Cell culture and cell cycle distribution
All chemicals were from Sigma-Aldrich unless otherwise noted. LY294004 and rapamycin were from Calbiochem. HeLa, 293T, and 293GP cells were cultured in DMEM (Cellgro) containing 10% bovine calf serum (Gemini). MCF10A cells were cultured in Ham's F12 medium (Cellgro) supplemented with 5% horse serum (Invitrogen), 2 mM glutamine (BioWhittaker), 20 ng/ml EGF (Invitrogen), 100 ng/ml cholera toxin (Calbiochem), 0.01 mg/ml bovine insulin (Invitrogen), and 500 ng/ml hydrocortisone. Leucine-free MCF10A medium and leucine were custom ordered (Specialty Media). Cell proliferation was assessed by plating 100,000 cells in six-well plates, performing siRNA transfection, and at the indicated times cells were trypsinized and counted in a Coulter counter (Beckman Coulter). Where indicated, growth rates were measured as an exponential trend of the growth curve. For cell cycle distribution analysis, treated MCF10A cells at similar densities were trypsinized, ethanol fixed, and labeled with 200 mg/ml propidium iodide in 38 mM sodium citrate in the presence of RNaseA. DNA histograms were acquired using a FACSCalibur flow cytometer (Becton Dickinson) and analyzed using ModFit LT (Ventry Software). Inhibition of the proteasome was achieved by treatment with 10 μM MGI32 for 6 h.
Membrane potential assay, autophagy induction, and MDC labeling
Mitochondrial membrane potential was measured by incubating control or eIF4GI-silenced cells for 30 min with 50 nM of tetramethyl rhodamine ethyl ester (TMRE; Invitrogen) followed by FACS analysis. Autophagy was induced by amino acid starvation. Indicated cells were washed three times with PBS and incubated in HBSS for 30 min at 37°C. Autophagic vacuoles in starved or siRNA-transfected MCF10A cells were labeled by incubation with 0.05 mM MDC (Invitrogen) in PBS at 37°C for 10 min (Biederbick et al., 1995
). Cells were washed three times with PBS and lysed in buffer containing 0.1% Triton X-100 and 10 mM Tris-HCl, pH 8. Incorporated MDC was measured by fluorescence photometry (excitation wavelength 380 nm, emission filter 525 nm) in a FluoroCount microplate reader, and normalized to DNA levels.
Confocal microscopy studies
After treatment, cells were seeded into 6-well plates with coverslips and grown for 24 h to ~70% confluency. After three PBS washes, cells were fixed in 4% paraformaldehyde (Electron Microscopy Sciences) in PBS for 10 min at 25°C. Fixed cells were washed once with PBS and permeabilized in 0.25% Triton in PBS for 10 min at 25°C, washed with PBS and blocked in 10% BSA in PBS for 1 h at 25°C. GFP-LC3 transfected or transduced cells were washed and mounted onto slides. For antibody staining for fluorescence microscopy, cells were incubated with 100 μl of primary antibody in 3% BSA in PBS (1:100 affinity-purified eIF4GI antiserum and 1:100 mouse monoclonal β-tubulin [BD Biosciences], or 1:100 rabbit anti-β-actin [AbCam] antibody) overnight in a humidified chamber at 4°C. After three PBS washes, cells were incubated with secondary antibody in 3% BSA in PBS (1:100 FITC-conjugated donkey anti–rabbit and 1:100 TRITC-conjugated donkey anti–mouse; Jackson ImmunoResearch Laboratories) at 25°C for 1 h. Coverslips were mounted on slides using Vectashield mounting medium with propidium iodide (Vector Laboratories). Confocal immunofluorescence microscopy and image acquisition were performed at room temperature using a confocal scanning laser microscope (LSM 510 META; Carl Zeiss, Inc.) using an oil immersion Plan Apochromat 1.4 NA 63x objective. The laser scan head is equipped with a META detector (multi-channel broad spectrum photomultiplier tube arrays) and two additional single photomultipliers that assemble the images pixel by pixel using Lasersharp 2000 software (Carl Zeiss, Inc.) and converted into JPEG files, pseudo-colored, and merged for image processing with Photoshop 7 software.
