|Home | About | Journals | Submit | Contact Us | Français|
The mammalian target of rapamycin (mTOR) complex 1 (mTORC1) functions as a rapamycin-sensitive environmental sensor that promotes cellular biosynthetic processes in response to growth factors and nutrients. While diverse physiological stimuli modulate mTORC1 signaling, the direct biochemical mechanisms underlying mTORC1 regulation remain poorly defined. Indeed, while three mTOR phosphorylation sites have been reported, a functional role for site-specific mTOR phosphorylation has not been demonstrated. Here we identify a new site of mTOR phosphorylation (S1261) by tandem mass spectrometry and demonstrate that insulin-phosphatidylinositol 3-kinase signaling promotes mTOR S1261 phosphorylation in both mTORC1 and mTORC2. Here we focus on mTORC1 and show that TSC/Rheb signaling promotes mTOR S1261 phosphorylation in an amino acid-dependent, rapamycin-insensitive, and autophosphorylation-independent manner. Our data reveal a functional role for mTOR S1261 phosphorylation in mTORC1 action, as S1261 phosphorylation promotes mTORC1-mediated substrate phosphorylation (e.g., p70 ribosomal protein S6 kinase 1 [S6K1] and eukaryotic initiation factor 4E binding protein 1) and cell growth to increased cell size. Moreover, Rheb-driven mTOR S2481 autophosphorylation and S6K1 phosphorylation require S1261 phosphorylation. These data provide the first evidence that site-specific mTOR phosphorylation regulates mTORC1 function and suggest a model whereby insulin-stimulated mTOR S1261 phosphorylation promotes mTORC1 autokinase activity, substrate phosphorylation, and cell growth.
The mammalian target of rapamycin (mTOR), an evolutionarily conserved serine/threonine protein kinase, senses and integrates signals from diverse environmental cues (14, 31, 50, 74). mTOR associates with different partner proteins to form functionally distinct signaling complexes (4). The immunosuppressive drug rapamycin acutely inhibits signaling by mTOR complex 1 (mTORC1) (22), which contains mTOR, mLST8/GβL, raptor, and PRAS40 (24, 33, 34, 54, 67). Rapamycin fails to acutely inhibit signaling by mTORC2, which contains mTOR, mLST8/GβL, rictor, mSin1, and PRR5/Protor (18, 32, 47, 55, 73, 76). mTORC1 promotes various biosynthetic processes, including protein synthesis, cell growth (an increase in cell mass and size), and cell proliferation (an increase in cell number) (14, 40, 74). During growth factor (e.g., insulin) and nutrient (e.g., amino acids and glucose) sufficiency, mTORC1 phosphorylates the translational regulators p70 ribosomal protein S6 kinase 1 (S6K1) and eukaryotic initiation factor 4E binding protein 1 (4EBP1) to coordinately upregulate protein biosynthesis (40). Both S6K1 and 4EBP1 contain a TOR signaling motif, which mediates their interaction with raptor and thus facilitates their recruitment to the mTOR kinase (10, 44, 57, 58). In addition to regulating protein synthesis, mTORC1-mediated phosphorylation of S6K1 and 4EBP also promotes cell growth and cell cycle progression (15, 16). While more recently identified and thus less well characterized than mTORC1, mTORC2 mediates the phosphorylation of AGC kinase family members (e.g., Akt [also known as protein kinase B, PKB], PKCα, and SGK1) on their hydrophobic motifs and modulates the organization of the actin cytoskeleton (20, 26, 32, 55, 56).
The insulin pathway represents the best-characterized activator of mTORC1 signaling to date, and thus many signaling intermediates that link insulin receptor activation to mTORC1 have been identified (12, 31). Complementary work using Drosophila melanogaster genetics and mammalian cell culture identified TSC1 (hamartin) and TSC2 (tuberin) as upstream negative regulators of mTORC1 (27). Inactivation of either the TSC1 or TSC2 genes, whose protein products heterodimerize to form a tumor suppressor complex, causes the development of benign tumors in diverse organs in both humans and rodents, a disease known as tuberous sclerosis complex (TSC) (36). TSC2 contains a GTPase-activating protein domain that acts on Rheb, a Ras-like GTP binding protein that activates mTORC1 (27). Thus, in TSC-deficient cells, constitutive Rheb-GTP leads to chronically high mTORC1 signaling. While the mechanism by which Rheb-GTP activates mTORC1 remains incompletely understood, Rheb coimmunoprecipitates with mTOR and directly activates mTORC1 kinase activity in vivo and in vitro when GTP bound (2, 38, 54). Rheb has been reported to augment the activity of PLD1, an enzyme that catalyzes the production of the lipid second messenger phosphatidic acid, which contributes to the mitogenic activation of mTORC1 signaling (13, 62). Additionally, Rheb-GTP was reported to induce the dissociation of the endogenous mTOR inhibitor FKBP38 (3), although aspects of this model have been questioned (72). Insulin/phosphatidylinositol 3-kinase (PI3K) signaling reduces the inhibitory effect of TSC on mTORC1 via Akt-mediated phosphorylation of TSC2 (29, 42, 64). Additionally, Ras-regulated signaling via mitogen-activated protein kinase (MAPK) and RSK also inhibits TSC via PI3K/Akt-independent phosphorylation of TSC2 (39, 51, 63). In contrast, glucose deprivation enhances TSC's inhibitory effect on mTORC1 signaling via AMP-activated protein kinase (AMPK)-mediated phosphorylation of TSC2 (on different sites) (30). Thus, TSC functions as a central nexus of diverse physiological signals to fine-tune mTORC1 signaling depending on environmental conditions (27). While the mechanism by which amino acids promote mTORC1 signaling has remained elusive, compelling new data reveal that the Rag GTPases link amino acid sensing to mTORC1 activation (35, 52, 53). During amino acid sufficiency, GTP-bound Rag heterodimers bind raptor and recruit mTORC1 to an endomembrane compartment that contains the mTORC1 activator Rheb; thus, amino acid sufficiency may function to prime mTORC1 for subsequent growth factor-mediated activation via a dynamic subcellular redistribution mechanism (52).
Despite the well-characterized regulation of mTORC1 signaling by growth factors (e.g., insulin), nutrients (e.g., amino acids and glucose), and cellular stress (e.g., hypoxia) and the identification of numerous signaling mediators of these pathways, the direct molecular mechanisms by which cellular signals modulate mTORC1 action remain obscure (31). While three phosphorylation sites (P-sites) on mTOR have been reported to date (T2446, S2448, and S2481), no function has yet been ascribed to any site (7, 43, 49, 59). Here we identify S1261 as a novel mTOR phosphorylation site in vivo in cultured mammalian cells and provide the first evidence that site-specific mTOR phosphorylation regulates mTORC1 function. We show that insulin signals via the PI3K/TSC/Rheb pathway in an amino acid-dependent and rapamycin-insensitive manner to promote mTOR S1261 phosphorylation, which regulates mTORC1 autokinase activity, biochemical signaling to downstream substrates, and cell growth to increased cell size, a major cellular function of mTORC1. Elucidation of the molecular mechanisms underlying mTORC1 regulation will enable us to better understand how mTORC1 senses environmental stimuli to control cellular physiology. As aberrantly upregulated mTORC1 signaling likely contributes to cancer, insulin-resistant diabetes, and cardiovascular diseases, understanding mTORC1 regulation may aid in the development of novel therapeutics for these prevalent human diseases (11, 21, 28).
