Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–29

Contents lists available at SciVerse ScienceDirect

Prostaglandins, Leukotrienes and EssentialFatty Acids

0952-32http://d

n CorrTel.: +49

E-m

journal homepage: www.elsevier.com/locate/plefa

Comparison of free serum oxylipin concentrations in hyper- vs.normolipidemic men

Jan Philipp Schuchardt a, Simone Schmidt a, Gaby Kressel a, Hua Dong b, Ina Willenberg c,Bruce D. Hammock b, Andreas Hahn a, Nils Helge Schebb c,n

a Institute of Food Science and Human Nutrition, Leibniz University Hannover, Germanyb Department of Entomology and Cancer Center, University of California, Davis, CA, USAc Institute of Food Toxicology and Analytical Chemistry, University of Veterinary Medicine Hannover, Germany

a r t i c l e i n f o

Article history:Received 27 November 2012Received in revised form5 February 2013Accepted 2 April 2013

Keywords:EicosanoidsPUFAArachidonic acidEicosapentaenoic acidHyperlipidemiaOmega-3 fatty acids

78/$ - see front matter & 2013 Elsevier Ltd. Ax.doi.org/10.1016/j.plefa.2013.04.001

espondence to: Bischofsholer Damm 15, 3017511 8567780; fax: +49 511 8567409.

ail address: schebb@wwu.de.(N.H. Schebb)

a b s t r a c t

Oxylipins, the oxidation products of unsaturated fatty acids (FA), are potent endogenous mediators beinginvolved in the regulation of various biological processes such as inflammation, pain and bloodcoagulation. Compared to oxylipins derived from arachidonic acid (AA) by cyclooxygenase action, i.e.prostanoides, only limited information is available about the endogenous levels of hydroxy-, epoxy- anddihydroxy-FA of linoleic acid (LA), AA, α-linolenic acid (ALA), eicosapentaenoic acid (EPA) anddocosahexaenoic acid (DHA) in humans. Particularly, it is unknown how metabolic disorders affectendogenous oxylipin levels in humans. Therefore, in the present study we compared the serumconcentrations of 44 oxylipins in 20 normolipidemic with 20 hyperlipidemic (total cholesterol4200 mg/dl; LDL-C4130 mg/dl; TG4150 mg/dl) men (age 29–51 y). The serum concentration variedstrongly among subjects. For most hydroxy-, epoxy- and dihydroxy-FA the concentrations werecomparable to those in plasma reported in earlier studies. Despite the significant change in blood lipidlevels the hyperlipidemic group showed only minor differences in oxylipin levels. The hyperlipidemicsubjects had a slightly higher serum concentration of 8,9-DiHETrE, 5-HEPE, 10,11-DiHDPE, and a lowerconcentration of 12,13-DiHOME, 12-HETE, 9,10-DiHODE, and 12,13-DiHODE compared to normolipidemicsubjects. Overall the hydroxy-, epoxy- and dihydroxy-FA levels were not changed suggesting that mildcombined hyperlipidemia has no apparent effect on the concentration of circulating oxylipins. Bycontrast, serum levels of several hydroxy-, epoxy-, and dihydroxy-FA are dependent on the individualstatus of the parent FA. Particularly, a strong correlation between the EPA content in the erythrocytemembrane and the serum concentration of EPA derived oxylipins was observed. Given that the synthesisof EPA from other n-3 FA in humans is low; this suggests that oxylipin levels can be directly influenced bythe diet.

& 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Omega-6 fatty acids (n-6 FA) such as linoleic acid (LA, C18:2) orarachidonic acid (AA, C20:4) and n-3 FA such as α-linolenic acid(ALA, C18:3), eicosapentaenoic acid (EPA, C20:5) and docosahex-aenoic acid (DHA, C22:6) are essential polyunsaturated FA (PUFA).Humans lack the desaturases to synthesize LA and ALA de novo,while LA and AA as well as ALA, EPA and DHA can be transformedfrom one into another in a multistage enzymatic chain elongationand desaturation process [1,2]. However, the conversion rate fromALA to EPA and especially to DHA is low (�1%), and thus the status

ll rights reserved.

