Insulin regulates Irf4 expression via effects on FoxO1

Adipose tissue is insulin-sensitive, while spleen is not; the fact that IRF4 expression is regulated in the former but not the latter suggested to us that insulin might repress IRF4. This proved to be the case, as cultured 3T3-L1 adipocytes treated with insulin showed a dose- and time-dependent reduction in Irf4 mRNA levels (Fig. 1D and Fig. S1C). To determine whether insulin plays a similar role in vivo, we examined adipose tissue from mice treated with streptozotocin, which causes β-cell failure and insulinopenia. In these animals, Irf4 mRNA expression in WAT rose significantly, and decreased when insulin was replaced (Fig. 1E). Similarly, Irf4 mRNA levels were elevated in mice lacking insulin receptors exclusively in adipose tissue [so-called Fat specific Insulin Receptor Knockout (FIRKO) mice (Bluher et al., 2002)] (Fig. 1F). Taken together, these findings indicate that insulin is a potent repressor of IRF4 expression, and implicate the rise and fall of insulin during feeding and fasting as a dominant determinant of IRF4 levels in fat. Consistent with this notion, IRF4 expression is repressed in whole WAT of three different rodent models of obesity, a hyperinsulinemic state (Fig. S1D). As previously reported (Joken et al., 2007; Kershaw et al., 2007), there was also a significant reduction of ATGL mRNA in whole WAT of high-fat diet induced obesity and ob/ob mice compared to controls (Fig. S1D). Fractionation reveals that the obesity-induced reduction of Irf4 in whole WAT predominantly reflects an effect in mature adipocytes; high-fat feeding induces an increase in stromal-vascular IRF4 content, consistent with adipose infiltration of macrophages (which also express a significant amount of IRF4) in obesity (Fig. S1E).

Many gene expression events affected by insulin are mediated by the transcription factor forkhead box O1 (FoxO1), which is phosphorylated and subsequently excluded from the nucleus upon insulin stimulation (Biggs et al., 1999; Kops et al., 1999; Nakae et al., 2000). To assess whether FoxO1 is an upstream regulator of Irf4 transcription, we introduced wild-type FoxO1 (WT), a constitutively active mutant FoxO1 (CA), a dominant negative FoxO1 mutant (DN), or EGFP control into mature 3T3-L1 ΔCAR cells using adenovirus. Ectopic overexpression of WT and CA FoxO1 caused a significant increase in basal Irf4 mRNA levels, while the DN mutant had the opposite effect (Fig. 1G). Although insulin stimulation could still diminish Irf4 mRNA levels in WT cells, this effect was blocked in CA cells. In silico analysis identified at least four predicted FoxO1 binding sites in the proximal 8 kb of the murine Irf4 upstream flanking sequence. We used serial deletion analysis of this region in a luciferase reporter to identify the region between 1250 and 1320 base pairs upstream of the Irf4 transcriptional start site as the insulin-responsive locus (Fig. 1H), and chromatin immunoprecipitation (ChIP) assay in 3T3-L1 adipocytes confirmed direct, insulin-sensitive FoxO1 binding to this region (Fig. 1I).

IRF4 is required for lipolysis in vitro

Given the profound induction of IRF4 in fasted adipocytes, we postulated a role for this transcription factor in lipolysis. To address this, we generated immortalized lines of wild-type (WT) and Irf4^−/− (KO) mouse embryonic fibroblasts (MEFs), and then used retroviral expression of C/EBPα to convert these cells into adipocytes. Oil red O staining and adipocyte marker gene expression demonstrated that equal degrees of differentiation were achieved in both cell lines (Fig. S2A and S2B). As predicted, isoproterenol-stimulated lipolysis was diminished in adipocytes lacking IRF4 (Fig. 2A). The converse was also true; forced IRF4 overexpression in mature WT and KO adipocytes enhanced lipolysis in the presence of isoproterenol (Fig. 2A). IRF4 is a transcription factor, so we hypothesized that expression of key lipases might be affected by its absence. Indeed, loss of IRF4 corresponded to a significant reduction in the mRNA (Fig. 2B) and protein levels (Fig. 2C) of Lipe (encoding HSL) and Pnpla2 (encoding ATGL). As in MEFs, forced IRF4 overexpression (Fig. S2C) in mature 3T3-L1 adipocytes significantly enhanced lipolysis in the presence of isoproterenol (Fig. S2D) and increased the expression of Lipe and Pnpla2 mRNA (Fig. S2E).

