PtdIns3P controls mTORC1 signaling through lysosomal positioning

Abstract

The mechanistic target of rapamycin complex 1 (mTORC1) is a protein kinase complex that localizes to lysosomes to up-regulate anabolic processes and down-regulate autophagy. Although mTORC1 is known to be activated by lysosome positioning and by amino acid-stimulated production of phosphatidylinositol 3-phosphate (PtdIns3P) by the lipid kinase VPS34/PIK3C3, the mechanisms have been elusive. Here we present results that connect these seemingly unrelated pathways for mTORC1 activation. Amino acids stimulate recruitment of the PtdIns3P-binding protein FYCO1 to lysosomes and promote contacts between FYCO1 lysosomes and endoplasmic reticulum that contain the PtdIns3P effector Protrudin. Upon overexpression of Protrudin and FYCO1, mTORC1-positive lysosomes translocate to the cell periphery, thereby facilitating mTORC1 activation. This requires the ability of Protrudin to bind PtdIns3P. Conversely, upon VPS34 inhibition, or depletion of Protrudin or FYCO1, mTORC1-positive lysosomes cluster perinuclearly, accompanied by reduced mTORC1 activity under nutrient-rich conditions. Consequently, the transcription factor EB enters the nucleus, and autophagy is up-regulated. We conclude that PtdIns3P-dependent lysosome translocation to the cell periphery promotes mTORC1 activation.

Figure 1.

Protrudin and FYCO1 mediate translocation of mTOR lysosomes to the cell periphery. (A) Confocal micrograph showing colocalization between endogenous LAMP1 and mTOR in HeLa cells grown in complete medium. (B and C) HeLa cells were transfected with GFP Protrudin for 20 h, stained with antibodies against LAMP1 and mTOR (B) or FYCO1 and mTOR (C), and analyzed by confocal microscopy. Insets show magnification of lysosomes in contact with Protrudin-positive ER. (D) HeLa cells were transfected (40 h) with the constructs indicated, labeled for endogenous mTOR, and analyzed by confocal microscopy. Asterisks indicate transfected cells. Arrows point at mTOR in the cell periphery. Insets show magnification of mTOR- and mCherry–FYCO1–positive lysosomes. Note that whereas mCherry–FYCO1WT colocalizes with mTOR in the cell periphery, mCherry–FYCO1ΔKIFBD colocalizes with mTOR in the perinuclear area. (E) Automated quantification of cells treated as in D, using the Olympus ScanR system. Graphs represent the relative total intensities of perinuclear localization of mTOR-positive lysosomes. The transiently transfected cells were compared directly with nontransfected neighboring cells (set to 1) in each sample. Error bars denote ± SEM from independent experiments: GFP–ProtrudinWT (n = 5), GFP–ProtrudinΔFYVE and ΔKIFBD (n = 4), and GFP–ProtrudinWT+mCh-FYCO1WT or ΔKIFBD (n = 3). **, P < 0.01; ***, P < 0.001. Myc–ProtrudinWT and Myc–ProtrudinWT+mChFYCO1WT, unpaired t test. Transfected cells versus control set to 1, one-sample t test. In total 200–800 transfected cells were analyzed per condition.

Figure 2.

Protrudin and FYCO1 increase S6K phosphorylation in a PtdIns3P-dependent manner. (A) HeLa cells were transfected (40 h) with GFP–Protrudin WT, ΔFYVE, or FYVE4A in combination with mCherry–FYCO1 and analyzed by immunoblotting using antibodies specific to the proteins indicated. GFP and mCherry were used as controls. Where indicated, the cells were starved for 2 h with EBSS. Asterisks indicate unspecific bands from anti-mCherry. Graph shows quantification of immunoblots, representing the relative intensity of phospho-S6K normalized to the total amount of S6K in each sample. Error bars denote ± SEM from independent experiments (n = 3). **, P < 0.01; ***, P < 0.001 (Protrudin WT vs. mutants; unpaired t test; fed control set to 1 vs. starved control or fed Protrudin WT; one-sample t test). (B) HeLa cells were cotransfected (40 h) with GFP–Protrudin WT and mCherry–FYCO1 and analyzed by immunoblotting using antibodies specific to the proteins indicated. Nontransfected cells were used as controls. Where indicated, the cells were starved for 2 h with EBSS. The immunoblot is representative of three independent experiments. (C) HeLa cells were treated with 0.1% DMSO or 3 µM SAR405 for 4 h, before washout of SAR405 for 2 h (all treatments in complete medium), stained with antibodies against LAMP1 and FYCO1, and analyzed by confocal microscopy. (D) RPE-1 cells were treated with 0.1% DMSO or 3 µM SAR405 for 4 h (all treatments in complete medium), stained with an antibody against mTOR, and analyzed by wide-field microscopy. (E) HeLa, HEK293, and RPE-1 cells were treated with 0.1% DMSO or 3 µM SAR405 in complete medium for 4 h and analyzed by immunoblotting. (F) Quantification of immunoblots (as in E), representing the relative intensities of phospho-S6K normalized to the total amount of S6K in each sample. Error bars denote ± SEM from independent experiments (n = 3). **, P < 0.01 (one-sample t test).

