Fulltext for PDF in Flash Viewer:

Toll-Like Receptor-4 Mediates Vascular Inflammation and Insulin Resistance in Diet-Induced Obesity Francis Kim, Matilda Pham, Ian Luttrell, Douglas D. Bannerman, Joan Tupper, Joshua Thaler, Thomas R. Hawn, Elaine W. Raines and Michael W. Schwartz Circ. Res. 2007;100;1589-1596; originally published online May 3, 2007; DOI: 10.1161/CIRCRESAHA.106.142851 Circulation Research is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX 72514 Copyright 2007 American Heart Association. All rights reserved. Print ISSN: 0009-7330. Online ISSN: 1524-4571 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circres.ahajournals.org/cgi/content/full/100/11/1589 Subscriptions: Information about subscribing to Circulation Research is online at http://circres.ahajournals.org/subscriptions/ Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax: 410-528-8550. E-mail: journalpermissions@lww.com Reprints: Information about reprints can be found online at http://www.lww.com/reprints Downloaded from circres.ahajournals.org by on April 26, 2008 Toll-Like Receptor-4 Mediates Vascular Inflammation and Insulin Resistance in Diet-Induced Obesity Francis Kim, Matilda Pham, Ian Luttrell, Douglas D. Bannerman, Joan Tupper, Joshua Thaler, Thomas R. Hawn, Elaine W. Raines, Michael W. Schwartz AbstractVascular dysfunction is a major complication of metabolic disorders such as diabetes and obesity. The current studies were undertaken to determine whether inflammatory responses are activated in the vasculature of mice with diet-induced obesity, and if so, whether Toll-Like Receptor-4 (TLR4), a key mediator of innate immunity, contributes to these responses. Mice lacking TLR4 (TLR4 / ) and wild-type (WT) controls were fed either a low fat (LF) control diet or a diet high in saturated fat (HF) for 8 weeks. In response to HF feeding, both genotypes displayed similar increases of body weight, body fat content, and serum insulin and free fatty acid (FFA) levels compared with mice on a LF diet. In lysates of thoracic aorta from WT mice maintained on a HF diet, markers of vascular inflammation both upstream (IKK activity) and downstream of the transcriptional regulator, NF- B (ICAM protein and IL-6 mRNA expression), were increased and this effect was associated with cellular insulin resistance and impaired insulin stimulation of eNOS. In contrast, vascular inflammation and impaired insulin responsiveness were not evident in aortic samples taken from TLR4 / mice fed the same HF diet, despite comparable increases of body fat mass. Incubation of either aortic explants from WT mice or cultured human microvascular endothelial cells with the saturated FFA, palmitate (100 mol/L), similarly activated IKK , inhibited insulin signal transduction and blocked insulin-stimulated NO production. Each of these effects was subsequently shown to be dependent on both TLR4 and NF- B activation. These findings identify the TLR4 signaling pathway as a key mediator of the deleterious effects of palmitate on endothelial NO signaling, and are the first to document a key role for TLR4 in the mechanism whereby diet-induced obesity induces vascular inflammation and insulin resistance. (Circ Res. 2007;100:1589-1596.) Key Words: nitric oxide Toll like receptor-4 free fatty acids obesity n conditions of nutrient excess such as obesity and diabetes, elevated free fatty acid (FFA) levels are implicated in the pathogenesis of both inflammation and insulin resistance in a variety of tissues, including endothelial cells.1 4 At the cellular level, nutrient excess is linked to insulin resistance via activation of IKK and, subsequently, NF- B, a key transcriptional mediator of inflammation.5 6 In peripheral insulin-sensitive tissues such as muscle and liver, the pattern recognition receptor, Toll-like receptor-4 (TLR4) is implicated as a mediator of this effect.7 The current work was undertaken to determine whether obesity induced by high-fat (HF) feeding causes inflammation and insulin resistance in vascular tissue in vivo, and whether TLR4 contributes to this effect. The bacterial endotoxin, lipopolysaccharide (LPS), is a potent activator of IKK and NF- B in most cell types. The majority of the biological activity of LPS is contained within a moiety (��ipid A�� that is acylated with saturated fatty acids, and removal of these fatty acids results in complete loss of endotoxic activity.8,9 Recently, TLR4 was shown to be I required not only for LPS-induced inflammatory responses, but for responses to nonbacterial ligands such as lauric acid (C 12:0), a saturated fatty acid.10,11 These in vitro studies suggest that activation of TLR4 by certain FFA species can trigger cellular inflammatory responses. Whether TLR4 signaling contributes to the link between nutrient excess, inflammation, and metabolic dysfunction in vivo is an important unanswered question. TLR4 is expressed on virtually all human cells and binds a wide spectrum of exogenous and endogenous ligands (including bacterial LPS) and is involved in innate immune responses to various infectious agents and stressors.12,13 In the presence of LPS, the TLR4 receptor complex (which includes CD-14 and an accessory protein, MD-2), recruits the adaptor protein, myeloid differentiation factor-88 (MyD88). MyD88 in turn recruits interleukin-1 receptorassociated kinase (IRAK) and, by activating IKK and NF- B, ultimately induces the expression of numerous inflammatory mediators. Recent evidence suggests that inflammatory processes induced by nutrient excess cause systemic insulin resistance via Original received October 17, 2006; revision received March 29, 2007; accepted April 23, 2007. From the Departments of Medicine (F.K., M.P., J. Tupper, J. Thaler, T.R.H., M.W.S.) and Urology (I.L.) Harborview Medical Center, and the Department of Pathology (E.W.R.), University of Washington, Seattle; and the Bovine Functional Genomics Laboratory (D.D.B.), U.S. Department of Agriculture, Agricultural Research Service, Beltsville, Md. Correspondence to Francis Kim, Department of Medicine, Box 359748, University of Washington, Seattle, WA 98104. E-mail fkim@u.washington.edu 2007 American Heart Association, Inc. Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.106.142851 1589 Downloaded from circres.ahajournals.org by on April 26, 2008 1590 Circulation Research June 8, 2007 inflammation and impaired endothelial insulin signaling in vivo and if so, whether TLR4 signaling is required for this effect. Compared with mice of either genotype fed a low-fat (LF) diet, HF feeding for 8 weeks increased body weight, body fat content, and serum levels of insulin, FFA, and cholesterol (supplemental Figure IA through IG). TLR4 deficiency had no detectable effect on body adiposity or humoral responses measured during consumption of either diet, with the exceptions of elevated serum FFA and triglycerides levels, which achieved statistical significance in TLR4 / but not in WT mice. To determine whether HF dietinduced obesity activates inflammatory signaling in vascular tissue in vivo, we measured IKK activity, as determined by the phosphorylation of I B in lysates of thoracic aorta removed from animals in each of the 4 groups (WT or TLR4 / mice fed a LF or HF diet). HF feeding for 8 weeks induced a significant 2-fold increase of IKK activity in aortic tissue taken from WT mice compared with LF-fed controls (Figure 1A). In TLR4 / mice maintained on a HF diet, however, no such increase of IKK activity was detected. Among many inflammatory genes induced by NF- B are IL-6 and ICAM1.17,18 To confirm that NF- B was indeed activated in thoracic aortic tissue exhibiting increased phospho-I B , we determined whether exposure to HF feeding increased IL-6 or ICAM expression, and whether absence of TLR4 signaling blocks this effect. As predicted, HF feeding was associated with increases of aortic ICAM1 and IL-6 mRNA levels in WT mice, but not in TLR4 / mice (Figure 1B and 1C). The induction of inflammatory genes was not attributable to atherosclerosis lesion formation in C57Bl/6 mice on a HF diet because previous reports show that these mice do not demonstrate evidence of atherosclerosis in the aorta, even after 40 weeks on a HF diet.19 To assess the effect of HF feeding on aortic insulin signaling, a subset of mice from each group received IP injections of either vehicle or insulin (2 U in 300 L of normal saline) and 15 minutes later, thoracic aorta was removed and aortic lysates analyzed for pAkt and peNOS activity. As expected, insulin-stimulated pAkt and peNOS activation were impaired in aortic tissue from WT mice fed the HF diet compared with LF-fed controls (Figure 1D and 1E). As in other insulin sensitive tissues,6 therefore, biochemical responses to insulin are impaired in arterial tissue taken from WT mice with DIO, including phosphorylation of eNOS serine 1177 (peNOS), an endothelial-specific measure. In contrast, HF feeding did not attenuate insulin-induced activation of Akt or eNOS in aortic tissue taken from TLR4 / mice. Collectively, these findings indicate that DIO induced by HF feeding promotes vascular inflammation and insulin resistance in vascular tissue via a mechanism that requires intact TLR4 signaling. Similarly, we found that insulin stimulation of pAkt in liver lysates was impaired by HFfeeding in WT mice, but not in mice lacking TLR4 (supplemental Figure II), consistent with a recent report.7 a mechanism involving TLR4,7 and a similar mechanism may also contribute to atherogenesis, as genetic deficiency of either TLR4 or MyD88 confers protection against atherosclerosis.14,15 Based on evidence that TLR4 can be activated by saturated FFAs,10,16 that TLR4 is required for LPS-mediated activation of NF- B in endothelial cells,17 and that palmitic acid, a saturated FFA, causes endothelial insulin resistance via a mechanism dependent on activation of IKK ,1 we sought to determine (1) whether the TLR4/MyD88 complex mediates the deleterious effects of fat excess on vascular tissue and, if so, (2) whether IKK and NF- B are downstream mediators of these responses. We demonstrate that in samples of thoracic aorta from mice made obese through consumption of a HF diet, IKK activity and subsequent NF- B dependent induction of IL-6 mRNA and ICAM protein expression are increased and that these responses are associated with cellular insulin resistance and impaired insulin-stimulated phosphorylation of eNOS (peNOS). Despite comparable increases of body weight and fat content, however, vascular inflammation and impaired insulin responsiveness were not evident in aortic samples taken from TLR4 / mice fed the same HF diet. To determine whether TLR4 signaling also mediates deleterious effects of FFAs in endothelial cells, we incubated human microvascular endothelial cells (HMECs) in