Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Lack of Interleukin-1 Receptor I (IL-1RI) Protects Mice From High-Fat Diet–Induced Adipose Tissue Inflammation Coincident With Improved Glucose Homeostasis

Lack of Interleukin-1 Receptor I (IL-1RI) Protects Mice From High-Fat Diet–Induced Adipose Tissue... ORIGINAL ARTICLE Lack of Interleukin-1 Receptor I (IL-1RI) Protects Mice From High-Fat Diet–Induced Adipose Tissue Inflammation Coincident With Improved Glucose Homeostasis 1 1 1 1 Fiona C. McGillicuddy, Karen A. Harford, Clare M. Reynolds, Elizabeth Oliver, 1 2 1 Mandy Claessens, Kingston H.G. Mills, and Helen M. Roche OBJECTIVE—High-fat diet (HFD)-induced adipose tissue in- flammation is a critical feature of diet-induced insulin resistance (IR); however, the contribution of interleukin-1 receptor I he emergent pandemic of obesity has demanded (IL-1RI)-mediated signals to this phenotype has not been defined. greater understanding of associated metabolic We hypothesized that lack of IL-1RI may ameliorate HFD-induced complications, including insulin resistance (IR) IR by attenuating adipose tissue inflammation. Tand type 2 diabetes. Immune cell infiltration into adipose during obesity has been documented with initial RESEARCH DESIGN AND METHODS—Glucose homeostasis was monitored in chow- and HFD-fed wild-type (WT) and infiltration of T cells (1–4) followed by macrophages (5,6). 2/2 IL-1RI mice by glucose tolerance and insulin tolerance tests. Proinflammatory cytokine release from infiltrating immune Macrophage recruitment and cytokine signature of adipose tissue cells in turn contributes to a chronic state of inflammation macrophages was evaluated. Insulin sensitivity and cytokine se- (7), with enhanced local secretion of proinflammatory cretion from adipose explants was quantified. Cytokine secretion cytokines, interleukin (IL)-1, IL-6, and tumor necrosis and adipocyte insulin sensitivity was measured in cocultures of 2/2 factor (TNF)-a, from the expanding adipose tissue mass WT or IL-1RI macrophages with 3T3L1 adipocytes. Synergistic (8–11), which in turn induces adipocyte IR in vitro (12–14). effects of IL-1b with tumor necrosis factor (TNF)-a on inflamma- 2/2 Furthermore, abrogation of TNF-a signaling alleviates tion was monitored in WT and IL-1RI bone-marrow macro- phages and adipose explants. high-fat diet (HFD)-induced IR in vivo (15), whereas lack of TLR4 protects against free fatty acid–induced IR (16,17). 2/2 RESULTS—Lean and obese IL-1RI animals exhibited en- These studies suggest that HFD-induced IR is driven by hanced glucose homeostasis by glucose tolerance test and insulin a pathological immune response; however, the contribu- tolerance test. M1/M2 macrophage number in adipose tissue was tion of IL-1 receptor I (IL-1RI) signaling to this phenotype comparable between genotypes; however, TNF-a and IL-6 secre- 2/2 tion was lower from IL-1RI adipose tissue macrophages. has not yet been deciphered, particularly in the context of 2/2 IL-1RI adipose exhibited enhanced insulin sensitivity, elevated adipose tissue biology. pAKT, lower cytokine secretion, and attenuated induction of There is much evidence to suggest that IL-1 has a path- phosphorylated signal transducer and activator of transcription 3 ogenic role in adipose tissue during obesity. Activation of and suppressor of cytokine signaling molecule 3 after HFD. Cocul- 2/2 IL-1b in vivo depends on processing of its pro- to active ture of WT, but not IL-1RI macrophages, with 3T3L1 adipocytes form by the NLRP3-caspase-1 inflammasome complex enhanced IL-6 and TNF-a secretion, reduced adiponectin secre- (18). Genetic deletion of NLRP3 (19), or pharmacological tion, and impaired adipocyte insulin sensitivity. TNF-a and IL-1b potently synergized to enhance inflammation in WT macrophages inhibition of caspase-1 (20), confer protection against and adipose, an effect lost in the absence of IL-1RI. obesity-induced IR, effects potentially attributable to re- duced IL-1b. Downstream IL-1 binds to IL-1RI and acti- CONCLUSIONS—Lack of IL-1RI protects against HFD-induced vates nuclear factor (NF)-kB and Jun NH2-terminal kinase IR coincident with reduced local adipose tissue inflammation, mitogen-activated protein kinase (JNK MAPK) (21,22), despite equivalent immune cell recruitment. Diabetes 60:1688– which have been implicated in IR (23,24). Furthermore, 1698, 2011 TNF-a, TLR4, and IL-1 signaling pathways converge at the inhibitor of NFkB kinase (IKK)/NFkB axis (25), with 2/2 marked protection of IKK mice against IR (26,27). IL-1b also potently induces production of other proinflammatory cytokines, including IL-6. Both IL-1b and IL-6 induce adi- From the Nutrigenomics Research Group, University College Dublin Conway pocyte IR in vitro (13,14,28), with IL-1b also inhibiting Institute, School of Public Health & Population Science, University College Dublin, Dublin, Ireland; and the Immune Regulation Research Group, adipogenesis (29). IL-6, via activation of JAK2/STAT3, po- School of Biochemistry and Immunology, Trinity College Dublin, Dublin, tently induces expression of suppressor of cytokine sig- Ireland. naling molecule (SOCS)-3, which is also associated with IR Corresponding author: Helen M. Roche, [email protected]. Received 9 September 2010 and accepted 18 March 2011. (30–32). Given this large body of evidence against IL-1, we DOI: 10.2337/db10-1278 hypothesized that lack of IL-1RI would protect against This article contains Supplementary Data online at http://diabetes. HFD-induced adipose inflammation and improve glucose diabetesjournals.org/lookup/suppl/doi:10.2337/db10-1278/-/DC1. F.C.M. and K.A.H. contributed equally to this work. homeostasis. 2/2 2011 by the American Diabetes Association. Readers may use this article as The current study demonstrates that IL-1RI mice are long as the work is properly cited, the use is educational and not for profit, partially protected from HFD-induced IR but not obesity. and the work is not altered. See http://creativecommons.org/licenses/by 2/2 -nc-nd/3.0/ for details. The inflammatory profile of IL-1RI adipose explants, 1688 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES were cultured in Dulbecco’smodified Eagle’s medium supplemented with 10% and isolated adipose tissue macrophages (ATMs), was FBS, 1% penicillin/streptomycin, and 30% L929 conditioned medium for 7 days markedly attenuated, coincident with the inability of TNF-a to differentiate to the bone marrow macrophage (BMM) phenotype (34). F4/80 / and IL-1b to synergize and augment inflammation in CD11B ATMs were seeded (200,000 cells/1 mL media) in 24-well plates and 2/2 IL-1RI macrophages and adipose. Attenuated inflamma- incubated in complete media for 24 h in the absence of any stimulus, and media 2/2 tion correlated with improved insulin sensitivity in IL-1RI were harvested for cytokine analysis. NIH-3T3 cells (Panomics, Cambridge, 2/2 adipose. Furthermore, coculture of IL-1RI macrophages U.K.) stably transfected with a luciferase reporter construct within the NFkB promoter region were maintained in complete media + hygromycin B (100 mg/mL) with 3T3L1 adipocytes resulted in a more favorable adipo- (Roche, Clare, Ireland). cytokine signature and maintenance of adipocyte insulin NIH-3T3 NFkB-luciferase activity assay. Conditioned media from adipose sensitivity compared with coculture with wild-type (WT) tissue explants were harvested and incubated with NIH-3T3 cells (13 10 cells/mL) macrophages. This study demonstrates that inflammatory for 6 h. Cells were washed in PBS and lysed, and luciferase levels were mea- signals transmitted via IL-1RI are primary mediators of sured by luminescence (Promega, Southampton, U.K.). adiposetissueinflammation during obesity. IL-1b and TNF-a synergistic studies. NIH-3T3 cells were treated with IL-1b (0.01–10 ng/mL), TNF-a (0.01–10 ng/mL), or IL-1b (0.005–5 ng/mL) and TNF-a (0.005–5 ng/mL) for 6 h, and NFkB-driven luciferase levels were monitored by RESEARCH DESIGN AND METHODS luminescence. The effect of IL-1b and TNF-a on 1) IL-6 protein secretion from 3 2 3T3L1 adipocytes (24-h treatment); 2) IL-6, IL-1, SOCS3, and TNF-a mRNA and Materials. Deoxy-D-glucose 2-[1,2- H(N)] was purchased from Perkin-Elmer 2/2 IL-6 protein secretion from WT and IL-1RI BMMs (24-h treatment); and 3) Analytical Sciences (Dublin, Ireland). Cell culture material was purchased IL-6 mRNA and protein secretion in adipose explants (100 mg tissue/mL com- from Lonza (Slough, U.K.). All other reagents unless otherwise stated were plete media) (6-h treatment) was monitored. from Sigma Aldrich (Dorset, U.K.). 2/2 2/2 Coculture study. WT and IL-1RI BMMs were seeded on Corning transwell Animals. IL-1RI breeding pairs, on a C57BL/6 background, were purchased filters (0.4 mm, Sigma Aldrich) and LPS-stimulated for 30 min. Lipopolysaccharide- from Jackson Laboratories and bred in the University College Dublin for 6–10 containing media was removed, cells were washed in PBS, and fresh media were generations under specific pathogen-free conditions. Ethical approval was added. Transwells were transferred to differentiated 3T3L1 adipocytes (day obtained from the University College Dublin Ethics Committee, and mice were 5 post-differentiation) and left coculture for 5 days. After 72 h, media were maintained according to the regulations of the European Union and the Irish 2/2 harvested for cytokine analysis and fresh media were added. After co- Department of Health. C57BL/6 WT and IL-1RI mice were fed HFD or culture, insulin-stimulated glucose transport into adipocytes was monitored. a nutrient-matched standard-fat chow diet (45% kcal from palm oil and 10% On separate identically treated plates, RNA or protein was harvested for kcal from palm oil, respectively; Research Diets, Brunswick, NJ) starting at further analysis. age 6 weeks for 12 weeks. Body weight and food intake were recorded weekly. In vitro glucose uptake assays. Insulin sensitivity of 3T3L1 adipocytes was Metabolic tests were performed at baseline (mice aged 6–8 weeks) and after monitored in vitro as previously described (35). Briefly, cells were serum 12 weeks of chow or HFD. starved for 4 h, and glucose was starved in PBS/0.2% BSA for 30 min before Intraperitoneal glucose and insulin tolerance tests. Mice were fasted for insulin (100 nmol/L) stimulation for 15 min. [ H]Deoxyglucose (0.5 mCi/mL 6 h and were injected intraperitoneally with 25% (wt/vol) glucose (1.5 g/kg; cold glucose [0.1 mmol/L]) was added for 30 min before washing with PBS. B. Braun Medical, Dublin, Ireland) for glucose tolerance test (GTT) or insulin Cells were lysed in RIPA buffer, and [ H]glucose uptake was measured by (0.75 units/kg; Actrapid, Novo Nordisk, Denmark) for insulin tolerance test liquid scintillation counting. Remaining lysates were stored for immunoblot (ITT), respectively. Glucose levels were monitored at baseline and at in- analysis. Fold increase in glucose uptake over basal (non-insulin–stimulated) dicated time points after the metabolic challenge via tail-vein blood sampling is presented. using a blood glucose meter from Accu-Chek (Roche, Dublin, Ireland). In- General laboratory methods. Detailed descriptions of plasma, gene ex- sulin secretory response was monitored in overnight fasted animals, and pression, immunohistochemistry, and immunoblot analysis are presented in the blood samples were collected at indicated times after glucose load (1.5 g/kg). Supplementary Data. Insulin levels were measured by enzyme-linked immunosorbent assay (ELISA). Statistical analysis. Data are reported as mean 6 SEM. For GTT/ITT studies Body mass composition analysis. Body composition (fat, lean, and water with multiple time points, we performed two-way repeated-measures ANOVA 2/2 to test for differences in means between WT and IL-1RI groups. When content) was analyzed using a Bruker’s minispecLF50bodycomposition ANOVA was significant, post hoc Bonferroni corrected t tests were applied. analyzer according to the manufacturer’s instructions (Bruker Optik, Area under the curve analysis was performed on GTT and ITT curves using Ettlingen, Germany). Graphpad Prism 5 software. For comparison of data between two groups at Stromal vascular fraction isolation and purification. Epididymal fat pads a single time point, unpaired t tests were performed. GraphPad Prism 5 were minced, and adipocytes and stromal vascular fractions were separated by (GraphPad Software, San Diego, CA) was used for statistical analyses. Sta- collagenase (2 mg/mL) digestion. Stromal vascular cells were filtered, blocked tistical significance is presented as *P , 0.05, **P , 0.01, and ***P , 0.001 in in PBS/2% BSA, and stained with fluorescently labeled antibodies: F4/80-FITC, all figures. CD11B-AF647/PE, CD11C-RPE, CD3-APC, CD4-FITC, CD8-PE, or CD117-PE as indicated (AbD Serotec, Kidlington, U.K.). Unstained single stains and fluo- rescence minus one (FMO) controls were used for setting compensations RESULTS and gates. Flow cytometry was performed on a Dako CyAn ADP platform 2/2 (Beckman-Coulter, Clare, Ireland) and analyzed using Summit v4.3 software. IL-1RI mice exhibit greater glucose tolerance and + + In separate studies, F4/80 /CD11B cells were isolated from stromal vascular are partially protected from HFD-induced IR. Glucose 2/2 fraction using a BD FACSAria cell sorter (BD Biosciences, Oxfordshire, U.K.). homeostasis was markedly improved in IL-1RI com- 2/2 Ex vivo adipose tissue culture. Adipose tissue from WT and IL-1RI mice pared with WT mice, both at baseline (lean, 6–8 weeks old) was harvested at baseline (aged 6–8 weeks) and after HFD. Adipose explants and after 12 weeks of HFD by GTT (Fig. 1A and C) and ITT were placed in 24-well plates (100 mg tissue/well) with 1 mL complete media (Fig. 1B and D). The GTTs for age-matched chow-fed and (Dulbecco’smodified Eagle’s medium, 10% FBS, and 1% penicillin/streptomycin) 2/2 for 24 h. Media were harvested, and cytokine secretion (TNF-a, IL-6, and IL-1b) HFD-fed IL-1RI mice were equivalent and significantly was analyzed by ELISA (Quantikine kits; R&D Systems Europe, Abingdon, lower than those for HFD WT mice (Supplementary Fig. U.K.). 