Figure 3.

Protrudin and FYCO1 depletion leads to perinuclear clustering of mTOR-positive lysosomes and decreases mTORC1 activity. (A) RPE-1 cells grown in complete medium were transfected with control RNA or siRNA targeting Protrudin or FYCO1, stained with an antibody to mTOR, and analyzed by high-content wide-field microscopy. (B and C) Automated quantification of images as in A using the Olympus ScanR system, representing the relative ratios of perinuclear/peripheral intensity of mTOR-positive lysosomes. Error bars denote ± SEM from independent experiments (B); n = 3; **, P ≤ 0.01 (one-sample t test); >1,400 cells per condition. (C) siRNA FYCO1 oligo 1 (n = 6), >5,000 cells. siRNA FYCO1 oligo 2 (n = 3), >900 cells. *, P < 0.05; **, P < 0.01 (one-sample t test). (D) Representative immunoblots showing the reduced levels of phosphorylated S6K and 4E-BP1 upon Protrudin depletion in RPE-1 cells grown in complete medium. Graphs represent quantifications from immunoblots. Error bars denote ± SEM from independent experiments. p-S6K (n = 6) or p-4E-BP1 (n = 4). **, P < 0.01; ***, P < 0.001 (one-sample t test). (E) Representative immunoblots showing the reduced levels of phosphorylated S6K and 4E-BP1 upon FYCO1 depletion in RPE-1 cells grown in complete medium. Graphs represent quantifications from immunoblots. Error bars denote ± SEM from independent experiments. p-S6K (n = 4) or p-4E-BP1 (n = 3). **, P < 0.01 (one-sample t test). (F) Representative immunoblots showing the reduced levels of phosphorylated S6K upon Protrudin depletion in HEK293 cells grown in complete medium. Graph represents quantifications from immunoblots. Error bars denote ± SEM from independent experiments (n = 4). *, P < 0.05; **, P < 0.01 (one-sample t test). (G) Representative immunoblots showing the reduced levels of phosphorylated S6K upon FYCO1 depletion in HEK293 cells grown in complete medium. Graph represents quantifications from immunoblots. Error bars denote ± SEM from independent experiments (n = 3). **, P < 0.01 (one-sample t test). (H) Representative immunoblots showing unaffected levels of phosphorylated AKT (mTORC2 site Ser473) upon Protrudin depletion in RPE-1 cells grown in complete medium. Graph represents quantifications from immunoblots. Error bars denote ± SEM from independent experiments (n = 4).

Figure 4.

Protrudin is required for amino acid–dependent lysosome redispersion after starvation. (A) RPE-1 cells were grown in complete medium, starved with EBSS for 4 h, and stimulated or not with 2× amino acids (AA) in DMEM with or without FCS for 2 h, stained with an antibody to LAMP1, and imaged by wide-field microscopy. (B) RPE-1 cells grown in complete medium were transfected with control RNA or siRNA targeting Protrudin, starved with EBSS for 4 h, and stimulated or not with 1× amino acids in DMEM for 1 h, stained with an antibody to LAMP1, and imaged by high-content wide-field microscopy. (C) Automated quantification of images as in B using the Olympus ScanR system, representing the relative ratios of peripheral/perinuclear intensity of LAMP1-positive lysosomes. Error bars denote ± SEM from independent experiments (n = 3). *, P < 0.05; **, P < 0.01. Treatments versus fed control set to 1 (one-sample t test), samples with equal variance (unpaired t test). Greater than 1,900 cells were analyzed per condition.

