Nutrient Modification of the Innate Immune Response: A Novel … · Nutrient Modification of the...

ISSN: 1524-4636 Copyright © 2010 American Heart Association. All rights reserved. Print ISSN: 1079-5642. Online

7272 Greenville Avenue, Dallas, TX 72514Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association.

DOI: 10.1161/ATVBAHA.109.201681 2010;

2010;30;802-808; originally published online Jan 28,Arterioscler Thromb Vasc BiolAnand, Phyllis S. Laine, Yali Su and Peter D. Reaven

Eric A. Schwartz, Wei-Yang Zhang, Sheetal K. Karnik, Sabine Borwege, Vijay R. Which Saturated Fatty Acids Greatly Amplify Monocyte Inflammation

Nutrient Modification of the Innate Immune Response: A Novel Mechanism by

http://atvb.ahajournals.org/cgi/content/full/ATVBAHA.109.201681/DC1Data Supplement (unedited) at:

http://atvb.ahajournals.org/cgi/content/full/30/4/802

located on the World Wide Web at: The online version of this article, along with updated information and services, is

http://www.lww.com/reprintsReprints: Information about reprints can be found online at

[email protected]. E-mail:

Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters

http://atvb.ahajournals.org/subscriptions/Biology is online at Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular

at UNIV OF CALIFORNIA LIB on April 13, 2010 atvb.ahajournals.orgDownloaded from

http://atvb.ahajournals.org/cgi/content/full/30/4/802

http://atvb.ahajournals.org/cgi/content/full/ATVBAHA.109.201681/DC1

http://atvb.ahajournals.org/subscriptions/

mailto:[email protected]

http://www.lww.com/reprints

http://atvb.ahajournals.org

Nutrient Modification of the Innate Immune ResponseA Novel Mechanism by Which Saturated Fatty Acids Greatly Amplify

Monocyte Inflammation

Eric A. Schwartz, Wei-Yang Zhang, Sheetal K. Karnik, Sabine Borwege, Vijay R. Anand,Phyllis S. Laine, Yali Su, Peter D. Reaven

Objective—Monocyte/macrophage inflammation is an important contributor to diabetes and cardiovascular disease.Studies have suggested saturated fatty acids (SFA) induce monocyte inflammation in a Toll-like receptor-4–dependentmanner, but recent data suggest SFA do not directly interact with Toll-like receptor-4. The present study tests the novelhypothesis that metabolism of SFA cooperatively amplifies Toll-like receptor-4–mediated inflammation.

Methods and Results—THP-1 monocytes exposed to 100 �mol/L SFA in vitro for 16 hours followed by 1 ng/mLlipopolysaccharide demonstrated enhanced IL-6 and IL-8 mRNA and protein expression (�3-fold higher than the sum ofindividual responses to SFA and lipopolysaccharide). SFA had similar effects on THP-1 macrophages and primary humanmonocytes. This amplified lipopolysaccharide response could be blocked by inhibition of SFA metabolism to ceramide andrestored by cell-permeable ceramide. Both SFA and ceramide activated PKC-� and the mitogen-activated protein kinases Erk,JNK, and p38. Inhibition of these pathways prevented the SFA-induced increase in cytokine expression.

Conclusion—These results provide evidence for potent amplification of monocyte/macrophage innate immune responsesby a novel pathway requiring metabolism of SFA to ceramide and activation of PKC-�/mitogen-activated proteinkinases. These findings demonstrate how nutrient excess may modulate innate immune system activation and possiblycontribute to development of diabetes and cardiovascular disease. (Arterioscler Thromb Vasc Biol. 2010;30:802-808.)

Key Words: ceramide � mitogen-activated protein kinase � monocytes � protein kinase C � Toll-like receptor

In vitro and animal studies have implicated chronic inflam-mation in the development and progression of insulin

resistance and type 2 diabetes, and in the excess cardiovas-cular disease risk associated with diabetes.1 These findingsclosely parallel results from epidemiological studies thatdemonstrate plasma inflammatory markers (such as the acutephase proteins, C-reactive protein and serum amyloid A, andthe proinflammatory cytokines, IL-6, IL-1�, and possiblytumor necrosis factor-�) as predictors, and potential media-tors, of these conditions.1–3

See accompanying article on page 692Monocytes and macrophages are critical cells in the devel-

opment of insulin resistance and cardiovascular disease. Theyaccumulate early in atherosclerotic lesion development andremain throughout the transition to advanced atheroma.4 Theyare closely linked to the chronic inflammatory processes thatunderlie plaque formation. Similarly, macrophages and mac-rophage-derived proinflammatory factors have been detectedin adipose and other tissues5 and appear to contribute todevelopment of insulin resistance and hyperglycemia.6,7 De-termining what factors play important roles in activating

monocytes/macrophages and contribute to systemic inflam-mation is essential to understanding the pathophysiology ofinsulin resistance and cardiovascular disease.

Bacterial lipopolysaccharide (LPS) is a potent inducer ofinflammation in many cell types via activation of Toll-likereceptor (TLR)-4. TLR4 activity is mediated in large part by thetranscription factor NF-�B, a known regulator of �200 proin-flammatory genes.8 LPS is frequently present in low concentra-tions in blood, resulting from its release by the bacterial flora ofthe intestines, from subclinical infections, and from the presenceof LPS in food. Plasma LPS is further elevated in the settings ofobesity, insulin resistance, and type 2 diabetes, as well as duringthe postprandial period, especially after high-fat meals.9,10 Al-though these low concentrations of LPS do not appear to causeacute systemic inflammation, several lines of investigation haveimplicated TLR signaling in chronic inflammatory diseaseprocesses. Recent studies have demonstrated that mice lackingTLR4 or related signaling components, such as MyD88, havereduced susceptibility to atherosclerosis.11–13 Furthermore, Caniet al9 have demonstrated that subacute LPS elevations causeinflammation, initiate insulin resistance, and even contribute toobesity in animals.

Received on: March 30, 2009; final version accepted on: December 23, 2009.From Phoenix VA Health Care System (E.A.S., W.-Y.Z., S.K.K., S.B., V.R.A., P.S.L., P.D.R.), Phoenix, Ariz; Arizona State University (E.A.S.),

Tempe, Ariz; Kronos Science Laboratories (Y.S.), Phoenix, Ariz; University of Arizona (P.D.R.), Tucson, Ariz.Correspondence to Eric A. Schwartz, PhD, Research Health Scientist, Phoenix VA Health Care System, 650 E. Indian School Rd, Phoenix, AZ

85012-1892. E-mail [email protected]© 2010 American Heart Association, Inc.

Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.109.201681

802 at UNIV OF CALIFORNIA LIB on April 13, 2010 atvb.ahajournals.orgDownloaded from


TLR also recognize numerous endogenous molecules.Studies by Hwang, Flier, and others14–16 have suggested thatelevated concentrations of saturated fatty acids (SFA) lead toactivation of TLR2 and TLR4. Because Western diets aretypically enriched in SFA, excess nutritional intake may in-crease plasma and tissue SFA concentrations, potentiallyinducing inflammation and contributing to development ofinsulin resistance.17 Reduced insulin action increases adiposetissue lipolysis and hepatic production of triglyceride-rich li-poproteins, this may further increase concentrations of plasmafree fatty acids (FFA), including SFA, Therefore, elevatedplasma concentrations of SFA may facilitate the initiation andprogression of insulin resistance18–20 and, through this mecha-nism, as well as direct vascular inflammation, contribute todevelopment of diabetes and cardiovascular disease.21

We have previously demonstrated direct proinflammatoryeffects of mixtures of FFA, as well as physiologicallyrelevant concentrations of SFA, on human monocytes andmacrophages in vitro, consistent with low-level activation ofTLR by SFA.22,23 However, recent studies have called intoquestion the ability of TLR to directly bind SFA as a ligandand suggest that SFA might modulate TLR activity via lipid raftchanges24 or that SFA may not induce inflammation via TLR-mediated signaling at all.25 We now describe the novel findingthat SFA greatly amplify the proinflammatory cytokine re-sponses of human monocytes and macrophages to low, physio-logically relevant concentrations of LPS through a mechanismseparate from, but parallel to, the TLR4 signaling pathway.

