Transient APC/C inactivation by mTOR boosts glycolysis during cell cycle entry

Cell culture

Human MCF-10A, human retinal pigment epithelial (RPE-1), human lung fibroblasts (HLFs) and transformed human embryonic kidney (HEK293T) cells were obtained from ATCC. MEFs were gift from J. Vidigal. Cell lines not obtained from ATCC were not authenticated. MEFs, HLFs and HEK293T cells were cultured in Dulbecco’s modified Eagles medium (DMEM) (Gibco, Life Technologies) containing 10% FBS (Gibco). MCF-10A cells were cultured in the following full-growth medium: phenol-red-free DMEM/F12 (Invitrogen) supplemented with 5% horse serum, 20 ng ml⁻¹ EGF, 10 µg ml⁻¹ insulin, 500 µg ml⁻¹ hydrocortisone, 100 ng ml⁻¹ cholera toxin and 1% penicillin–streptomycin. To induce quiescence, cells were mitogen-starved for 48–72 h with DMEM/F12 (Invitrogen) supplemented with 0.3% bovine serum albumin, 100 ng ml⁻¹ cholera toxin and 1% penicillin–streptomycin. For starvation experiments longer than 48 h, starvation medium was replaced in each 48 h interval. For contact inhibition mediated quiescence, cells were seeded at 90% confluence and allowed to grow in full growth medium for 48 h to induce quiescence. Where indicated, quiescent cells were transfected with 20% OPTI-MEM (Invitrogen) for 6 h, followed by replenishing with starvation medium for an additional 24 h before mitogen stimulation. To induce quiescence through MEK inhibition, cells in complete medium were treated with 100 nM trametinib for 48 h. RPE-1 cells were cultured in DMEM/F12 (Invitrogen) supplemented with 10% FBS and with 0.01 mg ml⁻¹ hygromycin B. For mitogen-removal experiments, RPE-1 cells were incubated with the above-described composition supplemented with 0.3% BSA and without FBS. MEFs were starved with DMEM containing 0% serum for 72 h. All tissue culture media were supplemented with 2 mM l-glutamine, 25 μg ml⁻¹ streptomycin and 25 U penicillin (Gibco). Cells were cultured in a humidified atmosphere with 5% CO₂ at 37 °C. For all pTet_on experiments, cells were pretreated with doxycycline for 12 h in starved medium followed by induction with mitogen and doxycycline if not otherwise mentioned. Cells were routinely tested for mycoplasma.

Constructs and stable cell lines

CSII-pEF1α-H2B-mTurquoise, CSII-pEF1α-mCherry-Geminin(amino acids 1–110) and CSII-pEF1α-DHB(amino acids 994–1087)-mVenus were described previously^8,13. HA–CDH1 was a gift from M. Santra, HA–CDH1(T129A) and HA–CDH1(T129D) were subcloned using Gibson cloning. CMV-Hylight was a gift from R. Goodman (Addgene, 193447)³⁰. Wild-type PFKFB3 and PFKFB3^KEN-mut (KEN box mutant PFKFB3) were gifts from J. P. Bolanos². Transduced cells were sorted on the BD Biosciences FACS-Aria Fusion system to obtain pure populations expressing the desired fluorescent biosensors. The plasmids were co-transfected with the packaging plasmid (pCAG-HIVgp) and the VSV-G- and Rev-expressing plasmid (pCMV-VSV-G-RSV-Rev) into HEK293T cells. High-titre viral solutions of the target proteins were prepared and used for co-transduction into several cell lines. Then, 72 h later, cells were FACS sorted to obtain a pure population of cells expressing the biosensors against the protein of interest. For making the inducible system, gene bodies were cloned in pSB-Tet on system (Addgene, 60496). Cells were transfected with sleeping beauty transposon, Tet-on gene and selected using puromycin as the selection marker.

