Research ArticleBIOCHEMISTRY

Mitochondrial protein import is regulated by p17/PERMIT to mediate lipid metabolism and cellular stress

See allHide authors and affiliations

Science Advances  11 Sep 2019:
Vol. 5, no. 9, eaax1978
DOI: 10.1126/sciadv.aax1978
  • Fig. 1 Induction of CerS1/C18-ceramide generation results in mitophagy.

    (A) Autophagic response evaluated by Cyto-ID in UM-SCC-22A-Tet On cells induced for expression of CerS1WT or CerS1H138A. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (B) Confocal images of UM-SCC-22A-Tet On cells induced for expression of CerS1WT (right) or CerS1H328A noncatalytic mutant (left) stained for LC3 (red) and mitochondria (MitoTracker, green). Images represent at least three independent experiments. Right panel shows quantification of colocalization estimated by calculating coefficient of colocalization (Rc) using Fiji Software. Scale bars, 100 μm (throughout the manuscript unless specifically noted). AU, arbitrary units. (C) Ceramide profiles of mitochondrial and soluble fractions of UM-SCC-22A cells upon treatment with vehicle or Tet were measured by lipidomics. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (D) Left: Confocal images of UM-SCC-22A-Tet On cells induced for expression of CerS1WT stained for ceramide (green) and mitochondria (Tom20, red). Images are representative of at least three independent experiments. Right: Quantification of the left panel. Colocolization correlation was estimated by calculating coefficient of colocalization using Fiji Software. Scale bars, 100 μm. (E) TEMs show fusion of mitochondria, gold-labeled with ceramide antibodies, in UM-SCC-22A-Tet On cells Tet-induced for expression of CerS1WT (+Tet) compared to untreated (−Tet) control. APH, autophagosome; M, mitochondria; sER, smooth ER. Top panel, 20,000× magnification; bottom panel, 80,000× magnification. Scale bars, 2 μm and 800 nm, respectively. Images represent at least three independent experiments. (F) Confocal images of UM-SCC-22A cells induced for expression of CerS1WT by SoSe and stained for ceramide (green) and mitochondria (Tom20, red). Images represent three independent experiments. Right: Quantification of left panel. Coefficient of colocalization was estimated using Fiji Software. Scale bars, 100 μm (throughout the manuscript unless specifically noted). (G) LC3 protein abundance in control (Scr) and CerS1 small interfering RNA (siRNA)–treated cells incubated with 5 μM SoSe for 3 hours. Images represent at least three independent experiments. (H) Quantification of confocal images of cells with silenced CerS1 or silenced LC3 UM-SCC-22A cells coloaded with 0.5 μM MTR for 60 min and LTG (0.5 μM for 20 min) upon treatment with 5 μM SoSe. Time points were selected to illustrate the onset and completion of mitochondrial digestion by autophagy. Data are means ± SD (n = 3 independent experiments, **P < 0.01). ns, not significant.

  • Fig. 2 Mitochondrial import of CerS1 from smooth ER mediates lethal mitophagy.

    (A) Confocal images of UM-SCC-22A cells induced for expression of CerS1 by Tet or SoSe and stained for CerS1 (green) and mitochondrial marker Tom20 (red) compared with nontreated cells. Yellow shows colocalization. Images represent three independent experiments. (B) Quantification of images in (A). Data are means ± SD (n = 3 different optical fields, **P < 0.01). (C) Distribution of CerS1 between mitochondria (M) and ER in UM-SCC-22A cells induced for CerS1 expression with Tet (left) or SoSe (right). Cox IV mitochondrial marker shows equal protein loading. Images are representative of three independent experiments. (D) Quantification of (C) at 72 hours for Tet-treated cells (left) and at 3 hours for SoSe-treated cells (right). Data are means ± SD (n = 3 independent experiments, **P < 0.01). (E) Quantification of confocal images of UM-SCC-22A cells treated with 5 μM SoSe for indicated periods of time and labeled with ER (PDI) and mitochondrial (Tom20) markers. Quantification of colocalization was performed with Fiji ImageJ software using at least three random fields of view from three independent experiments. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (F) Quantification of the confocal images of live MEFs, from CerS1WT and CerS1Top/Top animals, treated with SoSe for indicated periods of time. MEFs were coloaded with 0.5 μM MTR for 60 min and LTR (0.5 μM for 20 min) upon treatment with 5 μM SoSe. Time points selected to illustrate onset and completion of mitochondrial digestion by autophagy. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (G) Protein abundance of transiently expressed CerS1WT in MEFsTop/Top was detected by Western blotting. Actin was used as loading control. (H) Quantification of confocal microphotographs of live MEFs from animals expressing nonactive CerS1 mutant (MEFs CerS1Top/Top) and transiently transfected with empty vector control (CerS1Top/Top) or CerS1WT and treated with SoSe and stained with LTG and MTR. Time points selected to illustrate onset and completion of mitochondrial digestion by autophagy. Quantification was done using at least three images from three independent experiments. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (I) Mitochondrial versus ER localization of transiently expressed FLAG-tagged CerS1WT in CerS1Top/Top MEFs after induction with SoSe. Images represent at least three independent experiments.

