Experiments were carried out in BY4741 (MATa; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0) obtained from Euroscarf. Heterologous expression of C-terminally FLAG-tagged human αSyn under the control of a GAL10 promoter was performed using a pESC-URA plasmid (Stratagene). Therefore, a previously published αSyn-construct in pESC-HIS (Büttner et al., 2008, 2013b) was cut with SpeI and ClaI and ligated into pESC-URA. All oligonucleotides used in this study are listed in Supplementary Table S1. Constructs for overexpression of Vma1 and Vph2 in pESC-HIS were kindly provided by Ruckenstuhl et al. (2014). For overexpression of FLAG-tagged wild type Pep4 (Pep4^WT) and the double point mutant (D109A, D294A) of Pep4 (Pep4^DPM) under a galactose promoter, previously published constructs in pESC-HIS were used (Carmona-Gutierrez et al., 2011). The PMA1 gene was amplified and cloned with NotI and SpeI into pESC-HIS and PEP1 was inserted into a pESC-LEU plasmid using NotI. To visualize αSyn localization and aggregation, the C-terminally tagged variant αSyn-GFP was amplified from pUG23 (Zabrocki et al., 2005) and cloned with SpeI and ClaI into pESC-URA.

Transformation of plasmids into yeast cells was performed using the standard lithium acetate method (Gietz and Woods, 2002). Deletion mutants were either obtained from Euroscarf or generated via homologous recombination according to (Janke et al., 2004; for a detailed list see Supplementary Table S2). All oligonucleotides and plasmids used for this approach are listed in Supplementary Table S1. At least four different clones were tested after plasmid transformation or genomic replacement to rule out clonogenic variations.

Yeast strains were grown in synthetic complete (SC) medium containing 0.17% yeast nitrogen base (Difco, BD Biosciences), 0.5% (NH₄)₂SO₄ and 30 mg/L of all amino acids (except 80 mg/L histidine and 200 mg/L leucine), 30 mg/L adenine and 320 mg/L uracil with 2% D-glucose (SCD) or 2% D-galactose (SCG) for GAL10-driven expression of α-Syn, Pep4^WT or Pep4^DPM. All media were prepared with double distilled water and subsequently autoclaved (25 min, 121°C, 210 kPa). Amino acid mixtures were sterilized separately as 10× stocks and added after autoclaving. For solid media, 2% agar was admixed. Full media (yeast extract peptone dextrose, YEPD) agar plates contained 1% yeast extract (Bacto, BD Biosciences), 2% peptone (Bacto, BD Biosciences) and 4% D-glucose.

Experiments were performed using overnight cultures grown in SCD for 16–20 h at 28°C and 145 rpm. These cultures were inoculated in 10 mL SCD (in 100 mL Erlenmeyer flasks) to an OD₆₀₀ of 0.1 and grown to OD₆₀₀ 0.3. Subsequently, cells were transferred into 96-well deep well plates (VWR), whereat 500 μL of cell culture was used per well. After pelleting, cells were shifted to 500 μL SCG per well for induction of expression. For inhibition of Pep4 enzymatic activity, 50 μM pepstatin A dissolved in DMSO (Sigma) were added to cells upon galactose-mediated induction of expression.

To measure loss of membrane integrity, approximately 2 × 10⁶ cells were collected in 96-well plates via centrifugation at indicated time points after induction of expression and resuspended in 250 μL phosphate buffered saline (PBS, 25 mM potassium phosphate; 0.9% NaCl; adjusted to pH 7.2) containing 100 μg/L propidium iodide (PI). Cells were incubated for 10 min in the dark, pelleted via centrifugation and washed once in 250 μL PBS. PI staining was quantified via flow cytometry (BD LSR Fortessa), analyzing 30,000 cells with BD FACSDivia software.