Reagents were obtained from the following sources: protein A-Sepharose CL4B and protein G-Sepharose Fast Flow were from GE Healthcare; 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS) was from Pierce; Immobilon-P polyvinylidene difluoride (PVDF) membrane (0.45 μm) was from Millipore; autoradiography film (HyBlot CL) was from Denville Scientific; reagents for enhanced chemiluminescence (ECL) were from Millipore (Immobilon Western chemiluminescent horseradish peroxidase [HRP] substrate); all chemicals were from either Fisher Chemicals or Sigma.
AU1, Myc (9E10), and HA (HA.11) antibodies were from Covance. Flag-M2 antibody was from Sigma, and Rheb antibody was from AbCam. Donkey anti-rabbit antibody-HRP and sheep anti-mouse antibody-HRP were from GE Healthcare. The following antibodies were from Cell Signaling Technology: P-S6K1 (Thr389; rabbit monoclonal 108D2), S6, mTOR P-S2481, P-Akt (S473), total Akt, P-4EBP1 (T37/46), P-4EBP1 (S65), P-4EBP1 (T70), and total 4EBP1. Tubulin monoclonal antibody was a gift from K. Verhey (University of Michigan Medical School, Ann Arbor).
To generate immunogen, peptides, and P-peptides (70% pure; synthesized by Boston Biomolecules, Woburn, MA [now called Advanced Peptides, Inc., Boston, MA]) were coupled via an N-terminal cysteine to maleimide-activated mariculture keyhole limpet hemocyanin (mcKLH; Pierce). Antipeptide antibodies were affinity purified by positive selection on antigen peptide that was coupled to Affigel matrix (Bio-Rad). Phospho-peptide antibodies were affinity purified by positive selection on antigen phospho-peptide followed by negative selection on cognate antigen non-phospho-peptide and irrelevant phospho-peptide (either a P-Ser Jak2 or a P-Thr-Jak2 peptide). The following peptides were used to generate antibodies (“p” indicates a phosphorylated residue): (i) raptor (amino acids 885 to 901; human; CSSSLTNDVAKQPVSRDL); (ii) mTOR (amino acids 221 to 237; rat; CTQREPKEMQKPQWYRHT); (iii) P-S1261-mTOR [amino acids 1256 to 1266; rat; CKKLHV(pS)TINLQ]; (iv) rictor (amino acids 6 to 20; human; CRGRSLKNLRVRGRND); (v) S6K1 (C-terminal 17 amino acids 485 to 502 of the 70-kDa rat isoform; CKQAFPMISKRPEHLRMNL); (vi) P-S6 [amino acids 232 to 249; CRRL(pS)(pS)LRA(pS)TSK(pS)EE(pS)QK].
The pcDNA3/AU1-mTOR wild-type (WT), rapamycin-resistant (RR; S2035I), kinase-dead (KD; S2338A), and RR/KD (S2035I; D2338A) plasmids were generated by R. Abraham (Wyeth, Pearl River, NY); the pRK5/Myc-raptor plasmid was provided by D. Sabatini (MIT, Boston, MA); pRK5/Myc-mTOR and pRK5/Myc-mTOR-KD were obtained from D. Sabatini via Addgene (nos. 1861 and 8482, respectively); the pRK7/HA-S6K1, pKH3/HA-mLST8/GβL, and pRK7/Flag-Rheb plasmids were from J. Blenis (Harvard Medical School, Boston, MA), and pACTAG2/3HA-4EBP1 was from N. Sonenberg (McGill, Montreal, Canada).
HEK293 cells (20 15-cm plates per condition; ~15 × 106 cells/15-cm plate) were untransfected or transiently transfected with AU1-tagged mTOR (25 μg) and cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) or in growth factor and amino acid-stimulated medium (DMEM plus insulin; 30 min) after factor deprivation (via serum withdrawal for 24 h followed by incubation in Dulbecco's phosphate-buffered saline [D-PBS]-glucose for 60 min). AU1-mTOR was immunoprecipitated overnight, resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and stained with Coomassie blue R-250. AU1-mTOR protein was excised from the gel, digested with trypsin, and analyzed by tandem mass spectrometry.
Cubed Coomassie-stained gel bands of immunoprecipitated AU1-tagged rat mTOR were rinsed with water, destained with 50% acetonitrile (MeCN), 50% ammonium bicarbonate, dehydrated with 100% MeCN, and subjected to in-gel digestion with 6 ng/μl sequencing-grade modified trypsin (Promega) in 50 mM ammonium bicarbonate for 16 h at 37°C. Peptides were extracted once with 50% MeCN, 2.5% formic acid (FA), and once with 100% MeCN. Dried peptides were resuspended in 2.5% MeCN, 2.5% FA, and loaded using a Micro AS autosampler (Thermo Electron) and a Surveyor MS Pump Plus (Thermo Electron) onto a nano-electrospray microcapillary column packed with 14 cm of reverse-phase MagicC18 material (5 μm; 200Å; Michrom Bioresources, Inc.). Elution was performed with a 5 to 35% MeCN (0.15% FA) gradient over 45 min, after a 15-min isocratic loading at 2.5% MeCN, 0.15% FA. Solvent A was 2.5% MeCN, 0.15% FA, and solvent B was 99.85% MeCN, 0.15% FA. Mass spectra were acquired in an LTQ linear ion trap mass spectrometer (Thermo Electron). Throughout the entire run, 10 data-dependent tandem mass spectrometry (MS/MS) scans were acquired on the most abundant ions in dynamic exclusion mode following a precursor survey (MS1) scan. Mass spectral data were searched against a rat mTOR protein database using Turbo SEQUEST (Thermo Electron, version 27, revision 12) requiring no enzyme specificity and a 2-Da precursor mass tolerance. Cysteine residues were required to have a static increase of 71.0 Da for acrylamide (C3H5ON) adduction. Differential modification of 16.0 Da on methionine and 80.0 Da on serine, threonine, and tyrosine was permitted.