Materials and MethodsIn typical experiments, cells (monocytes and macrophages) wereexposed to FFA, LPS, specific pharmacological inhibitors, or com-binations of these reagents. Cells were typically incubated with FFAor inhibitors for 16 hours, washed in phosphate-buffered saline, andresuspended in fresh media with or without 1 ng/mL Escherichia coliLPS for 30 minutes (for phosphoprotein assays), 4 hours (forreal-time polymerase chain reaction), or 48 hours (for conditionedmedia enzyme-linked immunosorbent assay). Analysis of monocyteinflammation and gene expression was performed using standardlaboratory methods or commercially available kits. For detailedmaterials and methods, see the online Supplementary materialavailable at http://atvb.ahajournals.org.

ResultsCompared to control cells, THP-1 monocytes exposed to 1ng/mL LPS for 4 hours exhibited a 21-fold increase in IL-6mRNA expression and a 10-fold increase in IL-8 mRNAexpression (Figure 1A). Exposure of these cells to100 �mol/L palmitic acid (PA) for 16 hours induced a 7-foldand 2-fold increase in mRNA for these 2 cytokines, respec-tively. However, exposure of the cells to PA followed by LPSresulted in an 80-fold increase in IL-6 and a 53-fold increasein IL-8 mRNA expression, compared to controls. Theseresults were significantly (P�0.05) higher than the sum of theseparate effects of PA and LPS. Confirmatory experimentswith other saturated (Figure 1A) and unsaturated fatty acids,and with both macrophages and primary human monocytes,are described elsewhere (see Supplementary Results andSupplementaryFiguresI,II,availableathttp://atvb.ahajournals.org). Similar to LPS, PA-induced amplification of the mono-

cyte response to the TLR2 ligand Pam3CSK4 was alsoobserved (Supplemental Figure III).

Exposure of monocytes to SFA also enhanced LPS-inducedsecretion of cytokine proteins. Compared to controls, monocytesexposed to PA for 16 hours demonstrated increased IL-6 proteinsecretion over the subsequent 48 hours (Figure 1B). Exposure ofmonocytes to 1 ng/mL LPS alone for 48 hours did not increaseIL-6 protein secretion. In contrast, exposure of monocytes to PAfollowed by LPS increased IL-6 protein secretion by nearly4-fold; similar fold increases were observed for IL-8. Theseprotein changes confirm the existence of the synergistic increasein inflammation that was suggested by mRNA responses.Although there was a detectable increase in IL-6 and IL-8mRNA in response to 4 hours of LPS treatment alone, proteinsecretion in response to 48 hours of LPS alone was not differentfrom controls. In contrast, pretreatment with PA induced a

Figure 1. Exposure of monocytes to physiologically relevantconcentrations of SFA amplifies subsequent responses to LPS.A, THP-1 monocytes were exposed for 16 hours to media alone(C) or to media containing 100 �mol/L lauric acid (LaA), palmiticacid (PA), or stearic acid (SA); to 1 ng/mL LPS (LPS) for 4hours; or to LaA, PA, or SA followed by LPS (LaA LPS, PA LPS,or SA LPS, respectively). IL-6 (white bars) and IL-8 (black bars)mRNA was measured by real-time polymerase chain reaction.B, THP-1 monocytes were exposed to PA or LPS as in (A) for atotal of 20 hours, then they were washed and incubated in freshgrowth media for 48 hours. IL-6 and IL-8 proteins secreted intothese conditioned media were measured by enzyme-linkedimmunosorbent assay. Data are mean�SEM; n�4. *P�0.05 vsLPS alone and vs the respective fatty acid alone.

Schwartz et al Saturated Fats Amplify Monocyte Inflammation 803



prolonged and detectable increase in LPS-stimulated IL-6 andIL-8 protein secretion.

To elucidate the underlying mechanism by which SFAamplifies monocyte inflammation, we evaluated and ex-cluded several possibilities, including upregulation of TLRexpression or increased binding affinity of Toll-like recep-tors, or differentiation of monocytes to macrophages (Sup-plementary Results and Supplemental Figures IV, V). How-ever, experiments demonstrated that at least 8 to 16 hours ofexposure to THP-1 monocytes to SFA was required beforethe treatment amplified responses to LPS (data not shown).To test whether this time lag resulted from metabolic pro-cessing of PA by the monocytes, we examined the require-ment for incorporation of PA into cells. The fatty acyl CoAsynthetase inhibitor Triacsin-C significantly reduced the ef-fect of PA on the LPS response (Figure 2A). We nextassessed whether conversion of PA to its metabolic productceramide, an important proinflammatory molecule implicatedin the pathogenesis of insulin resistance,26,27 played a role inPA-mediated amplification of inflammation. As shown inFigure 2B, exposure of THP-1 monocytes to 100 �mol/L PAfor 16 hours significantly increased cellular C16 ceramide(the direct metabolic product of PA and the predominant

ceramide in the monocytes). C18 ceramide content alsoincreased significantly, although the increase was smaller.Furthermore, Triacsin-C also blocked PA-induced increasesin C16 and C18 ceramide (Figure 2B), consistent with itsinhibition of the amplified LPS response. Exposure of THP-1monocytes to 50 �mol/L C2 ceramide (a cell-permeableanalogue of PA-derived C16 ceramide) similarly amplifiedLPS responses (Figure 2C). In contrast, the cell-impermeableC2 ceramide precursor C2 dihydroceramide did not amplifysubsequent LPS responses. Inhibitors of different enzymaticsteps in the conversion of PA to C16 ceramide (specifically,the serine palmitoyltransferase inhibitor Myriocin; the keto-sphinganine synthetase inhibitor L-cycloserine; and the cer-amide synthase inhibitor Fumonisin-B1) also effectively in-hibited the ability of PA to enhance LPS-induced IL-6 andIL-8 expression (Figure 2A). Importantly, using C2 ceramide,but not C2 dihydroceramide, to bypass Triacsin-C restoredthe amplified responses to LPS (Figure 2D), demonstratingthat Triacsin-C was not inducing nonspecific suppression ofinflammation. Essentially identical results were observedwhen C2 ceramide was also used to bypass the otherinhibitors of ceramide formation (data not shown). These dataprovide considerable evidence that generation of ceramide

Figure 2. Amplification of LPS response by PA is ceramide-dependent. A, Monocytes were exposed to media alone (C), to media containing100 �mol/L PA, to 1 ng/mL LPS, or to 100 �mol/L PA followed by LPS (PA LPS) in the absence or presence of inhibitors of de novo cer-amide synthesis (100 �mol/L Triacsin C [TriC]; 1 �mol/L Myriocin [Myr]; 250 �mol/L L-cycloserine [LCS]; and 10 �mol/L Fumonisin B1 [FB1]).IL-6 (white bars) and IL-8 (black bars) mRNA was measured by real-time polymerase chain reaction. B, Cells were exposed to media or PA inthe absence (PA) or presence (PA Triascin-C) for 16 hours. C16 (black bars) and C18 (white bars) ceramide concentrations were measured bymass spectrometry. C, Cells were exposed to media, LPS, C2 ceramide (C2), C2 dihydroceramide (diC2), or C2 or diC2 followed by LPS (C2LPS or diC2 LPS). IL-6 and IL-8 mRNA was subsequently measured by real-time polymerase chain reaction. D, To determine whether cer-amide could restore the amplified inflammation when ceramide synthesis was blocked using pharmacological inhibitors, C2 ceramide wasalso added to cells that had been treated with Triacsin-C. Data are mean�SEM; n�3–4. *P�0.05 vs LPS alone and vs PA or C2 ceramidealone (A and C), vs control (B), or vs Triacsin-C alone (D). †P�0.05 vs PA�LPS. ‡P�0.05 vs PA.

804 Arterioscler Thromb Vasc Biol April 2010



from SFA is required for SFA-induced amplification of TLRsignaling, although they do not rule out contributions fromother SFA metabolites.