Whole cell extracts were generated by chemical lysis. Cells equivalent to an OD₆₀₀ of three (for general immunoblotting) or an OD₆₀₀ of eight (for detection of αSyn oligomers) were harvested 24 h after induction of expression, resuspended in 200 μL of 0.1 M NaOH and incubated shaking with 1400 rpm and 21°C for 5 min. After centrifugation with 4000 rpm for 5 min, pellets were resuspended in 150 μL 1× Laemmli buffer (50 mM Tris-HCl; 2% SDS; 10% glycerol; 0.1% bromophenol blue; 100 mM 2-mercaptoethanol; adjusted to pH 6.8) and again shaken with 1400 rpm and 21°C for 5 min. Of note, samples for detection of αSyn oligomers were prepared with 1× Laemmli buffer without 2-mercaptoethanol (semi-native approach). Samples were centrifuged with 13,000 rpm for 1 min and 15 μL of the supernatant was used for standard SDS-PAGE. To detect αSyn oligomers, polyacrylamide gels without SDS were applied and electrophoresis was performed at 4°C (semi-native approach). Immunoblotting was performed using standard protocols with antibodies directed against αSyn (Sigma, S3062), FLAG epitope (Sigma; F3165), influenza hemagglutinin protein (HA epitope; Sigma H3663), Pep1 (Abcam; ab113690), yeast glyceraldehyde 3-phosphate dehydrogenase (GAPDH; gift from Sepp Kohlwein, University of Graz) and the respective peroxidase-conjugated affinity-purified secondary antibodies (Sigma). A ChemiDoc™ Touch Imaging System (Bio-Rad) was used for detection, and subsequent densitometric quantification was performed with Image Lab 5.2 Software (Bio-Rad).

In vivo crosslinking experiments were performed with adapted protocols according to Klockenbusch and Kast (2010). All washing steps were accomplished with and all reagents were solubilized in 0.1 M sodium phosphate buffer (0.02 M Na₂HPO₄; 0.08 M NaH₂PO₄; adjusted to pH 7.4). In brief, cells equivalent to an OD₆₀₀ of five were harvested 24 h after induction of expression, washed once and resuspended in 1 mL of 1% formaldehyde. Of note, a negative control for every sample was resuspended in 1 mL buffer. Cells were incubated for 9 min and centrifuged for 1 min with 13,000 rpm. 1 mL of 1.25 M glycine was added to stop the reaction and incubated for 5 min. Cells were washed five times, followed by lysis and immunoblotting, conducted as described for semi-native detection of αSyn oligomers.

For densitometric quantification, signals were normalized to the respective GAPDH signal and fold change of αSyn oligomers upon expression of Pep4^WT or Pep4^DPM were plotted.

Indicated molecular weights in all shown immunoblots represent the apparent molecular weight (kDa) determined with a PageRuler prestained protein ladder (ThermoFisher Scientific) as indicated by the manufacturers migration pattern.

To measure the enzymatic activity of Pep4, a fluorometric CatD activity assay kit from Abcam (ab65302) was used and the protocol was adapted for yeast samples. Briefly, 2 × 10⁶ cells were harvested at specified time points after induction of expression. Protein extraction was performed with glass beads and the supplied CD cell lysis buffer and the resulting protein concentration was determined via a Bradford assay (Bio-Rad). Afterwards, 0.1 μg protein was used for the CatD activity assay. Reactions were incubated for 2 h at 28°C and the fluorescence signal was measured with a Tecan Genios pro microplate reader (ex: 328 nm, em: 460 nm) and plotted as fold change compared to the empty vector control. Of note, a Δpep4 strain was used as background control in each assay, being subtracted from presented Pep4 activity measurements.

