Akt-dependent Activation of mTORC1 Complex Involves Phosphorylation of mTOR (Mammalian Target of Rapamycin) by IκB Kinase α (IKKα)^*

Antibodies and Reagents

Antibodies were obtained from the following sources. Antibodies against IKKα, IKKβ, mTOR, and GST were obtained from Upstate Biotechnology. Raptor, Rictor, and GβL antibodies were obtained from Bethyl Laboratories. Anti-HA and anti-FLAG antibodies were obtained from Roche Applied Science and Sigma, respectively. The phospho-IKKα-T23 antibody is from the Abcam (ab38515). Anti-actin was obtained from Calbiochem. The anti-myc (9E-10), anti-tubulin, anti-S6K, anti-PTEN, and control rabbit IgG as well as horseradish peroxidase-labeled anti-mouse and anti-rabbit secondary antibodies were from Santa Cruz Biotechnology. Phosphoserine antibody was from BD Transduction Laboratories. All other antibodies were from Cell Signaling. Purified IKKα protein was from Upstate Biotechnology. Phospho-mTOR-Ser-1415 antibody is generated by with 21st Century Biochemicals. Other reagents were obtained from the following sources. Protease and phosphatase inhibitor cocktails were from Roche Applied Science. CHAPS was from Pierce. Protein A and protein G-agarose beads were from Invitrogen. All radiochemicals used were obtained from PerkinElmer Life Sciences.

Plasmids

GST-IKKα WT and KM were gifts from J. Hutti. All other plasmids below were obtained via Addgene: the pRK5/Myc-mTOR, pRK5/Myc-mTOR-KD, and HA-Raptor vectors were from D. Sabatini; the pRK7/HA-S6K1, FLAG-4E-BP1, GST-S6K, and GST-4E-BP1 were from J. Bleni; the FLAG-mTOR WT and kinase-inactive vectors were from J. Chen.

cDNA Mutagenesis and Sequencing

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-myc-mTOR backbones: S1415A, S1418A, and S1415A/S1418A. Mutations of S1415A, S1418A, S1415E, and S145A/S1418A in FLAG-mTOR were generated by DNA Express Inc.

Cell Lines, Cell Culture, Transient Transfection

IKKα wild type and IKKα^−/− MEFs were provided by I. Verma and M. Karin. Eker rat embryo fibroblast TSC2 wild type and TSC2^−/− cells were from J. Cheng and originally from R. Yeung. HEK293T and HeLa as well as prostate cancer cell lines PC3 and LNCaP and other PTEN mutated cancer cell lines were from American Type Culture Collection (ATCC). All cells were maintained in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mmol/liter glutamine, and 100 units/ml penicillin and streptomycin (Invitrogen). Transfections were done using Polyfect Transfection Reagent (Qiagen) or Lipofectamine Plus (Invitrogen) following the manufacturer's instructions. 3–4 h after transfection, cells were recovered in full serum for 36 h or in full serum for 24 h and then serum-starved for 16–24 h as indicated.

RNA Interference

Small interfering RNA (siRNA) SMARTpool IKKα, Akt1, and Akt2 were from Dharmacon. Each of these represents four pooled SMART-selected siRNA duplexes that target the indicated gene. PC3 cells were transfected with the indicated SMARTpool siRNA or nonspecific control pool using DharmaFECT 1 reagent (Dharmacon) according to the manufacturer's instructions. In brief, 20 nmol/liter final concentration of siRNA was used to transfect cells at 60–70% confluency. Twenty-four hours after transfection, cells were recovered in full serum or were serum-starved 16 h before harvest. Cells were harvested 48–72 h after siRNA transfection.

