Phosphorylation of Raptor by p38β Participates in Arsenite-induced Mammalian Target of Rapamycin Complex 1 (mTORC1) Activation^*

Materials

Sodium arsenite, rapamycin, insulin, and anisomycin were obtained from Sigma. PD98059 and SB203580 were obtained from Calbiochem. Immobilon-P PVDF membrane and Immobilon Western Chemiluminescent HRP Substrate were obtained from Millipore. Autoradiography film was obtained from Kodak.

Generation and Validation of p38α/β/γ/δ Conditional Knock-out Mouse Embryonic Fibroblasts (MEF)

p38α^lox/lox, p38β^lox/lox, p38γ^lox/lox, and p38δ^lox/lox primary MEFs were prepared from E13.5 embryos of respective mice with a locus flanked by LoxP sites and maintained in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen) supplemented with 10% fetal bovine serum (FBS) (Invitrogen) and 10 units/ml of penicillin/streptomycin. Cells were validated by genomic PCR and then immortalized by SV40 large T antigen. To generate gene knock-out MEF cells, the p38α^lox/lox, p38β^lox/lox, p38γ^lox/lox, and p38δ^lox/lox MEF cells were infected with Cre to excise the loxed gene. The gene knock-out was confirmed by genomic PCR.

DNA Constructs

Expression constructs of MKK6b(E), mTOR(KD), murine p38α, human p38β, human p38γ, and human p38δ were described previously (7). Full-length cDNAs of p38α, p38β, p38γ, and p38δ were subcloned into pBOBI vector for lentiviral gene delivery. p38β RNAi constructs were as described previously (35). Full-length or truncated cDNAs of Raptor, Deptor, mLST8/GβL, Rheb, S6K1, and p38β were cloned into mammalian expression vectors pcDNA6 or pCMV5, or a lentiviral vector pBOBI, or the bacterial expression vector pGEX-4T-1. The Raptor-Rheb(C20) expression plasmid was kind gift of Dr. Kunliang Guan (University of California, San Diego, CA). Point mutations of Raptor were generated by a PCR-based site-directed mutagenesis method using Pfu polymerase (Stratagene). The following oligonucleotides were used to create the point mutations; the underlined nucleotides indicate the mismatch (5′ → 3′): S771A-F, AGCAGCACCCTGGGCGCCCCCGAGAATGAGGA; S771A-R, TCCTCATTCTCGGGGGCGCCCAGGGTGCTGCT; S792A-F, ATGCGCCGCGCCAGCGCCTACTCCTCCCTCA; S792A-R, TGAGGGAGGAGTAGGCGCTGGCGCGGCGCAT; S795A-F, GCCAGCTCCTACTCCGCCCTCAACTCCCTCA; S795A-R, TGAGGGAGTTGAGGGCGGAGTAGGAGCTGGC;S863A-F, CAGTCGGCCCCCGCCGCCCCCACCAACAAGGG; S863A-R, CCCTTGTTGGTGGGGGCGGCGGGGGCCGACTG; T865A-F, GCCCCCGCCAGCCCCGCCAACAAGGGCGTGC; T865A-R, GCACGCCCTTGTTGGCGGGGCTGGCGGGGGC; S877A-F, CACCAGGCGGGGGGCGCCCCTCCGGCGTCCA; S877A-R, TGGACGCCGGAGGGGCGCCCCCCGCCTGGTG; S863A on T865A-F, CAGTCGGCCCCCGCCGCCCCCGCCAACAAGGG; S863A on T865A-R, CCCTTGTTGGCGGGGGCGGCGGGGGCCGACTG. shRNA targeted the 3′ UTR of human Raptor and were cloned into pll3.7 vectors. The target sequences were: 1) GGAAACCAAGAGAGAGGAA; 2) GCGAATTGAGACGGACTTT; 3)GCCTGCTGTACATAGTGAA; 4) GCAAGCAGGGACATTTCCT; 5) GGACATTTCCTAGCCAGCT; 6) GCTGTCGTGTCTGGAATGT; 7) GCCTAGAGGCTCATTATCT; 8) CCTAGAGGCTCATTATCTA; 9) CCATGTAATCAGAGCATTA; 10) GAGAGTGACCAACAGTAAA. The following oligonucleotides were used for qPCR (5′ → 3′): (m)p38α_qF, ACCGAGCCCCAGAGATCAT; (m)p38α_qR, CACGGACCAAATATCCACTGTCT; (m)p38β_qF, AGCACGAGAACGTCATAGGACTT; (m)p38β_qR, CATCAGGGTCGTCACGAGGTA; (m)p38γ_qF, TGCGCCTCCTCAAACACAT; (m)p38γ_qR, GAGACTCATCAGGTGTGAACACATC; (m)p38δ_qF, CAAGGCGGCCAAATCCTATA; (m)p38δ_qR, TGGCGCGTGGGAAGAG; (h)p38α_qF, GGTTACGTGTGGCAGTGAAG; (h)p38α_qR, AACGTCCAACAGACCAATCA; (h)p38β_qF, GCCATATGATGAGAGCGTTG; (h)p38β_qR, GTGAGCTCCTTCCACTCCTC; (h)p38γ_qF, AGATGATCACAGGCAAGACG; (h)p38γ_qR, TGTAGTTCTTGGCCTCATCG; (h)p38δ_qF, TCCTCAGCTGGATGCACTAC; (h)p38δ_qR, TGGTCCAGGTAATCTTTCCC. All constructs were verified by DNA sequencing.