Measurement of basal cytosolic calcium levels [Ca²⁺]_cyt was performed as described in Büttner et al. (2013a). Briefly, galactose-driven expression of αSyn was established in yeast strains carrying the vector pEVP11/AEQ89, coding for the bioluminescent reporter protein aequorin (kind gift from Kyle W. Cunningham) and cells were cultivated as described above. At indicated time points, approximately 1 × 10⁸ cells were transferred into a 96-well plate and harvested by centrifugation. The cell pellet was resuspended in 200 μL SCD medium (plus 4 μM colenterazine h, ThermoFisher Scientific) and incubated for 1 h in the dark. Excess colenterazine h was removed by washing once in SCD followed by a further incubation step of 30 min. The basal luminescence signal was measured in 0.5 s intervals for 20 s with an Orion II Microplate Luminometer (Berthold Detection Systems). The obtained signal was normalized to OD₆₀₀ of each well and aequorin protein levels in each transformant.

Approximately 1 × 10⁸ cells were harvested at indicated time points after induction of expression and washed with 500 μL of YEPD containing 100 mM HEPES (pH 7.6). After centrifugation, the pellet was resuspended in 500 μL of YEPD with 100 mM HEPES (pH 7.6) and 400 μM quinacrine. After incubation for 10 min at 28°C and 145 rpm, samples were incubated 5 min on ice. After centrifugation, cells were washed twice in 500 μL ice-cold HEPES buffer supplemented with 2% D-glucose. For PI co-staining, 500 μL of HEPES buffer with 2% D-glucose and 100 μg/L PI was added and incubated 10 min in the dark on ice. Cells were analyzed using a Zeiss Axioskop epifluorescence microscope. For quantification, 300–700 cells per genotype and experiment were manually counted.

Approximately 1 × 10⁶ cells from cultures were harvested at indicated time points after induction of expression and washed in 500 μL 10 mM HEPES buffer (pH 7.4) containing 5% D-glucose. Cells were resuspended in the same buffer containing 10 μM MDY-64 (Molecular Probes) and incubated for 3 min at room temperature in the dark. Subsequently, cells were washed in HEPES buffer and analyzed with a Zeiss Axioskop epifluorescence microscope. For quantification, 350–650 cells per genotype and experiment were manually counted.

Approximately 1 × 10⁶ cells from cultures were harvested at indicated time points after induction of expression and resuspended in 10 mM HEPES buffer (pH 7.4; supplemented with 5% glucose). Afterwards, CellTracker Blue CMAC (Molecular Probes) was added to a final concentration of 100 μM, cells were incubated in the dark for 20 min and subsequently analyzed with a Zeiss Axioskop epifluorescence microscope. For PI co-staining, PI (final concentration of 100 μg/L) was added to the CMAC staining solution. For analysis of vacuolar membrane permeabilization 2.5 μg/mL tunicamycin (Sigma) dissolved in DMSO was added to cells after the shift to galactose as positive control (adapted from Kim et al., 2012).

To determine mRNA levels, total RNA was extracted from respective strains (OD₆₀₀ 1) after 24 h of αSyn expression using TRIzol reagent (Thermo Scientific). Briefly, the cell pellet was resuspended in 500 μL TRIzol plus approximately 100 μl glass beads (500 μm diameter) and the samples were lysed mechanically in three cycles of 1 min. The subsequent steps were employed as described in the manual. Integrity of the isolated RNA was validated by visualizing rRNAs on an agarose gel (adapted from Aranda et al., 2012). Contaminating DNA was removed with the DNA-free Kit (Ambion) and the RNA concentration was determined spectrometrically with a NanoDrop (ND 1000 Spectrophotometer). 3.5 μg of total RNA were reverse transcribed using the M-MLV Reverse Transcriptase RNase H- (Solis Biodyne) and random hexamer primers (Thermo Scientific). Obtained cDNA was diluted accordingly and the actual q-RT-PCR was performed with the 5× HOT FIREPol EvaGreen qPCR Supermix (Solis Biodyne). Primers used in these experiments are listed in Supplementary Table S1. First, standard curve experiments were performed to ensure primer efficiency lying between 90% and 110%. The cDNA was used in duplicates for quantitative PCR amplification and the relative gene expression levels were calculated with the Comparative C_T Method (Livak and Schmittgen, 2001) with UBC6 used as housekeeping gene for normalization.