Ceramides activate members of the protein kinase C (PKC)family.28 In THP-1 monocytes exposed to PA, both conventionalPKC and PKC-� were activated (as measured by translocation tothe cell membrane and phosphorylation on Ser411, respectively;Figure 3A). Similar PKC activity was observed when cells wereexposed to C2 ceramide but not to C2 dihydroceramide. Impor-tantly, the activation of conventional PKC and PKC-� wasinhibited by Triacsin-C. Interestingly, the pan-conventional/novel PKC inhibitor Calphostin-C did not inhibit amplifica-tion of the LPS response, and the potent pan-conventional/novel PKC agonist PMA did not induce any synergisticamplification of THP-1 responses to LPS (Figure 3B). Incontrast, exposure of monocytes to PA in the presence of aninhibitory PKC-� pseudosubstrate peptide29 inhibited theability of PA to amplify monocyte proinflammatory geneexpression in response to LPS, demonstrating the requirementfor PKC-� activity in this amplification mechanism.

Because PKC can serve as a potent mediator of cellsignaling events via mitogen-activated protein kinase(MAPK) pathways,30 we determined whether SFA activatedMAPK signaling via ceramide-dependent and PKC-�–depen-dent pathways. Flow cytometry studies using activationstate-specific antibodies to MAPK demonstrated that expo-

sure of monocytes to PA induced activity of Erk1/2, JNK, andp38 (Figure 4A). Similarly, C2 ceramide also activated thesesame MAPK (Figure 4A, insets). We confirmed these flowcytometry results of PA-induced MAPK activation by West-ern blot techniques (Figure 4B). Importantly, the low con-centrations of LPS (1 ng/mL) used in these experiments didnot induce detectable JNK or p38 activity in the absence ofPA, although some activation of Erk1/2 was observed inresponse to LPS. In addition, Western blots showed activa-tion of JNK1 but not JNK2 in response to PA. Furtherexperiments demonstrated that phosphorylation of Erk1/2,JNK, and p38 could be suppressed by PKC-� inhibitor(Figure 4C) or by inhibitors of the metabolism of PA toceramide (Supplemental Figure VI). We also demonstratedPA-mediated activation of several MAPK-dependent tran-scription factors, including the AP-1 transcription factorsc-Jun (Figure 4B), JunD, and c-Fos (not shown), andC/EBP-� and C/EBP-� (Figure 4C). PA-induced C/EBPactivity was suppressed by the Erk1/2 inhibitor PD98059, theJNK inhibitor SP600125, or the p38 inhibitor SB202190,demonstrating this C/EBP activity is downstream of theMAPK. Importantly, the low concentration of LPS used inthe present study did not induce phosphorylation of AP-1 orC/EBP transcription factors, in agreement with our findingsthat 1 ng/mL LPS had minimal ability to activate MAPK.

To confirm the necessity for MAPK activation in PA-induced amplification of TLR-mediated inflammation, expo-sure of monocytes to PD98059, SP600125, or SB202190significantly reduced the ability of PA to amplify LPS-induced cytokine expression (Figure 4D). The combination ofthese MAPK inhibitors (each at a reduced concentration toavoid toxicity in combination) further reduced the ability ofPA to amplify monocyte responses to LPS. Similar results werealso obtained in freshly isolated primary human monocytesusing SB202190 and SP600125 (Supplemental Figure IIB),providing evidence that this same pathway may be relevant invivo. Taken together, these results provide evidence that MAPKsignaling is required for the SFA-amplified inflammation andsuggest that ceramide derived from the metabolism of SFA leadsto PKC-mediated activation of the MAPK.

DiscussionWork from several laboratories,14,31 including our own,22 hasdemonstrated that SFA induces monocyte/macrophage in-flammation, which appears mediated in part by TLR4-signaling. The present study now demonstrates the novelfinding that exposure of monocytes to elevated concentra-tions of SFA greatly amplifies the proinflammatory responsesof these cells to TLR ligands though a separate TLR4-independent pathway, dependent on the uptake and metabolicprocessing of SFA into ceramide.

Because elucidation of this pathway relied in part on theuse of 4 pharmacological inhibitors of different enzymaticsteps in the conversion of PA to ceramide, and because allinhibited amplification of the inflammatory response to LPS,it is highly unlikely that these results could be attributed tononspecific inhibitor effects. Importantly, C2 ceramide couldbypass the pharmacological inhibitors of SFA metabolismand augment the inflammatory action of LPS, further dem-

Figure 3. Activation of PKC-� is required for PA-induced amplifi-cation of LPS responses in monocytes. A, Monocytes wereexposed to media (C), 100 �mol/L PA (PA), PA plus 100 �mol/LTriacsin-C (PA TriC), 50 �mol/L C2 dihydroceramide (dC2), or50 �mol/L C2 ceramide (C2). Markers of the activation of con-ventional PKC (translocation to the plasma membrane) andPKC-� (phosphorylation) were measured by Western blot. B,Monocytes were exposed to PA, LPS, or PA and LPS, in thepresence or absence of PKC-� inhibitor (15 �mol/L; z),Calphostin-C (250 nM; CalC), or phorbol myristate acetate (100ng/mL; PMA). IL-6 (white bars) and IL-8 (black bars) mRNA weremeasured by real-time polymerase chain reaction. Data aremean�SEM; n�3–4. *P�0.05 vs C, vs PA alone, or vs LPSalone. †P�0.05 vs PA�LPS. Blots are representative of 3 inde-pendent experiments.




onstrating the specificity of the inhibitor effects and theimportance of ceramide in this pathway. We also demon-strated activation of PKC-� and MAPK signaling pathwayswhen monocytes were exposed to PA, and we showed thatinhibitors of ceramide synthesis also inhibited these path-ways. Moreover, direct inhibition of PKC-� and MAPK alsoattenuated the amplified inflammation. These data providestrong support for our hypothesis that SFA-enhanced inflam-mation is mediated by the conversion of SFA to ceramide andsubsequent ceramide-mediated activation of PKC-mediatedand MAPK-mediated signaling, and it is therefore unrelatedto any ability of SFA to act as TLR2 or TLR4 ligands.14,15,31

Experiments also demonstrated that PKC-� inhibitor simul-taneously inhibited PA-induced MAPK activity, providingstrong evidence that MAPK lie downstream of PKC-�.Although our data also showed activation of conventional PKCby PA and C2 ceramide, blockade of conventional PKC iso-forms with Calphostin-C did not inhibit the ability of PA toamplify monocyte responses to LPS. Activation of conventionalPKC using PMA also did not induce a synergistic amplificationof the monocyte response to LPS. These data suggest conven-tional PKC does not contribute to the enhanced LPS response,but they do not rule out other roles for these enzymes.

As further evidence of MAPK involvement, we also con-firmed that inhibitors of MAPK activity simultaneously reducedthe SFA-stimulated activity of MAPK-dependent AP-1 andC/EBP transcription factors and the amplification of the LPSresponse. Additionally, these data may also provide insight intohow MAPK activation could enhance LPS signaling. Severalinflammatory cytokines, including IL-6 and IL-8, have promoterbinding sites for members of the AP-1 and the C/EBP families,both of which have been reported to interact cooperatively withNF-�B to upregulate gene expression.32,33 We speculate that theobserved MAPK-mediated amplification of cytokine gene ex-pression therefore may be affected via these same transcriptionfactors enhancing the transcriptional effectiveness of LPS/TLR4-induced NF-�B activity.

