The mechanistic target of rapamycin (mTOR) signals through the mTOR complex 1 (mTORC1) and the mTOR complex 2 to maintain cellular and organismal homeostasis. Failure to finely tune mTOR activity results in metabolic dysregulation and disease. While there is substantial understanding of the molecular events leading mTORC1 activation at the lysosome, remarkably little is known about what terminates mTORC1 signaling. Here, we show that the AAA+ ATPase Thorase directly binds mTOR, thereby orchestrating the disassembly and inactivation of mTORC1. Thorase disrupts the association of mTOR to Raptor at the mitochondria-lysosome interface and this action is sensitive to amino acids. Lack of Thorase causes accumulation of mTOR-Raptor complexes and altered mTORC1 disassembly/re-assembly dynamics upon changes in amino acid availability. The resulting excessive mTORC1 can be counteracted with rapamycin in vitro and in vivo. Collectively, we reveal Thorase as a key component of the mTOR pathway that disassembles and thus inhibits mTORC1.

Lentivirus Generation

Lentiviruses expressing Thorase were generated by the transient transfection of HEK293T cells with lentiviral plasmids described above using FuGENE HD Transfection Reagent.  The cells were co-transfected with the Thorase lentiviral plasmids, the trans-complementation plasmids (pLP1 and pLP2), and the plasmid encoding the vesicular stomatitis virus envelope glycoprotein (VSVG) followed by sodium butyrate treatment 6 hours after transfection. The medium was replaced 24 hours after transfection. Viral particles were harvested by collecting medium 48 hours and 72 hours after transfection and centrifuged at 3000g for 10 minutes, then filtered through a 0.45 um membrane. Viral particles were then concentrated by ultracentrifugation (2 hours, 25,000 rpm, rotor SW28). The viruses were stored at -80° until needed. The expression of Thorase was verified by infecting HEK293 cells with viruses and cells examined under a fluorescence microscope 24–48 hours after infection.  Cells were harvested and lysates were resolved on SDS-PAGE and immunoblotted by probing with anti-Thorase.

Recombinant protein purification

GST- tagged fusion proteins were expressed in Escherichia coli strain BL21-CodonPlus (DE3)-RIPL bacteria.  Small cultures (100 ml) were grown overnight at 37°C and then transferred to 2-liter cultures for another 8 hours.  Cells were cold shocked at 4°C for 1 hour before adding 1mM IPTG for overnight induction at 16°C.  Cells were lysed using micro-Fluidizer in binding buffer (1x phosphate buffer, pH 7.5, 150 mM NaCl, 2.5 mM MgCl₂, 1 mM DTT, 5% glycerol) containing protease inhibitors and purified by using GST beads following the manufacturer’s instructions. To obtain untagged purified proteins, GST-tagged proteins bound to the GST beads were treated with Prescission protease to remove the GST tags. Samples were further purified using ion exchange chromatography and/or size-exclusion chromatography. Purity of the recombinant proteins was assessed by SDS-PAGE followed by Coomassie blue staining and immunoblotting. 

Pull down assays

For the identification of Thorase binding partners, recombinant GST-Thorase purified on beads was mixed with mouse whole brain cytosolic extract in the presence of non-hydrolysable ATP (ATPgS) in buffer A (50 mM HEPES, pH 7.5, 150 mM NaCl, 2.5 mM MgCl₂, 1 mM DTT, 5% glycerol) and incubated with end-over-end mixing for 2 hours. The beads were then extensively washed with buffer A containing ATPgS, and Thorase was cleaved from the GST beads with PreScission protease. The supernatant containing different protein complexes bound to Thorase was loaded into size-exclusion chromatography to separate the complexes. The different fractions eluted from the column were resolved by SDS-PAGE and individual bands (proteins) were excised for mass spectrometry analyses. Immunoblotting was used to confirm the presence of proteins identified by mass spectrometry.

Co-immunoprecipitation of endogenous Thorase and mTOR was carried out using lysates of whole mouse brains. Freshly isolated whole brains from wild type or Thorase KO mice were homogenized in buffer A containing protease inhibitors with or without 2 mM ADP, ATP or ATPγS. Triton X-100 was added to a final concentration of 1% followed by rotation for 2 hours at 4°C. Extracts were centrifuged at 15,000g for 30 minutes and supernatant was incubated for 3 hours at 4°C with Protein G beads pre-bound with anti-Thorase or anti-mTOR antibodies. The beads were washed 3 times with buffer A plus 1 mM ADP, ATP or ATPγS, and bound proteins were eluted from beads using 1x SDS-PAGE Laemmli buffer with DTT.  The eluted proteins were resolved by SDS-PAGE. Immunoblotting analyzes were carried out with antibodies to Thorase, mTOR, Raptor, S6K, and 4EBP1. 

