Main

The MIA machinery constitutes one of the essential protein translocation pathways into mitochondria and is responsible for the import of proteins targeted to the intermembrane space (IMS) of mitochondria1,2,3,4. RNA sequencing (RNA-seq) analysis of mia40-4int5,6 (a yeast temperature-sensitive mutant defective for protein import into the IMS) compared with wild-type cells was performed under permissive growth conditions to avoid cumulative consequences of broad mitochondrial dysfunction (Supplementary Table 1). Transcripts for cytosolic ribosomal proteins and proteins involved in translation, but not mitochondrial ribosomal proteins, decreased (Fig. 1a and Extended Data Fig. 1a–c). Thus, changes in the transcriptome could not be directly translated into protein abundance. Next, the cellular proteomes of mia40-4intS and wild-typeS cells were determined under permissive growth conditions using stable isotope labelling by amino acids in cell culture (SILAC)7 (Supplementary Table 2 and Extended Data Fig. 1d–g). Sensitive MIA substrates were significantly decreased in the mia40-4intS proteome (Supplementary Table 3 and Extended Data Fig. 1h–k) because they were not maintained upon mitochondrial import inhibition8.

Figure 1: The MIA pathway import efficiency regulates protein synthesis and proteasomal activity.
figure 1

a, b, Distribution of transcripts and proteins based on the GO term (black dots). c, d, Northern blot analysis of RNA and ratio of 25S/18S rRNAs in yeast grown in respiratory medium. Mean ± s.e.m., n = 3. e, Protein synthesis in cells grown in glycerol. f, Distribution of proteins based on GO terms (black dots). Dashed lines separate proteins of P value <0.05 and fold-change <−1.5 and >1.5. g, Ubiquitination analysis of cellular proteins. h, Proteasomal activity. Mean ± s.e.m., n = 7 (caspase-like activity), n = 4 (chymotrypsin-like activity). i, Cells were fractionated into total (T), post-mitochondrial supernatant (S) and mitochondria (M). Mia40↑, Mia40 overproducing strain. j, Proteasomal activity in cells grown at 24 °C. Mean ± s.e.m., n = 4 (caspase-like activity), n = 6 (chymotrypsin-like activity). c, h, j, *P < 0.05; **P < 0.03; ***P < 0.01. WT, wild-type. c–e, g, i, Uncropped blots are in Supplementary Information Fig. 1.

PowerPoint slide

Full size image

We determined the relative abundance of 2,564 proteins. A small fraction (11%) exhibited significant changes in abundance with 112 downregulated and 174 upregulated proteins in mia40-4intS (Supplementary Table 3 and Extended Data Fig. 1k). Localization and Gene Ontology (GO) analysis revealed the modulation of the structure and function of genetic information and reshaping of metabolic networks in mia40-4intS (Extended Data Fig. 2a, b). Consistent with the RNA-seq results (Fig. 1a), a class of proteins that constitute the cytosolic ribosome were downregulated (Fig. 1b and Extended Data Fig. 2b, c). Under restrictive growth conditions an imbalance in the ratio between 25S and 18S rRNA (Fig. 1c, d) and a change in ribosome content with less prominent polysomes in the mia40-4int mutants (Extended Data Fig. 2d–f) were observed. Interestingly, the mia40-4int mutant exhibited a decrease in protein synthesis (Fig. 1e), but extensive cellular death was excluded as a reason for this decrease (Extended Data Fig. 3a, b). The mia40-4int cells transformed with wild-type MIA40 exhibited a normal protein synthesis rate demonstrating that the translation defect depends on mia40-4int (Fig. 1e and Extended Data Fig. 3c).

The ubiquitin–proteasome system (UPS)9,10,11,12 is involved in the degradation of MIA substrate proteins8,13. Abundance of many proteins involved in the UPS did not change in mia40-4intS, including the core proteasomal subunits (Supplementary Table 3 and Extended Data Fig. 2d). Components of the proteasome assembly chaperone complex, Irc25 (also known as Poc3) and Poc4 (refs 14, 15, 16, 17), significantly increased at the protein level in mia40-4intS (Fig. 1f) despite non-significant changes at the mRNA level (Extended Data Fig. 3d). Other assembly chaperones, Ump1, Pba1 (also known as Poc1) and Add66 (also known as Poc2 or Pba2) (refs 14, 15, 16, 17), showed a tendency to increase, whereas the stress-inducible proteasome chaperone Adc17 (also known as Tma17) (ref. 18) did not change (Supplementary Table 3). Under restrictive growth conditions, ubiquitinated protein species decreased in mia40-4int and the effect was complemented by MIA40 (Fig. 1g). The abundance of ubiquitinated species may mirror translation efficiency because newly synthesized proteins constitute a substantial fraction of polyubiquitinated proteins in the cell19. However, the proteasomal inhibitor MG132 attenuated the difference in ubiquitination between wild-type and mutant cells, suggesting acceleration in mia40-4 protein turnover (Fig. 1g). The mia40-4int cells exhibited higher proteasomal activity than wild-type cells and this effect was rescued by MIA40 (Fig. 1h and Extended Data Fig. 4a–c). Applying restrictive temperatures to wild-type cells did not stimulate proteasomal activity (Extended Data Fig. 4d). Furthermore, other plasmid-borne conditional mia40 mutants5,6 with a phenotype milder than mia40-4int showed both a clear decrease in ubiquitinated species and an increase in proteasomal activity (Extended Data Fig. 4e, f).