Cell Lysis, Immunoblotting, and Coimmunoprecipitations

Cells growing in 100-mm dishes were rinsed twice with cold PBS and then lysed on ice for 20 min in 1 ml of lysis buffer (40 mmol/liter HEPES (pH 7.5), 120 mmol/liter NaCl, 1 mmol/liter EDTA, 10 mmol/liter pyrophosphate, 10 mmol/liter glycerophosphate, 50 mmol/liter NaF, 0.5 mmol/liter orthovanadate, and EDTA-free protease inhibitors; Roche Applied Science) containing 1% Triton X-100. After centrifugation at 13,000 × g for 10 min, samples containing 20–50 μg of protein were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and proteins were transferred to Pure Nitrocellulose Membrane (Bio-Rad), blocked in 5% nonfat milk, and blotted with the indicated antibodies. For immunoprecipitation experiments, the lysis buffer contained 0.3% CHAPS instead of 1% Triton. Four micrograms of the indicated antibodies were added to the cleared cellular lysates and incubated with rotation for 6–16 h. Then 25 μl of protein G-agarose were added, and the incubation was continued for 1 h. Immunoprecipitates captured with protein G-agarose were washed 3 times with the CHAPS lysis buffer and twice by wash buffer A (50 mmol/liter HEPES, (pH 7.5), 150 mmol/liter NaCl) and boiled in 4× SDS sample buffer for Western blot.

In Vitro mTOR Kinase Assay

Transfected HEK293T cells were grown in 100-mm dishes for 48 h in DMEM containing 10% FBS and lysed in 1 ml of lysis buffer with 0.3% CHAPS. Half of total cell lysate was incubated with anti-mTOR or FLAG antibody for 3 h followed by another hour of incubation with 25 μl of protein G-agarose beads. Immunoprecipitates were washed twice by lysis buffer, twice by wash buffer B (20 mmol/liter Tris (pH 7.5), 500 mmol/liter NaCl, 1 mmol/liter EDTA, 20 mmol/liter β-glycerophosphate, 5 mmol/liter EGTA, 1 mmol/liter DTT, 1 mmol/liter orthovanadate, 40 mg/ml phenylmethylsulfonyl fluoride (PMSF), 10 μg/ml leupeptin, 5 μg/ml pepstatin), once with wash buffer C (10 mmol/liter HEPES (pH 7.4), 50 mmol/liter glycerophosphate, 50 mmol/liter NaCl, 1 mmol/liter DTT, 1 mmol/liter orthovanadate, 40 mg/ml PMSF, 10 μg/ml leupeptin, 5 μg/ml pepstatin), and once with mTOR kinase assay buffer without ATP (10 mmol/liter HEPES (pH 7.4), 50 mmol/liter NaCl, 50 mmol/liter glycerophosphate, 1 mmol/liter DTT, 10 mmol/liter MgCl₂, 4 mmol/liter MnCl₂). Kinase assay toward recombinant GST-S6K1 (amino acids 308–400) in washed immunoprecipitates was done for 30 min at 30 °C in 30 μl of mTOR kinase buffer with 100 μmol/liter ATP unlabeled and 10 μCi [γ-³²P]ATP (PerkinElmer Life Sciences). To stop the reaction, 6 μl of 4× SDS sample buffer was added to each reaction, which was boiled for 10 min. The reaction was then separated by 4–12% SDS-PAGE and transferred to polyvinylidene difluoride membranes. ³²P incorporated into GST-S6K was assessed by autoradiography. In a cold in vitro kinase assay to GST-S6K, phosphorylation S6K was detected by phosphor-S6K-Thr-389 antibody.

In Vitro IKKα Kinase Assay

To map the phosphorylation sites in mTOR, the fragments of the mTOR coding sequence were cloned into pGEX vector (GE Healthcare). Purified GST-mTOR fusion proteins were immobilized to glutathione-agarose for kinase assay. Kinase assays were performed following a previously described protocol (Upstate Biotechnology). Kinase activity was determined by incubating purified IKKα with GST-mTOR fragments, GST-IκB, or immunoprecipitates of mTOR and Raptor from HEK 293T cells as indicated in the presence of 1 μCi ml⁻¹ [γ-³²P]ATP or cold ATP (100 μm) for 30 min at 30 °C. Reactions were resolved by SDS-PAGE (4–12%) and processed for autoradiography or protein immunoblotting.

Mass Spectrometric Analysis to Identify mTOR Phosphorylation Sites

Purified GST-mTOR-(1351–1650) protein was incubated with recombinant IKKα for cold kinase assay, resolved by SDS-PAGE, and stained with Coomassie Blue, excised from the gel, digested with trypsin, and analyzed by tandem mass spectrometry by Dr. John Asara of the Beth Israel Deaconess Medical Center.