3 Hannover, Germany.

of EPA and DHA relies largely on the dietary intake in man [1,3,4].Numerous studies demonstrate that the intake of EPA and DHAwith the diet is associated with various beneficial health effects[5–8].

The oxidation products of PUFA, so called oxylipins, are highlypotent mediators [9,10]. The formation and biology of AA derivedoxylipins, i.e. eicosanoids, has been intensively studied during thepast 40 years. Comparatively little attention has been paid to theanalogous oxidation products of other n-6 PUFA and n-3 PUFA[11,12].

In mammals, oxylipins are formed via four major routes/pathways(Fig. 1): (i) Conversion by cyclooxygenases (COXs) yields instableendoperoxides which are further converted to several highly potentmediators such as prostaglandins, thromboxanes and prostacyclins[10]. (ii) Lipoxygenases (LOXs) form regio- and stereo-selective hydro-peroxides, which serve as precursors for leukotrienes and hepoxylins

Fig. 1. Top—Simplified scheme of the formation of hydroxy-, epoxy- and dihydroxy-fatty acids from their PUFA precursors. Bottom—The structures of several oxylipins areexemplary shown for the generation of 12-HETE via the lipoxygenase (LOX) pathway and 14(15)-EpETrE and 14,15-DiHETrE by Cytochrom-P450 monooxygenases (CYP) andthe soluble epoxide hydrolase (sEH).

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–2920

and can be reduced to hydroxy-FA (HETE, HODE, HOTRrE, HEPE,HDoHE, Fig. 1) [13]. (iii) Metabolism of PUFA by Cytochrom-P-450monooxygenases (CYPs) leads to hydroxy-FA at ω and ω-1 position(CYP4 family) while family CYP2 dominantly forms epoxy-FA (EpOME,EpETrE, EpODE, EpETE, and EpDoPE) [9,14,15]. These highly potentmediators are chemically stable under physiological pH and stable tonucleophilic attack by glutathione. A major route of their degradationis the hydrolyzation to the corresponding dihydroxy-FA (DiHOME,DiHETrE, DiHODE, DiHETE, and DiHDPE, Fig. 1) by the soluble epoxidehydrolase (sEH) [16,17]. (iv) Non-enzymatic autoxidation of PUFA leadsto hydroperoxy-FA, hydroxy-FA and PG like prostanoids [18,19].Despite a pronounced substrate, regio- and stereospecifity of theoxylipin generating enzymes, the product pattern of the four processesoverlaps. For example, 15-HETE can not only be formed by reticulocyteand epidermis 15-LOX but also by COXs, CYPs and autoxidation[13,14,18,20]. Also human LOX enzymes do not only accept AA assubstrate. Via an intermediate hydroperoxide, 5-LOX conversion ofEPA leads to 5-HEPE [21]. Similarly, platelet type 12-LOX can catalyzethe formation of 12-HEPE from EPA [22]. Reticulocyte type 15-LOXconversion of LA leads to 13-HODE while EPA gives rise to 15-HEPE[23,24]. In an analogous manner it can be assumed that 13-HOTrE is a15-LOX product of ALA.

