Salicylate Blocks Lipolytic Actions of Tumor Necrosis Factor- in Primary Rat Adipocytes

Increased systemic free fatty acids (FFA) impair insulin sensitivity. In obese and diabetic subjects, production of tumor necrosis factor(TNF), a proinflammatory cytokine, is elevated. TNFhas a variety of effects by inducing inflammation, decreasing glucose utilization, and stimulating adipocyte lipolysis to release FFA to plasma. High doses of nonsteroidal anti-inflammatory drug salicylates have long been recognized to lower blood FFA and glucose in humans, although the mechanisms are not fully understood. In this report, we show that sodium salicylate at therapeutic concentrations directly blocks TNF-stimulated lipolysis and therefore inhibits FFA release from primary rat adipocytes. To elucidate the cellular basis of this action, we show that salicylate suppresses TNF-induced extracellular signal-related kinase activation and intracellular cAMP elevation, two early events during the lipolysis response to TNF. Furthermore, salicylate prevents the down-regulation of cyclic-nucleotide phosphodiesterase 3B, an enzyme responsible for cAMP hydrolysis. Perilipins coat intracellular lipid droplet surface by restricting lipase access to the triacylglycerol substrates. TNFdown-regulates perilipin but promotes its phosphorylation during lipolysis stimulation; these actions are efficiently reversed by salicylate. Salicylate slightly reduces basal but completely inhibits TNF-liberated lipase activity. In contrast, neither salicylate nor TNFalters the protein levels of hormone-sensitive lipase and adipose triglyceride lipase. In addition, sodium salicylate restricts basal lipolysis simulated by a high concentration of glucose and significantly diminishes the high glucose-enhanced lipolysis response to TNF. These results provide novel evidence that salicylate directly blocks TNF-mediated FFA efflux from adipocytes, hence reducing plasma FFA levels and increasing insulin sensitivity. Obesity and type 2 diabetes mellitus are associated with elevated levels of plasma FFA, which directly induce insulin resistance (Bergman and Ader, 2000). The increased systemic FFA is believed to result from dysregulated lipolysis of triacylglycerols in adipose cells. One mechanism that may contribute to elevated FFA release is an increase of TNFproduction in patients who are obese and diabetic (Hotamisligil et al., 1995). TNFis a proinflammatory cytokine that has multifunctional effects in inflammatory and metabolic disorders. Recent studies suggest that TNFis an important mediator in the development of insulin resistance (Hotamisligil et al., 1995; Uysal et al., 1997). For example, although controversial (Ofei et al., 1996), prolonged TNFneutralization by its antibodies effectively improves insulin resistance in patients with diabetes (Kiortsis et al., 2005). TNF-deficient obese mice have lower circulating FFA levels and are protected from obesity-related insulin sensitivity (Uysal et al., 1997). TNFhas important metabolic actions that stimulate chronic lipolysis in primary (Green et al., 1994; Ren et al., 2006) and differentiated adipocytes (Souza et al., 1998; Ryden et al., 2002; Green et al., 2004). The lipolytic action of TNFgoverns FFA efflux from adipocytes to plasma, thereby elevating systemic FFA levels and causing insulin resistance. This work was supported by grants 30670779 and 30370535 from the National Natural Science Foundation of China and grant 5072030 from the Beijing Natural Science Foundation. This work was supported by the Program for New Century Excellent Talents in the University, of the Education Ministry of China (NECT-04-0023), and by the Major State Basic Research Development Program of China (2006CB503903). Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org. doi:10.1124/mol.107.039479. ABBREVIATIONS: FFA, free fatty acids; TNF, tumor necrosis factor; ERK, extracellular signal-related kinase; PDE3B, phosphodiesterase 3B; HSL, hormone-sensitive lipase; ATGL, adipose triglyceride lipase; DMEM, Dulbecco’s modified Eagle’s medium; PKA, cAMP-dependent protein kinase; HRP, horseradish peroxidase; PCV, packed cell volume; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; LDH, lactate dehydrogenase; JNK, c-Jun-NH2-terminal kinase; H89, N-[2-(4-bromocinnamylamino)ethyl]-5-isoquinoline; IKK , I B kinase . 0026-895X/08/7301-215–223$20.00 MOLECULAR PHARMACOLOGY Vol. 73, No. 1 Copyright © 2008 The American Society for Pharmacology and Experimental Therapeutics 39479/3284521 Mol Pharmacol 73:215–223, 2008 Printed in U.S.A. 215 at A PE T Jornals on O cber 4, 2017 m oharm .aspeurnals.org D ow nladed from The nonsteroidal anti-inflammatory drugs sodium salicylate and acetylsalicylic acid (aspirin) are widely used to control pain, fever, and rheumatic arthritis. Aspirin is standard care for patients with diabetes with cardiovascular disease. In 1877, Ebstein found that high doses of sodium salicylate dramatically reduced glucosuria in patients with diabetes (Ebstein, 1877). Further early studies showed that high doses of salicylates also lowered blood glucose concentrations in rodents (Bizzi et al., 1965) and in diabetic humans (Reid et al., 1957; Carlson and Ostman, 1961). In contrast, conflicting results demonstrate that lower doses of aspirin (3 g/day for 3 days) may not improve glucose utilization in healthy (Newman and Brodows, 1983) and diabetic subjects (BratuschMarrain et al., 1985). Important discrepancies between these studies included lower salicylate dosages ( 3 g/day) and therapeutic duration (a few days) in the more recent studies than in the earlier studies (6–9 g/day for 1–3 weeks). Although salicylate reduces TNFproduction in rat macrophages (Vittimberga et al., 1999), administration of low-dose aspirin (325 mg/days) may result in a rebound increase in cytokine-induced synthesis of interleukin-1 and TNFin human (Endres et al., 1996), which can be expected to impair insulin sensitivity. Most recently, several studies indicate that high doses of salicylates attenuate deleterious effects of lipids and therefore improve lipid-induced insulin resistance both in rodents (Kim et al., 2001; Yuan et al., 2001) and humans (Hundal et al., 2002; Möhlig et al., 2006). Early studies indicated that salicylates lower serum FFA concentrations in healthy and diabetic subjects, which possibly contributes to their hypoglycemic effects. The FFA-lowering action may result from the suppression of FFA release from adipose tissue to plasma (Reid et al., 1957; Carlson and Ostman, 1961; Bizzi et al., 1965), because salicylates do not seem to affect FFA esterification and turnover. The FFA mobilization to plasma is governed by lipolytic reaction of adipocytes physiologically stimulated by catecholamines. Although the mechanism remains unidentified, salicylates can reduce catecholamine-stimulated lipolysis in isolated adipocytes (Stone et al., 1969; Schönhöfer et al., 1973). In persons who are obese, elevated TNFacts as a strong lipolytic stimulator (Hotamisligil et al., 1995). We (Ren et al., 2006) and others (Ryden et al., 2002) have indicated that activation of extracellular signal-related kinase (ERK) is a critical factor mediating TNF-induced lipolysis of adipocytes, whereas suppression of ERK signaling inhibits the lipolysis. Salicylates are able to inhibit ERK activation in neutrophils (Pillinger et al., 1998) and in TNF-stimulated fibroblasts (Schwenger et al., 1996). We hypothesize that salicylate may directly attenuate TNF-stimulated lipolysis response in adipocytes, which could be one of cellular bases for salicylate to reduce plasma FFA concentrations and thus improve insulin resistance. In this report, we show that sodium salicylate at therapeutic concentrations directly inhibits adipocyte lipolysis response to TNF. Salicylate attenuates TNF-induced ERK activation, cellular cAMP elevation, and the downregulation of cyclic-nucleotide phosphodiesterase 3B (PDE3B), thus restricting lipolysis. Furthermore, salicylate corrects TNF-dysregulated protein levels and phosphorylation state of perilipins