Cell Proliferation Assays

The indicated transfected cells were plated in 6-well plates at a density of 2.0 × 10⁴ cells/well. Cells were trypsinized and counted using a hematocytometer every day until confluency. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays were performed according to the manufacturer's recommendations (Promega, Madison, WI). The cells were plated in 96-well microtiter plates at a density of 1.0 × 10³ cells/well in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. The number of cells at 1, 2, and 3 days was determined using a cell counter and the colorimetric CellTiter96 AQueous (MTS) assay (Promega). Results were depicted as absorbance at 490 nm as a function of time.

In Vivo Experimentation

The conditional IKKα^loxp/loxp mice are described in Gareus et al. (42). Pb-cre mice were from Dr. Van Dyke's group. IKKα^loxp/loxp/C⁺ males were bred to IKKα^loxp/loxp females to obtain the IKKα^loxp/loxp/C⁺ genotype. Male IKKα^loxp/loxp/C⁺ mice between 6 and 15 weeks of age were used for experiments. All animal procedures were reviewed and approved by the University of North Carolina Institutional Animal Care and Use Committee.

Statistics

Data from the in vitro experiments are expressed as the mean ± S.E. from a minimum of three independent experiments. Comparison between groups were carried out by two-way analysis of variance or Student's t test, and a p value of <0.05 was considered significant.

IKKα Activates mTORC1 in a Kinase-dependent Manner

Our previous studies demonstrating the ability of IKKα to promote mTORC1 activity (37–39) prompted an examination of whether IKKα regulates mTORC1 through its kinase activity. To address this issue, S6K, a known mTOR substrate, was expressed as an HA-tagged version with FLAG-tagged IKKα wild type or kinase-inactive forms. Results from this experiment revealed that expression of IKKα wild type, but not a kinase mutant, enhances S6K phosphorylation at Thr-389 (Fig. 1A). Similarly, phosphorylation of 4E-BP1, another mTOR substrate, was enhanced with wild type IKKα expression but not with the kinase-dead variant (Fig. 1B). Next, we determined whether IKKα affects endogenous S6K phosphorylation and found that wild type IKKαsignificantly promoted, but kinase mutant IKKα inhibited, endogenous S6K phosphorylation in HEK 393T cells (Fig. 1C). These results demonstrate that IKKα activates mTOR in a kinase-dependent manner. To determine whether IKKα functions to promote mTOR activity through regulation of TSC2, it was determined whether IKKα affects the hyperactive mTORC1 signaling found in TSC2-null cells in which Rheb activates mTORC1 activity. It would be predicted that IKKα could not modulate mTORC1 activity in TSC2 null cells if IKKα regulates mTORC1 through TSC2. First, IKKα expression was silenced by siRNA-directed knockdown, and endogenous phosphorylation of S6K was analyzed. The data demonstrate that knockdown of IKKα significantly impaired mTORC1 activity in these cells (Fig. 1D). Moreover, expression of IKKα WT, but not kinase-inactive IKKα, promotes mTORC1 activity in both TSC2 wild type and TSC2-null cells (Fig. 1E). Collectively, the data indicate that IKKα activates mTORC1 in a kinase-dependent manner that is at least partly independent of TSC2. In addition, we tested the effects of wild type and kinase-inactive IKKα on insulin-induced mTORC1 activity. Endogenous IKKα was knocked down with siRNA (data not shown), and wild type and kinase-inactive FLAG-IKKα was transiently transfected into HeLa cells. As shown in Fig. 1F, insulin induces mTORC1 activity in IKKα wild type, but not the mutant, transfected cells (Fig. 1F). These data indicate that IKK also regulates insulin-induced mTORC1 activity in a kinase-dependent manner.

