p53 target genes sestrin1 and sestrin2 connect genotoxic stress and mTOR signaling

Abstract

The tumor suppressor p53 is activated upon genotoxic and oxidative stress and in turn inhibits cell proliferation and growth through induction of specific target genes. Cell growth is positively regulated by mTOR, whose activity is inhibited by the TSC1:TSC2 complex. Although genotoxic stress has been suggested to inhibit mTOR via p53-mediated activation of mTOR inhibitors, the precise mechanism of this link was unknown. We now demonstrate that the products of two p53 target genes, Sestrin1 and Sestrin2, activate the AMP-responsive protein kinase (AMPK) and target it to phosphorylate TSC2 and stimulate its GAP activity, thereby inhibiting mTOR. Correspondingly, Sestrin2-deficient mice fail to inhibit mTOR signaling upon genotoxic challenge. Sestrin1 and Sestrin2 therefore provide an important link between genotoxic stress, p53 and the mTOR signaling pathway.

Figure 1. Sesn1 and Sesn2 inhibit mTOR signaling

(A) Ectopic Sesn1, Sesn2 or TSC1:TSC2 expression decreases p70S6K phosphorylation. HEK293 cells were co-transfected with HA-p70S6K together with either Sesn1F, Sesn2F, TSC1F plus TSC2F or GFP expression vectors. After 48 hrs HA-p70S6K was immunoprecipitated (IP) and its phosphorylation was examined by immunoblotting. Sesn1/2 and TSC1/2 expression in total lysates was examined by immunoblotting (IB). (B) Induction of Sesn2 inhibits mTOR signaling. Sesn2F was induced in MCF7-tet OFF Sesn2F cells by doxycycline (Dox) removal. Control cultures were left with doxycycline. After 22 hrs, rapamycin (Rapa; 20 ng/ml) was added to some of the cultures. Cell lysates were prepared 2 hrs later and expression and phosphorylation of the indicated mTOR pathway components was examined by immunoblotting. (C) Sesn2 redox activity is not required for inhibition of mTOR signaling. H1299 cells were infected with wt or mutant Sesn2F-lentiviruses as well as GFP- and Prx1-lentiviruses and analyzed for phosphorylation and expression of the indicated proteins as above. (D) Sesn2 increases association of 4E-BP1 and eIF-4E with m⁷G-RNA Cap structures. Sesn2 was induced as in B and cell lysates were incubated with m⁷G-Sepharose beads for 6 hrs and bead-bound proteins were examined by immunoblotting (upper panel). Expression levels and phosphorylation state of the indicated proteins in the total lysates were analyzed as in B (lower panel). (E) Sesn2 induction inhibits cyclin D1 and c-Myc protein expression, but does not affect their mRNAs. Sesn2F was induced by doxycycline withdrawal and expression of cyclin D1 and c-Myc was examined 24 hrs later by immunoblotting (IB). RNA from the same cells was analyzed by Northern blotting (NB) with cyclin D1 or c-Myc probes. 18S RNA was used as a loading control.

Figure 2. Sestrins inhibit mTOR signaling upstream to the TSC1:TSC2 complex

(A) Rheb abrogates Sesn2-induced inhibition of p70S6K phosphorylation. HEK293 cells were co-transfected with HA-p70S6K and either Sesn2 or GFP plasmids together with either myristoylated (myr)-AKT, H-Ras or Rheb constructs or were treated without (Con.) or with insulin for 1 hr as indicated. HA-p70S6K was immunoprecipitated (IP) with anti-HA antibody and its phosphorylation was examined by immunoblotting. Total cell lysates were immunoblotted (IB) with the indicated antibodies. (B) Sesn2 expression decreases Rheb GTP loading. Sesn2 was induced in MCF7-tet OFF Sesn2F cells for 24 hrs and cells were metabolically labeled with ³²P for 4 hrs. Rheb was immunoprecipitated, guanine nucleotides were extracted, separated by thin layer chromatography, autoradiographed and quantified by QuantityOne software. Rheb recovery was examined by immunoblotting and used to normalize the calculated GTP/GDP ratio. The values shown are averages +/− S.D. of 3 experiments. (C) TSC2 is required for down-regulation of p70S6K phosphorylation by Sesn1/2. Tsc2^+/+ and Tsc2^−/− fibroblasts were co-transfected with HA-p70S6K together with Sesn1, Sesn2 or GFP expression vectors. HA-p70S6K was immunoprecipitated (IP) and its phosphorylation examined by immunoblotting. Sesn1/2 expression in total lysates was examined by immunoblotting (IB). NS - non-specific band. (D) Inhibition of mTOR signaling by Sesn2 requires TSC2. MCF7-tet OFF Sesn2F cells were infected with shTSC2- or shLuc-lentiviruses. After 48 hrs Sesn2F was induced by doxycycline removal and protein phosphorylation and expression was examined 24 hrs later.