on adipocyte lipid droplet surface, inhibits TNF-liberated lipase activity, but does not alter the protein expression of hormone-sensitive lipase (HSL) and adipose triglyceride lipase (ATGL). This study provides novel evidence that salicylate directly antagonizes TNF-stimulated FFA efflux from adipocytes to plasma, thus lowering systemic FFA levels and increasing insulin sensitivity. Materials and Methods Materials. Recombinant rat TNFwas purchased from PeproTech EC (London, UK). Sodium salicylate was from Beijing Chemical Reagents Co. (Beijing, China). Phenol red-free Dulbecco’s modified Eagle’s medium (DMEM) containing glucose (5 mM) and protein kinase A (PKA) inhibitor H89 were from Sigma Chemical (St. Louis, MO). Enzyme materials used for enzymatic assays were products of Totobo Co. (Tokyo, Japan). Antibodies against ERK-1, phosphoERK1/2, PDE3B, anti-Gi1 , actin, and horseradish peroxidase (HRP)-conjugated second antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit antibodies against rat perilipin and rat HSL (He et al., 2006) were generous gifts from Dr. Londos at the U.S. National Institutes of Health (Bethesda, MD). Antibody against phosphorylated PKA substrate (RRXS/T motif) was from Cell Signaling Technology (Danvers, MA). Rat anti-ATGL antibody was from Cayman Chemical (Ann Arbor, MI). Nitrocellulose blot membrane, prestained protein molecular weight marker, and ultrasensitive enhanced chemiluminescence (ECL) detection reagents were from Applygen Technologies Inc. (Beijing, China). Isolation and Culture of Primary Rat Adipocytes. Adipocytes were isolated from epididymal fat pads of Sprague-Dawley rats (150– 180 g) according to our laboratory method (He et al., 2006; Jiang et al., 2007). The fat pads were minced and digested in 5 ml of KrebsRinger solution containing 0.75 mg/ml type I collagenase, 200 nM adenosine, 25 mM HEPES, pH 7.4, and 1% defatted bovine serum albumin. After incubation for 40 min at 37°C in a water bath with shaking at 100 cycles/min, cells were filtered through a nylon mesh and washed 3 times with warmed DMEM containing 200 nM adenosine. Adipocytes floating on the top of the tube were packed by centrifuging at 200g for 3 min. Every 25 l of packed adipocytes was resuspended in 500 l of phenol redand serum-free DMEM containing 2% defatted bovine serum albumin and preincubated at 37°C for 1 h before treatments (He et al., 2006). Next, adipocytes were incubated in the presence or absence of the tested agents, followed by the assays described below. Fatty Acid Assay. The concentration of FFA in the culture medium was determined by colorimetric assay as described previously (Itaya, 1977) with some modifications. In brief, 50 l of culture medium was mixed with 120 l of isooctane and 80 l of cupric acetate-pyridine. The mixture was vortexed and centrifuged for 10 min at 12,000g at room temperature. The upper organic phase (80 l) was transferred to a clean tube. One hundred eighty microliters of the color development reagent consisting of diphenylcarbazone and diphenylcarbazide in methanol was then added to the tube. The mixture was vortexed for 5 s, and the color of the reaction was developed immediately. The absorbance of the color reaction at 540 nm was spectrophotometrically measured in a 96-well plate. Glycerol Assay. Glycerol content released in culture medium of adipocytes served as an index of lipolysis and was determined at the absorption at 490 nm (He et al., 2006; Ren et al., 2006), by use of a colorimetric assay (GPO Trinder reaction) kit from Applygen Technologies Inc. Lipolysis data were expressed as micromoles of glycerol or FFA per milliliter of packed cell volume of adipocytes. Western Blot. Adipocytes were packed and lysed in sample buffer containing 62 mM Tris-HCl, pH 6.8, 5% SDS, 0.1 mM sodium orthovanadate, and 50 mM sodium fluoride (He et al., 2006). After centrifugation at 12,000g for 10 min at 4°C, the lysate was transferred to a new tube and heated at 95°C for 5 min. Protein content in the extracts was determined by use of a