Site-directed mutagenesis was performed using QuikChange II XL (Stratagene), and cDNA inserts in mutated plasmids were fully sequenced. We generated the following mutations in WT and RR pcDNA3/AU1-mTOR backbones: S1261A, S1261D, and S2481A. We generated the following mutations in pRK5/Myc-mTOR: RR (S2035I), S1261A, RR/S1261A, RR/S1261D, and S2481A. The following oligonucleotides were used to create point mutations in the rat mTOR cDNA (accession number L37085) (capital letters indicate mismatches, and the three underlined nucleotides represent the codon mutated): S1261A, primer 1, 5′-gaagaaactgcatgtcGCcaccatcaacctcc-3′, and primer 2, 5′-ggaggttgatggtgGCgacatgcagtttcttc-3′; S1261D, primer 1, 5′-gaagaaactgcatgtcGAcaccatcaacctcc-3′, and primer 2, 5′-ggaggttgatggtgTCgacatgcagtttcttc-3′; S2481A, primer 1, 5′-tgccagaatccatccatgcCttcattggagatggtttgg-3′, and primer 2, 5′-ccaaaccatctccaatgaagGcatggatggattctggca-3′; S2035I, primer 1, 5′-ggcctagaagaggccATtcgcttgtactttggg-3′, and primer 2, 5′cccaaagtacaagcgaATggcctcttctaggcc-3′.
shRNA-encoding lentiviruses were generated via the protocol posted on the Addgene website (http://www.addgene.org/pgvec1). Briefly, 7 × 105 HEK293T cells on 60-mm plates were cotransfected with pLKO.1 shRNA plasmid (1 μg), pCMV-dR8.2 dvpr (Addgene number 8455) packaging plasmid (750 ng), and pCMV-VSVG (Addgene number 8454) envelope plasmid (250 ng) using FuGENE transfection reagent. The next morning, the cells were refed with 5 ml fresh medium (DMEM-10% FBS). After 24 h, the virus-containing medium was harvested, and another 5 ml of fresh medium was added. After 24 h, the viral-containing medium was harvested a second time, pooled with the first collection, and filtered (0.45 μm). Viral infection was performed by adding 500 μl of virus-containing medium to a ~70% confluent plate of TSC1−/− mouse embryonic fibroblasts (MEFs) that were fed with fresh medium containing 8 μg/ml Polybrene. After 24 h, cells were split 1:10 (due to the high proliferation rate) and selected for ~6 days in fresh medium containing 2.5 μg/ml puromycin. The following plasmids were used: pLKO.1-scramble shRNA (Addgene number 1864), pLKO.1 Rheb-1 (Open Biosystems number TRCN0000075603), and pLKO.1 Rheb-3 (Open Biosystems number TRCN0000075605) (the Rheb shRNA plasmids were kindly shared by John Blenis, Harvard Medical School, Boston, MA). The hairpin sequence for the mouse-specific Rheb-1 shRNA was the following: 5′-CCGGCCCGTCATCCTTGAAGATAAACTCGAGTTTATCTTCAAGGATGACGGGTTTTTG-3′. The hairpin sequence for the mouse-specific Rheb-3 shRNA was the following: 5′-CCGGCAGACATACTCCATAGATATTCTCGAGAATATCTATGGAGTATGTCTGTTTTTG-3′.
HEK293 cells and immortalized TSC1+/+ and TSC1−/− mouse embryonic fibroblasts (originally from D. Kwiatkowski, Brigham and Women's Hospital, Boston, MA) were cultured in DMEM that contained high glucose (4.5 g/liter), glutamine (584 mg/liter), and sodium pyruvate (110 mg/liter; Gibco/Invitrogen) supplemented with 10% FBS (HyClone). 3T3-L1 fibroblasts were differentiated into adipocytes by a standard protocol (41, 61). All cells were incubated at 37°C in a humidified atmosphere containing 5% CO2. All cells were serum deprived via incubation in DMEM supplemented with 20 mM HEPES (pH 7.2) for ~20 h. Insulin (100 nM; Invitrogen) was added to serum-deprived cells for 30 min. For drug treatment, serum-deprived cells were pretreated with rapamycin (20 ng/ml; Calbiochem) or wortmannin (100 nM; Upstate/Millipore) for 30 min prior to the addition of insulin; cells proliferating in DMEM-FBS (steady state) were pretreated with drugs for 1 to 2 h. To effect amino acid deprivation, HEK293 cells were incubated in Dulbecco's PBS containing d-glucose (1 g/liter) and sodium pyruvate (36 mg/liter; D-PBS/Glc) for 60 min. HEK293 cells on 60-mm plates were transfected using TransIT-LT1 (Mirus) and a total of 4 to 5 μg of DNA per plate; the specific amounts of experimental plasmid transfected are stated in the figure legends. Cells were lysed ~24 to 48 h posttransfection.
Cells were washed twice with ice-cold PBS and scraped into ice-cold lysis buffer A (10 mM KPO4, pH 7.2, 1 mM EDTA, 5 mM EGTA, 10 mM MgCl2, 50 mM β-glycerophosphate, 1 mM sodium orthovanadate [Na3VO4], 5 μg/ml pepstatin A, 10 μg/ml leupeptin, 40 μg/ml phenylmethylsulfonyl fluoride [PMSF]) containing CHAPS (0.3%) (to preserve the mTOR-Raptor interaction postlysis) (24, 33) or NP-40 (0.5%) and Brij35 (0.1%). Lysates were spun at 13,200 rpm for 5 min at 4°C, and the postnuclear supernatants were collected. The Bradford assay was used to normalize protein levels for immunoprecipitation and immunoblot analyses. For immunoprecipitation, whole-cell lysates (WCL) were incubated with antibodies for ~2 h (overnight for AU1 immunoprecipitations) at 4°C, incubated with protein A- or G-Sepharose beads for 1.5 h, washed three times in lysis buffer, and resuspended in 1× sample buffer (50 mM Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, 2% β-mercaptoethanol, 0.02% bromophenol blue). Samples were resolved on SDS-PAGE gels and transferred to PVDF membranes by using Towbin transfer buffer (25 mM Tris, 192 mM glycine, 10% methanol, 0.02% SDS). Immunoblotting was performed by blocking PVDF membranes in Tris-buffered saline with Tween 20 (TBST; 40 mM Tris-HCl, pH 7.5, 0.9% NaCl, 0.1% Tween 20) containing 3% nonfat milk and incubating the membranes in TBST with 2% bovine serum albumin containing primary antibodies or secondary HRP-conjugated antibodies. Blots were developed by ECL.
Following mTOR immunoprecipitation, beads were washed three times in CHAPS lysis buffer followed by two additional washes in ST buffer (50 mM Tris-HCl, pH 7.2; 150 mM NaCl). Beads were then resuspended in 1× phosphatase buffer (50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 2 mM dithiothreitol, 0.1 mM EGTA, 0.01% Brij 35) that contained 2 mM MnCl2. Samples were incubated at 30°C for 30 min in the absence or presence of λ-phosphatase (Invitrogen), and reactions were terminated by adding EDTA, pH 8.0, to a 50 mM final concentration and sample buffer to a 1× final concentration. Six percent SDS-PAGE gels were run at 8 mA for ~16 h.