Open in a separate window

Figure 2

IRF4 promotes lipolysis in vitro via effects on lipase expression

A. Basal and isoproterenol-stimulated glycerol release (6-hour stimulation) in WT and KO MEFs, with and without the addition of exogenous IRF4. *p<0.05 (n=6), Results expressed as mean ± SD. B. Pnpla2, Lipe, and Plin1 mRNA expression in WT and KO MEFs. *p<0.05 relative to WT control (n=3). Results expressed as mean ± SD. C. ATGL, HSL, and PLIN1 protein expression in WT and KO MEFs. D. Luciferase activity of Pnpla2 promoter reporter constructs transfected into 3T3-L1 adipocytes. Five days after adipogenic stimulation, 3T3-L1 cells were co-transfected with the indicated promoter construct and an IRF4 expression plasmid (or control). Luciferase activity was measured 24 hr after transfection. Results are expressed as mean ± SD (n = 6). *p < 0.05 relative to EGFP. E. ChIP analysis in 3T3-L1 adipocytes overexpressing IRF4. The signal in IgG is set as 1. Results expressed as mean ± SD.

To investigate possible direct regulation of lipase gene expression by IRF4, we performed transactivation assays in 3T3-L1 cells. These studies demonstrated that IRF4 induced reporter expression through predicted ISRE binding sites in the upstream flanking region of Pnpla2 (Fig. 2D), and ChIP assay confirmed IRF4 binding to one of these sites (Fig. 2E). While binding was detected at several putative ISRE sites in the Lipe promoter, IRF4 was able to induce reporter expression even in the absence of these motifs, suggesting the existence of an additional cryptic binding site in the proximal promoter (Fig. S2F and S2G). These data indicate that IRF4 is required for lipolysis, at least in part through direct induction of ATGL and HSL.

IRF4 inhibits lipogenesis

During feeding, adipose tissue stores excess calories as triglyceride. While some lipids are imported directly from the circulation, adipocytes also perform de novo lipogenesis. Because lipolysis and lipogenesis are functionally opposed, we speculated that IRF4 might act to suppress lipogenesis and triglyceride synthesis. This proved to be the case, as insulin was able to increase the rate of lipid synthesis in KO adipocytes to a greater degree than in WT cells (Fig. 3A). Furthermore, forced expression of IRF4 suppressed insulin-stimulated lipid synthesis, and was able to fully revert the lipogenic phenotype of cells lacking Irf4.

Open in a separate window

Figure 3

IRF4 inhibits lipogenesis in vitro

A. Insulin (100nM) stimulation of ¹⁴C glucose incorporation into lipid was determined in WT and KO MEFs, with and without the addition of exogenous IRF4. *p<0.05 (n=6), Results expressed as mean ± SD. from three independent experiments. *p<0.05 relative to WT control under insulin stimulation, ^#p<0.05 relative to KO MEF under insulin stimulation. B. WT and KO MEFs were transduced with lentivirus expressing IRF4 or EGFP 5 days after differentiation induction. mRNA expression of lipogenesis genes was measured using Q-PCR 7 days after infection. Results expressed as mean ± SD, n=6.

Consistent with the effect on lipogenesis, the expression of genes related to lipid synthesis was regulated by IRF4 (Fig. 3B). All lipid synthesis genes tested were repressed at least somewhat by exogenous IRF4, whereas ablation of IRF4 led to significant increases in select targets only, such as Scd1, Fasn, Srebf1, and Me1. Taken together, our data demonstrates coordinated regulation of lipid synthesis and breakdown in cultured adipocytes.