1A and B). Insulin secretion in response to glucose load Insulin-stimulated glucose uptake transport into adipose tissue 2/2 was significantly higher in lean IL-1RI mice than in WT explants. Lean and obese adipose explants were placed in 24-well plates mice, despite similar fasting levels (Fig. 1E). HFD-fed (100 mg tissue/well) in PBS + 0.2% BSA before stimulation 6 insulin (100 nmol/L) 2/2 3 3 IL-1RI mice exhibited higher fasting insulin levels for 15 min. [ H]Glucose (0.1 mmol/L 2-deoxyglucose + 0.5 mCi/mL [ H]deoxy- glucose) was added for 45 min. Tissue was washed, lysed in radioimmuno- compared with WT mice, but had a similar degree of in- precipitation assay (RIPA) buffer, and homogenized using a tissue lyser sulin secretion upon glucose load. There was no difference 2/2 (Qiagen, West Sussex, U.K.). An aliquot of lysate was used for liquid scintil- in total body weight between IL-1RI and WT mice lation counting, whereas the remaining aliquot was stored for immunoblot at baseline or after HFD (Fig. 1F), although epididymal analysis. and visceral adipose tissue depot weight was greater in Cell culture. 3T3L1 fibroblasts (European Collection of Cell Cultures, Salisbury, 2/2 obese IL-1RI animals (Fig. 1G). Body mass composition U.K.) were differentiated to adipocytes as previously described (33). Bone 2/2 marrow monocytes, isolated from femurs and tibias of WT and IL-1RI mice, analysis confirmed this increased fat mass (Fig. 1H). The diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1689 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 2/2 2/2 FIG. 1. GTT and ITT in WT and IL-1RI mice at baseline and after HFD. A: GTT (1.5 g/kg glucose) in 6-h fasted lean and obese WT and IL-1RI 2/2 2/2 2/2 animals (●, WT lean; ○, IL-1RI lean; ■, WT obese; □, IL-1RI obese; ***P < 0.001 with respect to [w.r.t.] IL-1RI obese; ###P < 0.001 w.r.t. 2/2 2/2 2/2 IL-1RI lean, n =27–31). B: ITT (0.75 units/kg insulin) in 6-h fasted lean and obese WT and IL-1RI animals (●, WT lean; ○, IL-1RI lean; ■, 2/2 2/2 WT obese; □, IL-1RI obese; *P < 0.05, ***P < 0.001 w.r.t. IL-1RI obese, n =22–31). C and D: Area under the curve (AUC) for lean and obese animals over course of GTT and ITT was calculated and is expressed as arbitrary units (AU) (*P < 0.5, ***P < 0.001 w.r.t. corresponding WT, ##P < 2/2 0.01, ###P < 0.001 w.r.t. corresponding lean, n =22–31). E: Insulin secretion response in overnight fasted lean and obese WT and IL-1RI animals 2/2 after intraperitoneal injection with 1.5 g/kg glucose (*P < 0.05 w.r.t. IL-1RI lean; #P < 0.05, ###P < 0.001 w.r.t. corresponding lean). F: Weight of animals at time of metabolic challenge (###P < 0.001 w.r.t. lean, n =22–31). G: Weight of epididymal adipose tissue (EAT) and visceral adipose 2/2 tissue (VAT) depots of lean and obese WT and IL-1RI mice (###P < 0.001 w.r.t. lean; *P < 0.05, ***P < 0.001 w.r.t. WT, n =9–12). H: Body mass composition analysis in lean and obese animals (*P < 0.05 w.r.t. WT obese, n =8). rate of weight gain on HFD (Supplementary Fig. 1D) and Altered immunogenic phenotype of ATMs from 2/2 food intake (data not shown) was comparable between IL-1RI mice despite equivalent macrophage strains. Fasting plasma insulin, leptin, triacylglycerol, and recruitment. Recruitment of macrophages into adipose nonesterified fatty acids were increased coincident with tissue was monitored to establish potential mechanisms of 2/2 reduced levels of adiponectin after HFD in both genotypes improved glucose homeostasis in IL-1RI mice. Whereas compared with lean counterparts (Table 1). Plasma IL-6 the number of adipose tissue M1 and M2 macrophages and TNF-a increased with HFD but was not significantly increased with HFD, little difference in recruitment different between genotypes. was evident between genotypes (Fig. 2A and B and 1690 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES TABLE 1 Plasma metabolic profile 2/2 2/2 WT lean WT obese IL-1RI lean IL-1RI obese Insulin (mg/L) 0.62 6 0.07 1.66 6 0.19‡ 0.55 6 0.09 2.90 6 0.6‖‡ Leptin (ng/mL) 1,091.1 6 171.2 20,944.3 6 2,623.8‡ 1,152.7 6 254.5 2,5631.1 6 1,178.6‡ TNF-a (ng/mL) 16.93 6 4.88 24.47 6 4.11 8.42 6 3.49 28.07 6 4.69† IL-6 (ng/mL) 419.1 6 131.7 1,054.2 6 188.8 146.4 6 52.7 1,060.6 6 214.4* Nonesterified fatty acid (mmol/L) 1.44 6 0.19 1.79 6 0.4 1.29 6 0.09 2.17 6 0.17* Triacylglycerol (mmol/L) 0.75 6 0.07 1.84 6 0.41† 1.02 6 0.18 2.09 6 0.2* Adiponectin (ng/mL) 2,733.9 6 365.4 1,654.1 6 189.9* 3,084.1 6 324.4 1,857.1 6 219.5* 2/2 Plasma was isolated from WT and IL-1RI mice by cardiac puncture, and levels of metabolic markers were analyzed enzymatically. *P , 0.05 with respect to (w.r.t.) respective lean. †P , 0.01 w.r.t. respective lean. ‡P , 0.001 w.r.t. respective lean. ‖P , 0.01 w.r.t. WT obese. Supplementary Fig. 2C). Furthermore, there was no differ- perilipin, and fatty acid synthase mRNA (Supplementary ence in the number of T cells or mast cells (Supplementary Fig. 3D). IL-1b also markedly induced IL-6 mRNA and Fig. 2A and B). We next monitored cytokine secretion protein secretion from 3T3L1 adipocytes (Supplementary 2/2 from obese-derived WT and IL-1RI ATMs and demon- Fig. 3E and F). IL-6 was also capable of inducing adipocyte strated markedly reduced IL-6 and TNF-a secretion by IR (Supplementary Fig. 3G) (13) and adipocyte lipolysis 2/2 ATMs from IL-1RI compared with WT mice (Fig. 2D (28). and E). Altered adipocyte-macrophage cross-talk between 2/2 2/2 Attenuated inflammatory profile of IL-1RI adipose IL-1RI macrophages and 3T3L1 adipocytes. Given 2/2 tissue compared with WT ex vivo. We monitored adi- the improved insulin-sensitive profile of IL-1RI adipose, 2/2 pose tissue inflammation in explants from lean and obese we speculated that the reduced immunogenicity of IL-1RI 2/2 WT and IL-1RI mice. Secretion of IL-6, TNF-a, and IL-1b macrophages may result in improved cross-talk with resi- was significantly higher from obese adipose explants dent adipocytes and preservation of adipocyte functional- compared with lean explants (Fig. 3A, C, and E); however, ity. Because the number of ATMs is limited for such 2/2 2/2 cytokine secretion from IL-1RI explants was signifi- mechanistic studies, we determined the effect of IL-1RI cantly reduced compared with WT (Fig. 3A and C). Fur- versus WT BMMs on adipocyte biology. BMMs from 2/2 thermore, expression of IL-6, TNF-a,and SOCS3 mRNA IL-1RI mice had significantly reduced basal expression 2/2 (Fig. 3B, D, and F) was lower in adipose from IL-1RI of IL-6, SOCS3,and IL-10 mRNA compared with WT (Fig. mice. HFD-mediated phosphorylation of STAT3 and up- 5H). Furthermore, IL-1b induced IL-6, SOCS3, and IL-10 regulation of the SOCS3 protein were also attenuated in mRNA expression in WT macrophages but had no effect 2/2 2/2 obese IL-1RI adipose, whereas there was no difference on IL-1RI BMMs (Supplementary Fig. 4A–D). There- 2/2 in serine-phosphorylated insulin receptor substrate (IRS)-1 fore, IL-1RI BMMs were similarly skewed as the ex vivo (Fig. 3G). Conditioned media from obese adipose explants ATMs toward a less inflammatory phenotype and thus induced greater activation of NFkB compared with lean represented a relevant macrophage phenotype to de- 2/2 explants; nonetheless, obese IL-1RI media had a signif- termine effects on adipocyte biology. icantly reduced capacity to drive NFkB compared with WT Coculture of WT BMMs with adipocytes significantly (Fig. 3H). reduced insulin-stimulated glucose transport into adipo- 2/2 2/2 IL-1RI adipose tissue exhibits greater insulin sen- cytes, while coculture with IL-1RI BMM resulted in sitivity than WT ex vivo. We next sought to estab- partial preservation of adipocyte insulin sensitivity (Fig. 5A 2/2 lish whether the attenuated inflammatory phenotype of and B). Coculture of WT BMMs, but not IL-1RI BMMs, 2/2 IL-1RI adipose would translate to improved adipose- with adipocytes significantly increased IL-6 and TNF-a specific insulin sensitivity. Insulin-stimulated [ H]glucose secretion, with a concomitant reduction in secretion of uptake was significantly higher into adipose of both lean insulin-sensitizing adiponectin (Fig. 5C–E). Furthermore 2/2 2/2 and obese IL-1RI mice compared with WT mice (Fig. 4A), WT, but not IL-1RI , BMMs reduced adipocyte GLUT4, coincident with increased insulin-induced tyrosine phos- coincident with increased IL-6 mRNA expression (Fig. 5F phorylation of IRS-1 (Fig. 4B). Expression of phosphor- and G). 2/2 ylated AKT was also elevated in lean and obese IL-1RI Synergy between IL-1 and TNF signaling pathways adipose; however, levels of phosphorylated insulin re- lost in the absence of IL-1RI. To further probe the 2/2 ceptor did not differ between strains (Fig. 4C and D). mechanisms of attenuated inflammation within IL-1RI Expression of IRS-1 and GLUT4 mRNA was significantly adipose and ATMs, we determined the synergistic effect of 2/2 higher in lean IL-1RI adipose compared with WT; how- IL-1b and TNF-a in driving adipocyte/macrophage and ever, no difference between genotypes was evident after adipose tissue inflammation, focusing on IL-6, since this HFD (Fig. 4E and F). Given this improved insulin-sensitive was the most abundant ex vivo adipokine. IL-1b in com- 2/2 phenotype of IL-1RI adipose, we further characterized bination with TNF-a induced greater activation of NFkB the direct effect of IL-1b on adipocyte biology in vitro. than either cytokine alone (Fig. 6A). There was marked Similar to previous studies (13,14), IL-1b dose-dependently synergy between IL-1b and TNF-a in promoting IL-6 se- attenuated insulin-stimulated glucose transport into 3T3L1 cretion by adipocytes (Fig. 6B). The synergistic effect of adipocytes (Supplementary Fig. 3A), coincident with down- IL-1b and TNF-a on IL-6 secretion was ablated in BMM 2/2 regulation of IRS-1 and GLUT4 mRNA and protein from IL-1RI mice (Fig. 6C). TNF-a and IL-1b also syn- (Supplementary Fig. 3B and C) and downregulation of ergized to enhance IL-6, IL-1b,and SOCS3, but not TNF-a, peroxisome proliferator–activated receptor (PPAR)-g, mRNA expression in WT BMMs (Fig. 6D–G), effects that 2/2 sterol regulatory element–binding protein (SREBP)-1c, were also abolished in IL-1RI BMMs. We also examined diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1691 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 2/2 FIG. 2. Recruitment of WT and IL-1RI macrophages into adipose tissue and immunogenic phenotype of isolated ATMs. Adipose tissue was 2/2 harvested from WT and IL-1RI animals and stromal vascular cells (SVCs) separated from tissue by collagenase digestion. SVCs were labeled + + + with antibodies to F4/80, CD11B, and CD11C and analyzed by flow cytometry. Cells triple positive (F4/80 /CD11B /CD11C ) were classified as M1 + + 2 macrophages, whereas cells double positive (F4/80 /CD11B /CD11C ) were classified as M2 macrophages. Infiltration of M1 (A) and M2 (B) macrophages into adipose tissue was monitored in lean and obese animals and is presented as percentage of total SVC (**P < 0.01, ***P < 0.001 + + with respect to [w.r.t.] lean, n =10–30). C–E: F4/80 /CD11B macrophages were sorted from obese adipose tissue and were seeded in complete media in 24-well plates (200,000 cells/mL media) and left culture for 24 h. Media were harvested and levels of IL-6 (D) and TNF-a (E) were measured by ELISA (***P < 0.001 w.r.t. WT, n =6). the effects of IL-1b and TNF-a on adipose IL-6 mRNA ex- IL-1 in regulation of energy homeostasis and weight 