinase A, suggesting that cytosolic fructose-1,6-bisphosphatase can also be degraded in the cytosol by an alternative mechanism (Schork et al., 1994a,b).

Selective, direct transport across the lysosomal membrane was first described in serum-deprived confluent fibroblasts. Imported proteins contained peptide sequences related to the KFERQ sequence (Dice 1990; Terlecky, 1994). This transport pathway appears to exist in rat liver under basal conditions but becomes progressively more important under starvation (Wing et al., 1991; Cuervo et al., 1995). The proteins to be degraded are recognized by a cytosolic heat shock cognate protein (hsc73), which binds the proteins and assists in recognition of the proteins by the lysosomes (Chiang et al., 1989; Cuervo and Dice, 1996; Hayes and Dice, 1996). In this respect, the role of hsc73 in targeting proteins to lysosomes resembles that of hsp70 in protein targeting to mitochondria. The lysosomal membrane protein Lamp-2 appears to be involved in the recognition of cytosolic proteins intended for degradation (Cuervo and Dice, 1996). Intralysosomal hsc73 is required for an efficient uptake of the cytosolic proteins into the lysosomal matrix (Agarraberes et al., 1997; Cuervo et al., 1997). These data suggest the presence of a lysosomal protein translocation machinery similar to that found in other cell organelles. In an alternative model, proteins are selectively taken up by lysosomes through membrane invaginations similar to those found in endocytic processes. More work is required to distinguish between the two models.

In this article, we describe the presence of a protein transport system for the uptake of cytosolic proteins in isolated yeast vacuoles. This system shares many features with direct protein import into mammalian lysosomes. Protein import is induced under starvation and depends on cytosolic chaperones of the hsp70 family as well as protease sensitive component(s) on the outer surface of vacuoles. The cytosolic isoform of yeast glyceraldehyde-3-phosphate dehydrogenase is imported and degraded via this pathway.

MATERIALS AND METHODS

Yeast Strains and Media

The following Saccharomyces cerevisiae strains were used: MW109 (Mata his3–11,15 leu2–3112 lys3 Δtrp1 ura3–52) (Werner-Washburne et al., 1987), MW123 (Matα his3–11,15 leu2–3112 lys3, Δtrp1 ura3–52 ssa1::HIS3 ssa2::LEU2) (Werner-Washburne et al., 1987), KHY31 (Matα leu2–3112 his4–5169, ade6 ura3–52 vph1::LEU2) (Graham et al., 1998); cl3-ABYSS-86 (Mata, ura3-Δ5, leu2–3, pra1–1, prb1–1, prc1–1, cps1–3, can^R) (Achstetter et al., 1984).

Yeast cells were grown in YPD (1% yeast extract, 2% peptone, 2% glucose), SD-N (0.17% yeast nitrogen base without amino acids, 2% glucose) (Takeshige et al., 1992), and SD (0.67% yeast nitrogen base without amino acids, 2% glucose). The powdered media were purchased from Life Technologies (Gaithersburg, MD).

Isolation of Vacuoles

Cells were grown in YPD at 30°C to an OD₆₀₀ of 0.8–1.0. For starvation, cells were grown in SD-N at 30°C to an OD₆₀₀ of 0.8–1.0. Cells were harvested by a 5 min spin at 4500 × g, resuspended in 0.1 M Tris/SO₄, pH 9.4, 10 mM DTT, incubated for 20 min at 30°C, and pelleted for 5 min at 4500 × g. The cell pellet was resuspended in spheroplasting buffer (1.2 M sorbitol, 50 mM Tris-Cl, pH 7.4, 0.5 mg Zymolyase 20T/50 OD₆₀₀ units [Seikagaku, Tokyo, Japan]) and incubated for 30 min at 30°C. Spheroplasts were pelleted for 3 min at 4°C at 500 × g. The pellet was carefully resuspended in 3 ml 15% Ficoll, 200 mM Sorbitol, 10 mM K-PIPES, pH 6.8. The spheroplasts were lysed with DEAE-dextran at a final concentration of 90 μg/100 OD₆₀₀ units of cells and incubated for 5 min on ice followed by a 2 min heat shock at 30°C. The spheroplasts were transferred to a Beckman SW-40 centrifuge tube (Beckman Instruments, Fullerton, CA). The suspension was overlaid with 2 ml 8% Ficoll, 2 ml 4% Ficoll, and 2 ml 2% Ficoll, and the tube was filled with 200 mM sorbitol, 10 mM K-PIPES, pH 6.8. The gradient was centrifuged for 2 h at 25,000 rpm at 4°C in the Beckman SW 40 rotor. Vacuoles accumulated at the 2–4% Ficoll interphase were collected, resuspended in 1 ml 200 mM sorbitol, 10 mM K-PIPES, pH 6.8, and layered on top of 1 ml 4% Ficoll solution in a Beckman TLS-55 centrifuge tube. The vacuoles were collected on the 0–4% interphase by centrifugation for 30 min at 110,000 × g at 4°C in a tabletop ultracentrifuge. The protein content was estimated using the Bio-Rad protein assay reagent (Bio-Rad, Hercules, CA). A 500-ml culture yields ∼350 μg vacuolar protein.

