Serum- and Glucocorticoid-Inducible Kinase 1 (SGK1) Regulates Adipocyte Differentiation via Forkhead Box O1

Abstract

The serum and glucocorticoid-inducible kinase 1 (SGK1) is an inducible kinase the physiological function of which has been characterized primarily in the kidney. Here we show that SGK1 is expressed in white adipose tissue and that its levels are induced in the conversion of preadipocytes into fat cells. Adipocyte differentiation is significantly diminished via small interfering RNA inhibition of endogenous SGK1 expression, whereas ectopic expression of SGK1 in mesenchymal precursor cells promotes adipogenesis. The SGK1-mediated phenotypic effects on differentiation parallel changes in the mRNA levels for critical regulators and markers of adipogenesis, such as peroxisome proliferator-activated receptor γ, CCAAT enhancer binding protein α, and fatty acid binding protein aP2. We demonstrate that SGK1 affects differentiation by direct phosphorylation of Foxo1, thereby changing its cellular localization from the nucleus to the cytosol. In addition we show that SGK1−/− cells are unable to relocalize Foxo1 to the cytosol in response to dexamethasone. Together these results show that SGK1 influences adipocyte differentiation by regulating Foxo1 phosphorylation and reveal a potentially important function for this kinase in the control of fat mass and function.

Imbalance between energy intake and its expenditure leads to obesity (1, 2), which is a major risk factor for development of type 2 diabetes and cardiovascular disease, part of a spectrum of diseases called “the metabolic syndrome” (3, 4). The efficient trafficking of calories in states of high or low food intake is orchestrated across multiple tissues by endocrine molecules and through signal transduction pathways that weaken or reinforce a number of key transcriptional effectors.

Peroxisome proliferator-activated receptor (PPAR)γ and CCAAT enhancer binding protein (C/EBP)α are the principal regulators of the transcriptional development of adipose tissue and control the conversion of undifferentiated mesenchymal cells into adipocytes. Studies involving gain- or loss-of-function of these factors have definitively demonstrated their necessity in promoting full adipogenesis (5–9). In addition to these two central players, a number of other factors have been described as contributing either positively or negatively to the adipogenic process (10). Among these, several forkhead winged helix family members have been shown to be important transcriptional integrators of developmental and metabolic processes (11). In particular, Foxo1, Foxa1 and 2, and Foxc2 counter the conversion of preadipocytes into adipocytes through distinct mechanisms involving repression of PPARγ’s gene transcription, activation of Pref1, or inhibition of PPARγ’s ability to activate its target genes (12–16).

The transcriptional regulation of adipogenesis is influenced by endocrine cues such as insulin and glucocorticoids, which activate intracellular pathways to increase messenger molecule levels, such as cAMP, or signaling kinases, such as those belonging to the AGC family. These kinases include arginine, serine kinase (RSK), ribosomal protein S6 kinase (S6K), MAPKs, protein kinase B/Akt, protein kinase C (PKC), and cAMP/GMP-dependent kinases and function to phosphorylate serine or threonine residues present at R-XRXXS/T consensus motifs. Gain- or loss-of-function studies implicate a unique role in adipogenesis for several of these AGC kinases. In particular, Akt, when overexpressed, increases differentiation in 3T3-L1 cells, whereas its absence reduces adipogenesis in primary fibroblasts (17). Furthermore, PKC isoforms β-I and ε promote differentiation (18), whereas PKC δ opposes the adipogenic process (19).

The serum- and glucocorticoid-inducible protein kinase 1 (SGK1) is the newest member of the AGC kinases, initially characterized in rat mammary tumor cells as an immediate early gene induced by serum and glucocorticoids (20). Although SGK1 expression is present in a variety of tissues, the physiological function of this kinase has been studied primarily in the kidney where it controls sodium reabsorption via regulation of the epithelial sodium channel ENaCα (21, 22). Mice lacking SGK1 are viable, confirming that this kinase is not required for survival (23). However, SGK1−/− mice show a reduced ability to retain salt after challenge by a salt-deficient diet, highlighting a role for SGK1 in salt metabolism (23, 24).

Dexamethasone priming of preadipocytes is necessary for their adipogenic commitment and differentiation (25), but the mechanisms through which glucocorticoids stimulate adipogenesis have not been fully elucidated. Therefore, we investigated the potential role of SGK1 as a mediator of glucocorticoid-initiated signals in adipocytes. Our results show that treatment with dexamethasone induces SGK1 during early adipogenesis. Ectopic expression of SGK1 enhances adipocyte differentiation, whereas its down-regulation via small interfering RNA (siRNA) reduces it. Furthermore, SGK1 directly phosphorylates Foxo1, leading to its relocalization to the cytosol. Dexamethasone promotion of adipogenesis appears to be mediated through SGK1 effects on Foxo1 because dexamethasone is unable to influence Foxo1’s subcellular localization in SGK1−/− cells.

