Activation of Peroxisome Proliferator–Activated Receptor-β/-δ (PPAR-β/-δ) Ameliorates Insulin Signaling and Reduces SOCS3 Levels by Inhibiting STAT3 in Interleukin-6–Stimulated Adipocytes

The PPAR-δ ligand GW501516 was obtained from Biomol Research Laboratories (Plymouth Meeting, PA). Other chemicals were from Sigma (St. Louis, MO).

3T3-L1 preadipocytes [American Type Culture Collection (ATCC)] were grown to confluence in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% bovine calf serum. Two days after confluence (day 0), differentiation of the 3T3-L1 cells was induced in DMEM containing 10% FBS, methylisobutylxanthine (500 μmol/L), dexamethasone (0.25 μmol/L), and insulin (5 µg/mL) for 48 h. The cells were then incubated in 10% FBS/DMEM with insulin for 3 days and this was then replaced with FBS/DMEM. Medium was changed every 2 days. Fat droplets were observed in more than 90% of cells after day 10. Adipocytes were then incubated with 10 μmol/L GW501516 and IL-6 (10 or 100 ng/mL) for the times indicated. After incubation, RNA and total and nuclear protein extracts were extracted from adipocytes as described below. Inhibitors were added 30 min before incubation with IL-6.

Obese male ZDF rats (ZDF/Gmi, fa/fa) and their lean littermates (fa/+ or ^+/+) were used. Both strains were maintained under standard light-dark cycle (12-h light/dark cycle) and temperature (21 ± 1°C) conditions and fed with Purina 5008 chow. Male ZDF and lean rats were killed at 12 weeks of age. CD-1 male mice (12 weeks old) were treated for 48 h with vehicle (100 μL PBS-0.1% BSA and 0.5% w/v carboxymethylcellullose medium viscosity), IL-6 (0.8 μg/g body wt i.p.), or IL-6 plus GW501516 (one daily oral gavage of 3 mg/kg/day GW501516 dissolved in carboxymethylcellullose). Epididymal white adipose tissue of rats and mice was rapidly removed, frozen in liquid nitrogen, and stored at −80°C. All procedures were conducted in accordance with the principles and guidelines established by the University of Barcelona Bioethics Committee, as stated in Law 5/1995, 21 July, passed by the Generalitat de Catalunya.

The generation of PPAR-δ–null mice was described previously (31). Eight male PPAR-δ–null mice and eight of their control male PPAR-δ wild-type mice were used (5 to 6 months of age). In agreement with the guidelines specified by the veterinary office of Lausanne (Switzerland), the mice were housed under standard light-dark cycle (12-h light/dark cycle) and temperature (21 ± 1°C) conditions and fed with Provimi Kliba 3436 chow. Epididymal white adipose tissue was rapidly removed, frozen in liquid nitrogen, and stored at −80°C.

Determination of 2-Deoxy-d-[¹⁴C]glucose (2-DG) uptake was performed as reported elsewhere (32).

Levels of mRNA were assessed by RT-PCR as previously described (33). Total RNA was isolated using the Ultraspec reagent (Biotecx, Houston, TX). The total RNA isolated by this method is nondegraded and free of protein and DNA contamination. The sequences of the sense and antisense primers used for amplification were: Socs3 (suppressor of cytokine signaling 3) 5′-TTTTCGCTGCAGAGTGACCCC-3′ and 5′-TGGAGGAGAGAGGTCGGCTCA-3′; early growth response (Egr-1), 5′-CTTCCTCTGCCTCCCAGAGCC-3′ and 5′-TGGGAACCTGGAAACCACCCT-3′, and Aprt (adenosyl phosphoribosyl transferase), 5′-GCCTCTTGGCCAGTCACCTGA-3′ and 5′-CCAGGCTCACACACTCCACCA-3′. Amplification of each gene yielded a single band of the expected size (Socs3: 250 bp, Egr-1: 210 bp, and Aprt: 329 bp). Preliminary experiments were carried out with various amounts of cDNA to determine nonsaturating conditions of PCR amplification for all the genes studied. Under these conditions, relative quantification of mRNA was then assessed by the RT-PCR method used in this study (34). Radioactive bands were quantified by video-densitometric scanning (Vilbert Lourmat Imaging). The results for the expression of specific mRNAs are always presented relative to the expression of the control gene (Aprt).