Cell Culture, Stimulation, Transfection, and Lentivirus Infection

Primary p38β^+/+ and p38β^−/− MEFs were obtained from E13.5 embryos of the respective mice. PRAK^+/+ and PRAK^−/− MEFs were described previously (35). TSC2^+/+ and TSC2^−/− cells were described previously (37). AMPKα1/α2 double knock-out cells were kindly provided by Drs. Benoit Viollet (Universite Paris Descartes) and Keith R. Laderoute (Stanford University School of Medicine). MK2^−/− cells were kindly provided by Dr. Matthias Gaestel (Hannover Medical School). The HEK293T cell line was obtained from ATCC. All cells were maintained in DMEM (Invitrogen) supplemented with 10% FBS (Invitrogen) and 10 units/ml of penicillin/streptomycin at 37 °C in a humidified incubator containing 5% CO₂.

For insulin or H₂O₂ stimulation, cells were serum-starved for 24 h followed by insulin or H₂O₂ stimulation, as indicated. For glucose starvation, cells were cultured with d-glucose-free DMEM (Invitrogen) containing 25 mm HEPES and 10% dialyzed FBS. Cells were stimulated with 25 mm glucose for 15–60 min as indicated. For amino acid starvation, cells were serum-starved for 24 h and incubated in DPBS (PBS containing 25 mm glucose) for 30 min. DMEM containing all amino acids was then added for 10–30 min as indicated.

293T cells were transfected with Lipofectamine 2000 and the corresponding empty vector was used for adjustment to ensure an equal amount of total DNA. 36 h after transfection, cells were collected using lysis buffer as indicated.

For lentivirus production, 293T cells were transfected with pLV-EF1α-Cre-IRES-Bsd, pBOBI-HA-Raptor, and pBOBI-MKK6b(E) with lentivirus-packaging plasmids (VSVG/PMDL/REV) using Lipofectamine 2000. The transfected cells were cultured at 37 °C for 48 h and the virus generated was used for infection.