Preparation of Cytosol

Cells were grown in YPD medium to an OD₆₀₀ of 1.0–1.2, pelleted for 5 min at 4.500 × g, and then resuspended in lysis buffer (25 mM K-PIPES, pH 6.8, 200 mM sorbitol, 50 mM KCl, 2 mM DTT, 10 mM MgCl₂, 0.5 mM PMSF, 20 μM leupeptin, and 20 μM pepstatin) at a concentration of 500 OD₆₀₀ units/50 ml and incubated for 5 min at 4°C. The cells were pelleted for 5 min at 4500 × g, and the pellet was resuspended in lysis buffer at a concentration of 500 OD₆₀₀ units/ml. The cells were broken by vortexing 15 times 30 s at the maximum speed with 1 g acid-washed glass beads. The material was pelleted for 5 min at 4°C at 2000 × g. The supernatant was centrifuged for 1 h at 160,000 × g in a tabletop ultracentrifuge. The supernatant contained the cytosol. The protein concentration was estimated using the Bio-Rad protein assay reagent. The cytosol was aliquoted, frozen in liquid nitrogen, and stored at −80°C. The cytosol was used within 2 wk.

Preparation of Radiolabeled Cytosol

Cells (0.2 ml) from a stationary preculture grown in YPD were used to inoculate 20 ml SD containing [³H]leucine (148 Ci/mmol, 2 μCi/ml) (Amersham, Arlington Heights, IL). Cells were labeled for 12 h at 30°C and pelleted for 5 min at 4500 × g, washed three times with ice-cold water, and resuspended in lysis buffer. The cytosol was prepared as described above. The protein concentration was measured with the Bio-Rad protein assay reagent. Finally, the radiolabeled cytosol was diluted with cold cytosol to a specific activity of 600 dpm/μg protein.

Standard Uptake Assay

Vacuoles (50 μg protein, as determined by the Bio-Rad protein assay reagent) were incubated for 30 min at 30°C with a cytosolic extract prepared from [³H]leucine-labeled cells (50 μg protein with a specific activity of 600 dpm/μg protein) in a total volume of 50 μl. The reaction was terminated by adding 1 vol of 20% (wt/vol) trichloroacetic acid (TCA). The sample was incubated for 30 min on ice and then pelleted in a tabletop ultracentrifuge for 30 min at 100,000 × g. Under these conditions, nondegraded proteins are precipitated, whereas proteolytic fragments remain soluble. To control for the integrity of the isolated vacuoles during the incubation period, vacuoles (60 μg protein) were incubated for 30 min at 30°C in a total volume of 30 μl and pelleted for 10 min in a tabletop ultracentrifuge. Twenty-five microliters of the supernatant were mixed with a cytosolic extract prepared from [³H]leucine-labeled cells (50 μg protein) and incubated for 30 min at 30°C. The reaction was stopped with TCA as described above. The radioactivity of the acid-soluble material was determined by liquid scintillation counting. Alternatively, the uptake reaction was terminated by pelleting the vacuoles at 4°C and removing the supernatant. The vacuoles were resuspended in buffer containing 100 μg/ml proteinase K. After 20 min on ice, the digest was stopped by adding PMSF to a final concentration of 1 mM. Proteins were precipitated with TCA as described above. The radioactivity of the TCA pellet containing the nondegraded proteins was determined by liquid scintillation counting. Variations and further experimental details of this protocol are given in the figure legends.

General Methods

Published methods were used for SDS-PAGE and immunoblotting (Horst et al., 1995). Electron microscopy was performed according to Aniento et al., 1993. α-mannosidase was determined as described by Yoshihisa and Anraku (1998). The ATP regenerating system used in some experiments consisted of 10 mM MgCl₂, 10 mM ATP, 2 mM phosphocreatine, and 50 μg/ml creatine phosphokinase. The antibody against glyceraldehyde-3-phosphate dehydrogenase is described by Aniento et al. (1993). Antibodies against the Ssa proteins were gifts from Dr. E. A. Craig (University of Wisconsin, Madison, WI). The carboxypeptidase Y and hexokinase antibodies were gifts from Dr. S. Schröder-Köhne (Max-Planck Institute of Biophysical Chemistry, Göttingen, Germany). Unless stated otherwise, all chemicals were from Sigma (St. Louis, MO).

RESULTS

Degradation of Cytosolic Proteins by Isolated Vacuoles

Vacuoles were purified from Saccharomyces cerevisiae according to a protocol described by Haas (1995) with minor modifications (for details see MATERIAL AND METHODS). The purity of the vacuole preparation was determined by two independent approaches: electron microscopy (EM) (Figure 1) and measuring enrichment of the specific activity of α-mannosidase as a vacuolar marker (Table 1). The EM pictures showed the absence of ER and mitochondrial contaminations. The specific activity of α-mannosidase was enriched in the vacuole preparation ∼75-fold compared with the spheroplast lysate, which is agreement with published values (Wiemken, 1975). By these criteria the vacuole preparation can be considered pure.

