A mitochondrial FUNDC1/HSC70 interaction organizes the proteostatic stress response at the risk of cell morbidity

Article27 December 2018free access Source DataTransparent process A mitochondrial FUNDC1/HSC70 interaction organizes the proteostatic stress response at the risk of cell morbidity Yanjun Li State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yanhong Xue National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Xiaojun Xu College of Life Science and Technology, HuaZhong University of Science and Technology, Wuhan, Hubei, China Search for more papers by this author Guopeng Wang School of Life Sciences, Peking University, Beijing, China Search for more papers by this author Yiqun Liu School of Life Sciences, Peking University, Beijing, China Search for more papers by this author Hao Wu State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Wenhui Li State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yueying Wang State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Ziheng Chen State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Weilin Zhang State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yushan Zhu orcid.org/0000-0002-5648-0416 College of Life Sciences, Nankai University, Tianjin, China Search for more papers by this author Wei Ji National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Tao Xu orcid.org/0000-0002-8260-9754 National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China College of Life Science and Technology, HuaZhong University of Science and Technology, Wuhan, Hubei, China Search for more papers by this author Lei Liu Corresponding Author [email protected] State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Quan Chen Corresponding Author [email protected] orcid.org/0000-0001-7539-8728 State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China College of Life Sciences, Nankai University, Tianjin, China University of Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yanjun Li State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yanhong Xue National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Xiaojun Xu College of Life Science and Technology, HuaZhong University of Science and Technology, Wuhan, Hubei, China Search for more papers by this author Guopeng Wang School of Life Sciences, Peking University, Beijing, China Search for more papers by this author Yiqun Liu School of Life Sciences, Peking University, Beijing, China Search for more papers by this author Hao Wu State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Wenhui Li State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yueying Wang State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Ziheng Chen State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Weilin Zhang State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Yushan Zhu orcid.org/0000-0002-5648-0416 College of Life Sciences, Nankai University, Tianjin, China Search for more papers by this author Wei Ji National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Tao Xu orcid.org/0000-0002-8260-9754 National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China College of Life Science and Technology, HuaZhong University of Science and Technology, Wuhan, Hubei, China Search for more papers by this author Lei Liu Corresponding Author [email protected] State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Search for more papers by this author Quan Chen Corresponding Author [email protected] orcid.org/0000-0001-7539-8728 State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China College of Life Sciences, Nankai University, Tianjin, China University of Chinese Academy of Sciences, Beijing, China Search for more papers by this author Author Information Yanjun Li1, Yanhong Xue2, Xiaojun Xu3, Guopeng Wang4, Yiqun Liu4, Hao Wu1, Wenhui Li1, Yueying Wang1, Ziheng Chen1, Weilin Zhang1, Yushan Zhu5, Wei Ji2, Tao Xu2,3, Lei Liu *,1 and Quan Chen *,1,5,6 1State Key Laboratory of Membrane Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China 2National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China 3College of Life Science and Technology, HuaZhong University of Science and Technology, Wuhan, Hubei, China 4School of Life Sciences, Peking University, Beijing, China 5College of Life Sciences, Nankai University, Tianjin, China 6University of Chinese Academy of Sciences, Beijing, China *Corresponding author. Tel: +86 10 64807329; E-mail: [email protected] author. Tel: +86 10 64807321; E-mail: [email protected] EMBO J (2019)38:e98786https://doi.org/10.15252/embj.201798786 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract Both protein quality and mitochondrial quality are vital for the cellular activity, and impaired proteostasis and mitochondrial dysfunction are common etiologies of aging and age-related disorders. Here, we report that the mitochondrial outer membrane protein FUNDC1 interacts with the chaperone HSC70 to promote the mitochondrial translocation of unfolded cytosolic proteins for degradation by LONP1 or for formation of non-aggresomal mitochondrion-associated protein aggregates (MAPAs) upon proteasome inhibition in cultured human cells. Integrative approaches including csCLEM, Apex, and biochemical analysis reveal that MAPAs contain ubiquitinated cytosolic proteins, autophagy receptor p62, and mitochondrial proteins. MAPAs are segregated from mitochondria in a FIS1-dependent manner and can subsequently be degraded via autophagy. Although the FUNDC1/HSC70 pathway promotes the degradation of unfolded cytosolic proteins, excessive accumulation of unfolded proteins on the mitochondria prior to MAPA formation impairs mitochondrial integrity and activates AMPK, leading to cellular senescence. We suggest that human mitochondria organize cellular proteostatic response at the risk of their own malfunction and cell lethality. Synopsis An interaction between mitochondrial outer membrane protein FUNDC1 and chaperone HSC70 promotes degradation or aggregation of unfolded cytosolic proteins on the mitochondria, thereby maintaining cellular proteostasis but reducing mitochondrial fitness. Upon proteostatic stress, FUNDC1 interacts with HSC70 to promote the mitochondrial translocation of proteasomal clients, which are subsequently degraded by LONP1. Mitochondrial translocation of proteasomal clients promotes the formation of mitochondrion-associated protein aggregates (MAPAs). MAPAs are distinct from aggresomes and contain ubiquitinated proteins, P62 and mitochondrial proteins. FUNDC1 mediates the autophagic degradation of MAPAs, linking mitophagy and aggrephagy. FUNDC1/HSC70-mediated mitochondrial translocation of proteasomal clients prior to MAPA formation contributes to cell senescence. Introduction Both protein quality and mitochondrial quality are vital for cellular activity and health. Protein quality is maintained through a coordinated network involving chaperones and two proteolytic systems, the ubiquitin–proteasome and autophagy–lysosome systems (Mizushima et al, 2008; Hartl et al, 2011; Koga et al, 2011). Proteasomes and autophagy share substrates, effectors, and regulators and compensate for one another in times of need (Kaushik & Cuervo, 2015). Chaperones, including heat-shock protein 70 (HSP70) family proteins, which assist the folding/unfolding, assembly/disassembly, and membrane translocation of their clients, as well as sorting them for degradation through the proteasomal or autophagic pathways (Kaushik & Cuervo, 2015). Misfolded proteins are either refolded, degraded, or sequestered to distinct cellular sites, such as aggresomes at the microtubule-organizing center, in a coordinated fashion (Johnston et al, 1998). When the proteostatic network becomes compromised during aging or stress conditions, aberrant proteins tend to accumulate as toxic aggregates, a process associated with numerous neurodegenerative diseases and other disorders (Lim & Yue, 2015). Mitochondria have their own genome (mtDNA), which encodes < 1% of the mitochondrial proteome. To maintain their vital functions for energy production, mitochondria require the import of nuclear DNA-encoded proteins via the TOM/TIM complexes, coordinated expression and integration of nuclear DNA- and mtDNA-encoded OXPHOS subunits (Ryan & Hoogenraad, 2007), and sophisticated quality control systems at both the protein and organellar levels. Mitochondria have their own independent protein quality control system, including their own chaperones and proteases, such as Lon and ClpXP in the mitochondrial matrix (Voos, 2009; Fischer et al, 2012). At the organellar level, functionally compromised mitochondria can be selectively removed by either ubiquitin- or receptor-mediated mitophagy to ensure mitochondrial quality (Youle & Narendra, 2011; Khaminets et al, 2016). Defective mitochondrial quality control leads to the accumulation of dysfunctional mitochondria, an etiological factor commonly observed in aging and age-related disorders. Apparently, proteostatic impairment and mitochondrial dysfunction work in concert and reinforce each other under stress conditions