the presence of palmitate. We found that palmitate-mediated activation of IKK was associated with NF- B transcriptional activation, as judged by induction of inflammatory markers ICAM and IL-6, and that these responses and the subsequent induction of insulin resistance in HMECs can be prevented either by decreasing TLR4 expression or by inhibiting MyD88 or IRAK signaling. Because inhibition of NF- B signaling also blocked the inhibitory effects of palmitate on endothelial insulin signaling and NO production, NF- B is strongly implicated as a mediator of palmitate induced endothelial dysfunction downstream of TLR4. These findings implicate the TLR4/MyD88 pathway in the deleterious effects of palmitate on endothelial insulin signaling and NO production, and are the first to document a key role for TLR4 in the mechanism whereby diet-induced obesity induces vascular inflammation and insulin resistance. Materials and Methods An expanded Materials and Methods section containing details for cell culture studies, Western blotting, siRNA transfection, NO measurement, ex vivo aortic studies, mouse studies, and analysis can be found in the online data supplement available at http://circres. ahajournals.org. Animals TLR4 / mice were kindly provided by Dr S. Akira (Osaka University, Osaka, Japan) and were backcrossed to the C57Bl/6 background (more than 9 generations) before study. Wild-type (WT) C57Bl/6 mice were purchased from Jackson Laboratories (Bar Harbor, Maine). Results Inflammation and Insulin Sensitivity of Aortic Tissue From WT and TLR4 / Mice Fed a HF Diet This study was undertaken to determine whether (dietinduced obesity, DIO) induced by a HF diet causes vascular Effect of Palmitate on Vascular Tissue Treated Ex Vivo As a first step to delineate mechanisms linking DIO to IKK activation and impaired insulin signaling in aortic tissue, we Downloaded from circres.ahajournals.org by on April 26, 2008 Kim et al TLR4 and Vascular Inflammation 1591 A low fat diet high fat diet B low fat diet high fat diet C low fat diet high fat diet A A 14 12 10 8 6 4 2 0 80 60 40 20 0 A IL-6 mRNA (Arb Units) * 100 ICAM (Arb Units) * 200 150 100 50 0 C57Bl6 TLR4 -/- * A C57Bl6 TLR4 -/- C57Bl6 TLR4 -/- D 350 300 pAKT (arb units) * veh ins E 10 8 6 4 2 0 veh * Ins 250 peNOS 200 A 150 100 50 0 Figure 1. The effect of 8 weeks of HF feeding on C57Bl/6 and TLR4 / murine aortic I B levels and aortic insulin signaling. Mice were given an IP injection of PBS or 2 U of insulin and then euthanized after 15 minutes. Thoracic aorta were removed and cell lysates were made. A, Western blot with an anti phospho-I B antibody were quantied using densitometry. B, Quantitation of ICAM-1 protein levels (by Western blot). C, IL-6 mRNA (by RT-PCR). D, pAkt levels as measured by ELISA. E, Western blot analysis using an anti-peNOS and total eNOS antibody. For each condition both peNOS and total eNOS levels were quantied using densitometry and the ratio of peNOS/eNOS was calculated. *P 0.05. I����p (Arb Units) C57Bl6 L H TLR4 -/L H C57Bl/6 L H TLR4 -/L H developed an ex vivo incubation assay system to determine whether brief exposure of aortic explants to palmitate, an abundant FFA in serum, mimics the deleterious effects of HF feeding. Using thoracic aorta taken from nonobese WT C57Bl/6 mice fed a standard chow diet, we found that ex vivo incubation with palmitate-BSA (100 mol/L) for 3 hours increased I B phosphorylation (a measure of IKK activity) by 5-fold compared with BSA alone (Figure 2A). Because aortic tissue samples contain many different cell types that might contribute to this inflammatory response, we next sought to determine whether palmitate exposure specifically impairs endothelial function in this assay. In support of this hypothesis, ex vivo insulin stimulation (100 nmol/L for 15 minutes) of thoracic aortic tissue taken from WT mice increased phosphorylation of eNOS serine 1177 (a response that is specific to endothelial cells, as other cell types do not express eNOS), and this effect was blocked by pretreatment with palmitate-BSA for 3 hours in the absence of any change of eNOS protein level. To determine whether these effects of palmitate are dependent on TLR4 signaling, we repeated these experiments using aortic samples obtained from chow-fed TLR4 / mice. Unlike the evidence of cellular inflammation and insulin resistance seen in vascular tissue from WT mice after incubation in palmitate for 3 hours, this intervention did not activate IKK , nor did it impair insulin-stimulated activation of peNOS (Figure 2A and 2B) in aortic tissue from TLR4-deficient mice. Thus, TLR4 signaling is required for vascular inflammation and insulin resistance induced by either chronic exposure to a HF diet or brief ex vivo exposure to palmitate. Role of TLR4 in the Response of Cultured Endothelial Cells to Palmitate To determine whether TLR4 signaling mediates palmitateinduced inflammation and insulin resistance in endothelial cells, we transfected HMECs with either TLR4 siRNA to decrease TLR4 expression or with scrambled siRNA as a Figure 2. The effect of 3 hour FFA treatment in isolated aortic tissue from C57Bl/6 and TLR4 / mice. A, Thoracic aorta from C57Bl/6 and TLR4 / were removed and placed into EBM media for 3 hours followed by treatment with BSA (control) alone or FFA (100 mol/L) for an additional 3 hours. Aortic segments were then treated with 100 nmol/L insulin