Lysates from HA-S6K1 transfected cells were immunoprecipitated with anti-hemagglutinin (HA) antibodies for 2 h and incubated with protein G-Sepharose beads for 1.5 h. Immune complexes were each washed in 1 ml of ice-cold buffer A (10 mM Tris-HCl, pH 7.2, 100 mM NaCl, 1 mM EDTA, 1.0% NP-40, 0.5% sodium deoxycholate, 1 mM Na3VO4, 40 μg/ml PMSF), buffer B (10 mM Tris-HCl, pH 7.2, 1 M NaCl, 1 mM EDTA, 0.1% NP-40, 1 mM Na3VO4, 40 μg/ml PMSF), ST (see above), and 1× kinase buffer (20 mM HEPES, pH 7.2, 100 mM MgCl2, 1 mg/ml bovine serum albumin). Immune complexes were then resuspended in 20 μl 1.5× kinase buffer. Kinase reactions were initiated with the addition of 10 μl of “start” solution containing cold ATP (50 μM final), [32P]ATP (10 μCi/reaction mixture), cyclic AMP-dependent protein kinase inhibitor (0.03 μg/reaction mixture; Sigma), and recombinant glutathione S-transferase (GST)-S6 purified from bacteria (2 μg/reaction mixture) and incubated at 30°C for 10 min. Kinase reactions were terminated by the addition of 30 μl of 2× sample buffer. The amount of 32P incorporated into GST-S6 was measured using a PhosphorImager (Typhoon 9400; Amersham Biosciences) and quantitated using the software ImageQuant TL.
HEK293 cells on 60-mm plates (80% confluent) were cotransfected with 1 μg green fluorescent protein (GFP)-spectrin and 10 μg Myc-mTOR plasmids. The next day, one-quarter of a 60-mm plate was transferred to a 10-cm plate and allowed to proliferate in DMEM-FBS in the absence or presence of rapamycin (20 ng/ml) for 72 h. For immunoblot analysis, one-half of the 60-mm culture was plated to a 10-cm plate and cultured for ~24 h. For cell size analysis, plates of subconfluent cells were washed once with PBS and incubated in 3 ml PBS-0.1% EDTA for ~5 min at room temperature. Cells were then gently triturated off the plastic and transferred to conical tubes on ice, centrifuged for 5 min at 1,000 rpm, and washed with 3 ml of PBS-1% calf serum. Cells were again centrifuged at 1,000 rpm for 5 min, resuspended in 0.5 ml PBS, fixed in 80% ethanol by adding 5 ml 88% ethanol, and stored overnight at 4°C. Immediately before fluorescence-activated cell sorting analysis, fixed cells were centrifuged at 1,600 rpm for 5 min, washed in 3 ml PBS-1% calf serum, and centrifuged at 1,000 rpm for 5 min. DNA was stained by resuspending fixed cells in 1 to 1.5 ml PI-RNase A solution and incubating them at 37°C for 30 min. The PI-RNase A solution contained the following: 10 μg/ml propidium iodide in 0.76 mM sodium citrate (pH 7.0), 250 μg/ml RNase A in 10 mM Tris-HCl (pH 7.5), 15 mM NaCl diluted in PBS-1% calf serum. For cell size analysis by flow cytometry, the mean forward scatter height (FSC-H) of 5,000 GFP-positive, G1-phase cells was determined by gating on GFP and PI fluorescence using a BD Biosciences FACSCalibur with CellQuest software.
For some figures, irrelevant lanes were removed from a scanned autoradiograph and flanking lanes juxtaposed using Adobe Photoshop. The presence of a thin, vertical black line indicates such a modification.
To quantitate ECL Western blot band densities, the software NIH Image J (version 1.41) was used after scanning/digitizing the autorads; a Student's unpaired t test (two-sample, assuming equal variances; two tails) was used (a P value of <0.05 was considered significant). For the cell size experiments, analysis of variance (ANOVA) followed by Tukey's post hoc tests was used to determine the significance of the population means (a P value of <0.01 was considered significant). The software Prism 5 from GraphPad was used for the ANOVA.
While growth factors, mitogens, and nutrients upregulate mTORC1 signaling, the biochemical mechanisms by which these stimuli directly activate mTORC1 remain unclear. As reversible protein phosphorylation controls the function of many signaling intermediates in the mTORC1 pathway, we examined the phosphorylation of mTOR itself. Indeed, λ-phosphatase treatment of mTOR increased its mobility on SDS-PAGE (Fig. (Fig.1A),1A), suggesting the extensive phosphorylation of mTOR in vivo. We used liquid chromatography-MS/MS to identify novel in vivo phosphorylation sites (P-sites) on AU1-tagged mTOR immunoprecipitated from transiently transfected HEK293 cells. Serine 1261 was identified as an unambiguous site of mTOR phosphorylation in cells growing in serum-containing medium as well as in factor-deprived cells that were stimulated with both insulin and amino acids (Fig. (Fig.1B).1B). While this work was in preparation, Daub et al. also identified S1261 as a site of mTOR phosphorylation in intact cells in a MS/MS-based phospho-proteomic screen, further confirming S1261 as a novel mTOR phosphorylation site (11a). Notably, our analysis identified two of the three previously identified P-sites that localize to the C terminus of mTOR-S2448 (a site phosphorylated by S6K1 via a feedback loop) and S2481 (a site of autophosphorylation) (8, 25, 43, 49) (data not shown). S1261 lies within HEAT-repeat motif 25T in the middle of mTOR (Fig. (Fig.1C)1C) (48) and is conserved among animals (human, rat, mouse, chick, zebrafish, and fly [in which it contains Thr rather than Ser]) but not in yeast or plants (Arabidopsis) (Fig. (Fig.1D).1D). To confirm mTOR S1261 phosphorylation (P-S1261) in vivo and to facilitate the study of P-S1261 regulation, we generated rabbit polyclonal, affinity-purified antibodies using an mTOR peptide phosphorylated on S1261. This antibody recognized WT AU1-mTOR but not an AU1-mTOR phosphorylation site-defective mutant bearing an Ala substitution for S1261 (S1261A) (Fig. (Fig.1E).1E). Additionally, P-S1261 antibodies failed to recognize λ-phosphatase-treated endogenous mTOR (Fig. (Fig.1F).1F). Taken together, these data demonstrate the site and phosphorylation state specificity of the P-S1261 antibody.
With this antibody in hand, we first asked whether S1261 is a site of autophosphorylation, similar to S2481. Consistent with published data (49), exogenously expressed, Myc-tagged WT-mTOR autophosphorylated on S2481 while KD-mTOR did not (Fig. (Fig.1G).1G). As both WT- and KD-mTOR exhibited similar S1261 phosphorylation, these data suggested that an autophosphorylation mechanism does not mediate mTOR P-S1261. Similar results were obtained using AU1-tagged mTOR alleles (data not shown). As mTORCs reportedly dimerize (71, 75), we next sought to exclude the possibility that P-S1261 occurs on KD-mTOR when a TORC containing endogenous WT-mTOR phosphorylates a TORC containing exogenous KD-mTOR via a trans-autophosphorylation mechanism. We thus pretreated cells with rapamycin in order to inhibit the activity of endogenous mTORC1 and found that rapamycin had no effect on P-S1261 on KD-mTOR, thus supporting the conclusion that S1261 is not a site of mTOR autophosphorylation (Fig. (Fig.1G1G).
