Ubiquitin-Dependent Degradation of Mitochondrial ... - Cell Press

138 downloads 0 Views 3MB Size Report
Jun 5, 2018 - Didier Lacombe,1,2,5 Jean-William Dupuy,2,6 Claude Lalou,1,2 ...... Perciavalle, R.M., Stewart, D.P., Koss, B., Lynch, J., Milasta, S., Bathina, M.,.
Report

Ubiquitin-Dependent Degradation of Mitochondrial Proteins Regulates Energy Metabolism Graphical Abstract

Authors Julie Lavie, Harmony De Belvalet, Sessinou Sonon, ..., Jean-William Dupuy, Claude Lalou, Giovanni Be´nard

Correspondence [email protected]

In Brief Lavie et al. show that ubiquitindependent degradation of key mitochondrial proteins regulates mitochondrial energy metabolism and that turnover of several key proteins is dependent upon the ubiquitin proteasome system (UPS). These findings support the idea that ubiquitindependent degradation is directly involved in the regulation of mitochondrial energy metabolism.

Highlights d

Ubiquitination occurs at the inner mitochondrial membrane

d

SDHA is ubiquitinated within the mitochondria

d

UPS-mediated SDHA degradation maintains proper balance of mitochondrial metabolites

Lavie et al., 2018, Cell Reports 23, 2852–2863 June 5, 2018 ª 2018 The Authors. https://doi.org/10.1016/j.celrep.2018.05.013

Cell Reports

Report Ubiquitin-Dependent Degradation of Mitochondrial Proteins Regulates Energy Metabolism Julie Lavie,1,2 Harmony De Belvalet,1,2 Sessinou Sonon,1,2 Ana Madalina Ion,1,2,3 Elodie Dumon,1,2 Su Melser,2,4 Didier Lacombe,1,2,5 Jean-William Dupuy,2,6 Claude Lalou,1,2 and Giovanni Be´nard1,2,7,* 1Laboratoire

Maladies Rares, Ge´ne´tique et Me´tabolisme-INSERM U1211, 33000 Bordeaux, France de Bordeaux, 146 rue Le´o-Saignat, 33076 Bordeaux Cedex, France 3Molecular Mechanisms of Disease, Radboud University, 65000 HC Nijmegen, the Netherlands 4INSERM, U1215 NeuroCentre Magendie, 33000 Bordeaux, France 5CHU Bordeaux, Service de Ge ´ ne´tique Me´dicale, 33076 Bordeaux, France 6Plateforme Prote ´ ome, Centre de Ge´nomique Fonctionnelle, Universite´ de Bordeaux, 146 rue Le´o Saignat, 33076 Bordeaux Cedex, France 7Lead Contact *Correspondence: [email protected] https://doi.org/10.1016/j.celrep.2018.05.013 2Universite ´

SUMMARY

The ubiquitin proteasome system (UPS) regulates many cellular functions by degrading key proteins. Notably, the role of UPS in regulating mitochondrial metabolic functions is unclear. Here, we show that ubiquitination occurs in different mitochondrial compartments, including the inner mitochondrial membrane, and that turnover of several metabolic proteins is UPS dependent. We specifically detailed mitochondrial ubiquitination and subsequent UPSdependent degradation of succinate dehydrogenase subunit A (SDHA), which occurred when SDHA was minimally involved in mitochondrial energy metabolism. We demonstrate that SDHA ubiquitination occurs inside the organelle. In addition, we show that the specific inhibition of SDHA degradation by UPS promotes SDHA-dependent oxygen consumption and increases ATP, malate, and citrate levels. These findings suggest that the mitochondrial metabolic machinery is also regulated by the UPS. INTRODUCTION Mitochondria produce energy in the form of ATP through a complex metabolic network that requires a high degree of regulation (Benard et al., 2010a). One of the main regulatory mechanisms is based on modulations of the metabolic machinery content. Such regulation is performed through biogenesis and degradation processes that can modify levels of individual enzymes or modulate overall amounts of mitochondria. Recently, our group showed that global degradation of mitochondria by mitophagy is a decisive event in maintaining the long-term efficiency of energy production (Melser et al., 2013). However, mitophagy is a general process leading to global degradation of membranes, DNA, and proteins. Degradation of single mitochondrial proteins

by proteases, such as LON protease, has also been described as being sensitive to mitochondrial energy status, but mainly related to acute mitochondrial protein-folding stress and hypoxic €nch and Harper, and oxidative stress (Bota and Davies, 2002; Mu 2016; Teng et al., 2013). This mechanism does not target specific proteins and is mostly a stress-induced response rather than a specific regulatory process associated with energy metabolism. Therefore, both mitophagic and LON-associated degradation processes are not appropriated to exert dynamic and precise control of mitochondrial energy metabolism. In cells, proteinspecific degradation occurs through the ubiquitin proteasome system (UPS). Ubiquitin is an 8.5-kDa peptide (76 amino acids) that specifically and covalently binds to lysine residues in a target protein. This binding involves different enzymes referred to as E1, E2, and E3 ubiquitin ligases. Among these, E3 ubiquitin ligases (E3s) are devoted to the specific recognition of target proteins and to the binding of ubiquitin to a lysine residue in the target. Other ubiquitin molecules are then successively attached to the first ubiquitin and, finally, the polyubiquitin-tagged protein is transported to the cytosolic proteasome for degradation (Hershko and Ciechanover, 1998). In several cellular processes, UPS-dependent degradation plays the role of an ‘‘on/off’’ switch, representing an efficient regulatory method for inhibiting processes and/or redirecting signaling cascades. For example, p53-mediated cell arrest is inhibited by the ubiquitin-dependent degradation of p53 (Kubbutat et al., 1997), and the induction of apoptosis is regulated by the elimination of caspases or BAX (Benard et al., 2010b; Yang et al., 2000). Recently, several studies reported a link between mitochondrial physiology and the UPS. Reduced UPS activity further deteriorates mitochondrial functions in models of mitochondrial diseases (Segref et al., 2014). Additionally, mutations in F-box and leucine-rich repeat protein 4 (FBXL4), a potential mitochondrial E3 ubiquitin ligase, are responsible for a mitochondrial encephalopathy associated with severe mitochondrial bioenergetic dysfunctions (Bonnen et al., 2013; Gai et al., 2013). Most studies investigating the link between the UPS and mitochondria focus on ubiquitination of outer mitochondrial membrane (OMM)

2852 Cell Reports 23, 2852–2863, June 5, 2018 ª 2018 The Authors. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

proteins. To date, several OMM proteins, including BAX, DRP1, MFN1/2, and VDAC, were found to be ubiquitinated with important consequences for mitochondrial physiology and cell health (Benard et al., 2010b; Burchell et al., 2013; Guan et al., 2013; Karbowski et al., 2007; Nakamura et al., 2006). By contrast, ubiquitination of proteins involved in mitochondrial energy machinery remains barely investigated, because these proteins are mainly located in the matrix and inner mitochondrial membrane (IMM). However, several studies have observed ubiquitination of many intramitochondrial proteins using whole-cell proteomics (Udeshi et al., 2013; Wagner et al., 2011, 2012). Unfortunately, these whole-cell approaches do not provide molecular details about these ubiquitination events; therefore, sites where these ubiquitination events occur, as well as their physiological relevance, remain uncharted. In this study, we postulated that ubiquitin-dependent degradation of key mitochondrial proteins participates in the regulation of mitochondrial energy metabolism. We demonstrated that turnover of several oxidative phosphorylation (OXPHOS) proteins is dependent upon the UPS. Additionally, we specifically detailed UPS-dependent degradation of succinate dehydrogenase subunit A (SDHA) and its physiological consequences for mitochondrial energy metabolism. Our findings support the idea that ubiquitin-dependent degradation is directly involved in the regulation of mitochondrial energy metabolism. RESULTS Machinery Involved in Mitochondrial Energy Metabolism Is Degraded by the UPS To determine whether the UPS is involved in mitochondrial metabolism through degradation of intramitochondrial proteins, we first assessed the presence of endogenous ubiquitination in the organelle. Cell-fractionation experiments revealed that ubiquitin accumulated in mitochondria-enriched fractions following treatment with the proteasome inhibitors epoxomicin (Figure 1A) or MG132 (Figure S1A). Additionally, immunofluorescence results showed that K48-polyubiquitinated proteins aggregated in the cytosol and nucleus upon proteasome-inhibitor treatment, and that, interestingly, some K48-polyubiquitinated proteins also colocalized to the mitochondria (Figures 1B and 1C). To evaluate potential accumulations of ubiquitinated proteins in different mitochondrial compartments, we performed sub-mitochondrial fractionations, resulting in no ubiquitin conjugates detected in the matrix following MG132 or epoxomicin treatment (Figure 1D); however, we found strong accumulations of K48 ubiquitin conjugates in both the IMM and OMM. Importantly, these results demonstrated that intramitochondrial proteins were endogenously ubiquitinated inside the organelle, and that a specific type of ubiquitination targeted these proteins for proteasomal degradation. To determine the mitochondrial proteins targeted for ubiquitination, we performed mass spectrometry analyses and identified 203 different mitochondrial proteins with a specific ubiquitin-remnant motif (i.e., di-glycine) present on lysines (Figure S1B). These results were consistent with previous proteomic studies (Lehmann et al., 2016; Udeshi et al., 2013; Wagner et al., 2011, 2012) and illustrated that ubiquitination might represent a