2/2 pression and protein secretion ex vivo and demonstrated (37,39). García et al. (37) showed IL-1RI mice main- 2/2 a loss of synergy in IL-1RI explants (Fig. 6H and I). tained on a chow diet develop mature-onset obesity Furthermore, the ability of TNF-a alone to induce IL-6 was coincident with moderate glucose intolerance and hyper- 2/2 2/2 reduced in IL-1RI adipose. insulinemia. On the contrary, IL-1Ra mice exhibit a lean phenotype with increased energy expenditure and hypo- insulinemia (39,40). These studies argue for a protective DISCUSSION effect of IL-1 against weight gain; however, differences Adipose tissue dysfunction due to excessive paracrine in- in insulin sensitivity in these mouse models could be flammation is widely recognized as the earliest and argu- solely attributable to differences in weight. In contrast, ably most important step in the etiology of obesity-induced NOD mice lacking IL-1RI exhibit slower progression whole-body IR. In the current study, we demonstrate that but not complete inhibition of diabetes (36), an effect signaling through the IL-1RI is a key mediator of potentially attributable to inhibition of IL-1b–mediated HFD-induced inflammation in adipose tissue with partial pancreatic b-cell destruction (41,42). Furthermore, and 2/2 protection of IL-1RI mice from HFD-induced IR. consistent with our findings, treatment of HFD-fed mice There have been conflicting findings to date on the role with IL-1Ra improved glucose homeostasis and enhanced of IL-1 during obesity and in the development of diabetes insulin secretion, with protection of b-cells from HFD- (36–38), which has been confounded by a central role for mediated toxicity (38). However, the effect of IL-1Ra on 1692 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES 2/2 FIG. 3. Inflammatory signature of WT and IL-1RI adipose. Adipose explants were cultured for 24 h in serum-containing media (100 mg/mL media) and levels of proinflammatory IL-6 (A), TNF-a (C), and IL-1b secretion into media (E) were measured by ELISA (*P < 0.05, **P < 0.01, ***P < 0.001 with respect to [w.r.t.] WT; #P < 0.05, ##P < 0.01, ###P < 0.001 w.r.t. corresponding lean, n =8–12). Adipose tissue mRNA levels of IL-6 (B), TNF-a (D), and SOCS3 (F) were analyzed by real-time PCR (*P < 0.5, **P < 0.01, ***P < 0.001 w.r.t. WT; #P < 0.05, ###P < 0.001 w.r.t. 2/2 lean, n =8). G: WT and IL-1RI adipose tissue protein lysates at weeks 0 and 12 post-HFD were probed for levels of phosphorylated STAT3, SOCS3, serine-phosphorylated IRS-1, and b-actin by immunoblot analysis. H: Adipose-tissue conditioned media were applied for 6 h to NIH-3T3 cells stably expressing an NFkB-luciferase promoter-reporter construct to monitor NFkB activity. Levels of NFkB-driven luciferase were quan- tified using a luminometer (*P < 0.05 w.r.t. corresponding WT; #P < 0.05, ###P < 0.001 w.r.t. lean AT media, n =8). adipose tissue biology and insulin sensitivity was not weeks. Intervention in high-risk prediabetic obese cohorts assessed. Recent human trials have shown that treatment may yield more promising results, wherein IL-1RI in- with IL-1Ra (anakinra) improves glycemia and b-cell se- hibition may delay, if not prevent, development of type 2 cretory function, coincident with reduced systemic in- diabetes by attenuating adipose inflammation and pro- flammation in patients with type 2 diabetes (43). This trial tecting functional b-cells. Our study adds insight into the did not demonstrate improved insulin sensitivity after 13 potential initial contribution of IL-1RI signaling during diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1693 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 2/2 FIG. 4. Ex vivo evaluation of adipose tissue insulin sensitivity. A: Adipose explants from lean and obese WT and IL-1RI mice were harvested and 3 3 stimulated 6 insulin (100 nmol/L) ex vivo for 15 min. Tracer [ H]glucose was added to explants for an additional 45 min. Levels of [ H]glucose were measured by liquid scintillation counting. Fold increase in [ H]glucose transport in response to insulin over basal (non–insulin-stimulated) into adipose tissue for each individual mouse was calculated and is presented (##P < 0.01, ###P < 0.001 w.r.t. lean; *P < 0.05, **P < 0.01 with 2/2 respect to [w.r.t.] WT; $$$P < 0.001 w.r.t. basal, n =8). B–D: Protein lysates were prepared from adipose tissue of lean and obese WT and IL-1RI mice after stimulation 6 insulin (100 nmol/L) for 1 h ex vivo. A: Levels of tyrosine phosphorylated IRS-1 in adipose protein lysates were measured using a PathScan ELISA kit (#P < 0.05 w.r.t. WT lean; *P < 0.05 w.r.t. WT obese; $P < 0.05, $$P < 0.01 w.r.t. basal, n = 8). Levels of phosphorylated AKT, phosphorylated insulin receptor, whole-cell AKT, and b-actin in lean (C) and obese (D) adipose protein lysates were assessed by immunoblot analysis. Adipose tissue mRNA levels of IRS-1 (E) and GLUT4 (F) were analyzed by real-time PCR analysis (**P < 0.01, ***P < 0.001 w.r.t. lean; ##P < 0.01, ###P < 0.001 w.r.t. WT, n =8). early phases of obesity-induced IR, particularly in adipose This phenotype was age-dependent; glucose tolerance was 2/2 tissue. equivalent in older chow-fed IL-1RI and WT mice, co- 2/2 Our in vivo findings demonstrate that lack of IL-1RI incident with enhanced weight gain in the IL-1RI mice as 2/2 partially protects mice from HFD-induced IR, with im- previously reported (37). Interestingly, HFD-fed IL-1RI proved glucose homeostasis and enhanced insulin sensi- mice did not develop further IR compared with age- 2/2 2/2 tivity observed in obese IL-1RI compared with WT mice. matched chow-fed IL-1RI mice, despite much greater Plasma concentrations of triacylglycerol, nonesterified weight gain, providing further evidence of protection against fatty acid, IL-6, and TNF-a did not differ between strains; HFD-induced IR but not obesity. however, consistent with a previous report (37), levels of Immune cell recruitment into adipose plays a key role 2/2 insulin were elevated in obese IL-1RI mice. This hyper- in the etiology of HFD-induced IR (6), prompting us to 2/2 insulinemia was not associated with IR and appears to be speculate that reduced immune cell number in IL-1RI an intrinsic characteristic of the animal model given the adipose may partially account for improved glucose ho- 2/2 converse hypoinsulinemic phenotype of IL-1Ra mice meostasis. Contrary to our hypothesis, the number of tis- 2/2 2/2 (39). Prediet lean IL-1RI mice cleared glucose more sue M1/M2 in adipose was not different between IL-1RI efficiently than WT during GTT, which may be attributable and WT mice after HFD, demonstrating little role for IL-1 to enhanced insulin secretion in response to glucose load, in immune cell chemotaxis. We next speculated the cyto- 2/2 together with heightened adipose tissue insulin sensitivity. kine signature of ATMs may be altered in IL-1RI mice. 1694 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES 2/2 2/2 FIG. 5. Cross-talk of WT and IL-1RI BMMs with 3T3L1 adipocytes. WT and IL-1RI BMMs, seeded on transwell (0.4 mmol/L thickness) inserts, were stimulated with LPS (100 ng/mL) for 30 min, washed with PBS before addition of fresh media, and transferred to 3T3L1 adipocyte plates where cells were left to coculture for 5 days. A: The effect of BMMs on insulin (100 nmol/L)-stimulated [ H]glucose transport into 3T3L1 adi- pocytes was evaluated. Fold increase in [ H]glucose transport into adipocytes in response to insulin over basal (non-insulin–stimulated) is pre- sented (*P < 0.05, ***P < 0.001 with respect to [w.r.t.] individual basal; $$$P < 0.001 w.r.t. control + insulin; ##P < 0.01 w.r.t. WT BMM + insulin, n =4). B: Protein lysates from cocultured adipocytes stimulated with insulin (100 nmol/L) for 15 min were harvested and probed for levels of phosphorylated AKT, whole-cell AKT, and b-actin by immunoblot analysis. C–E: Media from cocultured cells were harvested after 72 h and levels of TNF-a (C), adiponectin (D), and IL-6 (E) were measured by ELISA (*P < 0.05, ***P < 0.001 w.r.t. adipocytes alone, n =4). F and G: Effect of 2/2 coculture with WT or IL-1RI BMMs on adipocyte IL-6 (F) and Glut4 (G) mRNA expression by real-time PCR (**P < 0.01, ***P < 0.001 w.r.t. 2/2 2/2 control; #P < 0.05, ###P < 0.001 w.r.t. IL-1RI coculture, n =4). H: Baseline inflammatory mRNA signature of WT and IL-1RI BMMs. Basal 2/2 mRNA levels of IL-6, IL-10, TNF-a, SOCS3, SOCS1, caspase 1, and TLR4 in WT and IL-1RI BMMs as assessed by real-time PCR (*P < 0.05, **P < 0.01, ***P < 0.001 w.r.t. WT, n =4). diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1695 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE FIG. 6. IL-1b and TNF-a synergize to enhance adipose tissue inflammation and NFkB activity. A: NIH-3T3 cells stably expressing an NFkB- luciferase promoter-reporter construct were treated with increasing concentrations of IL-1b or TNF-a alone or both in combination for 6 h. Levels of NFkB-driven luciferase were quantified using a luminometer (***P < 0.001 with respect to [w.r.t.] control; ###P < 0.001 w.r.t. TNF alone; $$$P < 0.001 w.r.t. either cytokine alone, n =4). B: 3T3L1 adipocytes were treated 6 IL-1b6 TNF-a for 24 h, and levels of IL-6 secreted were measured by 2/2 ELISA (***P < 0.001 w.r.t. control; $$$P < 0.001 w.r.t. either cytokine alone, n =4). C–G: BMMs were harvested from WT and IL-1RI mice and stimulated 6 IL-1b6 TNF-a (10 ng/mL) for 24 h, and levels of IL-6 (C) secreted measured by ELISA and IL-6 (D), SOCS3 (E), IL-1b (F), and TNFa mRNA (G) were measured by real-time PCR (**P < 0.01, ***P < 0.001 w.r.t. control; #P < 0.05, ##P < 0.01, ###P < 0.001 w.r.t. WT; $P < 0.05, $$P < 0.01, $$$P < 0.001 w.r.t. either cytokine alone, n =3). H and I: Adipose tissue explants (100 mg) were harvested from lean WT and IL- 2/2 1RI mice and were cultured for 6 h 6 IL-1b6 TNF-a (10 ng/mL) and levels of IL-6 secretion (H) and IL-6 mRNA (I) were determined (***P < 0.001 w.r.t. control; #P < 0.05, ###P < 0.001 w.r.t. WT; $$$P < 0.001 w.r.t. either cytokine alone, n =5). 1696 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES Robust cytokine secretion from WT ATMs was evident suggesting that these cytokines work in concert to create after 24 h in culture in the absence of any external stim- a positive feedback loop. Interestingly, TNF-a mRNA ex- ulus, indicative of preactivation within the inflamed envi- pression was not enhanced, suggesting differential regula- ronment of the obese tissue before culture ex vivo. By tory mechanisms for TNF-a mRNA. These findings suggest 2/2 contrast, cytokine secretion from IL-1RI ATMs was that blocking IL-1 signaling not only ameliorates the ad- markedly lower, suggestive of altered immunogenicity. A verse effects of IL-1 on adipocyte/macrophage biology, but recent report demonstrated removal of HFD from WT also prevents the synergy between TNF-a and IL-1b, thus obese animals improved insulin sensitivity before reduc- alleviating adipose tissue inflammation. tion in macrophage number; however, the proinflamma- Recent reports demonstrating activation of the NLRP3 tory phenotype of these ATMs was markedly attenuated inflammasome following an HFD and protection of upon removal of HFD, demonstrating phenotypic plasticity inflammasome-compromised mice against HFD-induced IR (44). It is therefore arguable that it is not the number of (19,20) have given new insight into the upstream mecha- macrophages per se that determine poor outcome, but the nisms through which IL-1 is elevated during obesity. We activation state, cytokine signature, and functional effects have focused on the downstream consequences of IL-1RI 2/2 of ATMs that dictate severity of IR. Upon culture, IL-1RI signaling within adipose tissue during an HFD and have 2/2 ATMs secreted markedly reduced levels of both TNF-a and demonstrated that IL-1RI mice are protected against IL-6, and it is thus likely that altered immunogenic phe- HFD-induced IR coincident with attenuated adipose tissue 2/2 notype of recruited IL-1RI immune cells results in im- and ATM inflammation. There are numerous local cellular proved cross-talk with resident adipocytes and improved targets and functional consequences of IL-1 released from insulin sensitivity. Consistent with this, we demonstrated adipose tissue, including preadipocytes (inhibition of 2/2 improved cytokine cross-talk of IL-1RI BMMs with 3T3L1 adipogenesis), mature adipocytes (induction of IR), and adipocytes in vitro, coincident with an attenuated effect on infiltrated immune cells (enhanced cytokine secretion), all adipocyte insulin sensitivity compared with WT BMMs. of which may contribute to adipose tissue IR. However, Given the prominent role for adipose tissue dysfunction our study is limited in its ability to distinguish whether lack