IKKα Phosphorylates mTOR in Vitro and in Vivo

Previous studies demonstrated that IKKα interacts with mTORC1 downstream of active Akt (37), and the results above demonstrate that IKKα controls mTORC1 through its kinase activity. These results prompted us to investigate whether IKKα regulates mTOR through phosphorylation of a component of the mTORC1 complex. To this end it was determined whether IKKα can directly phosphorylate mTOR or Raptor. As expected, an antibody to mTOR immunoprecipitated mTOR, Raptor, and Rictor (Fig. 2A). An in vitro phosphorylation assay using the immunoprecipitate with purified IKKα led to phosphorylation of mTOR but not Raptor or Rictor. Similarly, an antibody to Raptor immunoprecipitated mTOR and Raptor but not Rictor. In vitro phosphorylation of this immunoprecipitate using IKKα yielded phosphorylation of mTOR but not Raptor or Rictor (Fig. 2A). To rule out the possibility that the phosphorylation of mTOR is due to mTOR autophosphorylation, FLAG-tagged wild type and kinase-inactive mTOR were expressed in 293T cells and immunoprecipitated with the FLAG antibody. As before, an in vitro IKKα kinase assay was performed which demonstrated that both wild type mTOR and kinase-inactive mTOR were phosphorylated by IKKα to a similar level indicating that the induced phosphorylation is not due to mTOR autophosphorylation (Fig. 2B). It was determined whether IKKα could directly phosphorylate mTOR in vitro in a nonradioactive kinase assay. As shown in Fig. 2C, recombinant IKKα phosphorylated immunoprecipitated mTOR as detected through recognition with an anti-phosphoserine antibody. Similarly, both immunoprecipitated WT and kinase-inactive mTOR were phosphorylated by IKKα (Fig. 2D). These data demonstrate that IKKα can phosphorylate mTOR in vitro. To determine whether IKKα phosphorylates mTOR in vivo, IKKα^−/− MEFs were transfected with IKKα, and endogenous mTOR was subsequently immunoprecipitated. As with the in vitro studies, the anti-phosphoserine antibody showed significantly enhanced reactivity for mTOR after IKKα transfection (Fig. 2E). Previously we demonstrated that the PTEN null prostate cancer cell line PC3 exhibited strong mTOR activity that is controlled partly by IKKα (37). Consistent with this, endogenous mTOR exhibits phosphorylation as recognized by the anti-phosphoserine antibody, and this phosphorylation is reduced by siRNA knockdown of IKKα (Fig. 2F). These results indicate that mTOR is a direct substrate of the kinase activity of IKKα.

IKKα Phosphorylates mTOR at Serine 1415

To identify a potential phosphorylation site(s) for IKKα on mTOR, nine fragments of mTOR that encompass the full-length protein were produced as GST fusions and were used as substrates in an IKKα-driven in vitro kinase assay. Results from this approach showed that the 1351–1650 fragment of mTOR was phosphorylated by IKKα, whereas other fragments were not phosphorylated significantly (Fig. 3A). When normalized to input, IκBα exhibits slightly higher phosphorylation by IKKα than mTOR (Fig. 3B). Moreover, further experimentation demonstrated that the purified wild type GST-IKKα, but not kinase-inactive IKKα, phosphorylates mTOR-(1351–1650) (Fig. 3C). These data ruled out the possibility that a contaminating kinase could phosphorylate mTOR. Using an in vitro IKKα kinase approach, mTOR-(1351–1650) was phosphorylated (nonradioactively) and was gel-purified for mass spectrometry analysis. The results demonstrated that both serine 1415 and serine 1418 of mTOR were phosphorylated by IKKα in vitro (Fig. 3D). Serines 1415 and 1418 are located in a conserved region of mTOR (Fig. 3E), and the sequence surrounding Ser-1415 exhibits homology with the human Foxo3a IKK site (33, 43). Mutation of Ser-1415 and Ser-1418 to alanines blocked in vitro phosphorylation of mTOR by IKKα (Fig. 3F), suggesting that this region is the predominant site of IKKα phosphorylation on mTOR. To investigate mTOR phosphorylation by IKKα in vivo, a rabbit polyclonal antibody against a peptide of mTOR including phospho-Ser-1415 was generated (p-mTOR-Ser-1415). Immunoblotting of exogenous immunoprecipitated WT and mutant mTOR revealed that the phospho-Ser-1415 antibody recognized the wild type protein but not mTOR mutated in serines 1415 and 1418 (Fig. 3G). PC3 prostate cancer cells have a high level of basal mTORC1 activity due to PTEN deletion and subsequent activation of Akt. Previously we showed that PC3 cells utilize IKKα to promote mTORC1 activity downstream of Akt (37). To determine if IKKα is involved in endogenous mTOR Ser-1415 phosphorylation, IKKα expression was silenced in PC3 cells, and endogenous Ser-1415 phosphorylation was detected. The results showed that Ser-1415 mTOR phosphorylation in PC3 cells is reduced when IKKα expression is knocked down with siRNA. Moreover, the decrease of phospho-mTOR-Ser-1415 is consistent with the reduction of phospho-S6K (Fig. 3H). Virtually identical results were found in other PTEN-null cancer cell lines (LNCaP, U87, Jurkat, K562, and U937) when IKKα expression was knocked down (Fig. 3H). Taken together, the results demonstrate that mTOR serine 1415 is a primary site of IKKα phosphorylation, although we cannot eliminate the possibility that there are other residues on mTOR that can be phosphorylated by IKKα.

