Plasmids were constructed using standard cloning protocols (Sambrook et al., 1989). All genes encoding expression constructs were sequenced in their entirety. All substrate proteins contain an engineered single hemagglutinin (HA) epitope tag at C-termini. The expression plasmids for all proteins were constructed by placing coding sequences under the strong constitutive TDH3 (encoding glyceraldehyde-3-phosphate dehydrogenase), inducible GAL1, or moderate constitutive GAS1 promoters as indicated, in yeast centromeric vectors. A plasmid list and oligonucleotide primers used in plasmid construction can be found in Tables S2 and S3, respectively.

S. cerevisiae strains used in this study are described in Table S1. Anti-HA mAb (HA.11) was purchased from Covance Research Products (Princeton, NJ). Anti-Kar2p and anti-Sec61p antibodies were provided by Peter Walter (University of California, San Francisco, CA). Anti-PrA was a gift from Tom Stevens (University of Oregon, Eugene, OR). Anti-Gas1p rabbit antiserum was previously described (Spear and Ng, 2003). Polyclonal anti-ubiquitin antibody was purchased from Abcam (Cambridge, United Kingdom), monoclonal anti-3-phosphoglycerate kinase (PGK) and anti-V5 antibody from Invitrogen, polyclonal anti-GFP antibody from Clontech (Palo Alto, CA), and monoclonal anti-FLAG antibody from Sigma-Aldrich (St. Louis, MO). Secondary antibodies labeled with Alexa Fluor 488 or Alexa Fluor 596 were purchased from Molecular Probes (Eugene, OR).

Cells were grown synthetic complete media (SC) lacking methionine, cysteine, and components for plasmid selection where applicable; 3.0 OD₆₀₀ units of cells were labeled with 82.5 μCi of [³⁵S]methionine/cysteine (EasyTag EXPRESS ³⁵S, Perkin Elmer-Cetus, Waltham, MA) and chased with excess cold amino acids for times indicated. For galactose induction, cells were incubated for 4 h in 2% galactose media before experiments. Protein immunoprecipitation and resolution by SDS-PAGE is carried as described (Vashist et al., 2001). Gels were exposed to phophor screens for 24–48 h and scanned and quantified using the TyphoonTM phosphorimager and ImageQuant TL software (GE Healthcare Life Sciences, Uppsala, Sweden). All data plotted reflect three independent experiments with the SD of the mean indicated.

Cells were grown to midlog phase in synthetic media. Cessation of protein synthesis was initiated by adding cycloheximide to 200 μg/ml to begin the chase. At each time point, the chase was terminated by transferring an aliquot of cells into 1 ml ice-cold 10% trichloroacetic acid (TCA). Detergent lysates were prepared by mechanical cell disruption and TCA precipitation as described previously (Vashist et al., 2001). SDS-PAGE–separated proteins were transferred to nitrocellulose membranes and incubated in LI-COR blocking buffer (Lincoln, NE). After incubation with appropriate primary antibodies, membranes were washed with phosphate-buffered saline (PBS) containing 0.1% Tween 20. Secondary antibodies labeled with IRDye 680 (infrared dye) or IRDye 800 were diluted 15,000-times in blocking buffer containing 0.1% Tween 20 and used for detecting bound primary antibodies. After washing, membranes were scanned and quantified by using the Odyssey infrared imaging system (LI-COR Biosciences).

Cells expressing misfolded proteins were resuspended in 10% TCA chilled on ice. After bead beating, precipitated proteins were pelleted by centrifugation at 14000 rpm for 10 min at 4°C. The pellet was resuspended in TCA resuspension solution (3% SDS, 100 mM Tris-base, 3 mM DTT). Protein sample, 50 μl, was mixed with 550 μl of IPS II (50 mM Tris-Cl, pH 7.4, 150 mM NaCl, and 1% Triton X-100), 6 μl of protease inhibitor cocktail (Roche, Nutley, NJ), and 6 μl of 100 mM PMSF. Misfolded proteins were immunoprecipitated and detected by anti-HA antibody. The ubiquitinated proteins were detected by using anti-ubiquitin antibody.

Cells expressing wild-type or mutant proteins were harvested and resuspended in cytosol buffer (20 mM HEPES, pH 7.4, 14% glycerol, 100 mM KOAc, and 2 mM MgOAc) and disrupted by bead beating for five 1-min full-speed cycles on a vortex mixer. For assays of PrA and ΔssPrA, Triton X-100 was added to lysates (to 1% vol/vol). After 5-min incubation at 30°C, trypsin was added at 5.0 μg/ml and incubated at 30°C. A portion was removed at each time point, and the reaction was terminated by adding TCA to 10% and proteins precipitated on ice. The recovered proteins were analyzed by SDS-PAGE/Western blotting analysis using the relevant antibodies.

