Saturated Fatty Acids Do Not Directly Stimulate Toll-Like Receptor Signaling
Objective— Toll-like receptors (TLRs) initiate inflammatory signaling in response to conserved microbial molecules. It has been proposed that dietary saturated fatty acids (SFAs) may also serve as endogenous ligands of TLR2 or TLR4, thereby promoting diseases associated with inflammation and dyslipidemia, including atherosclerosis and insulin resistance.
Methods and Results— We investigated the effects of SFAs on TLR-dependent signaling using a broad range of cell types and readouts. In HEK-293 cells transfected with TLR2, TLR4, or TLR5, SFAs complexed with fatty-acid-free bovine serum albumin (BSA)-stimulated TLR-dependent signaling. However, SFAs alone did not elicit a similar response. Further analysis showed that the effect seen with the complexed SFAs was attributable to LPS and lipopeptide contamination of fatty-acid-free BSA. Additional studies in macrophages, endothelial cells, smooth muscle cells, adipocytes, skeletal muscle cells, and human peripheral blood mononuclear cells confirmed the lack of stimulation of TLR-dependent signaling pathways or expression of TLR-target genes by SFAs.
Conclusions— SFAs do not directly stimulate TLR-dependent signaling, suggesting that alternative mechanisms link dietary fat intake with TLR-associated pathologies. LPS and lipopeptide contamination of the widely used reagent fatty-acid-free BSA explains the previously reported stimulation of TLR2 and TLR4 by SFAs.
Diets rich in saturated fat contribute to the development of disorders associated with dyslipidemia, such as insulin resistance and atherosclerosis.1,2 As inflammatory signaling pathways also contribute to these pathologies, a concerted effort is underway to identify potential mechanisms that may link dietary fat intake with inflammatory signaling. One proposed link has emerged from the observation that specific members of a family of receptors involved in innate immunity, termed Toll-like receptors (TLRs), appear to play a central role in the development of atherosclerosis and insulin resistance in animal models.3–6 For example, genetic deletion of TLR4 was shown to reverse high-fat diet–induced insulin resistance in mice,1,3,4,7 whereas deletion of TLR2, TLR4, or the shared TLR-signaling adaptor MyD88 significantly reduces atherosclerosis in apolipoprotein-E (ApoE)–deficient mice fed high-fat diets.2,5,6,8,9 It has therefore been suggested that modulation of TLR-signaling by dietary fats may partly explain the observed connection between dietary fat intake, insulin resistance, and atherosclerosis.1,10–13 Accordingly, identification of the ligands responsible for promoting TLR-signaling in these conditions is a key aim of current research in these fields.
The established ligands of the 10 human TLRs are conserved microbial molecules. For example, TLR2, with heterodimerization partners TLR1 or TLR6, detects bacterial lipopeptides (BLP), whereas TLR4, with the aid of the lipid-binding accessory protein MD2, detects enterobacterial lipopolysaccharide (LPS). TLR5 detects bacterial flagellin, whereas TLRs 3, 7, 8, and 9 appear to recognize nucleic acid motifs that are common in viral or bacterial pathogens (as reviewed in14). Depending on the cellular context, binding of such molecules to respective TLRs promotes the rapid induction of proinflammatory intracellular signaling pathways and the upregulation of cytokines, chemokines, and adhesion molecules that are required for an efficient antimicrobial response.14
Beyond these established microbial ligands, several molecules of endogenous (ie, host-derived) origin have also been proposed to serve as ligands for TLR2 or TLR4. Such proposals include molecules as diverse as fibronectin extra domain-A, heat-shock proteins, hyaluronan fragments, and high mobility box group-1 protein.15–17 Notably, a number of groups have independently proposed that saturated fatty acids (SFAs) may also directly stimulate TLR2- or TLR4-dependent signaling.1,10–13,18–25 This has led to the proposal that SFA-mediated induction of TLR-signaling is the likely mechanism linking dietary fat intake with diseases in which TLR-signaling plays a contributing role.1,10–13 The aim of the present study was to establish whether SFAs stimulate TLR-dependent signaling using a broad range of cell types and readouts. During the course of the studies, it emerged that contamination of a commonly used reagent to complex SFAs for use in such experiments may explain the previously reported effects of SFAs on TLR-signaling.