Nuclear extracts were isolated as previously described (35). Cells were scraped into 1.5 mL of cold PBS, pelleted for 10 s, and resuspended in 400 μL cold Buffer A (10 mM HEPES [pH 7.9 at 4°C], 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM DTT, 0.2 mM PMSF, and 5 μg/mL aprotinin) by flicking the tube. Cells were allowed to swell on ice for 10 min and then vortexed for 10 s. Samples were then centrifuged for 10 s, and the supernatant fraction was discarded. Pellets were resuspended in 50 μL of cold Buffer C (20 mM HEPES-KOH [pH 7.9 at 4°C], 25% glycerol, 420 mM NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.5 mM DTT, 0.2 mM PMSF, 5 μg/mL aprotinin, and 2 μg/mL leupeptin) and incubated on ice for 20 min for high-salt extraction. Cellular debris was removed by centrifugation for 2 min at 4°C, and the supernatant fraction (containing DNA-binding proteins) was stored at −80°C. Nuclear extract concentration was determined by the Bradford method.

Electrophoretic mobility shift assay (EMSA) was performed using double-stranded oligonucleotides (Santa Cruz) for the consensus binding site of the STAT3 nucleotide (5′-GATCCTTCTGGGAATTCCTAGATC-3′). Oligonucleotides were labeled in the following reaction: 2 μL of oligonucleotide (1.75 pmol/μL), 2 μL 5× kinase buffer, 1 μL T4 polynucleotide kinase (10 units/μL), and 2.5 μL [γ-³²P]ATP (3,000 Ci/mmol at 10 mCi/mL) incubated at 37°C for 2 h. The reaction was stopped by adding 90 μL of TE buffer (10 mmol/L Tris-HCl [pH 7.4] and 1 mmol/L EDTA). To separate the labeled probe from the unbound ATP, the reaction mixture was eluted in a Nick column (Amersham) following the manufacturer’s instructions. Crude nuclear protein (μg) was incubated for 10 min on ice in binding buffer (10 mmol/L Tris-HCl [pH 8.0], 25 mmol/L KCl, 0.5 mmol/L DTT, 0.1 mmol/L EDTA [pH 8.0], 5% glycerol, 5 mg/mL BSA, and 50 μg/mL poly[dI-dC]), in a final volume of 15 μL. Labeled probe (~75,000 cpm) was added, and the reaction was incubated for 30 min at 4°C. Where indicated, specific competitor oligonucleotide was added before the labeled probe and incubated for 20 min on ice. STAT3 antibody was added 15 min before incubation with the labeled probe at 4°C. Protein-DNA complexes were resolved by electrophoresis at 4°C on a 5% acrylamide gel and subjected to autoradiography.

Antibodies against total and phospho-ERK1/2 and phospho-STAT3 (Tyr⁷⁰⁵ and Ser⁷²⁷) were purchased from Cell Signaling. Antibodies against total STAT3 and Hsp90 were purchased from Santa Cruz.

To obtain total protein, cells and adipose tissue were homogenized in radioimmunoprecipitation assay (RIPA) buffer (Sigma) with phosphatase inhibitors (0.2 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L sodium orthovanadate, 5.4 μg/mL aprotinin). The homogenate was centrifuged at 16,700g for 30 min at 4°C. Protein concentration was measured by the Bradford method.