Treatment of cells

Cells were treated with vehicle (DMSO, Sigma-Aldrich, D2650), 100 nM of rapamycin (Tocris Biosciences, 1292/1), 10 nM of RMC6272 (Medchem Express, HY-134904), 100 nM of PD0325901 (MEKi) (Selleckchem, S1036), 10 µM of RO3306 (Calbiochem, SIG217714), 1 µM of palbociclib (Selleckchem, S1116), 2.5 µM PFK15 (Selleck Chem, S7289), 5 mM 2-deoxy glucose (Sigma-Aldrich, D8375), d-glucose (U-¹³C₆, 99%) (Cambridge Isotope Laboratories, CLM-1396-2), dialysed horse serum 10 kDa (Bioivt, HSE00SRM-0108261), 30 μM DRB18 (Cayman Chemical, 38217), 5–10 μM MG132 (Calbiochem, 133407-82-6) or a pan-phosphatase inhibitor (20 mg ml⁻¹ of sodium fluoride (Sigma-Aldrich, 7681-49-4) and 40 mg ml⁻¹ of sodium orthovanadate (Sigma-Aldrich, 13721-39-6)) for the indicated timepoints. For doxycycline induction of CDH1(T129A) or CDH1(T129D), cells were incubated with 5 μM doxycycline for 8 h in the starvation medium followed by mitogen stimulation with doxycycline. Similarly, for PFK15 and 2-DG treatment, cells were pretreated with respective drugs for 6 h in the starvation medium followed by mitogen stimulation with respective drugs. The cells were either collected after treatment and whole-cell extracts were prepared or imaging was continued. For experiments without glucose, cells were starved as described above with starvation medium, followed by replacement with starvation medium without glucose for the last 12–16 h. Cells were either stimulated with starvation medium containing 1,000 pg ml⁻¹ of EGF with or without glucose. For RPE-1 cells and MEFs, cells were starved for 48 h, followed by 12–16 h of starvation without glucose. Cells were stimulated to enter the cell cycle by replacing the starvation medium with 2% FBS containing DMEM-F12 or DMEM medium, respectively, with or without glucose.

Antibodies

The following antibodies were used in this study; CDH1 (FZR1) antibody (Abcam, Ab217038, 1:1,000, IP: 2 µg; Santa Cruz, sc-56312, IB: 1:800; Sigma-Aldrich, CC43-100UG, 1:2,000), mTOR antibody (CST, 2972, 1:1,000), p-mTOR (CST, 2971, 1:1,000), mCherry (Abcam, ab167453, 1:1,000), geminin (CST, 5165, 1:1,000), cyclin D1 (Thermo Fisher Scientific, MA5 14512, 1:750), p-Rb (Ser807/811) (CST, 8516, 1:1,000), Rb (CST, 9309, 1:2,000), vinculin (Sigma-Aldrich, V9131, 1:10,000), Ki-67 (Abcam, ab8191, 1:1,000), p21 (BD Biosciences, 556430, 1:1,000), p27 (CST, 3686, 1:1,000), His tag (Santa Cruz, sc-8036, 1:800), anti-DDK/Flag (Sigma-Aldrich, F3165, 1:1,000), p-S6 kinase (CST, 9205, 1:1,000), APC2 (CST, 12301, 1:1,000), APC6 (CST, 9499, 1:1,000), APC11 (CST, 14090, 1:1,000), S6 kinase (CST, 9202, 1:1,000), p-4EBP1 (CST, 2855, 1:1000), cyclin A2 (Santa Cruz, sc-271682, 1:500), EMI1 (Santa Cruz, sc-365212, 1:500), cyclin F (Santa Cruz, sc-515207, 1:500), βTrCP (CST, 4394, 1:1,000), pSer/Thr/Tyr (Thermo Fisher Scientific, 61-8300, 1:1,000), pThr (Abcam, ab9337, 1:500), pTyr (Abcam, ab10321, 1:1,000), pan anti-phospho-serine/threonine antibody (Phospho Solutions, PP2551; 1:2,000), HA (Santa Cruz, sc-7392, 1:800, IP: 2 µg; and CST, 3724, 1:1,000), GST (Santa Cruz, sc-138, 1:750), PFKFB3 (MBS, 9604769, IB: 1:750, IP: 2 µg), PFKFB3 (Abcam, AB181861-1001, 1:4,000), PFKFB2 (CST, 13029, 1:1,000), PFKFB1 (Abcam, ab155564, 1:1,000), β-actin (Abcam, ab6276, 1:2,000), RPTOR (CST, 2280, 1:1,000), RICTOR (CST, 2114, 1:1,000), mLST8 (CST, 3274, 1:1,000), ubiquitin (Santa Cruz, sc-8017, 1:800), TSC1 (CST, 6935, 1:1,000), NPRL2 (CST, 37344, 1:1,000), histone H1 (Abcam, 11079, 1:1,000), EGFR (Santa Cruz, sc-373746, 1:800), AKT (CST, 9272, 1:1,000), p-AKT (CST, 4060, 1:1,000), mouse anti-goat IgG-HRP (Santa Cruz, sc-2354, 1:10,000), anti-rabbit IgG, HRP-linked antibody (CST, 7074, 1:10,000), mouse anti-rabbit secondary-HRP (Santa Cruz, sc-2357, 1:2,500) or recombinant anti-mouse (Santa Cruz, sc-516102, 1:2,500), anti-mouse IgG, HRP-linked antibody (CST, 7076, 1:10,000), normal rabbit IgG (CST, 2729, IP: 2 µg), normal mouse IgG (Santa Cruz, sc-2025, IP: 2 µg).