  • Fig. 3 Insertion of CerS1 60 to 76 residues to CerS6 mediates mitochondrial import of CerS6.

    (A) Cellular lysates of UM-SCC-22A cells transfected with FLAG-tagged CerS6WT or CerS6 mutant with Hox domain replaced by the 60 to 76 amino acid sequence of CerS1 [CerSΔHox/CS1(60–76)] and treated with SoSe (5 μM, 3 hours) were differentially centrifuged, and resulting mitochondrial and ER fractions were analyzed using SDS–polyacrylamide gel electrophoresis (SDS-PAGE). Mitochondrial marker Cox IV was used as loading control. Images represent three independent experiments. (B) Quantification of confocal images of UM-SCC-22A treated as in (A) and labeled with FLAG antibody and mitochondrial marker Tom20. Quantification of FLAG-CerS6WT or FLAG-CerS6ΔHox/CS1(60–76) colocalization with Tom20 was performed using ImageJ software. Images represent at least three independent experiments. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (C) Abundance of C16-ceramide in mitochondrial and cytosolic fractions of shCerS1 UM-SCC-22A cells ectopically expressing CerS6WT and CerS6ΔHox/CS1(60–76) measured using mass spectrometry lipidomics. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (D) Abundance of LC3 lipidation in cells treated as in (C). Levels of FLAG and actin indicate equal expression levels of FLAG-tagged CerS6WT and CerS6ΔHox/CS1(60–76) proteins and equal protein loading. Images represent at least three independent experiments. (E) Quantification by ImageJ of confocal images of cells treated as in (C) and stained with LC3 and Tom20 antibodies. At least three random fields of view from at least three independent experiments were analyzed. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (F) CerS6-Hox translocated to mitochondria induces lethal mitophagy. Quantification of confocal images of live shCerS1 UM-SCC-22A cells ectopically expressing WT or ΔHox/CS1 (60 to 76) mutant of CerS6 coloaded with MTR (0.5 μM for 60 min) and LTR (0.5 μM for 20 min) upon treatment with 5 μM SoSe. Time points selected to illustrate onset and completion of mitochondrial digestion by autophagy. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (G) Abundance of CerS6WT-FLAG and mutant proteins in UM-SCC-22A cells was confirmed by Western blotting using anti-FLAG antibody. Empty vector (EV)–transfected cells were used as negative controls.

  • Fig. 4 p17/PERMIT mediates mitochondrial import of CerS1.