Antibodies

Antibodies to phospho-S6K(T389), S6K, phospho-S6(S235/236), phospho-4EBP1(T37/46), 4EBP1, phospho-p38(T180/Y182), p38β, Raptor, and mTOR were obtained from Cell Signaling Technology. Anti-phospho-Raptor(Ser⁸⁶³) and anti-HA(F-7) antibodies were obtained from Santa Cruz Biotechnology. Anti-FLAG M2 antibodies, anti-FLAG M2 beads, and anti-HA beads were obtained from Sigma. Polyclonal antibodies to phospho-Raptor (Ser⁷⁷¹) were generated in rabbits against keyhole limpet hemocyanin-coupled peptides (amino acids 767–775: STLG(pS)PENE) using Abmart's custom antibody service. Goat anti-mouse HRP and goat anti-rabbit HRP antibodies were from Thermo Scientific.

Cell Lysis, Immunoprecipitation, Western Blotting

Cells rinsed once with ice-cold PBS were lysed in ice-cold CHAPS lysis buffer (40 mm HEPES, pH 7.5, 120 mm NaCl, 1 mm EDTA, 10 mm pyrophosphate, 10 mm β-glycerophosphate, 50 mm NaF, 1.5 mm Na₃VO₄, 0.3% CHAPS, 1 μg/ml of leupeptin, and 1 mm PMSF) as described (38). The soluble fractions of cell lysates were isolated by centrifugation at 19,000 × g for 20 min at 4 °C in a microcentrifuge.

For immunoprecipitations, anti-FLAG M2 beads or anti-HA beads were added to the lysates and incubated with rotation for 3 h or overnight at 4 °C. The immunoprecipitates were then washed three times each with lysis buffer followed by addition of 1× SDS sample buffer, boiling for 10 min, resolved by SDS-PAGE, and analyzed by immunoblotting.

Calf Intestinal Alkaline Phosphatase (CIAP) Treatment

As described (24), anti-HA beads used for immunoprecipitating HA-Raptor deletions were washed three times in RIPA buffer (20 mm Tris-HCl, pH 7.5, 150 mm NaCl, 1 mm Na₂EDTA, 1 mm EGTA, 1% Nonidet P-40, 1% sodium deoxycholate, 2.5 mm sodium pyrophosphate, 1 mm β-glycerophosphate, 1 mm Na₃VO₄, 1 μg/ml of leupeptin, 0.1% SDS, and 1 mm PMSF) followed by two additional washes in CIAP buffer (100 mm NaCl, 50 mm Tris, pH 7.9, 10 mm MgCl₂, 1 mm dithiothreitol). Beads were then incubated with 1 μl of 30 units/μl of CIAP (Takara Bio, Inc.) in 50 μl of CIAP buffer for 1 h at 37 °C. Reactions were terminated by washing twice in CIAP buffer and boiling in 1× SDS sample buffer.

In Vitro Kinase Assay

p38β kinase assays were performed in a final volume of 20 μl consisting of p38 kinase buffer with 250 μm ATP and bacterial-derived GST-Raptor-(741–887) fragments as substrates for 30 min at 30 °C. The reactions were stopped by 2× SDS sample buffer, and subjected to SDS-PAGE and immunoblotting.

For the mTORC1 kinase assay, 293T cells were transfected with HA-Raptor WT or S771A, S863A, and S771A/S863A mutants. Transfected cells were treated with 1 mm arsenite for 30 min as indicated in the figures and then lysed in CHAPS lysis buffer followed by immunoprecipitation by HA antibodies. As described (39), the immunoprecipitates were washed three times in CHAPS lysis buffer and then twice in 25 mm HEPES, pH 7.4, 20 mm potassium chloride. Kinase assays were carried out in a volume of 15 μl consisting of mTORC1 kinase buffer (25 mm HEPES, pH 7.4, 50 mm KCl, 10 mm MgCl₂, 250 μm ATP) and bacterial-derived GST-S6K as a substrate for 30 min at 30 °C. The reactions were terminated by 2× SDS sample buffer, and subjected to SDS-PAGE and immunoblotting.