Figure 1.

Morphology of isolated vacuoles. Vacuoles were isolated from yeast cells and prepared for electron microscopy as described by Aniento et al. (1993). Bars, 1.5 μm; 0.3 μm (inset).

Table 1.

Enrichment of the vacuolar marker α-mannosidase

	Specific activity (units/g)
	Non-starved cells	Starved cells
Spheroplast lysate	1.2	9.7
Isolated vacuoles	89.9	703.1

To investigate whether isolated yeast vacuoles can take up cytosolic proteins for degradation, the following experiment was performed (Figure 2A). Cells were grown either in rich medium (YPD) or in nitrogen starvation medium (SD-N). It is well known that starvation induces vacuolar hydrolases (Hansen et al., 1977). To ensure starvation the induction of the vacuolar marker α-mannosidase was determined. α-mannosidase activity was induced eightfold (Table 1), which is in agreement with values reported in the literature (Hansen et al., 1977). Vacuoles were isolated and incubated for 30 min at 30°C with a cytosolic extract prepared from [³H]leucine-labeled cells. The reaction was terminated by adding TCA. Under these conditions the nondegraded proteins are precipitated, whereas peptides remain soluble. The radioactivity of the acid-soluble material was determined by liquid scintillation counting. Vacuoles isolated from starved cells were approximately five times more active in degrading cytosolic proteins compared with vacuoles isolated from nonstarved cells (Figure 2A, compare lanes 3 and 6), demonstrating that this degradation pathway is induced by starvation; however, it is also active at a basal level under normal growth conditions (Figure 2A, lane 3). Degradation of the radiolabeled cytosolic proteins in the uptake assay could be due to the leakage of vacuolar proteases from damaged vacuoles. The integrity of the vacuoles during the incubation period was assessed as follows. The vacuoles were incubated for 30 min at 30°C, intact vacuoles were pelleted, and the supernatant was incubated for 30 min at 30°C with the radiolabeled cytosol (Figure 2A, lanes 2 and 5). For vacuoles isolated from both starved and nonstarved cells, degradation caused by leakage of proteases from damaged vacuoles does not correspond to more than 15% of the total degradation observed after 30 min (Figure 2A, compare lanes 2 and 3 and lanes 5 and 6). This extravacuolar degradation was ∼2.5-fold higher using vacuoles isolated from starved cells compared with vacuoles isolated from cells grown in rich medium (Figure 2A, compare lanes 2 and 5). The reason for this difference is not known but suggests that vacuoles from starved cells are either more fragile than those from nonstarved cells or that a higher amount of hydrolases in those vacuoles leads to a higher hydrolytic background activity in the assay. Direct measurement of the latency of vacuoles isolated from starved cells by Western blotting for carboxypeptidase Y (CPY) revealed that even after 60 min at 30°C not more than 15% of CPY was found outside the vacuoles (Figure 2C, lane 4). These results are in agreement with the data obtained from the degradation experiments described above (Figure 2A, compare lanes 5 and 6).

Figure 2.

Cytosolic proteins are imported into and degraded by isolated vacuoles. (A) Vacuoles were isolated from yeast cells grown in either YPD (lanes 1–3) or nitrogen starvation medium (lanes 4–6). After 30 min incubation at 30°C with a cytosolic extract prepared from [³H]leucine-labeled cells, uptake reactions were terminated by adding TCA (lanes 3 and 6). To control for the integrity of the isolated vacuoles during the incubation period, vacuoles were incubated for 30 min at 30°C and pelleted, and the supernatant was incubated for an additional 30 min at 30°C with radiolabeled cytosol before the reaction was stopped with TCA (lanes 2 and 5). Background caused by lysed vacuoles and proteins already degraded in the cytosolic extract at the beginning of the incubation period was measured by TCA-precipitating samples at time 0 min (lanes 1 and 4). Radioactivity of the acid-soluble material in all samples was determined by liquid scintillation counting. Data and SD was calculated from three independent experiments performed in duplicate. (B) Vacuoles were isolated from yeast cells grown in nitrogen starvation medium. After a 30-min incubation at 30°C with a radiolabeled cytosolic extract, uptake reactions were terminated by pelleting the vacuoles. Vacuoles were resuspended in import buffer, and cytosolic proteins associated with the surface of the vacuoles were digested by externally added proteinase K for 30 min on ice. Proteinase K treatment was stopped by adding PMSF, and proteins were TCA-precipitated. Radioactivity associated with the pellet was determined by liquid scintillation counting (lane 2). In one sample, vacuolar proteases were inhibited by a 10-min pretreatment of the isolated vacuoles with the protease inhibitors E-64 (100 μM), leupeptin (20 μM), and pepstatin (20 μM) on ice (lane 3). To determine the background caused by the presence of proteinase K-resistant proteins on the vacuolar surface, vacuoles were incubated with the cytosol for 30 min on ice, vacuoles were reisolated by centrifugation, proteinase K-treated, and TCA-precipitated, and the radioactivity of the pellet was determined by liquid scintillation counting (lane 1). Data and SD were calculated from two independent experiments performed in duplicate. (C) Latency of the vacuoles isolated from starved cells was measured at 0 min and after 1 h at 30°C. Samples containing vacuoles plus cytosolic extracts were split into two aliquots. They were incubated with or without 1% Triton X-100 (TX-100; −) for 10 min on ice and pelleted, and the pellets were thoroughly resuspended in 1% TX-100 containing buffer. The proteins in the pellets (P) and supernatants (S) were TCA-precipitated, separated by SDS-PAGE, and immunoblotted with antibodies against carboxypeptidase Y and the 100-kDa subunit of the F_o complex of the vacuolar ATPase as a control.

