signature=1c1223d10236c18f125cb9f7a490588e,A chemical biology screen identifies a vulnerability of n...

Large-scale screen (482 cancer cell lines)

Cells were seeded in recommended growth media in black 384-well tissue culture treated plates at 500 cells per well. Cells were equilibrated in assay plates via centrifugation and placed in incubators for 24 h before treatment. At the time of treatment, a set of assay plates (which did not receive treatment) was collected and ATP levels were measured by adding ATPLite (Perkin Elmer) on Envision Plate Readers. Treated assay plates were incubated with compounds for 72 h and subsequently were developed for endpoint analysis using ATPLite. Each compound was assessed using a nine-point dilution curve. Cells were defined as sensitive to a given compound based on GI75 

SCLC cell lines

SCLC cell lines were obtained from the American Type Culture Collection (ATCC), with the following exceptions: LK2 cells were obtained from the Japanese Collection of Research Bioresources Cell Bank (JRCB), LU139 cells from Riken BRC, and SCLC-21H cells from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ). Dulbecco’s modified Eagle’s medium (DMEM)/F12 media (11320033, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (F2442, Sigma) in a humidified atmosphere (5% CO2) at 37 °C was used for all experiments involving SCLC cell lines. ACL4 medium was used in the experiments requiring serum-free medium.

All cell lines were routinely assessed for mycoplasma using the MycoAlert Mycoplasma Detection Kit (LT07–418, Lonza) and tested negative.

Drug sensitivity in the 42 SCLC cell line panel

For each cell line, cell growth was measured in triplicate across a ten-point dose curve of NB-598 or other cholesterol pathway inhibitor. The parameter mu was calculated as the log of the ratio of the CellTiter-Glo signal at 96 h to the signal at 0 h, and mu.max for each cell line was defined as the average mu value corresponding to DMSO-treated cells. For each cell line, a sigmoid function was fit to mu/mu.max as a function of log([drug]), and the AUC was defined as the integral of the fitted curve. For categorical analysis, AUC values <2.0 were considered sensitive, AUC values between 2.0 and 9.0 were considered moderate, and AUC values >9.0 were considered insensitive.

Individual cell viability assays

Cells were plated at 1 × 104 cells per well for suspension cell lines and 2 × 103 cells per well for adherent lines in 96-well plates (in triplicates) and were treated with increasing concentrations of inhibitors. CellTiter-Glo (CTG) cell viability assay was performed according to the manufacturer’s instructions (G7570, Promega) at T = 0 and at the end of the experiment, T = 96 h. An ATP standard curve was used for every CTG readout to normalize for batch-to-batch variation and slight deviations in incubation times, thus allowing rigorous comparisons of fold growth during the course of the assay. Growth rates (mu) were calculated using the following formula:

mu = LN(Tend-blank)/(T0-blank)]/time(h)

Mu/mu.max calculations were used to compare growth rates of drug-treated to DMSO-treated cells (mu.drug/mu.DMSO), where maximum growth is observed in DMSO-treated cells. Value of mu/mu.max = 1 corresponds to no effect of drug added. Values of mu/mu.max 

Small molecule inhibitors

The following compounds were purchased from commercial vendors: fluvastatin (S1909, Selleck Chemicals), atorvastatin (S2077, Selleck Chemicals), lapaquistat (SC-488705, Santa Cruz Biotechnology), Ro48–8071 (10006415, Cayman Chemical), AY 9944 (14611, Cayman Chemical), NB-598 (N4536, Sigma), and AZD-3988 (4837, Tocris).

Azalanstat, SR31747, and F12511 were synthesized using previously published methods

Generation of FDFT1 KO clones

Cas9 protein was transiently transfected into target cells using Lipofectamine CRISPRMAX reagent (CMAX00001, Life Technologies) followed by transient transfection of an in vitro transcribed sgRNA targeting exon 6 (ACGGCCAAGTCAATATTCTC) using RNAiMAX (13778075, Life Technologies). Forty-eight hours post transfection, single cells were selected for clonal expansion and target gene KO was assessed by FDFT1 antibody.

sgSQLE experiments

Cells expressing constitutive Cas9 protein were generated using the pRCCB-CMV-Cas9–2A-Blast plasmid (SVC9B-PS, Cellecta) packaged into VSV-G pseudotyped lentiviral particles. Cells were subsequently infected with sgRNA-expressing lentiviral vectors. After 48 h puromycin selection, cells were plated for colony-formation assays (see below).