As insulin stimulates mTORC1 and mTORC2 signaling (4, 31), we examined whether insulin signaling promotes mTOR P-S1261 in mTORC1 and mTORC2 in 3T3-L1 adipocytes, a classic insulin-responsive cell line. To examine P-S1261 in TORC1 versus TORC2, we immunoprecipitated raptor or rictor, respectively, as these partners uniquely define the complex. In mTORC1, insulin increased P-S1261 on raptor-associated mTOR, and pretreatment with the PI3K inhibitor wortmannin inhibited this insulin-induced increase (Fig. (Fig.2A,2A, left panel). In this experiment, we were unable to determine the sensitivity of P-S1261 to rapamycin, as rapamycin pretreatment resulted in the partial dissociation of mTOR from raptor, as reported previously (46), making analysis of the degree of mTOR P-S1261 in mTORC1 difficult to assess accurately. HEK293 experiments described below (Fig. (Fig.2C),2C), in which the levels of raptor-associated mTOR in the absence and presence of rapamycin were similar, suggest that mTOR P-S1261 in mTORC1 is insensitive to rapamycin. In mTOR immunoprecipitates, which contain a mixture of TORC1, TORC2, and possibly other not-yet-identified mTOR complexes, insulin increased P-S1261 in a wortmannin-sensitive but rapamycin-insensitive manner (Fig. (Fig.2B,2B, right panel). In mTORC2, insulin also increased P-S1261 on rictor-associated mTOR in a wortmannin-sensitive but rapamycin-resistant manner (Fig. (Fig.2B).2B). Collectively, these data show that in 3T3-L1 adipocytes, insulin/PI3K signaling promotes mTOR P-S1261 in both mTORC1 and mTORC2 in a rapamycin-insensitive manner.
As mTORC1 signaling has been extensively studied in HEK293 cells, the cell type in which we identified P-S1261, we also analyzed the regulation of P-S1261 in this cell type. While we have found it difficult to observe an insulin-induced increase in P-S1261 on mTOR associated with endogenous raptor (due to high basal P-S1261), coimmunoprecipitation of endogenous mTOR with exogenously expressed Myc-raptor revealed that in HEK293 cells, as in 3T3-L1 adipocytes, insulin promotes mTOR P-S1261 in a wortmannin-sensitive and thus PI3K-dependent manner (Fig. (Fig.2B,2B, left panel). In certain experiments, the rapamycin sensitivity of P-S1261 in mTORC1 could not be assessed accurately (e.g., experiment 1 in Fig. Fig.3A,3A, left panel) due to rapamycin-induced dissociation of mTOR from Myc-raptor (46). In certain experiments where rapamycin failed to dissociate mTOR from raptor (e.g., experiment 2 in Fig. Fig.3A,3A, right panel), we found that rapamycin has no effect on the phosphorylation of mTOR S1261 in mTORC1. Taken together, these data indicate that insulin promotes mTOR P-S1261 in mTORC1 and mTORC2 in a PI3K-dependent and rapamycin-insensitive manner in HEK293 cells and 3T3-L1 adipocytes. The remainder of the work presented here focuses on the regulation and function of mTOR P-S1261 in mTORC1.
As sufficient levels of amino acids are required for mTORC1 signaling, we investigated a role for amino acids in modulation of TOR P-S1261. HEK293 cells expressing exogenous Myc-raptor were cultured under steady-state conditions or incubated in D-PBS-glucose for 1 h (to effect amino acid deprivation). As expected, withdrawal of amino acids from the medium downregulated mTORC1 signaling, as shown by the dephosphorylation of the ribosomal protein S6 (Fig. (Fig.2D).2D). We found that withdrawal of amino acids from the medium resulted in reduced P-S1261 on Myc-raptor-associated mTOR (Fig. (Fig.2D).2D). These data indicate that maintenance of mTOR S1261 phosphorylation requires amino acids and are consistent with the idea that amino acid sufficiency primes mTORC1 for subsequent growth factor-mediated S1261 phosphorylation and activation.
Inactivation of the TSC tumor suppressor via mutation of either the TSC1 or TSC2 genes leads to constitutively high mTORC1 signaling (27). To determine if TSC suppresses mTOR P-S1261, we examined P-S1261 in 3T3-immortalized, TSC1 null MEFs (37). As for the phosphorylation of the mTORC1-dependent substrates S6K1 and the ribosomal protein S6, P-S1261 on raptor-associated mTOR was significantly higher in TSC1−/− cells compared to wild-type cells in the absence of serum growth factors (Fig. (Fig.3A).3A). Thus, TSC negatively regulates mTOR S1261 phosphorylation in mTORC1. We also examined a role for amino acids in the maintenance of mTOR P-S1261 in mTORC1 isolated from TSC1−/− fibroblasts. We found that withdrawal of amino acids from cycling TSC1−/− fibroblasts decreased the level of raptor-associated mTOR P-S1261 (Fig. (Fig.3B).3B). Thus, sufficient levels of amino acids are required to maintain mTOR P-S1261 in mTORC1 in both TSC null fibroblasts as well as HEK293 cells. To determine whether Rheb is required for mTOR S1261 phosphorylation, we knocked down Rheb by infecting TSC1−/− MEFs with lentiviruses encoding shRNA hairpins targeting two distinct Rheb sequences. Parental TSC1−/− MEFs as well as those infected with scrambled control, Rheb-1, and Rheb-3 lentiviruses were serum deprived, and S1261 phosphorylation status on raptor-associated mTOR was examined by coimmunoprecipitation and on total mTOR by immunoblotting of whole-cell lysate. As expected, parental TSC1−/− fibroblasts as well as those infected with the scrambled shRNA exhibited strong phosphorylation of S6K1 and S6 as well as strong raptor-associated and total mTOR P-S1261. TSC1−/− MEFs with reduced Rheb expression, however, exhibited the expected reduction in S6K1 and S6 phosphorylation, and raptor-associated and total mTOR P-S1261 were reduced as well (Fig. (Fig.3C).3C). These data demonstrate that phosphorylation of mTOR S1261 in mTORC1 requires Rheb.
