Reagents and drug treatment

PDGF (P3201) was purchased from Sigma and was used at 50 ng ml−1 to stimulate starved cells. MitoTracker red (M22425) and LysoTracker green (L7526) were purchased from Thermo Fisher Scientific and used at 100 nM final concentration. Torin 1 was purchased from TOCRIS (4247) and used at 1 µM for a 30-min pretreatment before imaging experiments were performed. H89 was used at 10 µM for a 10-min pretreatment. Gö6983 was used at 1 µM for a 10-min pretreatment. GDC-0068 was purchased from APExBIO (RG7440) and was used at 1 µM for a 10-min pretreatment. DMSO was purchased from Sigma (D2650) and used as vehicle. Fsk and IBMX were used at 50 µM and 100 µM, respectively. PMA was used at 50 µM. Doxycycline was purchased from Clontech (631311) and supplemented in medium at a final concentration of 100 ng ml−1 to induce gene expression. CTB was purchased from Sigma-Aldrich (C6499), dissolved in DMSO and used at a concentration of 50 µM. C646 was obtained from Sigma-Aldrich (SML0002), dissolved in DMSO and used at a final concentration of 10 µM. Leucine O-methyl ester (Leu-O-Me) was purchased from Sigma-Aldrich (L1002) and used at a final concentration of 7.5 mM. Methyl-β-cyclodextrin (MCD, 332615) and cholesterol (C8667) were purchased from Sigma-Aldrich and used at final concentrations of 50 or 250 µM for MCD or 20 µg ml−1 for cholesterol. Lipid-depleted serum was purchased from Kalen Biomedical (880100-1).

Constructs and cloning

TORCAR and its targeted versions (TORCAR-NES, Lyso-TORCAR and TORCAR-NLS) were previously described10. Mito-TORCAR was generated by the addition of a dAKAP mitochondrial targeting sequence40 to the N terminus of untargeted TORCAR. The PRAS40 sequence HA–PRAS40 wild type was a gift from D.-H. Kim (Addgene, plasmid 86758; http://n2t.net/addgene:86758; RRID:Addgene_86758). Primers were designed (Eton Bioscience) to generate different truncated fragments using PCR. Similarly, backbone fragments were amplified from a pcDNA3-mCherry plasmid. PRAS40 fragments and backbone fragments were Gibson assembled to generate different mCherry-tagged PRAS40 fragments, including pcDNA3-mCherry-114–245. Mutations in pcDNA-mCherry-114–245 were generated similarly by Gibson assembly using fragments amplified with primers containing the desired mutation(s), yielding pcDNA3-mCherry-114–245A (Cyto-TerminaTOR). pcDNA3-Cyto-TerminaTOR was digested with BamHI/EcoRI to obtain a TerminaTOR fragment and Lyso-TORCAR9 was digested with BamHI/EcoRI to obtain a backbone containing the lysosome targeting sequence. The fragments were ligated using T4 ligase to generate pcDNA3-Lyso-TerminaTOR. To generate Nuc-TerminaTOR, NES mutations were first introduced to generate pcDNA3-TerminaTOR (NES mutant), which was digested with BamHI/EcoRI to obtain a TerminaTOR (NES mutant) fragment. The backbone containing H2A was also digested with BamHI/EcoRI. Fragments were ligated using T4 ligase to generate pcDNA-Nuc-TerminaTOR. To generate lentiviral transfer plasmids, fragments from pcDNA were amplified and assembled with backbone fragments amplified from a lentiviral plasmid, which was a gift from E. Campeau and P. Kaufman (Addgene, plasmid 17452; http://n2t.net/addgene:17452; RRID:Addgene_17452). pcDNA-Cyto-TerminaTOR, pcDNA-Lyso-TerminaTOR and pcDNA-Nuc-TerminaTOR were subcloned into a doxycycline-inducible lentiviral-expressing vector (Addgene, backbone 27565) using Gibson assembly, yielding pLenti-TetOn-Cyto-TerminaTOR, pLenti-TetOn-Lyso-TerminaTOR and pLenti-TetOn-Nuc-TerminaTOR. GFP–LC3 (Addgene, 11546) was used to make RFP–GFP–LC3 through the insertion of RFP using Gibson assembly. smuRFP-tagged Lyso-TerminaTOR was also generated through the insertion of smuRFP using Gibson assembly. TerminaTOR and control construct amino acid sequences are listed in Supplementary Table 6.