Sequences of the sgRNAs used:

sgGFP: AAGATCGAGTGCCGCATCAC

sgPCNA: CCAGGGCTCCATCCTCAAGA

sgSQLE-1: GAAAACAATCAAGTGCAGAG

sgSQLE-2: GCAGCTGTGCTTTCCAGAGA

Immunoblotting

Cells were washed once with 1× phosphate-buffered saline (PBS) and harvested in 1× RIPA buffer (BP-115, Boston Bioproducts) containing phosphatase and protease inhibitor cocktail (5872S, Cell Signaling Technologies). Cell lysates were briefly sonicated and subsequently cleared by centrifugation at 14K rpm for 10 min at 4 °C. Protein quantification was done by Pierce BCA Assay (23225, Life Technologies). For immunoblotting analysis, lysates were loaded onto precast sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels (5671093, Bio-Rad) and subsequently transferred onto nitrocellulose membrane for detection. All primary antibodies were probed overnight at 4 °C, and membranes were washed with TBST and incubated with appropriate secondary antibodies for 1 h. Subsequently membranes were washed with TBST and visualized using Odyssey imaging system (LI-COR).

Primary antibodies used were FDFT1 (13128–1-AP, Proteintech, 1:1000), and β-actin (3700S, Cell Signaling, 1:5000). Secondary antibodies used were IRDye 680RD Donkey anti-Rabbit (926–68073, LI-COR, 1:5000) and IRDye 800CW Donkey anti-Mouse (926–32212, LI-COR, 1:5000).

Uncropped images are provided as Supplementary Fig. 9.

Lentiviral techniques

Lentivirus-based constructs were made using the standard protocol from The RNAi Consortium protocol from the Broad Institute (http://portals.broadinstitute.org/gpp/public/resources/protocols).

qPCR analyses

RNA was extracted using the Qiagen RNeasy Plus Mini Kit according to the manufacturer’s protocol (74136, Qiagen). cDNA was synthesized by converting extracted RNA using the High Capacity cDNA Synthesis Kit from Applied Biosystems according to the manufacturer’s protocol (4368813, Applied Biosystems). Relative gene expression levels were monitored using the following Taqman assays from Applied Biosystems:

HMGCR (Hs00168352_m1)

FDFT1 (Hs00926054_m1)

SQLE (Hs01123768_m1)

LDLR (Hs01092524_m1)

ASCL1 (Hs00269932_m1)

NEUROD1 (Hs01922995_s1)

Reactions used Advanced Fast Master Mix (4444557, Applied Biosystems) and CT values were normalized to β-actin (4326315E, Applied Biosystems), as the endogenous control.

CRISPR screen

A427 cells expressing constitutive Cas9 protein were generated using the pRCCH-CMV-Cas9–2A-Hygro plasmid (SVC9-PS, Cellecta) packaged into VSV-G pseudotyped lentiviral particles. A custom sgRNA library (Cellecta), containing 52,331 sgRNAs targeting 6461 genes, was similarly packaged into lentiviral particles. A427-Cas9 cells were infected with the sgRNA lentiviral library at multiplicity of infection = 0.3 and a cell density sufficient to maintain a minimum of 5.5 × 107 cells after 24 h of puromycin selection. After puromycin selection, the cells were cultured in the presence of 50 nM NB-598 or DMSO. Media was changed every 2 days until a minimum of 1 × 107 cells was reached in the NB-598-treated samples. Control samples (treated with 50 nM NB-598 but without the sgRNA library) were completely killed by drug treatment during the course of the experiment.

Upon completion of the screen, genomic DNA was extracted using phenol:chloroform method, followed by ethanol precipitation, and the sgRNA barcodes were amplified using the following primers:

FwdU6–3: ATTAGTACAAAATACGTGACGTAGAA

R2: AGTAGCGTGAAGAGCAGAGAA

The PCR product was sequenced using the Hiseq 4000 Sequencing Platform with 80–100 million single-end 50 bp reads per sample.

The constant fixed sequence immediately before 22 nucleotide barcode sequences was first removed before read mapping using AWK command in Unix. To determine the read count value for each sgRNA, the trimmed barcode reads were next mapped to custom sgRNA reference library using Bowtie2 (version 2.1.0) with option -L 22 -N 0 -k 1. These criteria retained only the perfectly matched reads for downstream analysis. The raw read count was normalized to the library size by dividing the read count by the total number of mapped reads.