in early stages of obesity-induced IR and the attenuated of IL-1RI signaling in immune cells or in adipocyte/ 2/2 immunogenic profile of IL-1RI ATMs, we next hypoth- preadipocyte fractions is primarily responsible for this im- esized that IL-1RI–transduced signals may contribute to proved phenotype. Metabolic studies in animals with ei- HFD-induced inflammation within whole adipose tissue. ther immune-specific or adipocyte-specific deletion of Previous findings in our laboratory highlighted attenuated IL-1RI would in turn provide greater insight into these inflammatory and cellular stress–related proteomic signa- mechanistic queries. ture exclusive to adipose tissue rather than skeletal mus- In conclusion, we ascribe a pathogenic role to IL-1RI– 2/2 cle or liver in obese IL-1RI mice compared with WT mediated signals during HFD-induced adipose tissue dys- 2/2 mice (45). The present work extends our understanding of function and IR. Improved glucose homeostasis in IL-1RI the paracrine effect associated with lack of IL-1RI in adi- mice correlated with a marked reduction in local adipose pose tissue with reduced IL-6 and TNF-a secretion from inflammation, with altered ATM cytokine signatures and explants coincident with reduced mRNA expression of pronounced attenuation in both IL-6 secretion and activa- TNF-a, IL-6, and SOCS3. Furthermore, we demonstrate tion of the IL-6 regulated proteins STAT3 and SOCS3. Loss 2/2 reduced activation of STAT3 coincident with reduced of synergy between TNF-a and IL-1b in IL-1RI adipose 2/2 SOCS3 protein in IL-1RI adipose, an effect potentially mayinturnhave contributedtothisattenuatedinflammatory attributable to reduced IL-6 secretion. This effect reduced profile. This study highlights the critical and nonredundant local adipose tissue inflammation in turn correlated with contribution of IL-1 to the medley of proinflammatory 2/2 improved insulin sensitivity of IL-1RI adipose ex vivo. insulin-desensitizing signals relevant to obesity-induced IR. Interestingly, this attenuated inflammatory profile was lo- calized to adipose tissue with equivalent concentrations of ACKNOWLEDGMENTS 2/2 systemic cytokines in IL-1RI and WT mice. The dis- This work was supported by the Science Foundation crepancy between systemic and local cytokine levels is Ireland PI Programme (06/IM.1/B105) (to H.M.R.). unclear, but may be attributable to almost undetectable No potential conflicts of interest relevant to this article plasma IL-1b levels in WT animals, even with obesity were reported. (46,47), compared with the concentration of IL-1b in F.C.M. researched data and wrote the manuscript. K.A.H. macrophages/adipose tissue. and C.M.R. researched data and proofread the manuscript. Whereas IL-1RI signaling clearly plays a nonredun- E.O. and M.C. researched data. K.H.G.M. reviewed data dant role during HFD-induced IR, it may be an oversim- and edited the manuscript. H.M.R. designed the study, re- plification to attribute the majority of adipose tissue viewed data, and edited the manuscript. inflammation to the actions of one cytokine. It is much The authors appreciate the technical support from the more likely that adipose-derived cytokines cross-talk to core technologies staff in the Conway Institute, University amplify adipose tissue inflammation. Consistent with this College Dublin, Belfield, Dublin, Ireland, and the technical hypothesis, we demonstrated a potent synergistic effect of support of Dr. Eileen Murphy during body composition the major adipocytokines IL-1b and TNF-a in enhancing analysis in the Alimentary Pharmabiotic Center, University NFkB activation and adipose IL-6 protein secretion. This College Cork, Cork, Ireland. synergistic effect on macrophage and adipose IL-6 secre- tion was lost in the absence of IL-1RI. Furthermore, the REFERENCES ability of TNF-a to induce IL-6 secretion was blunted in 2/2 1. Winer S, Chan Y, Paltser G, et al. Normalization of obesity-associated in- IL-1RI adipose, suggesting that TNF-a–induced IL-6 is sulin resistance through immunotherapy. Nat Med 2009;15:921–929 in part mediated via IL-1. Remarkably, TNF-a and IL-1b 2. Nishimura S, Manabe I, Nagasaki M, et al. CD8+ effector T cells contribute synergized to enhance IL-1b mRNA expression in WT to macrophage recruitment and adipose tissue inflammation in obesity. BMMs far beyond the effects of either cytokine alone, Nat Med 2009;15:914–920 diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1697 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 3. Kintscher U, Hartge M, Hess K, et al. T-lymphocyte infiltration in visceral 25. Verstrepen L, Bekaert T, Chau TL, Tavernier J, Chariot A, Beyaert R. TLR-4, adipose tissue: a primary event in adipose tissue inflammation and the IL-1R and TNF-R signaling to NF-kappaB: variations on a common theme. development of obesity-mediated insulin resistance. Arterioscler Thromb Cell Mol Life Sci 2008;65:2964–2978 Vasc Biol 2008;28:1304–1310 26. Arkan MC, Hevener AL, Greten FR, et al. IKK-beta links inflammation to 4. Rausch ME, Weisberg S, Vardhana P, Tortoriello DV. Obesity in C57BL/6J obesity-induced insulin resistance. Nat Med 2005;11:191–198 mice is characterized by adipose tissue hypoxia and cytotoxic T-cell in- 27. Yuan M, Konstantopoulos N, Lee J, et al. Reversal of obesity- and diet- filtration. Int J Obes (Lond) 2008;32:451–463 induced insulin resistance with salicylates or targeted disruption of Ikk- 5. Lumeng CN, Deyoung SM, Bodzin JL, Saltiel AR. Increased inflammatory beta. Science 2001;293:1673–1677 properties of adipose tissue macrophages recruited during diet-induced 28. Yang Y, Ju D, Zhang M, Yang G. Interleukin-6 stimulates lipolysis in porcine obesity. Diabetes 2007;56:16–23 adipocytes. Endocrine 2008;33:261–269 6. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW 29. Lagathu C, Yvan-Charvet L, Bastard JP, et al. Long-term treatment with Jr. Obesity is associated with macrophage accumulation in adipose tissue. interleukin-1beta induces insulin resistance in murine and human adipo- J Clin Invest 2003;112:1796–1808 cytes. Diabetologia 2006;49:2162–2173 7. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. 30. Emanuelli B, Peraldi P, Filloux C, et al. SOCS-3 inhibits insulin signaling J Clin Invest 2006;116:1793–1801 and is up-regulated in response to tumor necrosis factor-alpha in the adi- 8. Kern PA, Saghizadeh M, Ong JM, Bosch RJ, Deem R, Simsolo RB. The pose tissue of obese mice. J Biol Chem 2001;276:47944–47949 expression of tumor necrosis factor in human adipose tissue. Regulation 31. Ueki K, Kondo T, Kahn CR. Suppressor of cytokine signaling 1 (SOCS-1) by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest and SOCS-3 cause insulin resistance through inhibition of tyrosine phos- 1995;95:2111–2119 phorylation of insulin receptor substrate proteins by discrete mechanisms. 9. Fenkci S, Rota S, Sabir N, Sermez Y, Guclu A, Akdag B. Relationship of Mol Cell Biol 2004;24:5434–5446 serum interleukin-6 and tumor necrosis factor alpha levels with abdominal 32. Niemand C, Nimmesgern A, Haan S, et al. Activation of STAT3 by IL-6 and fat distribution evaluated by ultrasonography in overweight or obese IL-10 in primary human macrophages is differentially modulated by sup- postmenopausal women. J Investig Med 2006;54:455–460 pressor of cytokine signaling 3. J Immunol 2003;170:3263–3272 10. Wellen KE, Hotamisligil GS. Inflammation, stress, and diabetes. J Clin 33. Elmendorf JS, Chen D, Pessin JE. Guanosine 59-O-(3-thiotriphosphate) Invest 2005;115:1111–1119 (GTPgammaS) stimulation of GLUT4 translocation is tyrosine kinase- 11. Nov O, Kohl A, Lewis EC, et al. Interleukin-1b may mediate insulin re- dependent. J Biol Chem 1998;273:13289–13296 sistance in liver-derived cells in response to adipocyte inflammation. En- 34. Wang X, Collins HL, Ranalletta M, et al. Macrophage ABCA1 and ABCG1, docrinology 2010;151:4247–4256 but not SR-BI, promote macrophage reverse cholesterol transport in vivo. 12. Hotamisligil GS, Murray DL, Choy LN, Spiegelman BM. Tumor necrosis J Clin Invest 2007;117:2216–2224 factor alpha inhibits signaling from the insulin receptor. Proc Natl Acad Sci 35. McGillicuddy FC, Chiquoine EH, Hinkle CC, et al. Interferon gamma U S A 1994;91:4854–4858 attenuates insulin signaling, lipid storage, and differentiation in human 13. Tanaka T, Itoh H, Doi K, et al. Down regulation of peroxisome proliferator- adipocytes via activation of the JAK/STAT pathway. J Biol Chem 2009;284: activated receptorgamma expression by inflammatory cytokines and its 31936–31944 reversal by thiazolidinediones. Diabetologia 1999;42:702–710 36. Thomas HE, Irawaty W, Darwiche R, et al. IL-1 receptor deficiency slows 14. Jager J, Grémeaux T, Cormont M, Le Marchand-Brustel Y, Tanti JF. progression to diabetes in the NOD mouse. Diabetes 2004;53:113–121 Interleukin-1beta-induced insulin resistance in adipocytes through down- 37. García MC, Wernstedt I, Berndtsson A, et al. Mature-onset obesity in regulation of insulin receptor substrate-1 expression. Endocrinology 2007; interleukin-1 receptor I knockout mice. Diabetes 2006;55:1205–1213 148:241–251 38. Sauter NS, Schulthess FT, Galasso R, Castellani LW, Maedler K. The anti- 15. Uysal KT, Wiesbrock SM, Marino MW, Hotamisligil GS. Protection from inflammatory cytokine interleukin-1 receptor antagonist protects from obesity-induced insulin resistance in mice lacking TNF-alpha function. high-fat diet-induced hyperglycemia. Endocrinology 2008;149:2208–2218 Nature 1997;389:610–614 39. Somm E, Henrichot E, Pernin A, et al. Decreased fat mass in interleukin-1 16. Poggi M, Bastelica D, Gual P, et al. C3H/HeJ mice carrying a toll-like re- receptor antagonist-deficient mice: impact on adipogenesis, food intake, ceptor 4 mutation are protected against the development of insulin re- and energy expenditure. Diabetes 2005;54:3503–3509 sistance in white adipose tissue in response to a high-fat diet. Diabetologia 40. Matsuki T, Horai R, Sudo K, Iwakura Y. IL-1 plays an important role in lipid 2007;50:1267–1276 metabolism by regulating insulin levels under physiological conditions. 17. Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 links innate J Exp Med 2003;198:877–888 immunity and fatty acid-induced insulin resistance. J Clin Invest 2006;116: 41. Bendtzen K, Mandrup-Poulsen T, Nerup J, Nielsen JH, Dinarello CA, 3015–3025 Svenson M. Cytotoxicity of human pI 7 interleukin-1 for pancreatic islets of 18. Mills KH, Dunne A. Immune modulation: IL-1, master mediator or initiator Langerhans. Science 1986;232:1545–1547 of inflammation. Nat Med 2009;15:1363–1364 42. Maedler K, Sergeev P, Ris F, et al. Glucose-induced beta cell production of 19. Vandanmagsar B, Youm YH, Ravussin A, et al. The NLRP3 inflammasome IL-1beta contributes to glucotoxicity in human pancreatic islets. J Clin instigates obesity-induced inflammation and insulin resistance. Nat Med Invest 2002;110:851–860 2011;17:179–188 43. Larsen CM, Faulenbach M, Vaag A, et al. Interleukin-1-receptor antagonist 20. Stienstra R, Joosten LA, Koenen T, et al. The inflammasome-mediated in type 2 diabetes mellitus. N Engl J Med 2007;356:1517–1526 caspase-1 activation controls adipocyte differentiation and insulin sensi- 44. Li P, Lu M, Nguyen MT, et al. Functional heterogeneity of CD11c-positive tivity. Cell Metab 2010;12:593–605 adipose tissue macrophages in diet-induced obese mice. J Biol Chem 2010; 21. Maedler K, Dharmadhikari G, Schumann DM, Størling J. Interleukin-1 beta 285:15333–15345 targeted therapy for type 2 diabetes. Expert Opin Biol Ther 2009;9:1177– 45. de Roos B, Rungapamestry V, Ross K, et al. Attenuation of inflammation 1188 and cellular stress-related pathways maintains insulin sensitivity in obese 22. Stylianou E, Saklatvala J. Interleukin-1. Int J Biochem Cell Biol 1998;30: type I interleukin-1 receptor knockout mice on a high-fat diet. Proteomics 1075–1079 2009;9:3244–3256 23. Hirosumi J, Tuncman G, Chang L, et al. A central role for JNK in obesity 46. Jung C, Gerdes N, Fritzenwanger M, Figulla HR. Circulating levels of and insulin resistance. Nature 2002;420:333–336 interleukin-1 family cytokines in overweight adolescents. Mediators In- 24. Shoelson SE, Lee J, Yuan M. Inflammation and the IKK beta/I kappa B/NF- flamm 2010;2010:958403 kappa B axis in obesity- and diet-induced insulin resistance. Int J Obes 47. van der Poll T, Marchant A, Buurman WA, et al. Endogenous IL-10 protects Relat Metab Disord 2003;27(Suppl. 3):S49–S52 mice from death during septic peritonitis. J Immunol 1995;155:5397–5401 1698 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diabetes Pubmed Central

Lack of Interleukin-1 Receptor I (IL-1RI) Protects Mice From High-Fat Diet–Induced Adipose Tissue Inflammation Coincident With Improved Glucose Homeostasis

Download PDF
11 pages

Loading next page...