This has several important implications because NF-�B iswidely recognized as a “master switch” of inflammation.8

Therefore, we speculated that SFA may enhance transcriptionof a wide variety of inflammatory genes in a wide variety ofcell types. Exposure of monocytes to PA amplified LPS-induced expression of several genes, including IL-1�, IP-10,MCP-1, resistin, and tumor necrosis factor-�, but not COX-2(Supplemental Figure VII), suggesting that this response is

Figure 4. PA-dependent activation of MAPK sig-naling is required for amplification of LPSresponses. A and B, Monocytes were exposed tomedia alone (C and white histograms), 100 �mol/Lpalmitic acid (PA and black histograms), or 1ng/mL LPS. (Insets); monocytes were alsoexposed to media containing DMSO vehicle alone[white histograms]or media containing 50 �mol/LC2 ceramide [black histograms]). Phosphorylationof p38, JNK1/2, and Erk 1/2 was measured byflow cytometry. B, Western blot also confirmedthat exposure of monocytes to PA activatedErk1/2, JNK1, p38, and c-Jun in the presence andabsence of LPS. Activation of the JNK-dependenttranscription factor c-Jun was also measured.Beta-actin served as a loading control. C, Mono-cytes were exposed to control media (C), PA, orPA plus PKC-� inhibitor (PAz), and phosphorylationof Erk1/2 (white bars), JNK (black bars), and p38(striped bars) were measured by flow cytometry.D, Cells were exposed to PA and LPS plus inhibi-tors of Erk1/2 (20 �mol/L PD98059; PD), JNK(20 �mol/L SP600125; SP), and p38 (20 �mol/LSB202190; SB). C/EBP-� (white bars) andC/EBP-� (black bars) DNA binding activity wasmeasured by DNA-binding enzyme-linked immu-nosorbent assay. E, Real-time polymerase chainreaction was also used to measure IL-6 (whitebars) and IL-8 (black bars) mRNA in cells exposedto the individual MAPK inhibitors and to a mixtureof the 3 MAPK inhibitors (2 �mol/L each; Mix).Data are mean�SEM; n�3–4. *P�0.05 vs C, vsLPS alone, and vs PA alone. †P�0.05 vsPA�LPS. ‡P�0.05 vs PA. Blots and histogramsare representative from n�3 independentexperiments.




widespread among NF-�B–mediated proinflammatory genes.This also raises the possibility that SFA may prime cells toamplify responses not only to LPS but also to a variety of otherproinflammatory stimuli that activate NF-�B. Consistent withthis concept, we have also shown that exposure of monocytes toPA also amplifies their responses to the TLR2 ligandPam3CSK4. Importantly, whereas siRNA-mediated knockdownof TLR4 does inhibit PA-amplified responses to LPS (theexpected result because cells lacking TLR4 cannot respond toLPS), there is no inhibition of the ability of PA to amplifymonocyte responses to the TLR2 ligand Pam3CSK4. This resultclearly shows that PA-mediated amplified inflammation is notdependent on a specific TLR but must operate through a separateand more general mechanism. We illustrate in schematic form(Supplemental Figure VIII) the pathway that we believe our datasuggest underlies this amplification.

The synergy between LPS and FFA was specific for SFA(PA, lauric acid, and stearic acid) and was not found whenTHP-1 monocytes were exposed to unsaturated FFA. Inagreement with earlier reports,15,31 linoleic acid mildly sup-pressed the responses of monocytes to TLR ligands. Oleicacid neither amplified nor inhibited the responses of mono-cytes to LPS. This may be because of the preferentialconversion of SFA to ceramide34 or to the ability of unsatur-ated FFA to inhibit the underlying LPS-induced inflamma-tion.14,31 However, the amplification effect was maintained forTHP-1 cells and primary human monocytes when SFA weremixed with unsaturated FFA in physiological concentrations.These data are consistent with previous reports that SFA, but notunsaturated FFA, were proinflammatory.14–16,22,31 Although aprevious report indicated that lauric acid was not directlycapable of activating TLR-mediated signaling in mouseRAW296.7 macrophages,16 this finding actually supports ourinterpretation that SFA-enhanced inflammation occurs via anovel mechanism independent of direct SFA/TLR interactions.Importantly, the present results were obtained in human cellswhile using fatty acids and LPS at plasma levels commonlyfound in individuals with obesity or type 2 diabetes, or afterconsumption of high-fat meals,35–39 thereby providing evidencethat the ability of SFA to amplify TLR signaling may haveimportant physiological implications.

Consistent with this, our data also show that PA enhancesthe LPS response of macrophages as well as monocytes.Macrophage accumulation in tissue and production of proin-flammatory cytokines are believed critical in development ofinsulin resistance, diabetes, and cardiovascular disease. Wehave previously demonstrated that cytokines released frommacrophages induced adipocyte inflammation and reducedinsulin action and lipolysis.6 The resulting increase in FFAreleased from adipocytes can further enhance the macro-phage-initiated proinflammatory cascade.

Several methodological considerations deserve mention.Although the present study is consistent with in vivo studieslinking excess nutrition and elevated plasma SFA concentra-tions to inflammatory consequences,16,36,40 our results werelargely obtained using human cell lines. For this reason, theexistence of this synergistic effect, its dependence on SFAuptake, and its dependence on MAPK activity were alsoconfirmed ex vivo in primary human monocytes. The deci-

sion to conduct these experiments with human cells and celllines had advantages and disadvantages. Using human cells(monocytes and macrophages) provided considerably moreweight to the potential relevance of this phenomenon tohuman disease. More importantly, whereas it might provefruitful to examine specific signaling steps in this amplificationpathway using mouse models in the future, initial experimentsexposing mouse RAW296.7 macrophages to PA did not result inan amplified response to LPS (data not shown). These prelimi-nary data suggested differences in responses between mouse andhuman cells, possibly reflecting well-known species differencesin the sensitivity and specificity of human and murine TLR tocertain ligands.41 However, the necessity of using human mono-cytes limited the availability of genetically modified (eg, knock-out) models and generally complicated the use of molecularmethods. For example, whereas we successfully used siRNA toknock-down TLR4, our attempts to use siRNA against serinepalmitoyltransferase consistently led to considerable cell deathand a minimally effective knockdown in the remaining cells.Consistent with our experience, Tamehiro et al42 still needed touse Myricin to achieve substantial inhibition of ceramide syn-thesis even in THP-1 cells selected (using puromycin) forsuccessful transfection and expression of lentiviral shRNAagainst this enzyme. To overcome these difficulties, we studiedthis pathway using multiple, functionally unrelated inhibitors.

In conclusion, this study presents the novel finding thatSFA can greatly amplify the inflammatory responses ofmonocytes/macrophages to LPS via TLR4-independentmechanisms. These data suggest that uptake of SFA leads toenhanced ceramide generation, which in turn activates PKC-�and MAPK. Concentrations of SFA and LPS that are com-monly present in conditions resulting from excess nutrition(eg, obesity, insulin resistance, type 2 diabetes, etc) andhigh-fat diets can lead to inflammation that is several-foldgreater than the direct proinflammatory effects of each ofthese factors alone. The elevation of SFA and LPS in thesesettings could create a “perfect storm” of factors leading togreatly increased monocyte/macrophage innate immune re-sponses. Because monocytes/macrophages are critical cellsmediating tissue and systemic inflammation, a broad and“synergistic” amplification of their inflammatory activity bySFA may be an important contributor to the genesis orprogression of insulin resistance and type 2 diabetes, as wellas the cardiovascular sequelae of these conditions.

AcknowledgmentsThe authors thank Christopher B. Heward and the staff at KronosScience Laboratory for providing mass spectrometry equipment andexpertise.

Sources of FundingThis work was supported in part by the Department of VeteransAffairs, Biomedical Laboratory Research and Development, by aVISN 18 New Investigators Pilot grant (E.A.S.), and by a VA MeritReview grant (P.D.R., E.A.S.).

DisclosuresY.S. is employed by Kronos Science Laboratory. This work does notrepresent the views of the Department of Veterans Affairs or theUnited States Government.




References1. Barzilay J, Freedland E. Inflammation and its association with glucose

disorders and cardiovascular disease. Treat Endocrinol. 2003;2:85–94.2. Pickup JC, Crook MA. Is type II diabetes mellitus a disease of the innate

immune system? Diabetologia. 1998;41:1241–1248.3. Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive

protein, interleukin 6, and risk of developing type 2 diabetes mellitus.JAMA. 2001;286:327–334.

4. Navab M, Fogelman AM, Berliner JA, Territo MC, Demer LL, Frank JS,Watson AD, Edwards PA, Lusis AJ Pathogenesis of atherosclerosis. Am JCardiol. 1995;76:18C–23C.

5. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, FerranteAW Jr. Obesity is associated with macrophage accumulation in adiposetissue. J Clin Invest. 2003;112:1796–1808.

6. Permana PA, Menge C, Reaven PD. Macrophage-secreted factors induceadipocyte inflammation and insulin resistance. Biochem Biophys ResCommun. 2006;341:507–514.