For each sgRNA, log2-fold change of normalized read count between NB-598- and DMSO-treated conditions was calculated. For each gene, the median log2-fold change of its targeting sgRNAs between NB-598 and DMSO condition was determined to represent gene-level enrichment or depletion. The gene-level sgRNA abundance is calculated by taking the sum of the median sgRNA abundance of NB-598 and DMSO conditions in reads per million mapped reads. The gene-level median log2-fold change vs sgRNA abundance is plotted in a Bland–Altman MA plot. The sgRNA count was calculated using a custom script in Python (version 2.7.3) and Bland–Altman MA plot was generated using R (version 3.1.1). Additional information will be made available upon request.

Colony-formation assays

A427 and LK2 FDFT1-KO cells were plated at 1 × 104 cells per well in 6-well culture dishes in RPMI-1640 media (12–702F, Lonza) with 10% FBS (F2442, Sigma). After 24 h, the cells were treated with either 1% DMSO or 1 μM NB-598. After 10 days the cells were washed with PBS, stained with .5% crystal violet (C581, Fisher Scientific) in a 4% paraformaldehyde solution (F8775, Sigma) and then visualized using Odyssey imaging system (LI-COR).

Caspase activity assay

Cells were plated at a density of 1 × 104 cells per well in 96-well plates and treated with increasing concentrations of NB-598. DMSO 0.1% treatment was used as a control. Caspase activity was assessed at using Caspase-Glo 3/7 assay (G8090, Promega) following the manufacturer’s recommendations. Briefly, the reagent was added to the cells at a 1:1 ratio. The plates were then incubated at room temperature for 1 h, after which they were analyzed for luminescence in a luminometer. Values were normalized to 0.1% DMSO control after subtracting background signal.

BODIPY lipid droplet staining

Cell lines were plated on poly-d-lysine coverslips (1232B51, Thermo Fisher) in 24-well plates and treated with the indicated pharmacological inhibitors for 24 h. Following the treatment, cells were spun down for 5 min at 2500 rpm. The media was aspirated and the cells were carefully washed twice with PBS. BODIPY 493/503 neutral lipid dye (D3922, Thermo Fisher) was diluted in PBS to working concentration of 2 μM. Cells were then incubated in 1 mL staining solution for 15 min at 37 °C. Following staining, cells were carefully washed twice with PBS, spun down for 5 min at 2500 rpm, and then fixed in 4% paraformaldehyde (50–980–495, Fisher Scientific) for 30 min at room temperature. Following fixation, cells were carefully washed twice with PBS, and the coverslips were transferred onto glass microscope slides treated with VECTASHIELD Antifade Mounting Medium with 4,6-diamidino-2-phenylindole (H-1200, Vector Labs). Images were analyzed using the Image J software.

Transmission electron microscopy

Electron microscopy was performed in the Microscopy Core of the Center for Systems Biology/Program in Membrane Biology (Massachusetts General Hospital, Boston).

Human cell lines grown in suspension were fixed in 2.0% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.4 (Electron Microscopy Sciences), rinsed in 0.1 M sodium cacodylate buffer, scraped, and pelleted. Pellets were post-fixed in 1.0% osmium tetroxide in cacodylate buffer for 1 h, rinsed several times in cacodylate buffer, and stabilized in 2.0% agarose in PBS. The agarose blocks containing the pellets were dehydrated through a graded series of ethanol solutions to 100%, followed by a brief dehydration step in 100% propylene oxide. Specimens were then allowed to infiltrate in a 1:1 solution of Eponate resin (Ted Pella) and propylene oxide overnight on a gentle rocker at room temperature. The following day, specimens were transferred into fresh 100% Eponate resin and allowed to infiltrate for several hours, then embedded in flat molds in 100% Eponate, and resin allowed to polymerize 24–48 h at 60 °C.

Ultrathin (70 nm) sections were obtained using a diamond knife (ultra-grade, Diatome) and a Leica EM UC7 ultramicrotome and collected onto formvar-coated grids (Electron Microscopy Sciences). Thin sections were stained with uranyl acetate and Reynold’s lead citrate and examined in a JEOL JEM 1011 transmission electron microscope at 80 kV. Images were collected using an AMT digital imaging system with proprietary image capture software (Advanced Microscopy Techniques).