    (A) SDS-PAGE analysis of proteins bound to CerS1 in UM-SCC-22A cells treated with SoSe. Binding to IgG was used to account for nonspecific binding. Protein-containing gel bands were excised and analyzed by mass spectrometry. (B) p17/PERMIT silencing prevents CerS1 mitochondrial translocation. Distribution of CerS1 between mitochondrial and ER fractions in UM-SCC-22A control cells and cells with knocked-down p17/PERMIT expressing CerS1 upon Tet (top) or SoSe (5 μM, 3 hours) (bottom) induction. (C) Quantification of (B) using ImageJ. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (D) Quantification of confocal images of scrambled (Scr) control UM-SCC-22A cells and cells with silenced p17 treated with vehicle or SoSe (5 μM, 3 hours) and stained with CerS1 and Tom20 antibodies. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (E) Quantification of confocal images of live control or silenced p17 UM-SCC-22A cells coloaded with MTR (0.5 μM for 60 min) and LTR (0.5 μM for 20 min) upon treatment with 5 μM SoSe. Time points selected to illustrate onset and completion of mitochondrial digestion by mitophagy. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (F) Electron micrographs of UM-SCC-22A-Tet On cells with silenced p17 or Scr control induced for V5-tagged CerS1 and detected using gold-labeled anti-V5 antibody. Bottom panels are higher-magnification images of top panels. Arrows indicate junction between autophagosome and V5-CerS1–labeled mitochondria (Scr-siRNA, +Tet) and absence of junction and labeled mitochondria in cells with silenced p17 (p17 siRNA, +Tet). Labeling (V5-CerS1) localizes in sER. Images are representative of at least three independent experiments. Scale bars, 2 μm. (G) Live cell imaging was performed to detect transport of CerS1-expressing photoactivatable GFP (405 nm) from ER to mitochondria in response to SoSe in cells expressing Scr-shRNA or p17-shRNA. Mitochondria and ER were labeled with MitoTracker (far red, 633 nm) and red fluorescent protein (RFP) (565 nm), respectively. Colocalization of CerS1-GFP with ER versus mitochondria was measured using Fiji software. Rc is colocalization coefficient. (H) Mitochondrial potential measured by TMRE fluorescence in Scr control and in silenced p17 UM-SCC-22A cells treated with SoSe and compared to vehicle-treated control. Data are means ± SD (n = 3 independent experiments, **P < 0.01). (I) ATP measurements in control and p17-silenced UM-SCC-22A cells induced for CerS1 expression either with Tet (left, at 72 hours) or SoSe (right, at 3 hours). Data are means ± SD (n = 3 independent experiments, **P < 0.01).

  • Fig. 5 Amino acids 28R-29Y-30E of p17/PERMIT are key for CerS1 import to mitochondria.

    (A) shCerS1 UM-SCC-22A cells were transfected with FLAG-tagged CerS1WT or CerS1Δ60–76 and treated with SoSe. Precleared cell lysates (500 μg of protein) were immunoprecipitated with anti-p17 or anti-FLAG antibodies, and proteins were resolved and identified using SDS-PAGE and Western blotting. Images represent at least three independent experiments. (B) shCerS1 UM-SCC-22A cells were transfected with FLAG-tagged CerS6WT or CerS6ΔHox/CS1(60–76) and treated with SoSe. Precleared cell lysates (500 μg of protein) were immunoprecipitated with p17 or FLAG antibodies, and proteins were resolved and identified using SDS-PAGE and Western blotting. Images represent at least three independent experiments. (C) shp17 UM-SCC-22A cells were transfected with empty vector, p17WT, and p17 mutants, in which residues identified in fig. S6 were substituted with alanine (p17R28A;Y29A;E30A also labeled as p17RYE28-30AAA; p17R68A; p17R107A; p17R111A). Resulting lysates were subjected to IP with FLAG or CerS1 antibodies, and proteins were resolved and identified by SDS-PAGE and Western blotting. Images represent at least three independent experiments. (D) Stably knocked-down scrambled control or p17 shRNA UM-SCC-22A cells were transfected with either empty vector or p17 mutants as in (C). Resulting lysates were subjected to differential centrifugation, and distribution of CerS1 protein between mitochondrial and ER fractions was analyzed. Images represent at least three independent experiments. (E) Lysates of cells treated as in (D) were subjected to differential centrifugation to separate mitochondrial and ER fractions, and distribution of p17/PERMIT was analyzed by Western blotting. Images represent at least three independent experiments. (F) Quantification of live cell imaging of shp17 UM-SCC-22A cells transfected with either empty vector or p17WT or p17 mutants (p17R111A, p17RYE28-30AAA, and p17R68A) and coloaded with MTR (0.5 μM for 60 min) and LTR (0.5 μM for 20 min) upon treatment with 5 μM SoSe. Time points selected to illustrate onset and completion of mitochondrial digestion by autophagy. Data are means ± SD (n = 3 independent experiments, **P < 0.01).

  • Fig. 6 p17/PERMIT-mediated mitochondrial import of CerS1 is induced by Drp1 nitrosylation at C644.