major mechanism associated with regulation of the mitochondrial proteome. Interestingly, 82 of these ubiquitinated proteins were located in the IMM or the matrix and involved in mitochondrial energy metabolism (Figure S1B). Accordingly, we focused on the role played by the UPS in regulating subunit A of succinate dehydrogenase (SDHA), a key enzyme involved in both OXPHOS and the tricarboxylic acid cycle (TCA). First, we analyzed SDHA turnover using a combination of MG132 and a protein-synthesis inhibitor, cycloheximide (CHX). As a positive control, we used induced myeloid leukemia cell-differentiation protein 1 (MCL1), because this mitochondrial protein is reportedly degraded by the UPS (Perciavalle et al., 2012). Interestingly, SDHA turnover was sensitive to MG132 treatment similar to many subunits of OXPHOS complexes, including GRIM19 (complex I) or COX IV (complex IV) (Figures 1E and S1C). Similar results were obtained when we used epoxomicin (Figures 1F and S1D) or human primary skeletal muscle myoblasts (HSMMs; Figure 1G), which showed that these effects were not specific to the proteasomal inhibitor or to cell type. Under these conditions, we also showed that LON protease was not significantly involved in UPS-dependent SDHA turnover (Figures 1H and S2E). These results demonstrated that intramitochondrial proteins, including SDHA, were degraded by the UPS. Ubiquitination Occurs inside Mitochondria Because the majority of mitochondrial proteins are synthesized in the cytosol and then imported into the organelle, it remains unclear where these ubiquitination events occur. To discriminate between mitochondrial and cytosolic ubiquitination, we developed an HA-tagged ubiquitin coupled to a mitochondrial leader sequence (mt-Ubi). In HeLa cells, mt-Ubi localized in mitochondria, whereas HA-tagged ubiquitin without the leader sequence (whole-cell ubiquitin [wc-Ubi]) was distributed ubiquitously within the cell, including in mitochondria (Figures 2A and 2B). We measured the degrees of colocalization between HAtagged ubiquitin and mitochondria using Pearson’s correlation coefficient, with these values higher in HeLa cells expressing mt-Ubi (0.83 ± 0.01) as compared with HeLa cells expressing wc-Ubi (0.33 ± 0.05) (n = 4 replicates of >80 cells; p < 0.001, unpaired t test). This localization to the organelle was not affected by MG132 treatment (Figure 2A). To confirm mt-Ubi functionality in the mitochondria, we performed fractionation of cells expressing either wc-Ubi or mt-Ubi. In the presence of MG132, we found that wc-Ubi-transfected cells displayed HA-tagged ubiquitin conjugates in both mitochondrial and cytosolic fractions, whereas mt-Ubi-transfected cells exhibited accumulation of ubiquitinated proteins only in mitochondria (Figure 2C). Submitochondrial fractionation showed that the matrix addressed mt-Ubi accumulation in IMM following time-course treatment with epoxomicin (Figure 2D). Furthermore, these mt-Ubi conjugates included K48-ubiquitinated proteins (Figures 2E and 2F). Degrees of colocalization between K48 and hemagglutinin (HA) immunostaining were measured using Pearson’s correlation coefficient, with results of 0.74 ± 0.02 in the presence of epoxomicin as compared with 0.25 ± 0.05 in vehicle-treated cells. Similar results were obtained using MG132 (Figures S2A and S2B). Together, these results showed that proteasomal inhibition induced the accumulation of mt-Ubi conjugates at Cell Reports 23, 2852–2863, June 5, 2018 2853

Figure 1. Accumulation of Ubiquitin Conjugates in Mitochondria (A) Representative western blots showing accumulation of polyubiquitinated proteins in mitochondria-enriched fractions. Subcellular fractionations were obtained from HeLa cells treated with DMSO or epoxomicin. Ubiquitinated proteins were detected using an anti-ubiquitin antibody (ubi). Mitochondrial and cytosolic fractions were labeled using TOM20 and L-lactate dehydrogenase A (LDHA) antibodies, respectively. (B) Immunofluorescence showing subcellular localization of K48 conjugates in HeLa cells following proteasome-inhibitor treatment. Cells were treated for 8 hr with DMSO (Veh) or different proteasome inhibitors (MG132, epoxomicin, or lactacystin) before fixation. Mitochondria and ubiquitin K48-specific linkage were immunostained using antibodies against carnitine O-palmitoyltransferase (CPTA1, green) and K48 (red), respectively. Scale bars, 20 mm. (C) Line scales obtained from the images in (B) showing partial overlaps between mitochondria (red line) and K48 staining (green lines) upon different treatment. (D) Sub-mitochondrial localization of K48 conjugates analyzed by western blot. Sub-mitochondrial compartments (OMM, IMM, and matrix [Mx]) were obtained from HeLa cells treated with vehicle, MG132, or epoxomicin. MFN2, ATP5A, and ATAD3A are respective markers of the OMM, Mx, and IMM. (E) Representative western blots showing inhibition of mitochondrial protein degradation by MG132. HeLa cells were treated with MG132 or DMSO for 16 hr in the presence or absence of CHX. Levels of mitochondrial proteins were revealed by western blot of total cell extracts. (F and G) Representative western blot showing SDHA turnover following CHX treatment of (F) HeLa cells in the presence of epoxomicin or (G) in HSMMs in the presence of epoxomicin or MG132. (H) Effect of LONP1 downregulation on SDHA turnover analyzed in HeLa cells transfected with a cocktail of siRNA against luciferase (siCTRL) or LONP1 (siLONP1) and treated with MG132 and CHX. Western blot results are representative of four independent experiments. Cyt, cytosol-enriched fraction; Mito, mitochondria-enriched fraction; TCL, total cell lysate. See also Figure S1.

the IMM, and that these conjugates contained K48-specific linkages. In the presence of mt-Ubi, the electrophoretic profile of endogenous SDHA was modified by the accumulation of higher-molecular-weight proteins upon MG132 treatment (Figure S2C). These results suggested that SDHA was polyubiquitinated in the mito2854 Cell Reports 23, 2852–2863, June 5, 2018

chondrial compartment and accumulated upon proteasomal inhibition. To validate this possibility, we performed immunoprecipitation assays on cells expressing either wc-Ubi or mt-Ubi and treated with epoxomicin or vehicle. Pull down of HA-tagged ubiquitin showed that mt-Ubi-ubiquitinated mitochondrial proteins contained K48 aggregates (Figure 2G). Furthermore, we