p38β Participates in Arsenite-induced Activation of mTORC1

To study the effect of arsenite on mTOR activity, we measured phosphorylation levels of S6K1 and 4EBP1, substrates of mTOR, in arsenite-treated MEF cells. Arsenite-induced Thr³⁸⁹ phosphorylation of S6K1 and Thr^37/46 phosphorylation of 4EBP1 (Fig. 1A and supplemental Fig. S1A). The induction is rapamycin sensitive, indicating arsenite-activated mTORC1 (Fig. 1A). Arsenite can activate the ERK pathway, a pathway that plays a role in EGF-stimulated mTORC1 activation (24). To determine whether arsenite works via ERK to activate mTORC1, we used the MEK inhibitor PD98059 and found that inhibition of the ERK pathway had no effect on arsenite-induced mTORC1 activation (Fig. 1A). In contrast, the p38α/β inhibitor SB203580 effectively inhibited arsenite-induced mTORC1 activation (Fig. 1A), suggesting involvement of the p38 pathway. To unambiguously determine the participation of the p38 pathway in arsenite-induced mTORC1 activation, and to determine which p38 group member plays the role, we analyzed arsenite-induced mTORC1 activation in p38α^−/−, p38β^−/−, p38γ^−/−, p38δ^−/−, and their corresponding p38α^+/+, p38β^+/+, p38γ^+/+, and p38δ^+/+ MEF cells. These cells were generated by in vitro treatment of the large T immortalized p38α^lox/lox, p38β^lox/lox, p38γ^lox/lox, and p38δ^lox/lox MEF cells with a lentivirus-expressing Cre recombinase or a control virus as described previously (35). We treated these cells with different doses of arsenite and found that only p38β deletion impaired arsenite-induced S6K1 phosphorylation (Fig. 1B). To confirm this phenotype of p38β^−/− MEF, we tested two pairs of primary wild-type and p38β^−/− MEF cells and observed a similar deficiency in arsenite-induced mTORC1 activation (supplemental Fig. S1B). siRNA knockdown of p38β in 293T cells also reduced arsenite-induced mTORC1 activation (supplemental Fig. S1C). The importance of p38β is not due to the abundance of p38β because its expression is not higher than other p38 isoforms in MEF and 293T cells (supplemental Fig. S1, D and E). Thus, p38β is involved in arsenite-induced mTORC1 activation.

It was a surprise to us that p38β plays a promoting role in arsenite-induced mTORC1 activation, because our recent study showed that the p38β-PRAK cascade negatively regulates mTORC1 in 2-deoxyglucose-treated cells (35). We did a side by side comparison and confirmed that p38β is required for 2-deoxyglucose-induced inactivation of mTORC1 and arsenite-induced activation of mTORC1 (supplemental Fig. S2). In contrast to p38β, PRAK is required for 2-deoxyglucose-induced inactivation of mTORC1 but has no role in arsenite-induced mTORC1 activation (supplemental Fig. S3, A and B). It appears that p38β targets different downstream proteins to execute different functions in 2-deoxyglucose- and arsenite-treated cells.

p38β Activation Is Selectively Involved in Arsenite-induced Activation of mTORC1

Activation of mTORC1 has been reported in a number of conditions. Replenishment of insulin, glucose, amino acid, or serum all activates mTORC1. We found that p38β was not required for mTORC1 activation under these conditions (Fig. 2, A–D). The protein synthesis inhibitor anisomycin can activate a variety of intracellular signaling pathways, including the p38 and mTORC1 pathways. It was shown that serum refurnishing or anisomycin treatment both activated MK2, which in turn phosphorylates TSC2 on Ser¹²¹⁰ and enables TSC2 binding to 14-3-3, and thus promotes mTORC1 activation (33). We found that, as in serum refurbishment, p38β has no role in anisomycin-induced S6K1 phosphorylation, excluding the possibility that p38β functions via MK2 to activate mTORC1 (Fig. 2E). Under serum-starved conditions, H₂O₂ can activate mTOR, and we found that p38β has no role in H₂O₂-induced mTOR activation (Fig. 2F). Collectively, of all the activation of mTORC1 tested, arsenite-induced mTORC1 activation is the only one that requires p38β.

