Life Sciences 92 (2013) 984–992

Contents lists available at SciVerse ScienceDirect

Life Sciences

j ourna l homepage: www.e lsev ie r .com/ locate / l i fesc ie

Peroxidized unsaturated fatty acids stimulate Toll-like receptor 4 signaling inendothelial cells

Akiko Mutoh ⁎, Shinichiro UedaDepartment of Clinical Pharmacology & Therapeutics, Graduate School of Medicine, University of the Ryukyus, Okinawa, Japan

⁎ Corresponding author at: Department of Clinical PharmaSchool of Medicine, University of the Ryukyus, 207 UehOkinawa, Japan. Tel.: +81 98 895 1195; fax: +81 98 8

E-mail address: a-mutoh@umin.ac.jp (A. Mutoh).

0024-3205/$ – see front matter © 2013 Elsevier Inc. Allhttp://dx.doi.org/10.1016/j.lfs.2013.03.019

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 1 November 2012Accepted 25 March 2013

Keywords:Toll-like receptor 4Unsaturated fatty acidsSaturated fatty acidsPolyunsaturated fatty acidsInflammationLipotoxicityEndothelial cellsLipid peroxide

Aim: Although unsaturated fatty acids are assumed to be protective against inflammatory disorders that in-clude a pathway involving Toll-like receptor 4 (TLR4) activation, they might actually be toxic because oftheir high susceptibility to lipid peroxidation. Here we studied the effects of peroxidized unsaturated fattyacids on the TLR4–nuclear factor (NF)-κB pathway in endothelial cells.Main methods: Confluent cultured endothelial cells from bovine aorta were incubated for 1 h with fattyacids integrated into phosphatidylcholine vesicles. Lipopolysaccharide (LPS) or phosphatidylcholine vesi-cles without fatty acids were also applied as a positive control or a control for fatty acid groups, respectively.Activation of TLR4 and downstream signaling was assessed by membrane fractionation and Western blot-ting or immunofluorescent staining.Key findings: In the same way as LPS, application of sufficiently peroxidized unsaturated fatty acids like oleicacid or docosahexaenoic acid, acutely caused TLR4 translocation to caveolae/raft membranes, leading to ac-tivation of NF-κB signaling in endothelial cells. In contrast, saturated fatty acids did not show such effects.

Applying well-peroxidized unsaturated fatty acids, but not saturated fatty acids, acutely activates the TLR4/NF-κB pathway.Significance: Peroxidation of unsaturated fatty acid is essential for the acute activation of TLR4 by the fattyacids that follow the same pathway as the activation by LPS. Unsaturated fatty acids have been assumed tobe protective against inflammatory disorders, and drugs containing unsaturated fatty acids are now devel-oped and provided. Our result suggests that, for inflammatory disorders involving TLR4 signaling, using un-saturated fatty acids as anti-inflammatory drugs may cause contrary effects.

© 2013 Elsevier Inc. All rights reserved.

Introduction

Since 1970, when studies were conducted in an Eskimo population(Bang et al., 1971; Kromann and Green, 1980), possible benefits of intakeof polyunsaturated fatty acids (PUFAs) for the prevention of atheroscle-rotic cardiovascular diseases have been a focus of research. In fact, a seriesof randomized clinical trials did show that PUFA intake reduces cardiovas-cular event risk, albeit modestly (Burr et al., 1989; Marchioli et al., 2002;Yokoyama et al., 2007). Saturated fatty acids (SFAs), on the other hand,may have pro-atherogenic and pro-diabetic effects through activation ofToll-like receptor 4 (TLR4) signaling which is a crucial pathway in innateimmunity, to bacterial infection, butwhich alsomay promote progressionof atherosclerosis and insulin resistance by causing chronic inflamma-tion through the activation of the transcriptional factors such as NF-κB

cology&Therapeutics, Graduateara, Nishihara-cho, 903-0215,95 1447.

rights reserved.

resulting in secretion of pro-inflammatory cytokines and chemokines(Holland et al., 2011; Kim et al., 2007; Shi et al., 2006; Suganami et al.,2007). In contrast, PUFAs reportedly suppress such TLR4-mediated in-flammatory reactions (Lee et al., 2001, 2003). However, it is noteworthythat PUFAs may also have toxic effects, as they are highly susceptible tolipid peroxidation, which is implicated in the pathogenesis of variousdiseases (Guillen andGoicoechea, 2008; Serini et al., 2011). And indeed,there is reported experimental and clinical evidence that suggests anassociation of lipid peroxides from PUFAs with atherosclerosis(Esterbauer et al., 1992; Glavind et al., 1952; Jira et al., 1998; Stringeret al., 1989). Thus, the aim of the present study was to determinewhether peroxidized unsaturated fatty acids stimulate TLR4 signaling.