Prostaglandins and leukotrienes derived from AA by COX andLOX and their biology are among the most investigated endogen-ous mediators, and more than half of the currently sold pharma-ceutical drugs directly target one of these pathways [25]. Thehydroxy-, epoxy- and dihydroxy-FA also show a pronouncedbiological activity on various regulatory pathways. For example5-HETE and other hydroxy-FA play a key role in chemotaxis anddegranulation of neutrophils [26,27], being more active than theirn-3 PUFA analogs [28]. For 5-HEPE an effect on the glucose-dependent insulin secretion is discussed [29]. 12-HETE mediatesthe function of many growth factors and thus seems to be a keyfactor in cancer metastasis [30]. In many studies the bloodconcentration of hydroxy-FA is used as marker for the activity ofa LOX pathway because they are more stable than the initiallyformed hydroperoxides and short lived potent mediators asleukotrienes (5-LOX) and hepoxylins (12-LOX). However, the over-all biological role of many hydroxy-FA is not fully understood. Bycontrast, it is well accepted that epoxy-FA play an important rolein the regulation of vascular and cardiac function, in particular byacting as vasodilators of human coronary arteries [31,32]. More-over, epoxy-FA act anti-inflammatory and analgesic in variousanimal models, while their corresponding dihydroxy-FA show lowor no biological activity [17,33]. Recently, it was shown that thedihydroxy-FA play an important role in hematopoiesis [34], thusnot only the epoxy-FA but the epoxide/diol ratio might be theimportant factor that governs the regulation of cellular functions.

Up to date, only little information about endogenous levels ofhydroxy-, epoxy- and dihydroxy-PUFA in humans is available.

Particularly, only few studies comprehensively examined n-3 andn-6 oxylipin profiles including hydroxy-, epoxy-, and dihydroxy-FAin human blood samples [11,12,35,36]. Shearer et al. measured 46LA, AA, ALA, EPA, and DHA derived oxylipins in plasma aftersaponification of healthy subjects (n¼10) before and after 4 weeksof n-3 PUFA treatment [12]. Gomolka et al. focused on LOXproducts of AA, EPA, and DHA by measuring 29 oxylipins, mainlyhydroxy-FA, in plasma from healthy subjects (n¼5) [35]. Psycho-gios et al. reported levels of 76 oxylipins in human plasma ofwhich 56 were above the limit of quantification for a largecollection of 70 subjects [11]. These studies provide importantinformation on the concentration range of these oxylipins inhuman blood. However, these studies only examined healthysubjects, which were not further physiologically characterized e.g. bytheir blood lipid levels. In order to understand the physiological role ofhydroxy-, epoxy- and dihydroxy-FA, knowledge about the endogen-ous levels in humans and their regulation in response to disease isindispensable. For example, a recent study comprehensively monitor-ing more than 100 oxylipins could show significantly increasedplasma 5-HETE and 12-HETE levels as well as their analogous EPAmetabolites 5-HEPE and 12-HEPE following heart ischemia duringcardiac surgery. Despite the small sample size of only five patients, itcould be shown that these LOX metabolites potentially associatedwith the resolution phase of inflammation were elevated, whereas nochanges were observed in the prostaglandin and leukotriene levels[36]. Aside from such severe medical conditions, metabolic disordersmay impact the endogenous oxylipin patterns. In the rat model oftype-1 diabetes as well as obesity-induced hypertriglyceridemia,significant changes in the oxylipin pattern were observed, whichmay contribute to the pathophysiological states [37,38]. Hypertrigly-ceridemia and hypercholesterolemia (hyperlipidemia) are commonrisk factors for cardiovascular diseases [39] and associated with othermetabolic disorders such as the metabolic syndrome [40]. No data areavailable on if and how hyperlipidemia effects endogenous oxylipinlevels in humans. An altered oxylipin pattern could have strong effectson the physiology and thus for example on the development ofcardiovascular diseases. In order to investigate the impact of combinedhyperlipidemia on oxylipin levels, we compared patterns of 45 free(and non-covalently bound) hydroxy-, epoxy-, and dihydroxy-FA inserum samples of normo- and hyperlipidemic men. Moreover, thedata was correlated to the individual status of the five parent PUFA(LA, AA, ALA, EPA, and DHA) to normalize for interindividual varyingdiets and a potentially changed metabolism of PUFA in the differentgroups of subjects.