Metabolic labeling of cells and preparation of extracts
Cells were pulse-labeled for 30 min with 10 μCi of [35S]-methionine and cysteine mix per ml (Easytag Express Protein Labeling Mix; PerkinElmer) in DMEM without methionine and cysteine containing 2% dialyzed FBS. After treatment, cells were washed twice in ice-cold 1x PBS and lysed in ice-cold buffer A (0.5% igepal [vol/vol], 150 mM NaCl, 50 mM Hepes, pH 7.0, 2 mM Na3VO4, 25 mM β-glycerophosphate, and complete protease inhibitor mix; Roche). Debris was removed by centrifugation at 4°C for 10 min and the supernatants were stored at −80°C. Protein concentrations were determined by the Bradford method (Bio-Rad Laboratories). Specific activity was determined by trichloroacetic acid precipitation and liquid scintillation. For immunoprecipitation of 35S-labeled proteins, cells were labeled in Easytag mix as described above for 2 h, lysed in buffer A, and sonicated with four pulses of 5 s each. For determination of LC3, KLF4, and MYCBP levels, cells were lysed in RIPA buffer (1% igepal, 0.5% deoxycholate, 0.1% SDS, 150 mM NaCl, 50 mM Tris, pH 8.0, 2 mM Na3VO4, 25 mM β-glycerophosphate, and complete protease inhibitor mix; Roche) and protein concentrations were determined using the DC Protein Assay kit (Bio-Rad Laboratories).
Polysome gradients, RNA extraction, and microarray analysis
48 h after siRNA transfection, MCF10A cells were seeded in 15-cm plates and harvested at 50% confluency (three plates per condition). Cells were treated with 100 μg/ml cycloheximide (Calbiochem) in complete medium for 5 min at 37°C, collected in ice cold PBS supplemented with complete protease inhibitor without EDTA (Roche) and 100 μg/ml cycloheximide. All subsequent steps were performed at 4°C. Cells were washed twice in PBS with 100 μg/ml cycloheximide, pelleted, and resuspended in 700 μl polysome isolation buffer (200 mM Tris, pH 7.5, 100 mM NaCl, and 30 mM MgCl2). After 5 min incubation, 250 μl detergent buffer (1.2% Triton, 0.2 M sucrose in polysome isolation buffer) was added and cells were lysed with a dounce homogenizer with 10 strokes. Lysates were clarified by centrifugation at 13,000 g for 10 min at 4°C and supernatants were transferred to new tubes containing 100 μl heparin buffer (10 mg/ml heparin [Sigma-Aldrich], 1.5 M NaCl in polysome isolation buffer). Samples were layered onto 11-ml 10–50% sucrose (Sigma-Aldrich) gradients (in polysome extraction buffer supplemented with 100 μg/ml cycloheximide), sedimented at 36,000 rpm for 2 h in a SW40 rotor (Beckman Coulter) at 4°C. Gradients were collected as 15 × 750 μl fractions by pumping 60% sucrose into the bottom of the gradient and collecting from the top using an ISCO fraction collector while simultaneously monitoring absorbance at 254 nm. 30 μl of 0.5M EDTA were added to each fraction, RNA was isolated by extraction with phenol/chloroform followed by chloroform extraction, and precipitated overnight with 1 volume isopropanol and 0.1 volume NaOAc (pH 5.2) at −20°C. Fractions were centrifuged at 13,000 g for 10 min to collect the RNA, washed twice with 70% EtOH, and resuspended in 100 μl nuclease-free water. Fractions 6–12, representing polysomes, were combined and extracted twice with phenol and twice with chloroform, followed by isopropanol precipitation. 10 μg of pooled RNA were used for microarray analysis using the one-cycle cDNA synthesis GeneChip kit and hybridized to GeneChip Human Genome U133A 2.0 chips, both from Affymetrix, according to the manufacturer's instructions. For normalization, total RNA was extracted from cells using Trizol and further purified through RNeasy mini-elute columns (QIAGEN). RNA quantity and quality were determined using a bioanalyzer (Agilent Technologies), and samples were stored in nuclease-free water at −80°C. mRNA targets with splice variants that alter 5′ and 3′ UTRs were omitted from the UTR analyses.
MTT proliferation assay
Cell growth was measured by a standard MTT Cell Proliferation Assay. In brief, cells infected with shRNA-expressing lentiviruses and selected with puromycin and/or G418S were plated in quadruplicate (per time point) in 96-well format and subsequently quantified four times over a course of 72 h. Wells were treated with 20% sterile-filtered 5 mg/ml MTT (Sigma-Aldrich) in DMEM over 6 h at 37°C. The reduced formazan was then solubilized in a 1:1 DMF/20% SDS solution overnight. Colorimetric quantification was performed with a TECAN Sunrise microplate reader at an OD of 570 nm, with a negative reference at 450 nm. Relative growth rates were determined from the first-order logarithmic regression analysis of the colorimetric change over time per cell line (P < 0.01). Results were normalized to the nonsilencing shRNA control cell line.