siRNA transfection

The indicated cells were transfected using Dharmafect 1 (Horizon) according to the manufacturer’s instructions. The following siRNAs were used: On-Target plus control siRNA (nontargeting, Dharmacon), On-Target plus pooled set of four siRNAs for CCNA2 (M-003205-02-0005), FZR1 (L-004086-00-0005), EMI1 (M-012434-01-0005), CCNF (custom, 5′-GCACCCGGUUUAUCAGUAAUUUU-3′), BTRC (encoding βTrCP) (L-003463-00-0005 and L-003490-00-0005), PFKFB3 (L-006763-00-0005), TSC1 (L-003028-00-0005), NPRL2 (L-015645-00-0005), CDH1 3′ UTR (Qiagen, SI04955265|S1 (1027417) at final concentrations of 20 nM unless noted. Then, 6 h after transfection, the medium was replaced with starvation medium and imaging was started 24 h later.

Time-lapse microscopy

Before imaging, cells were plated in a 96-well plate (Ibidi, 89626) and allowed to grow in full growth medium for 24 h. Time-lapse imaging was conducted in 200 μl full growth medium, with images captured in CFP, YFP and RFP channels every 12 min using a Nikon Ti2-E inverted microscope (Nikon) with NIS elements (V5.11.00) equipped with ×10/0.45 NA and ×20/0.75 NA Plan Apo objectives. To minimize phototoxicity, the total light exposure time was kept under 300 ms (30 ms for CFP, 200 ms for YFP and 300 ms for mCherry) for each timepoint. The cells were imaged in a humidified chamber at 37 °C with 5% CO₂. To ensure minimal overlap, four or six sites were imaged per well with their positions spaced apart. For experiments involving drug treatments, cells were first imaged without drugs, and the video was then paused to exchange fresh medium in each well with the desired drug concentration. For experiments with the HYlight biosensor, the following filter sets were used: Sapphire: Chroma ET395/25x excitation filter, Chroma ET525/50m emission filter, Chroma T425lpxr dichroic; GFP: Chroma ET480/20X excitation filter, Chroma ET510/20 m emission filter, Chroma T495lpxr dichroic. Cell tracking and data analysis were performed using custom MATLAB scripts.

Cell lysate preparation and immunoblotting

The cells were washed twice with ice-cold PBS and collected. Whole-cell lysis buffer (50 mM Tris pH 7.4, 200 mM NaCl, 50 mM NaF, 1 mM Na₃VO₄, 0.5% Triton X-100 and protease inhibitor cocktail) was added to the cells, and the cells were lysed on ice for 30 min. The lysates were centrifuged at high speed (16,000g), and the clear supernatants were transferred into new tubes. The protein concentration was measured using the BCA method, with BSA used as a standard. The samples were prepared in SDS sample buffer and run in SDS–PAGE running buffer (Bio-Rad, 1610772). The separated proteins were transferred onto a PVDF membrane using the Bio-Rad semidry transfer system (Bio-Rad, 10026938). The membranes were incubated with the primary antibody overnight at 4 °C and subsequently with an HRP-conjugated secondary antibody for 1 h at room temperature. The blots were developed using the chemiluminescence method, and densitometry analysis of the immunoblots was performed using ImageJ software.