bicinchoninc acid protein assay kit from Applygen Technologies Inc. Equal amounts of proteins were loaded and separated by 10% SDS-polyacrylamide gel electro216 Zu et al. at A PE T Jornals on O cber 4, 2017 m oharm .aspeurnals.org D ow nladed from phoresis, then transferred to a nitrocellulose membrane. The membranes were blocked for 1 h in 5% nonfat milk in 150 mM NaCl, 20 mM Tris-HCl, pH 7.4, and 0.05% Tween 20) (Xu et al., 2005, 2006), then incubated with primary antibodies overnight at 4°C, followed by incubation for 1 h with HRP-conjugated secondary antibodies. The blots were developed by use of an enhanced chemiluminescence (ECL) detection kit. If required, the antibodies bound to membranes were removed by a commercial stripping solution from Applygen Technologies Inc. Blots were then reprobed with use of other antibodies and developed as described above. Densitometric analysis of protein bands involved use of NIH Image software (http://rsb.info. nih.gov/nih-image/). cAMP Immunoradioassay. Adipocytes were lysed in 150 l of ice-cold buffer containing 50 mM Tris-HCl, pH 7.4, and 1 mM EDTA. To solidify the fat-cake–enriched oil from lysed adipocytes, the lysate was incubated on ice for 15 min, vortexed vigorously, and centrifuged at 12,000g for 15 min at 4°C. The cytosolic fraction was collected from below the solidified fat cake in the tube (He et al., 2006; Jiang et al., 2007). The protein content in the cytosol fraction was determined. Then, 90 l of cytosol fraction was mixed with 30 l of 40% trichloroacetic acid. The tubes were vortexed and centrifuged at 12,000g for 5 min at 4°C. The supernatant was collected and used for cAMP assay according to the manufacturer-provided protocol from a commercial I radioimmunoassay kit (Isotope Laboratory of Shanghai University of Chinese Medicine, Shanghai, China). The value of cAMP concentrations was normalized and expressed as picomoles per milligram of cytosolic proteins. Assay of Adipose Lipase Activity. After the treatments, adipocytes were washed twice with warmed PBS buffer and packed by centrifugation. The 50 l of packed adipocytes was lysed in 120 l of buffer containing 50 mM Tris-HCl, pH 7.4, and 1 mM EDTA. After being vortexed vigorously, the lysate was centrifuged at 12,000g for 15 min at 4°C. The infranatant phase below the fat cake fraction was transferred to a new tube, then centrifuged at 12,000g for 5 min at 4°C. The supernatant was used for the determination of cellular lipase activity against emulsified triolein substrate (Peled and Krenz, 1981). The mixture was incubated for 30 min at 37°C, when the lipases hydrolyze emulsified triolein to produce glycerol. The release of glycerol from triolein hydrolysis represented the activity of adipose lipase and was assayed as described above. Cell Viability Assays. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and lactate dehydrogenase (LDH) assay were performed according to the manufacturer instructions of two commercial kits available from Applygen Technologies Inc. The MTT assay was based on the reduction of tetrazolium salt to colored formazan product by mitochondrial enzymes present only in living cells. Cell viability rates by MTT assay were directly proportional to the absorbance values of optical density measured at 570 nm in a 96-well plate and were presented as percentage of the control values. LDH activity was measured based on the conversion of lactate to pyruvate in the presence of LDH with parallel reduction of NAD , and the further reaction of pyruvate with 2,4-dinitrophenylhydrazine to form a colored hydrazone product that has a high optical density in the wavelength range of 400 to 500 nm. The absorbance at 440 nm was determined and used for calculating LDH activity. The cell viability index of LDH was presented as a percentage of LDH leakage in medium compared with total LDH activity. Statistical Analysis. Data are expressed as means S.E.M. One-way analysis of variance Tukey’s test involved use of GraphPad Prism version 4.0. p 0.05 was considered significant.