Results

SGK1 is induced during adipocyte differentiation

In the last few years it has been shown that SGK1 plays an important role in metabolism, specifically in the kidney (21). Given the function of SGK1 as a downstream effector of glucocorticoid signaling (26), we assessed the potential function of this kinase in adipose tissue. To determine whether SGK1 is expressed in murine adipose tissues and to quantify its levels in this tissue relative to other organs, we analyzed the SGK1 mRNA levels by real-time PCR, in metabolic tissues. As shown in Fig. 1A, the highest level of SGK1 expression was identified in the kidney. Interestingly, SGK1 RNA levels were detected also in adipose tissue, with white fat tissue expressing more than brown fat. To investigate whether SGK1 mRNA is regulated during adipogenesis in vitro, we analyzed SGK1 gene expression during different stages of differentiation in 3T3-L1 or 10T1/2 cells. The results shown in Fig. 1B indicate that SGK1 was induced during adipocyte differentiation in both cell lines. The levels of PPARγ, C/EBPα, and aP2 were measured to assess the extent of differentiation achieved in those cells (supplemental Fig. 1A published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). Interestingly, SGK1 mRNA levels were increased within a few hours from the induction of preadipocytes with the standard differentiation cocktail containing dexamethasone, insulin, and isobutylmethylxanthine (MDI). Conversely, cells induced to differentiate in the presence of insulin and PPARγ ligands showed an increase of the expression of SGK1 only at late stages. To determine which component of the differentiation mixture was responsible for the early induction of mRNA SGK1, we analyzed SGK1 levels in preadipocytes treated only with one of the three inducers. As shown in Fig. 1C, SGK1 mRNA levels were induced 3-fold by dexamethasone within 3 h of treatment. Insulin and 3-isobutyl-1-methylxanthine appeared to be able to increase SGK1 mRNA levels, but this induction occurred only after 6 h, in the case of insulin. Both dexamethasone and insulin increased the levels of total and phospho-SGK1 protein (supplemental Fig. 2). Because it is known that PPARγ can induce SGK1 expression in human kidney cells (27), we assessed whether PPARγ plays a role also in dexamethasone-mediated induction of SGK1. We therefore measured the induction of SGK1 in PPARγ−/− cells treated with MDI. The data in supplemental Fig. 3 show that SGK1 expression is induced even in the absence of PPARγ. Because dexamethasone has been shown to induce C/EBPδ during adipogenesis, we determined whether SGK1 could play a role in this process. As shown in supplemental Fig. 4, C/EBPδ was reduced when SGK1 is knocked down in differentiated 10T1/2, suggesting that SGK1 could play a potential role in mediating C/EBPδ induction during differentiation.

Fig. 1.

SGK1 is expressed in fat tissue and induced during adipogenesis. A, SGK1 mRNA levels in tissues obtained from C57BL/6J 8-wk-old male mice. Relative mRNA levels were determined by real-time PCR and normalized using 18S. B, mRNA levels of SGK1 were evaluated during differentiation (0–6 d) after MDI induction or after treatment with rosiglitazone and insulin (3T3-L1 cells) or after MDI or troglitazone and insulin treatment (10T1/2 cells) (*,#, P < 0.05; **,##, P < 0.01; ***,###, P < 0.001; ns, nonsignificant). C, SGK1 mRNA levels in 3T3-L1 after stimulation with a specific inducer (*, P < 0.05; **, P < 0.01, ***, P < 0.001; ns, nonsignificant). BAT, Brown adipose tissue; IBMX, isobutylmethylxanthine; WAT, white adipose tissue.

SGK1 affects adipogenesis in vitro

To investigate the potential role of SGK1 in adipogenesis, we ectopically expressed or silenced SGK1 in preadipocytes and measured the effects of the modulation of SGK1 on differentiation, both at the phenotypical and molecular level. Constitutively active SGK1 (^S422DSGK1) ectopically expressed in 10T1/2 cells (Fig. 2A) led to an increase in the number of lipid-storing cells, as shown by Oil Red O staining (Fig. 2B). The mRNA levels of markers of adipogenesis such as PPARγ, C/EBPα, and aP2 were significantly induced in ^S422DSGK1-expressing cells, indicating that this kinase plays a positive role in differentiation (Fig. 2C). In contrast, decreasing endogenous SGK1 levels by RNA interference (Fig. 2D and supplemental Fig. 5, A and B) led to diminished differentiation, as measured phenotypically by Oil Red O staining (Fig. 2E). SGK1 knockdown in preadipocytes cells induced to differentiate with either MDI or rosiglitazone led to a significant decrease in mRNA levels for markers of adipogenesis (Fig. 2F; supplemental Fig. 5, C and D; and supplemental Table I). These data suggest that modulation of SGK1 levels can affect adipogenesis.