Materials and methods

Cell culture

Fetal bovine aortic endothelial cells (ECs) were purchased fromJapanese Collection of Research Bioresources (JCRB, Osaka, Japan)or were isolated from bovine fetuses, as previously described(Mutoh et al., 2008). ECs were cultured in Medium 199 (GIBCO,

985A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

Life technologies Japan Ltd., Tokyo, Japan) supplemented with 20%complement-depleted fetal bovine serum (FBS) (Nichirei Biosci-ences Inc. Tokyo, Japan) in a humidified CO2 incubator at 37 °C,were passaged before full confluence, and were used before passage

Fig. 1. Effect of fatty acid peroxidation on TLR4/NF-κB signaling in endothelial cells. Experivesicle preparation sets divided into halves), immediately following preparation and afterPCFA (2.5 × 10−6 M as fatty acids), or UV-irradiated PC or PCFA (UV-PC, UV-PCFA), respectivetrifugation. The fractions are along the lines of sucrose gradient that start from lowest densitybranes, detected by Flot-1 and Cav-1, found in fractions 3 and 4. TLR4 blots showed ECs incubavesicles used in a and c analyzed by the TBARSmethod. (c) Expression of IκB-α in ECwhole cellthe difference being the freshness (indicating the level of peroxidation of fatty acids), as show

15 for all experiments. Normal cell functions (eNOS phosphorylationand nitric oxide production) of cultured ECs isolated from bovinefetuses were confirmed at passage 20, and therefore it was decidedthat it was reasonable to use ECs up to passage 15 in the current study.

ments presented in a, b and c were performed simultaneously, twice (using the sameaging, 5 days later. ECs were incubated with no stimulation (None), 10 ng/ml LPS, PC,ly for 1 h. (a) Distribution of TLR4 in plasmamembranes determined by sucrose ultracen-(fraction No. 1; 5% sucrose) to highest density (No. 9; 42.5% sucrose). Caveolae/raft mem-ted with freshly prepared (A) or 5-day-old (B) vesicle samples. (b) MDA concentration inlysates byWestern blotting in duplicate. The same vesicles were used in both panels, withed in a and b.

Fig. 1 (continued).

986 A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

Fatty acid preparation

In cell cultures, fatty acids integrated into phosphatidylcholine (PC)vesicles were used instead of bovine serum albumin (BSA)-conjugatedfatty acids because of a concern about contamination with lipopolysac-charide (LPS) in BSA and resultant non-specific BSA effects on ECfunction, including increased activity of endothelial nitric oxidesynthase (eNOS) (Maniatis et al., 2006). Hydrogenated phosphatidyl-choline (PC), chloroform, palmitic acid (PA), and oleic acid (OA) werepurchased from Wako Pure Chemicals Industries, Ltd., Osaka, Japan.Docosahexaenoic acid (DHA) was purchased from Cayman ChemicalCompany (MI, USA). Fatty acids were integrated into PC vesicles usinga protocol that was developed by reference to previously publishedreports (Anel et al., 1993; Huang and Thompson, 1974; Nojima, 1988),as follows: 50 mg PC was dissolved in 1 ml chloroform in a 50-mlround-bottom recovery flask. To prepare vesicles containing fattyacids, stock solutions in chloroform were added to the flask as follows:200 μl of 1 × 10−1 MPA (to yield PC–PA vesicles; because of its relativeinsolubility, twice as much PA as other fatty acid was loaded) or 100 μlof 1 × 10−1 M OA or DHA (to yield PC–OA and PC–DHA vesicles). Inaddition, to prepare vesicles containing a 1:1 molar ratio mixture ofPA and OA (to yield PC–FA vesicles), 100 μl of each stock solution wasplaced in the flask. The flask was plugged using a silicone stopperhaving connections to an N2 gas supply and a vacuum pump. Chloro-form was evaporated by intermittent addition of N2 gas. When visibleliquid chloroform had almost disappeared, the flask was maintainedunder vacuum for a few hours until completely dry. The resultant thinlayer on the inner flask was reconstituted in Hank's balanced salt solu-tion (HBSS, GIBCO) at 70 °C and used for vesicle formation by vigorousstirring with a teaspoon of glass beads in an N2 atmosphere. The flaskwas placed in a sonication bath for 1 min and the vesicle slurry wascentrifuged at 10,500 ×g for 60 min at 20 °C. Supernatants containingthe vesicles were passed through a 0.45 mm cellulose acetate filter(Sartorius Stedim Biotech, Goettingen, Germany) and were then storedin sealed glass tubes filled with N2 gas, and were protected from light.Using the ToxinSensor Endotoxin Detection System (GenScript, NJ,US), the LPS content in the 4 different lots of each PC and PC–FA (halfof these was freshly prepared and other half aged for 2 weeks) wasassessed and found to be lower than the detection sensitivity limit of

the assay (b0.005 EU/ml), whereas 2% BSA (RIA Grade, Sigma ChemicalCo., MO, US) HBSS (GIBCO) solution contained 0.57 EU/ml.