Figure 3. Sestrin inhibit p70S6K phosphorylation via an AMPK-dependent mechanism

(A) Sesn2 expression enhances AMPK phosphorylation. Sesn2F was induced in MCF7-tet OFF Sesn2F cells and phosphorylation of the indicated proteins was examined 24 hrs later. (B) Compound C abrogates Sestrin-induced inhibition of p70S6K phosphorylation. Sesn2F was induced in MCF7-tet OFF Sesn2F cells in the presence of compound C (CompC) or vehicle control (DMSO) and protein phosphorylation was examined 24 hrs later. (C) Knock-down of AMPKα1 abrogates Sesn2 inhibition of p70S6K and S6 phosphorylation. MCF7-tet OFF Sesn2F cells were infected with shAMPKα1- or shLuc-lentiviruses. Protein phosphorylation and expression were analyzed by immunoblotting 72 hrs after infection and 24 hrs after doxycycline removal.

Figure 4. Sestrins interact with TSC1:TSC2 and AMPK

(A) Co-immunoprecipitation of TSC1, TSC2 and AMPKα2 with Sesn2. HEK293 cells were co-transfected with Sesn2F or GFP expression vectors along with tagged TSC1, TSC2, TSC1 plus TSC2, or AMPKα2 plasmids as indicated. Sesn2F was immunoprecipitated with anti-Flag and presence of the indicated proteins in the immunecomplexes (IP) and total lysates (IB) was examined by immunoblotting. (B) Co-immunoprecipitation of endogenous TSC1, TSC2 and AMPKα1 with Sesn2. Sesn2F or GFP were induced in MCF7-tet OFF Sesn2F or GFP cells by incubation in low doxycycline concentration (0.01 µg/ml). After 24 hrs the cells were lysed. Sesn2F was immunoprecipitated with anti-Flag and presence of the indicated proteins in the immunoprecipitates (IP) and original lysates (IB) was examined. For comparison, MCF7 cells were incubated with camptothecin (Camp) for 12 hrs to induce Sesn2 expression that was examined by immunoblotting of the same amount of cell lysates as above. (C) Sesn2 co-elutes with TSC1, TSC2 and AMPKα in high molecular weight fractions. Extracts of H1299 cells were separated on a Superdex 200 gel filtration column and analyzed by immunoblotting. The elution positions (in kDa) of molecular weight standards are indicated at the top. (D) Co-immunoprecipitation of TSC1, TSC2 and AMPKα with endogenous Sesn2. Two months old wt male mice were injected with DEN (100 mg/kg). After 24 hrs, the liver was removed, homogenized and filtered. The lysate was separated by gel filtration as above and the high molecular weight fractions were combined and immunoprecipitated with anti-Sesn2 or control anti-rabbit (Con.) IgG. Presence of the indicated proteins in the immuneprecipitates was analyzed by immunoblotting.

Figure 5. Sesn2 activates AMPK and induces TSC2 phosphorylation

(A) Wt, but not N-terminally truncated, Sesn2 activates AMPK. HEK293 cells were cotransfected with AMPKα1F expression vector along with GFP, Sesn2F and Sesn2F-ΔN expression vectors. After 48 hrs the cells were lysed and AMPKα1F immune complexes were isolated and assayed for their ability to phosphorylate a TSC1:TSC2 complex isolated by immunoprecipitation from lysates of transiently transfected HEK293 cells. The amounts of AMPKα1F, Sesn2 and TSC2 in the kinase reaction were examined by immunoblotting (IB). (B) Sesn2 enhances TSC2 phosphorylation. HEK293 cells were co-transfected with TSC1Myc and TSC2F expression vectors along with either wt Sesn2F, N-terminally truncated Sesn2F or GFP expression vectors. After 48 hrs cells were metabolically labeled with ³²P-orthophosphate and the TSC1:TSC2 complex was immunoprecipitated 4 hrs later, gel separated and autoradiographed. TSC2 and Sesn2 expression were examined by immunoblotting (IB). (C) Sesn2-associated AMPKα is more extensively phosphorylated than AMPK that is not associated with Sesn2. Sesn2F or GFP were induced by doxycycline withdrawl in MCF7-tet cells. After 24 hrs cells were lysed and the lysates were subjected to immunoprecipitation with anti-Flag antibody. The supernatant of the immunoprecipitation reaction, the immune complexes and the total infractionated lysate were examined for AMPKα phosphorylation and content by immunoblotting.