 
/lp/pubmed-central/lack-of-interleukin-1-receptor-i-il-1ri-protects-mice-from-high-fat-Q10kY0xA0M

References (56)

Publisher
Pubmed Central
Copyright
© 2011 by the American Diabetes Association.
ISSN
0012-1797
eISSN
1939-327X
DOI
10.2337/db10-1278
Publisher site
See Article on Publisher Site

Abstract

ORIGINAL ARTICLE Lack of Interleukin-1 Receptor I (IL-1RI) Protects Mice From High-Fat Diet–Induced Adipose Tissue Inflammation Coincident With Improved Glucose Homeostasis 1 1 1 1 Fiona C. McGillicuddy, Karen A. Harford, Clare M. Reynolds, Elizabeth Oliver, 1 2 1 Mandy Claessens, Kingston H.G. Mills, and Helen M. Roche OBJECTIVE—High-fat diet (HFD)-induced adipose tissue in- flammation is a critical feature of diet-induced insulin resistance (IR); however, the contribution of interleukin-1 receptor I he emergent pandemic of obesity has demanded (IL-1RI)-mediated signals to this phenotype has not been defined. greater understanding of associated metabolic We hypothesized that lack of IL-1RI may ameliorate HFD-induced complications, including insulin resistance (IR) IR by attenuating adipose tissue inflammation. Tand type 2 diabetes. Immune cell infiltration into adipose during obesity has been documented with initial RESEARCH DESIGN AND METHODS—Glucose homeostasis was monitored in chow- and HFD-fed wild-type (WT) and infiltration of T cells (1–4) followed by macrophages (5,6). 2/2 IL-1RI mice by glucose tolerance and insulin tolerance tests. Proinflammatory cytokine release from infiltrating immune Macrophage recruitment and cytokine signature of adipose tissue cells in turn contributes to a chronic state of inflammation macrophages was evaluated. Insulin sensitivity and cytokine se- (7), with enhanced local secretion of proinflammatory cretion from adipose explants was quantified. Cytokine secretion cytokines, interleukin (IL)-1, IL-6, and tumor necrosis and adipocyte insulin sensitivity was measured in cocultures of 2/2 factor (TNF)-a, from the expanding adipose tissue mass WT or IL-1RI macrophages with 3T3L1 adipocytes. Synergistic (8–11), which in turn induces adipocyte IR in vitro (12–14). effects of IL-1b with tumor necrosis factor (TNF)-a on inflamma- 2/2 Furthermore, abrogation of TNF-a signaling alleviates tion was monitored in WT and IL-1RI bone-marrow macro- phages and adipose explants. high-fat diet (HFD)-induced IR in vivo (15), whereas lack of TLR4 protects against free fatty acid–induced IR (16,17). 2/2 RESULTS—Lean and obese IL-1RI animals exhibited en- These studies suggest that HFD-induced IR is driven by hanced glucose homeostasis by glucose tolerance test and insulin a pathological immune response; however, the contribu- tolerance test. M1/M2 macrophage number in adipose tissue was tion of IL-1 receptor I (IL-1RI) signaling to this phenotype comparable between genotypes; however, TNF-a and IL-6 secre- 2/2 tion was lower from IL-1RI adipose tissue macrophages. has not yet been deciphered, particularly in the context of 2/2 IL-1RI adipose exhibited enhanced insulin sensitivity, elevated adipose tissue biology. pAKT, lower cytokine secretion, and attenuated induction of There is much evidence to suggest that IL-1 has a path- phosphorylated signal transducer and activator of transcription 3 ogenic role in adipose tissue during obesity. Activation of and suppressor of cytokine signaling molecule 3 after HFD. Cocul- 2/2 IL-1b in vivo depends on processing of its pro- to active ture of WT, but not IL-1RI macrophages, with 3T3L1 adipocytes form by the NLRP3-caspase-1 inflammasome complex enhanced IL-6 and TNF-a secretion, reduced adiponectin secre- (18). Genetic deletion of NLRP3 (19), or pharmacological tion, and impaired adipocyte insulin sensitivity. TNF-a and IL-1b potently synergized to enhance inflammation in WT macrophages inhibition of caspase-1 (20), confer protection against and adipose, an effect lost in the absence of IL-1RI. obesity-induced IR, effects potentially attributable to re- duced IL-1b. Downstream IL-1 binds to IL-1RI and acti- CONCLUSIONS—Lack of IL-1RI protects against HFD-induced vates nuclear factor (NF)-kB and Jun NH2-terminal kinase IR coincident with reduced local adipose tissue inflammation, mitogen-activated protein kinase (JNK MAPK) (21,22), despite equivalent immune cell recruitment. Diabetes 60:1688– which have been implicated in IR (23,24). Furthermore, 1698, 2011 TNF-a, TLR4, and IL-1 signaling pathways converge at the inhibitor of NFkB kinase (IKK)/NFkB axis (25), with 2/2 marked protection of IKK mice against IR (26,27). IL-1b also potently induces production of other proinflammatory cytokines, including IL-6. Both IL-1b and IL-6 induce adi- From the Nutrigenomics Research Group, University College Dublin Conway pocyte IR in vitro (13,14,28), with IL-1b also inhibiting Institute, School of Public Health & Population Science, University College Dublin, Dublin, Ireland; and the Immune Regulation Research Group, adipogenesis (29). IL-6, via activation of JAK2/STAT3, po- School of Biochemistry and Immunology, Trinity College Dublin, Dublin, tently induces expression of suppressor of cytokine sig- Ireland. naling molecule (SOCS)-3, which is also associated with IR Corresponding author: Helen M. Roche, [email protected]. Received 9 September 2010 and accepted 18 March 2011. (30–32). Given this large body of evidence against IL-1, we DOI: 10.2337/db10-1278 hypothesized that lack of IL-1RI would protect against This article contains Supplementary Data online at http://diabetes. HFD-induced adipose inflammation and improve glucose diabetesjournals.org/lookup/suppl/doi:10.2337/db10-1278/-/DC1. F.C.M. and K.A.H. contributed equally to this work. homeostasis. 2/2 2011 by the American Diabetes Association. Readers may use this article as The current study demonstrates that IL-1RI mice are long as the work is properly cited, the use is educational and not for profit, partially protected from HFD-induced IR but not obesity. and the work is not altered. See http://creativecommons.org/licenses/by 2/2 -nc-nd/3.0/ for details. The inflammatory profile of IL-1RI adipose explants, 1688 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES were cultured in Dulbecco’smodified Eagle’s medium supplemented with 10% and isolated adipose tissue macrophages (ATMs), was FBS, 1% penicillin/streptomycin, and 30% L929 conditioned medium for 7 days markedly attenuated, coincident with the inability of TNF-a to differentiate to the bone marrow macrophage (BMM) phenotype (34). F4/80 / and IL-1b to synergize and augment inflammation in CD11B ATMs were seeded (200,000 cells/1 mL media) in 24-well plates and 2/2 IL-1RI macrophages and adipose. Attenuated inflamma- incubated in complete media for 24 h in the absence of any stimulus, and media 2/2 tion correlated with improved insulin sensitivity in IL-1RI were harvested for cytokine analysis. NIH-3T3 cells (Panomics, Cambridge, 2/2 adipose. Furthermore, coculture of IL-1RI macrophages U.K.) stably transfected with a luciferase reporter construct within the NFkB promoter region were maintained in complete media + hygromycin B (100 mg/mL) with 3T3L1 adipocytes resulted in a more favorable adipo- (Roche, Clare, Ireland). cytokine signature and maintenance of adipocyte insulin NIH-3T3 NFkB-luciferase activity assay. Conditioned media from adipose sensitivity compared with coculture with wild-type (WT) tissue explants were harvested and incubated with NIH-3T3 cells (13 10 cells/mL) macrophages. This study demonstrates that inflammatory for 6 h. Cells were washed in PBS and lysed, and luciferase levels were mea- signals transmitted via IL-1RI are primary mediators of sured by luminescence (Promega, Southampton, U.K.). adiposetissueinflammation during obesity. IL-1b and TNF-a synergistic studies. NIH-3T3 cells were treated with IL-1b (0.01–10 ng/mL), TNF-a (0.01–10 ng/mL), or IL-1b (0.005–5 ng/mL) and TNF-a (0.005–5 ng/mL) for 6 h, and NFkB-driven luciferase levels were monitored by RESEARCH DESIGN AND METHODS luminescence. The effect of IL-1b and TNF-a on 1) IL-6 protein secretion from 3 2 3T3L1 adipocytes (24-h treatment); 2) IL-6, IL-1, SOCS3, and TNF-a mRNA and Materials. Deoxy-D-glucose 2-[1,2- H(N)] was purchased from Perkin-Elmer 2/2 IL-6 protein secretion from WT and IL-1RI BMMs (24-h treatment); and 3) Analytical Sciences (Dublin, Ireland). Cell culture material was purchased IL-6 mRNA and protein secretion in adipose explants (100 mg tissue/mL com- from Lonza (Slough, U.K.). All other reagents unless otherwise stated were plete media) (6-h treatment) was monitored. from Sigma Aldrich (Dorset, U.K.). 2/2 2/2 Coculture study. WT and IL-1RI BMMs were seeded on Corning transwell Animals. IL-1RI breeding pairs, on a C57BL/6 background, were purchased filters (0.4 mm, Sigma Aldrich) and LPS-stimulated for 30 min. Lipopolysaccharide- from Jackson Laboratories and bred in the University College Dublin for 6–10 containing media was removed, cells were washed in PBS, and fresh media were generations under specific pathogen-free conditions. Ethical approval was added. Transwells were transferred to differentiated 3T3L1 adipocytes (day obtained from the University College Dublin Ethics Committee, and mice were 5 post-differentiation) and left coculture for 5 days. After 72 h, media were maintained according to the regulations of the European Union and the Irish 2/2 harvested for cytokine analysis and fresh media were added. After co- Department of Health. C57BL/6 WT and IL-1RI mice were fed HFD or culture, insulin-stimulated glucose transport into adipocytes was monitored. a nutrient-matched standard-fat chow diet (45% kcal from palm oil and 10% On separate identically treated plates, RNA or protein was harvested for kcal from palm oil, respectively; Research Diets, Brunswick, NJ) starting at further analysis. age 6 weeks for 12 weeks. Body weight and food intake were recorded weekly. In vitro glucose uptake assays. Insulin sensitivity of 3T3L1 adipocytes was Metabolic tests were performed at baseline (mice aged 6–8 weeks) and after monitored in vitro as previously described (35). Briefly, cells were serum 12 weeks of chow or HFD. starved for 4 h, and glucose was starved in PBS/0.2% BSA for 30 min before Intraperitoneal glucose and insulin tolerance tests. Mice were fasted for insulin (100 nmol/L) stimulation for 15 min. [ H]Deoxyglucose (0.5 mCi/mL 6 h and were injected intraperitoneally with 25% (wt/vol) glucose (1.5 g/kg; cold glucose [0.1 mmol/L]) was added for 30 min before washing with PBS. B. Braun Medical, Dublin, Ireland) for glucose tolerance test (GTT) or insulin Cells were lysed in RIPA buffer, and [ H]glucose uptake was measured by (0.75 units/kg; Actrapid, Novo Nordisk, Denmark) for insulin tolerance test liquid scintillation counting. Remaining lysates were stored for immunoblot (ITT), respectively. Glucose levels were monitored at baseline and at in- analysis. Fold increase in glucose uptake over basal (non-insulin–stimulated) dicated time points after the metabolic challenge via tail-vein blood sampling is presented. using a blood glucose meter from Accu-Chek (Roche, Dublin, Ireland). In- General laboratory methods. Detailed descriptions of plasma, gene ex- sulin secretory response was monitored in overnight fasted animals, and pression, immunohistochemistry, and immunoblot analysis are presented in the blood samples were collected at indicated times after glucose load (1.5 g/kg). Supplementary Data. Insulin levels were measured by enzyme-linked immunosorbent assay (ELISA). Statistical analysis. Data are reported as mean 6 SEM. For GTT/ITT studies Body mass composition analysis. Body composition (fat, lean, and water with multiple time points, we performed two-way repeated-measures ANOVA 2/2 to test for differences in means between WT and IL-1RI groups. When content) was analyzed using a Bruker’s minispecLF50bodycomposition ANOVA was significant, post hoc Bonferroni corrected t tests were applied. analyzer according to the manufacturer’s instructions (Bruker Optik, Area under the curve analysis was performed on GTT and ITT curves using Ettlingen, Germany). Graphpad Prism 5 software. For comparison of data between two groups at Stromal vascular fraction isolation and purification. Epididymal fat pads a single time point, unpaired t tests were performed. GraphPad Prism 5 were minced, and adipocytes and stromal vascular fractions were separated by (GraphPad Software, San Diego, CA) was used for statistical analyses. Sta- collagenase (2 mg/mL) digestion. Stromal vascular cells were filtered, blocked tistical significance is presented as *P , 0.05, **P , 0.01, and ***P , 0.001 in in PBS/2% BSA, and stained with fluorescently labeled antibodies: F4/80-FITC, all figures. CD11B-AF647/PE, CD11C-RPE, CD3-APC, CD4-FITC, CD8-PE, or CD117-PE as indicated (AbD Serotec, Kidlington, U.K.). Unstained single stains and fluo- rescence minus one (FMO) controls were used for setting compensations RESULTS and gates. Flow cytometry was performed on a Dako CyAn ADP platform 2/2 (Beckman-Coulter, Clare, Ireland) and analyzed using Summit v4.3 software. IL-1RI mice exhibit greater glucose tolerance and + + In separate studies, F4/80 /CD11B cells were isolated from stromal vascular are partially protected from HFD-induced IR. Glucose 2/2 fraction using a BD FACSAria cell sorter (BD Biosciences, Oxfordshire, U.K.). homeostasis was markedly improved in IL-1RI com- 2/2 Ex vivo adipose tissue culture. Adipose tissue from WT and IL-1RI mice pared with WT mice, both at baseline (lean, 6–8 weeks old) was harvested at baseline (aged 6–8 weeks) and after HFD. Adipose explants and after 12 weeks of HFD by GTT (Fig. 1A and C) and ITT were placed in 24-well plates (100 mg tissue/well) with 1 mL complete media (Fig. 1B and D). The GTTs for age-matched chow-fed and (Dulbecco’smodified Eagle’s medium, 10% FBS, and 1% penicillin/streptomycin) 2/2 for 24 h. Media were harvested, and cytokine secretion (TNF-a, IL-6, and IL-1b) HFD-fed IL-1RI mice were equivalent and significantly was analyzed by ELISA (Quantikine kits; R&D Systems Europe, Abingdon, lower than those for HFD WT mice (Supplementary Fig. U.K.). 