7. Solinas G, Vilcu C, Neels JG, Bandyopadhyay GK, Luo JL, Naugler W,Grivennikov S, Wynshaw-Boris A, Scadeng M, Olefsky JM, Karin M.JNK1 in hematopoietically derived cells contributes to diet-inducedinflammation and insulin resistance without affecting obesity. CellMetab. 2007;6:386–397.

8. Schwartz EA, Reaven PD. Molecular and signaling mechanisms of ath-erosclerosis in insulin resistance. Endocrinol Metab Clin North Am.2006;35:525–549, viii.

9. Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D, NeyrinckAM, Fava F, Tuohy KM, Chabo C, Waget A, Delmee E, Cousin B,Sulpice T, Chamontin B, Ferrieres J, Tanti JF, Gibson GR, Casteilla L,Delzenne NM, Alessi MC, Burcelin R. Metabolic endotoxemia initiatesobesity and insulin resistance. Diabetes. 2007;56:1761–1772.

10. Cani PD, Rodrigo B, Knauf C, Waget A, Neyrinck AM, Delzenne NM,Burcelin R. Changes in gut microbiota control metabolic endotoxemia-induced inflammation in high-fat diet-induced obesity and diabetes inmice. Diabetes. 2008;57:1470–1481.

11. Mullick AE, Tobias PS, Curtiss LK. Modulation of atherosclerosis inmice by Toll-like receptor 2. J Clin Invest. 2005;115:3149–3156.

12. de Kleijn D, Pasterkamp G. Toll-like receptors in cardiovascular diseases.Cardiovasc Res. 2003;60:58–67.

13. Michelsen KS, Wong MH, Shah PK, Zhang W, Yano J, Doherty TM,Akira S, Rajavashisth TB, Arditi M. Lack of Toll-like receptor 4 ormyeloid differentiation factor 88 reduces atherosclerosis and alters plaquephenotype in mice deficient in apolipoprotein E. Proc Natl Acad SciU S A. 2004;101:10679–10684.

14. Lee JY, Sohn KH, Rhee SH, Hwang D. Saturated fatty acids, but notunsaturated fatty acids, induce the expression of cyclooxygenase-2 mediatedthrough Toll-like receptor 4. J Biol Chem. 2001;276:16683–16689.

15. Lee JY, Zhao L, Youn HS, Weatherill AR, Tapping R, Feng L, Lee WH,Fitzgerald KA, Hwang DH. Saturated fatty acid activates but polyunsat-urated fatty acid inhibits Toll-like receptor 2 dimerized with Toll-likereceptor 6 or 1. J Biol Chem. 2004;279:16971–16979.

16. Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 linksinnate immunity and fatty acid-induced insulin resistance. J Clin Invest.2006;116:3015–3025.

17. Arkan MC, Hevener AL, Greten FR, Maeda S, Li ZW, Long JM,Wynshaw-Boris A, Poli G, Olefsky J, Karin M. IKK-beta links inflam-mation to obesity-induced insulin resistance. Nat Med. 2005;11:191–198.

18. Solinas G, Naugler W, Galimi F, Lee MS, Karin M. Saturated fatty acidsinhibit induction of insulin gene transcription by JNK-mediated phos-phorylation of insulin-receptor substrates. Proc Natl Acad Sci U S A.2006;103:16454–16459.

19. Hunnicutt JW, Hardy RW, Williford J, McDonald JM. Saturated fattyacid-induced insulin resistance in rat adipocytes. Diabetes. 1994;43:540–545.

20. Perseghin G, Petersen K, Shulman GI. Cellular mechanism of insulinresistance: potential links with inflammation. Int J Obes Relat MetabDisord. 2003;27(Suppl 3):S6–S11.

21. Posner BM, Cobb JL, Belanger AJ, Cupples LA, D’Agostino RB, StokesJ III. Dietary lipid predictors of coronary heart disease in men. TheFramingham Study. Arch Intern Med. 1991;151:1181–1187.

22. Laine PS, Schwartz EA, Wang Y, Zhang WY, Karnik SK, Musi N, ReavenPD. Palmitic acid induces IP-10 expression in human macrophages viaNF-kappaB activation. Biochem Biophys Res Commun. 2007;358:150–155.

23. Zhang WY, Schwartz E, Wang Y, Attrep J, Li Z, Reaven P. Elevatedconcentrations of nonesterified fatty acids increase monocyte expression

of CD11b and adhesion to endothelial cells. Arterioscler Thromb VascBiol. 2006;26:514–519.

24. Wong SW, Kwon MJ, Choi AM, Kim HP, Nakahira K, Hwang D. Fattyacids modulate toll-like receptor 4 activation through regulation ofreceptor dimerization and recruitment into lipid rafts in a ros-dependentmanner. J Biol Chem. 2009.

25. Erridge C, Samani NJ. Saturated fatty acids do not directly stimulateToll-like receptor signaling. Arterioscler Thromb Vasc Biol. 2009;29:1944–1949.

26. Haus JM, Kashyap SR, Kasumov T, Zhang R, Kelly KR, Defronzo RA,Kirwan JP. Plasma ceramides are elevated in obese subjects with type 2diabetes and correlate with the severity of insulin resistance. Diabetes.2009;58:337–343.

27. Schmitz-Peiffer C, Craig DL, Biden TJ. Ceramide generation is sufficientto account for the inhibition of the insulin-stimulated PKB pathway inC2C12 skeletal muscle cells pretreated with palmitate. J Biol Chem.1999;274:24202–24210.

28. Lozano J, Berra E, Municio MM, Diaz-Meco MT, Dominguez I, Sanz L,Moscat J. Protein kinase C zeta isoform is critical for kappa B-dependentpromoter activation by sphingomyelinase. J Biol Chem. 1994;269:19200–19202.

29. Fishelevich R, Malanina A, Luzina I, Atamas S, Smyth MJ, Porcelli SA,Gaspari AA. Ceramide-dependent regulation of human epidermal kerati-nocyte CD1d expression during terminal differentiation. J Immunol.2006;176:2590–2599.

30. Huwiler A, Brunner J, Hummel R, Vervoordeldonk M, Stabel S, van denBosch H, Pfeilschifter J. Ceramide-binding and activation defines proteinkinase c-Raf as a ceramide-activated protein kinase. Proc Natl Acad SciU S A. 1996;93:6959–6963.

31. Lee JY, Ye J, Gao Z, Youn HS, Lee WH, Zhao L, Sizemore N, HwangDH. Reciprocal modulation of Toll-like receptor-4 signaling pathwaysinvolving MyD88 and phosphatidylinositol 3-kinase/AKT by saturatedand polyunsaturated fatty acids. J Biol Chem. 2003;278:37041–37051.

32. Grassl C, Luckow B, Schlondorff D, Dendorfer U. Transcriptional reg-ulation of the interleukin-6 gene in mesangial cells. J Am Soc Nephrol.1999;10:1466–1477.

33. Hoffmann E, Dittrich-Breiholz O, Holtmann H, Kracht M. Multiple control ofinterleukin-8 gene expression. J Leukoc Biol. 2002;72:847–855.

34. Okere IC, Chandler MP, McElfresh TA, Rennison JH, Sharov V, SabbahHN, Tserng KY, Hoit BD, Ernsberger P, Young ME, Stanley WC.Differential effects of saturated and unsaturated fatty acid diets on car-diomyocyte apoptosis, adipose distribution, and serum leptin. Am JPhysiol Heart Circ Physiol. 2006;291:H38–H44.

35. Ma J, Folsom AR, Shahar E, Eckfeldt JH. Plasma fatty acid compositionas an indicator of habitual dietary fat intake in middle-aged adults. TheAtherosclerosis Risk in Communities (ARIC) Study Investigators. Am JClin Nutr. 1995;62:564–571.

36. Wang L, Folsom AR, Zheng ZJ, Pankow JS, Eckfeldt JH. Plasma fattyacid composition and incidence of diabetes in middle-aged adults: theAtherosclerosis Risk in Communities (ARIC) Study. Am J Clin Nutr.2003;78:91–98.