Determination of vesicle fatty acid and malondialdehyde concentrations

Concentrations of fatty acids in the vesicles were determined usingLabAssay™ NEFA (Wako Pure Chemical Industries, Ltd.). Average fattyacid concentrations (n = 4–6) in the resultant vesicles were 0.004 mM(undetected) in PC, 0.043 mM in PC–PA, 0.145 mM in PC–OA,0.047 mM in PC–DHA and 0.082 mM in PC–FA. Malondialdehyde(MDA) in the vesicles was measured by the TBARS assay kit (CaymanChemical Company).

Western blotting

Confluent ECs in 3.5 cm dishes were washed and culture mediumwas replaced with HBSS, cells were incubated for 1 h, and then LPS,PC vesicles only, or PC vesicles containing fatty acid were added tothe medium and incubated for 1 h. SDS-PAGE and Western blottingwere performed as previously described (Mutoh et al., 2008). Immuno-blotting of the membranes was done by incubation with IκB-α antibodyor NF-κB p65 antibody (1:2000) (Cell Signaling Technology Japan, K.K.,Tokyo, Japan) overnight at 4 °C. Immunoreactive proteins were detectedby HRP-conjugated secondary antibodies (1:20,000) (GE HealthcareJapan, Tokyo, Japan) and chemiluminescent reagent (ImmunoStar LD,Wako Pure Chemical Industries, Ltd.) by following the manufacturers'instructions. Detection of beta-actin was performed by incubating themembranes with HRP-conjugated anti-beta-actin (1:90,000) (Abcam,Cambridge, UK) for 20 min at room temperature followed by carryingout the chemiluminescent reaction as described above.

TLR4 activation assay by membrane fractionation

Isolation of detergent-resistantmembrane fractionswas performed aspreviously described (Sonnino and Prinetti, 2008). Briefly, ECs were cul-tured in 10 cm dishes. After treatment with LPS or vesicles as describedabove, ECs were washed and scraped off the plates with ice-cold PBScontaining protease inhibitor cocktail and phosphatase inhibitor cock-tail (Sigma-Aldrich Japan Co., LLC, Tokyo, Japan) and then centrifuged.

987A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

ECs pellets were suspended in lysis buffer containing 0.5% Triton X-100(ICN Biomedicals, Inc., Ohio, USA), 10 mM Tris–HCl (Invitrogen),150 mM NaCl (Wako), 5 mM EDTA (GIBCO) and protease and phos-phatase inhibitors, and then left on ice for 1 h. ECs were disrupted bypassing through a 27 G needle 6 times and were then centrifuged at1300 ×g for 5 min at 4 °C to obtain the post nuclear supernatant(PNS). Each 1 ml PNS was mixed with 1 ml 85% sucrose in a 13PAtube (Hitachi Koki, Co. Ltd., Tokyo, Japan) and then a 7 ml 30%–5%continuous sucrose gradient was layered on the PNS samples (now ina 42.5% sucrose solution) using a gradient maker (Sanplatec Corpora-tion, Osaka, Japan). The tubes were ultracentrifuged in a CP56G II(Hitachi Koki) equipped with a swing rotor (P40 S) at 200,000 ×g for18 h. The gradient was separated into 1 ml fractions from the top toobtain fraction Nos.1 (the lowest density fraction in the sucrosegradient) to 9 (the highest density fraction). Proteins in the fractionswere purified by trichloroacetic acid (TCA) precipitation and thensubjected to SDS-PAGE and Western blotting, as described above.Membranes were immunoblotted with polyclonal (1:1000) (Abnova

Fig. 2. Peroxidized PCFA dose–response effects on the TLR4/NF-κB pathway. (a) ECs incubateTLR4 activation. Percentage of TLR4 density in caveolae/raft membrane (fractions 3 and 4) rdose–response effect on IκB-α degradation. PC or PCFA at indicated concentrations or 10*p b 0.05, †p b 0.01 vs. PC. (c) Immunofluorescent staining for phospho-p65 (green). Nucle

Corporation, Taipei, Taiwan) or monoclonal (1:1000) (BD Biosciences,CA, USA) anti-TLR4 antibody, or anti-caveolin-1 or anti-flotillin-1 anti-body (1:5000) (BD Biosciences).