1A and B). Insulin secretion in response to glucose load Insulin-stimulated glucose uptake transport into adipose tissue 2/2 was significantly higher in lean IL-1RI mice than in WT explants. Lean and obese adipose explants were placed in 24-well plates mice, despite similar fasting levels (Fig. 1E). HFD-fed (100 mg tissue/well) in PBS + 0.2% BSA before stimulation 6 insulin (100 nmol/L) 2/2 3 3 IL-1RI mice exhibited higher fasting insulin levels for 15 min. [ H]Glucose (0.1 mmol/L 2-deoxyglucose + 0.5 mCi/mL [ H]deoxy- glucose) was added for 45 min. Tissue was washed, lysed in radioimmuno- compared with WT mice, but had a similar degree of in- precipitation assay (RIPA) buffer, and homogenized using a tissue lyser sulin secretion upon glucose load. There was no difference 2/2 (Qiagen, West Sussex, U.K.). An aliquot of lysate was used for liquid scintil- in total body weight between IL-1RI and WT mice lation counting, whereas the remaining aliquot was stored for immunoblot at baseline or after HFD (Fig. 1F), although epididymal analysis. and visceral adipose tissue depot weight was greater in Cell culture. 3T3L1 fibroblasts (European Collection of Cell Cultures, Salisbury, 2/2 obese IL-1RI animals (Fig. 1G). Body mass composition U.K.) were differentiated to adipocytes as previously described (33). Bone 2/2 marrow monocytes, isolated from femurs and tibias of WT and IL-1RI mice, analysis confirmed this increased fat mass (Fig. 1H). The diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1689 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 2/2 2/2 FIG. 1. GTT and ITT in WT and IL-1RI mice at baseline and after HFD. A: GTT (1.5 g/kg glucose) in 6-h fasted lean and obese WT and IL-1RI 2/2 2/2 2/2 animals (●, WT lean; ○, IL-1RI lean; ■, WT obese; □, IL-1RI obese; ***P < 0.001 with respect to [w.r.t.] IL-1RI obese; ###P < 0.001 w.r.t. 2/2 2/2 2/2 IL-1RI lean, n =27–31). B: ITT (0.75 units/kg insulin) in 6-h fasted lean and obese WT and IL-1RI animals (●, WT lean; ○, IL-1RI lean; ■, 2/2 2/2 WT obese; □, IL-1RI obese; *P < 0.05, ***P < 0.001 w.r.t. IL-1RI obese, n =22–31). C and D: Area under the curve (AUC) for lean and obese animals over course of GTT and ITT was calculated and is expressed as arbitrary units (AU) (*P < 0.5, ***P < 0.001 w.r.t. corresponding WT, ##P < 2/2 0.01, ###P < 0.001 w.r.t. corresponding lean, n =22–31). E: Insulin secretion response in overnight fasted lean and obese WT and IL-1RI animals 2/2 after intraperitoneal injection with 1.5 g/kg glucose (*P < 0.05 w.r.t. IL-1RI lean; #P < 0.05, ###P < 0.001 w.r.t. corresponding lean). F: Weight of animals at time of metabolic challenge (###P < 0.001 w.r.t. lean, n =22–31). G: Weight of epididymal adipose tissue (EAT) and visceral adipose 2/2 tissue (VAT) depots of lean and obese WT and IL-1RI mice (###P < 0.001 w.r.t. lean; *P < 0.05, ***P < 0.001 w.r.t. WT, n =9–12). H: Body mass composition analysis in lean and obese animals (*P < 0.05 w.r.t. WT obese, n =8). rate of weight gain on HFD (Supplementary Fig. 1D) and Altered immunogenic phenotype of ATMs from 2/2 food intake (data not shown) was comparable between IL-1RI mice despite equivalent macrophage strains. Fasting plasma insulin, leptin, triacylglycerol, and recruitment. Recruitment of macrophages into adipose nonesterified fatty acids were increased coincident with tissue was monitored to establish potential mechanisms of 2/2 reduced levels of adiponectin after HFD in both genotypes improved glucose homeostasis in IL-1RI mice. Whereas compared with lean counterparts (Table 1). Plasma IL-6 the number of adipose tissue M1 and M2 macrophages and TNF-a increased with HFD but was not significantly increased with HFD, little difference in recruitment different between genotypes. was evident between genotypes (Fig. 2A and B and 1690 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES TABLE 1 Plasma metabolic profile 2/2 2/2 WT lean WT obese IL-1RI lean IL-1RI obese Insulin (mg/L) 0.62 6 0.07 1.66 6 0.19‡ 0.55 6 0.09 2.90 6 0.6‖‡ Leptin (ng/mL) 1,091.1 6 171.2 20,944.3 6 2,623.8‡ 1,152.7 6 254.5 2,5631.1 6 1,178.6‡ TNF-a (ng/mL) 16.93 6 4.88 24.47 6 4.11 8.42 6 3.49 28.07 6 4.69† IL-6 (ng/mL) 419.1 6 131.7 1,054.2 6 188.8 146.4 6 52.7 1,060.6 6 214.4* Nonesterified fatty acid (mmol/L) 1.44 6 0.19 1.79 6 0.4 1.29 6 0.09 2.17 6 0.17* Triacylglycerol (mmol/L) 0.75 6 0.07 1.84 6 0.41† 1.02 6 0.18 2.09 6 0.2* Adiponectin (ng/mL) 2,733.9 6 365.4 1,654.1 6 189.9* 3,084.1 6 324.4 1,857.1 6 219.5* 2/2 Plasma was isolated from WT and IL-1RI mice by cardiac puncture, and levels of metabolic markers were analyzed enzymatically. *P , 0.05 with respect to (w.r.t.) respective lean. †P , 0.01 w.r.t. respective lean. ‡P , 0.001 w.r.t. respective lean. ‖P , 0.01 w.r.t. WT obese. Supplementary Fig. 2C). Furthermore, there was no differ- perilipin, and fatty acid synthase mRNA (Supplementary ence in the number of T cells or mast cells (Supplementary Fig. 3D). IL-1b also markedly induced IL-6 mRNA and Fig. 2A and B). We next monitored cytokine secretion protein secretion from 3T3L1 adipocytes (Supplementary 2/2 from obese-derived WT and IL-1RI ATMs and demon- Fig. 3E and F). IL-6 was also capable of inducing adipocyte strated markedly reduced IL-6 and TNF-a secretion by IR (Supplementary Fig. 3G) (13) and adipocyte lipolysis 2/2 ATMs from IL-1RI compared with WT mice (Fig. 2D (28). and E). Altered adipocyte-macrophage cross-talk between 2/2 2/2 Attenuated inflammatory profile of IL-1RI adipose IL-1RI macrophages and 3T3L1 adipocytes. Given 2/2 tissue compared with WT ex vivo. We monitored adi- the improved insulin-sensitive profile of IL-1RI adipose, 2/2 pose tissue inflammation in explants from lean and obese we speculated that the reduced immunogenicity of IL-1RI 2/2 WT and IL-1RI mice. Secretion of IL-6, TNF-a, and IL-1b macrophages may result in improved cross-talk with resi- was significantly higher from obese adipose explants dent adipocytes and preservation of adipocyte functional- compared with lean explants (Fig. 3A, C, and E); however, ity. Because the number of ATMs is limited for such 2/2 2/2 cytokine secretion from IL-1RI explants was signifi- mechanistic studies, we determined the effect of IL-1RI cantly reduced compared with WT (Fig. 3A and C). Fur- versus WT BMMs on adipocyte biology. BMMs from 2/2 thermore, expression of IL-6, TNF-a,and SOCS3 mRNA IL-1RI mice had significantly reduced basal expression 2/2 (Fig. 3B, D, and F) was lower in adipose from IL-1RI of IL-6, SOCS3,and IL-10 mRNA compared with WT (Fig. mice. HFD-mediated phosphorylation of STAT3 and up- 5H). Furthermore, IL-1b induced IL-6, SOCS3, and IL-10 regulation of the SOCS3 protein were also attenuated in mRNA expression in WT macrophages but had no effect 2/2 2/2 obese IL-1RI adipose, whereas there was no difference on IL-1RI BMMs (Supplementary Fig. 4A–D). There- 2/2 in serine-phosphorylated insulin receptor substrate (IRS)-1 fore, IL-1RI BMMs were similarly skewed as the ex vivo (Fig. 3G). Conditioned media from obese adipose explants ATMs toward a less inflammatory phenotype and thus induced greater activation of NFkB compared with lean represented a relevant macrophage phenotype to de- 2/2 explants; nonetheless, obese IL-1RI media had a signif- termine effects on adipocyte biology. icantly reduced capacity to drive NFkB compared with WT Coculture of WT BMMs with adipocytes significantly (Fig. 3H). reduced insulin-stimulated glucose transport into adipo- 2/2 2/2 IL-1RI adipose tissue exhibits greater insulin sen- cytes, while coculture with IL-1RI BMM resulted in sitivity than WT ex vivo. We next sought to estab- partial preservation of adipocyte insulin sensitivity (Fig. 5A 2/2 lish whether the attenuated inflammatory phenotype of and B). Coculture of WT BMMs, but not IL-1RI BMMs, 2/2 IL-1RI adipose would translate to improved adipose- with adipocytes significantly increased IL-6 and TNF-a specific insulin sensitivity. Insulin-stimulated [ H]glucose secretion, with a concomitant reduction in secretion of uptake was significantly higher into adipose of both lean insulin-sensitizing adiponectin (Fig. 5C–E). Furthermore 2/2 2/2 and obese IL-1RI mice compared with WT mice (Fig. 4A), WT, but not IL-1RI , BMMs reduced adipocyte GLUT4, coincident with increased insulin-induced tyrosine phos- coincident with increased IL-6 mRNA expression (Fig. 5F phorylation of IRS-1 (Fig. 4B). Expression of phosphor- and G). 2/2 ylated AKT was also elevated in lean and obese IL-1RI Synergy between IL-1 and TNF signaling pathways adipose; however, levels of phosphorylated insulin re- lost in the absence of IL-1RI. To further probe the 2/2 ceptor did not differ between strains (Fig. 4C and D). mechanisms of attenuated inflammation within IL-1RI Expression of IRS-1 and GLUT4 mRNA was significantly adipose and ATMs, we determined the synergistic effect of 2/2 higher in lean IL-1RI adipose compared with WT; how- IL-1b and TNF-a in driving adipocyte/macrophage and ever, no difference between genotypes was evident after adipose tissue inflammation, focusing on IL-6, since this HFD (Fig. 4E and F). Given this improved insulin-sensitive was the most abundant ex vivo adipokine. IL-1b in com- 2/2 phenotype of IL-1RI adipose, we further characterized bination with TNF-a induced greater activation of NFkB the direct effect of IL-1b on adipocyte biology in vitro. than either cytokine alone (Fig. 6A). There was marked Similar to previous studies (13,14), IL-1b dose-dependently synergy between IL-1b and TNF-a in promoting IL-6 se- attenuated insulin-stimulated glucose transport into 3T3L1 cretion by adipocytes (Fig. 6B). The synergistic effect of adipocytes (Supplementary Fig. 3A), coincident with down- IL-1b and TNF-a on IL-6 secretion was ablated in BMM 2/2 regulation of IRS-1 and GLUT4 mRNA and protein from IL-1RI mice (Fig. 6C). TNF-a and IL-1b also syn- (Supplementary Fig. 3B and C) and downregulation of ergized to enhance IL-6, IL-1b,and SOCS3, but not TNF-a, peroxisome proliferator–activated receptor (PPAR)-g, mRNA expression in WT BMMs (Fig. 6D–G), effects that 2/2 sterol regulatory element–binding protein (SREBP)-1c, were also abolished in IL-1RI BMMs. We also examined diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1691 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 2/2 FIG. 2. Recruitment of WT and IL-1RI macrophages into adipose tissue and immunogenic phenotype of isolated ATMs. Adipose tissue was 2/2 harvested from WT and IL-1RI animals and stromal vascular cells (SVCs) separated from tissue by collagenase digestion. SVCs were labeled + + + with antibodies to F4/80, CD11B, and CD11C and analyzed by flow cytometry. Cells triple positive (F4/80 /CD11B /CD11C ) were classified as M1 + + 2 macrophages, whereas cells double positive (F4/80 /CD11B /CD11C ) were classified as M2 macrophages. Infiltration of M1 (A) and M2 (B) macrophages into adipose tissue was monitored in lean and obese animals and is presented as percentage of total SVC (**P < 0.01, ***P < 0.001 + + with respect to [w.r.t.] lean, n =10–30). C–E: F4/80 /CD11B macrophages were sorted from obese adipose tissue and were seeded in complete media in 24-well plates (200,000 cells/mL media) and left culture for 24 h. Media were harvested and levels of IL-6 (D) and TNF-a (E) were measured by ELISA (***P < 0.001 w.r.t. WT, n =6). the effects of IL-1b and TNF-a on adipose IL-6 mRNA ex- IL-1 in