Fat-specific deletion of IRF4 results in increased adiposity

We sought next to extend our in vitro findings to a more physiologically relevant system. Because we wished to avoid confounding effects in macrophages or other immune cell types, we focused on a model that would allow observation of adipocyte-specific actions of IRF4 in vivo. To that end, we generated a BAC transgenic mouse in which Cre recombinase was inserted into the starting ATG of Adipoq, which encodes adiponectin (See Methods). These mice express Cre recombinase in a highly adipocyte-specific manner, with no expression in stimulated or resident macrophages (Fig. S3A-D). We crossed these Adipoq-Cre mice to Irf4^flox/+ mice to allow generation of fat-specific IRF4 knockout mice (hereafter referred to as FI4KO). FI4KO mice showed virtually complete absence of IRF4 mRNA and protein specifically in WAT and BAT, with intact IRF4 expression in spleen and macrophages (Figs. 4A and 4B). In addition to confirming the absence of IRF4, we also examined expression of the eight other IRF family members for possible compensatory regulation. As we reported previously, all IRF family members are detectable in WAT. We noted that IRF6, IRF7, and IRF9 were up-regulated in the absence of IRF4, however, the significance of this up-regulation is uncertain; IRF6 and IRF9 in particular are expressed at much lower absolute levels than IRF4 (Fig. S4E). FI4KO mice were born in a Mendelian ratio, and there were no obvious morphological differences from control mice (WT, Adipoq-Cre only, or Flox) in young animals (data not shown). On chow diet, total body weight was not different from controls, although MRI revealed a small but significant excess of adipose mass in 18-week old FI4KO animals (Fig. S4A and S4B). Histological analysis and cell size measurement of WAT showed slightly larger adipocytes in FI4KO mice than in littermate controls (Fig. S4C and S4D). When placed on a high-fat diet, the body weight of FI4KO mice was significantly greater than that of littermates (Figs. 4C and 4E), with the difference explained entirely by excess adiposity and larger fat cells (Figs. 4D, 4F, and 4G). Interscapular brown adipose tissue (BAT) mass was also enlarged in high-fat fed FI4KO mice (Figs. 4H). There was no obvious difference in hepatic lipid content between Flox and FI4KO animals (data not shown). FI4KO mice on high-fat diet showed impaired glucose and insulin tolerance to a degree consistent with their increased body weight (Fig. S5A,B). We were unable to detect a difference in food intake or locomotor activity; FI4KO mice showed somewhat reduced oxygen consumption and CO₂ production when normalized by body weight (Fig. S6A-C). This difference disappeared, however, when normalized on a per animal basis.

Open in a separate window

Figure 4

Characterization of Fat-specific IRF4 KO (FI4KO) mice

A. Irf4 mRNA expression in WAT, BAT, spleen, and peritoneal macrophages of male Irf4^f/f (Flox) and FI4KO (KO) mice (n=3). Data are normalized to 36B4. *p<0.05 relative to Flox mice. Results are expressed as mean ± SD. B. IRF4 protein expression in WAT, BAT, spleen, and peritoneal macrophages of male Adipoq-Cre and FI4KO (KO) mice. C. Body weights of male WT, Adipoq-Cre, Flox, and FI4KO (KO) mice on high-fat diet (n=8-10). Results are expressed as mean ± SD. D. Body composition analysis in 14-week-old male mice (n=8-10) by MRI. Results are expressed as mean ± SD, *p<0.05 relative to Flox mice. E. Morphology of FI4KO (KO) mice (16 month old) and fat pads after high-fat feeding. F. Hematoxylin and eosin staining of paraffin-embedded epididymal WAT sections from 16-month-old mice. Scale bars = 50μm. G. Representative cell size distribution of adipocytes in epididymal WAT from Flox and FI4KO (KO) mice. H. Gross appearance and hematoxylin and eosin staining of paraffin-embedded BAT sections from 16-month-old mice. Scale bars = 50μm.