Materials and Methods
Reagents and Antibodies
Free and sodiated forms of the common dietary SFAs lauric acid (C12:0), myristic acid (C14:0), palmitic acid (C16:0), and stearic acid (C18:0) were purchased from Sigma and Nucheck, respectively. Three types of bovine serum albumin (BSA) from Sigma were examined in detail, including low endotoxin BSA (cat No. A9543, BSA-1), fatty acid-free (FAF)-BSA (A0281, BSA-2), and standard BSA (A3912, BSA-3). Human (A1887, HSA-4) and low-endotoxin bovine (A8806, BSA-5) FAF-BSAs were also examined. LPS, Pam3CSK4, FSL-1, and flagellin of Bacillus subtilis were from Invivogen. Antibodies to phospho-p38 MAPK, GAPDH, and IκBα were from Cell Signaling and Santa Cruz Biotechnology.
Preparation of SFAs, BSA, and SFA/BSA Complexes
SFAs were reconstituted to 10 mmol/L in ethanol, and diluted directly into warmed culture medium for experiments examining uncomplexed SFAs. Alternatively, sodiated SFAs were dissolved in warmed culture medium directly and filter sterilized. BSA was freshly reconstituted to 5 mg/mL (≈75 μmol/L) in culture medium and sterile filtered immediately before experiments. SFA/BSA complexes were prepared by incubating 200 μmol/L of each SFA with 40 μmol/L of FAF-BSA (a 5:1 molar ratio) at 37°C for 4 hours before experiments.
HEK-293 Cell TLR-Transfection Assay
TLR-transfection assays were performed as described previously.26 Briefly, TLR-deficient HEK-293 cells were plated in 96-well plates at 104 cells per well and transfected after 24 hours using Genejuice (Novagen) according to manufacturer’s instructions. Amounts of construct per well were 30 ng of human TLRs 2, 4 (coexpressing MD-2), or 5 (Invivogen), 30 ng of pCD14, and 10 ng of luciferase-reporter construct driven by the NF-κB sensitive E-selectin promoter (pELAM), with the balance made up with empty pCDNA3. Cells were grown for 72 hours after transfection before 18-hour challenge with each treatment in triplicate. Promoter expression is represented as fold induction relative to cells cultured in medium alone ±SD. Cells transfected with CD14 and reporters alone were insensitive to all tested PAMPs and BSA-containing reagents (supplemental Figure I).
Measurement of Inflammatory Mediators
RAW 264.7 macrophages were grown in DMEM supplemented with 10% FCS and plated at 4×105 cells per well of 96-well plates 24 hours before challenge. Cells were challenged with 100 ng/mL LPS, Pam3CSK4 or FSL-1, 200 μmol/L of each SFA or 1 mg/mL of each BSA. For coincubation experiments, cells received 100 μmol/L SFA with 0.1 to 1000 ng/mL Pam3CSK4 or LPS concurrently. In some experiments, samples were preincubated with 10 μg/mL polymyxin B for 10 minutes before challenge. After 4 hours, TNF-α secretion was measured by L929-cell bioassay as described previously.27 For cytokine array experiments, peripheral blood mononuclear cells (PBMCs) were prepared from venous blood of healthy human subjects, obtained with informed consent according to local ethical guidelines, as described previously.27 PBMCs were cultured in RPMI (Sigma) with 1% FCS and vehicle (1% ethanol), or a mixture of 25 μmol/L each of 4 SFAs (C12, C14, C16, and C18) to total 100 μmol/L, or 1 μg/mL E. coli LPS for 18 hours. Expression of cytokines in the supernatant was measured by antibody-based array (R&D) according to the manufacturer’s protocol. Human umbilical vein endothelial cell (HUVEC) E-selectin expression was measured by cell-based ELISA as described previously.26
Measurement of Intracellular Signaling Intermediates
RAW macrophages, plated at 1×106 cells per well of 6-well trays, were challenged with 200 μmol/L of each SFA, 100 ng/mL of LPS, Pam3CSK4 or FSL-1 or 1 mmol/L of each BSA, with or without 10 μg/mL polymyxin B. Alternatively, HUVECs, or cells of the A10 smooth muscle line, were challenged with 100 μmol/L of each SFA or 100 ng/mL LPS. Cells were lysed after 30 minutes and analyzed for expression of GAPDH, IκBα, and phosphorylated p38 MAPK by Western blot as described previously.26
Limulus amoebocyte lysate (LAL) assay kits were purchased from Quadratech EC (UK). 50 μL of each sample was diluted in endotoxin-free water and added to 50 μL of LAL reagent. Absorbance was measured at 405 nm after 50 minutes incubation at 37°C.