Figure 5.

Protrudin is required for full mTORC1 activation after nutrient starvation–recovery. (A and B) RPE-1 (A) or HEK293 (B) cells grown in complete medium were transfected with control RNA or siRNA targeting Protrudin, starved in EBSS for 4 h, and stimulated or not with 1× amino acids (AA) in DMEM with or without growth factors for 2 h, before immunoblotting with the antibodies indicated. Graphs represent the relative levels of p-ULK1 or p-S6K. Error bars denote ± SEM from independent experiments. RPE-1: p-S6K (n = 3), p-ULK1 (n = 4). HEK293: p-S6K (n = 5), p-ULK1 (n = 4). *, P < 0.05; **, P < 0.01; ***, P < 0.001. No treatment (one-sample t test), samples with equal variance (unpaired t test). (C) RPE-1 cells were transfected with control RNA or siRNA targeting Protrudin. The cells were serum starved for 4 h and stimulated with FCS, insulin, and 1× amino acids in DMEM for 15 min and analyzed by immunoblotting. Error bars denote ± SEM from independent experiments (n = 3). *, P < 0.05 (one-sample t test). (D) RPE-1 cells were treated as in C, stained with antibodies against LAMP1 and p-AKT(Thr308), and analyzed by confocal microscopy.

Figure 6.

Amino acid–stimulated lysosomal FYCO1 recruitment and dispersion. (A) RPE-1 cells were starved with EBSS for 4 h and stimulated or not for 2 h with 2× amino acids (AA) in DMEM with or without 0.1% DMSO or 3 µM SAR405, stained with antibodies against FYCO1 and LAMP1, and analyzed by confocal microscopy. Note that upon amino acid stimulation, FYCO1 is recruited to LAMP1-positive lysosomes that spread, and this is prevented by SAR405. (B–D) High-content imaging and automated analysis of cells treated as in A, using the Olympus ScanR system, representing the relative total intensity of FYCO1 (B) or LAMP1 (C) per cell or the relative ratios of peripheral/perinuclear intensity of LAMP1-positive lysosomes (D). Error bars denote ± SEM from independent experiments (n = 3). *, P < 0.05; **, P < 0.01; ***, P < 0.001 (one-sample t test). More than 1,500 cells were analyzed per condition. (E) RPE-1 cells were starved in EBSS for 4 h and stimulated for 2 h with 2× amino acids in DMEM, stained with antibodies against FYCO1 and Protrudin, and analyzed by confocal microscopy.

Figure 7.

TFEB enters the nucleus in Protrudin- and FYCO1-depleted cells. (A) RPE-1 cells were grown in complete medium or starved in EBSS for 2 h, stained with an antibody to TFEB, and analyzed by high-content wide-field microscopy. (B) Automated quantification of images as in A of the mean intensity of nuclear TFEB using the Olympus ScanR system. Error bars denote ± SEM from independent experiments (n = 3). ***, P < 0.001 (unpaired t test). More than 1,700 cells were analyzed per condition. (C) RPE-1 cells grown in complete medium were transfected with control RNA or siRNA targeting Protrudin or FYCO1, stained with an antibody to TFEB, and analyzed by high-content wide-field microscopy. (D) Automated quantification of images as in C, of the mean intensity of nuclear TFEB using the Olympus ScanR system. Error bars denote ± SEM from independent experiments (n = 3). *, P < 0.05; **, P < 0.01 (unpaired t test). More than 1,200 cells were analyzed per condition. (E) HEK293 cells were transfected with control or Protrudin siRNA or treated with 500 nM Torin1 (1 h) or EBSS (2 h). Cytoplasmic and nuclear fractions were then prepared followed by immunoblotting analysis using indicated antibodies. ERK1/2 and histone H3 were used as the loading control for cytoplasmic and nuclear fractions, respectively. Graphs represent quantifications of TFEB levels in nuclear fractions from immunoblots. TFEB levels were normalized to histone H3 and presented relative to control. Error bars denote ± SEM from independent experiments (n = 3). **, P < 0.01; ***, P = 0.001 (one-sample t test). (F) HEK293 cells were transfected with control or FYCO1 siRNA or treated with Torin1 (1 h) or EBSS (2 h). Separation of cytoplasmic and nuclear fractions and subsequent immunoblotting were performed as described in E. Graphs represent quantifications of TFEB levels in nuclear fractions from immunoblots. TFEB levels were normalized to histone H3 and presented relative to control. Error bars denote ± SEM from independent experiments (n = 3). *, P < 0.05 (one-sample t test). (G) RPE-1 cells were treated as in E. Nuclear fractions were subjected to immunoblotting analysis with indicated antibodies. Graphs represent quantifications of TFEB levels in nuclear fractions from immunoblots. TFEB levels were normalized to histone H3 and presented relative to control. Error bars denote ± SEM from independent experiments (n = 5). *, P < 0.05; **, P < 0.01 (one-sample t test). (H) RPE-1 cells were treated as in F. Nuclear fractions were analyzed using immunoblotting with the indicated antibodies. Graphs represent quantifications of TFEB levels in nuclear fractions from immunoblots. TFEB levels were normalized to histone H3 and presented relative to control. Error bars denote ± SEM from independent experiments (n = 4). *, P < 0.05; **, P < 0.01; ***, P < 0.001 (one-sample t test). (I) Protrudin depletion does not induce ER stress. RPE-1 cells were transfected with control RNA or siRNA targeting Protrudin and analyzed with immunoblotting using antibodies against the ER stress–induced proteins indicated. Thapsigargin was used to induce ER stress in nontransfected cells.