Cell culture, transfection and starvation

NIH3T3 cells (CRL-1658, American Type Culture Collection (ATCC)) were cultured in DMEM (11885, Gibco) supplemented with 10% calf serum (30-2030, ATCC) and 1% penicillin–streptomycin (Sigma-Aldrich). Cal33 (CVCL_1108) cells derived from a 69-year-old male with tongue squamous cell carcinoma were obtained from the National Institute of Dental and Craniofacial Research Oral and Pharyngeal Cancer Branch cell collection87. Cal33 cells were cultured in DMEM (D-6429, Sigma-Aldrich), 10% FBS and 5% CO2 at 37 °C. Cells were routinely tested for Mycoplasma contamination and found negative. For transfection, cells were transfected with PolyJet transfection reagent (SL100688, SignaGen Laboratories) according to the manufacturer’s instructions and incubated in serum-free DMEM overnight (serum starvation). The next day, cells were incubated for 2 h in modified Hanks’ balanced salt solution (1× HBSS, diluted from 10× HBSS; 14065, Gibco) and supplemented with 20 mM HEPES pH 7.4 and 2 g L−1 glucose at 37 °C (referred to as ‘double starvation’) before live-cell imaging. For Leu-O-Me treatment, cells were transfected, double-starved and stimulated with 7.5 mM Leu-O-Me. For cholesterol starvation, cells were incubated in DMEM with 10% lipid-depleted serum and 0.5% MCD for 2 h. Cells were then stimulated with 20 µg ml−1 cholesterol precomplexed with 0.1% MCD.

Lentivirus production and stable cell line generation

For the generation of stable cell lines expressing TerminaTOR or control constructs, lentivirus was packaged in HEK293T cells. Specifically, HEK293T cells were cotransfected with the corresponding lentiviral transfer plasmids + psPAX2 + pMD2.G using PolyJet. After 48 h, the supernatant was collected, filtered with a 0.45-µm syringe filter and added to NIH3T3 cells in 35-mm glass-bottom dishes. Polybrene transfection reagent (TR-1003-G, Millipore) was supplemented to improve the transduction. After 48 h, cells were passed in fresh growth medium containing puromycin (2 µg ml−1) to select transduced cells for at least 1 week. Cells were maintained in selection medium and passed for related experiments. For doxycycline-inducible cell lines, NIH3T3 cells were cotransduced with lentivirus expressing rtTA transcription factor and vectors expressing doxycycline-inducible TetOn genes. Cells were then maintained in growth medium containing puromycin (2 µg ml−1) and blasticidin (2.2 µM) and passaged for related experiments. TSC-knockdown cell lines were generated as previously reported49.

IP–MS

NIH3T3 cells were washed with ice-cold PBS twice before lysing with 200 µl of ice-cold lysis buffer (40 mM HEPES pH 7.4, 2 mM EDTA, 10 mM sodium pyrophosphate, 0.3% CHAPS and one tablet of EDTA-free protease inhibitor (Roche, 11873580001) per 25 ml) per 15-cm dish to collect 1.2–1.5 mg of protein per sample. Samples were incubated on ice for 30 min and vortexed for 10 s once every 10 min. Samples were centrifuged for 20 min at 20,000g at 4 °C. The supernatant was moved to a new tube. A portion of sample was saved to run as input. Then, 1 mg of protein per sample was diluted in a 200-µl final volume of lysis buffer. The sample was incubated with 25 µl of RFP-trap beads (Chromotek, rtma) overnight. Before sample incubation, the beads were washed with 500 µl of wash buffer (40 mM HEPES pH 7.4, 2 mM EDTA, 10 mM sodium pyrophosphate, 0.3% CHAPS, one tablet of EDTA-free protease inhibitors per 25 ml and 150 mM NaCl) three times. Following overnight incubation, the beads were washed twice with 500 µl of wash buffer. The beads were washed again with detergent-free wash buffer and moved to fresh tubes. The beads were washed once more with detergent-free wash buffer for 5 min on a rotator at 4 °C. About 10% of the beads were collected to check mTOR, Raptor and TerminaTOR (RFP) presence by western blot. The remaining beads were used for MS.