To disprove the null-hypothesis (no difference between conditions), significance and p-values were calculated using one-way ANOVA, corrected with a Bonferroni post hoc test for one variable (type of expression). Significance is indicated with asterisks: ***p < 0.001, **p < 0.01, *p < 0.05. Statistics were conducted with Origin Pro 2016 (OriginLab) and figures were prepared with Origin Pro 2016 and Adobe Illustrator CS6 (Adobe). Microscopic pictures were processed with Fiji (Schindelin et al., 2012).

General lysosomal dysfunction as well as alterations in CatD levels and activity are implicated in the pathogenesis of PD (Qiao et al., 2008; Sevlever et al., 2008; Manzoni and Lewis, 2013; Matrone et al., 2016; Moors et al., 2016). To further analyze the impact of αSyn on lysosomal integrity in general and on CatD function in particular, we used a yeast model for PD based on the heterologous expression of human αSyn. While αSyn did not affect cell proliferation, it caused an increase of cell death in early stationary phase, as indicated by PI staining (Figure 1A). Biochemical quantification of yeast CatD (Pep4) proteolytic capacity in these cells revealed a massive reduction of Pep4 enzymatic activity upon αSyn expression (Figure 1B). Values obtained for Δpep4 cells have been subtracted as background. Performing immunoblotting, we observed an increase in Pep4 protein levels upon αSyn expression (Figures 1C,D), suggesting a compensatory upregulation of Pep4 as a futile cellular attempt to counterbalance the αSyn-triggered loss of Pep4 proteolytic activity. Pep4/CatD has been shown to be released into the cytosol as a result of vacuolar/lysosomal membrane permeabilization during acetic acid-induced apoptosis in yeast and mammalian tumor cell lines (Pereira et al., 2010; Marques et al., 2013). Thus, we analyzed the effects of αSyn on the integrity of the limiting vacuolar membrane and the localization of Pep4. However, αSyn did neither induce vacuolar membrane permeabilization (Supplementary Figure S1) nor a detectable vacuolar release of a Pep4-GFP chimera into the cytosol (Figure 1E). Instead, the larger fraction of Pep4-GFP was still targeted to the vacuole, visualized via CellTracker Blue CMAC, a dye that is rapidly sequestered into the vacuolar lumen. Interestingly, expression of αSyn led to the accumulation of Pep4 in prevacuolar compartments in some cells, suggesting that αSyn interferes with the trafficking of this protease from the trans-Golgi network to the vacuole, which might in part account for the observed reduction of Pep4 activity. Thus, we overexpressed Pep4 to antagonize the αSyn-induced decrease of vacuolar catabolic activity. Indeed, co-expression of Pep4^WT was sufficient to prevent cell death caused by αSyn upon entry into stationary phase (Figures 1F,G). This cytoprotection required Pep4 protease activity, as the proteolytically inactive double point mutant Pep4^DPM (Carmona-Gutierrez et al., 2011) showed no effect (Figures 1F,G). In addition, treatment with the CatD inhibitor pepstatin A completely eliminated cytoprotection (Figure 1G). Measurement of Pep4 activity revealed that high levels of Pep4^WT not only led to a general increase of its enzymatic activity but also abrogated the αSyn-mediated reduction of proteolytic capacity (Figure 1H). Interestingly, only a minor fraction of cells died after 16 h of αSyn expression, even though Pep4 activity was already reduced to only 0.2-fold of WT levels. This indicates that the decrease of Pep4 activity during exponential growth preceded the actual onset of cell death in early stationary phase (Figures 1G,H).