Overexpression of Mia40 improved protein import into the mitochondrial IMS in human cells20 and in yeast as determined by an in organello assay (Extended Fig. 5a–c) but did not influence the abundance of mitochondrial proteins (Extended Data Fig. 5d). However, when the MIA pathway was challenged by the overproduced Flag fusion protein Pet191Flag8, this protein was more efficiently localized to mitochondria (Fig. 1i and Extended Data Fig. 5e, f). The more productive removal of precursors from the cytosol by mitochondria with elevated Mia40 levels resulted in decreased proteasomal activity (Fig. 1j).

The import of a large fraction of mitochondrial proteins relies on the electrochemical potential of the inner mitochondrial membrane (IM potential)1,2,3,4. Cells treated with the chemical uncoupler carbonyl cyanide m-chlorophenyl hydrazine (CCCP) to dissipate the IM potential accumulated non-imported mitochondrial proteins in their longer unprocessed precursor forms (with the mitochondrial targeting sequence, called presequences1,2,3,4). During the chase in the absence of CCCP, precursor proteins diminished (Fig. 2a and Extended Data Fig. 6a, b). Following the accumulation of mitochondrial precursors in the cytosol, a significant stimulation of the proteasome was observed (Fig. 2b). Similar accumulation of precursor proteins was observed in temperature-sensitive mutants of the TIM23 translocase and motor complex, pam 18, pam16 and tim17 (refs 21, 22) (Fig. 2c and Extended Data Fig. 6c). These precursor forms were found in the cytosol (Fig. 2d and Extended Data Fig. 6d). The pam18 and pam16 mutants did not show any major changes in protein abundance (Extended Data Fig. 6e). However, similar to mia40, the pam18, pam16 and tim17 mutants exhibited a clear decrease in ubiquitinated species and increase in proteasomal activity (Fig. 2e, f and Extended Data Fig. 6f–h). Although we did not detect any change in ribosomal proteins or ribosomal RNA content (Extended Data Fig. 6e, i), protein synthesis was decreased and was accompanied by an increase in monosomes in pam16 cells (Fig. 2g and Extended Data Fig. 6j, k). Thus, the defective import of proteins into mitochondria leads to a response that regulates the cellular protein homeostasis through two arms, the attenuation of the cytosolic synthesis of proteins and activation of the proteasome to prevent the accumulation of mistargeted proteins.

Figure 2: Defects in the presequence import pathway modulate protein synthesis and proteasomal activity.
figure 2

a, b, Cells were treated for 30 min with CCCP (pulse) and chased without CCCP and analysed for protein content and proteasomal activity. Mean ± s.e.m., n = 6. c, The pam18, pam16 and wild-type were analysed for protein content. d, The pam16-3 cells after a 90 min shift to 37 °C were fractionated. Total (T); post-mitochondrial supernatant (S); mitochondria (M); precursor (p); mature (m). e, Cells were analysed for ubiquitination. f, Proteasomal activity. Mean ± s.e.m., n = 3 (caspase-like activity), n = 5 (chymotrypsin-like activity). g, Protein synthesis in cells grown in glycerol. b, f, *P < 0.05; **P < 0.03; ***P < 0.01. WT, wild-type. a, c–e, g, Uncropped blots are in Supplementary Information Fig. 1.

PowerPoint slide

Full size image

To determine whether mistargeted mitochondrial proteins are responsible for proteasomal regulation, we used Pet191Flag and Mix17Flag, which were partly mislocalized to the cytosol when ectopically expressed8 but remained mainly soluble (Fig. 1i and Extended Data Figs 5f and 7a, b). Proteins that formed aggregates were previously shown to inhibit the proteasome23. The steady-state levels of cellular proteins, including ribosomal and proteasomal subunits, rRNA levels and cytosolic translation were unchanged upon the Pet191Flag and Mix17Flag overexpression (Extended Data Fig. 7c–e). However, in the presence of mistargeted Pet191Flag or Mix17Flag proteasomal activity increased substantially (Fig. 3a and Extended Data Fig. 7f). Thus, mitochondrial protein mislocalization uncoupled protein degradation and protein synthesis effects.