For MS, triplicate samples were reduced and alkylated with DTT and iodoacetamide and digested on-bead with 200 ng of trypsin (Promega). The samples were desalted on stage tips, dried by vacuum centrifugation and then resuspended in 5% acetonitrile and 0.1% FA. Sample concentrations were determined by nanodrop. Next, 1 μg (for untargeted mCherry and TerminaTOR) or 500 ng (for H2A–mCherry and Nuc-TerminaTOR) of each sample was injected into a Thermo Orbitrap Eclipse coupled to an Easy nLC 1200. Peptides were separated on an integrated emitter column (20 cm × 75 μm inner diameter) packed with 1.9 μm of C18 beads (Dr. Maisch). The method separated peptides from 6% solvent B (80% acetonitrile, 0.1% formic acid in water) to 75% over 84 min. MS1 scans were measured in the Orbitrap at 60,000 resolution scanning from 350 to 2,000 m/z. The maximum injection time was set to 50 ms at an automated gain control (AGC) target of 4 × 105 ions. Quadrupole isolation width was 0.7 m/z, with no isolation offset. MS2 scans were measured in the Orbitrap at 15,000 resolution, with an injection time of 50 ms to reach 60% AGC. Dynamic exclusion was set to 45 s.

The data were searched with Maxquant against a UniProt (December 28, 2017) mouse database. Protein levels were quantified using label-free quantitation at the MS1 level. Protein N-terminal acetylation and methionine oxidation were set as dynamic modifications. Cysteine carbamidomethylation was set as a static modification. Digestion conditions were set to Trypsin/P. The instrument type was set to Orbitrap and peptides with a maximum length of 7 aa were searched with a first-search tolerance of 20 ppm and a main-search tolerance of 4.5 ppm. Next, a volcano plot was generated using the program Perseus. The resulting protein matrix was filtered to remove peptides identified by the reverse sequence and potential contaminants. Label-free quantification (LFQ) protein intensities were log2-transformed and the data were filtered such that proteins without at least two values in one condition were removed. The data were imputed to replace zero values with nonzero values close to the limit of detection of the instrument. P values and fold changes of Raptor and mTOR were set as the cutoff to identify proteins in the TerminaTOR conditions that may interact with the inhibitor.

Autophagy assays

NIH3T3 cells expressing mCherry or mCherry–Lyso-TerminaTOR were transfected with GFP–LC3 and kept under nonstarved nutrient rich conditions or double-starved. For double-starvation conditions, after transfection, NIH3T3 cells were cultured for 24 h in serum-free DMEM (Gibco, 11885) for serum starvation, followed by incubation for 2 h at 37 °C in modified 1× HBSS (diluted from 10× HBSS; Gibco, 14065) supplemented with 20 mM HEPES pH 7.4 and 2 g L−1 glucose for additional amino acid starvation. Nonstarved control cells were maintained in DMEM supplemented with 10% calf serum (ATCC, 30-2030) in parallel. The number of GFP puncta was counted using an ImageJ/Fiji macro88. For the RFP–GFP–LC3 assay, cells expressing smuRFP or smuRFP-tagged Lyso-TerminaTOR were transfected with RFP–GFP–LC3 and kept under nonstarved or double-starved conditions before imaging. Cells were treated with biliverdin (for smuRFP labeling) 3 h before imaging. The number of GFP-positive, RFP-positive or GFP–RFP-positive puncta were counted using an ImageJ/Fiji macro.