37. Chung BH, Hennig B, Cho BH, Darnell BE. Effect of the fat composition ofa single meal on the composition and cytotoxic potencies of lipolytically-releasable free fatty acids in postprandial plasma. Atherosclerosis. 1998;141:321–332.

38. Creely SJ, McTernan PG, Kusminski CM, Fisher M, Da Silva NF, KhanolkarM, Evans M, Harte AL, Kumar S. Lipopolysaccharide activates an innateimmune system response in human adipose tissue in obesity and type 2diabetes. Am J Physiol Endocrinol Metab. 2007;292:E740–E747.

39. Erridge C, Attina T, Spickett CM, Webb DJ. A high-fat meal induceslow-grade endotoxemia: evidence of a novel mechanism of postprandialinflammation. Am J Clin Nutr. 2007;86:1286–1292.

40. Nicholls SJ, Lundman P, Harmer JA, Cutri B, Griffiths KA, Rye KA,Barter PJ, Celermajer DS. Consumption of saturated fat impairs theanti-inflammatory properties of high-density lipoproteins and endothelialfunction. J Am Coll Cardiol. 2006;48:715–720.

41. Nahori MA, Fournie-Amazouz E, Que-Gewirth NS, Balloy V, ChignardM, Raetz CR, Saint Girons I, Werts C. Differential TLR recognition ofleptospiral lipid A and lipopolysaccharide in murine and human cells.J Immunol. 2005;175:6022–6031.

42. Tamehiro N, Mujawar Z, Zhou S, Zhuang DZ, Hornemann T, vonEckardstein A, Fitzgerald ML. Cell polarity factor Par3 binds SPTLC1and modulates monocyte serine palmitoyltransferase activity and che-motaxis. J Biol Chem. 2009;284:24881–24890.




SUPPLEMENT

SUPPLEMENTARY METHODS

Materials and Reagents

THP-1 cells were purchased from ATCC (Gaithersburg, MD). Cell culture media and

supplements were from Invitrogen (Gaithersburg, MD). C2 ceramide and

dihydroceramide were from Biomol (Plymouth Meeting, MA). Ceramide standards for

mass spectrometry were from Avanti Polar Lipids (Alabaster, AL). PKC-

psuedosubstrate inhibitor was from Tocris Bioscience (Ellisville, MO). PD98059, and

SB202190 were from VWR (South Plainfield, NJ). ELISAs were from R&D Systems

(Minneapolis, MO). DNA-binding ELISAs were from Active Motif (Carlsbad, CA).

Flow cytometry reagents were from Beckman-Coulter (Fullerton, CA). Real-time PCR

reagents were from Bio-Rad (Hercules, CA). Antibodies to phosphorylated and total

proteins were from Cell Signaling Technology (Danvers, MA). Antibody to TLR4 (clone

HTA125) was from eBioscience (San Diego, CA). Specific monoclonal antibody to

activated NF-B (clone 12H11) was from Millipore (Billerica, MA). Specific

monoclonal antibody to CCR2 (clone TG5/CCR2) was from Biolegend (San Diego, CA)

All other reagents not otherwise specified were from were from Sigma-Aldrich (St.

Louis, MO).

Cell culture and experimental design

THP-1 monocytes (passage 5-15) were routinely cultured in growth media (RPMI-1640

media supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 1 mM sodium

Schwartz, Saturated Fats Amplify Monocyte Inflammation

S.1 at UNIV OF CALIFORNIA LIB on April 13, 2010 atvb.ahajournals.orgDownloaded from


pyruvate, 100 mM HEPES, and 20 M -mercaptoethanol) under standard cell culture

conditions (37 oC, humidified, 95% air / 5% CO2), and reseeded every 2-3 days to

maintain a density of 250,000 to 1,000,000 cells/mL. All reagents were frequently tested

for the presence of LPS using a high-sensitivity Limulus amoebocyte assay kit (Lonza,

Walkersville, MD).

For experiments on inflammation responses, cells were reseeded in growth media with or

without FFA, LPS, specific pharmacologic inhibitors, and/or combinations of these

reagents. For all reagents not part of the routine culture of THP-1 cells, we performed

Trypan Blue exclusion and lactose dehydrogenase toxicity assays (Sigma) in the presence

of PA, and reagent concentrations used in this manuscript were not associated with

observable cell toxicity (data not shown). Cells were typically incubated with FFA

and/or inhibitors for 16 hours, washed in PBS, and resuspended in fresh growth media

with or without 1 ng/mL E. coli LPS for 30 minutes (for phosphoprotein assays), 4 hours

(for real-time PCR), or 48 hours (for conditioned media ELISA). In other experiments,

the TLR2 ligand Pam3CSK4 (5 ng/mL; Invivogen, San Diego, CA) was substituted for

LPS.

To produce THP-1 macrophages, aliquots of 1,000,000 monocytes were seeded in 2 mL

of growth media supplemented with 100 ng/mL phorbol ester for 24 hours, then

subsequently washed to remove non-adherent cells. Adherent cells were allowed to

mature into macrophages for 7 days prior to experiments, and were fed with 2 mL growth

media every 2 days.




Human monocytes were prepared from blood donated as part of a general laboratory

specimen collection protocol previously approved by the IRB of the Phoenix VA Health

Care System. Monocytes were isolated using dual Histopaque-1077 and Percoll density

gradient centrifugation, as previously described 1. Human monocytes were maintained in

modified growth media containing 10% human type AB serum (Sigma) in place of fetal

bovine serum.

Transfection

THP-1 monocytes were transfected using the Amaxa Nucleofector II and Nucleofector

Kit V (Lonza, Walkersville, MD), and following manufacturer’s protocols for THP-1

monocytes and optimized for our genes of interest. Briefly, THP-1 cells were pelleted by

centrifugation at 1000 RPM for 10 minutes and all media was aspirated from the pellet.

The cells were subsequently resuspended by gentle flicking to a density of 10,000,000

cells/mL in Nucleofector V transfection solution, and aliquots were mixed with siRNA to

TLR4 (5 g/mL; Dharmacon, Lafayette, CO), or SPTLC1 (5-10 g/mL; constructs from

both Dharmacon, and SA Biosciences, Frederick, MD), or with non-targeting control

siRNA (Dharmacon). Aliquots of 1,000,000 cells were eletroporated using Nucleofector

program V-001, and were immediately transferred to 500 mL cell growth media. All

aliquots were transfected within 15 minutes of cell resuspension, and allowed to recover

for 20 minutes. Cells were then transferred to 2 mL of growth media and incubated for 3

days to deplete internal stores of targeted protein. Cells transfected with the same siRNA

were pooled to ensure uniformity of subsequent aliquots before being used in




experiments.

Fatty acid preparation

Because we had previously found many common laboratory reagents were contaminated

with LPS, FFA were mixed to a final concentration of 40 mM in a solvent of 0.1 N

NaOH / 70% ethanol, and heated to 70 oC until dissolved. This alkaline buffer

hydrolyzes and detoxifies > 10 ng/mL LPS 2 and > 2.5 ng/mL of the lipopeptide

analogues FSL and Pam3CSK4 (unpublished data). In addition, because we had

previously detected ubiquitous endotoxin contamination of even low-endotoxin

preparations of bovine serum albumin, we added our FFA directly to the cell culture

media, where they could complex with the bovine serum albumin already present in the

10% fetal bovine serum. Except as otherwise noted, individual FFA were added at a final

concentration of 100 M. The mixture of FFA used in some confirmatory experiments

was 30% palmitic acid, 10% stearic acid, 32% oleic acid, 20% linoleic acid, 1%

eicosapentaenoic acid, and 1% docosahexaenoic acid (molar %), and was used at 200 M

final concentration. The final FFA:albumin ratio used in experiments was approximately

2.3:1.

Real-time PCR

RNA was extracted from cells using the Aurum mini column method (Bio-Rad

Laboratories, Hercules, CA) and following manufacturer’s instructions. Reverse

transcriptase PCR was performed using Bio-Rad iScript reagent. Real-time, semi-

quantitative PCR was performed on a Bio-Rad iCycler iQ instrument, using SYBR Green




chemistry (Bio-Rad) and primers as previously described 3.