To investigate whether the degradation of cytosolic proteins occurs inside the vacuoles, the uptake experiment was slightly modified (Figure 2B). After a 30-min incubation of the vacuoles with radiolabeled cytosol, vacuoles were pelleted, and cytosolic proteins associated with the surface of the vacuoles were digested by externally added proteinase K. Proteinase K was inactivated by the addition of 1 mM PMSF, and the proteins were TCA-precipitated. The radioactivity associated with the pellet was determined by liquid scintillation counting. Under these experimental conditions, only the nondegraded proteins are precipitated, whereas the proteolytic fragments stay TCA soluble. In one sample, vacuolar proteases were inhibited by pretreatment of the isolated vacuoles with protease inhibitors (Figure 2B, lane 3). In this sample, the imported and thus proteinase K-protected proteins were not degraded by the vacuolar proteases as they were in the control without pretreatment (Figure 2B, compare lanes 3 and 2). To determine the background attributable to proteinase K-resistant proteins on the vacuolar surface, the vacuoles were incubated with the cytosol for 30 min on ice. Under these conditions no uptake occurs (Figure 2B). The vacuoles were reisolated, proteinase K-treated, and TCA-precipitated, and the radioactivity of the TCA pellet was determined by liquid scintillation counting (Figure 2B, lane 1). In the uptake experiment in which the vacuoles were pretreated with the protease inhibitors, the background does not contribute to >5% of the radioactivity associated with the vacuoles (Figure 2B, compare lanes 1 and 3).

A time course of the proteolysis of cytosolic proteins in isolated vacuoles shows biphasic kinetics (Figure 3A). Cytosol was added to vacuoles isolated from starved cells, and the kinetics of protein uptake was measured at 30°C (Figure 3A, ●). Approximately 70% of the maximum observed proteolysis occurred within the first 10 min. Another 15% occurred during the next 30 min. The relative contribution of extravacuolar proteolysis (Figure 3A, ○) to total proteolysis increases with longer incubation periods. The biphasic kinetics could be explained by either a time-dependent protease inactivation or an increase of the vacuolar pH during the 30-min incubation.

Figure 3.

Uptake and degradation of cytosolic proteins by isolated vacuoles show biphasic kinetics and are temperature dependent. (A) Vacuoles isolated from yeast cells grown in nitrogen starvation medium were incubated with radiolabeled cytosol for the indicated time period at 30°C (●). To control for the integrity of the isolated vacuoles during the incubation, vacuoles were incubated for the indicated time periods at 30°C and pelleted, and the supernatants were incubated for indicated time periods at 30°C with radiolabeled cytosol (○). All reactions were stopped with TCA. Radioactivity of the acid-soluble material was determined by liquid scintillation counting. Data and SD were calculated from two independent experiments performed in duplicate. (B) Vacuoles were isolated from yeast cells grown in nitrogen starvation medium. The experiment was performed as described in A except that uptake was done for 30 min at different temperatures. Background values attributable to lysis of vacuoles during the 30-min incubation were subtracted from the values obtained in the uptake measurements. Data and SD were calculated from three independent experiments performed in triplicate.

The temperature dependency of the uptake/degradation of cytosolic proteins into isolated vacuoles was investigated (Figure 3B). The uptake/degradation efficiency increased linearly in the range of 10–30°C. Maximum efficiency was reached at 30°C. At higher temperatures the efficiency decreased. Some uptake occurred at temperatures below 15°C; therefore, a vesicular transport mechanism seems unlikely because these processes are usually completely blocked at 15°C (Pelham, 1989).

Uptake of Cytosolic Proteins Depends on Protease-sensitive Components on the Outer Surface of Vacuoles

A specific protein uptake mechanism would involve a receptor on the cytosolic surface of the vacuolar membrane. To investigate whether such a putative receptor is involved in protein uptake, vacuoles were treated with increasing concentrations of trypsin and subsequently inactivated by a 5 M excess of soybean trypsin inhibitor. Trypsin-treated vacuoles were tested for uptake of radiolabeled cytosol. Trypsin treatment of vacuoles reduced proteolysis in a dose-dependent manner (Figure 4, lanes 2–5). The integrity of the vacuolar membrane was not affected by trypsin as shown by latency measurements of CPY activity as well as by Western blotting for CPY (our unpublished results). In a control experiment in which trypsin and a fivefold molar excess of soybean trypsin inhibitor were added simultaneously to vacuoles, protein degradation and vacuolar integrity were not affected (Figure 4, lane 2).