For in vivo pull down of Thorase and mTOR, live cells were treated with or without 2.5 mg/ml DSP in the culture media and incubated for 30 min at 37°C. The cells were washed three times with HBSS prior to harvesting. Cell lysates were prepared in the presence of 2 mM ADP, ATP, or ATPgS in buffer A and incubated with mixing for 2 hours. The beads were then extensively washed with buffer A containing 1 mM ADP, ATP, or ATPgS. The eluted protein complexes were further separated on size exclusive chromatography (SEC).  Fractions from SEC were resuspended in SDS-PAGE sample buffer and resolved by SDS-PAGE. Immunoblotting was used to confirm the presence of protein complexes.

For in vitro pull down of Thorase and mTOR, recombinant GST-Thorase (wild type or variants) or GST-mTOR purified on beads were mixed with mouse whole brain lysates in the presence of 2 mM ADP, ATP, or ATPgS in buffer A and incubated with mixing for 2 hours. The beads were then extensively washed with buffer A containing 1 mM ADP, ATP, or ATPgS. The beads were resuspended in 1 x SDS-PAGE sample buffer and eluted samples resolved by SDS-PAGE. Immunoblotting was used to confirm the presence of proteins bound to GST-tagged proteins.  For assessing direct interaction between Thorase and mTOR, recombinant GST-Thorase purified on beads was mixed with purified non-tagged mTOR recombinant protein in the presence of 2 mM ATPgS in buffer A and incubated with mixing for 2 hours. The beads were then extensively washed with buffer A containing 1 mM ATPgS.  The sample eluted from the beads was loaded into size-exclusion chromatography to separate the complexes.  The different fractions eluted from the column were resolved by SDS-PAGE to confirm the presence of GST-Thorase and mTOR. Mass spectrometry was conducted at Harvard Medical School Taplin Spectrometry Facility.

Starvation and mTORC1 activation experiments

Starvation and stimulation in HEK-293T cells and MEFs was performed as previously described, with some modifications as follows. Cells cultured in complete DMEM (supplemented with glutamine, FBS and PS) were rinsed with and incubated in amino acid-free HBSS supplemented with dialyzed FBS (amino acid free), glucose, glutamine, and PS (CCM media) for 50 minutes and stimulated with a 10X mixture of amino acids in CCM for 15 minutes. Cells were fixed with 4% PFA or harvested at 5, 10, 20, 30, or 50 minutes after amino acid stimulation for 10 minutes.

For serum starvation, subconfluent cells were starved in DMEM for 16 hours and stimulated with insulin for 15 min, similar to described by Menon and colleagues, and Hoxhaj and colleagues.

For live imaging experiments, cells co-expressing a combination of Thorase-GFP, Thorase-RFP, mTOR-YFP, Lamp1-GFP, Lamp1-RFP, and/or Mito-RFP cultured in CCM media were rinsed with and incubated in amino acid-free HBSS supplemented with glucose, glutamine, and PS for 50 minutes (HEK293) or 2 hours (MEFs). Media was then replaced with a 10X mixture of amino acids in CCM and live imaged immediately using a Zeiss LSM 5 Duo confocal microscope with continuously perfused at a rate of 1 ml/min with CCM media with 10X mixture of amino acids to monitor co-localization of Lamp1-RFP and Thorase-GFP or mTOR-YFP. Cells were imaged for 3 minutes (1 image/15 sec) or 30 minutes (1 image/min). All images were analyzed using NIH ImageJ software (Rasband, W.S., NIH, http://rsb.info.nih.gov/ ij/, 1997–2007). 