Fig. 2.

SGK1 modulates adipogenesis. A, SGK1 mRNA levels were measured by RT-PCR in constitutively active SGK1 (^S422DSGK1)-expressing cells and compared with control vector cells (***, P < 0.001). B, Oil Red O staining of 10T1/2 ectopically expressing ^S422DSGK1 after 6 d of differentiation showed increased lipid accumulation. C, Ectopic expression of ^S422DSGK1 in 10T1/2 enhanced PPARγ, CEBPα, and aP2 mRNA levels in comparison to vector during MDI-induced differentiation. (***, P < 0.001). D, Analysis of SGK1 mRNA levels in 10T1/2 cells expressing either control siRNA or si-SGK1 (***, P < 0.001). E, Oil Red O staining of 10T1/2 cells expressing si-SGK1 showed decreased lipid accumulation compared with si-ctl expressing cells at d 6 of differentiation. F, PPARγ, CEBPα, and aP2 mRNA levels were reduced in SGK1 knockdown 10T1/2 cells compared with control-expressing cells induced to differentiate with MDI (***, P < 0.001). si-ctl, Small interfering control.

SGK1 phosphorylates Foxo1 in vitro and in vivo

It has been shown that Foxo1 acts as a negative regulator of adipogenesis (12, 28). To determine the mechanisms through which SGK1 affects adipocyte differentiation, we tested whether SGK1 could phosphorylate Foxo1. The in vitro protein kinase assay (Fig. 3A) shows that full-length glutathione-S-transferase (GST)-Foxo1 was directly phosphorylated by SGK1, whereas a GST control protein was not. To determine whether SGK1 phosphorylation of Foxo1 occurs in vivo, we cotransfected Foxo1 and SGK1 in U2OS cells and assessed the phosphorylation state of exogenously expressed Foxo1, using antibodies recognizing Foxo1 only when phosphorylated specifically at threonine 24 (T24), or serine 256 (S256) or serine 316 (S319) residues. As shown in Fig. 3B, in unstimulated conditions, Foxo1 showed a basal level of phosphorylation at all three sites (Fig. 3B, lane 3). However, when constitutively activated ^S422DSGK1 was cotransfected with Foxo1, phosphorylation was increased in vivo at these specific threonine and serine residues (Fig. 3B, lane 4). A triple alanine mutant of Foxo1 (Foxo1AAA) at T24, S256, and S316 sites could not be phosphorylated when coexpressed with the constitutively active form of ^S422DSGK1 (Fig. 3B, lane 6). Conversely, inactive SGK1 (^K127NSGK1) did not affect the phosphorylation levels of Foxo1 at any of these sites (Fig. 3B, lane 8), further supporting the notion that SGK1 can directly phosphorylate these residues in Foxo1. To assess whether SGK1 is able to phosphorylate endogenous Foxo1 in 10T1/2 cells, the constitutively active or the inactive forms of SGK1 (^S422DSGK1 and ^K127NSGK1) were transfected in this preadipocytic cell line, and the levels of phosphorylation of endogenous mouse Foxo1 were detected. As shown in Fig. 3C, Foxo1 phosphorylation at T24 and S253 (which corresponds to the human Foxo1 S256 residue) sites were increased, whereas the total level of Foxo1 remained unchanged. The level of phosphorylation at serine in position 319 could not be assessed due to the inability of the human Foxo1 phospho-specific antibody to cross-react with the mouse residue. To determine whether SGK1 is phosphorylated during adipogenesis and whether Foxo1 phosphorylation state is dependant on SGK1, we analyzed the levels of phosphorylation of both SGK1 and Foxo1 in 3T3-L1 cells. As shown in Fig. 3D, both the levels of SGK1 and Foxo1 phosphorylation are increased during 3T3-L1 differentiation. Taken all together, these data indicated that SGK1 is able to phosphorylate Foxo1 both in vitro and in cells.

Fig. 3.

SGK1 phosphorylates Foxo1. A, In vitro kinase assays performed in the presence of [γ-³²P]ATP and purified GST-Foxo1 or GST proteins. B, Western blot analysis of U2OS cells cotransfected with GFP-Foxo1 WT, GFP-Foxo1AAA, or control vector, in the presence or absence of either constitutively active SGK1 (^S422DSGK1) or SGK1 kinase-dead (^K127NSGK1). Phospho-specific antibodies recognizing phosphorylated Foxo1 protein at residues T-24, S-256, and S-319 were used to visualize phosphorylation. GFP and β-actin antibodies were used as controls. C, Phosphorylation levels of endogenous Foxo1 in 10T1/2 cells expressing constitutively active ^S422DSGK1 or kinase-dead ^K127NSGK1 or vector, using phospho-specific antibodies. D, Total and phosphoryated levels of endogenous SGK1 (P-SGK1) and Foxo1 (P-Foxo1) proteins during a time course of 3T3-L1 differentiation.