Immunofluorescent staining for phospho-p65

ECs were cultured on 15 mm in diameter micro coverslips. Afterstimulation with LPS (10 ng/ml, as the positive control) or vesicles, asdescribed in theWestern blotting section, ECs were washed with salineand fixed with 1% TCA for 15 min. TCA was removed by washing withTBST (an aqueous solution containing 137 mM NaCl, 2.7 mM KCl,24.7 mM Tris and 0.1% Tween 20, pH 7.4), and then ice cold 0.1% TritonX-100 in TBST was applied to the coverslips on ice for 5 min, followedby further TBST washing. After blocking with 5% BSA in TBST for30 min, phospho-Ser536 p65 rabbit monoclonal antibody (1:200)(Cell Signaling Technology) was applied overnight at 4 °C. The cov-erslips were washed 3 timeswith 0.5% BSA in TBST, followed by incu-bation for 1 h at room temperature with the secondary antibody,

d with PC or peroxidized PCFA at indicated concentrations for 1 h and then assayed forelative to total TLR4 in all fractions is shown. Mean (SD). n = 3. (b) Peroxidized PCFAng/ml LPS administered to ECs in duplicate and incubated for 1 h. Mean (SD). n = 4.i were visualized by staining with DAPI (red).

Fig. 2 (continued).

988 A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

Alexa Fluor 488-conjugated anti-rabbit IgG (1:100) (Invitrogen) and1 mg/ml 4′,6-diamidino-2-phenylindole (DAPI, Wako) for nuclearstaining. The coverslips were washed 3 times with TBST, rinsedwith pure water, and then mounted on glass slides using PermaFluoraqueous mounting medium (Thermo Fisher Scientific Inc., MA, USA).Fluorescent images of Alexa Fluor 488 and DAPI labeling wereobtained with a Leica confocal laser scanning microscope system(Leica Microsystems, Wetzlar, Germany). X–y fluorescent imageswere stacked along the z-axis from top to bottom in 5 planes andthen merged into 1 image.

Statistical analysis

The Fisher's PLSD test was applied to determine significant differ-ences between pairs after performing analysis of variance (ANOVA).Differences having p b 0.05 were considered significant.

Results

Fig. 1 summarizes results of an exploratory study to find out impor-tant factors in TLR4 activation by fatty acids. Fig. 1a, b, and c shows results

of TLR4 translocation assay; MDA concentration assay, and IκB-a degra-dation, respectively. These experiments were performed simultaneous-ly using half of one set of vesicles (PC, PCFA, UV-PC, UV-PCFA) just afterbeing prepared as a “fresh” set. When a period of 5 days had elapsedafter the vesicles' preparation, the same experiments were repeatedusing the other half as an “aged” set.

The results of Fig. 1a clarify conditions determining whether fattyacids act in amanner similar to LPS to activate TLR4 in ECs. The activationof TLR4was assessed by translocation of TLR4 from the non-caveolae/raftto the caveolae/raft domain, as reported by Triantafilou et al. (2002) andWalton et al. (2003). Fig. 1a shows TLR4 localization in detergent-resistant membrane fractions (low numbers indicate low-density mem-brane fractions). Fractions 3 and 4, which contained abundant flotillin-1(Flot-1) and caveolin-1 (Cav-1), are caveolae/raft membrane fractions.In unstimulated ECs, TLR4 was mainly distributed in non-caveolae/raftmembranes (Fig. 1a, top left). Incubation of ECs with 10 ng/ml LPS for1 h led to accumulation of TLR4 in caveolae/raft membranes (Fig. 1a,top right), consistent with previous reports. In a similar fashion, fattyacids were applied to EC for 1 h. A freshly prepared blend of PA and OA(PCFA) (2.5 × 10−6 M as fatty acid) did not activate TLR4 (Fig. 1a,middleright). However, the PCFA peroxidized either by aging for 5 days (aged

Table 1MDA concentration in unsaturated fatty acids increases with time.

PC PC–PA PC–OA PC–DHA

MDA concentration (×10−6 M) in 1 × 10−6 M of each fatty acids Fresh N.D. N.D. 0.020 (0.018) 0.067 (0.045)Aged N.D. N.D. 0.036 (0.015) 0.191 (0.031)

MDA concentrations in 1 × 10−6 M of each freshly prepared or aged (5 days old to 3 weeks) fatty acid as determined by the TBARS method. Mean (SD). (n = 3). N.D.: not detected.

989A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

PCFA) (Fig. 1a, middle right) or by UV irradiation for 10 min (UV-PCFA)(Fig. 1a, lower right) translocated TLR4 to the caveolae/raft domain. Iden-tically treated control PC vesicles had no effects (Fig. 1a,middle and lowerleft). In Fig. 1 results of Flot-1 and Cav-1 blots are shown only for the sam-ple to which “fresh” vesicles were applied. Application of “aged” vesiclesresulted in the same distribution. To confirm lipid peroxidation afteraging or UV radiation, concentrations of MDA in PCFA were measuredby the TBARS method, and were found to be elevated significantly, com-pared to control, after aging for 5 days (Fig. 1b).