Immunostaining

Wild type and Thorase KO mice were anesthetized with sodium pentobarbital, perfused with 4% PFA in PBS (phosphate buffer, pH 7.4), and immediately decapitated and the brains removed.  Post fixation was performed in 4% PFA in PBS overnight and fixed brains were then dehydrated in 30% glucose for another 48 hours. Brains were sectioned on a microtome at 30 μm, washed in PBS and then permeabilized with 0.3% Triton X-100 in PBS containing 5% goat serum for 1 hour at room temperature.  After three 5-minute washes with PBS, sections were incubated with primary antibodies in PBS containing 0.1% Triton X-100, 0.01% sodium azide, and 2.5% goat serum overnight at 4°C. Following washout with PBS, sections were incubated with secondary antibodies 3 hours at room temperature. Sections were washed three times with PBS and cell nuclei counterstained with DAPI (Invitrogen, Molecular Probes). Sections were mounted on slides and a Zeiss LSM 880 laser-scanning confocal microscope was used to acquire images under identical acquisition parameters for side-by-side comparison.  For co-localization experiments, we defined mTOR puncta as any mTOR accumulation at the lysosome that was at least 1µm in size. 3D images were generated and analyzed by Imaris 9.2.0 (Bitplane, AG www.imaris.com). For in vitro cultures grown on coverslips and fixed in 4% PFA, immunostaining followed the same protocol as described above.

De novo protein synthesis assessment

Fibroblasts derived from wild type or Thorase KO mouse embryos were grown in 12 well plates (Falcon) to around 80% confluency and pre-treated with vehicle DMSO or 500 nM rapamycin if needed. For S³⁵-labeling assays, 25 mCi of EasyTag™ EXPRESS³⁵S Protein Labeling Mix (Perkin Elmer) per well were added to the media, and cells were incubated at 37°C and 5% CO2 for 1 hour. Cells were then washed twice with cold PBS and lysed in extraction buffer (2% NP-40, 50 mM Tris-HCl, 150 mM NaCl, 5 mM EGTA, protease inhibitor cocktail), spun down and supernatants collected. Protein was concentrated with methanol/heparin, and protein pellets then resuspended in 8M Urea/150 mM Tris (pH 8.5). Samples were then analyzed in a LS 6500 Multi-purpose Scintillation Counter (Beckman Coulter), and radioactive counts normalized to sample protein concentration as measured by Pierce BCA protein assay (Thermo Scientific).  A fraction of the precipitated protein was run in 4–20% SDS-PAGE gels and the emitted radioactivity was images using Typhoon Imager (Amersham).

For SUnSET assays, per every milliliter of media 10 ug of Puromycin (Gibco), and in control conditions 100 ug of cycloheximide (Sigma) additionally, was added, and cells were incubated at 37°C and 5% CO2 for 20 minutes. The collected protein lysates were run in 4–20% SDS-PAGE gels and the extent of newly synthesized proteins was detected using an anti-puromycin antibody.

SiMPull experiments

HEK293 cells were transiently co-transfected with YFP-mTOR, and mCherry-Raptor followed by cell lysis in lysis buffer (40 mM Hepes, pH 7.5, 120 mM NaCl, 10 mM sodium pyrophosphate, 10 mM b-glycerophosphate, 1X protease inhibitor mixture and 0.3% CHAPS). YFP-mTOR- and mCherry-Raptor-only lysates were obtained by transiently transfecting cells with YFP-mTOR and mCherry-Raptor, respectively. Lysates were centrifuged at 12,000 X g for 12 min at 4°C and diluted 150-fold to obtain a surface density optimal for single-molecule analysis (~800 molecules in 5,000 mm² imaging area). For control experiments, HEK293 cells were transiently transfected with PKA isoforms PKA-RIIb-Flag-mCherry and PKA-Ca-HA-YFP. After 24 hours of expression, cells were lysed using a buffer containing 10mM Tris pH 7.5, 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM benzamidine, 10 mg/ml leupeptin, 1mM NaF, 1mM Na₃VO₄. The lysate thus obtained was centrifuged at 14,000 X g for 20 minutes and subsequently used for SiMPull. Recombinant Thorase was incubated with 20-fold molar excess of Cy5-maleimide dye at room temperature for an hour and at 4oC overnight. Unreacted dyes were removed by gel filtration using Zeba spin desalting columns.

Single-molecule experiments were performed on a prism-type TIRF microscope equipped with an electron-multiplying CCD camera (EM-CCD). For single-molecule pull-down experiments, quartz slides and glass cover slips were passivated with 5000 MW methoxy poly-(ethylene glycol) (mPEG, Laysan Bio) doped with 2–5% 5000 MW biotinylated PEG (Laysan Bio). Cell lysates were pulled down with biotinylated antibodies against YFP or RFP, already immobilized on the surface via neutravidin-biotin linkage. Thorase interactions with mTORC1 and its individual components were studied by incubation of pulled-down mTORC1 or its individual components with a pre-determined concentration of Cy5-Thorase in T200-BSA buffer supplemented with 5 mM AMP-PNP (or ATP), 20 mM MgCl₂ and 10% Glycerol. 15 frames were recorded from each of 20 different imaging areas (5,000 mm²) and isolated single-molecule peaks were identified by fitting a Gaussian profile to the average intensity from the first ten frames. Mean spot-count per image for YFP and mCherry was obtained by averaging 20 imaging areas using MATLAB scripts.