As the acidic pH in the vacuolar lumen is essential for proper function of Pep4 (Sørensen et al., 1994), disturbances of pH homeostasis might contribute to the reduction in Pep4 activity observed upon αSyn expression. Therefore, we visualized acidic cellular compartments via quinacrine staining after 16 h and 24 h of αSyn expression and used PI counterstaining to exclude dead cells from the analysis. While quinacrine accumulated within the vacuoles of cells harboring the vector control, αSyn-expressing cells mostly exhibited a rather diffuse fluorescence signal throughout the cell, with some cells displaying only a weak vacuolar quinacrine staining (Figures 2A,B and Supplementary Figure S2A). A mutant completely defective in vacuolar acidification due to a lack of Vph2, the assembly factor of the vacuolar H⁺-ATPase (V-ATPase), an evolutionary conserved multimeric protein complex that shuttles protons into the vacuole (Li and Kane, 2009), completely lost any vacuolar quinacrine staining (Figure 2A). Thus, the expression of αSyn resulted in clear cytosolic acidification, while the pH of the vacuoles seemed only slightly affected, indicated by decreased vacuolar quinacrine intensity.

Next, we assessed a possible effect of αSyn on vacuolar morphology, since the fission-fusion equilibrium of the vacuole is also tightly connected to acidification processes of this organelle (Coonrod et al., 2013). Visualization of vacuolar membranes using either the fluorescence dye MDY-64 (Figures 2C,D) or the vacuolar membrane protein Vph1 fused to mCherry (Figures 2E,F) suggested an αSyn-mediated shift of the vacuolar fission-fusion equilibrium towards fusion, a process known to require an acidic pH of the vacuole (Ungermann et al., 1999; Coonrod et al., 2013). The percentage of cells harboring one single and enlarged vacuole increased upon αSyn expression, while cells displaying several small vacuoles decreased significantly. Exponentially growing WT cells typically contain 1–5 vacuoles of intermediate size. Progressing age or starvation has been shown to decrease the vacuolar surface-to-volume ratio (Li and Kane, 2009; Michaillat and Mayer, 2013), leading to one single, enlarged vacuole similar to the phenotype observed upon αSyn expression. To further elucidate whether this premature age-associated vacuolar shape upon αSyn expression might be due to an inhibition of fission or rather to enhanced fusion events, we monitored osmotically induced vacuolar fission upon treatment with 0.4 M NaCl (Michaillat and Mayer, 2013). Within minutes, numerous small vacuoles, visualized via Vph1-mCherry, were detectable (Supplementary Figure S2B). The presence of αSyn did not impair fission efficiency, arguing against an inhibition of this process as causal for observed changes in vacuolar morphology.

We further examined whether an enforcement of proton transport out of the cell or into the vacuolar lumen might counteract cytosolic acidification induced by αSyn. Therefore, we overexpressed components of the V-ATPase, which governs vacuolar acidification (Li and Kane, 2009), as well as the plasma membrane H⁺-ATPase Pma1, the major regulator of cytosolic pH (Martínez-Muñoz and Kane, 2008). While high levels of the V-ATPase-subunit Vma1 and the V-ATPase assembly factor Vph2 have been shown to be sufficient to increase vacuolar acidity (Hughes and Gottschling, 2012) and to extend replicative and chronological lifespan in yeast cells (Hughes and Gottschling, 2012; Ruckenstuhl et al., 2014), co-expression of these proteins did not prevent αSyn-mediated cytosolic acidification or cell death (Figures 2G,H and Supplementary Figure S2C). Furthermore, co-expression of Pma1 had no effect either. In sum, αSyn altered vacuolar proteolytic function and morphology, which was associated with a disruption of cytosolic pH homeostasis, but probably not with major changes in vacuolar acidification.