Protein purification and in vitro kinase assays

For mCherry and TerminaTOR purification, Escherichia coli BL21 (DE3) chemically competent cells were transformed with His-tagged constructs in pRSET-B vector. A single colony from a Luria–Bertani (LB) agar plate with ampicillin (100 μg ml−1) was inoculated in 20 ml LB–ampicillin (100 μg ml−1) medium and cultured at 37 °C overnight. The seed culture was then inoculated in 1 L of LB–ampicillin medium and cultured at 37 °C with shaking (200 rpm) until reaching an optical density at 600 nm between 0.6 and 1.0. Expression was induced by adding IPTG (500 μM final concentration) and cultured for 18 h at 16 °C with shaking. Cells were harvested by centrifugation at 5,000g for 10 min at 4 °C. Cell pellets were chilled on ice then resuspended in lysis buffer (50 mM Tris pH 7.4, 300 mM NaCl and 1 mM DTT) containing cOmplete protease inhibitor cocktail (EDTA-free, one tablet per 50 ml; Roche). Cells were lysed using a high-pressure homogenizer and debris was removed by centrifugation (45,000g for 0.5 h, 4 °C). The supernatant was filtered through a 0.45-μm membrane filter before loading onto a prepacked Hispur Ni-NTA resin column (5 ml; GE Healthcare) through a syringe at roughly 1 ml min−1 at 4 °C. The column was washed thoroughly with washing buffers (lysis buffer containing imidazole) at increasing imidazole concentrations: 0 mM imidazole, 25 ml; 20 mM imidazole, 50 ml; 70 mM imidazole, 20 ml. mCherry or TerminaTOR was then eluted with lysis buffer supplemented with a total of 300 mM imidazole. Fractions were analyzed by SDS–PAGE and the pure fractions were pooled, concentrated and buffer-exchanged to 50 mM Tris pH 7.4, 300 mM NaCl and 10% glycerol.

To confirm whether TerminaTOR can inhibit activity on the nuclear mTORC1 substrate MAF1, an in vitro kinase assay was conducted with purified mCherry and TerminaTOR and recombinant mTORC1 (Sigma-Aldrich, SRP0364) and Maf1 (Antibodies Online, ABIN5711827). Briefly, 250 ng of mTORC1, 1 µg of MAF1 and 500 µM ATP were mixed with approximately 0.5 µg of mCherry or TerminaTOR in the kinase buffer (25 mM HEPES pH 7.4, 50 mM NaCl, 5 mM MnCl2 and one tablet of EDTA-free protease inhibitor (Roche, 11873580001) per 25 ml) in a final volume of 50 µl and then incubated for 30 min at 30 °C. Reactions were stopped with 16 µl of SDS buffer and incubated at 95 °C for 5 min before running a western blot and detecting phosphorylation using a phospho-S/T antibody.

To evaluate the effect of TerminaTOR on NEK7 activity, a commercially available NEK7 kinase assay kit was used (BPS Bioscience, 78850) according to the manufacturer’s protocol, using ADP-Glo (Promega, V6930) as the detection reagent. For the experimental conditions, purified mCherry or TerminaTOR titrated from 1 × 10−5 to 1 μM was included in the reaction mix. For the positive control, the reaction buffer was excluded. Kinase activity exhibited a linear relationship with the luminescence signal. For data processing, the kit’s manual was followed. In brief, the background luminescence was subtracted from each data point using the mean of three blank samples. Experimental conditions were normalized to the positive control condition. Four independent experiments were performed with three technical replicates per condition.