    (A) Scr-siRNA–treated control and Drp1 siRNA UM-SCC-22A-Tet On cells were induced for CerS1WT expression for the indicated period of time. Resulting lysates were subjected to differential centrifugation, and V5-tagged CerS1 expression in mitochondrial and ER fractions was evaluated. Images represent three independent experiments. (B) UM-SCC-22A CerS1-Tet On cells were transiently transfected with empty vector or Drp1DN or Drp1C644A mutants and simultaneously induced for CerS1WT expression. Mitochondrial and ER fractions were analyzed for CerS1-V5 expression. Images represent three independent experiments. (C) Autophagic response evaluated by Cyto-ID in UM-SCC-22A-Tet On cells treated as in (B). Quantification was done using three random fields of view from three independent experiments. (D) UM-SCC-22A-Tet On cells were transiently transfected with empty vector or GFP-Drp1WT or GFP-Drp1C644A mutants and simultaneously induced for CerS1 expression with Tet. Precleared lysates were immunoprecipitated with V5, p17, or Drp1 antibodies. Interactions between p17 and CerS1 or between p17 and Drp1 were analyzed using SDS-PAGE and Western blotting. Images represent at least three independent experiments. (E) Effects of the ectopic expression of GFP-Drp1WT or GFP-Drp1C644A or GFP-GFP-Drp1C644W mutants on p17 interaction in the absence or presence of SoSe was detected by co-IP in UM-SCC-22A cells that express shRNA against endogenous Drp1 (right). Equal expression and immunoprecipitation of proteins are shown as input (left). Data shown represent three independent studies. (F) Genotyping of the p17/PERMIT−/− mice was performed by PCR using genomic DNA. (G) Mitophagy induction in the absence or presence of SoSe measured in brain, muscle, pancreas, heart, spleen, eye, and liver tissues obtained from WT and p17/PERMIT−/− mice by PLA to detect the association of LC3 with ceramide and CerS1 with Tom20 using anti-LC3, anti-Tom20, anti-ceramide, or anti-CerS1 antibodies. PLA images shown are for brain, muscle, and pancreas. Quantification of PLA signals by PLA analysis software as described by the manufacturer (n = 3 mice per group). (H) Induction of mitophagy by SoSe in WT and p17/PERMIT−/− mice detected in brain, muscle, and pancreatic tissues measured by ACO2 expression with Western blot (n = 3 mice per group). Actin was used as loading control. (I) Levels of CerS1 in liver, spleen, eye, brain, pancreas, and muscle mitochondrial and soluble fractions of p17 KO versus WT mice treated with vehicle or SoSe for 3 hours.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/9/eaax1978/DC1

    Fig. S1. Roles of CerS1 versus CerS6 in ceramide generation and mitophagy induction.

    Fig. S2. CerS1 amino acids 60 to 76 are critical for Cers1 import to mitochondria.

    Fig. S3. Ceramide transport by CERT or FAPP2 is not involved in mitophagy.

    Fig. S4. CerS1 mitochondrial import is mediated by p17/PERMIT.

    Fig. S5. Mitochondrial import of CerS1 requires p17/PERMIT and involves amino acids RYE28-30.

    Fig. S6. p17/PERMIT-mediated CerS1 import to mitochondria induces mitophagy.

    Fig. S7. Roles of Drp1 nitrosylation at C644 in mitochondrial localization of CerS1.

    Fig. S8. Mitochondrial localization of CerS1 induces mitophagy via ATG, LC3, and Drp1 in response to SoSe.

    Fig. S9. Mitochondrial CerS1-dependent mitophagy is induced via activation of LC3 and Drp1.

    Fig. S10. Summary of the hypothesis and proposed mechanism for the mitochondrial trafficking of CerS1 by p17/PERMIT.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Roles of CerS1 versus CerS6 in ceramide generation and mitophagy induction.
    • Fig. S2. CerS1 amino acids 60 to 76 are critical for Cers1 import to mitochondria.
    • Fig. S3. Ceramide transport by CERT or FAPP2 is not involved in mitophagy.
    • Fig. S4. CerS1 mitochondrial import is mediated by p17/PERMIT.
    • Fig. S5. Mitochondrial import of CerS1 requires p17/PERMIT and involves amino acids RYE28-30.
    • Fig. S6. p17/PERMIT-mediated CerS1 import to mitochondria induces mitophagy.
    • Fig. S7. Roles of Drp1 nitrosylation at C644 in mitochondrial localization of CerS1.
    • Fig. S8. Mitochondrial localization of CerS1 induces mitophagy via ATG, LC3, and Drp1 in response to SoSe.
    • Fig. S9. Mitochondrial CerS1-dependent mitophagy is induced via activation of LC3 and Drp1.
    • Fig. S10. Summary of the hypothesis and proposed mechanism for the mitochondrial trafficking of CerS1 by p17/PERMIT.

    Download PDF

    Files in this Data Supplement:

Navigate This Article