DNA constructs, transfections, transductions, and luciferase assays
Plasmid pEGFP-LC3 was from T. Yoshimori (National Institute for Basic Biology, Okazaki, Japan). 293T cells were transfected by the CaPO4
method. pBabe-AcGFP-LC3 was generated by subcloning the LC3 fragment from EGFP-LC3 into the BglII and EcoRI sites of pAcGFP1-C1 (BD Biosciences), and further subcloning it from the SnaBI and SalI sites of AcGFP1-C1 to those same sites in pBABE-puro. The pRL-HCV-IRES-FL plasmid (Tischendorf et al., 2004
) was a gift from M. Krüger (Medizinische Hochschule Hannover, Hannover, Germany). Duplexed oligos spanning bases 17 to 105 of the 5′ UTR of mda-7/IL24 with our without uORFs (wild-type or AUG to UUG mutants) were purchased from Integrated DNA Technologies and ligated into the EcoRV site of the pRL-HCV-IRES-FL plasmid. pCMVLuc was generated by insertion of HindIII–XbaI fragment coding for luciferase from pSP-luc+ (Promega) into the same sites of pcDNA3.1 (Invitrogen). Duplexed oligos spanning the entire 5′ UTR of PPP3CB with our without uORFs (wild-type or AUG to UUG mutants) were ligated into blunt-ended HindIII-cut pCMVluc. Constructs were verified by DNA sequencing. pRhifF and pstemRhifF were provided by G. Goodall (Institute of Medical and Veterinary Science, Adelaide, Australia). The indicated concentrations of plasmids were transfected into MCF10A cells in 6-cm dishes at ~50% confluency 48 h after siRNA transfection using Fugene (Roche). Luciferase assays were performed 24 h later using either Firefly Luciferase or Dual Luciferase Reporter Assay Systems (Promega). HA-GST-p85 S6K1 pRK5 was from D.M. Sabatini (Whitehead Institute, Cambridge, MA; Addgene plasmid 8466). pSP4GI was a gift of R. Lloyd (Baylor College of Medicine, Houston, TX). The eIF4GI primers were used to amplify eIF4GIc (the nomenclature of Byrd et al. 
). The amplified product was cloned into pcDNA3.1/myc-His (Invitrogen) using the introduced XhoI and HindIII sites to generate myc-His-eIF4GI. pCAN-myc-DAP5 was a gift of F. McCormick (University of California, San Francisco, San Francisco, CA). This plasmid was used to amplify the DAP5 ORF which was cloned into pBMN-I-GFP (gift of G. Nolan, Stanford University, Stanford, CA) to generate pBMN-DAP5-I-GFP. MCF10A cells were transduced in the presence of polybrene (4 μg/ml) with viruses generated by cotransfection of either pBABE-AcGFP-LC3 or pBMN-DAP5-I-GFP and pCMV-VSVG into 293GP cells (gift of M. Pagano, NYU School of Medicine), which were harvested 72 and 96 h after transfection. 72 h after transduction, AcGPF-LC3 cells were selected with 2 μg/ml puromycin (InvivoGen). DAP5-expressing cells were enriched by sorting for GFP-positive cells 1 wk after transduction.
Measurement of ATP levels
ATP levels were measured by the luciferin/luciferase method using the ATP Bioluminescence Assay kit HS II, according to the instructions of the manufacturer (Roche). Luciferase reagent (25 μl) was injected into 100 μl of lysate from equal numbers of cells (105), and the luminescence was analyzed after a 1-s delay with a 1-s integration on an LMax Luminometer (MDS Analytical Technologies). A standard curve was generated from known concentrations of ATP and used to calculate the concentration of ATP in each sample. Luminescence increased linearly with the negative log of the ATP concentration in the samples over the range of concentrations measured.
shRNAs, antibodies, and mTOR kinase assays
For the siRNA experiments, 21 nucleotide complementary RNAs with 2 nucleotide overhangs were synthesized (QIAGEN) to the following regions: bases 3587–3607 and 5183–5203 of eIF4GI (GenBank/EMBL/DDBJ accession no. AY082886
), bases 5068–5088 of eIF4GII (AF012072
), 2856–2876 of eIF3S8 (BC001571
), 2015–2035 of eIF3S1 (NM_003758
), 3510–3530 of DAP5 (NM_001418
), 1744–1764 of Paip1 (NM_006451
), and 1542–1562 of eIF4E (NM_001968
). Control siRNAs (GFP and nonsilencing) were also from QIAGEN. Unless otherwise noted, cells were grown to 40% confluency and transfected twice 24 h apart with 5.6 μg siRNA per 10-cm plate using oligofectamine reagent (Invitrogen) according to the manufacturer's instructions. For lentivirus expression of shRNAs, target sequences were inserted into pLKO.1 (Stewart et al., 2003
). For raptor silencing we used pLKO.1 raptor_1 (D.M. Sabatini; Addgene plasmid 1857 [Sarbassov et al., 2005
]). To produce virus containing the shRNA-generating cassette, 293GP cells were transfected with 4 μg each of pCI-VSV-G, pCMVΔ8.2R', and pLK0.1 with Fugene reagent (Roche; plasmids were a gift of Dr. Weinberg, Whitehead Institute). Target cells were infected in the presence of polybrene (4 μg/ml) and selected with puromycin at 2.0 μg/ml for 72 h. For double silencing using the lentivirus system, the puromycin resistance cassette in pLK0.1 was substituted with that for neomycin. Cells infected with this vector were selected with 800 μg/ml G418S.