Treatment of Samples With Immobilized Lipases
A slurry of inert acrylic beads (Immobeads) covalently attached to lipase of Candida cylindracea or Candida antarctica (Sigma) was freshly prepared at a ratio of 1:10 (wt:vol) in PBS. 100 μL of PBS, or immobilized lipase beads of each type, was then added to 500 μL of 5 mg/mL FAF-BSA or 1 μg/mL Pam3CSK4 in sterile PBS, and incubated at 37°C with agitation for 18 hour. Beads were pelleted by centrifugation (13 000g, 10 minutes) and the TLR2-stimulating capacity of recovered supernatants was assayed using 293-TLR2 cells, or RAW macrophage TNF-α production.
Further methods are available in the supplemental materials (available online at http://atvb.ahajournals.org).
Effect of SFAs on TLR-Dependent Signaling
The previously reported capacity of SFAs complexed with fatty acid free (FAF)-BSA to stimulate TLR2- and TLR4-dependent signaling1,10–13,18–25 was readily confirmed using TLR-deficient HEK-293 cells transfected with TLR2 or TLR4 (Figure 1A and 1B). Each of the SFA/BSA complexes tested (from C12 to C18) stimulated significant induction of signaling in TLR2 and TLR4 transfectants, although this was not attributable to nonspecific cellular activation as TLR5 transfectants were not sensitive to SFA/BSA complexes (Figure 1C). However, when SFAs were applied to cells at up to 500 μmol/L in the absence of BSA, no TLR2-, TLR4-, or TLR5-dependent signaling was observed (Figure 1D through 1F), while respective transfectants were sensitive to as little as 10 ng/mL Pam3CSK4, 0.1 ng/mL LPS, and 100 ng/mL flagellin (supplemental Figure II). Similar results were observed using sodiated forms of each fatty acid (supplemental Figure III). To determine whether BSA itself may serve as a ligand for TLR2 or TLR4, sterile-filtered solutions of 3 different commercially sourced BSAs were assayed in TLR-transfected cells. Remarkably, fatty-acid–free BSA potently induced TLR2 and TLR4, but not TLR5-dependent signaling, whereas the other 2 BSA preparations did not promote detectable TLR2, TLR4, or TLR5 signaling (Figure 2A through 2C). Examination of 2 further fatty acid-free albumins (1 human and 1 bovine) revealed significant TLR2 or TLR4 contaminants in these products also (supplemental Figure IV).