Co-localization data were acquired from two or three separate movies from the same region of a slide using YFP, m-Cherry, and Cy5 excitation. The co-localization criterion was set to a diffraction-limited region of ~300 nm, which corresponds to 2 pixels for this TIRF setup. % co-localization was calculated as the % molecules co-aligned with the pulled down component, unless otherwise stated. YFP, mCherry, and Cy5 images taken from different areas were also overlapped to determine the probability of false co-localization arising from random spatial overlap of single molecules. Average % co-localization was calculated based on a minimum of 30 individual slide areas. For photobleaching analysis, single-molecule fluorescent time traces from individual YFP or mCherry spots were manually scored for the number of photobleaching steps and the stoichiometry of the molecules was assessed.  Briefly, the fluorescence trace of each molecule was classified as having one to four bleaching steps or was discarded if no clean bleaching steps could be identified. The intensity of molecules scored as bleaching in one and two steps was plotted to verify scoring: on average we expect the fraction of molecules bleaching in two steps to be twice as bright as the molecules bleaching in one step. The intensity of discarded molecules was also plotted to ensure unbiased scoring as observed via lack of enrichment of any specific intensities. To convert the photobleaching step distribution to monomer/dimer fraction, the percentage of molecules bleaching in two steps was compared with a calibration experiment. This conversion was performed only when >90% of the molecules bleached in one or two photobleaching steps.

Cell migration (scratch) assay

Wild type or Thorase KO fibroblasts were grown in tissue culture-treated glass bottom 12 well plates (Corning™ Costar™) to around 70% monolayer confluency.  A sterile p200 pipet tip was used to scrape the cell monolayer in a straight line to create a “scratch”.  The media with the cell debris were replaced with fresh growth medium containing vehicle DMSO or 500 nM rapamycin.  Live imaging of the scratched regions was performed using EVOS®-FL-Auto phase-contrast microscope in a tissue culture incubator at 37 °C with 5% CO2. Cells were imaged for 8 hours (1 image/15 minutes). All images were analyzed using NIH ImageJ software (Rasband, W.S., NIH, http://rsb.info.nih.gov/ ij/, 1997–2007).  For each image, distances between one side of scratch and the other were measured (nm) and rate of cell migration were determined by dividing the initial scratch width by the time spent in migration (time point to obtain scratch closure).

Rapamycin treatment in mice

Thorase homozygous KO mice and wild-type littermates were obtained by crossing heterozygous Thorase mice. Animal experiments were performed in compliance with the regulations of the Animal Ethical Committee of the Johns Hopkins University Animal Care and Use Committee.  Wild type or Thorase KO mice were injected intraperitoneally with 6mk/kg Rapamycin (#R-5000, LC laboratories) or vehicle (10% PEG400, 10% Tween 80 in water), 3 times per week starting at 10–14 days postnatally.

Study design and reproducibility

All biochemical and imaging experiments were repeated independently at least three times. For biochemical experiments involving brain tissue or mouse embryonic fibroblasts, 3 biological replicates per group were used per experiment. For mouse survival experiments and immunohistochemical analysis, enough sample size (as determined via post hoc power analysis) was used, and mice pertaining to different litters were used. We found our results reproducible when experiments were performed by 2 independent researchers.  For sample randomization, for biochemical experiments where 2 treatments were applied, 2 technical replicates per biological sample (3 biological replicates total per group) were used. For mouse experiments where either vehicle or rapamycin was injected, an independent researcher randomly allocated mice to either treatment group. Regarding blinding during data collection, for immunofluorescence, immunohistochemistry and electron microscopy imaging, researchers were blinded to genotypes. For mouse survival analysis, researchers were blinded to both genotype and treatment.

Data analysis and statistics

All experiments were repeated at least three times and quantitative data are presented as the mean ± standard error of the mean (SEM) by GraphPad prism6 software (Instat, GraphPad Software). Statistical significance was assessed by ANOVA. The significant differences were identified by post-hoc analysis using the Tukey-Kramer post-hoc method for multiple comparisons. Assessments were considered significant with a p < 0.05 and non-significant with a p > 0.05 (Table S2). To assess the sufficiency of sample size, post hoc power analysis was performed using GraphPad Prism, considering a power greater than 0.9 sufficient.