SGK1 affects Foxo1 subcellular localization

To determine whether SGK1 can affect the subcellular localization of Foxo1 and cause the inhibition of its activity by leading to its nuclear exclusion, we expressed green fluorescent protein (GFP)-Foxo1 in the presence or in the absence of constitutively active ^S422DSGK1 and monitored GFP localization in U2OS and in 10T1/2 cells. As shown in Fig. 4A, GFP-Foxo1 appeared to be predominantly localized in the cell nucleus in the absence of exogenous SGK1. In contrast, GFP-Foxo1 was excluded from the nucleus when the constitutive active ^S422DSGK1 was ectopically coexpressed. To better quantify these effects, we assessed the overall number of cells with nuclear or cytoplasmic expression of GFP-Foxo1 in the presence or absence of ectopically expressed ^S422DSGK1. To confirm the requirements of the Foxo1 phosphorylation sites for SGK1 action, we measured whether the cellular localization of the mutant Foxo1AAA, which cannot be phosphorylated, would be affected by SGK1. As shown in Fig. 4B, in the absence of SGK1, more than 60% of cells retained GFP-Foxo1 in the nucleus. Conversely, in the presence of SGK1, Foxo1 was localized in the nucleus in only 28% of the cells. Interestingly, GFP-Foxo1AAA appeared to be predominantly nuclear, independently of the presence or absence of SGK1. These data suggest that SGK1 limits the function of Foxo1 by secluding it to the cytoplasm.

Fig. 4.

Foxo1 subcellular localization is SGK1 dependent. Panel A, U2OS and 10T1/2 cells were transfected with GFP-Foxo1 in combination with constitutively active ^S422DSGK1 or vector and nuclear /cytoplasmic localization monitored by fluorescence. Panel B, GFP-Foxo1 WT or GFP-Foxo1AAA (T24A, S256A, S319A) were cotransfected in U2OS cells with constitutively active ^S422DSGK1 or vector. Quantification of the number of cells expressing GFP-Foxo1 WT or GFP-Foxo1AAA in the nucleus or in the cytoplasm were expressed as percentage relative to the total number of cells counted per subcellular compartment (***, P < 0.001; ns, not significant; N, nuclear localization; C, cytoplasmic localization). Panel C, Quantification of the number of cells showing endogenous Foxo1 localized either in the nucleus or in the cytoplasm, in WT or in SGK1−/− MEF cells in the presence of dexamethasone or vehicle alone. The values are expressed as percentage of number of cells with nuclear or cytoplasmic Foxo1 vs. total number of cells counted per field (***, P < 0.001; ns, not significant; N, nuclear localization; C, cytoplasmic localization; Dex, dexamethasone). Panel D, Subcellular localization of endogenous Foxo1 in WT or SGK1−/− MEF cells. Endogenous Foxo1 is visualized in red and nuclei with 4′,6-diamidino-2-phenylindole staining.

To determine whether dexamethasone affects Foxo1 localization and if this process is mediated by SGK1, we monitored GFP-Foxo1 subcellular localization in wild-type (WT) and SGK1−/− mouse embryonic fibroblast (MEF) cells, in the presence or absence of dexamethasone. As shown in Fig. 4, C and D, in the presence of vehicle, Foxo1 was primarily localized in the nucleus in both WT MEF and SGK1−/− cells. Conversely, in WT MEF cells treated with dexamethasone, the proportion of Foxo1 in the nucleus decreased whereas the percentage of cells with cytosolic Foxo1 staining increased. No changes in Foxo1 subcellular distribution occurred in SGK1−/− MEF treated with dexamethasone.