Effect of SFAs and BSA on Established TLR-Induced Mediators
Stimulation of signaling via any of the TLRs in macrophages results in the induction of a characteristic set of mediators, including the induction of NF-κB and p38 MAP-kinase signaling cascades and subsequent release of inflammatory cytokines, such as TNF-α.26 Treatment of RAW macrophages with SFAs did not induce these mediators, whereas fatty-acid free BSA, but not untreated BSA, induced phosphorylation of p38 MAPK, degradation of IκBα, and secretion of TNF-α (Figure 3A and 3B). SFAs did not induce expression of E-selectin in endothelial cells (Figure 3C), nor degradation of IκBα in endothelial cells or aortic smooth muscle cells (Figure 3D and 3E). SFAs also did not promote IκBα degradation, p38 MAPK phosphorylation, or induction of IL-6, TNF-α, or CCL-2 mRNA in cultured 3T3-L1 adipocytes or C2C12 skeletal muscle cells (supplemental Figure V). Likewise, whereas human PBMC produced TNF-α in response to ligands of TLR2, TLR4, or TLR5, they did not do so in response to SFAs (supplemental Figure VI). To further establish whether SFAs could upregulate markers of TLR-stimulation, cytokine array profiling was performed on supernatants from human PBMCs treated with vehicle (ethanol), 100 μmol/L of each SFA, or LPS. SFAs had clear biological effects on PBMCs, increasing their secretion of IL-1Ra, MIF, and IL-8. By contrast, TLR4-stimulation promoted a quite different profile of cytokine release, including IL-1β, TNF-α, GROα, CCL1, MIP-1α, IL-6, MIP1β, and Serpin E1 that were not induced by SFAs (supplemental Figure VII). SFAs also did not enhance the sensitivity of RAW macrophages or HEK-293 cells transfected with TLR2 or TLR4 toward Pam3CSK4 or LPS (supplemental Figures VIII & IX).
Effect of SFA/BSA Complexes on TLR-Dependent Signaling
We next sought to determine whether SFAs complexed with noncontaminated BSA could promote TLR-signaling. After removal of endogenous fatty acids from uncontaminated BSA-3 using activated charcoal at low pH, no fatty acids were detectable, in common with commercially-sourced FAF-BSAs (Figure 4A). The capacity of freshly prepared FAF-BSA-3 to complex SFAs was then compared with commercial FAF-BSA-2. After size exclusion chromatography to remove unbound SFAs, both types were found to efficiently form stable complexes containing ≈3 mols SFA per mol BSA (Figure 4A), similar to that observed in BSAs containing endogenous FAs (BSA-1,3). The freshly generated SFA/BSA complexes did not induce stimulation of TLR2- or TLR4-dependent signaling in HEK-293 transfectants, or TNF-α production in RAW macrophages (Figure 4B through 4D).
Effect of Polymyxin-B on Lipopeptide Signaling
Several previous studies of SFA function have discounted contamination of reagents with bacterial molecules through the use of polymyxin-B, an antibiotic that sequesters and thereby inactivates contaminating lipopolysaccharides (TLR4 agonists).22,23,25 We confirmed that polymyxin-B cotreatment efficiently blocked LPS-induced TNF-α secretion, IκBα degradation and p38 MAPK phosphorylation in RAW macrophages (Figure 5A and 5B). However, polymyxin-B treatment had no effect on the induction of these mediators in RAW macrophages treated with the synthetic di-acyl agonist of the TLR2/6 heterodimer FSL-1, or the synthetic triacyl agonist of the TLR2/1 heterodimer Pam3CSK4 (Figure 5A and 5B). In several other previous studies of SFA function, contamination of reagents with bacterial molecules has been ruled out by examining the capacity of reagents to stimulate the limulus (LAL) reaction.1,10,12 However, it was found that although this assay is highly sensitive to LPS contamination, it is insensitive to di-acyl or triacyl lipopeptides, or to flagellin (Figure 5C).
Effect of Lipase Treatment on Biological Activity of Lipopeptide Contaminants
As polymyxin-B treatment and LAL assays were found to be unsuitable controls for contamination of reagents with bacterial lipopeptide, alternative control methodologies were sought. Treatment of Pam3CSK4 with lipases derived from C antarctica or C cylindracea was therefore attempted, because covalent acylation of lipopeptides is a requirement for their capacity to stimulate TLR2 signaling.28 Accordingly, such treatment was found to reduce the capacity of Pam3CSK4 to stimulate TNF-α secretion from macrophages or TLR2-signaling in HEK-293-TLR2 cells by 99% and 96% respectively (Figure 6A and 6C). Likewise, lipase treatment of FAF-BSA reduced TNF-α and TLR2-signaling induced by this reagent by up to 97% and 92%, respectively (Figure 6B and 6D), indicating that the TLR2-dependent signaling of FFA-free BSA is likely to be attributable to contamination with bacterial lipopeptides.
