Figure 1.
SGK-1 inhibits FOXO3 transcriptional activity and subsequently LKB1 promoter activation.
(A) Illustration of the LKB1 promoter region, extending from nucleotide position -2537 to +727 relative to the transcription start site. Positions of the transcription start site (TSS) and binding sites of transcription factors involved in LKB1 regulation (FOXO3, NF-Y and Sp1) as well as their positions relative to the transcription start site are indicated in brackets. (B) LKB1 reporter activity in HEK293 cells after ectopic expression of SGK-1 (dark grey) is compared to the activity obtained from co-transfections with the empty vector (Ctrl, light grey). Reporter activity is expressed as fold of luciferase activity (relative light units normalized to renilla luciferase activity) obtained from co-transfection of the plasmid containing the LKB1 promoter (nucleotides −2537 to +727) together with the empty expression vector (pcDNA3). Instead of the empty vector the same amount of either a FOXO3 wild-type expression plasmid (FOXO3 WT) or a SGK-1 phosphorylation site deficient triple mutant (T32A/S253A/S315A) of FOXO3 (FOXO3 A3) has been co-transfected. Each bar represents the means ± standard deviation of three measurements. (C) Western blot analysis of the same lysates used in the LKB1 reporter assay described in (B) using antibodies against phospho-FOXO3 (T32), total FOXO3 and actin. The ratios of p-FOXO3/FOXO3 for each experimental condition are indicated. The shown blot is a representative of three independent experiments.
Figure 2.
Glucocorticoid-mediated SGK-1 induction inhibits LKB1 transcription through FOXO3 (T32) phosphorylation.
MCF-10A cells were treated with either vehicle (ethanol) (Ctrl), 1 µM dexamethasone (DEX) or a combination of 1 µM dexamethasone and 1 µM of the GR antagonist RU-486 (DEX/RU) for 0, 6, 24, and 48 h. (A) Relative mRNA levels of SGK-1, FOXO3 target genes (LKB1, TRAIL, IGFBP3) and the transcription factors involved in LKB1 regulation (Sp1, NF-YB, FOXO3). β-tubulin was used as internal control. RT-PCR products were analysed in a 1% agarose gel after ethidium bromide staining. (B) Relative protein levels of phospho-FOXO3 (T32), total FOXO3 and actin as internal control were analysed by western blotting. The ratios of p-FOXO3/FOXO3 for each experimental condition are indicated.
Figure 3.
Glucocorticoid receptor directly binds and activates the FOXO3 promoter.
(A) Illustration of the FOXO3 promoter region, extending from nucleotide position −4255 to +91 relative to the transcription start site, matched with deletion constructs generated by the indicated restriction enzymes. The 5′-extension relative to the transcription start site of each construct in base pairs (bp) is given on the right side (full length: −4255 to NotI: −11 respectively). The 3′-end at +91 is identical for each construct. The position of the TATA box, the transcription start site (TSS) and the potential glucocorticoid response elements (GRE 1, 2, 3) are indicated. The PCR products amplified in the ChIP assay (ChIP A, B, C) as well as the sequence of the potential GREs are marked above the illustration. (B) Not I restriction digestion of all constructed FOXO3 promoter deletion constructs separated in a 1% agarose gel. (C) FOXO3 promoter reporter assay in A549 cells treated with either vehicle (Ctrl, dark grey) or 1 µM dexamethasone (DEX, light grey) for 12 h. Luciferase activity of all FOXO3 promoter deletion constructs (relative light units normalized against renilla luciferase activity) is expressed as fold of the signal obtained with the plasmid containing the full length promoter (−4255 to +91) without dexamethasone treatment. Assays were performed in triplicates. The error bars denote mean ± standard deviation. (D) ChIP assay in MCF-10A cells treated with either 1 µM dexamethasone (DEX) or vehicle (Ctrl) for 3 h. DNA was sonicated to an average size of <500 bp and run on a 1% (wt/vol) agarose gel (upper panel). PCR products specific for the different potential GREs within the FOXO3 promoter (ChIP A, B, C) were amplified from sonicated DNA of 1/10 of the starting material (Input, positive control) as well as from sonicated DNA after ChIP with either a non-specific antibody (IgG, negative control) or with an antibody against the glucocorticoid receptor (α-GR). Products were visualized within the linear range of the reaction by ethidium bromide staining on a 1% (wt/vol) agarose gel (lower panel). The figure shows a representative of three independent experiments.