As high levels of Pep4 prevented cell death induced by αSyn, we evaluated the effects of this protease on αSyn-driven cytosolic acidification and vacuolar morphology after 16 h and 24 h of co-expression. Quinacrine staining revealed that indeed overexpression of Pep4^WT inhibited the αSyn-driven disturbances in pH homeostasis (Figures 3A,B and Supplementary Figure S3A). Again, this required the proteolytic activity of Pep4, as the inactive variant Pep4^DPM showed no effect (Supplementary Figure S3B). Moreover, the co-expression of Pep4 eliminated the effect of αSyn on vacuolar morphology observed after 16 h and restored WT organellar shape. The incidence of cells that displayed a small number (1–3) of large vacuoles, characteristic for αSyn expression, was largely reduced (Figures 3C,D). Upon entry into stationary phase, the WT vacuolar shape shifted to mostly one large vacuole per cell as expected, while Pep4-induced vacuolar fragmentation persisted (Supplementary Figure S3C). In aggregate, an increase of vacuolar hydrolytic capacity upon Pep4 overexpression was sufficient to restore cellular pH homeostasis, to correct changes in vacuolar shape and to inhibit cell death.

CatD has been shown to be critically involved in the degradation of αSyn (Sevlever et al., 2008), whose oligomers are believed to be the most toxic forms of this protein (Winner et al., 2011). Given that αSyn has already been shown to form aggregates in yeast cells (Outeiro and Lindquist, 2003; Zabrocki et al., 2005), we tested whether high levels of Pep4 would reduce the abundance of αSyn-positive protein inclusions. Microscopic analysis of an αSyn-GFP fusion protein revealed αSyn aggregates in about 40% of the cells after 24 h, while only 20% of the cells co-expressing Pep4^WT displayed GFP-positive foci (Figures 4A,B and Supplementary Figure S4A). Again, the co-expression of the proteolytically inactive Pep4^DPM as well as inhibition of Pep4^WT via pepstatin A completely abrogated this effect. Furthermore, immunoblot analysis demonstrated that co-expression of Pep4^WT caused a reduction of αSyn protein levels (Figures 4C,D). This most probably reflects enforced αSyn degradation rather than decreased αSyn expression, as: (i) administration of pepstatin A prevented this effect (Figures 4C,D); and (ii) the mRNA levels of αSyn were unaltered upon co-expression of Pep4^WT (Figure 4E). Next, we evaluated the effect of high Pep4 levels on oligomeric αSyn species using semi-native immunoblotting. We detected the αSyn monomer (approximately 18 kDa), a smaller fragment and several oligomeric species, whose levels were significantly reduced upon co-expression of Pep4^WT, while the proteolytically inactive variant Pep4^DPM had no effect (Figures 4F,G). During our experiments, samples for detection of αSyn oligomers appeared to be very sensitive to changes in temperature. Therefore, to verify the results obtained with semi-native immunoblotting, we applied in vivo crosslinking using 1% formaldehyde to stabilize αSyn oligomers. As expected, the cleavage product of αSyn (approximately 15 kDa), as well as oligomers with lower molecular weight were not visible with this approach, most likely due to crosslinking of these species with higher order αSyn forms (Figure 4H; a complete blot and quantifications are shown in Supplementary Figures S4B,C). However, bands with a molecular weight of about 100 and 120 kDa were detectable, which of course might not only represent pure αSyn oligomers but also stabilized interactions between αSyn and additional proteins (Figure 4H). Again, co-expression of Pep4^WT significantly reduced the levels of all observed αSyn species (Figure 4H and Supplementary Figures S4B,C).

There are several trafficking pathways that serve to deliver cellular material to the vacuole for subsequent proteolytic breakdown, including autophagy and the endocytic multivesicular body pathway. Microscopic analysis of αSyn-GFP cellular distribution and visualization of the vacuole with CellTracker Blue CMAC demonstrated that GFP-positive inclusions were indeed targeted to the vacuole (Figure 4I). As dysfunction of autophagy has been shown to contribute to αSyn cytotoxicity during PD (Winslow et al., 2010; Lynch-Day et al., 2012), we speculated that autophagy might be involved in the elimination of αSyn species via Pep4 and hence evaluated the requirement of a functional autophagic machinery for the observed cytoprotection. However, in strains lacking essential autophagy-related (ATG) genes, high levels of Pep4 still efficiently prevented αSyn-induced cell death (Figure 4J), indicating that mechanisms independent of autophagic key players are involved in the delivery of αSyn to the vacuole and subsequent Pep4-mediated degradation of αSyn species.