The next experiments were done to assess whether translocationof TLR4 by peroxidized fatty acids could lead to alterations down-stream of TLR signaling. Fig. 1c shows effects of PCFA on expressionof IκB-α in ECs as assessed by Western blotting. PCFA enhanced deg-radation of IκB-α after UV-radiation and aging (Fig. 1c), consistent

Fig. 3. Fatty acids with unsaturated bond(s) activate TLR4/NF-κB signaling. ECs incubated foas indicated. Aged vesicles used as described in Table 1 legend (i.e., PC–OA and PC–DHA acaveolae/raft membrane (fractions 3 and 4) relative to TLR4 level in all fractions. *p b 0.05 vsof IκB-α. Representative blots. Mean (SD). n = 4. *p b 0.05, †p b 0.001 vs. control or PC.

with the TLR4 translocation response shown in Fig. 1a, suggestingthat fatty acids have effects similar to those of LPS on TLR4 activation,but only when peroxidized.

Titration of peroxidized PCFA on TLR4 translocation is shown inFig. 2a. TLR4 significantly translocated to caveolae/raft membranes inPCFA concentrations greater than 0.5 × 10−6 M. Dose dependency wasalso observed in the degradation of IκB-α, as shown in Fig. 2b. PCFA atthe dose of 1 × 10−6 M which caused apparent translocation of TLR4and enhanced degradation of IκB-α, phosphorylated serine 536 of p65of NF-κB as shown by immunofluorescent staining (Fig. 2c). The MDAconcentration in the 1 × 10−6 M peroxidized PCFA used in these exper-iments was estimated to be 0.030 (±0.009) × 10−6 M.

The next experimentwas designed to ascertain differential effects ofSFA and unsaturated fatty acids, including PUFA, on the TLR4/NF-κB

r 1 h with no stimulation (None), 10 ng/ml LPS, PC, or PC containing a single fatty acid,re peroxidized). (a) TLR4 activation by aged fatty acid samples. Percentage of TLR4 in. PC. n = 4–7. (b) Western blotting to analyze phosphorylation of p65 and degradation

Fig. 3 (continued).

990 A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

pathway. As shown in Table 1, theMDA concentrationwas significantlyelevated in aged OA or DHA, but not in aged PC or PA. Using these agedsamples, the following experiments were performed. Incubation of ECswith PC–OA or PC–DHA for 1 h, caused translocation of TLR4, whereasPC–PA did not (Fig. 3a). PC–DHA significantly enhanced phosphoryla-tion of p65 and degradation of IκB-α (Fig. 3b). However, PC–OA hadno apparent effect on phosphorylation of p65 and only a modest effecton degradation of IκB-α. PC–PAdid not have any effects on downstreamcomponents of TLR4.

Discussion

The present study clearly showed that peroxidized fatty acids havingunsaturated bond(s), but not SFAs, dose-dependently translocate TLR4to the caveolae/raft domain with downstream signal activation, asevidenced by enhancement of IκB-α degradation and phosphorylationof serine 536 of NF-κB p65. Triantafilou and colleagues showed that re-ceptor molecules implicated in LPS- or bacteria-induced cellular events,including TLR4, accumulate in lipid rafts after stimulation with LPS orbacteria, and that disruption of the rafts inhibits TLR4 signaling(Triantafilou et al., 2002). TLR4 accumulation in the lipid microdomainappears to be a key event to initiate LPS- or bacteria-induced cell activa-tion; such accumulation may be more direct evidence of activated TLR4than alterations of downstream signals. To the best of our knowledge,this is the first report of fatty acids causing translocation of TLR4 asdirect evidence for activated TLR4 signaling. These results, therefore,raise the possibility that peroxidized fatty acids, as an endogenousligand, could activate TLR4 signaling directly. One could suggest thattranslocation of TLR4 is far from direct evidence for TLR4 activationand that NF-κB activation could occur through non-TLR pathway. Analternative interpretation is that translocation of TLR4 is not relatedto NF-κB activation and the activation occurs through the non-TLR

pathway. Activation through the non-TLR pathway cannot be experi-mentally excluded, because there is no TLR-specific antagonist. Howev-er, the dose–effect relationship of peroxidized fatty acids with %TLR4 inthe caveolae/raft membranes, IκB-α degradation, and phosphorylationof p65 is consistent with our interpretation of the present results.