SGK1 rescues the inhibitory effects of Foxo1 on differentiation

Foxo1 gain- and loss-of-function studies performed in 3T3-F442A and 3T3-L1 cells have shown that this forkhead factor acts as a negative regulator of adipogenesis (12, 28). To assess whether Foxo1 affects adipocyte differentiation also in the mesenchymal cell system in which we observed SGK1 proadipogenic function, we ectopically expressed or knocked down Foxo1 in 10T1/2 cells. Expression of exogenous Foxo1 (or Foxo1AAA) inhibited adipogenesis, whereas its knockdown via siRNA caused increased differentiation, as shown by Oil Red O staining (Fig. 5, A and B) and by the changes in the levels of expression of markers of adipogenesis (Fig. 5, C and D). We next checked whether SGK1 affects differentiation in 10T1/2 cells through Foxo1 by determining whether expression of activated ^S422DSGK1 could overcome Foxo1 inhibition of adipogenesis. As shown in Fig. 5E, active ^S422DSGK1, but not the kinase-dead ^K127NSGK1, rescued the differentiation defects present in adipocytes ectopically expressing Foxo1, as demonstrated by the increase in PPARγ, C/EBPα, and aP2 mRNA levels in cells expressing ^S422DSGK1 compared with the levels in control cells. This rescue effect of SGK1 does not occur when Foxo1AAA is expressed, further confirming that SGK1 effects on adipogenesis rely on its ability to phosphorylate Foxo1. Furthermore, Foxo1 had a greater inhibitory effect on adipogenesis in the absence of SGK1, as shown by the decreased mRNA levels of PPARγ, C/EBPα, and aP2 in cells expressing Foxo1 and small interfering SGK1 (si-SGK1) in comparison to cells expressing Foxo1 and small interfering control (si-control) (Fig. 5F). In contrast, Foxo1AAA affected differentiation independently of SGK1 expression. These results confirm that SGK1’s positive action on adipogenesis is mediated through the inactivation of Foxo1 via its phosphorylation.

Fig. 5.

SGK1 rescues Foxo1-inhibitory effect on adipogenesis. Oil Red O staining of 10T1/2 cells expressing (A) vector, Foxo1 WT, or Foxo1AAA or (B) si-control vs. si-Foxo1. C, PPARγ, C/EBPα, and aP2 levels in 10T1/2 cells expressing Foxo1 WT, Foxo1AAA, or control vector, at 3 d of differentiation. (***, P < 0.001). D, aP2 mRNA levels in 10T1/2 cells expressing si-Foxo1 or si-control. E, PPARγ, C/EBPα, and aP2 mRNA levels in 10T1/2 expressing Foxo1 WT or Foxo1AAA and constitutively active ^S422DSGK1 or inactive ^K127NSGK1 (***, P < 0.001; ns, nonsignificant). F, PPARγ, C/EBPα, and aP2 mRNA levels determined after 3 d of differentiation in 10T1/2 cells expressing si-control or si-SGK1 in the presence of vector, GFP-Foxo1 WT, or GFP-Foxo1AAA. (***, P < 0.001; ns, not significant). si-ctl, si-control.

Discussion

Adipogenesis has been amenable to detailed mechanistic studies because many of the features of fat cells can be reproduced in vitro and induced by a cocktail of hormones. In cells of mouse or human origin, in vitro differentiation is significantly enhanced by glucocorticoids. These molecules play a crucial role in the early phases of differentiation and are required for induction of preadipocytes to adipocytes (25, 29, 30). In vivo effects of corticosteroids are well known as seen in patients with Cushing’s syndrome or in those in whom glucocorticoids are used for medical therapy: they all exhibit significant weight gain as a result of increased adipogenesis and lipid storage. These tissue-specific effects have been elegantly studied in transgenic mice expressing the enzyme involved in cortisol production, 11β-hydroxyl steroid dehydrogenase type 1. Mice overexpressing 11β-hydroxyl steroid dehydrogenase type 1 in adipose tissue develop visceral obesity, presumably as a result of the direct action of cortisol on adipose tissue (31, 32). Despite evidence for the direct role of corticosteroids in fat formation, their specific mechanism of action in this regard remains ill defined. In this paper, we investigated the potential role of SGK1 in the regulation of adipogenesis and its function as a possible mediator of glucocorticoids’ effects on this tissue.

Our results demonstrate that dexamethasone treatment of preadipocytes rapidly induces SGK1 expression in the early inductive stages of adipogenesis. Early expression of SGK1 in differentiation appears to be critical because forced down-regulation of this kinase in preadipocytes, by siRNA knockdown studies, leads to inhibition of differentiation. Forced expression of SGK1 results in increased adipogenesis, phenocopying the effects of glucocorticoids both with regard to cell morphology and gene expression. Overall these results predict a role for SGK1 in adipocytes in vivo. In fact, in humans, SGK1 polymorphisms, which confer increased SGK1 activity, are associated with obesity in a Caucasian cohort (33).

Total ablation of SGK1 in mice does not reveal an immediately obvious adipose defect. Although this would imply that SGK1 is not absolutely necessary in normal development of adipose tissue, our results suggest that it is important to explore the possibility that SGK1 may yet contribute to an abnormal phenotype under certain metabolic conditions known to induce or require SGK1. This logic is supported by similar observations and approaches taken to demonstrate the importance of SGK1 in salt homeostasis in SGK1−/− mice. These animals manifest deficiencies in salt metabolism only when exposed to a low-salt diet, a known physiological stimulus that induces SGK1 mRNA levels (23).