or during aging (Ross et al, 2015). However, the underlying mechanisms and the primary cause for age-related disorders remain ill-defined. In an effort to address the mechanism of p62 translocation toward mitochondria, we have found that FUNDC1, a previously described mitophagy receptor, interacts with the cytosolic chaperone protein HSC70 to organize cellular proteostasis when proteasome is inhibited. Despite this, excessive mitochondrial accumulation of unfolded proteins prior to the formation of mitochondrion-associated protein aggregates (MAPAs) impairs mitochondrial integrity and activates AMPK, leading to cellular senescence. Results FUNDC1 interacts with HSC70 and promotes the mitochondrial recruitment of the proteasomal clients We have previously shown that FUNDC1 is a highly conserved mitochondrial membrane protein that harbors a LC3-interacting region (LIR) to directly interact with LC3 and to mediate hypoxia-induced mitophagy (Liu et al, 2012a). Transient expression of FUNDC1 resulted in the mitochondrial localization of LC3 in a manner that depended on the LIR motif (Appendix Fig S1). Unexpectedly, it promoted mitochondrial recruitment of p62/SQSTM1, a multifunctional protein which binds ubiquitin through its UBA domain (Ciani et al, 2003; Pankiv et al, 2007), in a LIR-independent manner (Fig 1A and Appendix Fig S2A). Additionally, a p62 mutant lacking the UBA domain was less robustly recruited to mitochondria than the full-length p62 in cells overexpressing FUNDC1 or FUNDC1-ΔLIR (compare Appendix Fig S2A and B), suggesting that ubiquitin-modified proteins may be responsible for such recruitment. Indeed, expression of FUNDC1 or FUNDC1-∆LIR resulted in the mitochondrial recruitment of ubiquitinated proteins (Fig 1B). Immunoblotting analysis of mitochondria isolated from HeLa cells revealed that overexpression of FUNDC1 or FUNDC1-∆LIR resulted in the mitochondrial accumulation and consequent global elevation of ubiquitinated proteins (Fig 1C). FUNDC1 or FUNDC1-∆LIR overexpression also resulted in the mitochondrial accumulation of p62, although had no visible effect on the level of this protein in total cell extract and cytosolic fractions (Fig 1C), likely due to the high abundance of p62 in cells. Figure 1. FUNDC1 interacts with HSC70 and promotes the mitochondrial recruitment of the proteasomal clients HeLa cells were co-transfected with Myc empty vector (Myc), Myc-tagged FUNDC1 (F1-Myc) or Myc-tagged FUNDC1-ΔLIR mutant (ΔLIR-Myc), and Mito-OM-cherry, which localizes to the mitochondrial outer membrane, and immunostained with fluorescent anti-p62 and anti-Myc antibodies. Scale bar = 10 μm. HeLa cells were co-transfected with Myc, F1-Myc or ΔLIR-Myc, and Mito-OM-cherry and immunostained with anti-ubiquitin (Ub) and anti-Myc antibodies. Scale bar = 10 μm. Fractionation analysis of HeLa cells transfected with Myc, F1-Myc, or ΔLIR-Myc. The loading controls (L.C.) were α-tubulin for total cell lysate (TCL) and cytosolic fraction (Cyto), and HSP60 for mitochondrial fraction (Mito). HeLa cells transfected with Myc or Myc/His6 double-tagged F1-Myc-His were lysed and precipitated with Ni-NTA beads. Samples were then subjected to SDS–PAGE. The differential band (upper arrow) was isolated and subjected to mass spectral analysis. The identified proteins with a score higher than 50 are listed. Co-IP of endogenous HSC70 with F1-Myc or ΔLIR-Myc. α-Myc, anti-Myc antibody. Co-IP of HSC70 with FUNDC1. α-F1, anti-FUNDC1 antibody. rIgG, control rabbit IgG. Co-IP analysis of endogenous FUNDC1 with Flag-HSP70 or Flag-HSC70. α-Flag, anti-Flag antibody. Flag,empty vector. HeLa cells were co-transfected with Myc, F1-Myc or ΔLIR-Myc, and Mito-OM-cherry and immunostained with anti-HSC70 and anti-Myc antibodies. Scale bar = 10 μm. Fractionation analysis of HeLa cells transfected with Myc or F1-Myc. HeLa cells were co-transfected with Ub-R-GFP and Myc or F1-Myc and immunostained with fluorescent anti-Myc and anti-GFP (α-GFP) antibodies. The α-GFP signals in F1-Myc-transfected cells were lowered to avoid the overexposure of the mitochondrially localized R-GFP, making such signals of the equally distributed R-GFP invisible. Green arrowheads show the mitochondrially localized R-GFP proteins, which were more abundant than the equally distributed R-GFP and showed stronger R-GFP fluorescence, and white arrowheads show the mitochondrially localized R-GFP proteins without stronger R-GFP fluorescence, although abundant. Scale bar = 10 μm. Fractionation analysis of HeLa cells co-transfected with Ub-R-GFP