or vehicle for 15 minutes. Cell lysates were analyzed for phospho-I B , and a representative blot is shown of 3 independent experiments. B, For each condition both peNOS and total eNOS levels were quantied using densitometry, and the ratio of peNOS/eNOS was calculated. *P 0.05. Downloaded from circres.ahajournals.org by on April 26, 2008 1592 Circulation Research June 8, 2007 Figure 3. Effect of inhibition of TLR4 by siRNA on FFA-mediated I B phosphorylation and endothelial insulin signaling. HMECs were transfected with either scrambled siRNA or TLR4 siRNA construct. A, Western blot with anti-TLR4 antibody. B, HMECs were treated with FFA (100 mol/L) for 3 hours, TNF- (10 ng/mL), or LPS (10 ng/mL) for 1 hour. Cell lysates were separated by SDSPAGE and Western blotted with an anti phospho-I B antibody. C, Transfected HMECs were stimulated with 100 nmol/L insulin or diluent control for 5 minutes, after 3 hours incubation with FFA (100 mol/L). Cell lysates were analyzed by Western blot with an antiphosphoAkt antibody. Densitometry was performed and fold increase over the untreated control (scrambled siRNA) was calculated (3 independent experiments). D, Insulin-mediated NO production as measured by cGMP assay using the same conditions as described in B. control. The reduction of TLR4 expression after transduction with siRNA to TLR4 was confirmed by Western blot with an anti-TLR4 antibody (Figure 3A). Transduced HMECs were treated with palmitate-BSA (100 mol/L for 3 hours) or with either of 2 other known activators of IKK TNF- (5 ng/mL for 1 hour), which does not signal via TLR4, or LPS (10 ng/mL for 1 hour), for which IKK activation is TLR4dependent (Figure 3B). In control HMECs (transduced with scrambled siRNA), palmitate, TNF- , and LPS each activated IKK , as assessed by increased I B phosphorylation. As expected, treatment with TLR4 siRNA blocked IKK activation in response to LPS, whereas the response to TNFremained intact despite reduced TLR4 signaling. Importantly, the ability of palmitate to activate IKK was also blocked by TLR4 siRNA, implicating TLR4 in the mechanism underlying IKK activation induced by palmitate and LPS, but not by TNF- , in endothelial cells. Thus, IKK appears to be a downstream mediator of TLR4-mediated inflammatory responses induced by palmitate. We next asked whether TLR4 is necessary for palmitatemediated impairment of endothelial insulin signaling and eNOS activity. Consistent with earlier results,1 insulinstimulated pAkt induction and NO production were inhibited in HMECs incubated with palmitate and scrambled control siRNA (Figure 3C). After treatment with TLR4 siRNA, however, these effects of palmitate were no longer detectable (Figure 3C and 3D), suggesting that TLR4 is necessary for palmitate-induced impairment of endothelial insulin signaling and NO production. Role of NF- B in Palmitate-Mediated Inhibition of NO Production in Endothelial Cells We previously reported that IKK is both necessary and sufficient for the inhibitory effect of palmitate on insulininduced NO production.1 One mechanism proposed to mediate this effect is that activation of IKK in turn activates NF- B, and that this response initiates a cascade of events that ultimately inhibit insulin-mediated Insulin-Receptor Substrate-1 (IRS-1) tyrosine phosphorylation, Akt serine 473 phosphorylation(pAkt), eNOS serine 1177 phosphorylation (peNOS), and NO production.1,20 According to this hypothesis, NF- B activation should be necessary for palmitateinduced inhibition of insulin-signaling and NO production in endothelial cells. To test this hypothesis, we used HMECs transduced with either a phosphorylation-resistant mutant of I B that blocks NF- B activation (NF- B super repressor)21 or control vector expressing GFP. Transduced HMECs were treated with palmitate-BSA (100 mol/L for 3 hours), TNF(5 ng/mL for 1 hour), or LPS (10 ng/mL for 1 hour). In control GFP-transduced HMECs, palmitate, TNF, and LPS each increased phosphorylation of I B . In contrast, this response was not seen in HMECs transduced with the Downloaded from circres.ahajournals.org by on April 26, 2008 Kim et al TLR4 and Vascular Inflammation 1593 A p-I����Con Control-GFP NF-�� super repressor FFA TNF LPS Con FFA TNF LPS B IP: IRS-1 IB: pY Con Ins FFA FFA/Ins Con Ins FFA FFA/Ins C Control-GFP NF-kB super repressor D 10 cGMP levels (Fold Increase) 8 6 4 2 0 Con Ins FFA FFA/Ins Control-GFP NF-kB super repressor Fold Increase in pAkt 25 20 15 10 5 0 Con Ins FFA FFA/Ins * * Figure 4. Effect of inhibition of NF- B signaling on Palmitate-mediated I B phosphorylation and endothelial insulin signaling. HMECs transduced with a phosphorylation resistant I B (NF- B super repressor) or GFP construct were treated with palmitate (100 mol/L), TNF- (10 ng/mL), or LPS (10 ng/mL). A, Cell lysates were analyzed by Western blot analysis with an antiphospho-I B antibody. B, Insulin mediated IRS-1 tyrosine phosphorylation C, Insulinmediated phospho-Akt (ser 473) quantied from Western blots (3 independent experiments) of HMECs transduced with GFP or NF- B super repressor. D, Insulin-mediated NO production as measured by cGMP assay. Fold increase over control (GFP-control) was calculated from 4 independent experiments. *P 0.05. I B -resistant mutant (Figure 4A), indicating that NF- B was not activated. Consistent with earlier results,1 insulinstimulated IRS-1 tyrosine phosphorylation, pAkt