Immunoblotting

Cells were washed with ice-cold PBS and then lysed in RIPA lysis buffer containing protease inhibitor cocktail, 1 mM PMSF, 1 mM Na3VO4, 1 mM NaF and 25 nM calyculin A. Total cell lysates were incubated on ice for 30 min and then centrifuged at 15,000g at 4 °C for 20 min. Equal amounts of total protein were separated by 4–15% SDS–PAGE and transferred to PVDF membranes. The membranes were blocked with TBS containing 0.1% Tween-20 and 5% BSA and then incubated with primary antibodies overnight at 4 °C. The next day, membranes were washed, incubated with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies and developed using HRP-based chemiluminescent substrate (34579 and 34076, Thermo Fisher Scientific). The following primary antibodies were used for immunoblotting: anti-pS6K1 (T389) (9205, rabbit polyclonal), anti-S6K1 (9202, rabbit polyclonal), anti-acetyl-H3K56 (4243, rabbit polyclonal), anti-H3 (9715, rabbit polyclonal), anti-p4EBP1 (T37/46) (2855, rabbit monoclonal), anti-4EBP1 (9452, rabbit polyclonal), anti-pGSK3β (S9) (9322, rabbit monoclonal), anti-pAkt (S473) (9271, rabbit polyclonal), anti-Akt (9272, rabbit polyclonal), anti-mTOR (2972, rabbit polyclonal), anti-Raptor (2280, rabbit monoclonal), anti-TSC2 (3612, rabbit polyclonal), anti-tubulin (2146, rabbit polyclonal), anti-ATG16L1B (8089, rabbit monoclonal), anti-ULK1 (8054, rabbit monoclonal), anti-pULK1 (S757) (6888, rabbit polyclonal) and anti-TFEB S211 (37681, rabbit monoclonal) antibodies from Cell Signaling Technology, anti-6×His tag (MA1-21315-HRP, mouse monoclonal) and anti-TFEB (A303-673A, rabbit polyclonal) antibodies from Thermo Fisher Scientific, anti-RFP fom Antibodies Online (ABIN129578, rabbit polyclonal), anti-ATG16L1 S278 from AbCam (ab195242, rabbit monoclonal), pan phospho-S/T antibody from ABclonal (AP1067, mouse monoclonal) and anti-GSK3β antibody from BD Biosciences (610201, mouse monoclonal). HRP-labeled goat anti-rabbit (PI31460) or anti-mouse (PI31430) secondary antibodies were purchased from Pierce. All primary antibodies were used at 1:1,000 dilution except anti-H3 (1:3,000). Secondary antibodies were used at 1:10,000 dilution.

Live-cell epifluorescence imaging

For live-cell imaging, cells were plated onto sterile glass-bottom 35-mm dishes (D35-14-1.5-N, CellVis) and grown to 40% confluency at 37 °C with 5% CO2. Cells were transfected, double-starved or not starved, washed once with HBSS and imaged in the dark at room temperature. Images were acquired on a Zeiss Axio Observer Z1 microscope equipped with a ×40 (numerical aperture: 1.3) objective (Carl Zeiss), Prime95B scientific complementary metal–oxide–semiconductor camera (Photometrics) and a motorized stage (Carl Zeiss). The microscope was controlled by MATLAB (Mathworks) and µmanager (Micro-Manager, an open-source microscope imaging software)-based MATScope imaging suite (GitHub: https://github.com/jinzhanglab-ucsd/MatScopeSuite). CFP images were acquired using an ET420/20x excitation filter with a T4551pxt dichroic and AT470/40m emission filter. YFP images were acquired using an ET495/10x excitation filter with a T5151p dichroic and ET535/25m emission filter. The C/Y FRET images were acquired using an ET420/20x excitation filter with T4551pxt dichroic and ET535/25m emission filter. mCherry imaging was performed using an HQ568/55x excitation filter with a Q600LPxr dichroic mirror and HQ653/95m emission filter. EGFP was imaged using an ET480/30x excitation filter with a T505dcxr dichroic and ET535/50m emission filter. Exposure times were 50–500 ms and images were taken every 30 s or 1 min. Imaging data were analyzed using custom MATLAB scripts. Specifically, fluorescence images were background-corrected and regions of interest (ROIs) were manually selected. The C/Y FRET emission ratio was calculated for each ROI on each frame of the image series. The ratio was normalized to the time point before the addition of PDGF (t = 0 min). For Lyso-TORCAR, TORCAR-NLS and TORCAR-NES data, cells with YFP intensity > 50 were analyzed. For mito-TORCAR, cells with YFP intensity > 200 were analyzed.