Indirect immunofluorescence was performed as described previously (Spear and Ng, 2003) with minor modifications. Briefly, cells were fixed with 3.7% formaldehyde at 30°C for 90 min and spheroplasted by zymolyase digestion (1 mg/ml zymolyase 20T (United States Biological, Marblehead, MA), 0.1 M potassium phosphate, pH 7.5, 1.4 M sorbitol). Spherolplasts were applied to each well of a poly-l-lysine–coated slide for 10 min and washed. Slides were immersed in methanol for 6 min and in acetone for 30 s at −20°C. Each well was blocked with PBST (PBS containing 0.1% Triton X-100) containing 5% bovine serum albumin. Primary antibodies and secondary antibodies were incubated in this buffer incubated for 90 min each. Slides were washed twice with PBS buffer after each application. Primary antibodies HA.11 mAb (Covance, Princeton, NJ), anti-V5 (Invitrogen) and polyclonal anti-Kar2p were diluted to 1:200, 1:200, and 1:500, respectively. Kar2p is an ER-resident protein and an established marker for the ER and nuclear envelope (Vogel et al., 1990). Secondary antibodies Alexa Fluor 488 goat anti-mouse and Alexa Fluor 596 goat anti-rabbit were diluted to 1:500. Nuclei were visualized by DAPI (4′,6′-diamidino-2-phenylindole) staining. Samples were examined by confocal microscopy using Axio Imager.M1 microscope with 100× 1.4 NA oil Plan- Aprochromat objective (Carl Zeiss MicroImaging, Oberkochen, Germany). Images were archived by LSM Image Browser (Zeiss) and Adobe Photoshop (San Jose, CA).

Cells overexpressing NLS-GFP-NES or NLS-GFP-P12 were grown at room temperature and shifted to 23, 30, or 37°C for 1 h. Cells were incubated for 5 min with Hoechst 33342 (Invitrogen) before viewing. Cells were examined by confocal microscopy using Axio Imager.M1 microscope with 100 × 1.4 NA oil Plan- Aprochromat objective (Carl Zeiss MicroImaging). Images were archived by LSM Image Browser (Zeiss) and Adobe Photoshop.

Cells expressing target substrates were harvested and suspended in lysis buffer (50 mM HEPES-KOH, pH 7.4, 250 mM sorbitol, 150 mM KOAc, and 5 mM MgOAc) containing 1 mM PMSF. Cells were mechanically disrupted with zirconium beads by vortexing at full speed (1 min vortex and 1 min on ice, 10 cycles). The lysate was clarified by centrifugation at 14,000 rpm for 5 min at 4°C twice. NaCl (final 0.5 M), Triton X-100 (final 0.5%), and protease inhibitor cocktail was added to the clarified lysate. IgG beads (GE Biosciences, Fairfield, CT) were added, and the mixture was gently rotated at 4°C for 2 h. Beads were washed three times with lysis buffer containing 0.5% Triton X-100. The bound proteins to beads were eluted, separated by SDS-PAGE, and detected on immunoblots.

Wolf and colleagues created ΔssCPY*, a model CytoQC substrate, by deleting the CPY* signal sequence (Park et al., 2007). In our hands, ΔssCPY* degrades rapidly and is dependent on Hsp70 and Ydj1p as previously reported (Figure S1A and data not shown). We also found that a fraction of ΔssCPY* translocates into the ER, making CytoQC-specific analyses difficult (Figure S1A) This observation is consistent with the behavior of signal sequence-deleted wild-type CPY (Blachly-Dyson and Stevens, 1987). Using ΔssCPY* as a guide, two novel substrates were developed to overcome the limitation. The first was made by deleting the signal sequence of vacuolar proteinase A (PrA), to direct mislocalization to the cytosol and cause its misfolding (Figure 1A) (Klionsky et al., 1988; Park et al., 2007). Next, a short internal deletion (Figure 1A, residues 25 through 36) was engineered in the GFP to disrupt its well-ordered β-barrel structure to create Δ2GFP (Ormo et al., 1996).