Figure 6. Sesn1/2 are required for negative regulation of mTOR under basal and stressed conditions

(A) Sesn1 negatively regulates mTOR signaling in mouse fibroblasts. 3T3 cells were infected with shLuciferase (shLuc)-, shSesn1-1- or shSesn1-2-lentiviruses, selected with puromycin and analyzed 48 hrs later for protein phosphorylation (IB) and Sesn1 mRNA expression by Northern blotting (NB). 18S RNA was used to control for RNA loading. (B) Cells (30,000) from (A) were plated onto 6 well plates and counted every 24 hrs for 6 days. (C) Sestrins are required for p53-induced mTOR inhibition. H1299 cells were infected with shLuciferase (shLuc)-, shSesn1- or shSesn2- lentiviruses and 48 hrs later were infected with p53- or GFP-lentiviral vectors. After 48 additional hrs the cells were lysed and analyzed by immunoblotting (IB) for p70S6K and S6 phosphorylation and p53 and Sesn2 expression. Sesn1 expression was examined by Northern blotting (NB). (D) Sesn1/2 are required for down-regulation of S6 phosphorylation in Nutlin-3 treated cells. U2OS cells were infected with shLuciferase (shLuc)-, shSesn1-, shSesn2- or shp53-lentiviruses. After 48 hrs the cells were incubated or not with Nutlin-3 (6 µM) for 12 hrs as indicated to induce p53 accumulation. Expression of p53, Sesn2, and p21^Waf1 and S6 phosphorylation were examined by immunoblotting (IB). Sesn1 mRNA was analyzed by Northern blotting (NB). (E) Sesn2 and p53 are required for inhibition of mTOR signaling during camptothecin-induced genotoxic stress. 3T3 cells of the indicated genotypes were treated with camptothecin (Camp, 20 µM) for 12 hrs and protein expression and phosphorylation were analyzed by immunoblotting. (F) Sesn2 is required for DEN-induced inhibition of S6 phosphorylation in mouse liver. Two months old Sesn2^+/− and Sesn2^−/− male mice were injected with DEN (100 mg/kg). After 24 hrs, protein expression and phosphorylation in liver homogenates were analyzed as above.

Figure 7. Sestrins inhibit cell growth and proliferation via mTOR

(A,B) Sesn1/2 expression decreases cell size and inhibits cell proliferation at G1. HEK293 cells were infected with Sesn1-, Sesn2- or GFP-lentiviruses. After 48 hrs cell size (A) and cell cycle distribution (B) were analyzed by flow cytometry. (C, D) Sesn2 expression decreases cell size and inhibits clonogenic growth in Tsc2^+/+ but not Tsc2^−/− cells. Tsc2^+/+ and Tsc2^−/− cells were infected with Sesn2- or GFP-lentiviruses. After 48 hrs cell size (C) was analyzed by flow cytometry. Colony formation (D) was analyzed 14 days after infection. (E) Sesn2 inhibits clonogenic growth upstream to Rheb. MCF7 cells were transfected with Sesn2 or GFP expression vectors along with Rheb, H-Ras or an empty expression vector. The cells were selected in puromycin-containing medium and colony formation was quantitated after 3 weeks. Results in (A, B, C, D, E) are averages ±S.E. * − p ≤ 0.05 by Student’s t test. (F) A model explaining our results. Sesn1/2 are induced upon stress-triggered p53 activation. Sesn1/2 dimers interact with TSC1:TSC2 and AMPKα and activate AMPK by induced proximity-dependent autophosphorylation. This results in TSC2 phosphorylation and stimulation of its GAP activity leading to inhibition of mTOR signaling, which is positively regulated by growth factors via Ras and PI3K.