The type I transmembrane sorting receptor Pep1 (Vps10) shuttles between the trans-Golgi network and prevacuolar endosome-like structures in order to transport vacuolar hydrolases to the vacuole (Marcusson et al., 1994; Cooper and Stevens, 1996). Interestingly, mislocalization of the mannose 6-phosphate receptor MPR300, the human ortholog of Pep1 (Whyte and Munro, 2001), has recently been shown to contribute to a reduction of CatD protein levels in SH-SY5Y cells and mice expressing αSyn (Matrone et al., 2016). To investigate the role of Pep1 in our yeast model, we co-expressed αSyn and Pep4 in a strain devoid of Pep1. PI staining revealed that Pep4-mediated cytoprotection was abolished under these conditions (Figure 5A). Furthermore, Pep4 overexpression in Δpep1 cells did no longer enhance Pep4 enzymatic activity, despite similar expression levels (Figures 5B,C). Interestingly, while the artificially high dosage of Pep4 obviously required Pep1 for proper activation, the endogenous Pep4 activity, as observed in cells harboring the empty vector controls, was largely unaffected by the lack of Pep1. This might be due to alternative shuttling routes in the absence of Pep1. For instance, Vth1 and Vth2, two proteins with strong similarity to Pep1, have been shown to mediate sorting of Pep4 into the vacuole to some extent (Westphal et al., 1996). Furthermore, additional pathways independent of these proteins seem to exist (Westphal et al., 1996), indicating a complex network of different routes to ensure correct localization of Pep4. Analyzing the endogenous Pep1 protein levels upon αSyn expression via immunoblotting, we observed an αSyn-induced upregulation of Pep1 levels that increased over time (Figures 5D,E). This is in line with the increased protein levels of Pep4 (Figures 1C,D) and might reflect the same compensatory mechanism in an attempt to counteract the actual lack of Pep4 activity within the vacuole. However, overexpression of Pep1 did not reduce αSyn-induced cell death (Figures 5F,G). Thus, stimulating the pathway for Pep4 trafficking into the vacuole via high levels of its sorting receptor Pep1 failed to mimic the cytoprotective effects observed upon enforced expression of Pep4. Still, Pep1 function was essential for proper activation of overexpressed Pep4 and the subsequent cytoprotection against αSyn toxicity.

As Pep1 acts as a sorting receptor for soluble vacuolar hydrolases including Pep4, efficient recycling of Pep1 from the late endosomal/prevacuolar compartments back to the trans-Golgi network is required to allow further rounds of transport. Thus, we tested for a potential effect of αSyn on the cellular distribution of Pep1 using a Pep1-GFP fusion protein expressed endogenously under its own promoter. In WT control cells, Pep1-GFP was detectable in punctate structures throughout the cell (Figure 5H), probably corresponding to endosomal compartments and the Golgi as reported previously (Seaman et al., 1997; Nothwehr et al., 1999). Upon expression of αSyn, Pep1-GFP accumulated at the limiting vacuolar membrane (Figures 5H,I). The recycling of Pep1 back to the trans-Golgi network depends on the retromer complex, and lack of components of this complex, such as Vps29 and Vps35, has been shown to cause mislocalization of Pep1 at the vacuolar membrane (Horazdovsky et al., 1997; Seaman et al., 1997; Nothwehr et al., 1999). In sum, αSyn might impair proper recycling of Pep1, which in consequence would cause insufficient sorting of Pep4 to the vacuole.