Figure 3: Mitochondrial precursor proteins stimulate proteasomal activity.
figure 3

a, Proteasomal activity upon expression of Pet191Flag or Mix17Flag at 24 °C. Mean ± s.e.m., n = 6. b, Proteasomal activity upon expression of mCox4Flag or Cox4Flag at 28 °C and mMdj1Flag or Mdj1Flag at 24 °C. Mean ± s.e.m., n = 5. c, Proteasomal activity upon Mix17Flag expression and CCCP treatment. Mean ± s.e.m, n = 3. d, Proteasomal activity in cells lacking Irc25 or Poc4 upon expression of Pet191Flag or Mix17Flag at 24 °C. Mean ± s.e.m, n = 3 (caspase-like activity), n = 4 (chymotrypsin-like activity). a–c, *P < 0.05; **P < 0.03; ***P < 0.01. WT, wild-type.

PowerPoint slide

Full size image

Other mitochondrial proteins, Cox4 and Mdj1 (substrates of the TIM23 pathway) and their non-imported variants that lacked presequences (mCox4 and mMdj1), were overexpressed (Extended Data Fig. 7g–i). The overexpression of Cox4 but not Mdj1, led to a visible accumulation of a cytosolic precursor form (Extended Data Fig. 7g, h). Cox4, mCox4, and mMdj1 (despite its low abundance) stimulated the proteasome, in contrast to fully imported Mdj1 which did not stimulate the proteasome (Fig. 3b). Non-mitochondrial model proteins, including mouse dihydrofolate reductase DHFR and its conformationally destabilized variant DHFRds, temperature sensitive Ubc9ts–GFP24 and native peroxisomal protein Pex22, were overexpressed and remained soluble (Extended Data Fig. 8a–c). They did not significantly affect proteasome activity (Extended Data Fig. 8d–g). To better mimic the stress caused by delayed mitochondrial import, we induced Mix17 in the presence of CCCP. The stimulatory effects of import inhibition and Mix17 overexpression on proteasome activity were additive (Fig. 3c and Extended Data Fig. 8h). Thus, proteins that were delayed in their transport to mitochondria—because of overexpression or the absence of mitochondrial targeting signals or chemical import inhibition—stimulated proteasomal activity.

No proteasomal activity stimulation by Mix17 was observed in cells lacking Irc25 or Poc4 (Fig. 3d and Extended Data Fig. 9a, b). Indeed, proteasomal subunits were not upregulated (in contrast to Irc25 and Poc4; Fig. 1f and Supplementary Table 3) and the purified proteasome was not stimulated in vitro by Pet191Flag or Mix17Flag (Extended Data Fig. 9c). The 26S proteasome is assembled from components, including the 20S catalytic core and the 19S regulatory particle10,11,12. Scl1TAP (core subunit) and Rpn6TAP (regulatory subunit) were used to purify the 20S core (represented by Pre10) and the 26S proteasome that comprised the core and regulatory particles (represented by Rpt1 and Rpt5) (Fig. 4a, b and Extended Data Fig. 9d). The proteasomal subunits were eluted more efficiently with Scl1TAP and Rpn6TAP upon Mix17 overexpression (Fig. 4c, d). Thus, the stimulatory effect was associated with proteasomal assembly.

Figure 4: Mistargeted proteins protect cells against stress.
figure 4

a, b, Affinity purification via Scl1TAP or Rpn6TAP. c, d, Affinity purification via Scl1TAP or Rpn6TAP upon 6 h induction of Mix17Flag at 24 °C. Load, 5%; eluate, 100%. The empty vector eluate was set to 1. Mean ± s.e.m, n = 4 (c), n = 5 (d). e, f, Lethality upon overexpression of mitochondrial proteins. Mean ± s.e.m, n = 4 (e), n = 3 (f). Mean lethality values of cells with empty vector: wild type, 19% (e) or wild type, 9.54%; Δirc25, 9.01%; Δpoc4, 7.13% (f), were set to 1. g, Cells were subjected to consecutive tenfold dilutions. h, Lethality of cells. Mean ± s.e.m, n = 3. Mean lethality values of cells with empty plasmid: Δpoc4, 5.33%; Δpoc1, 11.02%; Δpoc2, 11.68%; Δump1, 9.91%; Δpre9, 10.74%, were set to 1. c–e, f, h, *P < 0.05; **P < 0.03, ***P < 0.01. WT, wild-type. a–d, Uncropped blots are in Supplementary Information Fig. 1.