Previously we found that IKKα also contributes to mTORC1 activity downstream of insulin and TNF in a manner dependent on Akt (39). To determine whether IKKα is involved in insulin-induced mTOR activation in vivo, wild type and prostate IKKα-deleted mice were treated with insulin for 30 min. The results showed that insulin treatment led to significant phosphorylation of mTOR (Ser-1415) and S6K (Thr-389) in wild type mice but not in IKKα-deleted mice. The data indicate that insulin induces IKKα to phosphorylate mTOR to promote mTORC1 activity in mice (Fig. 3I).

Phosphorylation of mTOR at Ser-1415 by IKKα Promotes mTORC1 Kinase Activity

Given the involvement of IKKα in the activation of mTORC1 and its induction of phosphorylation of mTOR, we next examined whether the promotion of mTORC1 activity by IKKα is through phosphorylation of mTOR. To compare WT and S1415A mTOR, a transient expression assay was performed using exogenously expressed HA-S6K as a marker of mTOR activity. Expression of wild type mTOR enhanced S6K phosphorylation, as expected; however, expression of mTOR S1415A or S1415A/S1418A led to the inability of mTOR to promote S6K phosphorylation (Fig. 4A, left panel). Mutation of mTOR at Ser-1418 did not block mTOR activity, consistent with a key role for Ser-1415 and not Ser-1418 phosphorylation. To address whether it is the mutation of Ser-1415 that affects mTOR activity, we generated a phosphomimetic mutant (mTOR S1415E). The ability of mTOR wild type, mTOR-S1415A (A-form), and a phosphomimetic mutant of mTOR-S1415E (E-form) to phosphorylate S6K was then compared. The results demonstrated that the E-form of mTOR has stronger ability to promote S6K phosphorylation as compared with wild type mTOR (Fig. 4A). As before, mTOR S1415A did not promote S6K phosphorylation. In addition, WT mTOR was activated by insulin to induce phosphorylation of exogenous S6K, whereas mTOR S1415A was not (Fig. 4B). Moreover, expression of wild type mTOR increased, whereas the A-form of mTOR was functionally impaired relative to the ability to induce endogenous S6K phosphorylation (Fig. 4C). We then compared wild type mTOR with the E-form relative to induction of endogenous S6K phosphorylation (Fig. 4D). The results demonstrate that the E-form phosphomimetic is more active than wild type mTOR in promoting endogenous S6K phosphorylation. Because PC3 cells exhibit potent mTOR activity dependent on elevated Akt activity, we chose to knock down endogenous mTOR in these cells (Fig. 4E, left panel) and then re-express mTOR wild type, S1415A, and S1415A/S1418A. Expression of WT mTOR promoted S6K phosphorylation in PC3 cells, whereas S1415A expression was ineffective at promoting this response (Fig. 4E, right panel). Interestingly, and consistent with previous results, mutation of Ser-1418 did not significantly affect the ability of mTOR to phosphorylate S6K (Fig. 4E). These results indicate that IKKα enhances the ability of mTOR to phosphorylate S6K through phosphorylation of mTOR at serine 1415 but not serine 1418. In further experiments, wild type mTOR enhanced, but S1415A mTOR abolished, IKKα-induced exogenous S6K phosphorylation (data not shown). These data indicate that IKKα promotes mTORC1 activity through direct phosphorylation of mTOR Ser-1415.