Figure 2. Ubiquitination Occurs in the Mitochondrial Inner Compartment (A) Immunofluorescence performed on HeLa cells transfected with wc-Ubi or mt-Ubi. Mitochondria and ectopic ubiquitin were stained using TOM20 (red) and anti-HA (green), respectively. Cells were treated with MG132 or DMSO for 8 hr before fixation. (B) Line scale obtained from images in wc-Ubi (bottom) or mt-Ubi (upper) cells in the absence of MG132. Fluorescence intensities are shown for TOM20 (red line) and anti-HA (green line). (C) Cell fractionation performed on HeLa cells transfected with wc-Ubi, mt-Ubi, or control (Crtl) and in the presence or absence of MG132 treatment. Presence of HA-ubiquitin aggregates in cytosolic and mitochondria-enriched fractions were revealed by immunoblot against Ha-tag. *Indicates nonspecific bands. (D) Sub-mitochondrial fractionation performed on HeLa cells transfected with mt-Ubi and treated with epoxomicin for 0, 4, and 8 hr. Localization of mt-Ubi conjugates was revealed using immunoblot against the HA-tag. MFN2, MRPL45, and ATAD3A are respective markers of the OMM, Mx, and IMM. (E) Accumulation of K48-specific linkages in HeLa cells expressing mt-Ubi were analyzed by immunofluorescence. Cells were co-transfected with mt-Ubi and mtDsRed and treated with epoxomicin or vehicle. Mitochondria, mt-Ubi, and K48 ubiquitination were detected using DsRed fluorescence (red) and the anti-HA (green) and anti-K48 (blue) antibodies. (F) Degrees of colocalization between mt-Ubi and K48 conjugates in the presence or absence of epoxomicin treatment. Colocalization was evaluated by measuring Pearson’s correlation coefficient (n = 3, >50 counted cells; ***p < 0.001, unpaired t test). (G) Immunoprecipitation of HA-tagged ubiquitin performed using HeLa cells expressing wc-Ubi, mt-Ubi, or GFP (ctrl). Cells were treated with epoxomicin for 8 hr. Ha-tagged proteins, K48-ubiquitinated proteins, and SDHA and MFN1 levels were analyzed by immunoblot. Ha and SDHA expression were verified using 5-mg samples. Scale bar, 20 mm. See also Figure S2.

found that SDHA was also pulled down with mt-Ubi, indicating that SDHA was ubiquitinated inside the organelle. Notably, an 8-kDa modified-SDHA variant was found following immunoprecipitation of mt-Ubi-transfected cells (Figure 2G). Importantly, the OMM protein mitofusin-1 (MFN1), previously reported as being ubiquitinated (Xu et al., 2011), was recovered only upon wcUbi pull down. We obtained similar results following treatment of

cells with MG132 (Figure S2D). Because MCL1 can be recruited to different mitochondrial compartments (Perciavalle et al., 2012), this protein was recovered with both constructs (Figure S2D). These results showed that SDHA was ubiquitinated inside the organelle. To confirm these findings in endogenous ubiquitin conditions, we performed immunoprecipitations of ubiquitinated proteins Cell Reports 23, 2852–2863, June 5, 2018 2855

Figure 3. SDHA Is Ubiquitinated inside the Mitochondria (A) Immunoprecipitation of endogenous ubiquitin (ubi) performed on cytosolic and mitochondrial-enriched fractions obtained from HeLa cells treated with MG132 or DMSO for 16 hr. Control immunoprecipitation (Ctrl) was performed using an unrelated antibody coupled to agarose beads. Total cell lysate was used as the expression control. (B) Analysis of SDHA turnover in total cell lysate, mitochondrial, and cytosolic fractions. Fractions were obtained from HeLa cells treated with CHX for 0, 14, or 24 hr. After 14 hr of CHX treatment, epoxomicin or DMSO was added for an additional 10-hr incubation. Levels of SDHA in the different fractions were measured by western blot. LDHA and TOM20 are markers of cytosolic and mitochondrial fractions, respectively. (C) Accumulation of K48 conjugates in mitochondria was analyzed in HeLa cells transfected with SDHA-GFP or mito-GFP and treated with DMSO or epoxomicin for 8 hr. Transfected cells were detected using GFP labeling (Green) and immunostained for K48-specific ubiquitin linkages (blue). Mitochondria were stained using mito-DsRed (Red). (D) Pearson’s correlation coefficients between K48 conjugates and mito-DsRed staining obtained following epoxomicin treatment and measured in cells overexpressing mito-GFP or SDHA-GFP. Each mark represents the Pearson’s coefficient for one cell, and black marks represent the mean (n = 3, >50 counted cells). **p < 0.01, unpaired t test. (E) Immunofluorescence showing endogenous expression of SDHA in cells transfected with mt-Ubi. mt-Ubi and endogenous SDHA were labeled using anti-HA (red) and anti-SDHA (green). Cells were treated with either epoxomicin (epoxo) or DMSO (veh) for 8 hr. (F) Intensity of SDHA fluorescence measured in control (white circles) or mt-Ubi-transfected cells treated with vehicle (light gray circles) or epoxomicin (dark gray circles). Intensities were normalized to the cell surface. Each dot represents the value for one cell (n = 3–7, 60–100 cells). ***p < 0.001, one-way ANOVA, Dunn’s post hoc test. Scale bars, 20 mm. See also Figure S2.

from both cytosolic and mitochondrial-enriched fractions, finding that SDHA pulled down from the mitochondrial-enriched fractions (Figure 3A). We further confirmed these results by assaying accumulations of SDHA in mitochondrial or cytosolic fractions following pretreatment with CHX. For these experi2856 Cell Reports 23, 2852–2863, June 5, 2018

ments, cells were pretreated with CHX for 14 hr before adding epoxomicin in order to deplete cells from SDHA precursors (Figure 3B). Accordingly, no SDHA was detected in the cytosol, whereas the mitochondrial SDHA pool decreased upon CHX treatment, with this decrease inhibited by an additional 10-hr

incubation with epoxomicin. These results were consistent with previous studies showing that mitochondrial nascent proteins are directly coupled to their mitochondrial import, which strongly decreases their accessibility for ubiquitination in the cytosol (Gehrke et al., 2015; Gold et al., 2017; Lesnik et al., 2014). We then found that an artificial increase in SDHA expression in the mitochondria using SDHA-GFP transfection induced the accumulation of endogenous ubiquitinated K48 in the organelle (Figure 3C). Measurements of colocalization degrees between K48 labeling and GFP resulted in a value of 0.26 ± 0.02 in mitochondria expressing SDHA-GFP as compared with 0.15 ± 0.03 in mitochondria expressing control mito-GFP (Figures 3C and 3D). Upon MG132 treatment, we obtained values of 0.40 ± 0.06 and 0.12 ± 0.05 in cells expressing SDHA-GFP or mitoGFP, respectively (Figures S2E and S2F). We also found that mt-Ubi expression was sufficient to decrease endogenous SDHA levels, because fluorescence intensity associated with SDHA immunostaining decreased from 0.62 ± 0.03 to 0.45 ± 0.02 intensity/mm2 in control and mt-Ubi-expressing cells, respectively. Treatment with epoxomicin restored SDHA levels in mt-Ubi-expressing cells (0.76 ± 0.05 intensity/ mm2) (Figures 3E and 3F). These results demonstrated that SDHA was degraded by the UPS, and that this process included a ubiquitination step in the organelle. UPS-Dependent SDHA Degradation Is Related to Mitochondrial Energy Metabolism To determine whether SDHA degradation affects mitochondrial energy metabolism, we first analyzed SDHA turnover under different growing conditions. In the presence of glucose or glutamine as energy sources, UPS was involved in the SDHA turnover as shown by the strong effect of MG132 treatment (Figures 4A and 4B). By contrast, during the use of palmitoyl-D, L-carnitine to produce energy via b-oxidation, SDHA was highly stable and not degraded by the UPS (Figures 4A and 4B). Under these conditions, SDHA participated differentially in mitochondrial energy metabolism according to measurement of oxygen-consumption rate (OCR) in the presence of malonate, an SDH inhibitor. In the presence of palmitoyl-D, L-carnitine, OCR was inhibited by 65.4% ± 7.8% by malonate treatment (Figures 4C and S3A), whereas in the presence of glucose or glutamine, inhibition percentages were lower, at 15.9% ± 4.7% and 26.3% ± 5.5%, respectively (Figures 4C and S3A). Similar results were obtained in HSMMs (Figures S3B–S3D). Therefore, we postulated that high requirements of SDHA for mitochondrial energy metabolism resulted in a decreased SDHA-degradation rate. To confirm this hypothesis, we investigated mitochondrial energy metabolism under conditions where SDHA was artificially stabilized. Following remnant di-glycine proteomic analyses, we found that SDHA could be ubiquitinated on several lysines (Figure S4A). We then developed a Myc-tagged SDHA variant containing lysine-to-arginine substitutions at positions 179 and/or 541 (referred to as Myc-SDHAK179R, Myc-SDHAK541R, and Myc-SDHAK179R,K541R). Incorporation of these mutants resulted in lower SDHA-degradation rates, and MG132 treatment had less impact on SDHA degradation (Figures 4D and 4E). Additionally, these mutants showed fewer signs of ubiquitination in-