It is known that arsenite activates all p38 group members, including p38β (7). Because of the similar size of p38 family members, there is currently no method available to specifically detect p38β activation. To determine whether p38β is activated in our experimental setting, we used transiently expressed FLAG-p38β to determine whether arsenite activates p38β. 293T cells expressing FLAG-p38β were treated with arsenite and the phosphorylation of FLAG-p38β was detected by Western blotting with anti-phospho-p38 antibodies (Fig. 3A). Arsenite-induced p38β phosphorylation and rapamycin had no effect on arsenite-induced p38β activation, suggesting that p38β activation is upstream of mTORC1. Ectopically expressing constitutively active MKK6 (MKK6b(E)) in cells can also lead to activation of all p38 group members (40). We tested whether expression of MKK6b(E) in MEF cells can mimic arsenite treatment and found a low level increase in rapamycin-sensitive S6K1 phosphorylation (Fig. 3B), and found this phosphorylation to be p38β- but not PRAK-dependent (Fig. 3C). It appears that overactivation of p38β alone is sufficient to increase mTORC1 activity, but the increase of mTORC1 activity is limited. This data are consistent with the results showing that p38β deletion or SB203580 treatment did not completely inhibit arsenite-induced mTORC1 activation (Fig. 1).

Arsenite-induced Activation of mTORC1 Is Independent from TSC2 and AMPK

Because AMPK and TSC2 are important regulators of mTORC1, we determined their role in arsenite-induced mTORC1 activation. To examine whether AMPK is involved in arsenite-induced activation of mTORC1, we used AMPKα1/α2 double knock-out MEFs, and found that arsenite-induced mTORC1 activation was not affected by AMPK deletion (Fig. 4A). We then used TSC2^−/− cells to determine the involvement of TSC2 and found that arsenite-induced mTORC1 activation was not affected by TSC2 deletion (Fig. 4B). As anticipated, arsenite-induced mTORC1 activity can be inhibited in TSC2^−/− cells by SB203580 and rapamycin (Fig. 4C). Thus, arsenite-induced mTORC1 activation is independent from AMPK and TSC2.

Arsenite Induces Interaction between p38β and Raptor

To elucidate the pathway of p38β-mediated mTORC1 activation, we coexpressed p38β with different components of the mTORC1 complex in 293T cells and treated the cells with or without arsenite. We immunoprecipitated HA-tagged Raptor, mLST8/GβL, Deptor, and Rheb, finding that p38β did not interact with any of these proteins in the absence of arsenite treatment, but interacted with Raptor but not mLST8/GβL, Deptor, or Rheb in the presence of arsenite stimulation (Fig. 5A). To test whether the interaction between p38β and Raptor is p38β-specific, we compared p38β with p38α in their interaction with Raptor and found that only p38β can interact with Raptor in arsenite-treated cells (Fig. 5B). We then checked whether arsenite treatment induces interaction between endogenous p38β and Raptor. 293T cells were treated with or without 1 mm arsenite for 30 min and p38β was immunoprecipitated with anti-p38β antibodies. We detected Raptor in the immunoprecipitates from arsenite-treated cells, but not in the immunoprecipitates from control non-treated cells (Fig. 5C). The specificity of p38β immunoprecipitation was demonstrated by the absence p38β and Raptor in the immunoprecipitates using control IgG (Fig. 5C). Because arsenite induces p38β activation, we examined whether the activity of p38β is required for its interaction with Raptor. We coexpressed Raptor with wild-type p38β or p38β inactive mutant (p38β(AF)) in which the regulatory phosphorylation sites TGY were mutated to AGF, and treated the cells with or without arsenite. Both p38β and p38β(AF) could co-immunoprecipitate with Raptor after the cells were treated with arsenite (Fig. 5D). Consistently, treatment of the cells with SB203580 also did not affect arsenite-induced interaction between Raptor and p38β (Fig. 5D). We also examined whether the MKK6-p38β module has any role in arsenite-induced p38β-Raptor interaction and found that MKK6 was not associated with the p38β-Raptor complex (Fig. 5E). Our data indicated that arsenite-induced interaction between p38β and Raptor does not require p38β activating phosphorylation and p38β activity.