As no evidence currently exists to support a role for site-specific mTOR phosphorylation in regulation of mTORC1 signaling, we asked whether mTOR S1261 phosphorylation regulates mTORC1-mediated phosphorylation S6K1 on its hydrophobic motif site, T389. We took advantage of an RR mutant of mTOR that contains a point mutation (S2035I) in the FRB (FKBP12/rapamycin binding) domain, rendering the FKBP12/rapamycin complex unable to bind to and inhibit mTOR as part of mTORC1 (6, 9, 60). Thus, the use of RR-mTOR in conjunction with rapamycin permits the study of in vivo signaling by exogenously expressed mTOR isoforms in the absence of signaling by endogenous mTORC1. We compared the ability of RR and various RR/mutant AU1-mTOR isoforms to mediate the phosphorylation of cotransfected HA-S6K1 on T389 in the presence of rapamycin. As expected, rapamycin inhibited the phosphorylation of immunoprecipitated HA-S6K1 in cells that coexpressed WT-mTOR, expression of RR-mTOR rescued the rapamycin-mediated inhibition of HA-S6K1 phosphorylation, and a RR-mTOR allele bearing a kinase dead mutation (RR/KD) failed to rescue HA-S6K1 phosphorylation. Under either steady-state conditions (Fig. (Fig.4A)4A) or in response to acute insulin stimulation after growth factor deprivation (Fig. (Fig.4B),4B), the phosphorylation of HA-S6K1 on T389 was impaired in cells expressing RR/S1261A-mTOR relative to those expressing RR-mTOR. Under both conditions, S6K1 phosphorylation was reduced ~60% in cells expressing RR/S1261A-mTOR relative to those expressing RR-mTOR. We also assayed the ability of RR versus RR/S1261A mTOR to rescue rapamycin-induced inhibition of S6K1 activity by in vitro immune complex kinase assay using recombinant GST-S6 as substrate and [32P]ATP. Consistent with the decreased ability of RR/S1261A-mTOR to phosphorylate HA-S6K1 in vivo, HA-S6K1 immunoprecipitated from cells expressing RR/S1261A-mTOR was defective (~40%) in mediating the phosphorylation of GST-S6 in vitro compared to RR-mTOR (Fig. (Fig.4C).4C). Consistently, in vivo phosphorylation of endogenous ribosomal protein S6 was also reduced in response to RR/S1261A-mTOR signaling (Fig. (Fig.4C).4C). Together, these data indicate that the phosphorylation of mTOR S1261 contributes to the overall signaling capacity of mTORC1 toward S6K1.
Overexpression of Rheb activates mTORC1 signaling in the absence of serum growth factors (2, 19, 65). To determine whether phosphorylation of mTOR S1261 is required for Rheb-mediated activation of mTORC1 signaling, we compared the ability of RR-mTOR versus RR/S1261A-AU1-mTOR to mediate the phosphorylation of cotransfected HA-S6K1 in response to Rheb in the presence of rapamycin and absence of serum growth factors. As expected, overexpression of Rheb in serum-deprived cells led to a striking increase in phosphorylation of HA-S6K1 on T389, rapamycin completely inhibited this response, and coexpression with RR-mTOR rescued the rapamycin-mediated inhibition of P-S6K1 in response to Rheb overexpression (Fig. (Fig.4D).4D). Cells expressing Rheb and RR/S1261A-AU1-mTOR exhibited impaired phosphorylation of HA-S6K1 compared to those expressing Rheb and RR-mTOR (Fig. (Fig.4D).4D). These data reveal that phosphorylation of mTOR S1261 is required for Rheb-driven mTORC1 signaling. Lastly, we investigated whether RR-mTOR bearing a phospho-mimetic substitution at 1261 (S1261D) was sufficient to promote mTORC1 signaling to S6K1 in the absence of serum growth factors. We found that expression of RR/S1261D-mTOR in the presence of rapamycin failed to increase the phosphorylation of cotransfected HA-S6K1 in the absence of serum growth factors (Fig. (Fig.4E).4E). These data suggest that the phospho-mimetic Asp substitution at S1261 does not sufficiently mimic phosphorylation to promote insulin-stimulated S6K1 phosphorylation in the absence of other growth factor signals.
Insulin promotes mTORC1-dependent phosphorylation of 4EBP1, which contributes to initiation of cap-dependent translation (40). Two phosphorylation sites on 4EBP1 exhibit sensitivity to rapamycin (Ser 65 and Thr 70), while two others (Thr 37 and Thr 46) exhibit partial sensitivity (40). We therefore assayed the ability of RR versus RR/S1261A-mTOR to mediate the phosphorylation of cotransfected 3HA-4EBP1 on T70, S65, and T37/46 in the presence of rapamycin in response to insulin (Fig. (Fig.5).5). As expected, rapamycin reduced the insulin-stimulated phosphorylation of 4EBP1 on T70, S65, and T37/46 in cells coexpressing WT-mTOR, expression of RR-mTOR rescued this rapamycin-mediated inhibition of 4EBP1 phosphorylation, and expression of RR/KD failed to rescue the phosphorylation of these sites. We observed impaired phosphorylation of 4EBP1 on all three sets of sites (e.g., T70, S65, and T36/46) in cells expressing RR/S1261A-mTOR, which was most apparent on endogenous 4EBP1 and similar to the level of phosphorylation observed in cells expressing RR/KD-mTOR. These data indicate that phosphorylation of mTOR S1261 contributes to mTORC1-mediated phosphorylation of 4EBP1 in vivo.
To explore the mechanism by which P-S1261 promotes mTORC1 signaling, we examined the phosphorylation state of the mTOR S2481 autophosphorylation site in order to examine the effect of P-S1261 on mTORC1's intrinsic in vivo kinase activity. As the original characterization of P-S2481 occurred before the existence of different mTOR complexes was known (49), we first characterized the regulation of P-S2481 in response to insulin and Rheb overexpression specifically in mTORC1. We thus cotransfected HEK293 cells with Myc-raptor and AU1-mTOR in the absence and presence of Flag-Rheb and cultured the cells in the absence or presence of insulin and wortmannin. Insulin promoted a wortmannin-sensitive increase in P-S2481 on Myc-raptor-associated mTOR, and Rheb overexpression strongly increased the P-S2481 signal (Fig. (Fig.6A).6A). These data demonstrate that insulin/PI3K/Rheb signaling promotes mTOR S2481 autophosphorylation in mTORC1 and support the idea that P-S2481 reflects mTORC1's in vivo intrinsic catalytic activity. We next investigated the role of P-S1261 in regulation of mTORC1's intrinsic catalytic activity using P-S2481 as a readout. HEK293 cells were cotransfected with Myc-raptor and various AU1-mTOR alleles in the presence of Flag-Rheb, and Myc-raptor was immmunoprecipitated to detect the phosphorylation of S2481 on associated mTOR. As expected, P-S2481 was significantly reduced relative to WT-mTOR when mTORC1 contained KD- or S2481A-mTOR (Fig. (Fig.6B).6B). Notably, we observed reduced P-S2481 when mTORC1 contained S1261A-mTOR compared to mTORC1 that contained WT-mTOR. These data reveal that phosphorylation of mTOR S1261 contributes to the activation of mTORC1's in vivo intrinsic catalytic activity. Additionally, we found that the absence of TSC1 (Fig. (Fig.3A)3A) or overexpression of Rheb (Fig. (Fig.6B)6B) retarded the mobility of raptor on SDS-PAGE. Moreover, the Rheb-induced mobility shift of Myc-raptor was significantly reduced when mTORC1 contained KD-mTOR (Fig. (Fig.6B).6B). These data suggest that in addition to promoting mTOR S1261 phosphorylation and S2481 autophosphorylation, TSC null status and Rheb promote extensive phosphorylation of raptor on many sites in a manner that requires mTORC1's kinase activity. These observations are consistent with recent reports demonstrating site-specific phosphorylation of raptor (5, 23, 70).