Figure 4.

Uptake of cytosolic proteins depends on protease-sensitive component(s) on the surface of vacuoles. Vacuoles were isolated from yeast cells grown in nitrogen starvation medium and treated with increasing concentrations of trypsin for 20 min on ice. Trypsin was inhibited with a fivefold molar excess of soybean trypsin inhibitor for 10 min on ice. Uptake experiments (lanes 3–5) as well as the appropriate controls (lanes 1 and 6) were performed as described in the legend to Figure 2A. To determine whether trypsin and soybean trypsin inhibitor interfere with the integrity of the vacuoles and the uptake process, vacuoles were treated with 100 μg/ml trypsin in the presence of a fivefold molar excess of soybean trypsin inhibitor for 20 min on ice (lane 2, Mock Treatment). The uptake assay was performed as described in Figure 2A. Data and SD were calculated from two independent experiments performed in triplicate.

Influence of Cytosolic hsp70 on the Uptake of Cytosolic Proteins by Vacuoles

In the mammalian lysosomal system, hsc73 plays a role in the uptake of cytosolic proteins. In the yeast cytosol there are four hsp70 isoforms: the constitutively expressed Ssa1 and Ssa2 isoforms and the heat-inducible Ssa3 and Ssa4 isoforms. Ssa3 and Ssa4 are not expressed under normal growth conditions (Werner-Washburne et al., 1988). The uptake efficiency of vacuoles isolated from starved cells increased by 20% when radiolabeled cytosol prepared from yeast that had undergone a 2-h heat shock at 37°C was used (Figure 5A, compare lanes 2 and 4). When this cytosol was used together with 10 mM Mg-ATP and an ATP-regenerating system, the uptake efficiency increased by 80% (Figure 5A, compare lanes 2 and 3). After heat shock the total amount of cytosolic hsp70 proteins is increased ∼2.5-fold as shown by Western blotting using an antisera that recognizes three of the four hsp70 isoforms (the constitutively expressed Ssa1 and Ssa2 and the heat-induced Ssa3; E. A. Craig, personal communication) (Figure 5B). When radiolabeled cytosol prepared from a strain that is deficient in Ssa1 and Ssa2 (MW123; see MATERIAL AND METHODS) or Ssa-immunodepleted cytosol was used, the uptake efficiency decreased by >50% (Figure 5A, compare lane 2 with lanes 5 and 6).

Figure 5.

Uptake of cytosolic proteins depends on cytosolic hsp70s. (A) Vacuoles were isolated from yeast cells grown in nitrogen starvation medium. Uptake was performed as described in the legend to Figure 2A, except that different cytosolic extracts from [³H]leucine-labeled cells were used (lanes 2–6). For the reaction in lane 2, a cytosolic extract from wild-type cells grown at 30°C was used; for the sample in lane 3, a cytosolic extract from cells heat-shocked for 2 h at 37°C was used together with an ATP-regenerating system; in lane 4 the same cytosol was used as in lane 3 except that an ATP-regenerating system was not present; for the reaction in lane 5, a cytosol from cells in which the genes for the two constitutively expressed cytosolic hsp70s in yeast had been deleted was used; for the reaction in lane 6, a cytosol immunodepleted with an antiserum recognizing Ssa1, Ssa2, and Ssa3 was used. Radioactivity of the acid-soluble material was determined by liquid scintillation counting. Data and SD were calculated from two independent experiments performed in triplicate. (B) The amounts of cytosolic hsp70s in the cytosol prepared from heat-shocked cells (HS) compared with non–heat-shocked cells (−) and the amount of cytosolic hsp70s in immunodepleted cytosol (DEP) compared with control cytosol (−) were estimated by Western blot analysis using an antiserum that recognized Ssa1, Ssa2, and Ssa3 (HSP70s; E. A. Craig, personal communication). A Western blot with an antiserum raised against carboxypeptidase Y (CPY) was used to check whether the same amount of protein was loaded in each lane of the gel.

Uptake and Degradation of Glyceraldehyde-3-Phosphate Dehydrogenase by Isolated Vacuoles

Glyceraldehyde-3-phosphate dehydrogenase is a soluble protein that has two isoforms in yeast: GPD1 and GPD2 (Albertyn et al., 1994; Eriksson et al., 1995). GPD1 is found exclusively in the cytosol, whereas GPD2 is found in mitochondria. GPD1 is a key enzyme in the degradation of glucose. In mammalian cells and especially in hepatocytes, the enzyme is taken up by lysosomes where it is degraded. This uptake has been successfully reconstituted in vitro. (Aniento et al., 1993). On the basis of these results we postulated that GPD1 would follow the same degradation pathway in starved yeast cells.