As diets rich in SFAs are established to promote insulin resistance and atherosclerosis in mice,1,2,5,7 and beacause it has emerged that the innate immune receptors TLR2 and TLR4 play contributing roles in these diseases,1–9 it has been suggested that SFAs may promote these conditions via direct stimulation of TLR2- or TLR4-dependent signaling.1,10–13 Indeed, several groups have reported that SFAs directly induce either TLR2- or TLR4-dependent signaling in cultured macrophages and transfected cells.1,10–13,18–25
However, an inherent difficulty with studies of novel potential TLR-ligands is the fact that bacterial LPS and lipopeptides, the established ligands of TLR2 and TLR4, are frequent contaminants of the reagents required to perform these studies. For example, the proposed proinflammatory properties of both heat shock protein-60 and C-reactive protein were shown subsequently to be attributable to LPS contamination of the recombinant proteins examined.29,30 The present study therefore aimed to establish whether SFAs may indeed directly stimulate TLR-dependent signaling in cell types relevant to atherosclerosis, or whether LPS and lipopeptide contamination may instead account for the previously reported properties of SFAs.
In a screen of a wide variety of cell types, including macrophages, adipocytes, smooth muscle cells, and endothelial cells, we found that SFAs did not induce expression of gene products that are typically upregulated by TLR-stimulation, such as IL-1β, TNF-α, CCL-2, and E-selectin. SFAs also did not promote the phosphorylation of p38 MAPK or degradation of IκBα, which are established to be universal features of TLR-stimulation in macrophages.14 Finally, in transfected HEK-293 cells, which are sensitive to very low concentrations of respective TLR-ligands, SFAs, even at supraphysiological levels, neither induced nor enhanced TLR2 or TLR4 signaling (Figure 1 and supplemental Figure IX). Thus, although our studies are limited to the examination of isolated cells in vitro and therefore cannot rule out potential effects in vivo, our results support recent biochemical evidence that SFAs do not physically bind to TLR4 or MD2,13 and suggest that SFAs do not directly stimulate TLR-signaling in these systems.
Instead, we present evidence that fatty-acid–free BSA, the reagent used in previous studies to present SFAs to cells in a physiological form, can be contaminated with both LPS and lipopeptide, and therefore likely accounts for the results of these previous studies.1,10–13,18–25 Notably, the issue of potential contamination of reagents with bacterial products was addressed in several of these studies by use of either the limulus assay,1,10,12 which detects LPS, or polymyxin-B,22,23,25 which neutralizes LPS (as summarized in supplemental Table I). However, we found that these 2 widely used control methodologies neither detect nor neutralize TLR2-stimulating lipopeptide contaminants (Figure 5). Instead, we show that treatment of samples with immobilized lipases can efficiently remove the TLR2-stimulating activity of lipopeptides from contaminated reagents (Figure 6). This technique may therefore be of use in future studies of novel candidate TLR-agonists, as a screen of other commonly used laboratory reagents revealed that TLR2 or TLR4 stimulants can also be found in restriction enzymes, occasional azide-free antibodies and casein (supplemental Figure X), suggesting that lipopeptide contamination of laboratory reagents may be more widespread than previously appreciated.