The following antibodies were used: rabbit anti-eIF4GI C terminus, 1:500 dilution (Schneider laboratory); rabbit anti-eIF4GI N terminus, 1:1,000 (gift of R. Lloyd, Baylor College of Medicine); rabbit anti-eIF4GII and rabbit anti-Paip1, both at 1:1,000 (gifts of N. Sonenberg, McGill University, Montreal, Canada); monoclonal anti-eIF4A, 1:1,000 (gift of W. Merrick, Case Western Reserve University, Cleveland, OH); rabbit anti-LC3, 1:1,000 (gift of T. Yoshimori, National Institute for Basic Biology, Okazaki, Japan); mouse anti-eIF4E, 1:1,000, mouse anti-DAP5 1:500 (BD Transduction Laboratories); rabbit anti-DAP5, 1:1,000 (gift of M. Holcik, Children's Hospital of Eastern Ontario Research Institute); rabbit anti-PABP, 1:2,000, rabbit anti-eIF3S8, 1:15,000, rabbit anti-KLF4 1:500 (AbCam); rabbit anti-eIF3S1, 1:800 (Proteintech Group); rabbit anti-eIF4G, rabbit anti-phospho eIF4G (S1108), rabbit anti-mTOR, rabbit anti-raptor, rabbit anti-phospho S6K1 (T389), rabbit anti-S6K1, rabbit anti-S6, rabbit anti-phospho S6 (S240/244), rabbit anti-4E-BP1, rabbit anti-phospho-AMPKα (T172), rabbit anti-AMPKα, rabbit monoclonal anti-caspase-3, all at 1:1,000 (Cell Signaling Technology); goat anti-Hsp27, 1:1,000 (Santa Cruz Biotechnology, Inc.); rabbit anti-MYCBP, 1:2,000 (AVIVA Systems Biology); mouse anti-β-tubulin, 1:2,000 (BD Biosciences); mouse anti-Skp2, 1:500 (Zymed Laboratories); mouse anti-HA, 1:1,000 (Covance); mouse anti-myc, 1:1,000 (Roche). HRP-labeled goat anti–rabbit and anti–mouse secondary antibodies were used at 1:10,000 dilution (GE Healthcare). Samples were separated by SDS-PAGE and transferred to PVDF membranes (Millipore). The enhanced chemiluminescence (ECL; GE Healthcare) procedure was used to detect protein signals. For cap-affinity chromatography, 500 μg of lysate were combined with 30 μl of m7GTP-Sepharose (GE Healthcare). Affinity purifications were performed for 3 h at 4°C with rotation.
For immunoprecipitation, mouse anti-Skp2 antibody (Zymed Laboratories) and goat anti-Dap5 (Santa Cruz Biotechnology, Inc.) were used at 1:50; mouse anti-HA at 1:150 (Covance); rabbit anti-myc at 1:250 (AbCam); and rabbit anti mTOR at 1:100 (Cell Signaling Technology). 30 uL of Protein A/G-agarose slurry were used for each immunoprecipitation (Santa Cruz Biotechnology, Inc.). Kinase assays were performed essentially as described, with modifications (Kim et al., 2002
). 293T cells were transfected with HA GST S6K1 or myc-His-eIF4GI plasmids and 24 h later incubated with 10 μM LY294002 and 100 nM rapamycin for 12 h. For mTOR immunoprecipitation, MCF10A cells were lysed in buffer B (40 mM Hepes, pH 7.4, 120 mM NaCl, 1 mM EDTA, and 0.3% CHAPS) supplemented with complete protease inhibitor tablet (Roche) and 1x Halt Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific). Lysates were clarified by centrifugation at 0.8 g
for 10 min. After preclearing, 500 μg of MCF10A lysate were immunoprecipitated with mTOR antibody and protein A/G-agarose slurry. Treated 293T cells expressing either S6K1 or eIF4GIc were lysed in Buffer A, and immunoprecipitated with either HA or myc antibodies and protein A/G-agarose slurry. Captured immunoprecipitates were washed in their respective buffers three times, then washed two more times in buffer B without EDTA. mTOR immunoprecipitates were combined with either S6K1 or eIF4GI immunocomplexes and incubated with 40 μl of kinase buffer (25 mM Hepes-KOH, pH 7.4, 50 mM KCl, 20% Glycerol, 10 mM MgCl2
, 4 mM MnCl2
, 1 mM DTT, and 500 μM ATP) for 15 min at 37°C. The reaction was stopped by addition of 10 μl 6x SDS-loading buffer and proteins were separated by SDS-PAGE followed by immunoblot. Quantification of bands was performed by densitometry using QuantityOne software (Bio-Rad Laboratories).