A monoclonal antibody raised against rabbit glyceraldehyde-3-phosphate dehydrogenase recognizes on a Western blot of total yeast proteins only one band with a molecular mass of 40 kDa. The antibody does not recognize any protein on a Western blot of mitochondrial proteins (Figure 6, lane 1), demonstrating that the antibody does not recognize GPD2 (Figure 6, lane 2). The antibody also recognizes commercially available GPD1 (Figure 6, lane 3).

Figure 6.

A monoclonal antibody raised against rabbit glyceraldehyde-3-phosphate dehydrogenase recognizes glyceraldehyde-3-phosphate dehydrogenase isoform 1 in yeast. Proteins from cell extracts (CE, lane 1) and proteins from isolated mitochondria (MITO, lane 2) were separated by SDS-PAGE and transferred to a nitrocellulose membrane that was then immunodecorated (BLOT) with a monoclonal antibody raised against rabbit glyceraldehyde-3-phosphate dehydrogenase. As a control, commercially available glyceraldehyde-3-phosphate dehydrogenase isoform 1 of yeast (GPD1) was either immunoblotted with glyceraldehyde-3-phosphate dehydrogenase antibodies (GPD1, lane 3) or separated by SDS-PAGE and stained with Coomassie Blue (GPD1, lane 4).

To investigate whether GPD1 is taken up by vacuoles, the following experiment was performed (Figure 7A). Vacuoles prepared from starved cells were incubated with a cytosolic extract for 30 min at 30°C in the presence of an ATP-regenerating system. The vacuoles were pelleted, and the supernatant containing the cytosol was TCA-precipitated (Figure 7, S). The vacuoles were resuspended in import buffer containing proteinase K to digest nonimported proteins. After a 20-min incubation on ice, the digest was stopped by adding PMSF, and the vacuoles were TCA-precipitated (Figure 7, P). The TCA-precipitated cytosolic proteins (S) and the TCA-precipitated vacuoles (P) were separated by SDS-PAGE and immunoblotted with antibodies against glyceraldehyde-3-phosphate dehydrogenase. In the sample in which vacuolar proteolysis was inhibited by pretreatment with protease inhibitors, GPD1 (13% of the total GPD1) was found in the vacuole and was not degraded by the vacuolar proteases (Figure 7A, lane 4). In the control sample in which protease inhibitors were omitted, only 2% of the total GPD1 was present in the vacuole (Figure 7A, lane 2). Vacuoles were also isolated from the yeast strain ΔABYS (cl3-ABYSS-86; see MATERIAL AND METHODS), where four major vacuolar proteases (protease A, protease B, carboxypeptidase Y, and carboxypeptidase S) were deleted. In these vacuoles some of the imported GPD1 was detected inside the vacuoles (5% of total GPD1) (Figure 7A, lane 6); however, the vacuoles from the ΔABYS strain were still able to degrade the majority of the imported GPD1 (Figure 7A, compare lanes 4 and 6).

Figure 7.

Glyceraldehyde-3-phosphate dehydrogenase is imported into and degraded by isolated vacuoles. (A) Wild-type cells (lanes 1–4) and cells in which four major vacuolar proteases had been deleted (ΔABYS, lanes 5 and 6) were grown in nitrogen starvation medium. Vacuoles were prepared from both strains. In the uptake shown in lanes 3 and 4, the isolated vacuoles were pretreated for 10 min on ice with E-64 (100 μM), leupeptin (20 μM), and pepstatin (20 μM) (INH) to block vacuolar protein degradation. Cytosolic extract from wild-type cells was added to the vacuoles at 30°C for 30 min, and the uptake was terminated by pelleting the vacuoles. The supernatant was TCA-precipitated, and 50% of the supernatant (S) was separated by SDS-PAGE and immunoblotted for GPD1 (lanes 1, 3, and 5). The vacuolar pellet was resuspended in import buffer, and cytosolic proteins associated with the surface of the vacuoles were digested by externally added proteinase K (100 μg/ml) for 30 min on ice. The digest was stopped by adding PMSF, and proteins were TCA-precipitated (P), separated by SDS-PAGE, and immunoblotted for GPD1 (lanes 2, 4, and 6). The percent values given in the figure correspond to the amount of GPD1 associated with the vacuolar fraction. (B) Wild-type cells (lanes 1–4) and cells in which the 100-kDa subunit of the vacuolar ATPase subcomplex F_o had been deleted (vph1, lanes 5 and 6) were grown in nitrogen starvation medium. Vacuoles were prepared from both strains. In the uptake shown in lanes 3 and 4, the isolated wild-type vacuoles had been pretreated for 10 min on ice with 1 μM bafilomycin (BA), an inhibitor of the vacuolar ATPase. The uptake, the proteinase K treatment, and the processing of the samples were performed as described in A.