PowerPoint slide

Full size image

Interestingly, heat-shock conditions revealed a gain of stress resistance upon mitochondrial precursor appearance (Fig. 4e, f and Extended Data Fig. 10a). In the cells without Irc25 or Poc4, Mix17Flag and Cox12Flag exerted toxicity under temperature-stress conditions (Fig. 4f). The induction of Pet191Flag, Mix17Flag and Cox12Flag but not DHFRFlag in these cells led to a synthetic growth defect (Fig. 4g and Extended Data Fig. 10b). Cells without Poc1, Poc2, Ump1 or the proteasomal component Pre9 did not exhibit an increase in lethality (Fig. 4h). Thus, the response that was activated by protein mistargeting relied on proteasome assembly governed by the Irc25–Poc4 complex.

The mitochondrial protein import slowdown and precursor over-accumulation stress triggers processes that maintain cellular proteostasis. This includes the inhibition of protein synthesis and activation of the proteasome that originates from mistargeted proteins (Extended Data Fig. 10c). This unfolded protein response activated by protein mistargeting (UPRam) exerts a protective effect on the cells. UPRam differs from UPRmt, another response pathway to compromised mitochondrial function present in higher Eukaryotes, because UPRmt acts by regulating the abundance of chaperones and proteases inside mitochondria25,26,27. Proteasomal activation and mild mitochondrial dysfunction have both been strongly implicated in the positive regulation of longevity28,29,30. The newly identified UPRam causally links defects in mitochondrial biogenesis with proteasomal activity, thus providing a possible explanation for the beneficial effects of mild mitochondrial dysfunction on ageing.

Methods

Data reporting

No statistical methods were used to predetermine sample size. The experiments were not randomized. The investigators were not blinded to allocation during experiments and outcome assessment.

Yeast strains

Yeast Saccharomyces cerevisiae strains are derivatives of YPH499 (MATa, ade2-101, his3-Δ200, leu2-Δ1, ura3-52, trp1-Δ63, lys2-801) or BY4741 (MATa, his3Δ1; leu2Δ0; met15Δ0; ura3Δ0). The temperature-sensitive mia40-4int (Fomp2-7int; 305), mia40-4 (YPH-fomp2-7; 176), mia40-3 (YPH-BG-fomp2-8; 178), mia40-F311E (660) and corresponding wild-type (398) strains were described previously5,6,8. The mia40-4int was generated by integration of the plasmid-borne mia40-4 allele into the genome6. This mutant showed a stronger decrease in the accumulation of intermembrane space proteins than mia40-4 (refs 5, 6, 8), but due to a single genomic copy of mia40 was more suitable for SILAC. The SILAC strains are arg8::kanMX4 derivatives of YPH499 (524) and mia40-4int (305) and are referred as to WTS (654) and mia40-4intS (558; a gift from Bernard Guiard). The pam16-1 (YPH-BG-mia1-1; 733), pam16-3 (YPH-BG-mia1-3; 734), pam18-1 (YPH-BG-Mdj3-66; 739), tim17-4 (YPH-BG17-9d; 83), tim17-5 (YPH-BG17-21-7; 85) and corresponding wild-type (738, 524) strains were described previously21,22,31. Strains expressing PRE3 (843) or RPN6 (905) with a TAP tag and deletion strains of PRE9, POC1, POC2, IRC25, POC4, UMP1 in the BY4741 genetic background were purchased from Euroscarf. The strain expressing SCL1 with TAP tag was a gift from Michael Glickman (916). To rescue mia40-4int, the wild-type and mia40-4int yeast cells were transformed with a plasmid containing MIA40 under an endogenous promoter (pGB8220; 85p). To overproduce Mia40, yeast cells were transformed with the TRP1-containing plasmid (pAC8-3, 97p) harbouring MIA40 (YKL195W) under control of the GAL10 promoter. The plasmid for Ubc9ts–GFP overproduction was purchased from Addgene (314p)24. To express PET191 (YJR034W), MIX17 (YMR002W), COX12 (YLR038C), POC1 (YLR199C), POC2 (YKL206C), IRC25 (YLR021W), POC4 (YPL144W), COX4 (YGL187C), MDJ1 (YFL016C), PEX22 (YAL055W) and DHFR encoding mouse dihydrofolate reductase or its destabilized version DHFRds in frame with the FLAG tag under control of the GAL10 promoter, yeast cells were transformed with URA3- or LEU2-containing pAG1 (53p), pAG2 (54p), pEJ1 (334p), pAG3 (55p), pPCh6 (320p), pPCh1 (315p), pMS126 (151p), pMS127 (152p), pMS120 (145p), pMS121 (146p), pUT9 (333p), pMS128 (162p), pMS119 (144p) plasmids. Mix17 (YMR002W) was previously known as Mic17 (refs 8, 32, 33). The mMdj1 and mCox4 constructs correspond to 17–511 and 25–155 amino acid residues of Mdj1 and Cox4, respectively. The sequence for DHFRds contains three destabilizing amino acid substitutions C7S, S42C, N49C (gift from Wolfgang Voos34,35).