It should be noted that translocation of TLR4 occurred after a relative-ly short incubation (1 h) of cells with a relatively low concentration(0.5–2.5 × 10−6 M) of fatty acids. Given that LPS rapidly stimulatesTLR4 at 10 ng/ml (within a few minutes) (Hambleton et al., 1996),these findings support our hypothesis. On the other hand, previous stud-ies demonstrated thatmuch longer incubationswith SFAs (for 3 to 16 h)at much higher concentrations (ranging from 1 × 10−4 to 1 × 10−3 M)were required to activate the TLR4/NF-κB pathway in various tissues andcells (Holland et al., 2011; Kim et al., 2007; Shi et al., 2006; Suganami etal., 2007). Lack of effects of SFA (PA) on TLR4 signaling in the current ex-periments, therefore, may be explained by the possibility that there wasless exposure of cells to the SFA than there was in previous studies. Thismight also suggest that SFAs, unlike LPS, enhances TLR4-related signalingthrough chronic displacement of fatty acids in membrane phospholipidsrather than by acting as an agonist to acutely stimulate TLR4.

Although TLRs serve in the innate immune system as sensors ofnon-self pathogenic molecules, more than 20 molecules, includingSFAs, have been proposed as endogenous ligands. However, whetherthesemolecules can act as ligands for TLR4 has been questioned recentlybecause of a concern that stimulation of TLRs by such candidate ligandsmight actually be due to minor LPS contamination in the albumin towhich they were complexed (Erridge and Samani, 2009). In fact, Erridgeand Samani showed that SFA–BSA complexes, but not SFA alone, activat-ed TLR-dependent signaling (Erridge, 2010). In the present experiments,the PC vesicle method was adopted to exclude the likelihood of contam-ination by LPS in BSA. This could explain the lack of effect of PA on TLR4signaling in this study.

991A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

Elevated free fatty acid levels are implicated in the development ofmetabolic and atherosclerotic diseases. While SFAs seem to serve as“the bad guys,” facilitating inflammatory reactions through TLR4 activa-tion, causing insulin resistance and atherosclerosis, PUFAs serve as “thegood guys,” protecting against adverse effects of SFAs. Such centraldogma has been fairly well established by substantial experimentaland clinical evidence (Bang et al., 1971; Burr et al., 1989; Holland etal., 2011; Kim et al., 2007; Kromann and Green, 1980; Marchioli et al.,2002; Shi et al., 2006; Suganami et al., 2007; Yokoyama et al., 2007).However, it is of note that PUFAs are highly susceptible to peroxidation,which has also been implicated in various diseases. Lipid peroxideswere first detected in atherosclerotic human aorta tissue more than50 years ago (Glavind et al., 1952). Several groups have since reportedelevated plasma lipid peroxides in patients with atherosclerosis,dyslipidemia and diabetes (Dormandy et al., 1973; Sato et al., 1979;Stringer et al., 1989; Uysal et al., 1986). More recently,Walter et al. dem-onstrated that circulating lipid hydroperoxides have a predictive valuefor adverse cardiovascular outcome in patients with stable angina(Walter et al., 2008). TLR4 activation by peroxidized fatty acids shownin the present study, as well as possible effects mediated through thenon-TLR pathway, may be related to such links between lipid peroxida-tion and cardiovascular diseases. In fact, TLR4 was expressed in macro-phages and endothelial cells of atherosclerotic lesion (Edfeldt et al.,2002). It can be assumed that secretion of cytokines and chemokines fol-lowing the activation of TLR4 receptors in such cells contribute to the de-velopment of atherosclerosis and plaque activation. Considered togetherwith previous reports, the results presented here suggest that PUFAs arenot necessarily protective against atherosclerosis. Several clinical trialsshowedmodest effects of n−3 fatty acids on cardiovascular events. A re-cent meta-analysis, however, of carefully designed clinical trials of n−3fatty acids has not demonstrated any benefit for cardiovascular outcome(Kwak et al., 2012). A very recent double-blinded, placebo-controlledstudy showed that daily supplementation with 1 g of n−3 fatty acidsdid not reduce the rate of cardiovascular events in patients withdysglycemia (The ORIGIN Trial Investigators, 2012). The present resultspartly explain such lack of efficacy of n−3 fatty acids for cardiovascularevent prevention. Interestingly, it was recently reported that omega-3fatty acids, such as DHA, were rather harmful in patients with acutelung injury (Rice et al., 2011), in which TLR4 signaling is known to be in-volved (Xiang and Fan, 2010).

In this work there are some limitations to be noted. First of all,our experiments were performed on cultured cells. Strictly speak-ing, our experimental results do not tell anything definite aboutthe effects of PUFA on the vascular function of human whole body.Another limitation is that we have studied only acute effect ofperoxidized unsaturated fatty acids on TLR4 signaling. Chronic ef-fects also need to be studied, if one wants to elucidate the develop-ment of atherosclerosis.