Cytosolic acidification has been shown to stimulate the influx of Ca²⁺ (Burns et al., 1991), and increased cytosolic Ca²⁺ levels contribute to αSyn cytotoxicity (Büttner et al., 2013a). In addition, both deletion and overexpression of the Ca²⁺/calmodulin-dependent phosphatase calcineurin increased αSyn toxicity, suggesting that moderate levels of calcineurin activity are cytoprotective (Caraveo et al., 2014). Thus, we evaluated a possible role of calcineurin in the protective effects of Pep4 and co-expressed Pep4 and αSyn in strains either devoid of both its catalytic subunits (Cna1 and Cna2) or of its regulatory subunit (Cnb1). In these cells, the toxicity of αSyn was slightly increased and, most importantly, Pep4 completely lost its protective function (Figures 6A,B). In addition, we applied quinacrine staining to analyze potential effects of dysfunctional calcineurin signaling on the disturbances of cellular pH homeostasis caused by αSyn. While αSyn-driven cytosolic acidification was largely unaffected upon lack of Cnb1, Pep4 completely lost its ability to counteract this defect (Figures 6C,D). As these data suggest an essential role for calcineurin in Pep4-mediated cytoprotection, we screened for known calcineurin targets that might contribute to the observed effects (Goldman et al., 2014). However, Pep4 co-expression efficiently protected against αSyn-mediated cell death in all deletion mutants tested, including cells lacking the main calcineurin-activated transcription factor Crz1, distinct calcineurin-regulated Ca²⁺-transporters as well as numerous candidates identified to be dephosphorylated by calcineurin via phosphoproteomics (Goldman et al., 2014; Figure 6E). For paralogous pairs of calcineurin targets, double deletion mutants were analyzed.

To test for a more direct connection between calcineurin signaling and Pep4 function, we assessed the cellular distribution of a Pep4-GFP fusion protein in cells lacking Cnb1. Pep4 accumulated in prevacuolar compartments in a small portion of Δcnb1 cells (Figure 7A). In line with this partially impaired delivery of Pep4 to the vacuolar lumen, the lack of functional calcineurin led to a massive decrease of proteolytic Pep4 activity (Figure 7B). Monitoring the cellular distribution of the Pep1-GFP chimera in Δcnb1 cells, we observed a reduction in the number of Pep1-GFP positive endosomal structures (Figure 7C). In addition, a minor accumulation of this sorting receptor at the vacuolar membrane, indicative of inefficient recycling, was detectable. Immunoblotting demonstrated that the Pep1 protein levels were decreased upon lack of Cnb1 (Figures 7D,E). Furthermore, we detected enforced proteolytic clipping at the N-terminus of Pep1 in Δcnb1 cells (Figures 7E,F). This phenotype has been frequently observed in cells defective in retromer-dependent recycling or the multivesicular body pathway (Marcusson et al., 1994; Cooper and Stevens, 1996; Babst et al., 2002; Balderhaar et al., 2010). Thus, the lack of functional calcineurin impairs endosomal trafficking and vacuolar targeting, however only to an extent that does not impair growth (data not shown). Under standard growth conditions, calcineurin activity is low and the resting cytosolic Ca²⁺ concentration is not affected by inactivation of calcineurin (Supplementary Figure S5B; Forster and Kane, 2000; Cyert, 2003), indicating that the impediment of endosomal trafficking and sorting in Δcnb1 cells is most likely not due to elevated basal Ca²⁺ levels.

In line with hindered Pep1-mediated sorting of Pep4 to the vacuolar lumen upon lack of Cnb1, Pep4 overexpression in Δcnb1 cells resulted in insufficient enhancement of Pep4 activity compared to WT cells (Figure 8A). Consistently, Pep4-mediated reduction of αSyn oligomers was absent in Δcnb1 cells (Figure 8B, for quantification see Supplementary Figure S5A). Microscopic analysis of αSyn-GFP revealed that the lack of Cnb1 did not affect the abundance of αSyn aggregates per se, but prevented the decrease of these protein inclusions upon Pep4 co-expression (Figure 8C). In sum, these results indicate an essential role of calcineurin in maintaining proper Pep4 activity and in the subsequent protection against αSyn-inflicted cellular demise.