induction, and NO production were also inhibited by palmitate in control HMECs transduced with GFP (Figure 4B through 4D), whereas in HMECs transduced with NF- B super repressor, these responses to palmitate were no longer detectable (Figure 4B through 4D). Collectively, these results suggest that NF- B signaling acts downstream of TLR4 and is necessary for palmitate-induced impairment of endothelial insulin signaling and NO production. control vector was also attenuated by palmitate, and this inhibitory effect was prevented by transfection with DNMyD88 (Figure 5B through 5D). These data collectively indicate that, like TLR4 and NF- B, MyD88 is required for the deleterious effects of both palmitate and LPS on endothelial function. Role of IRAK in Endothelial Cell Responses to Palmitate IRAK-1 functions as an adapter protein within the TLR signaling complex. After receptor activation MyD88 binds to IRAK-1, which in turn activates IKK , and subsequently NF- B, a step known to be essential for cellular inflammation induced by LPS. To investigate the role of IRAK-1 in palmitate-induced activation of IKK in endothelial cells, HMECs were stably transfected with a gene encoding a truncated mutant of IRAK-1 that interrupts LPS-induced NF- B activation.22 As shown in Figure 5A, exposure to palmitate for 3 hours increased IKK activity in HMECs transduced with GFP, but not with the DNIRAK mutant. Similarly, insulin-stimulated IRS-1 tyrosine phosphorylation, serine phosphorylation of Akt, and production of NO were inhibited by exposure to palmitate in control, but not in DN-IRAK-transfected, HMECs (Figure 5B through 5D). To investigate whether NF- B is activated downstream of TLR4, we determined whether exposure to inflammatory stimuli increases IL-6 or ICAM expression, and whether inhibition of TLR4 signaling blocks this effect. Toward this end, HMECs were incubated with palmitate, LPS, or TNFand levels of IL-6 and ICAM were determined. Each of these inflammatory stimuli induced both IL-6 and ICAM expression, and in the case of palmitate and LPS, but not TNF- , Role of MyD88 Signaling on Palmitate-Induced Endothelial Dysfunction MyD88, by virtue of its direct recruitment to TLR4, is a proximal signaling molecule involved in LPS-induced NF- B activation. To determine whether MyD88 is also required for FFA-mediated activation of IKK , we used HMECs stably transfected with either a dominant negative (DN) mutant of MyD8822 or a control construct expressing GFP. After treatment with palmitate-BSA (100 mol/L for 3 hours), TNF- (5 ng/mL for 1 hour), or LPS (10 ng/mL for 1 hour; Figure 5A), GFP-transfected HMECs exhibited the expected increase of I B phosphorylation, indicating increased IKK activity. As predicted, HMECs transfected with DN-MyD88 displayed increased IKK activity in response to TNF- , because TNF- activates IKK via a mechanism that is not dependent on TLR4 or MyD88 signaling.23 In contrast, activation of IKK in response to either LPS or FFA was strongly attenuated by transfection with DN-MyD88. Furthermore, the ability of insulin (100 nmol/L insulin for 5 minutes) to increase IRS-1 tyrosine phosphorylation, Akt serine phosphorylation, and NO production in HMECs transfected with Downloaded from circres.ahajournals.org by on April 26, 2008 1594 Circulation Research June 8, 2007 A Control-GFP p-I����Con FFA TNF LPS Con FFA TNF LPS Con FFA TNF LPS DN-MyD88 DN-IRAK B IP: IRS-1 IB: pY Con Control-GFP DN-MyD88 DN-IRAK INS FFA FFA/INS Con INS FFA FFA/INS Con INS FFA FFA/INS C pAkt (fold increase) Ins Veh D Ins Veh * cGMP levels * Figure 5. Effect of dominant negative MyD88 and IRAK on FFA-mediated I B phosphorylation and endothelial insulin signaling. A, HMECs transduced with a DN-MyD88, DN-IRAK, or GFP construct were treated with FFA (100 mol/L), TNF- (10 ng/mL), or LPS (10 ng/mL). Cell lysates were analyzed by Western blot analysis with an antiphospho-I B antibody. B, Insulin-mediated IRS-1 tyrosine phosphorylation. C, Insulinmediated phospho-Akt (ser 473) quantied from Western blots (3 independent experiments) of HMECs transduced with either GFP, DN-MyD88, or DN-IRAK. D, Insulin-mediated NO production as measured by cGMP assay. Fold increase over control (GFP-control) was calculated as the mean of 4 independent experiments. *P 0.05. Con FFA Con FFA Con FFA Con FFA Con FFA Con FFA Con-GFP DN-MyD88 DN-IRAK Con-GFP DN-MyD88 DN-IRAK these effects were inhibited by DN-MyD88 and DN-IRAK compared with GFP controls (supplemental Figure III). These results indicate that like LPS, palmitate activates NF- B mediated gene transcription through a TLR4-dependent pathway in endothelial cells. Discussion Growing evidence implicates the cellular inflammatory response to nutrient excess as a key mechanism linking cardiovascular disease to obesity and related metabolic disorders.24 26 A steady increase in the prevalence of these disorders has heightened the need for an improved understanding of how nutrient excess affects the vasculature. Toward this end, the current work was undertaken to determine whether HF feeding induces vascular inflammation and insulin resistance in vivo, and to ascertain the role played by TLR4 in these effects. First, we replicated the observation that in WT mice, obesity induced by HF feeding is associated with elevated circulating levels of both insulin and FFA, consistent with the development of insulin resistance.27 In this