Xenograft studies

Five-to-6-week-old female mice were obtained from either Taconic Laboratories (ICR SCID), Jackson Laboratories (NSG) or Charles River Laboratories (Nu/Nu) and maintained in ventilated caging. Experiments were carried out under an Institutional Animal Care and Use Committee–approved protocol, and institutional guidelines for the proper and humane use of animals were followed. Xenografts (H1963 cells in ICR SCID mice; H446 cells in Nu/Nu mice) were established by subcutaneously injecting 100 µL of cells at a concentration of 5e6 per mouse in the right flank of each mouse. Similarly, xenografts (LU139 and H2171 cells in NSG mice; H146 cells in ICR SCID mice) were established by subcutaneously injecting 100 µL of cells at a concentration of 10e6 cell per mouse also into the right flank of each mouse. All cell lines were prepared for implantation by resuspending cells in DMEM/F12 and Matrigel at a 1:1 ratio. Upon establishment of tumors (150–250 mm3), mice were randomized into two groups: vehicle (0.5% methylcellulose, 4% HPMC-AS) or NB-598 (300 mg kg−1). Each group received treatment once daily via oral administration at 10 mL kg−1. Tumor volume was measured twice weekly by caliper, and volume was calculated using the formula 0.5 × W2 × L with the results presented as means and standard error of measurement (s.e.m.). At the end of the study, mice were sacrificed by CO2 inhalation; mice were bled via cardiac puncture, and blood sampled into EDTA tubes. Blood samples were spun at 10,000 rpm for 5 min and plasma decanted into labeled Eppendorf tubes. Tumors were aseptically removed and flash frozen with liquid nitrogen and both plasma and tumor were stored at −80 °C until analysis.

For in vivo D2O-labeling studies, mice bearing LU139 xenografts (tumor volume between 300 and 600 mm3) were dosed with three daily treatments of vehicle or NB-598 to reach a steady-state inhibitor concentrations. At 24 h post last dose, animals received an intraperitoneal bolus injection of 19.8% D2O at 20 mL kg−1. Tumors were harvested after 6 h of D2O administration.

Immunohistochemistry

Frozen tumors were sectioned and stained with the indicated antibodies using standard methods (Tufts Discovery Pathology, Cummings School of Veterinary Medicine).

In vitro labeling studies

For in vitro D2O labeling, cells were grown in the DMEM/F12 media and supplemented with 0.5% D2O (151882, Sigma) for the indicated amounts of time.

For in vitro 13C2-acetate labeling, cells were grown in DMEM/F12 media and supplemented with 500 μM 13C2-sodium acetate (CLM-440–1, Cambridge Isotope Laboratories) for 48 h.

MS analysis of cholesterol pathway

For quantitative squalene analysis, samples were quenched with a mixture of methanol and water (80:20%, v/v) and extracted with ethyl acetate containing stable labeled internal standard. The supernatant was dried down under nitrogen and then reconstituted with 100 μl of 0.4 mg mL−1 BHT in acetonitrile (ACN). An aliquot of 20 μl was injected into the ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) system. The instrument setup consisted of an AB Sciex API-4000 QTrap Mass Spectrometer (AB Sciex, Framingham MA, USA) equipped with a Waters UPLC Acquity (Waters, Milford, MA, USA). The UPLC separation was performed on an ACQUITY UPLC BEH C18 (2.1 × 50 mm2, 1.7 µm, Waters) at 40 ℃. Formic acid in water (0.1%, v/v, mobile phase A) and a mixture of ACN and isopropanol (80:20%, v/v with 0.1% formic acid, mobile phase B) were employed as the mobile phase. A gradient elution over 4 min with flow rate set at 0.6 mL min−1 was used for chromatographic separation. Squalene and 2,3-oxidosqualene were ionized under a positive ion spray mode via atmospheric chemical ionization (APCI) and detected through the multiple-reaction monitoring of a mass transition pair at m/z 411.3 → 95.1 and 427.4 → 409.5, respectively.

For 13C isotopomer analysis of the cholesterol pathway, cells were extracted with ethyl acetate and analyzed via UPLC high-resolution MS. The instrument set up consisted of a Thermo QExactive Mass Spectrometer (Thermo, San Jose, CA, USA). The UPLC separation was performed on an ACQUITY UPLC BEH C18 (2.1 × 50 mm2, 1.7 µm, Waters) at 40 ℃. Formic acid in water (0.1%, v/v, mobile phase A) and a mixture of ACN and isopropanol (80:20%, v/v with 0.1% formic acid, mobile phase B) were employed as the mobile phase. An isocratic elution of 98% mobile phase B was used and run time was 1.5 min. The flow rate of mobile phase was set at 0.6 mL min−1. Squalene and 2,3-oxidosqualene were ionized under a positive ion spray mode via APCI. Peak areas were calculated in El Maven (https://elucidatainc.github.io/ElMaven/) and stable isotopic measurements were corrected against the naturally occurring isotopes for each metabolite measured.