Figure 8.

Protrudin depletion promotes initiation of autophagy. (A) Immunoblot showing an increased level of LC3-II in fed RPE-1 cells transfected with siRNA targeting Protrudin compared with control. The graph represents the relative level of LC3-II normalized to the loading control HSP90 quantified from immunoblots as in A. Error bars denote ± SEM from independent experiments (n = 5). *, P < 0.02; **, P < 0.01 (one-sample t test). (B) RPE-1 cells were siRNA transfected and either grown in complete medium or starved for 2 h in EBSS. ConA was given for 2 h where indicated. The cells were stained with an antibody to LC3 and analyzed by high-content wide-field microscopy. (C) Automated quantification of images as in B of the total intensity of LC3 dots per cell using the Olympus ScanR system. Error bars denote ± SEM from independent experiments (n = 4). ***, P < 0.001 (unpaired t test). More than 1,400 cells were analyzed per condition. (D) Automated quantification of images as in B of the number of LC3 dots per cell using the Olympus ScanR system. Error bars denote ± SEM from independent experiments (n = 4). **, P < 0.01; ***, P < 0.001 (unpaired t test). More than 1,400 cells were analyzed per condition. (E) Automated quantification of images as in B of the amount of cells having LC3 dots with size >50 pixels, using the Olympus ScanR system. Error bars denote ± SEM from independent experiments (n = 4). **, P < 0.01; ***, P < 0.001 (unpaired t test). More than 1,400 cells were analyzed per condition. (F) Imaris representations of confocal z-stacks visualizing the number and size of LC3 positive dots in control- or siRNA-treated RPE-1 cells grown in complete medium with or without ConA. (G) Quantification of the rate of LLPD in control and siRNA-treated RPE-1 cells. Error bars denote ± SEM from independent experiments (n = 3). ***, P < 0.001 (unpaired t test).

Figure 9.

Model for Protrudin- and FYCO1-mediated mTOR activation. (A) Under nutrient starvation conditions, lysosomes localize perinuclearly, mTORC1 is predominantly cytosolic, and VPS34 activity on LyLEs is low. The lack of PtdIns3P prevents Protrudin-mediated ER–endosome contacts. (B) The presence of amino acids stimulates the recruitment of the mTORC1 complex to lysosomes. In addition, amino acids activate VPS34 to produce PtdIns3P at the lysosomal membrane. The PtdIns3P-binding protein FYCO1 is recruited to lysosomes, and Protrudin can mediate PtdIns3P-dependent ER–lysosome contact. In such contact sites, the microtubule motor Kinesin-1 is transferred from Protrudin to FYCO1. (C) Lysosomes loaded with Kinesin-1 are translocated along microtubules to the cell periphery. This brings the lysosomal mTORC1 complex in close apposition to nutrient signaling complexes at the plasma membrane and increases mTORC1 activity.