Our results show that Foxo1 is a direct target of SGK1 in adipocytes and provide a potential mechanistic explanation for the proadipogenic effects exerted by SGK1. SGK1 phosphorylates Foxo1 at threonine 24, serine 256, and serine 319 residues, leading to subcellular redistribution of Foxo1 from the nucleus to the cytosol. Interestingly, Akt phosphorylates Foxo1 at the same target sites, also leading to its nuclear exclusion (34). This suggests that different extracellular stimuli may converge on Foxo1 to achieve diverse effects through common mechanisms.

Foxo1 has been clearly demonstrated to negatively influence adipocyte differentiation (12, 28). Our observation, that dexamethasone-induced subcellular relocalization of Foxo1 is SGK1 dependent, argues strongly in favor of this pathway in mediating the proadipogenic properties of glucocorticoids. In this regard, SGK1 may serve as a critical nexus between glucocorticoids and Foxo1. We postulate that in pathological states characterized by high amounts of circulating glucocorticoids, SGK1 levels in adipose tissue may be increased leading to promotion of fat differentiation. This could be tested by assessing whether the obesity phenotype so characteristic of subjects with Cushing’s syndrome is due to higher SGK1 expression in fat tissue. If so, it is possible that the use of SGK1-specific inhibitors may be potentially very useful therapeutic agents in the treatment of this type of obesity.

Materials and Methods

Materials and reagents

Antibodies against total and phospho-Foxo1 (T-24, S-256, or S-319), total and phospho-SGK1 (S-422), β-actin, GFP, Alexa Fluor 488 goat antirabbit, troglitazone, and rosiglitazone were purchased respectively from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), Cell Signaling Technology (Danvers, MA), Sigma (St. Louis, MO), Stressgen Biotech Corp. (Victoria, British Columbia, Canada), Invitrogen (Carlsbad, CA), and Cayman Chemical Co. (Ann Arbor, MI).

Plasmids

Human GFP-Foxo1, GFP-Foxo1AAA, and GST-Foxo1 were purchased from Addgene, Inc. (Cambridge, MA). Mouse SGK1 cDNA was cloned into pCR2.1-TOPO vector (Invitrogen) after PCR amplification of a mouse kidney cDNA library using Triple Master Taq Polymerase (Eppendorf) with the following primers: mSGK1 forward (F), 5′-AACAGCCACCATGGCCGTCAAAGCCGAGGC-3′; and mSGK1 reverse (R), 5′-GGCGAGACTGCCAAGCTTCC-3′. The mSGK1 cDNA was subsequently subcloned into pCR3.1 (Invitrogen) using EcoRI. Constitutively active SGK1 (^S422DSGK1) and inactive SGK1 (^K127NSGK1) were generated according to Dieter et al. (33), using the QuikChange II XL Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) with the following primers: S422D F, 5′-GCCTTCCTCGGCTTCGATTATGCACCTCCTGTGG-3′, S422D R, 5′-CCACAGGAGGTGCATAATCGAAGCCGAGGAAGGC-3′, K127N F, 5′-GAAGTATTCTATGCAGTCAACGTTTTACAGAAGAAAGCCATCCTG-3′, K127N R, 5′-CAGGATGGCTTTCTTCTGTAAAACGTTGACTGCATAGAATACTTC-3′ (Mutagenesis site indicated by bold).

Cell culture

3T3-L1, 10T1/2, and U2OS cell lines (American Type Culture Collection, Manassas, VA) and PPARγ−/− cells (kind gift of Dr. Rosen) were grown in DMEM (Mediatech, Inc., Manassas, VA) supplemented with 10% fetal bovine serum (FBS) (Hyclone Laboratories, Logan, UT) and 1% penicillin/streptomycin (Mediatech) in 5% CO₂.

Animals

C57BL/6J male mice (6–8 wk of age) were housed in 12 h light, 12-h dark cycle (light on at 0700 h), at 22 C, and allowed ad libitum access to diet and water. Mice were fed a normal diet that contained 9% of calories as fat (Zeigler Bros, Gardners, PA). SGK1-loxP mice were generated as described elsewhere (24). Mice were killed for tissue collection by CO₂ narcosis using an approved chamber receiving CO₂ from a pressurized gas cylinder followed by cervical dislocation, according to National Institutes of Health ACUC animal study-approved procedures. Tissues were snap frozen in liquid nitrogen and stored at −80 C for subsequent analysis.

Adipocyte differentiation

3T3-L1 and 10T1/2 cells were induced to differentiate either with MDI medium consisting of 0.5 μm 3-isobutyl-1-methylxanthine, 1 μm dexamethasone, and 5 μg/ml insulin, in addition to DMEM with 10% FBS and penicillin/streptomycin or with 5 μg/ml insulin and PPARγ ligand troglitazone (10 μm) or rosiglitazone (100 nm). After 48 h of MDI induction, cells were grown in maintenance medium consisting of DMEM supplemented with 10% FBS, 1% penicillin/streptomycin, and 5 μg/ml insulin.