and Myc or F1-Myc. HeLa cells were co-transfected with Ub-R-GFP and F1-Myc. Mitochondria were isolated and then treated with or without proteinase K (Pro. K), or proteinase K in combination with digitonin or Triton X-100 (TX-100). Samples were then subjected to immunoblot analysis. HSP60, TIM23, and TOM20 are representatives of mitochondrial matrix-, mitochondrial inner membrane-, and mitochondrial outer membrane-localized proteins, respectively. Fractionation analysis of HeLa cells stably transfected with control (Ctrl) shRNA vector or HSC70 shRNA and transiently co-transfected with Ub-R-GFP, Flag or Flag-HSC70 (shRNA-resistant), and Myc, or F1-Myc, as indicated. Data information: In panels (C and I), three replicates were analyzed using Student's two-tailed t-test and are presented as the means ± SEM. *P < 0.05; **P < 0.01; NS, non-significant. Source data are available online for this figure. Source Data for Figure 1 [embj201798786-sup-0003-SDataFig1.JPG] Download figure Download PowerPoint We speculated that FUNDC1 may interact with other molecules for such recruitments. Mass spectral analysis of the proteins co-precipitated with FUNDC1 revealed HSC70, a constitutively expressed cytosolic chaperone protein belonging to the HSP70 family (Liu et al, 2012b), as a novel FUNDC1-interacting partner (Fig 1D). Co-IP assay further confirmed the interaction between FUNDC1 and HSC70 (Fig 1E and F). The interaction seems specific, as HSP70, an inducible member of the HSP70 family that shares 85% homology with HSC70 (Liu et al, 2012b), has limited capacity to interact with FUNDC1 (Fig 1G). In cells transiently expressing FUNDC1 or FUNDC1-ΔLIR, HSC70 was recruited to the mitochondria (Fig 1H). Fractionation analysis also revealed that overexpression of FUNDC1 resulted in the mitochondrial recruitment of HSC70 (Fig 1I). CHIP is an HSC70-interacting E3 ubiquitin ligase which promotes the ubiquitination and subsequent proteasomal degradation of HSC70 clients (Connell et al, 2001). Confocal analysis revealed that CHIP was also recruited to mitochondria in response to FUNDC1 or FUNDC1-ΔLIR overexpression (Fig EV1A). Knockdown of either HSC70 or CHIP diminished the FUNDC1-induced mitochondrial localization of ubiquitin and p62 (Appendix Figs S3 and S4). These data suggest that overexpressed FUNDC1 may recruit HSC70 and its unfolded clients, including proteasomal substrates, to mitochondria; the client proteins are then subjected to ubiquitination by CHIP and hence endowed with the ability to attract p62. Click here to expand this figure. Figure EV1. HSC70, but not CHIP, is required for the FUNDC1 overexpression-resulted mitochondrial accumulation of proteasomal clients HeLa cells were co-transfected with Flag-CHIP, Mito-OM-cherry, and Myc, F1-Myc, or ΔLIR-Myc and immunostained with anti-Flag and anti-Myc antibodies. Scale bar = 10 μm. Cells transfected with GFP or Ub-R-GFP were lysed and subjected to Co-IP analysis. The samples were immunoprecipitated with rabbit anti-GFP antibodies (Rb α-GFP) and immunoblotted using mouse anti-HSC70 or mouse anti-GFP antibodies. HeLa cells co-transfected with Flag-HSC70 or Flag-HSC70-ΔPBD and Ub-R-GFP or control empty vectors were lysed and subjected to Co-IP analysis. HeLa cells were co-transfected with Myc or F1-Myc and GFP or Ub-R-GFP and then immunostained with fluorescent anti-GFP and anti-Myc antibodies. Scale bar = 10 μm. HeLa cells stably transfected with Ctrl, HSC70, or CHIP shRNA and transiently co-transfected with Ub-R-GFP and Myc or F1-Myc were immunostained with α-GFP and anti-Myc antibodies. Scale bar = 10 μm. Fractionation analysis of the HeLa cells stably transfected with Ctrl, HSC70, or CHIP shRNA and transiently co-transfected with Ub-R-GFP and Myc or F1-Myc. Download figure Download PowerPoint To prove this, Ub-R-GFP, a GFP-based reporter for proteasomal degradation, was employed for the following study. The fused ubiquitin moiety of Ub-R-GFP will be cleaved in cells, producing an N-end rule substrate R-GFP (Dantuma et al, 2000). We found that R-GFP, but not stable GFP, can be recognized by HSC70, depending on the latter's peptide-binding domain (PBD; Fig EV1B and C), a region required for the binding of HSC70 clients (Fourie et al, 1994). Additionally, FUNDC1 overexpression resulted in the recruitment of R-GFP proteins to FUNDC1-labeled mitochondria as expected; these R-GFP proteins could be clearly visualized by immunostaining using anti-GFP-antibody, although some of their own fluorescence were undetectable (Figs 1J and EV1D), which is likely