Figure 4.
Transcriptional activation of Foxo3 by glucocorticoids in mice.
Semi-quantitative (left panel) and quantitative real-time PCR analysis (right panel) of Foxo3 mRNA levels in livers of C57Bl6 mice treated with dexamethasone (DEX) (1 mg/kg/day) or saline (Ctrl) for a period of (A) 3 h or (B) 3 weeks. Bars represents the means ± standard deviation (n: number of animals; panel A, n = 6; panel B, n = 7).
Figure 5.
FOXO3 coordinates LKB1 gene expression depending on the metabolic status of the cell.
MCF-10A cells were treated with AMPK activating stimuli (1 mM AICAR, 50 mM 2- Deoxyglucose or 2 µM Oligomycin) in combination either with vehicle (Ctrl), 1 µM dexamethasone (DEX) or a mixture of 1 µM dexamethasone and 1 µM RU-486 (DEX/RU) for 18 hours. (A) Relative mRNA levels of SGK-1, LKB1, FOXO3 and GAPDH as internal control were analysed by semi-quantitative RT-PCR and separated in a 1% agarose gel after ethidium bromide staining. (B) Relative mRNA levels of LKB1 and FOXO3 were analysed by quantitative real time PCR and normalized to GAPDH (mean ±SD, n = 3). (C) Western blot analyses of the relative protein levels of phospho-AMPK (T172), phospho-FOXO3 (S413), phospho-FOXO3 (T32), total FOXO3 in comparison to actin as internal control. The ratios of p-FOXO3/FOXO3 for each experimental condition are indicated. The data shown is representative of three independent experiments. (D) Schematic illustration underlying the glucocorticoid-dependent modulation of FOXO3 transcriptional activity.
Figure 6.
Glucocorticoid-induced transcriptional activation of FOXO3 is further enhanced by AMPK activation.
(A) MCF-10A cells were treated with 1 mM of specific AMPK activator AICAR in combination with either vehicle (Ctrl), 1 µM dexamethasone (DEX) or a mixture of 1 µM dexamethasone and 1 µM RU-486 (DEX/RU) for 3, 6, 12 and 24 h. (B) MCF-10A cells were treated with vehicle (Ctrl), 1 µM dexamethasone (DEX) or a combination of 1 µM dexamethasone and 1 mM AICAR (DEX+AICAR) in the presence or absence of 10 µg/ml cycloheximide (CHX) for 6 h. Relative mRNA levels of FOXO3 were analysed by quantitative real-time PCR (mean ±SD, n = 3). Relative protein levels of FOXO3 in comparison to actin as internal control were analysed by western blotting in the presence or absence of cycloheximide.
Figure 7.
FOXO3 binds to its own promoter.