context, our finding of IKK activation and impaired insulin signaling in aortic tissue from WT mice fed a HF diet suggests that DIO causes inflammation and insulin resistance within the vasculature as well as in conventional insulinsensitive tissues such as muscle, liver, and fat.26,28 Moreover, our finding that these vascular consequences of DIO did not occur in mice lacking TLR4, despite comparable increases of body weight and body fat mass, identifies TLR4 as a key mediator of vascular inflammation and insulin resistance in this setting. The hypothesis that TLR4 plays a key role to impair vascular insulin signaling in mice with DIO is consistent with findings from recent studies. For example, FFAs were re- cently shown to activate TLR4 signaling in adipocytes and macrophages, and the ability of FFA to activate inflammatory signaling in these cells was blocked in the absence of TLR4.7 Moreover, DIO-induced systemic insulin resistance is attenuated in TLR4 / mice7 and, in the current work, we found that insulin-mediated pAkt stimulation in liver is inhibited by DIO in WT mice, but not in mice lacking TLR4. Collectively, these findings implicate TLR4 as a key mediator of inflammation leading to insulin resistance under conditions of nutrient excess in many key tissues, including the vasculature. To determine whether exposure to palmitate, an abundant FFA in mammalian plasma, can recapitulate the vascular inflammation and insulin resistance induced in WT mice by DIO, and to determine whether this effect involves TLR4, we incubated samples of aortic tissue from WT and TLR4 / mice fed a LF diet in palmitate-containing media. As predicted, we found that whereas brief ex vivo incubation of aortic tissue from normal mice in palmitate-containing media increased IKK activity and reduced both insulin signaling and eNOS activity, these effects, like the response to HF feeding, were absent in vascular tissue from mice lacking TLR4. Deleterious effects of palmitate on isolated thoracic aorta, therefore, require intact TLR4 signaling. A key question raised by these findings is whether TLR4mediated vascular inflammation is mediated directly through vascular cells, through an indirect mechanism involving macrophages or other inflammatory cells, or whether both mechanisms contribute. Although additional studies are needed to address this question, our data indicate that in endothelial cells, TLR4 signaling is a critical mediator of palmitate-induced IKK and NF- B activation and subsequent decreases of insulin signaling and NO production. Downloaded from circres.ahajournals.org by on April 26, 2008 Kim et al The cellular mechanism responsible for palmitate-induced impairment of NO production appears to involve impaired activation of IRS-1, which in turn induces cellular insulin resistance.20 The IRS-1/PI3-kinase/pAkt/peNOS pathway is important for endothelial NO production because transfection with inhibitory mutants of IRS-1, PI3-kinase, or Akt prevent insulin-stimulated NO production.29,30 Furthermore, many of the metabolic abnormalities associated with diabetes and obesity, including elevated circulating TNF- and FFA levels, impair the IRS-1/PI3-kinase signaling pathway in a variety of cell types,20 and we have shown that exposing endothelial cells to TNF- impairs insulin- and fluid shear dependent NO production.31 Interestingly, excess glucose also impairs NO production through its inhibition of the IRS-1/pAkt/peNOS pathway in much the same way as shown for palmitate and TNF- .32 Conversely, NO signaling in endothelial cells for which IRS-1 is not required, such as via VEGF-dependent activation of peNOS, is not impaired by palmitate (data not shown). These observations collectively suggest that inhibition of IRS-1/pAkt/peNOS signaling is an important mechanism whereby endothelial insulin signaling and NO production are impaired by nutrient excess. In support of this hypothesis, we demonstrated that in cultured endothelial cells, reduced IRS-1 signaling and eNOS activation induced by palmitate are dependent on each of the key proteins in the TLR4 signaling pathway: TLR4, MyD88, IRAK, IKK , and NF- B. These results constitute strong direct support for a model in which palmitate activates TLR4, which in turn engages MyD88 and IRAK, subsequently activating IKK and NF- B. Activation of NF- B then inhibits IRS-1 tyrosine phosphorylation via an as yet unidentified mechanism. Among several candidate mediators of NF- Bmediated inhibition of IRS-1 signaling are S6kinase-1, mTOR (mammalian target of rapamycin),20 suppressor of cytokine signaling (SOCS3),33 c-JUN-NH2 (JNK), Akt,34 and several PKC isoforms. Each of these enzymes share in common the ability to inhibit IRS-1 signaling by phosphorylating one or more of its 70 IRS-1 serine residues, an IRS-1 modification known to impair IRS-1 signaling.20 Based on these collective results, TLR4 is strongly implicated in the development of endothelial inflammation, insulin resistance, and impaired NO production in lipotoxic conditions of FFA excess.35 Whether other cell types within the vasculature display similar responses in states of nutrient excess, and whether such effects are also dependent on TLR4 signaling, await further study. In this context, we note that our results do not exclude alternative mechanisms whereby TLR4 signaling may impair NO production. Among these is the possibility that