An in vitro trypsinization assay was applied to assess the folding states of ΔssPrA and Δ2GFP. Unfolded proteins typically exhibit protease hypersensitivity compared with folded proteins (Taniuchi and Anfinsen, 1969). Consistent with this criterion for unfolded proteins, both molecules degraded by the first time point, whereas folded PrA and GFP were stable over the course of the experiment (Figure 1B). In vivo, a metabolic pulse-chase experiment showed that ΔssPrA and Δ2GFP are highly unstable compared with their stable, folded forms (Figure 1C). Thus far, all model CytoQC substrates require Hsp70, Hsp40, and the 26S proteasome for their degradation (Zhang et al., 2001; McClellan et al., 2005; Park et al., 2007; Metzger et al., 2008). To determine further whether ΔssPrA and Δ2GFP are representative model substrates of CytoQC, metabolic pulse-chase assays were performed in wild-type, ssa1-45^ts, and Δydj1 cells. ssa1-45^ts cells contain an ssa1 temperature sensitive allele with the SSA2, SSA3, and SSA4 genes deleted (Becker et al., 1996). This genetic background is necessary because SSA genes are essential only when simultaneously deleted. In addition, substrate stability was tested in the presence and absence of proteasome inhibitor MG132 in drug-sensitized Δpdr5 cells (Fleming et al., 2002; Lipford et al., 2005). As shown in Figure 1, D and E, the substrates are strongly stabilized in the mutant strains and in the presence of MG132. Together, these data show that ΔssPrA and Δ2GFP are bona fide substrates of CytoQC.

Next, the intracellular localization of the model substrates was analyzed by indirect immunofluorescence combined with confocal imaging. Surprisingly, ΔssPrA displayed strong nuclear staining in addition to the expected, but weaker, cytosolic staining (Figure 2, ΔssPrA). Δ2GFP also localized to the nucleus but in a less concentrated distribution (Figure 2, Δ2GFP). ΔssCPY* distribution is similar to ΔssPrA except that a fraction is also localized to the ER (data not shown). Nuclear localization observed for all three substrates suggested that it is a part of the CytoQC mechanism. Alternatively, it could be explained by the presence of cryptic NLSs. No classical NLS is predicted for Δ2GFP according to the PSORT II algorithm (Horton et al., 2007). ΔssPrA contains a sequence, PVRRK, that resembles the monopartite NLS motif P(B3X), where B is a basic amino acid and X is any residue. To test if the motif is functional, the conforming sequence was destroyed by mutation to PVAAA (ΔssPrA-3A). ΔssPrA-3A localizes to the nucleus, indicating that its trafficking is not dependent on the NLS-like sequence (Figure S1B). We next explored the functional significance of substrate nuclear localization.

E3 ubiquitin ligases are a broad class of factors often responsible for substrate recognition (Kerscher et al., 2006). Gardner et al. (2005) demonstrated that the nuclear San1p E3 ligase is required for the degradation of damaged nuclear proteins. If misfolded cytosolic proteins specifically traffic to the nucleus, it seemed plausible that San1p also directs their turnover. To test this notion, wild-type and Δsan1 cells expressing the model substrates were analyzed by metabolic pulse-chase analysis. As shown in Figure 3A, their turnover is strongly curtailed in the Δsan1 mutant compared with wild type (similarly, ΔssPrA-3A degradation is also dependent on San1p; Figure S1C). In addition, stabilized ΔssPrA and Δ2GFP accumulated in Δsan1 nuclei, showing that substrate localization is independent of the E3 enzyme (Figure 3B). We next determined if San1p is required for substrate ubiquitination. ΔssPrA and Δ2GFP were immunoprecipitated from wild-type and Δsan1 cells, resolved by SDS-PAGE, blotted onto nitrocellulose, and probed with anti-ubiquitin antibodies. Both substrates displayed a significant reduction in polyubiquitination when expressed in the Δsan1 mutant compared with wild type (Figure 3C). These data show that San1p is required for efficient polyubiquitination of these substrates. However, residual degradation was observed in Δsan1 cells suggesting other pathways can partially compensate in its absence (see Figure 6).

To determine if San1p interacts with substrates, coimmunoprecipitation experiments were performed. For this purpose, a functional V5-tagged San1p was introduced into MG132-sensitized cells expressing ΔssPrA, Δ2GFP, or no substrate. Cells were treated with MG132 for 4 h to stabilize substrates, and detergent extracts were prepared under nondenaturing conditions. Cell extracts were next subjected to immunoprecipitation with anti-HA antibody. Isolated proteins were resolved by SDS-PAGE and blotted onto nitrocellulose filters probed with anti-V5 and anti-HA antibodies. As shown in Figure 4A, San1p-V5 is specifically detected in complexes containing ΔssPrA and Δ2GFP (top panel).