(A) Illustration of the FOXO3 promoter region extending from nucleotide position -2500 to +91 relative to the transcription start site. Positions of the TATA box, the transcription start site (TSS) and potential FOXO3 binding sites (FOXO A-H) as well as their sequences are indicated. (B) 32P-labeled double-stranded oligonucleotides containing the FOXO3 binding sites of the LKB1 promoter (LKB1 FOXO) or the different potential FOXO3 binding sites of the FOXO3 promoter (FOXO3 FOXO A-H) were incubated with 25 ng of recombinant GST-tagged FOXO3 protein, separated in a 4% polyacrylamide gel and visualized by autoradiography after 6 h (upper panel) or 20 min (lower panel) exposure. Formation of sequence specific protein complexes was confirmed by competition with a 500-fold molar excess of unlabeled oligos containing either the wild-type (Wt) or the mutant (Mut) FOXO3 site of the LKB1 promoter. Complexes containing either one (1x FOXO3) or two FOXO3 complexes (2x FOXO3) are marked. (C) Complex formation between GST-FOXO3 and 32P-labeled double-stranded oligonucleotides containing the FOXO3 binding sites of the LKB1 promoter (LKB1 FOXO), the wild-type E2-binding site of the FOXO3 promoter (FOXO3 E2 Wt). Addition of a GST specific antibody (α-GST, lanes 2 and 4) confirms that the composition of protein-DNA complexes consists of FOXO3 molecules. (D) Endogenous complex formation between 4 µg of nuclear extracts from MCF-10A cells and the FOXO3 E2 wild-type site in the presence of unlabeled wild-type E2 (Wt, lane 2) or mutant E2 (Mut, lane 3) competitor. Addition of a FOXO3 specific antibody (α-FOXO3, lane 5) confirms that the composition of protein-DNA complexes consists of endogenous FOXO3 molecules.
Table 1.
Oligonucleotides used in EMSA.
Figure 8.
FOXO3 activates its own promoter via a positive autoregulatory feedback loop in combination with GR.
(A) FOXO3 reporter activity in A549 cells after treatment with either vehicle (Ctrl, light grey) or 1 µM dexamethasone (DEX, dark grey) for 12 h. Reporter activity is expressed as fold of luciferase activity (relative light units normalized to renilla luciferase activity) obtained from co-transfection of the plasmid containing the full length FOXO3 promoter (nucleotides −4255 to +91) together with the empty expression vector (pcDNA3) without dexamethasone treatment. Instead of the empty vector the same amount of either a FOXO3 wild-type expression plasmid (FOXO3 WT) or a SGK-1 phosphorylation site deficient triple mutant (T32A/S253A/S315A) of FOXO3 (FOXO3 A3) was co-transfected. Each bar represents the means ± standard deviation of a triplicate experiment. (B) Quantitative real-time PCR analysis of FOXO3 mRNA levels in MCF-10A cells which were starved of growth factors in medium containing 2% of charcoal stripped FBS for 72 h (Ctrl), transferred into normal growth medium containing 10 µg/ml human insulin and 5% horse serum and stimulated with 0.5 µg/ml hydrocortisone (HC+Serum) for 6, 24 and 48 h, in the presence or absence of 50 µM of the specific PI3-kinase inhibitor LY294002.
Figure 9.
AMPK is essential for FOXO3 induced activation of LKB1 and FOXO3 transcription.
(A) MCF-10A cells were either treated with vehicle (Ctrl) and 1 µM dexamethasone (DEX) or in combination with either 1 mM AICAR (left panel) or 50 mM 2-deoxyglucose (right panel) in the presence or absence of 20 µM of the specific AMPK inhibitor “Compound C” for 18 h. Relative mRNA levels of LKB1, FOXO3 and GAPDH as internal control were analysed by semi-quantitative RT-PCR in a 1% agarose gel after ethidium bromide staining. The data shown is representative of three independent experiments. (B) Illustration of metabolic regulation of FOXO3’s transcriptional activity and subsequent effects on FOXO3 and LKB1 expression. Activating events are indicated in green colour, inhibiting events in red. High glucocorticoid levels immediately activate FOXO3 and SGK-1. Following stimulation of FOXO3 transcription, the protein is phosphorylated at threonine 32 (Thr 32), leading to its inactivation. Treatment with AMPK activating stimuli (metabolic stress) triggers phosphorylation of FOXO3 by AMPK at position 413 of the serine residue (Ser 413), thereby counteracting inactivation of FOXO3 by SGK-1. This results in a temporally delayed further enhancement of FOXO3 transcription since the FOXO3 protein can bind and activate its own gene promoter via a positive autoregulatory feedback loop in the presence of glucocorticoids. Functional FOXO3 in turn induces the master upstream kinase that further actives AMPK by threonine phosphorylation at position 172 (Thr 172). Restored ATP levels or inhibition of AMPK by compound C interrupts this circuit.