side the organelle. Indeed, in cells co-expressing both mt-Ubi and wild-type (WT)/mutant Myc-SDHA variants, HA-tag immunoprecipitation revealed strong reductions in recovered MycSDHA mutants as compared with WT Myc-SDHA, with recovered Myc-SDHAK179R,K541R corresponding to 34% ± 14% that of recovered WT Myc-SDHA (Figures S4B and S4C). Accordingly, inhibition of UPS-dependent degradation was particularly strong with the double mutant, which was subsequently used to analyze the impact of inhibition of UPS-dependent SDHA degradation on mitochondrial energy metabolism. Ectopic expression of Myc-SDHAK179R,K541R significantly increased OCR sensitivity to malonate (Figures 4F and S4D), with the percentage of OCR inhibition by malonate increasing in cells transfected with Myc-SDHAK179R,K541R (44.4% ± 4.3%) as compared with that observed in cells transfected with Myc-SDHA WT (31.5% ± 2.6%). Importantly, this double mutation did not impact SDHA mitochondrial localization (Figures S4E and S4F), and steady-state levels of the SDH complex barely increased (Figure S4G). Similarly, intrinsic SDH enzyme properties were unaffected by these mutations (Figure S4H), with maximal activities (Vmax) of 371 ± 137 and 341 ± 109 variation of optical density (DOD)/min/mg protein for Myc-SDHA WT and MycSDHAK179R,K541R, respectively, and apparent Michaelis constants (km-app) of 16.3 ± 3.6 and 12.7 ± 2.7 mM for Myc-SDHA WT and Myc-SDHAK179R,K541R, respectively. Therefore, expression of a stable SDHA variant promoted the participation of SDH in mitochondrial energy metabolism. Furthermore, impaired SDHA degradation induced broader changes in mitochondrial energy metabolism in the presence of glutamine. Additionally, significant increases in ATP concentration were observed (+20.2%) in Myc-SDHAK179R, K541R cells as compared with Myc-SDHA WT cells (Figure 4G). There were also significant increases in concentrations of TCA-cycle metabolites, including malate (+45.1%) and citrate (+21%), in Myc-SDHAK179R,K541R cells as compared with Myc-SDHA WT cells, although no significant succinate accumulation was observed. These results showed that impairing SDHA UPS-dependent degradation increased SDH-specific flux. DISCUSSION The link between the UPS and mitochondria has been insufficiently documented; however, numerous studies now support that mitochondria are not the exception regarding ubiquitinrelated activities. Considering ubiquitination and UPS-dependent degradation of mitochondrial proteins, most studies focus on the turnover of OMM proteins. Indeed, there remains a conceptual barrier to considering that intramitochondrial proteins (i.e., IMM and matrix proteins) could be degraded by the cytosolic proteasome. In this study, we demonstrated that ubiquitination occurs inside the organelle, and that turnover of several intramitochondrial proteins depends upon the proteasome. Whole-cell proteomic investigations demonstrated that 25% to 30% of mitochondrial proteins are ubiquitinated (Udeshi et al., 2013; Wagner et al., 2011), although ubiquitination rates and protein targets vary according to tissue (Wagner et al., 2012). These data agreed with our proteomics data showing that 203 mitochondrial proteins were ubiquitinated. Recently, Cell Reports 23, 2852–2863, June 5, 2018 2857

Figure 4. Links between SDHA Degradation and Mitochondrial Metabolism (A) SDHA turnover investigated in HeLa cells grown in the presence of different energy-related substrates. Cells were treated with CHX in the presence of MG132 or DMSO for 16 hr. Substrate conditions involved the presence of glucose (Gluc), glutamine (Glut), or palmitoyl-D, L carnitine (PC). (B) Quantification of SDHA levels in the presence of different energy-related substrates according to western blots as in (A). SDHA levels were normalized against actin. Black, white, and gray bars represent the presence of DMSO, CHX, and CHX+MG132, respectively. *p < 0.05; #p < 0.05 CHX versus CHX+MG132, Sidak’s multiple-comparison test (n = 6). (C) Inhibition of complex II-dependent respiration by malonate under different energy-related substrate conditions and expressed as a percentage of their respective controls in HeLa cells (n = 6). **p < 0.01; ***p < 0.001, Sidak’s multiple-comparison test. (D) Effect of MG132 on the accumulation of Myc-SDHA, Myc-SDHAK179R (K179), Myc-SDHAK541R (K541), and Myc-SDHAK179R,K541R analyzed by western blot. HeLa cells were labeled as no transfection (NT) or transfected with mutant or WT Myc-SDHA and treated with MG132 (8 hr). Expression of ectopic SDHA was revealed using an anti-Myc antibody. (E) Quantification of ectopic SDHA levels obtained from western blot upon treatment with vehicle (white bars) or MG132 (gray bars). *p < 0.05; unpaired t test (n = 4). (F) Inhibition of complex II-dependent respiration by malonate under different energy-related substrate conditions and expressed as the percentage of their respective controls in HeLa cells expressing Myc-SDHA or Myc-SDHAK179R,K541R. *p < 0.05, unpaired t test (n = 7). (G) Levels of different cellular metabolites analyzed in HeLa cells transfected with Myc-SDHA (light bars) or Myc-SDHAK179R,K541R (dark bars) (n = 3–4). *p < 0.05, two-way ANOVA and multiple-comparison test. See also Figures S3 and S4.

Lehmann et al. (2016) estimated that 62% of the mitochondrial proteome might be ubiquitinated. These proportions show that ubiquitination of mitochondrial proteins should impact mitochondrial physiology; however, these studies failed to identify whether this ubiquitination occurs inside the organelle. In the 2858 Cell Reports 23, 2852–2863, June 5, 2018

present study, we specifically investigated ubiquitination and degradation of the intramitochondrial protein SDHA, and showed that its ubiquitin-dependent degradation affected mitochondrial energy metabolism. Despite these findings and regarding intramitochondrial-protein ubiquitination and

ubiquitin-dependent degradation, several important questions remain unclear. Among the most challenging questions regarding ubiquitindependent degradation of mitochondrial proteins are those related to the trafficking of ubiquitinated proteins. A recent study showed that impaired import of mitochondrial proteins induces accumulation of mitochondrial precursors in the cytosol and, thus, specifically stimulates proteasomal activity (Wrobel et al., 2015). In the absence of altered import, nascent mitochondrial proteins are directly coupled to their import, which strongly decreases their accessibility for cytosolic ubiquitination (Gehrke et al., 2015; Gold et al., 2017; Lesnik et al., 2014). Our results demonstrated that ubiquitin conjugates accumulated in both mitochondrial membranes. Whether this ubiquitination occurs during the import of proteins or at the end of their life cycle, as well as how IMM-ubiquitinated proteins reach the cytosolic proteasome, remains unknown. If UPS-dependent protein degradation occurs at the end of their life cycle in the mitochondrial matrix, specific mechanisms should exist to unfold and export the protein to the cytosol. To date, no evidence supporting the existence of an IMM retro-translocation system has been made available; therefore, the simplest model involves their ubiquitination during import. We believe that ubiquitination of intramitochondrial proteins occurs before their assembly into native complexes, with newly synthetized and unfolded proteins ubiquitinated at the level of the IMM during import. This ubiquitination might occur at the matrix or the intramembrane space, after which ubiquitinated proteins accumulate in this membrane. Ubiquitinated proteins should be then extracted from the IMM and released into the cytosol for proteasomal degradation. In this process, the IMM represents a sorting/regulatory barrier that coordinates intramitochondrial-protein content, and this further supports the intricate link between the import machinery and degradation of mitochondrial proteins. Previous studies hypothesized this mechanism (Azzu and Brand, 2010; Margineantu et al., 2007); however, to date, there has been no clear presentation of a molecular mechanism or signaling details supporting such activity. By contrast, there is more information available regarding retro-translocation through the OMM. Indeed, several studies demonstrate the existence of mitochondria-associated degradation (Bragoszewski et al., 2015; Karbowski and Youle, 2011; Taylor and Rutter, 2011), where P97/valosin-containing protein (VCP) appears to play a key role. This ATPases associated with diverse cellular activities (AAA)-ATPase extracts ubiquitinated proteins from endoplasmic reticulum membranes, followed by presentation of these proteins to the cytosolic proteasome for degradation. Xu et al. (2011) showed that P97/ VCP is also associated with mitochondria in human cells. At these mitochondrial sites, the AAA-ATPase extracts mitochondrial proteins from the OMM for their cytosolic degradation by the proteasome (Xu et al., 2011). This process also implicates other proteins in other species. In yeast, Heo et al. (2010) showed that Vms1 is required for mitochondrial activity and localization of Cdc48, the yeast homolog of P97/VCP (Heo et al., 2010). Doa1 is also needed to maintain Cdc48-related mitochondrial activity (Wu et al., 2016). Mitochondrial proteins can be eliminated by different degradation systems, including mitophagy and proteases. Therefore, the