Live-cell and fixed-cell spinning disk confocal microscopy

NIH3T3 cells expressing Lyso-TerminaTOR were incubated with 100 nM of LysoTracker green in modified HBSS, including 1× HBSS with 2 g L−1 glucose pH 7.4, (diluted from 10× HBSS; 14065, Gibco) for 30 min before imaging to check Lyso-TerminaTOR localization. For immunofluorescence to check mTOR colocalization with LAMP1 in cells expressing mCherry or Lyso-TerminaTOR, NIH3T3 cells plated on glass coverslips expressing either construct were fixed with 4% paraformaldehyde (15710S, Electron Microscopy Sciences) for 30 min at room temperature, washed with PBS, permeabilized using PBS with 0.1% Triton X-100 for 15 min and then blocked with PBS containing 0.1% Triton X-100 and 5% BSA at room temperature for 1 h. Coverslips were incubated overnight with anti-mTOR (Cell Signaling, 2972) and anti-LAMP1 (Santa Cruz, sc20011) primary antibodies diluted 1:100 at 4 °C. Following the overnight incubation, cells were incubated with secondary antibodies anti-rabbit AlexaFluor 488 (Life Technologies, A-11001) and anti-mouse AlexaFluor 562 (Life Technologies, A-21422) diluted 1:1,000 for 1 h at room temperature. Coverslips were mounted on glass microscope slides with Prolong glass antifade mountant with NucBlue (P36981, Invitrogen). A Nikon Ti2 Scope (Nikon) equipped with a W1 confocal scanhead (Yokogawa Electric), Dual Prime95B (Photometrics) cameras, LUN-F-XL laser engine (Nikon), Nano-Drive (Mad City Labs), Galvo XY scanner, Sola light box (Lumencor), Piezo z-stage (Mad City Labs), bandpass filter cubes (Chroma) and an Apo TIRF ×100 (NA: 1.49) objective was used to acquire images with the NIS Elements AR software (Nikon). Laser lines at 488 nm and 561 nm were used to coimage LysoTracker green and mCherry (red) TerminaTOR or mTOR and LAMP1 colocalization. Pearson’s coefficient of correlation between mTOR and LAMP1 (lysosomes) was calculated using the Coloc2 plug-in on Fiji (ImageJ).

mRNA sequencing and analyses

Stable NIH3T3 cells were cultured in growth medium supplemented with doxycycline at a final concentration of 100 ng ml−1 to induce Nuc-TerminaTOR expression or H2A–mCherry expression as control. After 24 h, cells were serum-starved for another 24 h in doxycycline-containing medium. On the day of sample collection, cells were starved of amino acids for 1.5 h in HBSS and treated with CHX (25 μg ml−1) and 100 nM rapamycin or DMSO for 0.5 h to inhibit translation and mTORC1, respectively. Optimal conditions for translation inhibition were determined by imaging mCherry expression in cells with various durations of CHX treatment. Lastly, PDGF (50 ng ml−1) treatment was conducted for 3 h before RNA extraction. Total RNA was extracted using Direct-zol RNA miniprep kits (R2051, Zymo Research) from triplicate samples for each group and sent for RNA sequencing (Novogene, NovaSeq 6000 PE150). Samples that passed quality control were proceeded to paired-end mRNA sequencing. Raw data were analyzed using the web-based platform Galaxy (https://usegalaxy.org) and open-source R software (version 4.1.0; https://www.r-project.org). Briefly, high-quality raw reads were mapped to the reference mouse genome mm10 using HISAT2 (version 2.2.1) and aligned reads were counted using htseq-count (version 0.9.1). The R package DESeq2 (version 1.34.0) was used to conduct DEG analysis. Pathway analysis was conducted to identify enriched pathways among DEGs using the R package clusterProfiler.