Immunoprecipitation

The cells were lysed as described above (see the ‘Cell lysate preparation and immunoblotting’ section). A total of 300–500 μg of whole-cell lysate was co-immunoprecipitated with 2 μg of antibody and IgG control in 500 µl of modified IP lysis buffer (50 mM Tris pH7.4, 200 mM NaCl, 50 mM NaF, 1 mM Na₃VO₄, 0.1% Triton X-100 and protease inhibitor cocktail). The protein–antibody mixture was kept at 4 °C in a rotor with gentle rocking for 12–16 h. The next day, the mixture was allowed to bind to protein G agarose beads for 2 h at 4 °C with gentle rocking. The immunoprecipitated proteins were eluted from the beads using Laemmli buffer for 3–5 min and boiled before resolving on SDS–PAGE. In all immunoprecipitation assays, 3% of the proteins taken in the immunoprecipitation experiment were used as input.

For co-IP of purified protein, 200 ng of His–CDH1 was allowed to bind to the beads overnight followed by washing and addition of purified mTOR. The whole complex was kept at 4 °C with rotation for 1 h followed by washing and elution using 1× SDS-Laemmli buffer. For untagged purified protein co-IPs, magnetic Flag beads (A36797, Thermo Fisher Scientific) were used and DDK–mTOR (Sigma-Aldrich, SRP0364)/DDK–DYRK4 (Carna Biosciences, 04-434) was allowed to bind to the beads overnight at 4 °C followed by washing and addition of purified WT or mutant CDH1 or His–SUMO–CDH1 (Mybiosource, MBS1437931). The whole complex was kept at 4 °C with rotation for 1 h followed by washing and elution using 1× SDS-Laemmli buffer.

In vitro kinase assay

A total of 200 ng of untagged human CDH1 or His–SUMO–CDH1 purified protein was incubated with 200 ng of DDK-tagged human mTOR protein (Carna Biosciences or Sigma-Aldrich), GST–CDK1–cyclin B (Carna Biosciences, 04-102), GST–ERK (Reaction Biology, 0883-0000-1) or Flag–DYRK4 (Carna Biosciences, 04-434) in 1× kinase buffer (Cell Signaling Technology, 9802). The assay was performed with 100 µM of ATP (Cell Signaling Technology, 9804) at 37 °C for 30 min, followed by addition of 4×-reducing SDS sample buffer and heating at 95 °C for 10 min. The reaction was resolved by SDS–PAGE and immunoblotting was performed as stated above.

Ubiquitination assay

Before collection, cells were treated with 10 μM of MG132 for 3–4 h (except where otherwise mentioned). After collection, whole-cell lysates were prepared and 300–500 μg of lysate was subjected to immunoprecipitation using the indicated antibodies. The immunoprecipitate obtained was separated using SDS–PAGE and probed with anti-ubiquitin antibodies to determine the levels of ubiquitylation.

In vitro phosphorylation of CDH1 for ubiquitination assay

CDH1 purified from baculoviral infected insect cells was subjected to phosphorylation by recombinant Flag–mTOR + mLST8 (Carna Biosciences, 11-431). Then, 1 µM CDH1 WT, T129A or T129D was mixed with buffer or 10 ng µl⁻¹ mTOR and 200 µM of MgCl₂-ATP for 30 min at 30 °C, after which the reactions were stored on ice. All of the reactions (control or containing mTOR) were treated equally). All further steps were performed either on ice or at 4 °C unless otherwise stated. The reactions were quenched by removal of ATP through desalting with a Zeba Spin column (Thermo Fisher Scientific, PI89882) into 20 mM HEPES pH 7.0, 400 mM AmSO₄, 100 mM NaCl with or without 2.5% glycerol according to the manufacturer’s specifications. Flag–mTOR was removed from the reactions, desalted into glycerol-containing buffer through addition of 40 µl of 50:50 Flag-resin slurry (GenScript, L00432) equilibrated with glycerol-containing buffer. The slurry was rotated in batch for 1 h on ice, then Flag resin was pelleted by centrifugation at 1,500g. The supernatant was removed and used for further reactions. The concentration of reacted CDH1 was checked by Nanodrop and confirmed with Coomassie staining. Phosphorylation was confirmed by western blotting with pan-anti-phospho-serine/threonine antibody (Phospho Solutions, PP2551; 1:2,000) or CDH1 (Sigma-Aldrich, CC43-100UG; 1:2,000), mouse anti-rabbit secondary-HRP (Santa Cruz, sc-2357; 1:2,500) or recombinant anti-mouse (Santa Cruz, sc-516102; 1:2,500) and visualized with Clarity ECL (Bio-Rad, 1705060).