p38β Phosphorylates Raptor

The inducible interaction between p38β and Raptor suggests that Raptor is a target of p38β in arsenite-treated cells. To explore the effect of p38β on Raptor, we coexpressed Raptor with p38β and included MKK6b(E) in some samples to activate p38β in 293T cells. We found that activation of p38β by coexpressed MKK6b(E) led to a band shift of Raptor, and this shift is at least partially due to phosphorylation, because treatment of the sample with CIAP reduced the band shift (Fig. 6A). To identify the phosphorylation site(s) we created a number of deletion mutants of Raptor and determined which fragment could be modified by MKK6b(E)-activated p38β. Western blotting analysis of Raptor mutants that coexpressed with MKK6b(E) and p38β in 293T cells revealed that the band shift of Raptor is primarily caused by a fragment of 741–887 amino acids (supplemental Fig. S4, A–C). The band shift of this fragment is due to phosphorylation, because treatment of immunoprecipitated Raptor with CIAP eliminated the band shift (supplemental Fig. S4D).

We mutated each of the serine and threonine sites (41) to alanine in the 741–887 region of Raptor and determined which sites are involved in the phosphorylation-mediated band shift. Converting Ser⁸⁶³ and Ser⁷⁷¹ to Ala both slightly reduced the band shift, and simultaneous mutation of Ser⁸⁶³ and Ser⁷⁷¹ completely abolished the shift (Fig. 6B). We then checked whether mutation of Ser⁷⁷¹ and Ser⁸⁶³ to Ala affected the MKK6b(E)/p38β-mediated shift of full-length Raptor and found that the S771A/S863A mutant has much less band shift in comparison with wild-type Raptor (Fig. 6C). Thus, Ser⁸⁶³ and Ser⁷⁷¹ constitute the major phosphorylation sites that are responsible for the p38β-induced band shift of Raptor.

We obtained anti-Ser⁸⁶³ phospho-Raptor and anti-Ser⁷⁷¹ phospho-Raptor antibodies. An in vitro kinase reaction using recombinant His-p38β and GST-Raptor-(741–887) was performed and the reaction mixtures were analyzed by immunoblotting using the anti-phospho-Raptor antibodies. Phosphorylation of Ser⁸⁶³ and Ser⁷⁷¹ were both detected in the products of the kinase reaction (Fig. 6D). The weak signal of Ser⁷⁷¹ phosphorylation is most likely due to the titer of the anti-Ser⁷⁷¹ phospho-Raptor being low, and we concentrated anti-Ser⁷⁷¹ antibody 40-fold by using an Amicon ultra filter for further studies. We determined that the concentrated antibody can specifically detect Ser⁷⁷¹ phospho-Raptor and the anti-Ser⁸⁶³ phospho-Raptor antibody can specifically detect Ser⁸⁶³ phospho-Raptor in arsenite-treated cells expressing HA-Raptor (supplemental Fig. S5A), although the titers of these antibodies are still not high enough to detect endogenous protein. By using these antibodies, we analyzed wild-type and p38β^−/− MEF cells that had infected HA-Raptor expressing lentivirus, and found that arsenite treatment induced Ser⁷⁷¹ and Ser⁸⁶³ phosphorylation of Raptor in wild-type but not p38β^−/− cells (Fig. 6E). The requirement of p38β in arsenite-induced Raptor phosphorylation was also observed when different doses of arsenite were used (supplemental Fig. S5B). Thus, arsenite induces p38β-dependent phosphorylation of Raptor on Ser⁷⁷¹ and Ser⁸⁶³.