Cell culture

3T3-L1 cells (American Type Culture Collection) or 3T3-L1ΔCAR cells (Orlicky et al., 2001) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen) with 10% Bovine Calf Serum (BCS; Hyclone) in 5% CO₂. Two days post confluence, cells were exposed to DMEM/10% Fetal Bovine Serum (FBS; Invitrogen) with 1μM dexamethasone (Sigma), 5μg/ml insulin (Sigma), and 0.5mM isobutylmethylxanthine (Sigma). After 2 days, cells were maintained in medium containing FBS only. For insulin regulation experiments, fully differentiated 3T3-L1 adipocytes were incubated in serum-free DMEM containing 1% fatty acid-free BSA (Sigma) with and without bovine insulin at the doses and times indicated.

Generation and differentiation of immortalized mouse embryonic fibroblasts (MEFs)

Irf4^+/− mice were generated as previously described (Mittrucker et al., 1997). Male and female Irf4^+/− mice were mated, and embryos were harvested at e12.5. Primary MEFs were generated and plated as monolayer cultures in DMEM with 10% FBS as described previously (Rosen et al., 2002). Primary MEFs were immortalized by plating 3 × 10⁵ cells per 60-mm dish every 3 days as described (Todaro and Green, 1963). The C/EBPα-pMSCV construct was transfected into Phoenix packaging cells using the CellPhect transfection kit (Amersham Biosciences). Immortalized MEFs were transduced with viral supernatants and selected in hygromycin. Two days post confluence, cells were exposed to DMEM/10% FBS with dexamethasone (1μM), insulin (5μg/ml), isobutylmethylxanthine (0.5mM), and rosiglitazone (1μM). After 2 days, cells were maintained in medium containing FBS only.

Animals

Fasting and feeding studies were performed using FVB mice obtained from Charles River Laboratories and fed a standard diet (8664, Harlan Teklad). Mice were maintained under a 14 hr light/10 hr dark cycle at constant temperature (22°C) with free access to food and water. To determine the effect of nutritional status, 10-week-old male FVB mice were fed ad libitum, fasted for 24h, or re-fed for 24h after the 24-h fast (n=7-8). To determine the effect of insulin deficiency and replacement on IRF4 mRNA expression, 10-week old male C57BL/6J mice were treated with streptozotocin (STZ) with or without insulin replacement as described (Kershaw et al., 2006). All animal studies were approved by the Institutional Animal Care and Use Committee of Beth Israel Deaconess Medical Center. White adipose tissue (WAT) of Fat-specific insulin receptor knockout (FIRKO) mice were generously provided by Dr. C. Ronald Kahn. The 16-week-old ob/ob and db/db mice were purchased from the Jackson Laboratory. For DIO mice, C57BL/6J mice were fed a high-fat diet (D12331; Research Diets Inc.) for 12 weeks. Body weight and food intake were measured weekly. Mice were subjected to Magnetic Resonance Imaging (MRI) (Echo Medical Systems) to examine body composition.

Generation of Adipoq-Cre transgenic mice

A C57Bl/6 mouse bacterial artificial chromosome (90G21) containing the Adipoq gene was transformed into the recombinogenic EL250 bacteria cells, and homologous recombination was performed as described elsewhere (Lee et al., 2001). The Cre-FRT-Kan-FRT cassette was transformed into the Adipoq BAC host EL250 cells and recombined to insert the Cre ATG into the Adipoq ATG with a resulting deletion of 222 bp of the Adipoq gene. Adipoq-Cre-FRT-Kan-FRT BAC host EL250 clones were identified by PCR screening. The FRT-Kan-FRT cassette was removed and an Adipoq-Cre BAC host EL250 clone without mutation in the Cre coding sequence was obtained. The loxP site present in the vector sequence of the Adipoq-Cre BAC was removed. The transgenic construct was microinjected into pronuclei of fertilized one-cell stage embryos of FVB mice (Jackson Laboratories) using standard methods. Twelve founders were obtained. The Adipoq-Cre founder was backcrossed >8 times onto C57BL/6J prior to use in these experiments. Adipoq-Cre mice have been deposited at the Jackson Laboratory Repository with JAX Stock No. 010803.