specific function of UPS-dependent degradation of mitochondrial proteins remains an open question. Mitophagy and proteases are mainly related to damage or stress-induced responses and exhibit low specificity (Haynes et al., 2007; Jin and Youle, 2013). By contrast, ubiquitin-dependent degradation is a highly specific mechanism that controls numerous essential cellular functions (Kubbutat et al., 1997; Yang et al., 2000). Accordingly, it is likely that specific degradation of key metabolic enzymes might represent an efficient mechanism for inhibiting or promoting specific metabolic pathways under specific conditions. Such regulation is exerted by the UPS on non-mitochondrialspecific metabolic pathways, such as those related to cholesterol synthesis (Hwang et al., 2016) or hepatic lipid storage (Yoshizawa et al., 2014). In the present study, our results identified ubiquitination of 82 mitochondrial metabolic enzymes (Figure S1B); therefore, many aspects of mitochondrial metabolic pathways, including the use of specific energy substrates or production of specific metabolites, are potentially regulated by UPS-dependent degradation. Regarding mitochondrial energy metabolism, recent studies correlate mitochondrial metabolic dysfunction with the UPS. Segref et al. (2014) demonstrated that acute inhibition of mitochondrial respiration using rotenone, antimycin, or azide modulates proteasomal degradation, and inhibition of proteasomal degradation was also observed in the fibroblasts of patients carrying cytochrome c oxidase deficiency (Segref et al., 2014). Sun and Denko (2014) showed that UPS-dependent degradation of a-ketoglutarate dehydrogenase (OGDH2) occurs under hypoxic conditions, and that this degradation involves the E3 ligase SIAH2. This degradation leads to decreased conversion of a-ketoglutarate into succinate and promotes accumulation of citrate via isocitrate dehydrogenase activity. As a result, there is a shift in mitochondrial energy metabolism from energy production through the TCA cycle toward accumulation of fatty acids through citrate buildup (Sun and Denko, 2014). Our results showed that impairing UPS-dependent SDHA degradation promoted the contribution of SDH activity to mitochondrial energy metabolism. Additionally, inhibiting SDHA degradation promoted SDH-dependent oxygen consumption and ATP production, resulting in increased levels of TCA-cycle metabolites (malate and citrate). In both OGDH2 and SDHA cases, degradation of one specific mitochondrial protein was enough to induce a particular remodeling of mitochondrial metabolism, revealing the efficiency of this UPS-dependent regulation for metabolism. Furthermore, these results suggest that the UPS senses mitochondrial-metabolic status and responds by degrading specific targets. Another open question concerns how mitochondria communicate with the UPS. Collapse of mitochondrial-membrane potential stabilizes the E3 ligase PARKIN in the mitochondria, thereby promoting ubiquitination of OMM proteins (Narendra et al., 2008). ATP levels also play important roles in PARKIN-dependent ubiquitination. First, PARKIN stabilization requires protein kinase A-dependent phosphorylation of the IMM protein MIC60 (Akabane et al., 2016), and second, phosphorylation of ubiquitin is needed to activate PARKIN (Kane et al., 2014; Koyano et al., 2014). These phosphorylation steps require ATP, the levels of which are highly dependent upon mitochondrial energy metabolism. In cells containing healthy mitochondria, there Cell Reports 23, 2852–2863, June 5, 2018 2859

are likely other signaling mechanisms capable of stabilizing E3 ligases at the IMM. For example, degradation of OGDH2 occurs upon changes in oxygen levels, whereas SDHA degradation is dependent upon energy-related substrates. Future identification and characterization of mitochondrial-specific ubiquitin ligases will provide further details concerning communication between mitochondria and the UPS. The E1 ubiquitin ligase has been observed in mitochondria, but this remains a debated issue (Schwartz et al., 1992). By contrast, several E2 and E3 ubiquitin ligases, as well as deubiquitylases, have been found in mitochondria (Benard et al., 2010b; Bingol et al., 2014; Haddad et al., 2013; Karbowski et al., 2007; Leboucher et al., 2012). Identification of a possible intramitochondrial E3 ubiquitin ligase (FBXL4) confirmed that the UPS was capable of controlling intramitochondrial metabolic protein content (Bonnen et al., 2013; Gai et al., 2013; Huemer et al., 2015). Mutations in this gene have been found in several patients and are responsible for mitochondrial encephalopathy accompanied by severe deficiencies in mitochondrial bioenergetics and alterations in OXPHOS proteins. Finally, the mitochondrial unfolded-protein response (UPRmt) likely interacts with mitochondrial ubiquitination (Zhao et al., 2002). The UPRmt is a proteotoxic stress response that promotes degradation of unfolded or misfolded proteins and involves mitochondrial proteases and chaperones (Haynes et al., €nch and Harper, 2016). The UPR was first identified 2013; Mu at the endoplasmic reticulum, and characterization of ER-associated degradation (ERAD) pathways subsequently illustrated how the UPR mobilizes the UPS (Friedlander et al., 2000; Gauss et al., 2006; Smith et al., 2011). It is possible that the UPRmt involves the mitochondria-associated degradation (MAD). The existence of MAD has been established (Taylor and Rutter, 2011); however, its involvement in the UPRmt remains unclear. ERAD components (P97/VCP or Doa1) are recruited to mitochondria in order to extract OMM-ubiquitinated proteins (Heo et al., 2010; Wu et al., 2016). On the other hand, activation of the UPRmt promotes expression of UPS-associated proteins, such as UBL5 (Haynes et al., 2007) and PINK1 (Thomas et al., 2014). Similar to the ER, the UPRmt might promote mitochondrial ubiquitination and activate MAD. It remains unknown whether the UPRmt and accumulation of ubiquitinated proteins in the IMM are interconnected processes. These mechanisms might be associated with different physiological functions or act in a coordinated manner. In the first case, the UPRmt would eliminate unfolded/damaged proteins, with ubiquitinationdependent degradation regulating levels of mitochondrial proteins. In a case where both mechanisms are coordinated, activation of the UPRmt might drive ubiquitination responses, or the accumulation of ubiquitinated proteins might activate the UPRmt. This regulatory activity might be independently activated by specific physiopathological conditions. For example, oxidative stress generated by the accumulation of unfolded/misfolded proteins activates the UPRmt (Kaufman et al., 2017). By contrast, our results and previous studies showed that degradation of specific mitochondrial proteins is related to the presence of specific metabolic substrates (Sun and Denko, 2014). We are confident that future work will provide experimental evidence supporting one of these hypotheses. 2860 Cell Reports 23, 2852–2863, June 5, 2018