In vitro ubiquitination assays

Ubiquitination assays were performed with recombinant APC/C, UBE2C, UBA1, substrates, ubiquitin and CDH1 as previously described^38,39. Then, 100 nM of fluorescently labelled cyclin A2 (FAM-cyclin A) or the N-terminal domain (NTD) of cyclin B (FAM-cyclin B^NTD) was incubated on ice with 100 nM UBA1, 300 nM UBE2C, 5 mM Mg-ATP, 30 nM APC/C and 30 nM CDH1. The components were equilibrated to room temperature for 5 min before reactions were started with addition of 100 µM ubiquitin. The reactions were quenched after 20 min with 4× SDS–PAGE loading buffer, separated on 4–12% Bis-Tris gels, and imaged on the Amersham Typhoon imager. For in vitro ubiquitination assays with recombinant phosphorylated CDH1, 40 nM CDH1 was mixed with 30 nM APC/C, 100 nM UBA1, 300 nM UBE2C, 250 nM CycB-NTD* and 5 mM MgCl₂-ATP on ice. The reaction was allowed to come to room temperature, and 100 µM Ub was added. The reactions were quenched after 20 min with 4× SDS buffer. Proteins were separated on 4–12% Bis-Tris SDS–PAGE gels (GenScript, M00654) and CycB-NTD* ubiquitination was visualized on the Typhoon fluorescence scanner. Ubiquitination was quantified in ImageQuant software and normalized to CDH1 WT without mTOR.

Image analysis

Custom MATLAB scripts were used for all image analyses, according to previously described methods⁸. In brief, optical illumination bias was empirically determined by sampling background areas in all wells during an imaging session, and this information was then used to flatten all images. This enabled a global background to be measured and subtracted from each image. Cells were segmented based on their nuclei using either Hoechst staining for fixed-cell imaging or H2B–mTurquoise for live-cell imaging. The procedure for determining APC/C and CDK2 activity was previously reported¹³.

OPP assay

Cells were seeded and starved for 48 h. After 24 h of the starvation, empty vector, CDH1(T129A) or CDH1(T129D) mutant was ectopically expressed in the cells and after 6 h of transfection medium was replaced with starvation medium. Cells were released as described above and incubated with 20 µM of O-propargylpuromycin (OPP) for 30 min followed by fixation with 4% paraformaldehyde for 15 min at room temperature in the dark. Further procedures were performed according to the manufacturer’s instructions (Thermo Fisher Scientific, C10458).

Glycolysis and ATP rate measurements by extracellular flux assays

To measure glycolytic function of cells, the Glycolysis Stress Test Kit (Agilent technologies, 103020-100) was used. To measure the glycolytic bioenergetics of cells, the Glycolytic Rate Assay Kit (Agilent technologies, 103344-100) was used. To distinguish between the fraction of ATP produced from mitochondrial OXPHOS and glycolysis, the ATP Rate Assay Kit (Agilent technologies, 103591-100) was used. For all of the assays, cells were initially seeded at a density of 10,000 cells per well in an XF96-well plate precoated with poly-lysine. For ectopic overexpression or silencing, Lipofectamine 3000 or Dharmafect 1 (Horizon) were used to transfect cells according to the manufacturer’s instructions. After 24 h of transfection, cells were released with complete medium with or without drugs for another 3 h. After release, for the glycolysis stress test, cells were washed and incubated with Seahorse XF DMEM medium, pH 7.4 (Agilent Technologies, 103575-100) supplemented with 2 mM glutamine. Glucose, oligomycin and 2-DG were added to cells sequentially at the specific timepoint with final concentrations of 10 mM, 1 μM and 50 mM, respectively. For the glycolytic rate assay, cells were washed and incubated with Seahorse XF DMEM medium, pH 7.4 (Agilent Technologies, 103575-100) supplemented with 10 mM glucose, 1 mM pyruvate and 2 mM glutamine. Antimycin/rotenone and 2-DG were added to cells sequentially at the specific timepoint with final concentrations of 0.5 μM and 50 mM, respectively. For the ATP rate test, cells were washed and incubated with Seahorse XF DMEM medium, pH 7.4 (Agilent Technologies, 103575-100) supplemented with 10 mM glucose, 1 mM pyruvate and 2 mM glutamine. Oligomycin and antimycin/rotenone were applied with final concentrations of 1.0 μM and 0.5 μM, respectively. The oxygen-consumption rate and ECAR was collected using the Agilent Seahorse Wave 2.6.1 desktop software. Data were normalized to the cell number determined using the crystal violet assay and analysed using the Wave software.