Oil Red O staining

To measure lipid accumulation, cells were washed twice with PBS and fixed with 10% buffered formalin (Electron Microscopy Sciences, Hatfield, PA) for 30 min. The fixed cells were incubated for 1 h in freshly diluted Oil Red O solution (Sigma), prepared by mixing six parts of Oil Red O stock solution (0.5% Oil Red O in isopropanol) and four parts of distilled water and washed three times with distilled water.

Transfection assays

Cells (3 × 10⁵) were transfected with Nucleofector or Shuttle devices (Amaxa) according to manufacturer’s instructions. Cells were transfected with either 1 μg of cDNA or 100 nm of si-RNA (ON-TARGETplus SMART pool or individual si-RNAs; Dharmacon) and induced to differentiate. The sequences of the si-RNA are represented in supplemental Table I published as supplemental table data. All experiments were performed in duplicate, three to five times. The empty vector cDNA (pcDNA3.1) or the ON-TARGETplus Non-targeting pool were used as negative controls.

Real-time PCR

RNA was extracted from cultured cells or animal tissues using Trizol reagent (Invitrogen), according to the manufacturer’s instructions. cDNA was obtained from 1 μg of total RNA using High-Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA). Real-time PCR was performed using ABI Prism 9700HT Sequence Detection System Instrument (Applied Biosystems) connected to Sequence Detector Software (Applied Biosystems) for collection and analysis of data. According to the recommendations of the manufacturer, 25 μl/96-well reactions were set up in a MicroAmp Optical 96-well reaction plate using FastStart SYBR Green (Roche, Indianapolis, IN), 5 μm reverse and forward primers (IDT DNA), and 20 ng of cDNA. Real-time PCR was performed according to the following protocol: 50 C for 2 min and 95 C for 15 min, and subsequently 40 cycles at 94 C for 15 sec, 59 C for 30 sec, and 72 C for 30 sec. 18S rRNA was used to normalize the data. For the analysis of SGK1, PPARγ, C/EBPα, aP2, and 18S the following primers were used: mSGK1 F, 5′-CTTGGGCTATCTGCACTCCC-3′; mSGK1 R, 5′-GCCCAAAGTCAGTGAGGACG-3′; mPPARγ F, 5′-AGTCTGCTGATCTGCGAGCC-3′; mPPARγ R, 5′-CTTTCCTGTCAAGATCGCCC-3′; mC/EBPα F, 5′-GAACAGCAACGAGTACCGGGTA-3′; mC/EBPα R, 5′-GCCATGGCCTTGACCAAGGAG-3′; maP2 F, 5′-TCGATGAAATCACCGCAGAC-3′; maP2 R, 5′-TGTGGTCGACTTTCCATCCC-3′; 18S F, 5′-AGTCCCTGCCCTTTGTACACA; 18S R, 5′-CGATCCGAGGGCCTCACTA-3′.

In vitro kinase assay

BL21 cells (Stratagene) were transformed with GST-Foxo1 or GST expression plasmids and GST or GST-Foxo1 fusion protein were purified. Briefly, BL21 bacteria were initially grown at 37 C overnight, and the following morning a starter culture was added to 200 ml LB broth containing 100 μg/ml ampicillin (Sigma). Bacteria were grown at room temperature, induced with 1 mm isopropyl-β-δ-thio-galactopyranoside (Fisher Scientific, Pittsburgh, PA) for 2 h and centrifuged at 7700 × g for 10 min at 4 C (Sorvall). Bacterial pellets were resuspended in cold PBS and lysed using a sonicator (Misonix) for 2 min with 30-sec intermittent bursts and 10-sec pauses. After sonication, 1% TritonX-100 was added to the solution, and lysates were rocked at room temperature for 30 min, followed by centrifugation at 20,000 × g for 20 min at 4 C. The GST fusion proteins were purified on glutathione-sepharose beads 4B (Amersham Pharmacia Biotech, Piscataway, NJ) and quantified on SDS-PAGE gels by Coomassie staining (Bio-Rad Laboratories, Inc., Hercules, CA) by comparison with BSA standards. In vitro kinase assays were performed by incubating 0.5 μg GST-Foxo1 protein, or GST alone, for 10 min at 30 C, in kinase buffer (Upstate Biotechnology, Inc., Lake Placid, NY), in the absence or presence of 0.5 μl of purified active ^S422DSGK1 and a cocktail containing inhibitor and MgAc/ATP (Upstate) and 1 μl of [γ-³²P]ATP (PerkinElmer, Wellesley, MA). A thermomixer was used to keep the beads in suspension (Eppendorf). The reaction was stopped on ice, and samples were centrifuged at 800 × rpm for 10 min at 4 C. The proteins were resolved by SDS-12% PAGE, and protein phosphorylation was detected using Biomax MR Film (Eastman Kodak Co., Rochester, NY).