We next investigated the role of phosphorylation of Ser-1415 mTOR by IKKα in regulating catalytic activity of mTORC1. To examine whether IKKα promotes mTOR kinase activity, HEK293T cells were cotransfected with FLAG-mTOR and HA-IKKα, and FLAG-mTOR was immunoprecipitated for an in vitro kinase assay using GST-S6K. Phosphorylation of S6K (Thr-389) was significantly increased with mTOR co-transfected with IKKα relative to mTOR without IKKα cotransfection (Fig. 4F). To compare wild type and S1415A mTOR in phosphorylating GST-S6K in vitro, kinase assays of the immunoprecipitated wild type mTOR and mTOR A1415A were performed using GST-S6K as the substrate, and the phosphorylation of Thr-389 were detected by the phospho-S6K antibody. A dramatic reduction of S6K Thr(P)-389 was observed with S1415A mTOR as compared with wild type mTOR (Fig. 4G). These data indicate that phosphorylation of mTOR Ser-1415 promotes mTOR catalytic activity.

IKKα-dependent Effects on mTOR and Raptor Interaction

Previous studies have observed that nutrients such as amino acids and glucose, which activate mTORC1, lead to a reduction in the affinity (but not stoichiometry) of mTOR-Raptor association, as measured by coimmunoprecipitation (40, 41). These observations suggest that a conformational change or modification within the mTORC1 complex is related to mTORC1 activity. To determine whether IKKα modulates mTOR-raptor interaction, IKKα^−/− MEFs and HEK293T cells were transfected with wild type HA-IKKα, and mTOR was immunoprecipitated from the cell lysates. The results demonstrate that IKKα destabilizes the interaction of mTOR with Raptor but did not affect mTOR-GβL interaction (Fig. 5A). Additionally, IKKα expression did not change mTOR-Rictor interaction in the mTORC2 complex (Fig. 5A). Next, to determine if IKKα could alter mTOR-raptor interaction in vitro, recombinant IKKα was incubated with mTOR immunoprecipitated from HEK 393T cells in an in vitro IKKα kinase assay. The results demonstrated that the affinity of interaction of mTOR and Raptor decreased in the presence of IKKα (Fig. 5B) consistent with the observation found in vivo. Further experiments showed that expression of wild type IKKα, but not kinase-inactive IKKα, weakened the mTOR-raptor interaction (Fig. 5C). To further address the ability of IKKα to modulate mTOR-raptor interaction, HA-Raptor was transfected with WT or S1415A FLAG-mTOR. Results from this experiment showed that wild type mTOR exhibits weaker interaction with Raptor as compared S1415A mTOR (Fig. 5D). Moreover, the E-form mTOR shows reduced association with exogenous and endogenous Raptor as compared with wild type mTOR (Fig. 5, E and F). These results imply that IKKα regulates mTOR-raptor interaction through phosphorylation of mTOR. To address this hypothesis, IKKα was transfected with myc-mTOR and HA-Raptor in 293T cells, and mTOR was immunoprecipitated. These data revealed that IKKα-induced mTOR phosphorylation was accompanied by a decreased mTOR-Raptor association (Fig. 5G). Consistent with this result, knockdown of IKKα decreased mTOR Ser-1415 phosphorylation and enhanced mTOR-Raptor interaction (Fig. 5H). Conversely, overexpression of IKKα increased mTOR phosphorylation and weakened the mTOR-Raptor interaction (Fig. 5I). These data indicate that IKKα-induced Ser-1415 phosphorylation promotes a reduced affinity of Raptor within the mTORC1 complex.