RNA was extracted using Trizol (Invitrogen), and in the case of luciferase reactions, DNase treated using Turbo DNA-free kit (Ambion). Single step qRT-PCR reactions were performed using SYBR Green Quantitative RT-PCR kit (Sigma-Aldrich) in a Lightcycler (Roche) according to manufacturer's instructions. Primer pair sequences used are available upon request. mRNA changes were quantified using the ΔΔCt method using β-actin mRNA as control (Winer et al., 1999
). For qRT-PCR analysis of light and heavy polysomes, fractions were pooled into unbound mRNA and subunits (1–4), monosomes and light (5–8), and heavy (9–12) polysomal pools. Virtually no RNA was amplifiable in the first pool, so comparisons were made between light and heavy polysomes only, normalized to β-actin and setting the sum of mRNA in the two pools to 100%.
Ribosome half-transit time measurements
Measurement of ribosome half-transit time was performed as described, with modifications (Ruvinsky et al., 2005
; Sivan et al., 2007
). Equal number of control or eIF4GI-silenced cells were resuspended in 2 ml labeling medium supplemented with 5% dialyzed FBS and 10 μCi of [35
S]-methionine and cysteine mix per ml. At the indicated times, 0.25 ml aliquots were mixed with 1.75 ml ice-cold PBS with 100 μg/ml cycloheximide. Cells were lysed in ribosome buffer (20 mM Hepes [KOH], pH 7.2, 10 mM NaCl, 3 mM MgCl2
, 0.5% Igepal, 0.1 mM DTT, 100 μg/ml cycloheximide, and protease inhibitor cocktail) for 20 min, nuclei and mitochondria were pelleted by centrifugation at 4°C, maximum speed for 10 min. 50% of this post-mitochondrial supernatant (PMS) was depleted of ribosomes through ultracentrifugation at 100,000 g
for 20 min at 4°C to generate post-ribosomal supernatant (PRS). Incorporated radioactivity was measured in equal aliquots of PMS and PRS samples, and the ribosome half-transit time was determined as the displacement in time of the PMS (total protein) and PRS (released protein) linear regression trends.
DNA microarray data
The DNA array analysis was performed at the NYU Cancer Institute Genomics Facility. Samples were processed in triplicate for each experimental condition. The preparation of cRNA probes and hybridization to GeneChip HGU133A 2.0 arrays followed the recommendations of the manufacturer (Affymetrix). Raw data were normalized by model-based expression index algorithm using dChip (Li and Wong, 2001
) filtered on presence/absence calls using GeneSpring software version 7.2 (Silicon Genetics/Agilent). Differentially abundant mRNAs were identified in a cross-validation approach using TM4 analysis software (Saeed et al., 2003
) by combining t
test with P value cut-off of 0.05 and/or SAM (Tusher et al., 2001
) with false discovery rate (FDR) set to 5%. The complete set of microarray data was deposited for public access to the NCBI Gene Expression Omnibus database under the accession number GSE11011
Online supplemental material
Fig. S1 shows that eIF4GI silencing was sustained and in most cells. Fig. S2 demonstrates that eIF4GI silencing does not strongly decrease protein synthesis in primary and highly transformed cells, despite disassembly of eIF4F complexes, and can be rapidly recovered after hypoxia or hypertonic stresses that inhibit initiation and elongation. Fig. S3 demonstrates that ribosome transit time on mRNAs is not significantly altered by eIF4GI silencing. Fig. S4 shows effects of silencing different eIF4G family members on protein synthesis and cellular proliferation rates. Fig. S5 provides evidence that eIF4GI silencing particularly impairs translation of mRNAs containing uORFs. Online supplemental material is available at http://www.jcb.org/cgi/content/full/jcb.200710215/DC1