Western blot analysis

For protein analysis, cells were washed twice with cold PBS and harvested in lysis buffer containing 50 mm Tris/HCl (pH 7.4), 150 mm NaCl, 1 mm EDTA, 1% Triton X-100, 25 mm sodium pyrophosphate, and a protease inhibitor cocktail (Roche). Whole-cell extracts were obtained by three cycles of freezing and thawing followed by centrifugation at 14,000 rpm for 5 min at 4 C, to remove insoluble material. Protein (30 μg) was separated by SDS-PAGE and transferred to polyvinylidene difluoride membrane (Pierce Chemical Co., Rockford, IL). Blots were blocked for 1 h at room temperature in Tris-buffered saline-Tween 20 with 5% nonfat dry milk and subsequently incubated with primary antibodies overnight at 4 C. After three washes with Tris-buffered saline-Tween 20, membranes were incubated at room temperature for 1 h with secondary antibodies. Immune complexes were visualized by ECL Plus detection reagent (Pierce) following the manufacturer’s instructions.

Generation of MEF cells

WT and SGK1-loxP MEF cells were isolated from 13-d-old embryos derived from SGK1-loxP heterozygous pregnant females crossed with SGK1-loxP heterozygous males, using standard methods (24). DNA was obtained from each embryo’s head and tail to confirm their genotype. MEF cells were maintained in DMEM, containing 10% FBS supplemented with penicillin and streptomycin, and grown in 5% CO₂. The WT and SGK1-loxP genotypes were confirmed by PCR using the following primers: F, 5′-CTCATTCCAGACCGCTGACAA-3′; R, 5′-AAAGCTTATCTCAAACCCAAACCAA-3′.

Immunofluorescence

MEF cells (5 ×10⁵) were plated in four-chamber slides (Thermo Fisher) and infected with 2 × 10⁷ PFU of Cre-GFP adenovirus (Vector BioLabs, Burlingame, CA) to obtain recombination at the SGK1 locus. Cells were treated 48 h after infection with 2 μm of dexamethasone or vehicle and 24 h later fixed and permeabilized following the manufacturer’s instruction (Invitrogen). Fixed cells were treated for 1 h at room temperature with 3% BSA blocking solution and incubated for 1 h at room temperature with anti-mFoxo1 antibody at dilution of 1:200 in 3% BSA and visualized by red fluorescence with an antirabbit IgG secondary antibody Alexa Fluor. To visualize nuclei, slides were mounted with ProLong Gold Antifade reagent containing 4′,6-diamidino-2-phenylindole (Invitrogen). To assess the overall number of cells with cytoplasmic or nuclear localization of Foxo1, each field was divided by a grid into 12 sections, and the number of cells with nuclear or cytoplasmic Foxo1 in each section was determined.

Statistical analysis

All experiments were repeated at least three times. Results are presented as means ± sem. Student’s t test and ANOVA performed with Prism version 5.0 (GraphPad Software, Inc., San Diego, CA) were used as needed. P < 0.05 was considered statistically significant.

Acknowledgments

We thank Rick Proia for helpful comments on the manuscript and Pasha Sarraf for critical discussions. We also thank William R. Sellers for kindly providing GFP-Foxo1, GFP-FoxoAAA, and GST-Foxo1. We thank Rick Dreyfuss from the Imaging Facility for the cell pictures he took for us.

This research was supported by the Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health (to E.M.) and by National Institutes of Health Grants DK 41481 (to A.N.F.T.) and DK 58898 (to G.F.T.).

Disclosure Summary: All authors have nothing to declare.

First Published Online December 4, 2006

1

N.D.P. and V.P. contributed equally to this work.

Abbreviations

 
• C/EBP

  CCAAT enhancer-binding protein;

 
• F

  forward;

 
• FBS

  fetal bovine serum;

 
• GFP

  green fluorescent protein;

 
• GST

  glutathione-S-transferase;

 
• MDI

  dexamethasone, insulin, and isobutylmethylxanthine;

 
• MEF

  mouse embryonic fibroblast;

 
• PPAR

  peroxisome proliferator-activated receptor;

 
• R

  reverse;

 
• SGK1

  serum- and glucocorticoid-inducible protein kinase 1;

 
• si-control

  small interfering control;

 
• siRNA

  small interfering RNA;

 
• si-SGK

  small interfering SGK;

 
• WT

  wild type.