reactive oxygen species are generated via the association between TLR4 and NADPH oxidase,36 and that these highly reactive molecules impair eNOS activity. Along with oleic and stearic acid, palmitic acid is 1 of 3 FFA species that constitute 70% of the total circulating FFA pool and is normally present in concentrations between 10 to 50 mol/L.37,38 Our cell culture and ex vivo studies used palmitic acid complexed with BSA at a concentration of 100 mol/L, slightly higher than is usually found in plasma. In other studies, however, we found that IKK is activated in endothelial cells by palmitate/BSA at concentrations as low TLR4 and Vascular Inflammation 1595 as 10 mol/L,1 which supports the physiological relevance of our findings and raises the interesting but untested possibility that endothelial cell insulin sensitivity and nitric oxide production are constrained by FFA at circulating levels that are within the physiological range. Combined with our current results, we hypothesize that TLR4 signaling plays a key role to mediate vascular responses to FFAs. If prolonged, vascular inflammation and endothelial dysfunction may contribute to atherosclerosis in the setting of obesity. A role for TLR4 in the pathogenesis of atherosclerosis is suggested by evidence that TLR4 is present both in human atherosclerotic plaque39 and in murine models of atherosclerosis, and that expression of this receptor in both endothelial cells and macrophages is upregulated by oxidized LDL, a key proinflammatory agent present during the development of atherosclerosis.40 Furthermore, deficiency of either TLR4 or MyD88 attenuates diet-induced atherosclerosis in apolipoprotein E / mice.15,41 Although TLR4 likely contributes to atherosclerosis progression via effects in macrophages and other inflammatory cells, results of the present study suggests that consumption of a diet high in saturated fats can induce vascular inflammation and insulin resistance through activation of local TLR4 signaling in endothelial and perhaps in other vascular cell types, and that these responses can impair nitric oxide production and favor atherosclerosis progression. Based on these collective observations, we conclude that the TLR4/MyD88 pathway is a key mediator in the mechanism linking DIO with vascular dysfunction. Sources of Funding This study was supported by NIH grants DK073878, HL04346 (to F.K.), DK52989, DK68384 (to M.W.S.), HL18645, HL67267 (to E.R.) and by the Royalty Research Fund and the Diabetes Endocrinology Research Center at the University of Washington (to F.K.). Body composition studies in mice were performed with support from the Clinical Nutrition Research Unit at the University of Washington. Disclosures None. References 1. Kim F, Tysseling KA, Rice J, Pham M, Haji L, Gallis BM, Baas AS, Paramsothy P, Giachelli CM, Corson MA, Raines EW. Free fatty acid impairment of nitric oxide production in endothelial cells is mediated by IKKbeta. Arterioscler Thromb Vasc Biol. 2005;25:989 994. 2. Roden M, Price TB, Perseghin G, Petersen KF, Rothman DL, Cline GW, Shulman GI. Mechanism of free fatty acid-induced insulin resistance in humans. J Clin Invest. 1996;97:2859 2865. 3. Shoelson SE, Lee J, Yuan M. Inflammation and the IKK beta/I kappa B/NF-kappa B axis in obesity- and diet-induced insulin resistance. Int J Obes Relat Metab Disord. 2003;27(Suppl 3):S49 S52. 4. Hotamisligil GS. Inflammatory pathways and insulin action. Int J Obes Relat Metab Disord. 2003;27(Suppl 3):S53S55. 5. Cai D, Yuan M, Frantz DF, Melendez PA, Hansen L, Lee J, Shoelson SE. Local and systemic insulin resistance resulting from hepatic activation of IKK-beta and NF-kappaB. Nat Med. 2005;11:183190. 6. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest. 2006;116:17931801. 7. Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 links innate immunity and fatty acid-induced insulin resistance. J Clin Invest. 2006;116:30153025. 8. Kitchens RL, Ulevitch RJ, Munford RS. Lipopolysaccharide (LPS) partial structures inhibit responses to LPS in a human macrophage cell line Downloaded from circres.ahajournals.org by on April 26, 2008 1596 Circulation Research June 8, 2007 26. Yki-Jarvinen H. Insulin resistance and endothelial dysfunction. Best Pract Res Clin Endocrinol Metab. 2003;17:411 430. 27. Burcelin R, Crivelli V, Dacosta A, Roy-Tirelli A, Thorens B. Heterogeneous metabolic adaptation of C57BL/6J mice to high-fat diet. Am J Physiol Endocrinol Metab. 2002;282:E834 E842. 28. Woods SC, Seeley RJ, Rushing PA, D��lessio D, Tso P. A controlled high-fat diet induces an obese syndrome in rats. J Nutr. 2003;133: 10811087. 29. Zeng G, Nystrom FH, Ravichandran LV, Cong LN, Kirby M, Mostowski H, Quon MJ. Roles for insulin receptor, PI3-kinase, and Akt in insulinsignaling pathways related to production of nitric oxide in human vascular endothelial cells. Circulation. 2000;101:1539 1545. 30. Montagnani M, Ravichandran LV, Chen H, Esposito DL, Quon MJ. Insulin receptor substrate-1 and phosphoinositide-dependent kinase-1 are required for insulin-stimulated production of nitric oxide in endothelial cells. Mol Endocrinol. 2002;16:19311942. 31. Kim F, Gallis B, Corson MA. TNF-alpha inhibits flow and insulin signaling leading to NO production in aortic endothelial cells. Am J Physiol Cell Physiol. 2001;280:C10571065. 