Growth conditions

The strains were grown on minimal synthetic medium (0.67% (w/v) yeast nitrogen base, 0.079% (w/v) CSM amino acid mix) containing respiratory carbon source (3% (v/v) glycerol supplemented with 0.05 – 0.2% (v/v) glucose) or fermentable carbon source (2% (v/v) sucrose, 2% (v/v) galactose, 2% (v/v) glucose). To induce the GAL10 promoter for the expression of MIA40, cells were grown on minimal selective medium with 1.5% (v/v) glycerol supplemented with 0.5% (v/v) galactose. To induce the GAL10 promoter for the expression of other proteins, cells were grown on minimal selective medium with 2% (v/v) galactose. In the experiments in which Mia40 and Pet191 were simultaneously overproduced, cells were grown on minimal selective medium with 2% (v/v) galactose. The yeast strains were usually grown at permissive temperature (19 °C or 24 °C) to the early logarithmic growth phase and analysed before or after a shift to a restrictive temperature (37 °C).

RNA-seq

Wild-type and mia40-4int cells were grown at 19 °C in YPG (1% (w/v) yeast extract, 2% (w/v) peptone, 3% (v/v) glycerol) medium to the mid-logarithmic phase (D600nm (OD600) of ∼0.65) in three biological replicates per genotype. RNA was extracted with MagNALyser/MagNA Pure Compact RNA Isolation Kit (Roche) from frozen yeast preserved with RNAlater (Ambion). DNA was digested with TURBO DNA-free Kit (Life Technologies). The RNA quality and concentration were tested with Agilent 2100 Bioanalyzer and Qubit RNA BR Assay (Life Technologies). RNA was spiked with ERCC RNA Spike-In Mix (Life Technologies) and rRNA depleted with Ribo-Zero Gold rRNA Removal Kit (Illumina). RNA-seq libraries were prepared with Ion Total RNA-Seq Kit v2 (Life technologies) with 2 + 10 cycles amplification. Equimolar pool of barcoded libraries was used to prepare sequencing template with Ion PI Template OT2 200 Kit v3 (Life Technologies), followed by sequencing with Ion Proton on Ion PI v2 chip (Life Technologies). The reads were adaptor trimmed and aligned using Torrent Suite and Partek Flow as recommended by Life Technologies. The reads were quality trimmed by Torrent Suite and resulting files were aligned to the Ensembl S. cerevisiae R64 reference genome (GCA_000146045.2) by TopHat2 (ref. 36). Not aligned reads were submitted to the Bowtie2 (ref. 37) alignment with the ‘very sensitive’ setting. Alignments were combined and coordinate sorted with SAMtools38. Differential expression analysis was performed with Partek Genomics Suite (Partek). Transcripts were quantified to the Ensembl R64 GTF annotation file and resulting RPKM values were used for statistical analysis in Partek ANOVA (General Linear Model).

SILAC labelling and LC/MS sample preparation

Yeast strains of WTS and mia40-4intS were grown at 19 °C on respiratory minimal liquid medium (0.67% (w/v) yeast nitrogen base, 0.068% (w/v) CSM amino acid mix lacking arginine and lysine, 3% (v/v) glycerol, 0.02% (w/v) l-proline) supplemented with 18.6 mg l−1 [13C6/15N4] l-arginine and 19.35 mg l−1 [13C6/15N2] l-lysine (Sigma-Aldrich) or non-labelled l-arginine and l-lysine. Three separate cultures were grown and differentially labelled using SILAC amino acids including a label-switch. Following mixing of differentially SILAC-labelled WTS and mia40-4intS cells and cell lysis, proteins were acetone-precipitated and analysed using a gel-free and a gel-based approach. For the gel-free approach, proteins were redissolved in 60% MeOH/20 mM NH4HCO3 (pH 7.8) followed by reduction and alkylation of cysteines and tryptic digest as described39. For the gel-based approach, proteins were resuspended in urea buffer (7 M urea/2 M thiourea/30 mM Tris-HCl, pH 8.5) and separated (15 µg protein per replicate) on a 4 – 12% NuPAGE Bis-Tris gel. Lanes were cut into 15 slices, destained and washed as reported40 followed by reduction and alkylation of cysteines39. Proteins were in-gel digested with trypsin and peptides reconstituted in 0.1% TFA.

LC/MS analysis

Peptides obtained by in-solution digestion were analysed in triplicates per biological replicate using an RSLCnano/LTQ-Orbitrap XL system (Thermo Scientific, Bremen, Germany) as described39 with the modification that peptides were separated with a 150-min gradient (4 – 35% acetonitrile in 0.1% formic acid). Peptides derived from the gel were analysed on an RSLCnano/Orbitrap Elite system using a 45-min gradient and parameters as follows: survey scan range, m/z 370–1,700; resolution of 120,000 (m/z 400); top15 method (CID); automatic gain control and maximum fill time, 1 × 106 and 200 ms for the orbitrap, 5 × 103 and 150 ms for the linear ion trap; normalized collision energy, 35%; activation q, 0.25; activation time, 10 ms; dynamic exclusion, 45 s.