In this study, ΔssPrA and Δ2GFP were expressed from the strong TDH3 promoter to facilitate localization and for consistency to prior studies using similarly active promoters (Kaganovich et al., 2008). We wondered if the San1p pathway is also used for moderate substrate loads. For this, substrate coding sequences were placed under the control of the moderate GAS1 promoter (Nuoffer et al., 1991). Under its control, ΔssPrA and Δ2GFP expression levels are reduced 4.7- and 5.2-fold, respectively (Figure S2A). Pulse-chase experiments showed their rapid degradation in wild-type cells and stabilization in ssa1-45, and Δydj1 cells (Figure S2, B and C). Importantly, degradation remains highly dependent on San1p (Figure S2D). From these data, we conclude that the San1p pathway is a constitutive mechanism of cytosolic quality control.

In the above experiment, substrate turnover rates were increased marginally in cells bearing reduced substrate loads, an indication of limiting factors in CytoQC (compare Figure S2D to Figure 1C). The steady state localization of high-expression substrates in the nucleus suggested that functions such as ubiquitination or proteasomal proteolysis within the nucleus could be a bottleneck (Figure 2). During ERAD, increasing the level of the rate-limiting Hrd1 E3 enzyme elevates substrate turnover rates (Gardner et al., 2000; Garza et al., 2009). To assess whether San1p-dependent ubiquitination might be rate limiting in CytoQC, SAN1 coding sequences were placed under the control of the inducible GAL1 promoter. Wild-type cells expressing ΔssPrA and Δ2GFP were transformed with the GAL1-SAN1 plasmid or empty vector. After a shift to galactose media, substrate degradation accelerated significantly in GAL1-SAN1 cells compared with vector control (Figure 4C). To determine if overexpression affects San1p localization, GAL1-SAN1 was appended with the V5 epitope tag. Under identical assay conditions, galactose-induced San1p-V5 accelerates substrate degradation as well as San1p and San1p-V5 localizes exclusively to nuclei (Figure S3). Intracellularly, substrate localization patterns did not change significantly, other than the expected reduced signal, suggesting that nuclear import is not rate-limiting (Figures 4B and 4D).

To visualize the site of decay, a cycloheximide chase assay was combined with indirect immunofluorescence. In wild-type cells, ΔssPrA localizes mostly to the nucleus before cycloheximide addition (Figure 5, wild type, 0 min). Nuclear localization is observed over the time course even as the signal diminishes, with staining virtually absent at 60 min. By contrast, the nuclear signal persists strongly throughout the time course when ubiquitination is disrupted by the Δsan1 mutation (Figure 5, Δsan1 panels). Substrate degradation was also analyzed in the xpo1–1 mutant that exhibits broad defects in nuclear export (Stade et al., 1997). XPO1/CRM1 encodes the major nuclear export receptor responsible for the trafficking of mRNA, preribosomal subunits, and proteins bearing nuclear export signals (Macara, 2001; Cullen, 2003; Zemp and Kutay, 2007). In cells deficient for XPO1/CRM1 function, CytoQC substrate degradation is as efficient as wild type, showing that this major export pathway is not required (Figure S4). However, the experiment does not rule out the existence of an XPO1/CRM1-independent pathway used for their transport. Taken together, these data show that some misfolded cytosolic proteins traffic to the nucleus where they are ubiquitinated by San1p and degraded by the 26S proteasome.

Ubr1p, the E3 ubiquitin ligase for substrates of the “N-end rule” degradation mechanism (Bartel et al., 1990), was recently reported to stabilize a CytoQC reporter protein called ΔssCL*myc (Eisele and Wolf, 2008). ΔssCL*myc is a hybrid containing the majority of ΔssCPY* fused to myc-tagged Leu2p. To assess the requirement of Ubr1p in ΔssPrA and Δ2GFP degradation, turnover assays were performed with wild-type, Δubr1, and Δubr1Δsan1 strains. By contrast to Δsan1, cells lacking UBR1 degraded these substrates nearly as well as wild type and did not alter substrate localization (Figure 6, A and B). However, deleting both genes stabilized the substrates to a greater extent than single gene knockouts, suggesting that Ubr1p can partially compensate if San1p is absent (cf. Figure 6A with Figure 3A). In addition, ΔssPrA and Δ2GFP accumulate strongly in nuclei of Δubr1Δsan1 cells (Figure 6B). Taken together, these data show that the San1p pathway is the major mechanism of quality control for ΔssPrA and Δ2GFP. Interestingly, neither E3 enzyme is required for substrate localization to the nucleus suggesting that the process is ubiquitin-independent.