2. Experimental procedures

This investigator initiated study was designed and conductedaccording to the principles of the Good Clinical Practice Guidelines

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–29 21

laid down in the Declaration of Helsinki. The study was approvedby the Freiburger ethic committee (Freiburg, Germany). Allincluded subjects gave their written informed consent to takepart in the study.

2.1. Subjects and design

The recruitment of subjects was performed by advertisementsin Hannover, Germany. One hundred and six subjects were pre-selected via telephone interviews according to the followingexclusion criteria: female; body-mass-index 435; smoker; intakeof any corticosteroids, lipid-lowering or anti-inflammatory drugs;diagnosed chronic, cardiovascular or liver diseases; gastrointest-inal disorders; blood coagulation disorders and intake ofcoagulation-inhibiting drugs; renal failure; periodic intake oflaxatives; ingestion of supplements enriched with n-3 PUFA,phytosterols, polyglucosamines, other lipid-binding ingredients.The pre-selected subjects were invited for a screening examinationto collect fasting blood and determine serum lipid levels.Among these subjects, 20 men with mild combined hyperlipide-mia (threshold value: total cholesterol TC4200 mg/dl;LDL4130 mg/dl; TG > 150 mg/dl [41]) and 20 normolipidemicmen (TCo200 mg/dl; LDLo130 mg/dl; TGo150 mg/dl) werechosen (Table 1). The recruitment of the study population andselection of the subjects is shown in Fig. 2 in more detail. Thehyperlipidemic group and the normolipidemic group showed nosignificant differences in age, height and weight. However, thehyperlipidemic group had a slightly higher mean BMI (p¼0.034).

2.2. Sample collection and analysis of clinical parameters

Blood samples were collected in the morning between 6:00and 9:30 a.m. after overnight fasting. The samples were obtainedby venipuncture of an arm vein into 10 ml BD Vacutainer BloodCollection Tubes (article no. 367896, Becton Dickinson, Heidelberg,Germany) containing silicate particles as clot activators. After30 min incubation at room temperature, tubes were centrifugedfor 10 min at 2000� g and serum was transferred into 15 mLfalcon tubes (Becton Dickinson) and immediately frozen andstored at −80 1C upon extraction and LC-MS analysis. Other setsof blood samples collected simultaneously were sent to externallaboratories for the measurement of clinical parameters (Table 1).Serum lipid levels were determined in LADR laboratory, Hann-over; Germany. Erythrocyte membrane PUFA composition was

Table 1Description of the study population.

Parameter Total population (n¼40) Normolipidemmean7SD mean7SD

Age [years] 3978 3878Body height [cm] 18177 18077Body weight [kg] 85.2713.7 81.1712.8Body mass index [kg/m²] 26.173.49 24.973.40Total cholesterol [mg/dl] 223755 185717.7Triacylglycerol [mg/dl] 1887189 100744.6High density lipoprotein [mg/dl] 51.4710.3 56.0710.9Low density lipoprotein [mg/dl] 133737.5 110719.0LDL-C/HDL-C quotient 2.770.9 2.0770.62LA [% of total FA] 15.572.24 15.072.17AA [% of total FA] 14.772.20 15.871.13ALA [% of total FA] 0.3270.23 0.2570.09EPA [% of total FA] 0.9370.35 0.8570.25DHA [% of total FA] 4.1471.06 4.2770.86

All variables were normally distributed.n Significant differences in group means were determined by independent sample t

determined in EDTA stabilized whole blood at Omegametrix,Martinsried, Germany as previously described [42].