Whole-cell lysates and nuclear extracts were mixed with various antibodies (as specified under 14results) and protein A coupled to agarose beads. Proteins from whole-cell lysates, nuclear extracts, and immunoprecipitates were separated by SDS-PAGE and then transferred to immobilon polyvinylidene diflouride membranes (Millipore, Bedford, MA) and blotted with various antibodies (as specified in 14results). Detection was achieved using the EZ-ECL chemiluminescence kit (Amersham). Size of detected proteins was estimated using protein molecular-mass standards (Invitrogen, Barcelona, Spain).

Data are presented as mean ± SD of five separate experiments. Significant differences were established by one-way ANOVA, using the GraphPad InStat program (GraphPad Software V2.03, GraphPad Software, San Diego, CA). When significant variations were found, the Tukey-Kramer multiple comparisons test was applied. Differences were considered significant at P < 0.05.

We first examined the effects of the PPAR-δ agonist GW501516 on IL-6–induced insulin resistance, which was assessed as the inhibition of insulin-stimulated Akt phosphorylation and glucose uptake. Differentiated 3T3-L1 adipocytes were stimulated with IL-6 in the absence or in the presence of GW501516, a selective ligand for PPAR-δ with a 1,000-fold higher affinity toward PPAR-δ than PPAR-α and PPAR-γ (36). Exposure of adipocytes to IL-6 for 24 h caused a reduction in insulin-stimulated Akt phosphorylation (Fig. 1A). In contrast, when cells preincubated with 10 μmol/L GW501516 were exposed to IL-6 the inhibitory effect of this cytokine on insulin-stimulated Akt phosphorylation was prevented. The effect of GW501516 on insulin-stimulated Akt phosphorylation was also observed at lower concentrations and the effect attained at 10 μmol/L was dependent on PPAR-δ since it was abolished by coincubation with the PPAR-δ antagonist GSK0660 (Fig. 1B). Similarly, GW501516 significantly reversed the reduction observed in glucose uptake in IL-6–stimulated cells (Fig. 1C). Drug treatment in the absence of insulin did not affect the phosphorylation status of Akt (data not shown). Thus GW501516 treatment offered protection against a reduction in insulin responsiveness by IL-6.

Because IL-6–induced insulin resistance in 3T3-L1 adipocytes has been attributed to SOCS3 (22), we then examined the effect of PPAR-δ activation on the mRNA levels of the STAT3-target gene SOCS3. Differentiated 3T3-L1 adipocytes were stimulated with 10 ng/mL of IL-6 for 1 h in the absence or in the presence of GW501516. Under these conditions, the increase in SOCS3 mRNA levels caused by IL-6 exposure (23-fold induction; P < 0.001) was reduced in cells coincubated with IL-6 plus GW501516 (sevenfold induction; P < 0.001 vs. IL-6–stimulated cells) (Fig. 2A). Furthermore, because of recent reports showing suppression of IL-6 signaling through inhibition of STAT3-Hsp90 interaction by the selective Hsp90 inhibitor geldanamycin (16,37,38), we also evaluated the effects of the latter. Geldanamycin treatment significantly reduced the increase in SOCS3 mRNA levels caused by IL-6 (P < 0.05 vs. IL-6–stimulated cells). Overall, these findings suggest that PPAR-δ activation inhibits STAT3. Dimerization, nuclear translocation, and increase in transcriptional activity of STAT3 require its phosphorylation on tyrosine residue 705. In addition, STAT3 phosphorylation on Ser⁷²⁷ is required for its maximal activation (13,14). When we analyzed the phosphorylation status of STAT3 we observed that IL-6 exposure increased both Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation, whereas in the presence of GW501516 these changes were prevented (Fig. 2B). Because IL-6 activates ERK1/2 (10), which has been reported to be a kinase for STAT3 phosphorylation on Ser⁷²⁷ (15), and we have previously reported that GW501516 prevents LPS-induced ERK1/2 activation in adipocytes (36), we evaluated the effect of the PPAR-δ agonist on the activation of this kinase. IL-6 exposure caused a slight increase in ERK1/2 phosphorylation, whereas GW501516 strongly suppressed ERK1/2 protein phosphorylation (Fig. 2C). Consistent with these changes, the increase in Egr-1 mRNA levels, which has been attributed to IL-6–mediated activation of ERK1/2 (10), was abolished by GW501516 (Fig. 2D). To demonstrate the involvement of the ERK1/2 activation in IL-6–induced insulin resistance in adipocytes, we took advantage of U0126, a potent and specific ERK1/2 inhibitor, which binds to mitogen-activated protein kinase (MAPK)–ERK1/2 (MEK1/2), thereby inhibiting its catalytic activity as well as phosphorylation of ERK1/2. Similarly to GW501516, U0126 prevented the reduction in insulin-stimulated Akt phosphorylation (Fig. 3A) and glucose-uptake (Fig. 3B) and prevented the increase in STAT3 phosphorylation on Ser⁷²⁷ (Fig. 3C) caused by IL-6. U0126 treatment alone did not affect the phosphorylation status of Akt (data not shown).