Reverse transcription (RT)–qPCR

Stable NIH3T3 cells were treated the same way as in RNA-sequencing experiments. Total RNAs were extracted using Trizol (15596026) and equal amounts of RNA from each sample were reverse-transcribed to complementary DNA (cDNA) using PrimeScript RT master mix (Takara, RR036A) according to the manufacturer’s instructions. Synthesized cDNA was diluted and used to perform real-time PCR using iTaq Universal SYBR green supermix (Bio-rad, 1725121) on a Bio-Rad CFX96 Touch real-time detection system. Ct values from each reaction were collected and the ΔΔCt method was used to calculate the normalized fold change of relative mRNA expression. The primers used for respective genes are listed in Supplementary Table 5.

Luciferase assay

CCAAT motifs from the Rhob gene (268 bp, from −24 to –291 of 5′ untranslated region) were cloned into a pGL3 firefly luciferase reporter construct. Both pGL3-CCAAT-FLuc and a control Renilla luciferase reporter constructs were transfected into NIH3T3 cells, which were incubated with doxycycline for 24 h to express H2A–mCherry or Nuc-TerminaTOR. At the time of transfection, the medium was changed to serum-free medium containing doxycycline. Then, 24 h after transfection, cells were starved in HBSS for 2 h and stimulated with PDGF (50 ng ml−1) for 3 h. A dual-luciferase reporter assay kit (Promega, E1910) was used to measure the luciferase activity. Briefly, cells were lysed and firefly luciferase activity was first measured and then quenched before measuring the Renilla luciferase activity, which was used as internal control to normalize the firefly luciferase activity.

Cell growth assay

NIH3T3 cells were infected with lentivirus to produce stable cell lines that could be induced by doxycycline to express mCherry or Nuc-TerminaTOR. Cells were cultured in growth medium supplemented with 100 ng ml−1 doxycycline. Cells were seeded at 10,000 cells per well, then collected and counted using an automated cell counter (CountLess II, Thermo Fisher) at 48-h time points. Cell numbers were normalized to the seeding number.

Cell viability assay

For the cell viability assay conducted with Cal33 cells, 3,000 cells per well were seeded in 96-well plates and treated with doxycycline (1 µg ml−1) to induce Nuc-TerminaTOR or H2A–mCherry expression after cells were attached to the plates. After treatment for 72 h, the cell culture medium was supplemented with 1:100 of the culture volume of Aquabluer reagent (6015, MultiTarget Pharmaceuticals) for 2 h. The fluorescence was recorded (excitation: 540 nm, emission: 590 nm) using a Spark microplate reader (Tecan).

Colony formation assay

For the colony formation assay conducted with Cal33 cells, 1,000 cells per well were seeded in 12-well plates and doxycycline was added after they were attached to induce Nuc-TerminaTOR or H2A–mCherry expression. Cells were treated for 7–9 days and the medium was changed every 2–3 days. Cell culture plates containing colonies were gently washed with PBS twice, fixed for 5 min with a solution of methanol and acetic acid (3:1) and stained for 15 min with 0.5% crystal violet solution diluted in methanol. Excess stain was removed by washing repeatedly with PBS. The colony area percentage was calculated using ImageJ.

AlphaFold3 structure prediction

The AlphaFold89 server was used to generate structure predictions for TerminaTOR interactions with mTORC1 components. Truncated sequences of mTOR and Raptor or full-length mLST8 were used individually with the TerminaTOR sequence, each as separate entities, to visualize predicted interactions. Interactions were visualized in ChimeraX90.

Statistical analysis

The data were analyzed using GraphPad Prism 9. All data were obtained from at least three independent biological replicates. Unless otherwise noted, an unpaired two-tailed t-test was used for two-group comparisons and a one-way analysis of variance (ANOVA) with Dunnett’s multiple-comparison test was used for multigroup comparisons (****P < 0.0001, ***P < 0.001, **P < 0.01 and *P < 0.05; NS, not significant, P > 0.05). As indicated in the figures, legends and main text, n represents the number of experiments or the number of cells. Shaded areas in average curves indicate the s.e.m. Violin plots and box plots show the upper and lower adjacent values, interquartile range and the median.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.