Since an altered interaction of raptor or mLST8/GβL with S1261A-mTOR could explain the defect in mTORC1 autophosphorylation or mTORC1-mediated substrate phosphorylation, we examined the coimmunoprecipitation of S1261A-mTOR with raptor and mLST/GβL. We cotransfected HEK293 cells with various isoforms of Myc-mTOR together with HA-raptor and HA-mLST/GβL. We found no difference in the ability of HA-raptor or HA-mLST8/GβL to coimmunoprecipitate with Myc-tagged S1261A-mTOR (in either the WT or RR backbones) compared to WT-mTOR (or RR-mTOR) under either steady-state conditions (data not shown) or in the absence or presence of insulin (Fig. (Fig.6C).6C). Thus, phosphorylation of mTOR S1261 does not modulate the interaction of mTOR with its partners. Moreover, as S1261A-mTOR interacts with its partners normally, substitution of Ala for Ser at amino acid residue 1261 did not grossly impair mTOR's protein structure. We noted additionally that when we examined either transfected (Fig. (Fig.6C)6C) or endogenous mTOR and raptor (data not shown), insulin mediated a moderately weakened interaction of raptor with mTOR while it had no effect on the interaction of mLST8/GβL with mTOR. This observation is reminiscent of the nutrient-regulated modulation of the mTOR-raptor interaction reported by Kim et al., who observed a weak mTOR-raptor interaction in nutrient-replete cells (high mTORC1 signaling) and a stronger mTOR-raptor interaction in nutrient-depleted cells (low mTORC1 signaling) (33). Taken together, these observations suggest that insulin as well as nutrients induce allosteric conformational changes in mTORC1 that affect the mTOR-raptor interaction and that may contribute to the regulation of mTORC1 signaling. Our data indicate that failure to phosphorylate S1261 is not sufficient to abrogate the insulin-induced weakening of the mTOR-raptor interaction.
We reported previously that the level of mTORC1 signaling controls the size at which cells proliferate, which is a measure of the overall level of cell growth (16). To elucidate the molecular mechanisms underlying mTORC1-regulated cell growth, we asked whether S1261 phosphorylation plays a role in this major mTORC1 function. We therefore cotransfected HEK293 cells with GFP-tagged spectrin and a panel of Myc-tagged mTOR isoforms and used flow cytometry to measure the mean FSC-H (a measure of relative cell size) of the transfected (GFP-positive), G1-phase population in the absence and presence of rapamycin. Combined cell size data from three experiments, each performed in quadruplicate (n = 12), are shown in Fig. Fig.7A.7A. As reported previously (16), inhibition of mTORC1 signaling with rapamycin under steady-state conditions decreased the mean FSC-H of cells expressing WT mTOR by ~10%, expression of RR-mTOR rescued the rapamycin-mediated decrease in cell size, and expression of RR/KD-mTOR failed to rescue the rapamycin-mediated decrease in cell size (Fig. (Fig.7A).7A). Strikingly, cells expressing RR/S1261A-mTOR in the presence of rapamycin were statistically smaller than those expressing RR-mTOR, indicating a defect in the ability of RR/S1261A-mTOR to drive cell growth to increased cell size (Fig. (Fig.7A)7A) (ANOVA followed by Tukey's post hoc tests confirmed the statistical significance of the FSC-H means). These data indicate that phosphorylation of mTOR S1261 contributes to mTORC1-driven cell growth. Interestingly, cells expressing the phospho-mimetic RR/S1261D allele exhibited a size that was statistically larger than those expressing RR/S1261A but not statistically as large as those expressing RR-mTOR. These data suggest that in cycling cells, the phospho-mimetic S1261D allele indeed mimics phosphorylation, albeit weakly. It is important to note that for each cell size experiment, we confirmed equal expression of the various Myc-mTOR alleles (the immunoblot corresponding to one representative cell size experiment is shown in Fig. Fig.7A).7A). To further analyze the ability of S1261D to mimic phosphorylation at S1261, we compared the ability of the Myc-tagged RR/S1261A versus RR/S1261D-mTOR alleles to rescue rapamycin-inhibited phosphorylation of S6K1 under steady-state conditions (as done for Fig. Fig.4A)4A) (Fig. (Fig.7B).7B). As expected, RR-S1261A demonstrated impaired signaling to cotransfected HA-S6K1 compared to RR-mTOR. While the signaling capacity of RR/S1261D-mTOR was not statistically higher than RR/S1261A-mTOR, mean HA-S6K1 phosphorylation in cells expressing RR/S1261D was higher than in cells expressing RR/S1261A. These observations are consistent with the cell size data and suggest that the Asp (D) S1261 substitution weakly mimics phosphorylation.
Here we have elucidated the poorly understood mechanisms underlying mTORC1 regulation and provided the first evidence that site-specific mTOR phosphorylation (on S1261) regulates mTORC1 function. We show that insulin signals via PI3K to promote the autophosphorylation-independent (Fig. (Fig.1G),1G), rapamycin-insensitive phosphorylation of mTOR S1261 in both mTORC1 and mTORC2 (Fig. 2A to C). Thus, mTOR S1261 phosphorylation is not mTOR complex specific. In mTORC1, maintenance of mTOR P-S1261 requires sufficient levels of amino acids in HEK293 cells and MEFs (Fig. (Fig.2D2D and and3B),3B), similar to the requirement for amino acids in mTORC1-mediated substrate phosphorylation. As recent work suggests that amino acid sufficiency induces the subcellular relocalization of mTOR to a Rab7-positive endomembrane compartment that contains Rheb (52), an intriguing hypothesis presents itself: perhaps mTOR S1261 phosphorylation occurs in this unique Rheb-containing subcellular compartment. In the absence of amino acids, mTORC1 may relocalize out of this compartment, and thus S1261 phosphorylation could not be maintained. Moving downstream of PI3K, we find that TSC suppresses while Rheb promotes mTOR P-S1261, as fibroblasts lacking TSC1 exhibit elevated P-S1261 in the absence of serum growth factors (Fig. (Fig.3A)3A) and Rheb knockdown reduces mTOR P-S1261 in TSC1 null MEFs (Fig. (Fig.3C).3C). By using rapamycin-resistant mTOR alleles containing either wild-type or phosphorylation site-defective S1261 (S1261A) combined with rapamycin to inhibit signaling by endogenous mTORC1, we showed that mTOR P-S1261 is required for phosphorylation and activation of S6K1 and phosphorylation of 4EBP1 (Fig. 3A to C and and5).5). Thus, P-S1261 promotes mTORC1-mediated biochemical signaling to its best-characterized substrates. Additionally, Rheb-driven S6K1 phosphorylation (Fig. (Fig.3D),3D), mTOR S2481 autophosphorylation in vivo (Fig. (Fig.6B),6B), and mTORC1-driven cell growth (Fig. (Fig.7A)7A) require P-S1261.