Generation of FI4KO mice

Irf4^flox/flox mice were generated as described previously (Klein et al., 2006). Adipoq-Cre mice were mated with Irf4^flox/flox mice, and cohorts were established by mating F1 Irf4^flox/+; Cre⁺ mice to littermate Irf4^flox/+; Cre⁻ mice. Mice were maintained under a 14 hr light/10 hr dark cycle at constant temperature (22°C) with free access to food and water and were fed either a standard chow diet (8664, Harlan Teklad) or a high fat diet (D12331; Research Diets Inc.).

Isolation of primary cells

Adipocytes were isolated from epididymal fat depots in Krebs-Ringer HEPES (KRH) buffer by collagenase digestion as described (Joost and Schurmann, 2001). Peritoneal macrophages from Adipoq-Cre, Irf4^flox/flox mice and FI4KO mice were isolated by peritoneal lavage 4 days after injection of 3% thioglycollate (2.5ml i.p. ; Difco; BD Diagnostics).

Measurement of adipocyte size

Epididymal WAT were isolated and fixed in osmium tetroxide solution (Sigma-Aldrich). The destribution of adipocyte size in WAT were analyzed by a Coulter counter (Multisizer II, Coulter Electronics) as described previously (DiGirolamo and Fine, 2001).

Serum analysis

After collection of serum from tail vein, β-hydroxybutyrate (StanBio) and nonesterified fatty acids (NEFAs) (Wako Chemicals) were measured in duplicate using colorimetric assays.

Cold-induced thermogenesis

Rectal temperature was measured using a rectal probe (Yellow Spring Instruments Co.) in Flox and FI4KO (KO) mice under the cold exposure.

Indirect Calorimetry

Metabolic rate of mice was measured by indirect calorimetry in open-circuit oxymax chambers that are a component of the Comprehensive Lab Animal Monitoring System (CLAMS; Columbus Instruments, Columbus, OH). Mice were housed individually and maintained at 22°C under a 12:12-h light-dark cycle. Food and water were available ad libitum.

Human subjects

Healthy male adults ages 18-60, BMI 19-39 kg/m² without significant medical illness were recruited by advertisement for participation. Individuals with diabetes, heart disease, liver disease, renal insufficiency, thyroid disease, gout or a history of kidney stones were excluded. Individuals with a history of alcohol abuse or marijuana use were also excluded. Subjects reported stable weight (+/− 3kg) for 6 months prior to participation. All subjects underwent a screening visit that included a medical history, physical exam and baseline laboratory tests. Patients gave written informed consent to all clinical investigations, which were performed in accordance with the principles embodied in the Declaration of Helsinki. The protocols were conducted in accordance with the ethical standards of the Committee on Clinical Investigation and were approved by the Beth Israel Deaconess Medical Center Institutional Review Board. The studies were conducted from June, 2007 to April, 2010 at the Beth Israel Deaconess Medical Center, General Clinical Research Center, Boston, MA ClinicalTrials.gov identifier: NCT00476125. Following a screening visit, 5 healthy, lean male subjects reported to the GCRC at 8am for a standard 500 calorie breakfast. One hour after the meal, subjects underwent a subcutaneous fat biopsy and then began fasting. Subjects were typically fasted for 72 hours, except one individual who terminated the fast after 50 hours due to fatigue. Body weight and body composition were measured daily. A second subcutaneous fat biopsy was done at the end of the fasting period, before the subject was fed a standard meal. Clinical data are presented in Supplemental Table 1.

Analysis of Gene Expression by Q-PCR

Total RNA was extracted from cells or tissues using TRIzol reagent (Invitrogen) following the manufacturer’s instructions. First-strand cDNA synthesis was performed using RETROscript (Ambion). Total RNA was converted into first-strand cDNA using oligo(dT) primers as described by the manufacturer. PCR was performed using cDNA synthesized from 1.5 μg total RNA in an Mx3000P Q-PCR system (Stratagene) with specific primers and SYBR Green PCR Master Mix (Stratagene). The relative abundance of mRNAs was standardized using 36B4 mRNA as the invariant control. Primers used are listed in Supplemental Table 2.