In conclusion, regulation of mitochondrial functions by the UPS represents an emerging question, with numerous aspects requiring clarification. However, based on the recent literature and according to the data presented in this study, we believe that the role of the UPS in mitochondrial energy metabolism has important physiopathological implications. EXPERIMENTAL PROCEDURES Cell Culture, Transfection, and Pharmacological Treatments HeLa cells were cultured in glucose media consisting of DMEM containing 25 mM glucose. All media were supplemented with 10% heat-inactivated fetal bovine serum, 1 mM sodium pyruvate, non-essential amino acids, 100 U/mL penicillin, and 100 mg/mL streptomycin. Cells were cultured in a 5% CO2 atmosphere at 37 C. Cells were transfected using FuGENE HD (Roche, Basel, Switzerland) according to the manufacturer’s protocol. For experiments in the presence of different energy-related substrates, high-glucose media were removed and replaced with DMEM glucose-free medium containing 10 mM galactose and supplemented with 4 mM glutamine (glutamine media), 24 mM palmitoyl-D, L-carnitine (PC media), or 5 mM glucose (glucose media). Cells were cultured for at least 24 hr prior to experiments under these conditions. MG132 was used at 10 mM (final concentration) in media. Epoxomicin and lactacystin were used at 1 and 5 mM (final concentration), respectively. CHX was used at a final concentration of 50 mg/mL. HSMMs were cultured in skeletal muscle cell growth medium-2 supplemented with serum and growth factor (Lonza, Basel, Switzerland). Chemicals Proteasome inhibitors were purchased from UBPbio (Aurora, CO, USA). Other chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA). DNA Constructs Full-length SDHA clones were obtained by reverse transcription PCR of human fibroblast cDNA and cloned in-frame with a C-terminal Myc tag into the pCMV6 vector (Origene, Rockville, MD, USA). For the SDHA mutants (K179R, K541R, and the double mutant K179R,K541R), a QuikChange II XL site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA, USA) was used to introduce each point mutation. mt-Ubi was obtained by in-frame fusion of the mitochondrial-localization signal of pyruvate dehydrogenase a-1 (50 -atga ggaagatgctcgccgccgtctcccgcgtgctgtctggcgcttctcagaagccggcaagcagagtgctg gtagcatcccgtaattttgcaaatgatgctacatttctgcagtcgacggtaccgcgggcccg-30 ), the HA tag, and ubiquitin, followed by cloning into the pRRsin-PGK-MCS-WPRE vector (kindly provided by Plateforme de Vectorologie, Vect’UB, Universite´ de Bordeaux, Bordeaux, France). All constructed plasmids were verified by DNA sequencing (Beckmann-Coulter, Brea, CA, USA). Western Blot, Immunoprecipitation, and Cell Fractionation Total cell protein lysates and subcellular fractions were prepared using 23 sample buffer supplemented with protease inhibitors. Samples were analyzed by western blot using conventional methods. In brief, 10–40 mg protein was separated by electrophoresis (120 V for 1 hr) and transferred to polyvinylidene difluoride membranes using TransBlot Turbo (Bio-Rad, Hercules, CA, USA). Membranes were blocked with 5% milk in PBS-Tween (0.05%). Proteins were detected using specific antibodies diluted in 5% milk/PBSTween (0.05%). Antibodies were purchased from Abcam (SDHA, CPT1A, UBQCRC2, ATP5A, and cytochrome c; Cambridge, UK), Santa Cruz Biotechnology (TOM20 and total ubiquitin [P4D1]; Dallas, TX, USA), Roche (MYC-tag), Millipore (ubiquitin K48-specific linkage; Billerica, MA, USA), and SigmaAldrich (MCL1 and b-actin). Horseradish peroxidase (HRP)-conjugated antirabbit or anti-mouse antibody (Bio-Rad) was used as secondary antibodies. HRP signals were visualized using chemiluminescent substrates (Thermo Fisher Scientific, Waltham, MA, USA) and acquired using a Chemidoc MP imaging system (Bio-Rad) or Odyssey imaging system (Li-Cor Biosciences, Lincoln, NE, USA). For coimmunoprecipitation assays, cells were broken in lysis buffer (1% Triton X-100, 50 mM Tris [pH 7.4], 150 mM NaCl, 10 mM EDTA, and

protease inhibitors), and debris were removed by centrifugation for 20 min at 16,000 3 g. Supernatants were incubated for 4 hr with anti-Ha or anti-Myc agarose beads (Thermo Fisher Scientific) or anti-Ubi agarose beads (Santa Cruz Biotechnology) at 4 C. The beads were washed five times with Tris-buffered saline/Tween (0.05%), and proteins were eluted with 23 Laemmli sample buffer (Sigma-Aldrich). The different steps of cell fractionation were performed at 4 C. Cells were harvested in mitochondrial-isolation buffer (10 mM Tris HCl [pH 7.4], 210 mM mannitol, 70 mM sucrose, and 1 mM EDTA) supplemented with protease inhibitor (Sigma-Aldrich) and homogenized by passing through a 26G syringe (20 strokes). The samples were centrifuged at 500 3 g for 5 min at 4 C, followed by centrifugation of the supernatant at 10,000 3 g for 15 min at 4 C to obtain a heavy membrane pellet. The resulting supernatant was stored as the cytosolic fraction. The pellets were resuspended in 1 mL of isolation buffer and subjected to another round of centrifugation, with the final pellet stored as the mitochondria-enriched fraction. Sub-mitochondrial fractions were obtained by treating isolated mitochondria with digitonin (0.5%–0.75%, w/v final) for 15 min at 4 C. After centrifugation at 10,000 3 g, supernatants were saved as OMMenriched fractions, with the pellets containing mitoplasts. Mitoplasts were treated with 1% Triton X-100 and centrifuged at 100,000 3 g, and the resulting pellets and supernatants contained IMM- and matrix-enriched fractions, respectively. SDH native-complex analyses were performed on isolated mitochondria from HeLa cells. Native complexes were obtained by solubilization of mitochondria with 1.0% digitonin for 30 min on ice, followed by centrifugation at 16,000 3 g (4 C). The supernatant was collected and supplemented with 0.25% Coomassie blue G and protease inhibitors. Proteins were then separated on 4%–16% gradient native polyacrylamide gels (Invitrogen, Lyon, France).

SDH-dependent respiration was evaluated by adding a complex II inhibitor (5 mM malonate, final concentration), and OCR was measured before and after malonate addition. SDH activities were measured spectrophotometrically, as described previously (Benard et al., 2006). The protocol was slightly modified in order to use succinate concentrations ranging from 0 to 50 mM. Metabolomics were performed by Metabolomics & Fluxomics Facilities (MetaToul, Toulouse, France) based on methods described previously (Martano et al., 2015). HeLa cells were grown in the presence of glutamine on coverslips. At 80% confluence, coverslips were washed in cold milliQ water and soaked in cold quenching solution containing methanol:acetonitrile:H2O (2:2:1) plus 0.1% formic acid. Quenching solution also contained isotopic standards, except for blanks. Cells were harvested from coverslips using cell scrapers. Samples were sonicated for 30 s and frozen in liquid nitrogen prior to further analysis by ion chromatography-MS/MS or LChigh-resolution MS. Quantification and Statistical Analysis All values represent the mean ± SEM. Statistical analyses were performed using GraphPad Prism 6 software (GraphPad Software, La Jolla, CA, USA). Normality was determined using the D’Agostino normality test. Statistical tests used are described in the legend of each figure. Contact for Reagent and Resource Sharing Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Giovanni Be´nard ([email protected]).

SUPPLEMENTAL INFORMATION Proteomics Analyses Proteomics analyses were performed on whole-cells extracts. Enrichment of K-ε-GG peptides was performed using the PTMScan ubiquitin-remnant motif kit according to manufacturer’s protocol (Cell Signaling Technology, Danvers, MA, USA). Peptides were identified by tandem mass spectrometry nano-liquid chromatography (LC)-tandem mass spectrometry (MS/MS) analyses using an Ultimate 3000 RSLC Nano-UPHLC system (Thermo Fisher Scientific) coupled to a nanospray Q Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Fisher Scientific). Mascot, SEQUEST, and Amanda algorithms through Proteome Discoverer 1.4 software (Thermo Fisher Scientific) were used for protein identification in batch mode by searching against a UniProt Homo sapiens database (68,421 entries; Reference Proteome Set, release 2015_04; http://www.uniprot.org/). Two missed enzyme cleavages were allowed. Mass tolerances in MS and MS/ MS were set to 10 ppm and 0.02 Da. Gly-Gly addition to lysines was searched as dynamic modifications. Identification of mitochondrial proteins was performed using an Excel macro, and the functions of each identified mitochondrial protein were analyzed using the UniProt database (http:// www.uniprot.org/). Immunofluorescence Microscopy Immunofluorescence experiments were performed according a standard protocol using paraformaldehyde fixation and Triton X-100 permeabilization. We used the following primary antibodies: anti-HA and anti-TOM20 (Santa Cruz Biotechnology), CPTA1 (Abcam), K48-specific linkage (Millipore), and antiMyc (Roche). Mitochondrial staining with MitoTracker (Thermo Fisher Scientific) was performed according to manufacturer’s instructions. Images were acquired using a Zeiss microscope (AxioVision; Carl Zeiss, Oberkochen, Germany) with a 633 objective. Z sections (interval: 0.2 mm) covering the entire depth of the cell were acquired. Quantifications (levels of fluorescence and Pearson’s correlation coefficient) were performed using Zeiss colocalization software (Carl Zeiss).