We have previously reported that PPAR-δ expression and activity is reduced, whereas ERK1/2 phosphorylation is increased in white adipose tissue from an animal model of obesity and insulin resistance, the ZDF (fa/fa) rat (36). In the current study we found that STAT3 phosphorylation at Ser⁷²⁷ and SOCS3 protein levels were increased in white adipose tissue of ZDF rats compared with lean animals (Fig. 4A and B). Hence, we hypothesized that reduced PPAR-δ activity and enhanced ERK1/2 activation may contribute to the increase in STAT3 activity in the white adipose tissue of the ZDF rat. We then evaluated whether GW501516 might prevent the increase in STAT3-SOCS3 pathway in white adipose tissue after IL-6 stimulation in vivo in a similar fashion as observed in vitro. When we examined the SOCS3 protein levels in mice exposed to IL-6 (Fig. 4C) we observed that the increase caused by IL-6 treatment was prevented in those mice treated with the PPAR-δ agonist. Similarly, drug treatment prevented the increase in the phosphorylation status of STAT3 in both Tyr⁷⁰⁵ and Ser⁷²⁷ (Fig. 4D). In agreement with data obtained in vitro, GW501516 inhibited the increase in phospho-ERK1/2 levels induced by IL-6 treatment (Fig. 4E).

To clearly demonstrate the involvement of PPAR-δ in the regulation of STAT3 activity in white adipose tissue we took advantage of the PPAR-δ–null mouse. In the absence of PPAR-δ, Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation of STAT3 and SOCS3 protein levels were increased compared with wild-type mice (Fig. 5A). Consistent with these changes, the DNA-binding activity of STAT3 was strongly increased in nuclear extracts of white adipose tissue of PPAR-δ–null mice compared with wild-type animals (Fig. 5B).

Hsp90 is thought to contribute to many steps in STAT3 activation, such as binding to its docking sites on gp130 and subsequent phosphorylation by associated JAKs, as well as enhanced trafficking of the activated cytosolic STAT3 to the nucleus (16,37). Thus we examined whether PPAR-δ suppressed IL-6 signaling through inhibition of STAT3-Hsp90 interaction. This interaction was studied by using nuclear extracts of 3T3-L1 adipocytes stimulated with IL-6 in the presence or in the absence of GW501516 and the Hsp90 inhibitor geldanamycin, which were immunoprecipitated with anti-Hsp90, and analyzed by Western blot (Fig. 6A). IL-6 stimulation caused an increase in the interaction of Hsp90 with STAT3, whereas in the presence of GW501516 or geldanamycin the IL-6–induced recruitment of Hsp90 to STAT3 was blocked. Consistent with this, the association of STAT3 with Hsp90 was very faint in white adipose tissue of wild-type mice but strongly increased in PPAR-δ–null mice (Fig. 6B). Overall, these findings indicate that PPAR-δ may inhibit IL-6–induced STAT3 activation by promoting STAT3 dissociation from Hsp90.