due to their aberrant conformation. Subcellular fractionation analysis also revealed that FUNDC1 overexpression resulted in the accumulation of the R-GFP proteins and p53 and β-catenin, two endogenous proteasomal substrates, in mitochondria (Fig 1K). Proteinase K protection assays showed that the mitochondrially recruited R-GFP, β-catenin, and p53 proteins showed a certain degree of resistance to digestion even when treated with digitonin, which permeabilizes the outer membrane of mitochondria, but not when treated with Triton X-100, which permeabilizes both outer and inner mitochondrial membranes (Fig 1L). This indicates that a portion of the recruited client proteins were imported into the mitochondrial matrix. The mitochondrially recruited p62 and ubiquitinated proteins, however, showed hardly any resistance to proteinase K digestion (Fig 1L), suggesting that most of these proteins were on the surface of mitochondria. Additionally, we found that FUNDC1-induced mitochondrial recruitment of R-GFP was diminished in HSC70-knockdown cells and was partially restored by reintroduction of shRNA-resistant Flag-HSC70 (Figs 1M, and EV1E and F). CHIP knockdown, however, had no obvious effect on the mitochondrial recruitment of R-GFP caused by FUNDC1 overexpression (Fig EV1E and F). These data suggest that HSC70 may reroute its client proteins from fast proteasomal degradation to mitochondria via interaction with the overexpressed FUNDC1. FUNDC1 and HSC70 are required for the translocation of unfolded cytosolic proteins to mitochondria in proteasome-inhibited cells Given the critical role of HSC70 in recognition of cytosolic unwanted proteins, we hypothesized that the interaction between FUNDC1 and HSC70 may be the missing link between mitochondrial and cytosolic proteostasis network. We thus checked whether the cytosolic proteasomal clients could be recruited to mitochondria in a FUNDC1-/HSC70-dependent fashion under proteostatic stress conditions. In R-GFP-expressing cells treated with the proteasome inhibitor MG132, the accumulation of R-GFP proteins could be visualized by immunostaining using anti-GFP antibody (Fig EV2A, upper rows). The mitochondrially recruited R-GFP proteins were discernible only when the cytosolic proteins were extracted by digitonin (Fig EV2A, lower rows), but their own fluorescence was undetectable. Fractionation analysis also revealed that R-GFP, p53, and β-catenin proteins were recruited to mitochondria in the presence of the proteasome inhibitors MG132 or lactacystin (Figs 2A and EV2B). Proteinase K protection assays showed that most of the R-GFP and some of p53 and β-catenin proteins that were recruited to mitochondria were imported into the mitochondrial matrix (Fig 2B). The mitochondrially localized R-GFP proteins isolated from TOM20-, TOM22-, and TOM70-knockdown cells were much less resistant to proteinase digestion compared to those isolated from control cells (Figs 2C and EV2C), suggesting that TOM complex is required for the import of the proteasomal clients into mitochondria. The mitochondrial accumulation of R-GFP upon MG132 treatment was augmented in the LONP1-knockdown cells, and this was reversed by reintroduction of shRNA-resistant LONP1 (Figs 2D and EV2D), suggesting that LONP1 may degrade the proteasomal clients imported into the mitochondrial matrix. These results are in agreement with the findings in yeast (Ruan et al, 2017). Click here to expand this figure. Figure EV2. R-GFP proteins were recruited to mitochondria in proteasome-inhibited cells HeLa cells co-transfected with Ub-R-GFP and Mito-OM-cherry were treated with DMSO or 10 μM MG132 for 8 h. Cells were then treated with digitonin, or not, and then fixed and immunostained with α-GFP. Scale bar = 10 μm. Fractionation analysis of HeLa cells transfected with Ub-R-GFP and treated with lactacystin as indicated. Quantification of Fig 2C. Quantification of the R-GFP proteins in the Mito isolated from the MG132-treated cells in Fig 2D. Quantification of Fig 2E. Fractionation analysis of the wild-type (WT) and FUNDC1-knockout (F1−/−) HeLa cells transfected with Ub-R-GFP and treated with DMSO or MG132. Quantification of Fig 2F. Quantification of Fig 2G. Graphic description of the in vitro mitochondrial import assay. RT, room temperature. Quantification of Fig 2J. Quantification of Fig 2K. Data information: In panels (C-E, G, H, J, and K) three replicates were analyzed using Student's two-tailed t-test and are presented as the means ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; NS, non-significant. Download