Supplemental Information includes four figures and can be found with this article online at https://doi.org/10.1016/j.celrep.2018.05.013. ACKNOWLEDGMENTS Mito-DsRed was kindly provided by Dr. Roman Serrat and Ha-Ubiquitin by Dr. Olga Corti. MetaToul (Metabolomics & Fluxomics Facilities, Toulouse, France; https://www6.toulouse.inra.fr/metatoul), and its staff members (Lara Gales and Tony Palama) are gratefully acknowledged for performing metabolome analyses. MetaToul is part of the national infrastructure related to MetaboHUB-ANR-11-INBS-0010 (The French National Infrastructure for Metabolomics and Fluxomics; http://www.metabohub.fr). MetaToul is supported by grants from the Re´gion Midi-Pyre´ne´es, the European Regional Development Fund, the SICOVAL, the Infrastructures en Biologie Sante et Agronomie (IBiSa, France), the Centre National de la Recherche Scientifique (CNRS), and the Institut National de la Recherche Agronomique (INRA). This project was supported by ‘‘agence national de la recherche’’ (ANR; n ANR-17-CE140039-01). We thank Dr. Rodrigue Rossignol and Dr. Johan Garaude for providing comments on the manuscript. AUTHOR CONTRIBUTIONS J.L., S.S., A.M.I., D.L., and G.B. conceived the study. J.L., H.D.B., S.S., A.M.I., and G.B. performed the biochemical experiments. J.L. and S.M. designed and conceived the genetic constructs. J.-W.D. performed mass spectrometry analyses. C.L. and E.D. performed the activity and oxygraphy assays. G.B. wrote the manuscript along with contributions from all authors. DECLARATION OF INTERESTS The authors declare no competing interests.

Metabolomic Analyses, Respiration Assays, and SDHA Enzyme Assays Mitochondrial oxygen-consumption assays were performed using the highresolution respirometry system Oxygraph-2k (Oroboros, Innsbruck, Austria). Cellular respiration was measured from 1 3 106 to 5 3 106 cells/mL at 37 C.

Received: October 9, 2017 Revised: December 21, 2017 Accepted: May 3, 2018 Published June 5, 2018

Cell Reports 23, 2852–2863, June 5, 2018 2861

REFERENCES Akabane, S., Uno, M., Tani, N., Shimazaki, S., Ebara, N., Kato, H., Kosako, H., and Oka, T. (2016). PKA regulates PINK1 stability and Parkin recruitment to damaged mitochondria through phosphorylation of MIC60. Mol. Cell 62, 371–384. Azzu, V., and Brand, M.D. (2010). Degradation of an intramitochondrial protein by the cytosolic proteasome. J. Cell Sci. 123, 578–585. Benard, G., Faustin, B., Passerieux, E., Galinier, A., Rocher, C., Bellance, N., Delage, J.P., Casteilla, L., Letellier, T., and Rossignol, R. (2006). Physiological diversity of mitochondrial oxidative phosphorylation. Am. J. Physiol. Cell Physiol. 291, C1172–C1182. Benard, G., Bellance, N., Jose, C., Melser, S., Nouette-Gaulain, K., and Rossignol, R. (2010a). Multi-site control and regulation of mitochondrial energy production. Biochim. Biophys. Acta 1797, 698–709. Benard, G., Neutzner, A., Peng, G., Wang, C., Livak, F., Youle, R.J., and Karbowski, M. (2010b). IBRDC2, an IBR-type E3 ubiquitin ligase, is a regulatory factor for Bax and apoptosis activation. EMBO J. 29, 1458–1471. Bingol, B., Tea, J.S., Phu, L., Reichelt, M., Bakalarski, C.E., Song, Q., Foreman, O., Kirkpatrick, D.S., and Sheng, M. (2014). The mitochondrial deubiquitinase USP30 opposes parkin-mediated mitophagy. Nature 510, 370–375. Bonnen, P.E., Yarham, J.W., Besse, A., Wu, P., Faqeih, E.A., Al-Asmari, A.M., Saleh, M.A., Eyaid, W., Hadeel, A., He, L., et al. (2013). Mutations in FBXL4 cause mitochondrial encephalopathy and a disorder of mitochondrial DNA maintenance. Am. J. Hum. Genet. 93, 471–481. Bota, D.A., and Davies, K.J. (2002). Lon protease preferentially degrades oxidized mitochondrial aconitase by an ATP-stimulated mechanism. Nat. Cell Biol. 4, 674–680. Bragoszewski, P., Wasilewski, M., Sakowska, P., Gornicka, A., Bo¨ttinger, L., Qiu, J., Wiedemann, N., and Chacinska, A. (2015). Retro-translocation of mitochondrial intermembrane space proteins. Proc. Natl. Acad. Sci. USA 112, 7713–7718. Burchell, V.S., Nelson, D.E., Sanchez-Martinez, A., Delgado-Camprubi, M., Ivatt, R.M., Pogson, J.H., Randle, S.J., Wray, S., Lewis, P.A., Houlden, H., et al. (2013). The Parkinson’s disease-linked proteins Fbxo7 and Parkin interact to mediate mitophagy. Nat. Neurosci. 16, 1257–1265. Friedlander, R., Jarosch, E., Urban, J., Volkwein, C., and Sommer, T. (2000). A regulatory link between ER-associated protein degradation and the unfolded-protein response. Nat. Cell Biol. 2, 379–384. Gai, X., Ghezzi, D., Johnson, M.A., Biagosch, C.A., Shamseldin, H.E., Haack, T.B., Reyes, A., Tsukikawa, M., Sheldon, C.A., Srinivasan, S., et al. (2013). Mutations in FBXL4, encoding a mitochondrial protein, cause early-onset mitochondrial encephalomyopathy. Am. J. Hum. Genet. 93, 482–495. Gauss, R., Jarosch, E., Sommer, T., and Hirsch, C. (2006). A complex of Yos9p and the HRD ligase integrates endoplasmic reticulum quality control into the degradation machinery. Nat. Cell Biol. 8, 849–854. Gehrke, S., Wu, Z., Klinkenberg, M., Sun, Y., Auburger, G., Guo, S., and Lu, B. (2015). PINK1 and Parkin control localized translation of respiratory chain component mRNAs on mitochondria outer membrane. Cell Metab. 21, 95–108. Gold, V.A., Chroscicki, P., Bragoszewski, P., and Chacinska, A. (2017). Visualization of cytosolic ribosomes on the surface of mitochondria by electron cryotomography. EMBO Rep. 18, 1786–1800. Guan, K., Zheng, Z., Song, T., He, X., Xu, C., Zhang, Y., Ma, S., Wang, Y., Xu, Q., Cao, Y., et al. (2013). MAVS regulates apoptotic cell death by decreasing K48-linked ubiquitination of voltage-dependent anion channel 1. Mol. Cell. Biol. 33, 3137–3149. Haddad, D.M., Vilain, S., Vos, M., Esposito, G., Matta, S., Kalscheuer, V.M., Craessaerts, K., Leyssen, M., Nascimento, R.M., Vianna-Morgante, A.M., et al. (2013). Mutations in the intellectual disability gene Ube2a cause neuronal dysfunction and impair parkin-dependent mitophagy. Mol. Cell 50, 831–843.