ATP measurements

Cells were seeded at a density of 5,000 cells per well in a 96-well plate and incubated with complete growth medium for 24 h. The medium was then replaced with starvation medium and incubated for another 24 h. The cells were then transfected with either siControl or siPFKFB3 (using Dharmafect, Horizon) or CDH1(T129A) and CDH1(T129D) (using Lipofectamine 3000, Invitrogen). Then, 6 h after transfection, the medium was replaced with starvation medium. After another 24 h, the cells were mitogen-stimulated to promote cell cycle entry. Then, 4 h after stimulation, the medium was replaced with chilled PBS and the sample was kept on an iron slab on ice to stop or slow down the metabolic turnover^40,41. The cells were collected and the assays were performed according to the manufacturer’s protocols (Abcam, ab113849).

qPCR with reverse transcription

Total RNA was extracted from cultured cells using the RLT RNeasy 96 Qiacube HT Kit (Qiagen, 74171) and on-column DNA digestion was performed using the Qiagen DNA digestion kit (79254). The cDNA was synthesized using the Bio-Rad cDNA preparation kit according to the manufacturer’s instructions, starting with 1 µg of total RNA. Quantitative PCR (qPCR) was performed using the Bio-Rad SYBR green qPCR Mix. The expression of GAPDH was used as a reference gene to normalize the mRNA level of the gene of interest. The relative mRNA levels were determined by comparing the treated or transfected samples to the untreated or vector control, which was considered as 1. The following primers were used: GAPDH forward, AATCCCATCACCATCTTCCA; GAPDH reverse, TGGACTCCACGACGTACTCA; PFKFB3 forward, GGTACCGAATCAAGCAGAGC; PFKFB3 reverse, GCAGTAGGAGGACGAGTTGG.

Mathematical model

The incoherent feedforward model involving mTOR and a delayed protein phosphatase activity towards the APC/C was mathematically modelled using a set of ordinary differential equations (ODEs) as shown below:

1. (1)

   dMTOR/dt = 1/τ_mTOR(\(S^(n_1)/(k_1^(n_1)+S^(n_1))-\rmm\rmT\rmO\rmR\))

2. (2)

   dPP/dt = 1/τ_PP(\(S^(n_2)/(k_2^(n_2)+S^(n_2))-\rmPP\))

3. (3)

   dpCDH1/dt = 1/τ_pCDH1(\((\rmm\rmT\rmO\rmR^(n_3)/(k_3^(n_3)+\rmm\rmT\rmO\rmR^(n_3)))(k_4^(n_4)/\)\((k_4^(n_4)+\rmP\rmP^(n_4)))-\rmp\rmC\rmD\rmH1\))

4. (4)

   dAPC/dt = 1/τ_APC(\(k_5^(n_5)/(k_5^(n_5)+\rmp\rmC\rmD\rmH1^(n_5))-\rmA\rmP\rmC\))

5. (5)

   dGEMININ/dt = 1/τ_GEMININ(\(k_6^(n_6)/(k_6^(n_6)+\rmA\rmP\rmC^(n_6))-\rmG\rmE\rmM\rmI\rmN\rmI\rmN\)).

The parameters for the model were as follows: k₁ = 0.2, n₁ = 1, t_mTOR = 0.5; k₂ = 0.2, n₂ = 1, t_PP = 0.5; k₃ = 0.3, n₃ = 1, t_pCDH1 = 0.5; k₄ = 0.2, n₄ = 2, t_APC = 0.5; k₅ = 0.8, n₅ = 1, t_GEMININ = 0.3; k₆ = 0.4, n₆ = 3.

The initial conditions for the model were as follows: mTOR = 0, PP = 0, pCDH1 = 0, APC = 1, GEMININ = 0.

The following notation is used: S, mitogens (for example, serum); mTOR, mTOR activity; PP, protein phosphatase activity; pCDH1, phosphorylated CDH1; APC, APC/C activity; GEMININ, APC/C substrate levels (for example, geminin).