Vacuolar ATPase mutants that are unable to acidify the vacuole are defective in the degradation of substrates in the vacuole and in the transport of proteins into the vacuole (Yaver et al., 1993). We checked whether adding the ATPase-inhibitor bafilomycin (Figure 7B, lanes 3 and 4) or using vacuoles from a mutant strain missing the 100-kDa subunit of the F_o complex of the vacuolar ATPase (vph1) (Figure 7B, lanes 5 and 6) interferes with import and degradation of GPD1. In both cases GPD1 is taken up by the vacuoles (12% of total GPD1 in the bafilomycin-treated vacuoles and 11% in the vacuoles from the vph1 strain) but is not degraded.

Cytosolic hsp70s are involved in the uptake and degradation of cytosolic proteins in vacuoles (Figure 5). We investigated whether these chaperones are also involved in the uptake and degradation of GPD1 (Figure 8). GPD1 uptake was measured using vacuoles prepared from wild-type cells grown at 30°C (Figure 8, lanes 1–4) or heat-shocked cells (to increase the amount of hsp70s) (Figure 8, lanes 5 and 6) in the presence of an ATP-regenerating system (i.e., maximum protein uptake conditions as seen in Figure 5; compare lanes 2 and 3). Hsp70 in the presence of ATP enhanced GPD1 uptake (Figure 8, compare lanes 4 and 6: 12% uptake using cytosol from cells grown at 30°C versus 17% uptake using cytosol from cells grown at 37°C).

Figure 8.

Glyceraldehyde-3-phosphate dehydrogenase uptake depends on cytosolic hsp70s. Vacuoles were prepared from starved wild-type cells (lanes 1–4) and from starved wild-type cells that had been heat-shocked for 2 h at 37°C (lanes 5 and 6). In the reactions shown in lanes 3–6, the isolated vacuoles had been pretreated for 10 min on ice with E-64, leupeptin, and pepstatin (INH) to block vacuolar protein degradation. The uptake, the proteinase K treatment, and the processing of the samples was performed as described in the legend to Figure 7A.

DISCUSSION

In this article we describe the presence of a direct protein transport into yeast vacuoles. As in mammalian lysosomes, protein uptake in yeast vacuoles is dependent on cytosolic hsp70 proteins as well as unknown protease-sensitive component(s) on the vacuolar membrane. Protein uptake in both lysosomes and vacuoles is induced by starvation.

Yeast vacuoles were prepared using a modification of a published method (Haas, 1995). The enrichment of vacuoles measured by the increase of the specific activity of α-mannosidase was ∼75-fold (Table 1). This is in agreement with published values for the isolation of yeast vacuoles (Wiemken, 1975). EM pictures showed the absence of ER and mitochondrial contaminations. EM indicated that most of the vacuoles were intact (Figure 1). This is supported by the latency measurements shown in Figure 2. The EM data and enzyme activity measurements suggest that the uptake experiments were performed with highly enriched and intact vacuoles.

It is well established that starvation induces the proteolytic activity of yeast vacuoles (Hansen et al., 1977). Forty-five percent of all cellular proteins are degraded in the vacuole within 24 h (Teichert et al., 1989). Enhanced protein degradation in vacuoles under starvation is thought to reflect the need for higher protein turnover in these cells. We show that uptake and degradation efficiency of isolated vacuoles is stimulated approximately fivefold under starvation (Figure 2A). This induction level is similar to that reported in mammalian lysosomes (Cuervo et al., 1995).

How starvation induces protein uptake by vacuoles is unclear. Interestingly, starvation induces the expression of cytosolic hsp70s (Satyanarayana et al., unpublished observations). Vacuolar (Figures 5 and 8) and lysosomal protein uptake (Chiang et al., 1989) are induced by hsp70s. It is therefore possible that starvation induces protein uptake in vacuoles through activation or elevation of hsp70 levels. Starvation could activate hsp70 by covalent modifications or interactions with cytosolic cofactors.

What is the mechanism of the uptake of proteins into the vacuole? Autophagocytosis (Dunn, 1994) and direct protein import similar to that described in other cell organelles such as mitochondria (for review, see Schatz and Dobberstein, 1996) are possible mechanisms.

Autophagocytosis describes two distinct protein transport processes: microautophagy and macroautophagy. Both are induced by starvation. Microautophagy is a pathway involved in degradation of cytoplasmic proteins and organelles. The cytosolic constituents are directly taken up at the vacuolar surface. Engulfment might occur by membrane invagination or through the formation of vacuolar protrusions (Dunn, 1994). Macroautophagy is also a degradative process, but the sequestration event does not occur at the vacuolar membrane. Double membrane structures (autophagosomes) are formed in the cytosol, and these engulf proteins or organelles (Dunn, 1994). The autophagosomes fuse with the vacuole, releasing unilamellar vesicles into the vacuolar lumen. These vesicles are subsequently degraded by vacuolar hydrolases (Takeshige et al., 1992). Both autophagocytic pathways appear to be nonspecific, as shown by the uptake of several cytosolic proteins into the vacuole under starvation conditions (Egner et al., 1993). In our system, the cytosolic proteins tested so far were glyceraldehyde-3-phosphate dehydrogenase and hexokinase, a noncytosolic protein like the mitochondrial Dna-J homologue Mdj1, which was not taken up by isolated vacuoles (our unpublished results). Additional studies have to investigate whether a recognition mechanism on yeast vacuoles can distinguish between cytosolic proteins and non-cytosolic proteins. In mammalian cells a consensus sequence (KFERQ) is recognized by the cytosolic heat shock protein hsc73 and is believed to be responsible for the specific uptake of proteins by lysosomes (Chiang et al., 1989). In higher eukaryotes, ∼20% of the cytosolic proteins contain this consensus sequence. None of the cytosolic proteins used in our in vitro system contains a KFERQ consensus motif.