It should be noted that the current findings do not refute the possibility that TLR2 or TLR4 may be involved in the potentiation of insulin resistance or atherosclerosis by fatty diets. Indeed, the data from in vivo investigations of the roles of TLR2 and TLR4 in atherosclerosis and insulin resistance seem well controlled, robust, and conclusive. Diet-induced insulin resistance is clearly reduced in TLR4-deficient animals,1,3,4,7 and aortic plaque burden has been shown to be significantly reduced in TLR2-, TLR4-, or MyD88-deficient mice by several independent groups.2,5,6,8,9 Moreover, numerous studies have shown that experimental administration of the established ligands of TLR2 or TLR4 (ie, lipopeptide or LPS), in the purified forms to animals markedly enhances insulin resistance31 and atherosclerosis.5,6,9
Thus, although the present study indicates that SFAs do not promote these pathologies by direct stimulation of TLR-signaling, it seems clear that an alternative TLR-dependent mechanism links dietary fat intake with insulin resistance and atherosclerosis. One possibility that is currently being investigated by ourselves and other researchers is that the large quantities of lipopeptide and LPS derived from the commensal organisms of the mammalian intestine may contribute to systemic stimulation of TLR2 or TLR4 signaling. We showed recently that a high-fat meal markedly increases circulating concentrations of LPS in healthy human subjects postprandially.32 Likewise, in mice, it has been shown that the mechanism of absorption of intestinal endotoxin is similar to that of fat-soluble vitamins, being via incorporation into chylomicrons and subsequent transport to the circulation via the thoracic duct.33 With this mechanism in mind, and the recent observation that around 0.2% of orally-ingested radiolabeled LPS can cotranslocate with dietary fat into the circulation,33 we were interested to find that standard mouse chow can contain large quantities of both lipopeptide and LPS (supplemental Figure X). However, although this finding is likely to be of relevance to studies of germ-free animals, as the human intestine contains up to 1 gram of LPS, the commensal-derived pools of these agents seem more likely to play a greater role in modulating diet-dependent inflammatory processes.
Taken together, the findings of the present study suggest that SFAs do not directly stimulate TLR-signaling and that the mechanisms linking high-fat diets with TLR-associated pathologies such as atherosclerosis and insulin resistance remain to be discovered. Further studies will be required to investigate whether alternative mechanisms, such as diet-induced translocation of intestinal endotoxins, may contribute to the TLR-dependent component of common chronic inflammatory diseases.
Sources of Funding
This work was supported by a University of Leicester Fellowship (awarded to C.E.). N.J.S. is supported by a Chair funded by the British Heart Foundation.
Received March 10, 2009; revision accepted July 28, 2009.
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 signalling pathways. Nat Med. 2004; 10: 416–421.
Poggi M, Bastelica D, Gual P, Iglesias MA, Gremeaux T, Knauf C, Peiretti F, Verdier M, Juhan-Vague I, Tanti JF, Burcelin R, Alessi MC. C3H/HeJ mice carrying a toll-like receptor 4 mutation are protected against the development of insulin resistance in white adipose tissue in response to a high-fat diet. Diabetologia. 2007; 50: 1267–1276.
Tsukumo DM, Carvalho-Filho MA, Carvalheira JB, Prada PO, Hirabara SM, Schenka AA, Araújo EP, Vassallo J, Curi R, Velloso LA, Saad MJ. Loss-of-function mutation in Toll-like receptor 4 prevents diet-induced obesity and insulin resistance. Diabetes. 2007; 56: 1986–1998.
Kim F, Pham M, Luttrell I, Bannerman DD, Tupper J, Thaler J, Hawn TR, Raines EW, Schwartz MW. Toll like receptor-4 mediates vascular inflammation and insulin resistance in diet-induced obesity. Circ Res. 2007; 100: 1589–1596.
Michelson KS, Wong MH, Shah PK, Zhang W, Yano J, Doherty TM, Akira S, Rajavashisth, 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.
Staiger H, Staiger K, Stefan N, Wahl HG, Machicao F, Kellerer M, Haring HU. Palmitate-induced interleukin-6 expression in human coronary artery endothelial cells. Diabetes. 2004; 53: 3209–3216.
Senn JJ. Toll-like receptor-2 is essential for the development of palmitate-induced insulin resistance in myotubes. J Biol Chem. 2006; 281: 26865–26875.
Nguyen MT, Favelyukis S, Nguyen AK, Reichart D, Scott PA, Jenn A, Liu-Bryan R, Glass CK, Neels JG, Olefsky JM. A subpopulation of macrophages infiltrates hypertrophic adipose tissue and is activated by free fatty acids via Toll-like receptors 2 and 4 and JNK-dependent pathways. J Biol Chem. 2007; 282: 35279–35292.