The loss of cytoplasmic Hsp70 function shuts down the turnover of misfolded cytosolic proteins (Zhang et al., 2001; McClellan et al., 2005; Park et al., 2007; Metzger et al., 2008). Substrates aggregate under these conditions suggesting that Hsp70 proteins are required to maintain the solubility of substrates for degradation. To analyze the role of Hsp70 in CytoQC and various other functions, the ssa1-45^ts mutant has proven to be particularly useful (Becker et al., 1996). However, due to the specific nature of the mutant (simultaneous deletion SSA2, SSA3, and SSA4), it remains unclear which Hsp70 isoforms are involved in PQC. To better understand the role Hsp70 isoforms, we analyzed the contribution of SSA genes in PQC systematically.

First, single deletion mutants were generated and tested for their ability to degrade ΔssPrA and Δ2GFP. Pulse-chase analysis shows that no single SSA gene is essential for CytoQC (Figure S5A). Next, double mutants were generated in every combination and turnover of the substrates again measured. Strikingly, only one combination displayed a clear defect. In strains deleted of SSA1 and SSA2, both substrates were dramatically stabilized at the normal growth temperature of 30°C (Figure 7A and Figure S5B). However, the extent of stabilization was less than the ssa1-45 strain at the restrictive temperature, which causes a complete loss of Hsp70 function (Figure 1D). This suggests that Ssa3p and Ssa4p (SSA3 and SSA4 genes strongly induce in a Δssa1Δssa2 strain (Werner-Washburne et al., 1987) can partially compensate, albeit poorly, in the absence of Ssa1p and Ssa2p. These data show that Ssa1p and Ssa2p are the primary Hsp70 isoforms in constitutive CytoQC and are functionally redundant.

To determine whether Ssa proteins bind CytoQC substrates before degradation, FLAG-tagged Ssa1p was introduced into the Δssa1Δssa2 strain. In this strain, substrate degradation was fully restored confirming that the tagged protein is functional (data not shown). In coimmunoprecipitation experiments, Ssa1-FLAG was recovered in a specific complex with ΔssPrA and Δ2GFP (Figure 7B, bottom panels).

Next, we sought to understand the roles Ssa1p and Ssa2p play in CytoQC. For this, we performed indirect immunofluorescence to visualize the fate of substrates stabilized in the Δssa1Δssa2 strain. Through this analysis, two major observations were made. The first shows substrates forming large intracellular inclusions instead of the normally diffuse patterns (Figure 7C). This is consistent with the established role of Hsp70 in keeping unfolded proteins soluble (Skowyra et al., 1990; Cyr, 1995). More notable is the distribution of substrates. More of the proteins were found in large cytosolic inclusions instead of the nucleus. Although the induction of SSA3 and SSA4 genes in this mutant is sufficient for viability (Werner-Washburne et al., 1987), it is not enough to facilitate efficient transport of the substrates into the nucleus. The defect is not due to substrate excess because they accumulate primarily in the nucleus in the Δsan1Δubr1 strain (Figures 6B). These data show that Ssa1p and Ssa2p are required for efficient transport of ΔssPrA and Δ2GFP into the nucleus for degradation.

The Hsp70 nucleotide exchange factor Sse1p (Hsp110) is required for the degradation of misfolded VHL tumor suppressor in yeast cells (McClellan et al., 2005). Sse1p is a nucleotide exchange factor for Hsp70 that mediates the switch from high- to low-affinity binding. Sse1p also has the capacity to bind peptide substrates (Goeckeler et al., 2008; Polier et al., 2008). For these reasons, we wanted to learn if it is required for nuclear import and degradation of ΔssPrA and Δ2GFP. In a Δsse1 mutant, both substrates are stabilized as strongly as in Δsan1 (Figure 8A). To assess its role in substrate nuclear localization, indirect immunofluorescence was performed using Δsan1 and Δsse1Δsan1 mutants. The Δsan1 background was used to stabilize proteins able to bypass an import block, if one exists. In Δsan1 cells, substrate accumulation in nuclei was observed as usual (Figure 8B). In Δsse1Δsan1 cells, however, we observed substrate-specific effects. For ΔssPrA, the distribution becomes slightly more cytosolic suggesting that Sse1p may play a role for its efficient nuclear import. For Δ2GFP, there is no discernable difference between the strains (Figure 8B). It should be noted that nuclear import is not strongly blocked for either substrate in the absence of Sse1p and that the resulting intracellular distribution becomes similar for both substrates in the Δsse1Δsan1 strain background. Although Sse1p is required for substrate degradation, we conclude that, at best, it enhances the import of some substrates.