Insulin resistance and type 2 diabetes are closely associated with low-grade chronic inflammation characterized by abnormal proinflammatory cytokine production (39), such as TNF-α (1) and IL-6 (4,5,21). Of these cytokines, IL-6 plasma levels correlate more strongly with the severity of insulin resistance in insulin-resistant patients than with TNF-α (3,6). Adipocytes play an important role in IL-6–induced insulin resistance since adipose tissue is an important source (40) and a major target of this cytokine. IL-6 acts primarily by activating STAT3 and upregulating the transcription of its target gene SOCS3, which causes insulin resistance by interfering with insulin receptors and/or IRS-1 (19,41–43). Our findings demonstrate that PPAR-δ activation by GW501516 prevents IL-6–induced expression of SOCS3 in adipocytes, showing a similar effect to that reported for the PPAR-γ activators rosiglitazone (21) and pioglitazone (44). These data suggest that PPAR-δ prevents STAT3 activation in adipocytes, which is in agreement with a previous report showing that GW501516 prevented IL-6–induced STAT3 activation in hepatocytes (30), although in this latter study the molecular mechanism involved was not elucidated. The activity of STAT3 is dependent on its phosphorylation status. Thus STAT3 phosphorylation on Tyr⁷⁰⁵ leads to dimerization and translocation to the nucleus, where it regulates the transcription of its target genes (12), whereas phosphorylation of Ser⁷²⁷ is required to achieve maximal activity (13,14). Our findings show that the PPAR-δ activator GW501516 inhibits IL-6–induced STAT3 phosphorylation on both residues. The increase in STAT3 phosphorylation, STAT3 DNA-binding activity, and SOCS3 protein levels in white adipose tissue from PPAR-δ−null mice compared with wild-type animals clearly indicates that PPAR-δ inhibits STAT3 activation in adipocytes. Interestingly, white adipose tissue from ZDF rats showed increased STAT3 serine phosphorylation and SOCS3 protein levels. These changes could be related to the reduction in PPAR-δ expression observed in white adipose tissue of ZDF rats (36). In addition, because overexpression of SOCS3 in adipose tissue causes local but not systemic insulin resistance (45), these findings suggest that the increase in SOCS3 levels in white adipose tissue from ZDF rats might, at least, exacerbate insulin resistance in this tissue, thereby contributing to the metabolic alterations observed in this animal model of type 2 diabetes.

The prevention of IL-6–induced STAT3 activation after PPAR-δ activation might be the result of several mechanisms of action. First, protein kinases responsible for STAT3 serine phosphorylation include, among others, ERK1/2 (46), and we have previously reported that GW501516 inhibits ERK1/2 phosphorylation in adipocytes (36). GW501516 abolished both ERK1/2 phosphorylation and the expression of Egr-1, which is upregulated after ERK1/2 activation (10), suggesting that inhibition of this protein kinase by PPAR-δ activation might be involved in the suppression of STAT3 Ser⁷²⁷ phosphorylation in IL-6–exposed cells. These data are consistent with previous studies reporting that the MEK/ERK inhibitor PD98059 suppresses STAT3 serine phosphorylation, whereas STAT3 tyrosine phosphorylation was blunted by the JAK2 inhibitor AG490 (47). However, our findings also showed a reduction of STAT3 Tyr⁷⁰⁵ phosphorylation after GW501516 treatment, suggesting that additional mechanisms were involved. This second mechanism might involve a reduction in the interaction of STAT3 with Hsp90. In fact, activation of STAT3 requires its association with Hsp90 in many steps, including binding to its docking sites on gp130 and subsequent phosphorylation by associated JAKs and its translocation to the nucleus (16). Thus it has been reported that both geldanamycin, a selective Hsp90 inhibitor, and pyrrolidine dithiocarbamate (PDTC) inhibit STAT3 tyrosine and serine phosphorylation and its translocation to the nucleus by reducing the association of STAT3 with Hsp90 (16). Our immunoprecipitation studies in adipocytes exposed to IL-6 indicate that PPAR-δ activation by GW501516 dissociates Hsp90 from STAT3, a mechanism that may prevent STAT3 activation by phosphorylation. In agreement with this, white adipose tissue from PPAR-δ knockout mice showed enhanced STAT3 phosphorylation and DNA-binding activity. Our data do not explain how PPAR-δ activation reduces the physical interaction between STAT3 and Hsp90. However, because it has been reported that PPAR-δ interacts with Hsp90 (48), it is likely that competition between PPAR-δ and STAT3 for binding to Hsp90 may reduce the availability of this heat shock protein to interact with STAT3.