MS data analysis

Data were processed using MaxQuant/Andromeda (version 1.3.0.5)41,42 and searched against the Saccharomyces Genome Database (http://www.yeastgenome.org). LC/MS data from mitochondrial samples were included to support peptide identifications. The database search was performed with default settings including deamination of asparagine and Pro6 as variable modifications. Protein identifications were based on at least one unique peptide. Mass tolerances for precursor and fragment ions were 6 p.p.m. and 0.5 Da, respectively. A false discovery rate of 1% was applied on peptide and protein level. SILAC ratios were calculated based on unique peptides using MaxQuant default settings and “match between runs”. SILAC ratios for proteins were log-transformed and mean log10 ratios across all replicates and P values (two-sided t-test) were determined. Protein groups exhibiting a posterior error probability (PEP) < 0.01, a P value <0.05 and a minimum fold change ±1.5 were classified as up- or downregulated in mia40-4intS cells. Variances of the mean log10 SILAC ratios were consistent across all three biological replicates yielding uniform data. Sample size, data analysis approach and the statistical test chosen fulfill the requirements for quantitative proteome analyses employing similar LC/MS instrumentation and methods. Statistical methods to determine sample size were not applied. The acquired data were normally distributed, thus meeting the assumption of the statistical test. All raw data and original MaxQuant result files have been deposited to the ProteomeXchange Consortium43 via the PRIDE partner repository44 with the data set identifier PXD001495. GO term enrichment analysis of proteins classified as upregulated or downregulated was performed using the BiNGO45 plug-in 2.44 for Cytoscape 2.8 (ref. 46) and the yeast genome as background. Using the built-in Benjamini–Hochberg procedure, raw P values calculated for GO enrichment analysis were corrected for multiple testing. GO terms with a corrected P value <0.05 were considered enriched.

Proteasome activity assays

Proteasome activity assays were carried out as previously described47 with few modifications. 20 D600 nm (OD600) units of cells were shaken in lysis buffer (50 mM Tris-HCl pH 7.4, 10 mM MgCl2, 10% glycerol, 2 mM ATP, 2 mM PMSF, 1 mM DTT) with glass beads (Sigma-Aldrich) for 10–20 min at 4 °C. After a clarifying spin (20,000g, 15 min, 4 °C), protein concentration was determined by Bradford assay with bovine serum albumin as standard. Activity assays were performed in a final volume of 200 µl of lysis buffer with 50 µg of soluble total protein extracts in a 96-well plate by adding 100 µM Suc–Leu–Leu–Val–Tyr–AMC peptide substrate (chymotrypsin-like activity; Bachem, I-1395) or 100 µM Ac–NIe–Pro–NIe–Asp–AMC (caspase-like activity; Bachem, I-1850). Fluorescence (excitation wavelength 380 nm, emission wavelength 460 nm) was measured every 5 min for 1–2 h at 25 °C using a microplate fluorometer (Infinite M1000, Tecan). For in vivo inhibition of proteasome activity, cells were grown as described previously8. Cultures were supplemented with either 75 µM MG132 or DMSO.

Stress assays

Cells were grown to the early logarithmic phase and expression of proteins was induced for 6 h at 24 °C in minimal medium (0.67% yeast nitrogen base, 0.077% amino acid supplement mixture) containing 2% (v/v) galactose. Only cultures with similar D600nm (OD600) were compared for their viability after heat shock. One D600nm (OD600) unit of yeast cells was washed with PBS (137 mM NaCl, 12 mM phosphate, 2.7 mM KCl, pH 7.4). Cells were stained with 3 µg ml−1 propidium iodide in PBS for 15 min at room temperature in dark and analysed within 1 h by flow cytometry (FACSCalibur). 10,000 cells for each sample were measured.

Mitochondria procedures and cell fractionation

Yeast were usually grown on YPG medium. The yeast with overproduced Mia40 were grown on YPGal (1% (w/v) yeast extract, 2% (w/v) peptone, 2% (w/v) galactose). Mitochondria were isolated by differential centrifugation according to the standard method48. Mitochondria were solubilized in Laemmli buffer with 50 mM DTT (reducing) or 50 mM iodoacetamide (non-reducing conditions), and analysed by SDS–PAGE followed by western blotting. [35S]methionine radiolabelled precursors were prepared by in vitro transcription (mMESSAGE mMACHINE, Ambion) and translation (rabbit reticulocyte lysate system, Promega) and imported into the isolated mitochondria as previously described6. Import experiments were performed in at least two different mitochondrial isolations (biological replicates) and repeated at least three times (technical replicates). For fractionation, yeast were grown on minimal medium with fermentable carbon source (2% glucose) to D600nm (OD600) 0.5, induced for 6 h or overnight by 2% galactose for overexpression of MIA40 and 35 D600nm (OD600) units were fractionated by differential centrifugation8.