Arsenite-induced Raptor Ser⁸⁶³ Phosphorylation Is mTORC1 Independent

Ser⁸⁶³ phosphorylation of Raptor was previously reported to be executed by mTORC1 during insulin stimulation (31). Here we need to determine whether mTORC1 has any role in arsenite-induced Ser⁸⁶³ phosphorylation of Raptor. To determine whether arsenite-induced phosphorylation of Raptor is mTORC1-dependent, we treated the cells with the mTORC1 inhibitor rapamycin. Rapamycin inhibited arsenite-induced S6K1 phosphorylation, but it did not block arsenite-induced Raptor Ser⁸⁶³ phosphorylation (Fig. 7A). Similar results were obtained in MKK6b(E)-expressing cells (Fig. 7B). We also evaluated whether inhibition of mTOR by overexpressing inactive mutant of mTOR (mTOR(KD)) in 293T cells can affect arsenite-induced phosphorylation of Raptor, and found that overexpression of FLAG-mTOR(KD) inhibited arsenite-induced S6K1 phosphorylation but not Ser⁸⁶³ and Ser⁷⁷¹ phosphorylation on Raptor (Fig. 7C). Collectively, these data demonstrated that arsenite-induced Raptor phosphorylation is independent from mTORC1 activity.

mTOR-mediated Ser⁸⁶³ phosphorylation was shown to enhance mTORC1 activity (31, 32). We examined whether S863A, S771A, and S863A/S771A mutations affect on the function of Raptor in cells treated with arsenite. We expressed wild-type and mutants of HA-Raptor in 293T cells and stimulated the cells with or without arsenite. The cells were lysed with CHAPS buffer, because in this buffer mTOR is still associated with Raptor (38). Raptor was immunoprecipitated with anti-HA antibodies and the Raptor-mTOR complexes were subjected to a kinase assay using recombinant GST-S6K1 as a substrate as described (39). Wild type-Raptor complex isolated from arsenite-treated cells had increased activity in the phosphorylation of S6K1 (Fig. 8A), which reflects the activation of mTOR by arsenite treatment. S771A, S863A, and S863A/S771A mutations all reduced arsenite-induced mTOR activity in the immunocomplex (Fig. 8A).

Because fusion of Raptor with the C-terminal 20-amino acid portion of Rheb creates a dominant-active Raptor (Raptor-Rheb(C20)), we mutated Ser⁸⁶³ and Ser⁷⁷¹ to Ala in this dominant-active Raptor and analyzed the effect of these mutations on the activity of Raptor-Rheb(C20). The S792A mutant was included as a control. Mutation of Ser⁸⁶³ or Ser⁷⁷¹, but not Ser⁷⁹², reduced Raptor-Rheb(C20)-induced increase of S6K1 phosphorylation (Fig. 8B).

Knockdown of Raptor and then reconstituting the cells with wild-type or phosphorylation site mutants of Raptor is an approach to evaluate the effect of the phosphorylation site mutations, and was used in our study. We designed a number of shRNA to target 3′ untranslated region (3-UTR) of Raptor and identified effective shRNAs. As shown in Fig. 8C, short hairpin Raptor 10 effectively reduced Raptor protein in 293T cells. We transfected expression plasmid containing the coding region of wild-type Raptor, S771A Raptor, S863A Raptor, or S771A/S863A Raptor into Raptor knockdown cells, and examined arsenite-induced phosphorylation of S6K1. The cells with the reconstitution of S771A Raptor, S863A Raptor, or S771A/S863A Raptor had a lower level of arsenite-induced phospho-S6K1 in comparison with the cells that were reconstituted with wild-type Raptor (Fig. 8D). Collectively, the data shown in Fig. 8 suggest the role of phosphorylation on Ser⁸⁶³ and Ser⁷⁷¹ of Raptor in arsenite-induced activation of mTORC1.