The chaperone-binding activity of the mitochondrial surface receptor Tom70 protects the cytosol against mitoprotein-induced stress

Abstract

Most mitochondrial proteins are synthesized as precursors in the cytosol and post-translationally transported into mitochondria. The mitochondrial surface protein Tom70 acts at the interface of the cytosol and mitochondria. In vitro import experiments identified Tom70 as targeting receptor, particularly for hydrophobic carriers. Using in vivo methods and high-content screens, we revisit the question of Tom70 function and considerably expand the set of Tom70-dependent mitochondrial proteins. We demonstrate that the crucial activity of Tom70 is its ability to recruit cytosolic chaperones to the outer membrane. Indeed, tethering an unrelated chaperone-binding domain onto the mitochondrial surface complements most of the defects caused by Tom70 deletion. Tom70-mediated chaperone recruitment reduces the proteotoxicity of mitochondrial precursor proteins, particularly of hydrophobic inner membrane proteins. Thus, our work suggests that the predominant function of Tom70 is to tether cytosolic chaperones to the outer mitochondrial membrane, rather than to serve as a mitochondrion-specifying targeting receptor.

Keywords: Tom70; chaperones; mitochondria; outer membrane; protein translocation; proteostasis; prototoxicity.

Graphical Abstract

Figure 1. Identification of Tom70/71 clients

(A and B) Radiolabeled Atp1, Hsp60, Pet9, and Oac1 were incubated with isolated wild-type and Δtom70/71 mitochondria for the times indicated at 25°C. Non-imported protein was removed by treatment with proteinase K, and samples were analyzed by SDS-PAGE and autoradiography. Graphs show mean values and standard deviations from three independent experiments. (C) The proteomes of different mutants were compared using quantitative proteomics and multiplexing (see also Figure S1A). (D–G) The proteomes of Δtom70/71 cells carrying either empty or Tom70-expressing plasmids (three biological replicates each) were measured by mass spectrometry. Shown are the mean values of the ratios obtained from Δtom70/71 (30°C) to Tom70-expressing cells (30°C) plotted against their statistical significances (p values). The points in the top left corner show the highest Tom70 dependence. The data point for Tom70 is shown in Figure S1B. Different groups of proteins are indicated in the same dataset. IM, inner membrane. (H) The relative depletion of proteins in the Δtom70/71 to Tom70 comparison (log₂ fold changes [FCs]) were taken as proxy for the Tom70 dependence of proteins. Shown are the distributions of these Tom70 dependence values for different groups of mitochondrial proteins (Morgenstern et al., 2017).

Figure 2. Mitochondrial proteins strongly differ in their Tom70 dependence

(A) Scheme of the systematic visual screen of GFP-tagged mitochondrial proteins. (B–D) The mitochondrial localization of 113 N-terminally GFP-tagged mitochondrial proteins (all lacking an MTS) were visualized. Proteins shown in (B) showed a strongly reduced mitochondrial localization in the absence of Tom70 and moderately reduced levels if Tom71 was deleted. Thus, these proteins depend to some degree on both receptors. Proteins shown in (C) were unaffected if Tom71 was deleted but still required Tom70. For proteins shown in (D), Tom70 and Tom71 were hardly, if at all, relevant. (E) The whole-cell GFP signal change in Δtom70Δtom71 compared with wild-type cells measured for different mitochondrial protein classes. See Table S3 for details. Scale bars, 10 μm. OM, outer membrane.

Figure 3. Tom70/71 supports biogenesis of aggregation-prone mitochondrial proteins

(A) The aggregation propensities (Conchillo-Sole et al., 2007) and the presence of iMTS-L sequences in proteins (Boos et al., 2018) were calculated. Plotted are the distributions of these values for Tom70-dependent (log₂ FC, <—0.2) and -independent (log2 FC, >0.2) proteins. (B) Aggregation propensities were calculated for different groups of mitochondrial proteins. The dotted line shows the mean value of Tom70-independent proteins as a reference. (C) The indicated strains were precultured in galactose-containing medium at 30°C and spotted on galactose medium, following 3 days of incubation at 30°C, 34°C, or 37°C. WT, wild-type; ev, empty vector. (D and E) The influence of temperature (log₂ FC of Tom70 37°C as compared to Tom70 30°C) and the absence of Tom70 (log₂ FC of Δtom70/71 as compared to Tom70 at 30°C) were analyzed. Blue circles show the isobaric distribution of mitochondrial proteins, whereas black ones show the distribution of the entire proteome. Enrichment of mitochondrial proteins among proteins with a log₂ FC below a certain threshold was calculated, and significance of this enrichment was plotted (side panels).