For 2H isotopomer analysis of the cholesterol pathway, cells were extracted from cells or tissues with 100 µL 1 N KOH in 80% ethanol. Samples were heated for 65 °C for 1 h. Thereafter, the pH of the samples were adjusted to below pH 2 with the addition of 6 N HCl (25 µL). Lipids were extracted with 150 µL chloroform and centrifuged for 10 min (2000 × g). The chloroform layer was removed and evaporated to dryness. Acetate esters of cholesterol was prepared by adding 100 µL acetic anhydride:pyridine (2:1, v/v) to the dried samples. Samples were heated for 1 h at 75 °C. Reagents were evaporated slowly by nitrogen gas. Samples were reconstituted with 100 µL with ethyl acetate and transferred to autosampler vial for analysis.

Cholesterol ester preparations were analyzed for deuterated-palmitate with a Thermo Finnigan Delta V IRMS coupled to a Thermo Trace GC Ultra with a GC combustion interface III and Conflow IV. The acetate ester of 2H-cholesterol was analyzed using a split-less injection with CTC Pal autosampler (1 µL), at an injection temperature of 250 °C, and using a Zebron ZB-5 column of 30 m × 0.25 mm × 0.50 µm film thickness (Phenomenex, Torrance, CA). The GC oven was programmed with an initial column temperature of 100 °C with a 1-min hold, followed by a ramp of 10 °C per minute to 150 °C and a final ramp of 30 °C min−1 to 340 °C. Compounds eluting off the column were directed into the pyrolysis reactor, heated at 1450 °C, and converted to hydrogen gas. The deuterated enrichment was first initially expressed in delta values compared to a calibrated hydrogen gas and then converted to atom % D by standard equations.

All IRMS analyses involving deuterium labeling were performed at Metabolic Solutions (Nashua, NH).

SCLC proteomics

SCLC cells were grown in DMEM/F12 media (11320033, Thermo Fisher Scientific) and harvested in triplicate at 5e6 cells per pellet.

Unless otherwise stated, all chemicals were from Sigma Aldrich. All water and solvents were Optima LC/MS grade from Thermo Scientific.

Proteomics: protein extraction and digestion

Cell pellets were lysed in 8 M urea/50 mM HEPES pH8.5 (Alfa Aesar) and treated with nuclease (Thermo Scientific) for 10 min at room temperature with constant shaking in a Thermomixer (Eppendorf). Lysates were reduced with 5 mM dithiothreitol (DTT) for 30 min at 37 °C and cysteine residues alkylated with 15 mM iodoacetamide for 30 min at room temperature in the dark. Excess iodoacetamide was quenched with 10 mM DTT for 15 min at room temperature in the dark. Protein was extracted by methanol−chloroform precipitation and 2× 1 mL methanol washes. Pellets were dried and resuspended in 8 M urea/50 mM HEPES, pH 8.5. Protein concentrations were measured by BCA assay (Thermo Scientific) prior to protease digestion. Two hundred micrograms of protein were diluted to 2 M urea and digested with LysC (Wako) in a 1:100 enzyme:protein ratio overnight at 30 °C. The next morning, trypsin (Promega) was added to a final 1:100 enzyme:protein ratio for 6 h at 37 °C. Digests were acidified with 10% trifluoroacetic acid (TFA) to a pH ∼2 and subjected to solid-phase extraction (SPE) with HyperSep Retain PEP Cartridges (Thermo Scientific). Peptides were resuspended in 200 mM HEPES pH8.5/10% ACN, and concentrations were measured by microBCA assay (Thermo Scientific).

Proteomics: tandem mass tag labeling

Isobaric labeling of peptides was performed using 10-plex tandem mass tag (TMT) reagents (Thermo Scientific). TMT reagents (0.8 mg) were dissolved in 41 μL of anhydrous ACN and 10 μL was added to 50 µg of peptide. Samples were labeled for 2 h at room temperature and quenched by the addition of hydroxylamine to 0.5%, v/v, for 15 min. Samples were pooled, acidified with 10% TFA to a pH ∼2 and subjected to SPE. In all, 1.08 mg of peptides from 36 protein digests labeled with TMT-131 was spiked into each 10-plex to normalize across 10-plex experiments.