Ceramide measurements

Aliquots of 1x106 monocytes were collected by centrifugation and washed twice in PBS

to remove extracellular lipids. Cell pellets were spiked with an internal standard of 25 ng

deuterated C16-d31 ceramide. Total polar lipids were extracted from each cell pellet

using chloroform:methanol (2:1) and supernatants were dried under nitrogen. Dried

lipids were resuspended in water/methanol mobile phase and analyzed using LC/MS-MS

on an API2000 triple stage quadrupole mass spectrometer (Applied Biosystems, Foster

City, CA) with the positive electrospray ionization method using the multiple reaction

monitoring mode (ion pairs of 538.4/264.4, 566.4/164.4, and 569.3/265.4 were used to

monitor C16, C18 and C16-d31, respectively). Concentrations of these ceramides were

quantified by comparison of the ratio of the mass chromatogram peak areas of each

ceramide vs. C16-d31 with a 7-point external standard calibration curve for each

ceramide, using Analyst 1.41 software.

ELISA

DNA binding activity of AP-1 and C/EBP- and – was measured using DNA-binding

ELISA kits (Active Motif, Carlsbad, CA) following manufacturer’s instructions, using a

VersaMax plate reader (Molecular Dynamics, Sunnyvale, CA). IL-6 and IL-8 proteins

were similarly measured by ELISA (R&D Systems; Minneapolis, MN).

Western blots




Activity of mitogen-activated protein kinases (MAPK) was measured using Western

blots, following standard laboratory methods using NuPAGE reagents (Invitrogen).

Phosphoprotein bands were visualized using primary antibodies against phosphorylated

Jun N-terminal kinase-1 and -2 (JNK1/2), p38 MAPK, and extracellular-regulated kinase-

1 and -2 (Erk1/2), as well as phosphorylated c-Jun, and -actin (Sigma).

Flow Cytometry

To measure cell surface expression of TLR4 and CD14, THP-1 monocytes were fixed in

4% paraformaldehyde in PBS for 30 minutes. Cells were subsequently washed in PBS,

then blocked in PBS containing 2% fetal bovine serum for 10 minutes. CCL2, TLR4 and

CD14 were labeled using specific monoclonal antibodies (clones TG5/CCL2, HTA125,

and 61D3, respectively), followed by biotinylated secondary antisera (eBioscience).

Cells were stained using allophycocyanin-labeled streptavidin (SouthernBiotech,

Birmingham, AL). TLR4 binding activity was measured by exposing unfixed monocytes

to 1, 10, and 100 ng/mL fluorescein isothiocyanate-labeled LPS (Sigma) for 30 minutes;

preliminary experiments showed no concentration-related difference in LPS binding, and

data are presented using 10 ng/mL. MAPK activity was also assayed using fluorescently-

labeled, activation-state-specific antibodies to p38, JNK, or Erk (Beckman-Coulter,

Fullerton, CA), following manufacturer’s instructions. NF-B activity was measured

using a specific monoclonal antibody that recognizes the exposed IB-binding site on the

activated NF-B p65 subunit (clone 12H11). Staining was as for TLR4, except that cells

were fixed in 3.7% formaldehyde for 10 minutes, then permeabilized in 90% methanol

for 30 minutes prior to the blocking step. Fluorescent labeling of TLR4 and CD14




surface expression, fluorescein isothiocyante-labeled LPS binding, as well as MAPK and

NF-B activity was measured on an FC500 flow cytometer (Beckman-Coulter).

Statistics

Except as otherwise noted, experiments were repeated at least in triplicate. Results

consisting of raw numerical data (e.g., DNA-binding ELISA) were analyzed by ANOVA

followed by post-hoc Tukey tests. Results consisting of samples normalized to same-

experiment controls (e.g., real-time PCR) were analyzed using Z-statistics with

Bonferroni’s correction. Data were presented as mean ± SEM; significance was taken as

p ≤ 0.05.

SUPPLEMENTARY RESULTS

Our central results, namely that the saturated fatty acid PA amplifies TLR signaling in a

human monocyte cell line, were confirmed and expanded upon with other saturated fatty

acids and different cells of monocytic lineage. First, lauric acid and stearic acid induced

effects similar to those caused by PA (Fig. 1). In contrast, exposure of cells to the

unsaturated fatty acids linoleic acid or oleic acid either suppressed or failed to enhance

(respectively) IL-6 and -8 mRNA responses to LPS (Suppl. Fig. I). It is important to note

that amplification of the response to LPS was also seen when the monocytes were

incubated with an elevated concentration of a mixture of FFA approximating the

distribution of saturated and unsaturated FFA in human plasma (Suppl. Fig. IIA).




Confirmatory experiments performed with primary human monocytes pre-treated with

this FFA mixture also showed ~4.5-fold amplification of LPS-induced IL-6 and a nearly

3-fold increase in IL-8 expression (Suppl. Fig. IIB) compared to LPS alone. Importantly,

Triacsin-C inhibited the ability of physiological mixtures of FFA to amplify LPS-induced

inflammation in primary human monocytes ex vivo (Suppl. Fig. II B), suggesting that the

same ceramide-dependent mechanism demonstrated in THP-1 cells might also be

relevant in vivo. Similarly, exposure of THP-1 macrophages (differentiated from THP-1

monocytes using 100 M phorbol ester) to PA induced a reproducible increase in LPS-

stimulated IL-6 and IL-8 mRNA levels, although this required slightly longer incubation

(24 hours) with PA (Suppl. Fig. IIC).

Importantly, siRNA to TLR4 inhibited the response of the monocytes to LPS, but did not

alter either the response of the cells to Pam3CSK4 or the amplification of that response

by PA (Suppl. Fig. III). These supplementary data demonstrate that PA amplification of

TLR signaling does not depend on TLR4 specifically, and points out the distinct

possibility that SFA may amplify inflammatory signal pathways for multiple TLRs.

To elucidate the underlying mechanism(s) by which SFA amplifies monocyte TLR-

dependant inflammation, we first examined several obvious possibilities. Both Western

blot (data not shown) and flow cytometry studies showed no increase in total or cell

surface TLR4 expression (Suppl. Fig. IV A). Similar results were observed for CD14

(data not shown). Following PA treatment, LPS binding activity was actually slightly

decreased (Suppl. Fig. IV B). Furthermore, although we previously reported that

exposure of monocytes to PA alone induced NF-B activity 1, 2, PA-induced NF-B




activity was not additive with that induced by LPS (Suppl. Fig. IV C). These data

indicated that (even if SFA are themselves a weak TLR4 ligand, as suggested by some 4,

5, but not all 6 investigators) the ability of SFA to amplify LPS-induced TLR4-mediated

inflammation is not due to upregulation of TLR4 expression or activity. Monocytes

exposed to PA showed no morphological changes consistent with macrophage

differentiation, and in fact demonstrated increased gene expression (with no change in

protein) of CCR2, a receptor that characteristically decreases in expression during

differentiation into macrophages (Suppl. Fig. V) These data suggest that PA does not

increase inflammation by inducing differentiation of the monocytes to macrophages, a

conclusion further strengthened by the observation that PA markedly amplifies LPS-

induced inflammation in cells already differentiated to macrophages (Suppl. Fig. IIC).

SUPPLEMENTARY REFERENCES

1. Zhang WY, Schwartz E, Wang Y, Attrep J, Li Z, Reaven P. Elevated concentrations of nonesterified fatty acids increase monocyte expression of CD11b and adhesion to endothelial cells. Arterioscler Thromb Vasc Biol. 2006;26:514-519.

2. Laine PS, Schwartz EA, Wang Y, Zhang WY, Karnik SK, Musi N, Reaven PD. Palmitic acid induces IP-10 expression in human macrophages via NF-kappaB activation. Biochem Biophys Res Commun. 2007;358:150-155.

3. Zhang WY, Schwartz EA, Permana PA, Reaven PD. Pioglitazone Inhibits the Expression of Inflammatory Cytokines From Both Monocytes and Lymphocytes in Patients With Impaired Glucose Tolerance. Arterioscler Thromb Vasc Biol. 2008.

4. 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.

5. 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;116:3015-3025.