Each system of ODEs was solved in RStudio (v.1.3.1093) using the ode function (from the deSolve R package) using the LSODA algorithm. Most parameters were chosen on the basis of previous models^9,42, but mTOR and phosphatase activation kinetics were chosen to match data from this study. To model the effect of mitogen stimulation, the initial conditions of the model were set as listed above and the model was solved for S = 3. To model the delay in protein phosphatase activity as measured experimentally, the PP term was kept at zero until 4 h after mitogen stimulation.

In vitro purification of CDH1

Baculovirus-induced expression of 3×Myc–6×His–CDH1 wild type and variants was performed by infecting Trichoplusia ni cells in ESF921 medium (Expression Systems). All purification steps were completed at 4 °C. Cells were collected about 2.5 days later, pelleted and resuspended in 20 ml of lysis buffer (20 mM HEPES pH 8.0, 100 mM AmSO₄, 2.5% glycerol, 10 mM imidazole supplemented with 10 µg ml⁻¹ leupeptin, 20 µg ml⁻¹ aprotinin, 5 U ml⁻¹ benzonase, 2 mM benzamidine, 2.5 mM PMSF and 1 Roche EDTA-free protease inhibitor tablet per 50 ml buffer) per litre. Cells were lysed by sonication and clarified by centrifugation at 30,000g for 1 h. The lysates were incubated in batch with Ni-NTA resin (Genesee Scientific) for 1 h before centrifugation of beads at 500g for 10 min. After resuspending and washing the resin with 10 column volumes of wash buffer (20 mM HEPES pH 7.0, 100 mM AmSO₄, 2.5% glycerol and 10 mM imidazole), CDH1 was eluted with wash buffer supplemented with 300 mM AmSO₄ and 250 mM imidazole. The affinity tags were removed by the addition of HRV-3C protease for 1 h on ice. The protein solution was diluted back to 100 mM AmSO₄ with wash buffer lacking salt or imidazole and further purified by cation-exchange chromatography with SP-Sepharose resin (Thermo Fisher Scientific). Protein was eluted with 20 mM HEPES pH 7.0, 100 mM AmSO₄, 400 mM NaCl, 2.5% glycerol and 2 mM dithiothreitol and flash frozen with liquid nitrogen in small aliquots.

Protein digestion

Recombinant protein (500 ng) was digested by Arg-C (Promega) at a ratio of 1:50 (Promega) and incubated overnight at 37 °C. The digested peptide samples were acidified by formic acid to a final concentration of 1% and desalted using Pierce peptide desalting columns according to the manufacturer’s protocol (Thermo Fisher Scientific). Peptides were eluted from the columns using 50% acetonitrile/0.1% formic acid, vacuum centrifuged to dry and stored at −80 °C until analysed by MS.

MS acquisition and data analysis

Dried peptide fractions were reconstituted in 0.1% TFA and subjected to nanoflow liquid chromatography (Thermo Ultimate 3000 RSLC nano LC system, Thermo Fisher Scientific) coupled to an Orbitrap Eclipse mass spectrometer (Thermo Fisher Scientific). Peptides were separated using a low-pH gradient using a 5–50% acetonitrile over 120 min in mobile phase containing 0.1% formic acid at a flow rate of 300 nl min⁻¹. Full MS1 scans were performed in the Orbitrap at a resolution of 120,000 with an ion accumulation target set at 4 × 10⁵ and a max IT set at 50 ms over a mass range of 400–1,600 m/z. Ions with determined charge states between 2 and 5 were selected for MS2 scans in the orbitrap with HCD fragmentation (NCE 30%; maximum injection time, 22 ms; AGC 5 × 10⁴) at a resolution of 15,000.

Acquired MS/MS spectra were searched against FZK-Human fasta along with a contaminant protein database, using a SEQUEST and Fixed Value PSM Validator in the Proteome Discoverer v.2.4 software (Thermo Fisher Scientific). The precursor ion tolerance was set at 10 ppm and the fragment ions tolerance was set at 0.02 Da along with methionine oxidation and phosphorylation of serine, threonine and tyrosine included as dynamic modification. Arg-C was specified as the proteolytic enzyme, with up to two missed cleavage sites allowed. The HCD fragmentation pattern of the peptide [KGLFT*YSLSTKR] is shown. The red colour peaks shown in Extended Data Fig. 5 represent the matched b+ ion from the peptide. The blue colour peaks represent the y+ ions from the fragmented peptide. Peptide fragmentation ladder confirms the sequence of the peptide along with phosphorylation site of Thr129. Furthermore, the neutral loss of the labile phosphate group from Thr129 confirms the phosphorylation site, shown as the yellow highlighted peak in the spectra.