Luciferase Reporter Assay

The indicated regions of the murine Irf4, Pnpla2 (ATGL), and Lipe (HSL) promoters were PCR amplified and ligated into pGL3-Basic (Promega). Five days after adipogenic stimulation, 3T3-L1 adipocytes were detached with trypsin and transfected using the Amaxa nucleofection system (Amaxa Biosystems). Transfections were performed using 2 μg of reporter construct along with 100ng of IRF4-pCDH expression vector or EGFP-pCDH, with 100ng of galactosidase expression vector as a transfection control. Luciferase activity was measured 24 hr after transfection using the Galacto-Star luciferase reporter assay (Roche) according to the manufacturer’s instructions.

Chromatin Immunoprecipitation

3T3-L1 cells were treated with 1% formaldehyde for 10 min at room temperature to crosslink DNA-protein complexes. Genomic DNA was sheared using a Sonic Dismembrator Model 100 (Fisher Scientific) to obtain fragments ranging from 200 bp to 1 kb. ChIP was performed using a kit from Upstate, with 3 μg of primary antibody, goat IgG, or rabbit IgG. The following primary antibodies were used: goat anti-IRF4 (sc-6059), or rabbit anti-FoxO1 (sc-11350) (Santa Cruz Biotechnology), Crosslinking was reversed and purified DNA was subjected to Q-PCR using SYBR green fluorescent dye (Stratagene). Primers used are listed in Supplemental Table 2.

Protein Extraction and Western Blot Analysis

Tissue and cell lysates were prepared using RIPA buffer (Boston BioProducts) supplemented with complete protease inhibitor cocktail (Roche). For Western blot analyses, 50 ug of protein was subjected to SDS-PAGE under reducing conditions, transferred, and blotted using the anti-IRF4 antibody described above.

Lipolysis assay

For in vitro studies, adipocytes were washed twice with DMEM and then incubated for 6 h in serum-free DMEM containing 1% fatty acid–free BSA in the presence or absence of 10μM isoproterenol. Conditioned medium was then analyzed for glycerol using the Free Glycerol Determination Kit (Sigma). Adipocytes were then rinsed with PBS and harvested in RIPA buffer. The protein content of cell lysates was determined using the DC Protein Assay kit (BioRad). For ex vivo studies, adipocytes were isolated from epididymal fat pads by collagenase digestion. Cells were resuspended in Krebs-Ringer HEPES (KRH) buffer at 2 × 10⁵ cells/ml. Cells were incubated in the presence or absence of 1μM isoproterenol for 0-2h at 37°C with gentle shaking. Glycerol and NEFA content of conditioned medium was determined as above. For in vivo studies, mice were fasted for 4h and injected with isoproterenol (10 mg/kg body weight). Blood was collected from the tail vein before and 15 min after injection, and glycerol and NEFA content of serum was measured as above.

Lipogenesis assay

Lipogenesis assays were performed as previously described (Wiese et al., 1995). Briefly, after serum starvation, cells were incubated in the presence or absence of 10nM insulin for 15 min, and incubated with 5mM [¹⁴C]-glucose for 1h at 37°C. Cells were then washed three times with ice-cold PBS. Radiolabeled glucose incorporation into lipid was assessed by scraping cells into 1 ml PBS and shaking vigorously with 5ml of Betaflour Liquid Scintillation Fluid (National Diagnostics). After samples settled overnight, radioactivity partitioning into the organic phase was determined by scintillation counting. The protein content of cell lysates was determined using the DC Protein Assay kit.

The authors wish to thank Dr. C.R. Kahn for use of samples from FIRKO mice, Dr. Tak Mak for Irf4^−/− mice, Drs. Shingo Kajimura, Tetsuya Hosooka, Choel Son, Junji Kawashima, Linh Vong, and Mark Herman for technical advice, Dr. Domenico Accili for FoxO1 adenoviral constructs, and members of the Rosen lab for helpful discussion. We thank Drs. Leo Otterbein and Eva Csizmadia for help with alveolar lavage and immunohistochemistry. This work was funded by the Kanae Foundation for the Promotion of Medical Science, American Heart Association Award AHA0825928D (to J.E.), and NIH R01 DK63906 and NIH R01 DK085171 (to E.D.R.).