Fructose-1,6-bisphosphatase (FBPase), one of the key enzymes of the gluconeogenic pathway, is rapidly degraded by an autophagocytic process when cells are shifted from a poor carbon source to glucose-containing media. Under these conditions, Chiang and Schekman (1991) could localize the enzyme in the vacuole and demonstrate that its degradation requires protease A, one of the major hydrolytic enzymes in the vacuole. In contrast, Schork et al. (1994a,b) found that FBPase degradation requires the proteasome and under certain conditions is independent of protease A; however, in the presence of glucose, at least part of the FBPase is degraded by autophagy. Immunofluorescence experiments show a punctate FBPase pattern after addition of glucose (Chiang et al., 1996) that could be explained by the formation of autophagocytic vesicles. In some FBPase transport mutants, the protein is found inside small vesicles that are believed to be intermediates in the degradation pathway of FBPase (Hoffman and Chiang, 1996). These vesicles are distinct from the vacuole and other endosomal transport vesicles (Huang and Chiang, 1997). Cytosolic proteins such as GPD1 could have been transported to the vacuole in the same vesicles as those used to transport FBPase. For several reasons we consider this possibility rather unlikely. The way the cytosol and the vacuoles were prepared for the uptake experiments makes it very unlikely that those vesicles were present (Figure 1A). FBPase transport into the vacuole could only be reconstituted in a semi-intact cell system and not in an in vitro system containing purified organelles (Shieh and Chiang, 1998). Vesicular protein transport processes are completely blocked at 15°C (Pelham, 1989). The uptake of proteins into isolated vacuoles occurs, albeit at a low level, below 15°C (Figure 3B), suggesting that no vesicular intermediate is involved in the uptake.

Protein uptake by isolated lysosomes was suggested to have similarities with the posttranslational transport of proteins from the cytosol into cell organelles, e.g., mitochondria (Dice, 1990; Terlecky, 1994). Protein uptake by isolated vacuoles also shows similarities to these transport systems. Hsp70s are involved in the targeting of proteins to their correct destinations in several systems. Cytosolic hsp70s are known to assist in the translocation of proteins into the ER (Chirico et al., 1988; Deshaies et al., 1988) and into mitochondria (Hachiya et al., 1995). They are also known to interact with cytoplasmic proteins that are destined to the lysosomes for degradation (Chiang et al.,1989; Chiang et al., 1991; Terlecky et al., 1992). As shown in this article, hsp70s are involved in the uptake of cytosolic proteins into yeast vacuoles (Figures 5 and 8). Cytosolic heat shock proteins of the 70-kDa family play a major role on the cis-side of the organellar membrane by keeping the transported proteins in a loosely folded, import-competent conformation. Receptor proteins on the cis-side of the target membrane are presumably responsible for the targeting specificity. We have shown that the vacuolar membrane contains a protease-sensitive component involved in uptake (Figure 4). On the trans-side of ER and mitochondrial membranes, other chaperones exist that are generating the driving force for the membrane translocation (Horst et al., 1997). In the mammalian lysosomes, an hsp70 was also found to be associated with the lysosomes (Agarraberes et al., 1997; Cuervo et al., 1997). Whether a luminal vacuolar hsp70 exists in yeast awaits further investigation.

Hsp70s prevent the accumulation of denatured proteins generated as a result of exposure to high temperatures or some other type of stress. There are two ways by which this is achieved: hsp70-bound proteins are either transferred to the proteasome machinery or their aggregation is prevented by hsp70s that assist in their renaturation (Parsell and Lindquist, 1993). Hsp70s may in fact have a third function: the delivery of misfolded proteins to the lysosomes/vacuoles for degradation.

Degradation of GPD1 is dependent on a functional vacuolar ATPase. Bafilomycin and a vph1 mutation block GPD1 degradation (Figure 7B). Interestingly, maturation of most vacuolar hydrolases occurs, albeit at reduced levels, in strains in which subunits of the vacuolar ATPase had been deleted or in the presence of ATPase inhibitors. GPD1 degradation could require denaturation of GPD1 by the low vacuolar pH, turning the protein into a better substrate for the vacuolar proteases, whereas protease precursor processing does not depend on a low pH (Wolff et al., 1996).

The availability of an in vitro vacuolar protein transport system should allow a better characterization of this protein degradation pathway. Comparison of the lysosomal and vacuolar systems will assist in our understanding of the evolution of this complex biological process.