Mitochondrial precursor protein analysis

Yeast cells were grown in YPGal with addition of 0.2% glucose at 19 °C. Upon consumption of glucose, yeast were grown for additional 2–3 h to D600nm (OD600) 1.5–3. Cultures were then supplemented with either 40 µM CCCP or DMSO. After 1.5 h, samples were analysed by SDS–PAGE and western blotting. For pulse-chase experiments, upon consumption of glucose, yeast were grown for an additional 1 h to D600nm (OD600) 1.1–1.3. Cultures were then supplemented with either 10 µM CCCP or DMSO for the pulse. For pulse-chase experiments with CCCP in combination with expression of Mix17Flag cells were grown at 24 °C in minimal selective medium containing 2% (v/v) sucrose to logarithmic phase. Cells were washed and resuspended in YPGal medium and grown for 4 h at 24 °C. Cells were shifted for 15 min at 19 °C before supplementing the culture with either 10 µM CCCP or DMSO and samples were chased at 19 °C.

Aggregation assay

Cells were grown for 4 h at 28 °C with either 75 µM MG132 or the corresponding volume of DMSO8. Then 10 D600nm (OD600) units of cells were shaken in lysis buffer (30 mM Tris-HCl pH 7.4, 20 mM KCl, 5 mM EDTA, 0.5 mM PMSF, 0.5% Trition X-100) with glass beads for 10 min at 4 °C. After clarifying centrifugation (4,000g for 10 min, 4 °C), total (T) samples were separated into pellet (P) of aggregated proteins and supernatant (S) by ultracentrifugation for 1 h at 125,000g. Total and supernatant fractions were precipitated by 10% trichloroacetic acid.

Affinity purifications

For purification of Flag-tagged proteins, cells were grown at 28 °C in 3% (v/v) glycerol minimal medium with 0.5% (v/v) galactose. Cells (100U D600nm (OD600)) were shaken in buffer A (20 mM Tris-HCl pH 7.4, 300 mM NaCl, 2 mM PMSF) with glass beads for 20 min at 4 °C. The solution was supplemented with 1% (w/v) digitonin and incubated for 15 min on ice. After clarifying centrifugation (20,000g, 15 min, 4 °C), supernatants were incubated with anti-Flag M2 affinity gel (Sigma-Aldrich) at 4 °C for 1 h. The column was washed with buffer A, followed by elution with 0.1 M glycine pH 3.5 for 5 min at room temperature. The pH of the elution fraction was adjusted to pH 7.4.

For proteasome complex purification, 100 D600nm (OD600) units of cells were shaken in buffer B (50 mM Tris-HCl pH 7.4, 10 mM MgCl2, 10% glycerol, 2 mM ATP, 2 mM PMSF, 1 mM DTT) with glass beads for 10 min at 4 °C. After clarifying centrifugation (20,000g, 15 min, 4 °C) supernatants were incubated with immunoglobulin G Sepharose (GE Healthcare) for 1 h at 4 °C. The column was washed and resuspended in 100 µl of buffer B (for purified proteasome activity assay) or proteins were eluted by Laemmli buffer with 50 mM DTT. Proteasome components co-purified with TAP-tagged proteins were quantified and the ratio of signals from cells with Mix17 versus empty plasmid was calculated and normalized to the efficiency of Scl1TAP or Rpn6TAP recovery in the eluate.

Ribosome profile analysis

Yeast were grown on YPGal to early logarithmic phase and shifted to 37 °C. Cells were disrupted in lysis buffer (40 mM Tris-HCl pH 7.4, 100 mM NaCl, 30 mM MgCl2, 100 µg ml−1 cycloheximide, 300 µg ml−1 heparin, 1 mM DTT, 0.5 mM PMSF, 2 µM pepstatin A, 2 µg ml−1 chymostatin, 1 µM leupeptin, 2 mM benzamidie HCl and 5 µM chymostatin) by shaking with glass beads followed by Triton X-100 addition to 0.2% and incubation for 15 min on ice. After clarifying spin (20,000g, 10 min, 4 °C), samples were separated on 10–50% (w/v) sucrose gradients in polysome buffer (20 mM Tris-HCl pH 7.4, 100 mM NaCl, 5 mM MgCl2, 100 µg ml−1 cycloheximide, 2 mM DTT; 222,000g, 2 h, 4 °C). Gradients were fractionated with a continuous absorbance measurement at 254 nm. The peak areas (volume (ml) × absorbance (mAU)) were quantified with Unicorn 5.2 (GE Healthcare).