Okamura Y, Watari M, Jerud ES, Young DW, Ishizaka ST, Rose J, Chow JC, Strauss JF III. The extra domain A of fibronectin activates Toll-like receptor 4. J Biol Chem. 2001; 276: 10229–10233.
Termeer C, Benedix F, Sleeman J, Fieber C, Voith U, Ahrens T, Miyake K, Freudenberg M, Galanos C, Simon JC. Oligosaccharides of Hyaluronan activate dendritic cells via toll-like receptor 4. J Exp Med. 2002; 195: 99–111.
Park JS, Svetkauskaite D, He Q, Kim JY, Strassheim D, Ishizaka A, Abraham E. Involvement of toll-like receptors 2 and 4 in cellular activation by high mobility group box 1 protein. J Biol Chem. 2004; 279: 7370–7377.
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.
Lee JY, Ye J, Gao Z, Youn HS, Lee WH, Zhao L, Sizemore N, Hwang DH. Reciprocal modulation of Toll-like receptor-4 signaling pathways involving MyD88 and phosphatidylinositol 3-kinase/AKT by saturated and polyunsaturated fatty acids. J Biol Chem. 2003; 278: 37041–37051.
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: 16971–16979.
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: 16683–16689.
Weatherill AR, Lee JY, Zhao L, Lemay DG, Youn HS, Hwang DH. Saturated and polyunsaturated fatty acids reciprocally modulate dendritic cell functions mediated through TLR4. J Immunol. 2005; 174: 5390–5397.
Suganami T, Tanimoto-Koyama K, Nishida J, Itoh M, Yuan X, Mizuarai S, Kotani H, Yamaoka S, Miyake K, Aoe S, Kamei Y, Ogawa Y. Role of the Toll-like receptor 4/NF-kappaB pathway in saturated fatty acid-induced inflammatory changes in the interaction between adipocytes and macrophages. Arterioscler Thromb Vasc Biol. 2007; 27: 84–91.
Suganami T, Nishida J, Ogawa Y. A paracrine loop between adipocytes and macrophages aggravates inflammatory changes: Role of free fatty acids and tumor necrosis factor alpha. Arterioscler Thromb Vasc Biol. 2005; 25: 2062–2068.
Erridge C, Kennedy S, Spickett CM, Webb DJ. Oxidised phospholipid inhibition of toll-like receptor (TLR) signalling is restricted to TLR2 and TLR4 - roles for CD14, LPS-binding protein and MD2 as targets for specificity of inhibition. J Biol Chem. 2008; 283: 24748–24759.
Delahooke D, Barclay G, Poxton I. TNF-α induction by an aqeous phenol-extracted lipopolysaccharide complex from Bacteroides species. Infect Immun. 1996; 63: 840–846.
Gao B, Tsan MF. Recombinant human heat shock protein 60 does not induce the release of tumor necrosis factor alpha from murine macrophages. J Biol Chem. 2003; 278: 22523–22529.
Taylor KE, Giddings JC, van den Berg CW. C-reactive protein-induced in vitro endothelial cell activation is an artefact caused by azide and lipopolysaccharide. Arterioscler Thromb Vasc Biol. 2005; 25: 1225–1230.
Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D, Neyrinck AM, Fava F, Tuohy KM, Chabo C, Waget A, Delmée E, Cousin B, Sulpice T, Chamontin B, Ferrières J, Tanti JF, Gibson GR, Casteilla L, Delzenne NM, Alessi MC, Burcelin R. Metabolic endotoxemia initiates obesity and insulin resistance. Diabetes. 2007; 56: 1761–1772.
Erridge C, Attina T, Spickett CM, Webb DJ. A high-fat meal induces low-grade endotoxemia: evidence of a novel mechanism of postprandial inflammation. Am J Clin Nutr. 2007; 86: 1286–1292.
Ghoshal S, Witta J, Zhong J, de Villiers W, Eckhardt E. Chylomicrons promote intestinal absorption of lipopolysaccharides. J Lipid Res. 2009; 50: 90–97.