Figure 4. Tom70 can be replaced by a chaperone tether on the mitochondrial surface

(A) Schematic representation of the different domains of Tom70 formed by 11 TPR domains. (B and C) The indicated sequences of yeast TPR proteins were fused to the membrane anchor of Tom70 and expressed in the Δtom70/71 mutant. (D) A Δsam37 Δtom70 double mutant carrying SAM37 on a URA3-containing plasmid was transformed with plasmids for the expression of the indicated fusion proteins. Upon addition of 5-fluoroorotic acid (5°FOA), only cells that lost the URA3-containing SAM37 plasmid could grow. (E) Cells of the Δtom70/71 mutant carrying the mt-Tah1 expression or an empty plasmid were grown in galactose medium to mid-log phase. Cells were washed, gently lysed with Triton X-100, and incubated with Sepharose beads carrying HA-specific antibodies (to pull out mt-Tah1). Samples from four independent replicates for each strain were analyzed by mass spectrometry. The full dataset can be found in Table S4. (F) Radiolabeled Atp1 was incubated with mitochondria isolated from the indicated mutants. Non-imported Atp1 was removed by adding proteinase K after the times indicated. Mt-Tah1-C2-C3 is a fusion protein in which the C2 and C3 domains of Tom70 were fused to mt-Tah1. (G) Radiolabeled Atp1 was incubated with mitochondria after the membrane potential was depleted by treatment with carbonyl cyanide m-chlorophenyl hydrazone (CCCP). When indicated, CCCP was quenched by dithiothreitol (DTT) to restore the membrane potential. The presence of the C2-C3 domains was essential to keep Atp1 bound to the mitochondria (indicated by the red arrow). (H) Model of the chaperone binding property of the C1 domain of Tom70/71 and of mt-Tah1. The C2 and C3 domains facilitate direct substrate binding that is particularly relevant under the conditions of the in vitro import reaction.

Figure 5. Chaperone binding by Tom70 is important for different cellular activities

(A) The Δtom70 allele was introduced into a systematic yeast deletion library by automated genetic manipulations. Colony sizes were measured, and the 100 most-affected deletion mutants (Table S5) were analyzed (see STAR Methods for details). (B) Schematic illustration of the CRISPRi strategy used to knock down TOM70. (C) TOM70 transcript levels were measured by qPCR 6.5 h after addition of anhydrotetracyclin (ATc). Shown are mean values of three replicates. (D) Tom70 levels were analyzed by western blotting of the indicated strains at different time points after addition of ATc. (E) Growth curves of the following strains: indicated single deletions without addition of ATc (mock); TOM70 is knocked down through the addition of 960 ng/μl ATc (TOM70 ↓); and TOM70 is knocked down through the addition of 960 ng/μl ATc, but mt-Tah1 rescues the synthetic growth defect of some mutants (TOM70 ↓ + mt-Tah1). Shown are mean values and standard deviations from three replicates.

Figure 6. Chaperone binding by Tom70 is crucial for the biogenesis of small inner membrane proteins

(A and B) Protein levels in mitochondria isolated from either wild-type cells or the indicated Δtom70/71 mutants were analyzed by western blotting. Six data points from three biological repeats were analyzed for each protein. The error bars refer to standard deviations. The p values were generated from the two-tailed paired t test. (C) The volcano plot shows the comparison of the proteomes of Δtom70/71 cells that express the mt-Tah1(K8A) to those with mt-Tah1. The positions of several small inner membrane proteins (brown) and of carriers (purple), which are considerably stabilized by mt-Tah1 but not by mt-Tah1(K8A), are indicated. (D) The effects by which Tom70 and mt-Tah1 influence the cellular proteomes are plotted against each other. (E) Relative log₂ FCs of Tom70-dependent mitochondrial proteins that are rescued by either mt-Tah1 or its variant. (F) Graphical overview of the number of Tom70-dependent proteins that are rescued by expression of mt-Tah1 near to Tom70 full-length levels.

Figure 7. Chaperone binding by Tom70 prevents mitoprotein-induced toxicity

(A) Proteins that are enriched by the expression of mt-Tah1 compared with Δtom70/71 are shown. Only proteins with masses smaller than 18 kDa and positive enrichment factors were considered. Mitochondrial proteins are indicated in blue. (B) Schematic representations of small inner membrane proteins for which information about their overall structure and targeting information exists. Blue regions show presequences, and black boxes indicate transmembrane domains. (C and D) Cox5a-HA, Tim11-HA, Atp17-HA, and Atp18-HA were expressed under GAL1 control from multi-copy plasmids in wild-type and Δtom70/71 cells. The times indicate how long cells were shifted to 0.5% galactose-containing medium. (E) Radiolabeled Cox5a was incubated with isolated mitochondria for the times indicated at 30°C. The membrane potential (Δψ) was depleted in control samples by addition of CCCP. Mitochondria were reisolated and incubated with or without proteinase K. (F and G) The indicated strains were transformed with plasmids to express Atp17-HA, Tim11-HA, Cox5a-HA, and Pet9-HA under the control of the GAL1 promoter. All cultures were grown on lactate medium to mid-log phase, induced with 0.5% galactose for 4.5 h, and dropped onto galactose plates. (H) Rpn4-driven gene expression was measured using a yellow fluorescent protein (YFP) reporter system (Boos et al., 2019). (I) Tom70 supports the biogenesis of aggregation-prone mitochondrial membrane proteins by recruiting cytosolic chaperones to the mitochondrial surface, thereby generating a “mitochondria-associated proteophilic zone.”