Proteomics: basic pH reverse-phase high-performance liquid chromatography (HPLC) fractionation

TMT-labeled peptides were subjected to orthogonal basic-pH reverse-phase fractionation. Peptides were solubilized in buffer A (10 mM ammonium bicarbonate, pH 8.0/5% ACN) and separated on a Biobasic C18 column (5 µm particle size, 4.6 mm ID, and 250 mm length, Thermo Scientific) using an Ultimate 3000 HPLC (Thermo Scientific) and a 44 min linear gradient from 12% to 36% ACN in 10 mM ammonium bicarbonate pH 8.0 (flow rate of 0.8 mL min−1) into a total of 96 fractions. The 96 fractions were consolidated into 24 samples in a checkerboard manner, acidified with TFA to pH ~2, and vacuum-dried. Each sample was dissolved in 20 µL 5% formic acid (FA)/5% ACN and 2 µL was analyzed by MS.

Proteomics: Orbitrap fusion parameters

Spectra were acquired on an Oribtrap Fusion (Thermo Scientific) coupled to an Easy-nLC 1200 ultrahigh pressure liquid chromatography pump (Thermo Scientific). Peptides were separated on a 50 μM C18 EASYspray column (Thermo Scientific) using a 70 min, 8−28% ACN (constant 0.1% FA) gradient with a 300 nL min−1 flow rate. MS1 spectra were collected in the Orbitrap at a resolution of 60,000, automated gain control (AGC) target of 5e5, and a max injection time of 100 ms. The 10 most intense ions were selected for MS/MS in a data-dependent manner. Precursors were filtered according to charge state (2–6 z) and monoisotopic peak assignment, and previously selected peaks were excluded using a dynamic window of 60 s with a mass error ±10 ppm. MS2 precursors were isolated with a quadrupole mass filter set to a width of 0.5 m/z and detected in the ion trap operated at rapid scan rate. MS2 spectra were collected at an AGC of 1e4, max injection time of 150 ms, and CID collision energy of 35%. Synchronous precursor selection was enabled to include the top 10 MS2 fragment ions for MS3 analysis in the Orbitrap at a resolution of 60,000, AGC target of 5e5, and a max injection time of 250 ms. 50% HCD collision energy was used to ensure TMT reporter detection.

Proteomics: data processing

All .RAW files were processed using Proteome Discoverer 2.1.0.81. MS2 spectral assignment was performed using the SEQUEST algorithm using Uniprot Human reference proteome (UP000005640, downloaded 10/05/2016) and a list of known contaminants (CRAPome.org). Mass tolerances were set to 10 ppm for precursor ions, 0.6 Da for MS2 ions, and 20 ppm for MS3 reporter ions. MS2 false discovery rate of <1% was calculated using the Percolator algorithm. Reporter ion intensities were adjusted to correct for the isotopic impurities of the different TMT reagents (manufacturer's specifications). TMT tags on peptide N-termini per lysine residues (+229.162932 Da) and carbamidomethylation of cysteine residues (+57.02146 Da) were set as static modifications and methionine oxidation (+15.99492 Da) was set as a variable modification. Signal-to-noise values for all peptides were summed within each TMT channel and each channel was scaled according to the highest channel sum so that the sum abundance of each channel is equal. Peptides were filtered for a minimum sum signal-to-noise value of 160 across all 10 channels. Quantitation data from razor peptides were excluded and only unique peptides were used for protein quantitation.

Statistical analysis

Analyses were performed using the GraphPad Prism 7 software package. All data are expressed as mean of multiple measurements (n, indicating the number of replicates) from a representative experiment. Each experiment was replicated multiple times. Error bars represent s.e.m. for all in vivo studies and standard deviation (s.d.) for all the in vitro studies. Student’s t test was used to assess statistical significance. Exact values and cutoff are specified within each figure or figure legend.

Code availability

Additional statistical analyses and visualizations of cellular growth, RNA-seq, and proteomics data were performed using code written in Python using the standard NumPy, pandas, matplotlib, and seaborn packages. Code is available upon request.

Reporting summary

Further information on experimental design is available in the Nature Research Reporting Summary linked to this Article.