regulation of energy homeostasis and weight 2/2 pression and protein secretion ex vivo and demonstrated (37,39). García et al. (37) showed IL-1RI mice main- 2/2 a loss of synergy in IL-1RI explants (Fig. 6H and I). tained on a chow diet develop mature-onset obesity Furthermore, the ability of TNF-a alone to induce IL-6 was coincident with moderate glucose intolerance and hyper- 2/2 2/2 reduced in IL-1RI adipose. insulinemia. On the contrary, IL-1Ra mice exhibit a lean phenotype with increased energy expenditure and hypo- insulinemia (39,40). These studies argue for a protective DISCUSSION effect of IL-1 against weight gain; however, differences Adipose tissue dysfunction due to excessive paracrine in- in insulin sensitivity in these mouse models could be flammation is widely recognized as the earliest and argu- solely attributable to differences in weight. In contrast, ably most important step in the etiology of obesity-induced NOD mice lacking IL-1RI exhibit slower progression whole-body IR. In the current study, we demonstrate that but not complete inhibition of diabetes (36), an effect signaling through the IL-1RI is a key mediator of potentially attributable to inhibition of IL-1b–mediated HFD-induced inflammation in adipose tissue with partial pancreatic b-cell destruction (41,42). Furthermore, and 2/2 protection of IL-1RI mice from HFD-induced IR. consistent with our findings, treatment of HFD-fed mice There have been conflicting findings to date on the role with IL-1Ra improved glucose homeostasis and enhanced of IL-1 during obesity and in the development of diabetes insulin secretion, with protection of b-cells from HFD- (36–38), which has been confounded by a central role for mediated toxicity (38). However, the effect of IL-1Ra on 1692 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES 2/2 FIG. 3. Inflammatory signature of WT and IL-1RI adipose. Adipose explants were cultured for 24 h in serum-containing media (100 mg/mL media) and levels of proinflammatory IL-6 (A), TNF-a (C), and IL-1b secretion into media (E) were measured by ELISA (*P < 0.05, **P < 0.01, ***P < 0.001 with respect to [w.r.t.] WT; #P < 0.05, ##P < 0.01, ###P < 0.001 w.r.t. corresponding lean, n =8–12). Adipose tissue mRNA levels of IL-6 (B), TNF-a (D), and SOCS3 (F) were analyzed by real-time PCR (*P < 0.5, **P < 0.01, ***P < 0.001 w.r.t. WT; #P < 0.05, ###P < 0.001 w.r.t. 2/2 lean, n =8). G: WT and IL-1RI adipose tissue protein lysates at weeks 0 and 12 post-HFD were probed for levels of phosphorylated STAT3, SOCS3, serine-phosphorylated IRS-1, and b-actin by immunoblot analysis. H: Adipose-tissue conditioned media were applied for 6 h to NIH-3T3 cells stably expressing an NFkB-luciferase promoter-reporter construct to monitor NFkB activity. Levels of NFkB-driven luciferase were quan- tified using a luminometer (*P < 0.05 w.r.t. corresponding WT; #P < 0.05, ###P < 0.001 w.r.t. lean AT media, n =8). adipose tissue biology and insulin sensitivity was not weeks. Intervention in high-risk prediabetic obese cohorts assessed. Recent human trials have shown that treatment may yield more promising results, wherein IL-1RI in- with IL-1Ra (anakinra) improves glycemia and b-cell se- hibition may delay, if not prevent, development of type 2 cretory function, coincident with reduced systemic in- diabetes by attenuating adipose inflammation and pro- flammation in patients with type 2 diabetes (43). This trial tecting functional b-cells. Our study adds insight into the did not demonstrate improved insulin sensitivity after 13 potential initial contribution of IL-1RI signaling during diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1693 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 2/2 FIG. 4. Ex vivo evaluation of adipose tissue insulin sensitivity. A: Adipose explants from lean and obese WT and IL-1RI mice were harvested and 3 3 stimulated 6 insulin (100 nmol/L) ex vivo for 15 min. Tracer [ H]glucose was added to explants for an additional 45 min. Levels of [ H]glucose were measured by liquid scintillation counting. Fold increase in [ H]glucose transport in response to insulin over basal (non–insulin-stimulated) into adipose tissue for each individual mouse was calculated and is presented (##P < 0.01, ###P < 0.001 w.r.t. lean; *P < 0.05, **P < 0.01 with 2/2 respect to [w.r.t.] WT; $$$P < 0.001 w.r.t. basal, n =8). B–D: Protein lysates were prepared from adipose tissue of lean and obese WT and IL-1RI mice after stimulation 6 insulin (100 nmol/L) for 1 h ex vivo. A: Levels of tyrosine phosphorylated IRS-1 in adipose protein lysates were measured using a PathScan ELISA kit (#P < 0.05 w.r.t. WT lean; *P < 0.05 w.r.t. WT obese; $P < 0.05, $$P < 0.01 w.r.t. basal, n = 8). Levels of phosphorylated AKT, phosphorylated insulin receptor, whole-cell AKT, and b-actin in lean (C) and obese (D) adipose protein lysates were assessed by immunoblot analysis. Adipose tissue mRNA levels of IRS-1 (E) and GLUT4 (F) were analyzed by real-time PCR analysis (**P < 0.01, ***P < 0.001 w.r.t. lean; ##P < 0.01, ###P < 0.001 w.r.t. WT, n =8). early phases of obesity-induced IR, particularly in adipose This phenotype was age-dependent; glucose tolerance was 2/2 tissue. equivalent in older chow-fed IL-1RI and WT mice, co- 2/2 Our in vivo findings demonstrate that lack of IL-1RI incident with enhanced weight gain in the IL-1RI mice as 2/2 partially protects mice from HFD-induced IR, with im- previously reported (37). Interestingly, HFD-fed IL-1RI proved glucose homeostasis and enhanced insulin sensi- mice did not develop further IR compared with age- 2/2 2/2 tivity observed in obese IL-1RI compared with WT mice. matched chow-fed IL-1RI mice, despite much greater Plasma concentrations of triacylglycerol, nonesterified weight gain, providing further evidence of protection against fatty acid, IL-6, and TNF-a did not differ between strains; HFD-induced IR but not obesity. however, consistent with a previous report (37), levels of Immune cell recruitment into adipose plays a key role 2/2 insulin were elevated in obese IL-1RI mice. This hyper- in the etiology of HFD-induced IR (6), prompting us to 2/2 insulinemia was not associated with IR and appears to be speculate that reduced immune cell number in IL-1RI an intrinsic characteristic of the animal model given the adipose may partially account for improved glucose ho- 2/2 converse hypoinsulinemic phenotype of IL-1Ra mice meostasis. Contrary to our hypothesis, the number of tis- 2/2 2/2 (39). Prediet lean IL-1RI mice cleared glucose more sue M1/M2 in adipose was not different between IL-1RI efficiently than WT during GTT, which may be attributable and WT mice after HFD, demonstrating little role for IL-1 to enhanced insulin secretion in response to glucose load, in immune cell chemotaxis. We next speculated the cyto- 2/2 together with heightened adipose tissue insulin sensitivity. kine signature of ATMs may be altered in IL-1RI mice. 1694 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES 2/2 2/2 FIG. 5. Cross-talk of WT and IL-1RI BMMs with 3T3L1 adipocytes. WT and IL-1RI BMMs, seeded on transwell (0.4 mmol/L thickness) inserts, were stimulated with LPS (100 ng/mL) for 30 min, washed with PBS before addition of fresh media, and transferred to 3T3L1 adipocyte plates where cells were left to coculture for 5 days. A: The effect of BMMs on insulin (100 nmol/L)-stimulated [ H]glucose transport into 3T3L1 adi- pocytes was evaluated. Fold increase in [ H]glucose transport into adipocytes in response to insulin over basal (non-insulin–stimulated) is pre- sented (*P < 0.05, ***P < 0.001 with respect to [w.r.t.] individual basal; $$$P < 0.001 w.r.t. control + insulin; ##P < 0.01 w.r.t. WT BMM + insulin, n =4). B: Protein lysates from cocultured adipocytes stimulated with insulin (100 nmol/L) for 15 min were harvested and probed for levels of phosphorylated AKT, whole-cell AKT, and b-actin by immunoblot analysis. C–E: Media from cocultured cells were harvested after 72 h and levels of TNF-a (C), adiponectin (D), and IL-6 (E) were measured by ELISA (*P < 0.05, ***P < 0.001 w.r.t. adipocytes alone, n =4). F and G: Effect of 2/2 coculture with WT or IL-1RI BMMs on adipocyte IL-6 (F) and Glut4 (G) mRNA expression by real-time PCR (**P < 0.01, ***P < 0.001 w.r.t. 2/2 2/2 control; #P < 0.05, ###P < 0.001 w.r.t. IL-1RI coculture, n =4). H: Baseline inflammatory mRNA signature of WT and IL-1RI BMMs. Basal 2/2 mRNA levels of IL-6, IL-10, TNF-a, SOCS3, SOCS1, caspase 1, and TLR4 in WT and IL-1RI BMMs as assessed by real-time PCR (*P < 0.05, **P < 0.01, ***P < 0.001 w.r.t. WT, n =4). diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1695 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE FIG. 6. IL-1b and TNF-a synergize to enhance adipose tissue inflammation and NFkB activity. A: NIH-3T3 cells stably expressing an NFkB- luciferase promoter-reporter construct were treated with increasing concentrations of IL-1b or TNF-a alone or both in combination for 6 h. Levels of NFkB-driven luciferase were quantified using a luminometer (***P < 0.001 with respect to [w.r.t.] control; ###P < 0.001 w.r.t. TNF alone; $$$P < 0.001 w.r.t. either cytokine alone, n =4). B: 3T3L1 adipocytes were treated 6 IL-1b6 TNF-a for 24 h, and levels of IL-6 secreted were measured by 2/2 ELISA (***P < 0.001 w.r.t. control; $$$P < 0.001 w.r.t. either cytokine alone, n =4). C–G: BMMs were harvested from WT and IL-1RI mice and stimulated 6 IL-1b6 TNF-a (10 ng/mL) for 24 h, and levels of IL-6 (C) secreted measured by ELISA and IL-6 (D), SOCS3 (E), IL-1b (F), and TNFa mRNA (G) were measured by real-time PCR (**P < 0.01, ***P < 0.001 w.r.t. control; #P < 0.05, ##P < 0.01, ###P < 0.001 w.r.t. WT; $P < 0.05, $$P < 0.01, $$$P < 0.001 w.r.t. either cytokine alone, n =3). H and I: Adipose tissue explants (100 mg) were harvested from lean WT and IL- 2/2 1RI mice and were cultured for 6 h 6 IL-1b6 TNF-a (10 ng/mL) and levels of IL-6 secretion (H) and IL-6 mRNA (I) were determined (***P < 0.001 w.r.t. control; #P < 0.05, ###P < 0.001 w.r.t. WT; $$$P < 0.001 w.r.t. either cytokine alone, n =5). 1696 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org F.C. MCGILLICUDDY AND ASSOCIATES Robust cytokine secretion from WT ATMs was evident suggesting that these cytokines work in concert to create after 24 h in culture in the absence of any external stim- a positive feedback loop. Interestingly, TNF-a mRNA ex- ulus, indicative of preactivation within the inflamed envi- pression was not enhanced, suggesting differential regula- ronment of the obese tissue before culture ex vivo. By tory mechanisms for TNF-a mRNA. These findings suggest 2/2 contrast, cytokine secretion from IL-1RI ATMs was that blocking IL-1 signaling not only ameliorates the ad- markedly lower, suggestive of altered immunogenicity. A verse effects of IL-1 on adipocyte/macrophage biology, but recent report demonstrated removal of HFD from WT also prevents the synergy between TNF-a and IL-1b, thus obese animals improved insulin sensitivity before reduc- alleviating adipose tissue inflammation. tion in macrophage number; however, the proinflamma- Recent reports demonstrating activation of the NLRP3 tory phenotype of these ATMs was markedly attenuated inflammasome following an HFD and protection of upon removal of HFD, demonstrating phenotypic plasticity inflammasome-compromised mice against HFD-induced IR (44). It is therefore arguable that it is not the number of (19,20) have given new insight into the upstream mecha- macrophages per se that determine poor outcome, but the nisms through which IL-1 is elevated during obesity. We activation state, cytokine signature, and functional effects have focused on the downstream consequences of IL-1RI 2/2 of ATMs that dictate severity of IR. Upon culture, IL-1RI signaling within adipose tissue during an HFD and have 2/2 ATMs secreted markedly reduced levels of both TNF-a and demonstrated that IL-1RI mice are protected against IL-6, and it is thus likely that altered immunogenic phe- HFD-induced IR coincident with attenuated adipose tissue 2/2 notype of recruited IL-1RI immune cells results in im- and ATM inflammation. There are numerous local cellular proved cross-talk with resident adipocytes and improved targets and functional consequences of IL-1 released from insulin sensitivity. Consistent with this, we demonstrated adipose tissue, including preadipocytes (inhibition of 2/2 improved cytokine cross-talk of IL-1RI BMMs with 3T3L1 adipogenesis), mature adipocytes (induction of IR), and adipocytes in vitro, coincident with an attenuated effect on infiltrated immune cells (enhanced cytokine secretion), all adipocyte insulin sensitivity compared with WT BMMs. of which may contribute to adipose tissue IR. However, Given the prominent