In conclusion, peroxidation of fatty acids having unsaturated bond(s)is essential for inducing their acute action on TLR4, which follows thesame pathway as LPS. This would suggest that lipotoxicity on vascu-lar function requires fatty acids having unsaturated bond(s) and anoxidative microenvironment. Additionally, use of PUFAs as an anti-inflammatory agent may cause opposite effects in patients with in-flammatory disorders involving TLR4 signaling.

Conflict of interest statement

The authors report no conflicts of interest.

Sources of funding

Financial support for this study was provided by two Ministry ofEducation, Science and Culture Grants-in-Aid for Scientific Research(16590439 and 13670734 to S.U.).

Acknowledgment

We thank OkinawakenMeat Center Company and Dr. Kouichi Yabikufor their help with the dissection of bovine fetuses and Dr. Naoki Mori forletting us to use their confocal microscope.

References

Anel A, Richieri GV, Kleinfeld AM. Membrane partition of fatty acids and inhibition of Tcell function. Biochemistry 1993;32(2):530–6.

Bang HO, Dyerberg J, Nielsen AB. Plasma lipid and lipoprotein pattern in GreenlandicWest-coast Eskimos. Lancet 1971;1(7710):1143–5.

Burr ML, Fehily AM, Gilbert JF, Rogers S, Holliday RM, Sweetnam PM, et al. Effectsof changes in fat, fish, and fibre intakes on death and myocardial reinfarction:diet and reinfarction trial (DART). Lancet 1989;2(8666):757–61.

Dormandy JA, Hoare E, Khattab AH, Arrowsmith DE, Dormandy TL. Prognostic significance ofrheological and biochemical findings in patients with intermittent claudication. BrMed J1973;4(5892):581–3.

Edfeldt K, Swedenborg J, Hansson GK, Yan ZQ. Expression of toll-like receptors inhuman atherosclerotic lesions: a possible pathway for plaque activation. Cir-culation 2002;105(10):1158–61.

Erridge C. Endogenous ligands of TLR2 and TLR4: agonists or assistants? J Leukoc Biol2010;87(6):989–99.

Erridge C, Samani NJ. Saturated fatty acids do not directly stimulate Toll-like receptorsignaling. Arterioscler Thromb Vasc Biol 2009;29(11):1944–9.

Esterbauer H, Gebicki J, Puhl H, Jurgens G. The role of lipid peroxidation and antioxidants inoxidative modification of LDL. Free Radic Biol Med 1992;13(4):341–90.

Glavind J, Hartmann S, Clemmesen J, Jessen KE, Dam H. Studies on the role oflipoperoxides in human pathology. II. The presence of peroxidized lipids in the ath-erosclerotic aorta. Acta Pathol Microbiol Scand 1952;30(1):1–6.

Guillen MD, Goicoechea E. Toxic oxygenated alpha, beta-unsaturated aldehydes andtheir study in foods: a review. Crit Rev Food Sci Nutr 2008;48(2):119–36.

Hambleton J, Weinstein SL, Lem L, DeFranco AL. Activation of c-Jun N-terminal kinasein bacterial lipopolysaccharide-stimulated macrophages. Proc Natl Acad Sci U S A1996;93(7):2774–8.

Holland WL, Bikman BT, Wang LP, Yuguang G, Sargent KM, Bulchand S, et al.Lipid-induced insulin resistance mediated by the proinflammatory receptor TLR4requires saturated fatty acid-induced ceramide biosynthesis in mice. J Clin Invest2011;121(5):1858–70.

Huang C, Thompson E. Preparation of homogenous, single-walled phosphatidylcholinevesicles. Methods Enzymol 1974;32(45):485–9.

Jira W, Spiteller G, Carson W, Schramm A. Strong increase in hydroxy fatty acids de-rived from linoleic acid in human low density lipoproteins of atherosclerotic pa-tients. Chem Phys Lipids 1998;91(1):1-11.

Kim F, Pham M, Luttrell I, Bannerman DD, Tupper J, Thaler J, et al. Toll-like receptor-4mediates vascular inflammation and insulin resistance in diet-induced obesity.Circ Res 2007;100(11):1589–96.

Kromann N, Green A. Epidemiological studies in the Upernavik district, Greenland. Inci-dence of some chronic diseases 1950–1974. Acta Med Scand 1980;208(5):401–6.

Kwak SM, Myung SK, Lee YJ, Seo HG. Efficacy of omega-3 fatty acid supplements(eicosapentaenoic acid and docosahexaenoic acid) in the secondary prevention of car-diovascular disease: a meta-analysis of randomized, double-blind, placebo-controlledtrials. Arch Intern Med 2012.