figure Download PowerPoint Figure 2. FUNDC1/HSC70 interaction is required for the mitochondrial translocation of proteasomal clients under stress conditions Fractionation analysis of the HeLa cells transfected with Ub-R-GFP and treated with DMSO (control vehicle) or 10 μM MG132 for 4 h. Proteinase K protection analysis of the mitochondria isolated from the HeLa cells transfected with Ub-R-GFP and treated with MG132. Proteinase K protection analysis of the mitochondria isolated from HeLa cells co-transfected with Ub-R-GFP and Ctrl, TOM20, TOM22, or TOM70 shRNA and treated with MG132. Fractionation analysis of HeLa cells stably transfected with Ctrl or LONP1 shRNA and transiently co-transfected with Ub-R-GFP and Myc or LONP1 (shRNA-resistant) and then treated with DMSO or MG132. Fractionation analysis of HeLa cells stably transfected with Ctrl shRNA, F1 shRNA, or F1 shRNA together with shRNA-resistant F1-Myc or ΔLIR-Myc, and transiently transfected with Ub-R-GFP and then treated with DMSO or MG132. Fractionation analysis of HeLa cells stably transfected with Ctrl or HSC70 shRNA and transiently co-transfected with Ub-R-GFP and Flag or shRNA-resistant Flag-HSC70 and then treated with DMSO or MG132. Fractionation analysis of HeLa cells stably transfected with Ctrl or HSP70 shRNA and transiently co-transfected with Ub-R-GFP and Flag or shRNA-resistant Flag-HSP70 and then treated with DMSO or MG132. Mitochondria were isolated from the untransfected cells, cytosols were isolated from the cells transfected with GFP or Ub-R-GFP and treated with DMSO or MG132 for 8 h, and the isolated mitochondria and cytosols were then subjected to in vitro protein import analysis. Mitochondria isolated from the untransfected cells were resuspended in the cytosol isolated from the cells transfected with Ub-R-GFP and incubated at RT for 1 hour, and the mitochondria were then isolated from the in vitro system and then treated with or without proteinase K and subjected to immunoblot analysis. Mitochondria were isolated from the cells transfected with Ctrl shRNA, F1 shRNA, or F1 shRNA together with F1-Myc, cytosols were isolated from the cells transfected with Ub-R-GFP, and the isolated mitochondria and cytosols were then subjected to in vitro protein import analysis. Cytosol isolated from the cells transfected with Ub-R-GFP was mixed with the cytosols isolated from the cells transfected with Ctrl shRNA, HSC70 shRNA, or HSC70 shRNA together with Flag-HSC70, mitochondria were isolated from the untransfected cells, and the isolated mitochondria and cytosol mixtures were then subjected to in vitro protein import analysis. Source data are available online for this figure. Source Data for Figure 2 [embj201798786-sup-0004-SDataFig2.JPG] Download figure Download PowerPoint We next addressed the question of whether FUNDC1 and HSC70 are required for the mitochondrial translocation of the proteasomal clients. First, we found that the MG132-induced mitochondrial accumulation of R-GFP was reduced in FUNDC1-knockdown or FUNDC1-knockout cells (Figs 2E, and EV2E and F). Reintroduction of shRNA-resistant FUNDC1 or FUNDC1-ΔLIR in FUNDC1-knockdown cells restored the mitochondrial accumulation of R-GFP (Figs 2E and EV2E). HSC70 knockdown also diminished the MG132-induced mitochondrial translocation of R-GFP, which was partially restored by expression of shRNA-resistant Flag-HSC70 (Figs 2F and EV2G). In contrast, HSP70 knockdown aggravated the MG132-induced mitochondrial translocation of R-GFP, and this was counteracted by expression of shRNA-resistant Flag-HSP70 (Figs 2G and EV2H). These data suggest that both FUNDC1 and HSC70 are required for the mitochondrial recruitment of proteasomal clients. HSP70, however, may prevent such recruitment. To further substantiate our finding that FUNDC1 interacts with HSC70 to recruit proteasomal clients to mitochondria, we used an in vitro assay in which we incubated the mitochondria isolated from untransfected cells with the cytosol isolated from cells expressing R-GFP or GFP (Fig EV2I). More R-GFP proteins were recruited to mitochondria compared to stable GFP proteins and were imported into mitochondria as judged by their resistance to protease digestion (Fig 2H and I). As expected, both FUNDC1 and HSC70 were required for the mitochondrial recruitment of R-GFP even in this in vitro system (Figs 2J and K, and EV2J and K). These results collectively suggested that in response to proteostatic stress, HSC70 may take its clients via interaction with FUNDC1 to mitochondria for impo