role for adipose tissue dysfunction our study is limited in its ability to distinguish whether lack in early stages of obesity-induced IR and the attenuated of IL-1RI signaling in immune cells or in adipocyte/ 2/2 immunogenic profile of IL-1RI ATMs, we next hypoth- preadipocyte fractions is primarily responsible for this im- esized that IL-1RI–transduced signals may contribute to proved phenotype. Metabolic studies in animals with ei- HFD-induced inflammation within whole adipose tissue. ther immune-specific or adipocyte-specific deletion of Previous findings in our laboratory highlighted attenuated IL-1RI would in turn provide greater insight into these inflammatory and cellular stress–related proteomic signa- mechanistic queries. ture exclusive to adipose tissue rather than skeletal mus- In conclusion, we ascribe a pathogenic role to IL-1RI– 2/2 cle or liver in obese IL-1RI mice compared with WT mediated signals during HFD-induced adipose tissue dys- 2/2 mice (45). The present work extends our understanding of function and IR. Improved glucose homeostasis in IL-1RI the paracrine effect associated with lack of IL-1RI in adi- mice correlated with a marked reduction in local adipose pose tissue with reduced IL-6 and TNF-a secretion from inflammation, with altered ATM cytokine signatures and explants coincident with reduced mRNA expression of pronounced attenuation in both IL-6 secretion and activa- TNF-a, IL-6, and SOCS3. Furthermore, we demonstrate tion of the IL-6 regulated proteins STAT3 and SOCS3. Loss 2/2 reduced activation of STAT3 coincident with reduced of synergy between TNF-a and IL-1b in IL-1RI adipose 2/2 SOCS3 protein in IL-1RI adipose, an effect potentially mayinturnhave contributedtothisattenuatedinflammatory attributable to reduced IL-6 secretion. This effect reduced profile. This study highlights the critical and nonredundant local adipose tissue inflammation in turn correlated with contribution of IL-1 to the medley of proinflammatory 2/2 improved insulin sensitivity of IL-1RI adipose ex vivo. insulin-desensitizing signals relevant to obesity-induced IR. Interestingly, this attenuated inflammatory profile was lo- calized to adipose tissue with equivalent concentrations of ACKNOWLEDGMENTS 2/2 systemic cytokines in IL-1RI and WT mice. The dis- This work was supported by the Science Foundation crepancy between systemic and local cytokine levels is Ireland PI Programme (06/IM.1/B105) (to H.M.R.). unclear, but may be attributable to almost undetectable No potential conflicts of interest relevant to this article plasma IL-1b levels in WT animals, even with obesity were reported. (46,47), compared with the concentration of IL-1b in F.C.M. researched data and wrote the manuscript. K.A.H. macrophages/adipose tissue. and C.M.R. researched data and proofread the manuscript. Whereas IL-1RI signaling clearly plays a nonredun- E.O. and M.C. researched data. K.H.G.M. reviewed data dant role during HFD-induced IR, it may be an oversim- and edited the manuscript. H.M.R. designed the study, re- plification to attribute the majority of adipose tissue viewed data, and edited the manuscript. inflammation to the actions of one cytokine. It is much The authors appreciate the technical support from the more likely that adipose-derived cytokines cross-talk to core technologies staff in the Conway Institute, University amplify adipose tissue inflammation. Consistent with this College Dublin, Belfield, Dublin, Ireland, and the technical hypothesis, we demonstrated a potent synergistic effect of support of Dr. Eileen Murphy during body composition the major adipocytokines IL-1b and TNF-a in enhancing analysis in the Alimentary Pharmabiotic Center, University NFkB activation and adipose IL-6 protein secretion. This College Cork, Cork, Ireland. synergistic effect on macrophage and adipose IL-6 secre- tion was lost in the absence of IL-1RI. Furthermore, the REFERENCES ability of TNF-a to induce IL-6 secretion was blunted in 2/2 1. Winer S, Chan Y, Paltser G, et al. Normalization of obesity-associated in- IL-1RI adipose, suggesting that TNF-a–induced IL-6 is sulin resistance through immunotherapy. Nat Med 2009;15:921–929 in part mediated via IL-1. Remarkably, TNF-a and IL-1b 2. Nishimura S, Manabe I, Nagasaki M, et al. CD8+ effector T cells contribute synergized to enhance IL-1b mRNA expression in WT to macrophage recruitment and adipose tissue inflammation in obesity. BMMs far beyond the effects of either cytokine alone, Nat Med 2009;15:914–920 diabetes.diabetesjournals.org DIABETES, VOL. 60, JUNE 2011 1697 IL-1 AND OBESITY-INDUCED INSULIN RESISTANCE 3. Kintscher U, Hartge M, Hess K, et al. T-lymphocyte infiltration in visceral 25. Verstrepen L, Bekaert T, Chau TL, Tavernier J, Chariot A, Beyaert R. TLR-4, adipose tissue: a primary event in adipose tissue inflammation and the IL-1R and TNF-R signaling to NF-kappaB: variations on a common theme. development of obesity-mediated insulin resistance. Arterioscler Thromb Cell Mol Life Sci 2008;65:2964–2978 Vasc Biol 2008;28:1304–1310 26. Arkan MC, Hevener AL, Greten FR, et al. IKK-beta links inflammation to 4. Rausch ME, Weisberg S, Vardhana P, Tortoriello DV. Obesity in C57BL/6J obesity-induced insulin resistance. Nat Med 2005;11:191–198 mice is characterized by adipose tissue hypoxia and cytotoxic T-cell in- 27. Yuan M, Konstantopoulos N, Lee J, et al. Reversal of obesity- and diet- filtration. Int J Obes (Lond) 2008;32:451–463 induced insulin resistance with salicylates or targeted disruption of Ikk- 5. Lumeng CN, Deyoung SM, Bodzin JL, Saltiel AR. Increased inflammatory beta. Science 2001;293:1673–1677 properties of adipose tissue macrophages recruited during diet-induced 28. Yang Y, Ju D, Zhang M, Yang G. Interleukin-6 stimulates lipolysis in porcine obesity. Diabetes 2007;56:16–23 adipocytes. Endocrine 2008;33:261–269 6. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW 29. Lagathu C, Yvan-Charvet L, Bastard JP, et al. Long-term treatment with Jr. Obesity is associated with macrophage accumulation in adipose tissue. interleukin-1beta induces insulin resistance in murine and human adipo- J Clin Invest 2003;112:1796–1808 cytes. Diabetologia 2006;49:2162–2173 7. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. 30. Emanuelli B, Peraldi P, Filloux C, et al. SOCS-3 inhibits insulin signaling J Clin Invest 2006;116:1793–1801 and is up-regulated in response to tumor necrosis factor-alpha in the adi- 8. Kern PA, Saghizadeh M, Ong JM, Bosch RJ, Deem R, Simsolo RB. The pose tissue of obese mice. J Biol Chem 2001;276:47944–47949 expression of tumor necrosis factor in human adipose tissue. Regulation 31. Ueki K, Kondo T, Kahn CR. Suppressor of cytokine signaling 1 (SOCS-1) by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest and SOCS-3 cause insulin resistance through inhibition of tyrosine phos- 1995;95:2111–2119 phorylation of insulin receptor substrate proteins by discrete mechanisms. 9. Fenkci S, Rota S, Sabir N, Sermez Y, Guclu A, Akdag B. Relationship of Mol Cell Biol 2004;24:5434–5446 serum interleukin-6 and tumor necrosis factor alpha levels with abdominal 32. Niemand C, Nimmesgern A, Haan S, et al. Activation of STAT3 by IL-6 and fat distribution evaluated by ultrasonography in overweight or obese IL-10 in primary human macrophages is differentially modulated by sup- postmenopausal women. J Investig Med 2006;54:455–460 pressor of cytokine signaling 3. J Immunol 2003;170:3263–3272 10. Wellen KE, Hotamisligil GS. Inflammation, stress, and diabetes. J Clin 33. Elmendorf JS, Chen D, Pessin JE. Guanosine 59-O-(3-thiotriphosphate) Invest 2005;115:1111–1119 (GTPgammaS) stimulation of GLUT4 translocation is tyrosine kinase- 11. Nov O, Kohl A, Lewis EC, et al. Interleukin-1b may mediate insulin re- dependent. J Biol Chem 1998;273:13289–13296 sistance in liver-derived cells in response to adipocyte inflammation. En- 34. Wang X, Collins HL, Ranalletta M, et al. Macrophage ABCA1 and ABCG1, docrinology 2010;151:4247–4256 but not SR-BI, promote macrophage reverse cholesterol transport in vivo. 12. Hotamisligil GS, Murray DL, Choy LN, Spiegelman BM. Tumor necrosis J Clin Invest 2007;117:2216–2224 factor alpha inhibits signaling from the insulin receptor. Proc Natl Acad Sci 35. McGillicuddy FC, Chiquoine EH, Hinkle CC, et al. Interferon gamma U S A 1994;91:4854–4858 attenuates insulin signaling, lipid storage, and differentiation in human 13. Tanaka T, Itoh H, Doi K, et al. Down regulation of peroxisome proliferator- adipocytes via activation of the JAK/STAT pathway. J Biol Chem 2009;284: activated receptorgamma expression by inflammatory cytokines and its 31936–31944 reversal by thiazolidinediones. Diabetologia 1999;42:702–710 36. Thomas HE, Irawaty W, Darwiche R, et al. IL-1 receptor deficiency slows 14. Jager J, Grémeaux T, Cormont M, Le Marchand-Brustel Y, Tanti JF. progression to diabetes in the NOD mouse. Diabetes 2004;53:113–121 Interleukin-1beta-induced insulin resistance in adipocytes through down- 37. García MC, Wernstedt I, Berndtsson A, et al. Mature-onset obesity in regulation of insulin receptor substrate-1 expression. Endocrinology 2007; interleukin-1 receptor I knockout mice. Diabetes 2006;55:1205–1213 148:241–251 38. Sauter NS, Schulthess FT, Galasso R, Castellani LW, Maedler K. The anti- 15. Uysal KT, Wiesbrock SM, Marino MW, Hotamisligil GS. Protection from inflammatory cytokine interleukin-1 receptor antagonist protects from obesity-induced insulin resistance in mice lacking TNF-alpha function. high-fat diet-induced hyperglycemia. Endocrinology 2008;149:2208–2218 Nature 1997;389:610–614 39. Somm E, Henrichot E, Pernin A, et al. Decreased fat mass in interleukin-1 16. Poggi M, Bastelica D, Gual P, et al. C3H/HeJ mice carrying a toll-like re- receptor antagonist-deficient mice: impact on adipogenesis, food intake, ceptor 4 mutation are protected against the development of insulin re- and energy expenditure. Diabetes 2005;54:3503–3509 sistance in white adipose tissue in response to a high-fat diet. Diabetologia 40. Matsuki T, Horai R, Sudo K, Iwakura Y. IL-1 plays an important role in lipid 2007;50:1267–1276 metabolism by regulating insulin levels under physiological conditions. 17. Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 links innate J Exp Med 2003;198:877–888 immunity and fatty acid-induced insulin resistance. J Clin Invest 2006;116: 41. Bendtzen K, Mandrup-Poulsen T, Nerup J, Nielsen JH, Dinarello CA, 3015–3025 Svenson M. Cytotoxicity of human pI 7 interleukin-1 for pancreatic islets of 18. Mills KH, Dunne A. Immune modulation: IL-1, master mediator or initiator Langerhans. Science 1986;232:1545–1547 of inflammation. Nat Med 2009;15:1363–1364 42. Maedler K, Sergeev P, Ris F, et al. Glucose-induced beta cell production of 19. Vandanmagsar B, Youm YH, Ravussin A, et al. The NLRP3 inflammasome IL-1beta contributes to glucotoxicity in human pancreatic islets. J Clin instigates obesity-induced inflammation and insulin resistance. Nat Med Invest 2002;110:851–860 2011;17:179–188 43. Larsen CM, Faulenbach M, Vaag A, et al. Interleukin-1-receptor antagonist 20. Stienstra R, Joosten LA, Koenen T, et al. The inflammasome-mediated in type 2 diabetes mellitus. N Engl J Med 2007;356:1517–1526 caspase-1 activation controls adipocyte differentiation and insulin sensi- 44. Li P, Lu M, Nguyen MT, et al. Functional heterogeneity of CD11c-positive tivity. Cell Metab 2010;12:593–605 adipose tissue macrophages in diet-induced obese mice. J Biol Chem 2010; 21. Maedler K, Dharmadhikari G, Schumann DM, Størling J. Interleukin-1 beta 285:15333–15345 targeted therapy for type 2 diabetes. Expert Opin Biol Ther 2009;9:1177– 45. de Roos B, Rungapamestry V, Ross K, et al. Attenuation of inflammation 1188 and cellular stress-related pathways maintains insulin sensitivity in obese 22. Stylianou E, Saklatvala J. Interleukin-1. Int J Biochem Cell Biol 1998;30: type I interleukin-1 receptor knockout mice on a high-fat diet. Proteomics 1075–1079 2009;9:3244–3256 23. Hirosumi J, Tuncman G, Chang L, et al. A central role for JNK in obesity 46. Jung C, Gerdes N, Fritzenwanger M, Figulla HR. Circulating levels of and insulin resistance. Nature 2002;420:333–336 interleukin-1 family cytokines in overweight adolescents. Mediators In- 24. Shoelson SE, Lee J, Yuan M. Inflammation and the IKK beta/I kappa B/NF- flamm 2010;2010:958403 kappa B axis in obesity- and diet-induced insulin resistance. Int J Obes 47. van der Poll T, Marchant A, Buurman WA, et al. Endogenous IL-10 protects Relat Metab Disord 2003;27(Suppl. 3):S49–S52 mice from death during septic peritonitis. J Immunol 1995;155:5397–5401 1698 DIABETES, VOL. 60, JUNE 2011 diabetes.diabetesjournals.org

Journal

Diabetes – Pubmed Central

Published: May 21, 2011

There are no references for this article.