Sample preparation for metabolomics

Cells were serum-starved for 48 h to induce quiescence. After starvation, cells were transfected with either empty vector or CDH1(T129A). The medium was replaced with fresh starvation medium at 6 h after transfection. Cells were washed twice with 1× PBS 24 h after transfection and the medium was replaced with glucose-free starvation medium for 1 h. Cells were then incubated with either standard starvation medium or mitogen-rich medium containing dialysed horse serum supplemented with labelled glucose (Cambridge Isotope Laboratories, CLM-1396-2). Cells were washed with 1× PBS, placed on ice, and collected by scraping followed by quick freezing in liquid nitrogen for subsequent analysis.

Reversed-phase ion-pairing LC–MS/MS analysis for cell ¹³C₆-glucose isotope tracing

The unlabelled glycolytic metabolite reference compounds were purchased from Sigma-Aldrich. OmniSolv LC–MS grade acetonitrile and methanol were obtained from EMD Millipore. Tributylamine (TBA), LC–MS grade acetic acid and formic acid were purchased from Thermo Fisher Scientific. All chemicals and solvents used in this study were HPLC or reagent grade unless otherwise noted.

Metabolite extraction was performed by adding 200 µl chilled 80% methanol–water solution to the cell pellet as previously described⁴³. The sample was vortexed vigorously for 30 s and centrifuged at 14,000g for 10 min. Then, 50 µl supernatant was transferred to an autosampler vial. The sample was dried with the SpeedVac vacuum concentrator (Thermo Fisher Scientific) and then reconstituted in 60 µl 3% (v/v) methanol in water. Then, 10 µl sample was injected for reversed-phase ion-pairing LC–MS/MS analysis. The LC–MS/MS analysis was performed using the Thermo TSQ Quantiva triple quadrupole mass spectrometer (Thermo Fisher Scientific) coupled to the NexeraXR LC system (Shimadzu Scientific Instruments). Both the HPLC and mass spectrometer were controlled by the Xcalibur software v.4.1 (Thermo Fisher Scientific). Reversed-phase ion-pairing liquid chromatography was carried out on a 100-mm-long, 2.1-mm-inner-diamter Synergi Hydro-RP C18 column with 2.5 µm particles and a 100 Å pore size (Phenomenex) and kept at 40 °C. The mobile phase, operating at a flow rate of 200 µl min⁻¹, consisted of 10 mM TBAA in water as solvent A and methanol as solvent B. For the current analysis, a linear gradient was held at a B/A solvent ratio 3:97 for 3 min, followed by a B/A solvent ratio of 80:20 for 14 min. After washing with 98% B for 3 min, the column was re-equilibrated with a mobile-phase composition B/A of 3:97 for 10 min before the next injection. The general MS conditions were as follows: source: ESI; ion polarity: negative; spray voltage: 2,500 V; sheath and auxiliary gas: nitrogen; sheath gas pressure: 40 arbitrary units; auxiliary gas pressure: 5 arbitrary units; ion transfer capillary temperature, 350 °C; scan type: selected reaction monitoring (SRM); collision gas: argon; collision gas pressure: 2 mTorr. The ¹³C₆-glucose isotope tracing of the targeted cell glycolysis isotopomer/isotopolog analysis peak identifications and integrations was carried out using Xcalibur Quan Browser (Thermo Fisher Scientific). The SRM conditions and natural isotope abundance corrections analyses were adapted from the previous publications^44,45. The final peak area data were normalized to the sample protein concentration.

Statistical analysis and reproducibility

Statistical analyses were performed in MATLAB (MathWorks, vR2020b) and Prism (GraphPad 9, v.9.2.0). Specific statistical tests used are noted in the figure legends. No statistical method was used to determine sample size. No data were excluded from the analyses. The experiments were not randomized, but unbiased, automated analysis was used to ensure that observer bias did not influence the experimental results.

Reporting summary

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