6. Erridge C, Samani NJ. Saturated Fatty Acids Do Not Directly Stimulate Toll-Like Receptor Signaling. Arterioscler Thromb Vasc Biol. 2009.




SUPPLEMENTARY FIGURE LEGENDS

Supplementary Figure I. Exposure of monocytes to unsaturated FFA does not

amplify subsequent responses to LPS. A. THP-1 monocytes were exposed to media

alone (C), or to 100 M of linoleic acid (LA) or oleic acid (OA) for 16 hours, followed

by exposure to 1 ng/mL LPS (LPS) for 4 hours as described in Methods. IL-6 (white

bars) and IL-8 (black bars) mRNA was measured by real-time PCR. Data are mean ±

SEM; n = 3.

Supplementary Figure II. Exposure of cultured or primary human monocytes, or

cultured human macrophages, to physiological mixtures of SFA in combination with

unsaturated FFA also show amplified responses to LPS. A. THP-1 monocytes were

exposed to media for 16 hours (C), media followed by 1 ng/mL LPS for 4 hours (LPS),

200 M non-esterified fatty acids (NEFA; fatty acid composition as described in

Methods) for 16 hours, or NEFA followed by LPS (NEFA LPS). Expression of cytokine

(IL-6 and IL-8) mRNA was measured by real-time PCR. B. Freshly-isolated primary

human monocytes were also exposed to NEFA and/or LPS similar to the THP-1 cells,

above. Additionally, primary human monocytes were exposed NEFA and LPS in the

presence of the acyl-CoA transferase inhibitor Triacsin-C (TriC), the p38 MAPK

inhibitor SB202190 (SB), and the JNK inhibitor SP600125 (SP). C. Similar experiments

were also performed exposing THP-1 macrophages (differentiated with 100 ng/mL

phorbol ester) to PA for 24 hours followed by 1 ng/mL LPS for 4 hours. Data are mean ±

SEM; n = 3 - 6; * p < 0.05 both vs. LPS alone and vs. PA or NEFA alone; † p < 0.05 vs.

NEFA LPS.




Supplementary Figure III. PA amplifies TLR2 responses, and siRNA-mediated

knockdown of TLR4 does not affect the amplification response. THP-1 monocytes

were sham-transfected with a non-targeting siRNA (black bars) or with siRNA against

TLR4 (grey bars; knockdown efficiency 60% by qPCR). Cells were subsequently

maintained under control conditions (C), or exposed to PA, the TLR4 ligand LPS (LPS),

or the TLR2 ligand Pam3CSK4 (Pam3), alone or in combination. Pre-treatment with PA

amplified monocyte responses to Pam3CSK4 (PA Pam3) as well as LPS. Knockdown of

TLR4 inhibited monocyte responses to LPS, however, monocyte responses to

Pam3CSK4, and amplification of Pam3CSk4 response by PA, were unaffected. Data are

mean ± SEM; n = 3; * p < 0.05 both vs. LPS or Pam3 alone.

Supplementary Figure IV. Exposure of monocytes to SFA did not amplify TLR4

expression, LPS binding capacity, or LPS-stimulated NF-B activity. THP-1

monocytes were exposed to media (C-black histograms), or media containing 100 M

palmitic acid (PA-white histograms) for 16 hours. A. TLR4 surface expression was

measured by flow cytometry using a monoclonal antibody to TLR4 and a

biotin/allophycocyanin-streptavidin labeling system. Similar results were also found for

CD14 surface expression (not shown). B. Aliquots of these cells were also stained with

fluorescein isothiocyanate-labeled LPS for 30 minutes. LPS binding was measured using

flow cytometry. C. Cells were exposed to media for 16 hours (C; and the control image

replicated for comparison in other frames), media followed by 1 ng/mL LPS for 30

minutes (LPS), 100 M PA for 16 hours (PA), or PA followed by LPS (PA LPS). NF-




B activity was measured by flow cytometry using a monoclonal antibody specific to

activated NF-B. Histograms and data are representative from n ≥ 3 independent

experiments counting 10,000 cells each; all histograms in C. are from the same

representative experiment.

Supplementary Figure V. Exposure of monocytes to SFA does not induce

differentiation of monocytes to macrophages. To confirm that the amplified

expression of proinflammatory genes wasn’t due to SFA-mediated transformation of the

monocytes to macrophages, THP-1 monocytes were exposed to media (C) or palmitic

acid (PA) for 16 hours. Expression of the monocyte marker CCR2 was measured by flow

cytometry (white bars) and real-time PCR (black bars). Data are mean ± SEM; n = 3.

Supplementary Figure VI. Pharmacologic antagonists of the metabolism of PA to

ceramide inhibit PA-induced MAPK phosphorylation. Monocytes were maintained in

media (C) or exposed to 100 M palmitic acid (PA) for 16 hours in the presence or

absence of the serine palmitoyltransferase inhibitor Myriocin (Myr), the ceramide

synthase inhibitor Fumonisin B1 (FB1), the ketosphinganine synthetase inhibitor L-

cycloserine (LCS), or the acyl-CoA synthetase inhibitor Triacsin-C (TriC).

Phosphorylation of the mitogen-activated protein kinases p38, JNK, and Erk1/2 was

measured by flow cytometry. Data are mean ± SEM; n = 3; * p < 0.05 vs. PA.

Supplementary Figure VII. LPS responses of multiple proinflammatory cytokine

genes are upregulated by exposure of monocytes to PA. THP-1 monocytes were




exposed to media (C-black histograms), or media containing 100 M palmitic acid (PA-

white histograms) for 16 hours, then were subsequently exposed to LPS for 4 hours.

Expression of expression of IL-1, IP-10, Resistin, and mRNA was measured by real-

time PCR. In addition, MCP-1 and TNF- (not shown) demonstrated trends towards

amplification but did not reach significance vs. PA and/or LPS. Data are mean ± SEM; n

= 3; * p < 0.05 both vs. LPS alone and vs. PA alone; † p < 0.05 vs. the sum of PA alone +

LPS alone.

Supplementary Figure VIII. A schematic of a proposed amplification mechanism.

TLR4 binds LPS and activates NF-B-mediated gene expression. However, in the

presence of physiologically relevant doses of SFA, excess SFA are converted to

ceramide, which subsequently activates PKC-. PKC- then activates the MAPK

signaling pathway, leading to activation of MAPK-dependent transcription factors such

as c-Jun/AP1 and the C/EBP family. These transcription factors then enhance the

transcriptional effectiveness of NF-B, leading to greatly enhanced expression of several

proinflammatory genes. Solid arrows indicate synthesis steps; dashed arrows indicate

activation steps; diamond boxes represent kinases; arrow boxes represent transcription

factors.




Supplementary Figure I

essi

on

on

tro

l)

25

30

35IL-6 IL-8

ne

Gen

e E

xpr

nd

uct

ion

vs.

co

10

15

20

25

CLPS LA

LPSOA

LPS

Cyt

oki

n(f

old

in

0

5

10

LLA L O

OA L




Supplementary Figure II

A

BB

CC




Supplementary Figure III

ssio

nco

ntr

ol)

100

120

140

sham siTLR4

A*

*

6 G

ene

Exp

res

d i

nd

uct

ion

vs.

c

40

60

80

100

C PALPS

PALPS

Pam3

PAPam3

IL-6

(fo

ld

0

20

P

sio

no

ntr

ol)

800

1000

sham siTLR4

B**

Gen

e E

xpre

ssin

du

ctio

n v

s. c

o

400

600

C PALPS

PALPS

Pam3

PAPam3

IL-8

(f

old

i

0

200

P PA




Supplementary Figure IV




Supplementary Figure V

rol)

4

5

Protein mRNA

Exp

ress

ion

ctio

n v

s. c

on

tr

2

3

4 mRNA

CC

R2

(fo

ld in

du

c

0

1

2

C PA0




Supplementary Figure VI




Supplementary Figure VII




Supplementary Figure VIII




Nutrient Modification of the Innate Immune Response: A Novel … · Nutrient Modification of the...

Documents

Transcript of Nutrient Modification of the Innate Immune Response: A Novel … · Nutrient Modification of the...