Lee JY, Sohn KH, Rhee SH, Hwang D. Saturated fatty acids, but not unsaturated fattyacids, induce the expression of cyclooxygenase-2 mediated through Toll-like re-ceptor 4. J Biol Chem 2001;276(20):16683–9.

Lee JY, Ye J, Gao Z, Youn HS, Lee WH, Zhao L, et al. Reciprocal modulation of Toll-likereceptor-4 signaling pathways involvingMyD88 and phosphatidylinositol 3-kinase/AKTby saturated and polyunsaturated fatty acids. J Biol Chem 2003;278(39):37041–51.

Maniatis NA, Brovkovych V, Allen SE, John TA, Shajahan AN, Tiruppathi C, et al. Novelmechanism of endothelial nitric oxide synthase activation mediated by caveolaeinternalization in endothelial cells. Circ Res 2006;99(8):870–7.

Marchioli R, Barzi F, Bomba E, Chieffo C, Di Gregorio D, Di Mascio R, et al. Early protec-tion against sudden death by n−3 polyunsaturated fatty acids after myocardial in-farction: time-course analysis of the results of the Gruppo Italiano per lo Studiodella Sopravvivenza nell'Infarto Miocardico (GISSI)-Prevenzione. Circulation2002;105(16):1897–903.

Mutoh A, Isshiki M, Fujita T. Aldosterone enhances ligand-stimulated nitric oxide pro-duction in endothelial cells. Hypertens Res 2008;31(9):1811–20.

Nojima S. The liposomes. Tokyo: Nankodo; 1988.ORIGIN Trial Investigators. n−3 fatty acids and cardiovascular outcomes in patients

with dysglycemia. N Engl J Med 2012;367(4):309–18.Rice TW, Wheeler AP, Thompson BT, deBoisblanc BP, Steingrub J, Rock P. Enteral

omega-3 fatty acid, gamma-linolenic acid, and antioxidant supplementation inacute lung injury. JAMA 2011;306(14):1574–81.

Sato Y, Hotta N, Sakamoto N, Matsuoka S, Ohishi N, Yagi K. Lipid peroxide level in plasmaof diabetic patients. Biochem Med 1979;21(1):104–7.

Serini S, Fasano E, Piccioni E, Cittadini AR, Calviello G. Dietary n−3 polyunsaturatedfatty acids and the paradox of their health benefits and potential harmful effects.Chem Res Toxicol 2011;24(12):2093–105.

Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 links innate immunity andfatty acid-induced insulin resistance. J Clin Invest 2006;116(11):3015–25.

Sonnino S, Prinetti A. Membrane lipid domains and membrane lipid domain preparations:are they the same thing? Trends Glycosci Glycotechnol 2008;20(116):315–40.

992 A. Mutoh, S. Ueda / Life Sciences 92 (2013) 984–992

Stringer MD, Gorog PG, Freeman A, Kakkar VV. Lipid peroxides and atherosclerosis. BMJ1989;298(6669):281–4.

Suganami T, Tanimoto-Koyama K, Nishida J, Itoh M, Yuan X, Mizuarai S, et al. Role ofthe Toll-like receptor 4/NF-kappaB pathway in saturated fatty acid-induced in-flammatory changes in the interaction between adipocytes and macrophages.Arterioscler Thromb Vasc Biol 2007;27(1):84–91.

Triantafilou M, Miyake K, Golenbock DT, Triantafilou K. Mediators of innateimmune recognition of bacteria concentrate in lipid rafts and facilitatelipopolysaccharide-induced cell activation. J Cell Sci 2002;115(Pt 12):2603–11.

Uysal M, Bulur H, Sener D, Oz H. Lipid peroxidation in patients with essential hyperten-sion. Int J Clin Pharmacol Ther Toxicol 1986;24(9):474–6.

Walter MF, Jacob RF, Bjork RE, Jeffers B, Buch J, Mizuno Y, et al. Circulating lipid hydroper-oxides predict cardiovascular events in patients with stable coronary artery disease:the PREVENT study. J Am Coll Cardiol 2008;51(12):1196–202.

Walton KA, Cole AL, Yeh M, Subbanagounder G, Krutzik SR, Modlin RL, et al. Specificphospholipid oxidation products inhibit ligand activation of toll-like receptors 4and 2. Arterioscler Thromb Vasc Biol 2003;23(7):1197–203.

Xiang M, Fan J. Pattern recognition receptor-dependent mechanisms of acute lunginjury. Mol Med 2010;16(1–2):69–82.

Yokoyama M, Origasa H, Matsuzaki M, Matsuzawa Y, Saito Y, Ishikawa Y, et al. Effectsof eicosapentaenoic acid onmajor coronary events in hypercholesterolaemic patients(JELIS): a randomised open-label, blinded endpoint analysis. Lancet 2007;369(9567):1090–8.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具