Based on our data, we propose the following simple model. Insulin signaling via PI3K mediates the TSC/Rheb-dependent phosphorylation of mTOR S1261, which contributes to activation of mTORC1's intrinsic catalytic activity (as measured by S2481 autophosphorylation) and thereby promotes mTORC1-mediated substrate phosphorylation (e.g., S6K1, 4EBP1), biochemical signaling, and cell growth (Fig. (Fig.8).8). Elucidation of the precise relationship between TSC/Rheb and mTOR P-S1261 awaits the identification and molecular cloning of the unknown P-S1261 regulator. While the model shown in Fig. Fig.88 suggests that TSC/Rheb lies upstream of the P-S1261 regulator in a linear pathway in which Rheb modulates the activation state of the P-S1261 regulator, it is possible that TSC/Rheb and the P-S1261 regulator lie on parallel pathways that converge on mTORC1 to cooperatively activate the complex. Such an idea has precedent, as Akt mediates mTORC1 activation via parallel pathways, by the direct phosphorylation of TSC2, which activates Rheb (27), and by the direct phosphorylation of PRAS40, which relieves PRAS40's inhibitory effect on mTORC1 via induced dissociation (54, 67). It also remains unclear as to whether insulin promotes mTOR P-S1261 by modulating the action of a kinase or a phosphatase. The basic nature of the amino acids located in the S1261 consensus (KKLHVpS) suggests that a basophilic kinase mediates mTOR S1261 phosphorylation.
We are not the first to investigate the phosphorylation of mTOR. Prior to this work, three phosphorylation sites on mTOR had been identified: S2448, S2481, and T2446 (in order of discovery) (7, 43, 49). While S2448 was originally thought to be phosphorylated by Akt (43, 59), it is now believed that S6K1 mediates S2448 phosphorylation as part of a feedback loop whose function is currently unknown (8, 25). S2481 was identified as a site that underwent autophosphorylation in vitro and in vivo; additionally, growth factors were shown to weakly increase P-S2481 (49). This work proposed that P-S2481 is a measure of mTOR's in vivo autokinase activity (49). T2446 was identified as a site whose phosphorylation increased upon nutrient withdrawal (possibly via AMPK) and decreased upon insulin stimulation (7). Analysis of these phosphorylation sites failed to provide evidence that site-specific mTOR phosphorylation regulates mTOR signaling (49, 59). It is important to note, however, that the ability of mTOR to activate S6K1 was the only assay of mTOR function utilized in these studies. It remains possible, however, that these three P-sites participate in regulation of mTOR functions not yet identified, or alternatively, if several P-sites cooperatively regulate mTOR signaling, then analysis of individual P-sites may not be sufficient to observe a functionally defective phenotype. As our data reveal a modest defect in S1261A-mTOR signaling to mTORC1 substrates, an outstanding question remains as to whether P-S1261 cooperates with any of the three previously identified P-sites, or with other not-yet-identified P-sites, to coordinately regulate mTORC1 function.
Our data demonstrating that insulin/PI3K signaling and Rheb overexpression strongly promote mTOR S2481 autophosphorylation extend the work of Peterson et al. (49) and support a model whereby growth factors activate mTORC1 signaling by increasing mTORC1's intrinsic catalytic activity. This model has not had strong experimental support, however, as attempts to observe significant insulin-stimulated activation of mTORC1 by in vitro kinase assay have proven historically unsuccessful. Thus, it was postulated that insulin may activate mTORC1 signaling by a mechanism other than regulation of mTORC1's catalytic activity. Recently, however, Sancak et al. (54) overcame technical problems inherent in commonly employed mTORC1 in vitro kinase assay protocols and demonstrated consequently that insulin greatly augments mTORC1's kinase activity, as assayed by in vitro phosphorylation of recombinant S6K1 and 4EBP1 (54). Furthermore, addition of recombinant Rheb-GTP to mTORC1 isolated from serum-deprived cells also potently activates mTORC1's in vitro kinase activity toward substrates (54). Collectively, the data demonstrating significant insulin-stimulated activation of mTORC1 kinase activity, as monitored by in vivo mTOR S2481 autophosphorylation (Fig. (Fig.6)6) and by in vitro kinase assay (54), support the notion that insulin increases mTORC1's kinase activity toward itself and toward exogenous substrates. As Rheb-driven mTOR S2481 autophosphorylation requires mTOR S1261 phosphorylation (Fig. (Fig.6B),6B), these data indicate a role for mTOR S1261 phosphorylation in regulation of mTORC1's intrinsic catalytic activity.
In addition to our observations that phosphatase treatment increases mTOR's electrophoretic mobility on SDS-PAGE (Fig. (Fig.1A),1A), which suggests the extensive phosphorylation of mTOR in vivo, we have also observed the extensive phosphorylation of raptor (Fig. (Fig.2C2C and and6B,6B, our observations reported here, and also unpublished data). Consistent with these observations, recent work reports the phosphorylation of raptor: AMPK, a kinase activated by energy stress, directly phosphorylates raptor to attenuate mTORC1 signaling (23); RSK, a mitogen-activated kinase in the MAPK pathway, directly phosphorylates raptor to promote mTORC1 signaling (5); mTOR itself phosphorylates raptor to promote mTORC1 signaling (70). Moreover, several groups have reported that Akt and mTOR phosphorylate PRAS40, an mTORC1 interactor that displays properties of both an inhibitor and substrate, which results in the dissociation of PRAS40 from mTORC1 and promotes mTORC1 signaling to other substrates (e.g., S6K1 and 4EBP1) (17, 45, 54, 66-69). Collectively, these data suggest that multiple phosphorylation events on mTORC1 components (e.g., mTOR, raptor, PRAS40) cooperate to regulate mTORC1 signaling, thus ensuring precise mTORC1 regulation in response to diverse and dynamic environmental cues. It will be important in the future to identify the basophilic S1261 kinase, to determine whether P-S1261 cooperates with other P-sites on mTOR or its partners to regulate mTORC1 signaling, and to determine whether mTOR P-S1261 occurs in a distinct subcellular compartment. As insulin/PI3K signaling promotes mTOR P-S1261 in mTORC2, it will be critical to determine whether S1261 phosphorylation promotes mTORC2-mediated phosphorylation of Akt on its hydrophobic motif, S473, or the phosphorylation of other AGC kinase family members (e.g., PKCα, SGK1) on their respective hydrophobic motifs, S657 and S422 (1, 20, 26, 55, 56). Lastly, as mTOR S1261 phosphorylation does not appear to modulate the insulin-induced destabilization of the mTOR-raptor interaction (Fig. (Fig.6C),6C), it will be important to identify the biochemical mechanism underlying this regulated interaction.
We thank all members of the laboratory as well as Martin G. Myers, Jr., for critical reading of the manuscript. We are indebted to John Blenis for encouragement. Thanks go to David Sabatini, Robert Abraham, Nahum Sonenberg, Ormond MacDougald, David Kwiatkowski, and John Blenis for generously sharing reagents.
This work was supported by NIH R01 DK-078135, the American Diabetes Association, the American Cancer Society, and the Michigan Diabetes Research and Training Center (D.C.F.); the American Heart Association (G.A.S. and J.A.K.); and the Vermont Genetics Network (NIH P20 RR16462 from the Institutional Development Award Networks of Biomedical Research Excellence Program of the National Center for Research Resources) (B.A.B.). This work utilized the Cell and Molecular Biology Core(s) of the the Michigan Diabetes Research and Training Center funded by NIH5P60 DK20572 from NIDDK at the University of Michigan.
Published ahead of print on 1 June 2009.