The mechanisms of action responsible for STAT3 inhibition by PPAR-δ agonists reported here are different to that previously reported for PPAR-γ activators (49,50). Thus ligand binding to PPAR-γ promotes its dissociation from the corepressor silencing mediator for retinoid and thyroid hormone receptors (SMRT), which, in turn, interacts with STAT3 and inhibits its transcriptional activity. It remains a matter of further study whether PPAR-γ ligands inhibit STAT3 by promoting its dissociation from Hsp90. However, this possibility seems unlikely since it has been reported that, in contrast with PPAR-δ and PPAR-α, PPAR-γ does not interact with Hsp90 (48). Further studies are necessary to evaluate whether PPAR-δ–mediated inhibition of STAT3 involves enhanced interaction between SMRT and STAT3.

It has been suggested that the proinflammatory cytokine IL-6 is one of the mediators linking obesity-derived chronic inflammation with insulin resistance. In addition, it has previously been reported that STAT3 activation is required for IL-6 inhibition of insulin signaling in hepatocytes (51) and that the negative effect of IL-6 on insulin signaling in adipocytes is linked to the upregulation of SOCS3 (21). Thus we wanted to explore whether PPAR-δ activation prevented IL-6–induced insulin resistance in adipocytes. In agreement with the reported inhibition of the STAT3-SOCS3 by GW501516, PPAR-δ activation by this drug prevented the reduction in insulin-stimulated Akt phosphorylation and glucose uptake caused by IL-6 exposure. These findings suggest that the inhibition of STAT3 and the subsequent reduction in SOCS3 levels after PPAR-δ activation in IL-6–stimulated adipocytes might contribute toward preventing IL-6–induced insulin resistance. Given that impairment of insulin resistance in adipocytes by exposure to TNF-α (52) and IL-1α (53) has been largely associated with IL-6 production and SOCS3 induction, it is likely that the effects of PPAR-δ on the improvement in insulin sensitivity can also be extended to these cytokines.

In summary, on the basis of our findings in adipocytes, we suggest that PPAR-δ activation prevents IL-6–induced STAT3 activation and SOCS3 upregulation, thereby contributing toward preventing the cytokine-mediated development of insulin resistance.

This study was partly supported by funds from the Swiss National Science Foundation, the Spanish Ministerio de Ciencia e Innovación (SAF2006-01475 and SAF2009-06939), and European Union ERDF funds. CIBER de Diabetes y Enfermedades Metabólicas (CIBERDEM) is an Instituto de Salud Carlos III project. L.S.-M. was supported by an FPI grant from the Spanish Ministerio de Ciencia e Innovación. R.R.-C. was supported by a grant from the Fundación Ramón Areces.

No potential conflicts of interest relevant to this article were reported.

L.S.-M., R.R.-C., and I.E.K. performed experiments. X.P. performed experiments and contributed to discussion. L.M. and W.W. developed the PPAR-β/-δ–null mice and contributed to discussion. M.V.-C. wrote the manuscript and performed experiments.

The authors thank the University of Barcelona’s Language Advisory Service for its help.