IKKα Modulates mTOR-Raptor Interaction Downstream of Akt

Our previous studies demonstrated that Akt promotes IKKα association with mTORC1, which prompted us to investigate whether Akt controls IKKα phosphorylation and regulation of mTORC1. To address this point, starved HeLa cells were treated with insulin, and mTOR association with Raptor and IKKα was measured. Results from this experiment demonstrated that insulin induces reduced affinity of interaction between Raptor and mTOR while inducing the association between IKKα and mTOR, which are consistent with phosphorylation of both Akt and S6K (Fig. 6A). Inhibition of PI3K/Akt suppresses the reduced interaction between mTOR and Raptor and blocked the association between IKKα and mTOR (Fig. 6B) while blocking Akt phosphorylation as well as the Akt-dependent phosphorylation of TSC2. IKKα was expressed with tagged mTOR and Raptor in HeLa cells, and stimulation with insulin led to enhanced phosphorylation of mTOR at Ser-1415 along with enhanced interaction between IKKα and mTOR (Fig. 6C). Inhibition of PI3K/Akt blocked the insulin-induced phosphorylation of mTOR and promoted mTOR-Raptor interaction (Fig. 6D). These data indicate that insulin, functioning through Akt, controls IKKα association with and phosphorylation of mTOR to activate mTORC1. To further address the ability of Akt to regulate mTORC1 activity and phosphorylation, additional experiments were performed. Inhibition of PI3K/Akt in PC3 cells reduced mTOR phosphorylation at Ser-1415 and promoted Raptor interaction with mTOR (Fig. 6E). Expression of PTEN in the PTEN-negative PC3 cells reduces mTOR Ser-1415 phosphorylation and leads to enhanced mTOR-Raptor interaction (Fig. 6F). Knockdown of either Akt1 or Akt2 by siRNA decreased mTOR phosphorylation at Ser-1415 and enhanced mTOR-Raptor interaction (Fig. 6G). These data demonstrated that the IKKα-mediated mTOR-Raptor interaction is regulated by Akt.

It has been reported that Akt phosphorylates IKKα at threonine 23 to activate NF-κB in response to TNFα treatment or DNA damage (46). Thus, we determine if Akt regulates IKKα-mTORC1 interaction and promotes mTORC1 activity via phosphorylation of IKKα at threonine 23. To examine if Akt phosphorylates IKKα Thr-23, we treated PC3 prostate cancer cells, which have high basal levels of Akt and IKKα, with a PI3K inhibitor, immunoprecipitated IKKα, and immunoblotted with a phospho-IKKα-Thr-23 antibody. Phosphorylation of IKKα at Thr-23 was reduced by the PI3K inhibitor (Fig. 6H). The results are consistent with the ability of Akt to phosphorylate IKKα at Thr-23. To determine if Akt promotes an interaction between IKKα and mTOR via phosphorylation at Thr-23, Akt was transfected with myc-mTOR and with either FLAG-tagged WT IKKα or T23A IKKα. The results show that Akt promotes an interaction between both WT and the T23A mutant IKKα (Fig. 6I). Consistent with these results, transfection of IKKα-T23A enhances mTORC1 activity identically to that generated by WT IKKα (Fig. 6J). These data indicate that Akt-mediated IKKα activation of mTORC1 is not through phosphorylation of IKKα at Thr-23.

IKKα Phosphorylates mTOR to Promote Cell Proliferation

Previously it was shown that mTOR phosphorylation is involved in cell growth and cell proliferation in certain cells (41). Given the functional interaction of IKKα and mTORC1, we hypothesized that IKKα will induce prostate cancer cell proliferation through mTOR phosphorylation. To determine whether IKKα and mTORC1 affect PC3 cell proliferation, siRNA was targeted against IKKα, mTOR, and Raptor, and cell proliferation (cell number) was measured at 48 h after transfection. Consistent with previous studies (38), knockdown of IKKα, mTOR, and Raptor impaired PC3 cell proliferation (Fig. 7A). To determine if Ser-1415 mTOR phosphorylation may contribute to growth of PC3 cells, WT mTOR or S1415A mTOR was stably expressed in PC3 cells (Fig. 7B). Consistent with our hypothesis that phosphorylation of mTOR by IKKα is important for mTOR function, expression of WT mTOR activated and mTOR S1415A reduced endogenous phosphorylation of S6K and 4E-BP1 (Fig. 6B). Expression of WT mTOR led to an increase in cell number, whereas expression of S1415A reduced cell number (Fig. 7C). Additionally, the phosphomimetic S1415E mTOR further promotes cell proliferation as compared with wild type mTOR (Fig. 7D). Next, mTOR wild type and the S1415A mutant were cotransfected with IKKα, and cell proliferation was measured at 3 days after transfection. Results demonstrate that transfection of IKKα with mTOR wild type, but not the S1415A mutant, enhances cell proliferation over IKKα expression alone (Fig. 7E). Collectively, the data indicate that IKKα promotes PTEN-deficient cancer cell proliferation through mTORC1 by mTOR phosphorylation.