32. Kim F, Tysseling KA, Rice J, Gallis B, Haji L, Giachelli CM, Raines EW, Corson MA, Schwartz MW. Activation of IKKbeta by glucose is necessary and sufficient to impair insulin signaling and nitric oxide production in endothelial cells. J Mol Cell Cardiol. 2005;39:327334. 33. Ueki K, Kondo T, Kahn CR. Suppressor of cytokine signaling 1 (SOCS-1) and SOCS-3 cause insulin resistance through inhibition of tyrosine phosphorylation of insulin receptor substrate proteins by discrete mechanisms. Mol Cell Biol. 2004;24:5434 5446. 34. Gao Z, Zuberi A, Quon MJ, Dong Z, Ye J. Aspirin inhibits serine phosphorylation of insulin receptor substrate 1 in tumor necrosis factortreated cells through targeting multiple serine kinases. J Biol Chem. 2003;278:24944 24950. 35. Unger RH. Lipotoxicity in the pathogenesis of obesity-dependent NIDDM. Genetic and clinical implications. Diabetes. 1995;44: 863 870. 36. Park HS, Chun JN, Jung HY, Choi C, Bae YS. Role of NADPH oxidase 4 in lipopolysaccharide-induced proinflammatory responses by human aortic endothelial cells. Cardiovasc Res. 2006;72:447 455. 37. Yli-Jama P, Meyer HE, Ringstad J, Pedersen JI. Serum free fatty acid pattern and risk of myocardial infarction: a case control study. J Intern Med. 2002;251:19 28. 38. Knopp RH, Retzlaff B, Walden C, Fish B, Buck B, McCann B. One-year effects of increasingly fat-restricted, carbohydrate-enriched diets on lipoprotein levels in free-living subjects. Proc Soc Exp Biol Med. 2000; 225:191199. 39. Edfeldt K, Swedenborg J, Hansson GK, Yan ZQ. Expression of toll-like receptors in human atherosclerotic lesions: a possible pathway for plaque activation. Circulation. 2002;105:1158 1161. 40. Xu XH, Shah PK, Faure E, Equils O, Thomas L, Fishbein MC, Luthringer D, Xu XP, Rajavashisth TB, Yano J, Kaul S, Arditi M. Toll-like receptor-4 is expressed by macrophages in murine and human lipid-rich atherosclerotic plaques and upregulated by oxidized LDL. Circulation. 2001;104:31033108. 41. Bjorkbacka H, Kunjathoor VV, Moore KJ, Koehn S, Ordija CM, Lee MA, Means T, Halmen K, Luster AD, Golenbock DT, Freeman MW. Reduced atherosclerosis in MyD88-null mice links elevated serum cholesterol levels to activation of innate immunity signaling pathways. Nat Med. 2004;10:416 421. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. without inhibiting LPS uptake by a CD14-mediated pathway. J Exp Med. 1992;176:485 494. Munford RS, Hall CL. Detoxification of bacterial lipopolysaccharides (endotoxins) by a human neutrophil enzyme. Science. 1986;234: 203205. Lee JY, Sohn KH, Rhee SH, Hwang D. Saturated fatty acids, but not unsaturated fatty acids, induce the expression of cyclooxygenase-2 mediated through Toll-like receptor 4. J Biol Chem. 2001;276: 1668316689. Lee JY, Plakidas A, Lee WH, Heikkinen A, Chanmugam P, Bray G, Hwang DH. Differential modulation of Toll-like receptors by fatty acids: preferential inhibition by n-3 polyunsaturated fatty acids. J Lipid Res. 2003;44:479 486. Takeda K, Kaisho T, Akira S. Toll-like receptors. Annu Rev Immunol. 2003;21:335376. Akira S, Uematsu S, Takeuchi O. Pathogen recognition and innate immunity. Cell. 2006;124:783 801. Pasterkamp G, Van Keulen JK, De Kleijn DP. Role of Toll-like receptor 4 in the initiation and progression of atherosclerotic disease. Eur J Clin Invest. 2004;34:328 334. Michelsen KS, Wong MH, Shah PK, Zhang W, Yano J, Doherty TM, Akira S, Rajavashisth TB, Arditi M. Lack of Toll-like receptor 4 or myeloid differentiation factor 88 reduces atherosclerosis and alters plaque phenotype in mice deficient in apolipoprotein E. Proc Natl Acad Sci U S A. 2004;101:10679 10684. Lee JY, Zhao L, Youn HS, Weatherill AR, Tapping R, Feng L, Lee WH, Fitzgerald KA, Hwang DH. Saturated fatty acid activates but polyunsaturated fatty acid inhibits Toll-like receptor 2 dimerized with Toll-like receptor 6 or 1. J Biol Chem. 2004;279:1697116979. Li X, Tupper JC, Bannerman DD, Winn RK, Rhodes CJ, Harlan JM. Phosphoinositide 3 kinase mediates Toll-like receptor 4-induced activation of NF-kappa B in endothelial cells. Infect Immun. 2003;71: 4414 4420. Foo SY, Nolan GP. NF-kappaB to the rescue: RELs, apoptosis and cellular transformation. Trends Genet. 1999;15:229 235. Tabibiazar R, Wagner RA, Spin JM, Ashley EA, Narasimhan B, Rubin EM, Efron B, Tsao PS, Tibshirani R, Quertermous T. Mouse strainspecific differences in vascular wall gene expression and their relationship to vascular disease. Arterioscler Thromb Vasc Biol. 2005;25: 302308. Zick Y. Ser/Thr phosphorylation of IRS proteins: a molecular basis for insulin resistance. Sci STKE. 2005;2005(268):pe4. Ferri N, Garton KJ, Raines EW. An NF-kappaB-dependent transcriptional program is required for collagen remodeling by human smooth muscle cells. J Biol Chem. 2003;278:1975719764. Bannerman DD, Tupper JC, Erwert RD, Winn RK, Harlan JM. Divergence of bacterial lipopolysaccharide pro-apoptotic signaling downstream of IRAK-1. J Biol Chem. 2002;277:8048 8053. Wesche H, Henzel WJ, Shillinglaw W, Li S, Cao Z. MyD88: an adapter that recruits IRAK to the IL-1 receptor complex. Immunity. 1997;7: 837 847. Beckman JA, Creager MA, Libby P. Diabetes and atherosclerosis: epidemiology, pathophysiology, and management. Jama. 2002;287: 2570 2581. Williams IL, Wheatcroft SB, Shah AM, Kearney MT. Obesity, atherosclerosis and the vascular endothelium: mechanisms of reduced nitric oxide bioavailability in obese humans. Int J Obes Relat Metab Disord. 2002;26:754 764. Downloaded from circres.ahajournals.org by on April 26, 2008