2862 Cell Reports 23, 2852–2863, June 5, 2018

Haynes, C.M., Petrova, K., Benedetti, C., Yang, Y., and Ron, D. (2007). ClpP mediates activation of a mitochondrial unfolded protein response in C. elegans. Dev. Cell 13, 467–480. Haynes, C.M., Fiorese, C.J., and Lin, Y.F. (2013). Evaluating and responding to mitochondrial dysfunction: the mitochondrial unfolded-protein response and beyond. Trends Cell Biol. 23, 311–318. Heo, J.M., Livnat-Levanon, N., Taylor, E.B., Jones, K.T., Dephoure, N., Ring, J., Xie, J., Brodsky, J.L., Madeo, F., Gygi, S.P., et al. (2010). A stress-responsive system for mitochondrial protein degradation. Mol. Cell 40, 465–480. Hershko, A., and Ciechanover, A. (1998). The ubiquitin system. Annu. Rev. Biochem. 67, 425–479. Huemer, M., Karall, D., Schossig, A., Abdenur, J.E., Al Jasmi, F., Biagosch, C., Distelmaier, F., Freisinger, P., Graham, B.H., Haack, T.B., et al. (2015). Clinical, morphological, biochemical, imaging and outcome parameters in 21 individuals with mitochondrial maintenance defect related to FBXL4 mutations. J. Inherit. Metab. Dis. 38, 905–914. Hwang, S., Hartman, I.Z., Calhoun, L.N., Garland, K., Young, G.A., Mitsche, M.A., McDonald, J., Xu, F., Engelking, L., and DeBose-Boyd, R.A. (2016). Contribution of accelerated degradation to feedback regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase and cholesterol metabolism in the liver. J. Biol. Chem. 291, 13479–13494. Jin, S.M., and Youle, R.J. (2013). The accumulation of misfolded proteins in the mitochondrial matrix is sensed by PINK1 to induce PARK2/Parkin-mediated mitophagy of polarized mitochondria. Autophagy 9, 1750–1757. Kane, L.A., Lazarou, M., Fogel, A.I., Li, Y., Yamano, K., Sarraf, S.A., Banerjee, S., and Youle, R.J. (2014). PINK1 phosphorylates ubiquitin to activate Parkin E3 ubiquitin ligase activity. J. Cell Biol. 205, 143–153. Karbowski, M., and Youle, R.J. (2011). Regulating mitochondrial outer membrane proteins by ubiquitination and proteasomal degradation. Curr. Opin. Cell Biol. 23, 476–482. Karbowski, M., Neutzner, A., and Youle, R.J. (2007). The mitochondrial E3 ubiquitin ligase MARCH5 is required for Drp1 dependent mitochondrial division. J. Cell Biol. 178, 71–84. Kaufman, D.M., Wu, X., Scott, B.A., Itani, O.A., Van Gilst, M.R., Bruce, J.E., and Crowder, C.M. (2017). Ageing and hypoxia cause protein aggregation in mitochondria. Cell Death Differ. 24, 1730–1738. Koyano, F., Okatsu, K., Kosako, H., Tamura, Y., Go, E., Kimura, M., Kimura, Y., Tsuchiya, H., Yoshihara, H., Hirokawa, T., et al. (2014). Ubiquitin is phosphorylated by PINK1 to activate parkin. Nature 510, 162–166. Kubbutat, M.H., Jones, S.N., and Vousden, K.H. (1997). Regulation of p53 stability by Mdm2. Nature 387, 299–303. Leboucher, G.P., Tsai, Y.C., Yang, M., Shaw, K.C., Zhou, M., Veenstra, T.D., Glickman, M.H., and Weissman, A.M. (2012). Stress-induced phosphorylation and proteasomal degradation of mitofusin 2 facilitates mitochondrial fragmentation and apoptosis. Mol. Cell 47, 547–557. Lehmann, G., Udasin, R.G., and Ciechanover, A. (2016). On the linkage between the ubiquitin-proteasome system and the mitochondria. Biochem. Biophys. Res. Commun. 473, 80–86. Lesnik, C., Cohen, Y., Atir-Lande, A., Schuldiner, M., and Arava, Y. (2014). OM14 is a mitochondrial receptor for cytosolic ribosomes that supports cotranslational import into mitochondria. Nat. Commun. 5, 5711. Margineantu, D.H., Emerson, C.B., Diaz, D., and Hockenbery, D.M. (2007). Hsp90 inhibition decreases mitochondrial protein turnover. PLoS ONE 2, e1066. Martano, G., Delmotte, N., Kiefer, P., Christen, P., Kentner, D., Bumann, D., and Vorholt, J.A. (2015). Fast sampling method for mammalian cell metabolic analyses using liquid chromatography-mass spectrometry. Nat. Protoc. 10, 1–11. Melser, S., Chatelain, E.H., Lavie, J., Mahfouf, W., Jose, C., Obre, E., Goorden, S., Priault, M., Elgersma, Y., Rezvani, H.R., et al. (2013). Rheb regulates mitophagy induced by mitochondrial energetic status. Cell Metab. 17, 719–730. €nch, C., and Harper, J.W. (2016). Mitochondrial unfolded protein response Mu controls matrix pre-RNA processing and translation. Nature 534, 710–713.

Nakamura, N., Kimura, Y., Tokuda, M., Honda, S., and Hirose, S. (2006). MARCH-V is a novel mitofusin 2- and Drp1-binding protein able to change mitochondrial morphology. EMBO Rep. 7, 1019–1022. Narendra, D., Tanaka, A., Suen, D.F., and Youle, R.J. (2008). Parkin is recruited selectively to impaired mitochondria and promotes their autophagy. J. Cell Biol. 183, 795–803. Perciavalle, R.M., Stewart, D.P., Koss, B., Lynch, J., Milasta, S., Bathina, M., Temirov, J., Cleland, M.M., Pelletier, S., Schuetz, J.D., et al. (2012). Antiapoptotic MCL-1 localizes to the mitochondrial matrix and couples mitochondrial fusion to respiration. Nat. Cell Biol. 14, 575–583. Schwartz, A.L., Trausch, J.S., Ciechanover, A., Slot, J.W., and Geuze, H. (1992). Immunoelectron microscopic localization of the ubiquitin-activating enzyme E1 in HepG2 cells. Proc. Natl. Acad. Sci. USA 89, 5542–5546. Segref, A., Kevei, E´., Pokrzywa, W., Schmeisser, K., Mansfeld, J., Livnat-Levanon, N., Ensenauer, R., Glickman, M.H., Ristow, M., and Hoppe, T. (2014). Pathogenesis of human mitochondrial diseases is modulated by reduced activity of the ubiquitin/proteasome system. Cell Metab. 19, 642–652. Smith, M.H., Ploegh, H.L., and Weissman, J.S. (2011). Road to ruin: targeting proteins for degradation in the endoplasmic reticulum. Science 334, 1086– 1090. Sun, R.C., and Denko, N.C. (2014). Hypoxic regulation of glutamine metabolism through HIF1 and SIAH2 supports lipid synthesis that is necessary for tumor growth. Cell Metab. 19, 285–292. Taylor, E.B., and Rutter, J. (2011). Mitochondrial quality control by the ubiquitin-proteasome system. Biochem. Soc. Trans. 39, 1509–1513.

Udeshi, N.D., Svinkina, T., Mertins, P., Kuhn, E., Mani, D.R., Qiao, J.W., and Carr, S.A. (2013). Refined preparation and use of anti-diglycine remnant (K-ε-GG) antibody enables routine quantification of 10,000s of ubiquitination sites in single proteomics experiments. Mol. Cell. Proteomics 12, 825–831. Wagner, S.A., Beli, P., Weinert, B.T., Nielsen, M.L., Cox, J., Mann, M., and Choudhary, C. (2011). A proteome-wide, quantitative survey of in vivo ubiquitylation sites reveals widespread regulatory roles. Mol. Cell. Proteomics 10, M111.013284. Wagner, S.A., Beli, P., Weinert, B.T., Scho¨lz, C., Kelstrup, C.D., Young, C., Nielsen, M.L., Olsen, J.V., Brakebusch, C., and Choudhary, C. (2012). Proteomic analyses reveal divergent ubiquitylation site patterns in murine tissues. Mol. Cell. Proteomics 11, 1578–1585. Wrobel, L., Topf, U., Bragoszewski, P., Wiese, S., Sztolsztener, M.E., Oeljeklaus, S., Varabyova, A., Lirski, M., Chroscicki, P., Mroczek, S., et al. (2015). Mistargeted mitochondrial proteins activate a proteostatic response in the cytosol. Nature 524, 485–488. Wu, X., Li, L., and Jiang, H. (2016). Doa1 targets ubiquitinated substrates for mitochondria-associated degradation. J. Cell Biol. 213, 49–63. Xu, S., Peng, G., Wang, Y., Fang, S., and Karbowski, M. (2011). The AAAATPase p97 is essential for outer mitochondrial membrane protein turnover. Mol. Biol. Cell 22, 291–300. Yang, Y., Fang, S., Jensen, J.P., Weissman, A.M., and Ashwell, J.D. (2000). Ubiquitin protein ligase activity of IAPs and their degradation in proteasomes in response to apoptotic stimuli. Science 288, 874–877.

Teng, H., Wu, B., Zhao, K., Yang, G., Wu, L., and Wang, R. (2013). Oxygensensitive mitochondrial accumulation of cystathionine b-synthase mediated by Lon protease. Proc. Natl. Acad. Sci. USA 110, 12679–12684.

Yoshizawa, T., Karim, M.F., Sato, Y., Senokuchi, T., Miyata, K., Fukuda, T., Go, C., Tasaki, M., Uchimura, K., Kadomatsu, T., et al. (2014). SIRT7 controls hepatic lipid metabolism by regulating the ubiquitin-proteasome pathway. Cell Metab. 19, 712–721.

Thomas, R.E., Andrews, L.A., Burman, J.L., Lin, W.Y., and Pallanck, L.J. (2014). PINK1-Parkin pathway activity is regulated by degradation of PINK1 in the mitochondrial matrix. PLoS Genet. 10, e1004279.

Zhao, Q., Wang, J., Levichkin, I.V., Stasinopoulos, S., Ryan, M.T., and Hoogenraad, N.J. (2002). A mitochondrial specific stress response in mammalian cells. EMBO J. 21, 4411–4419.

Cell Reports 23, 2852–2863, June 5, 2018 2863