Northern blot and qPCR assays

Total RNA was extracted from cells as described49. For northern blot 5 μg of RNA was separated on agarose or acrylamide denaturing gels, transferred to Hybond N+ (GE Healthcare) membrane and immobilized by the UVP CL-1000 crosslinker. Oligonucleotides used for RNA hybridizations (25S rRNA (007) 5′-CTCCGCTTATTGATATGC; 18S rRNA (008) 5′-CATGGCTTAATCTTTGAGAC; 5.8S rRNA (017) 5′-GCGTTGTTCATCGATGC; 5S rRNA 5′-CTACTCGGTCAGGCTC) were labelled with [γ-32P]ATP using T4 polynucleotide kinase (New England Biolabs). All northern blot steps were performed using PerfectHyb Plus Hybridization Buffer (Sigma-Aldrich). Radioactive signals were detected by digital autoradiography. For qPCR analysis RNA extracts were treated with TURBO DNase (Ambion) and re-isolated. 1 µg of RNA was reverse transcribed using SuperScript III First-Strand Synthesis SuperMix (Invitrogen) with 2.5 µM oligo(dT)20 and 1.8 µM random hexamers as primers. Quantitative PCR reactions were performed using 7900HT Real-Time PCR (LifeTechnologies) and SensiFAST SYBR Hi-ROX Kit (Bioline) with standard protocol in triplicates. Standard curves were generated by the serial dilution of the pooled cDNA. To normalize nonspecific variations, the normalization factor was calculated as the geometric mean of transcript levels of 3 control genes, ALG9, FBA1 and TUB2 that were selected on the basis of stability in transcriptomic and proteomic data sets and published observations50,51,52. Primers used for qPCR: ALG9 5′-CACGGATAGTGGCTTTGGTGAACAATTAC, 5′-TATGATTATCTGGCAGCAGGAAAGAACTTGGG; FBA1 5′-ACGAAGGTCAAAATGCTTCCAT, 5′-TGGCACAGTGGTCAGAGTGTAAG; IRC25 5′-CACTGGCAGGAAACCTAGGAAA, 5′-CTCCCAACTTGGTAACCACTTGA; POC4 5′-CTCTACCTGCAGACGATCGTTCT, 5′-GGGAATGGCGTAGTAATAACATGA; TUB2 5′-ATCTTGTCCCATTCCCACGTT, 5′-TTGCTGTGTTAATTCAGGGACAGT.

Statistical analysis

For statistical analysis, a two-tailed, paired t-test was used. For statistical analysis of SILAC, polysome profiles, northern blots and qPCR, a two-tailed, unpaired t-test, assuming unequal variance, was performed. A P value <0.05 was considered significant. Proteasomal activity was calculated as a slope value and data are represented in a fold change compared to wild-type strain or wild-type containing the empty plasmid (with the exception for the pulse-chase experiments, in which the CCCP-treated samples were compared to the DMSO-treated control). Represented fold changes are the mean of fold changes obtained from independent biological replicates ± standard error of the mean (s.e.m.). ImageQuant was used for the quantification of western blotting results. The sample size used was estimated based on previous experiments and statistical tests were chosen based on published data with comparable methodology. Experiments have been replicated at least three times.

Other details

Volcano plot distribution of transcripts and proteins was based on the GO Slim terms provided by the Saccharomyces genome database (SGD). Protein extracts were prepared from amount of cells corresponding to 0.2–0.4 D600nm (OD600) units by alkaline lysis53 and resuspended in Laemmli buffer containing 50 mM DTT. To assess protein synthesis, cells were grown in Met-free minimal medium before incubating with [35S]-labelled methionine and cysteine (EXPRESS; Perkin Elmer) that were added to a final concentration of 6 µCi ml−1 for 20 min to 2 h. The samples were analysed by SDS–PAGE and digital autoradiography. Proteins were separated by SDS–PAGE on 15%, 12% or 10% gels. Primary antibodies were custom-raised in rabbits and individually controlled for specificity. Commercially available antibodies for mouse monoclonal anti-ubiquitin (SC-8017, Santa Cruz), anti-Flag M2 (F1804, Sigma-Aldrich), rabbit polyclonal anti-Rpt5 (PW8245, Enzo Life Science), anti-Rpt1 (PW8255, Enzo Life Science), anti-alpha7/Pre10 (PW8110, Enzo Life Science) and anti-Hsp104 (ADI SPA 1040, Enzo Life Science) were used. Chemiluminescence protein signals were detected by use of X-ray films. Autoradiography signals were processed with ImageQuant (GE Healthcare). The images were processed digitally using Adobe Photoshop CS4. In some figures, non-relevant gel parts were excised digitally. Protein concentration was determined using Roti-Quant (Carl Roth GmbH) with bovine serum albumin as a standard. The nomenclature of proteins is according to the SGD.