2.3. Analysis of oxylipins

Analysis of oxylipin levels in serum was carried out by liquidchromatography–electrospray ionization tandem mass spectro-metry (LC–ESI-MS/MS) as described [25,37,43]. The serum samples(950 mL) were spiked with 20 mL internal standard solution con-taining 50 nM of 2H11-14,15-DiHETrE, 2H4-9-HODE, 2H8-5-HETE,2H11-11(12)-EpETrE in methanol (MeOH). This was followed byaddition of antioxidants (10 mL of a 0.2 mg/mL solution of buty-lated hydroxytoluene and EDTA). In order to extract the freeoxylipins in serum, solid phase extraction (SPE) without subse-quent protein precipitation was performed. It should be noted thatthis procedure may also co-extract oxylipins which are non-covalently bound to proteins. In brief, SPE was carried out asfollows: Cartridges (60 mg waters Oasis-HLB, Waters, Milford, MA)were cleaned with one column volume of ethyl acetate (EA) andone of MeOH. After the serum samples were loaded on precondi-tioned SPE columns, the columns were washed with two columnvolumes of 5/95 MeOH/water acidified with 0.1% acetic acid. Afterdrying the SPE column applying low vacuum (0.2 bar), the ana-lytes were eluted with 500 mL of MeOH and 1500 mL of EA. Theeluted samples were evaporated to dryness and reconstitutedin 50 μL of MeOH containing 100 nM 1-cyclohexyluriedo-3-dodecanoic acid for the calculation of IS recovery. The resultingextract was filtered using Ultrafree-MC-VV polyvinylidene fluoridefilter (pore-size 0.1 μm; Millipore, Bedford, MA), transferred inamber glass vials with insert (Waters) and directly analyzedby LC–ESI-MS/MS. Separation was carried out in 21 min on anAgilent 1200 SL LC system (Palo Alto, CA), utilizing an AgilentZorbax Eclipse Plus C18-reversed phase column (dimensions2.1�150 mm, particle size 1.7 μm) using a gradient of 0.1% aceticacid as solvent A and 80/15/0.1 acetonitrile/methanol/acetic acidas solvent B. The detection was carried out on a 4000 QTRAPinstrument (Applied Biosystems, Foster City, CA) operating innegative ion mode. The oxylipins were monitored in so called“scheduled multiple reaction monitoring” mode with a detectionwindow of 90 s for each compound around the expected retentiontime and a maximal cycle time of 0.6 s. Recovery rates of the IS inserum samples were 80710% for 2H11-14,15-DiHETrE, for 2H4-9-HODE 70710%, 70710% for 2H8-5-HETE and 60710% for 2H11-11(12)-EpETrE and in the same range as previously described[11,25,36]. The monitored oxylipins, their transitions and

ic group (n¼20) Hyperlipidemic group (n¼20) P* (normo vs. hyper)mean7SD

4079 0.33718177 0.79989.3713.6 0.05727.373.26 0.034261753.2 o0.0012757235 0.00447.278.04 0.008160735.7 o0.0013.3770.71 o0.00116.072.26 0.17713.672.46 0.0010.4070.29 0.0461.070.42 0.1874.071.23 0.430

-test.

Fig. 2. Flow chart of recruitment of the human subjects.

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–2922

quantification limits are presented in the supplementary materials(SI, Table S1). Only oxylipins which exceeded the limit of quanti-fication in ≥ 95% of the serum samples were used for data analysis.

2.4. Data analysis and statistics

Results for anthropometrical measures, lipid levels as well aserythrocyte membrane PUFA levels are presented as mean7SD.Serum oxylipin levels are presented as mean7SE. The influence ofthe PUFA status on oxylipin levels was investigated by subdividingstudy participants into two groups with low and high PUFA status.For this purpose, the total study population was divided intotertiles based on PUFA levels in erythrocyte membranes and in thefollowing analysis, groups with low and high PUFA status werecompared. Ratios between epoxides/diols as well as between LOXproducts were calculated on individual level and results are shownas mean7SD. The sample sets were analyzed for their distributionby the Kolmogorov–Smirnov test. A two tailed t-test for indepen-dent samples was used to reveal significant differences betweennormal distributed groups. In cases of skewed distribution, theMann–Whitney U test was used. Statistical significance was gen-erally accepted at p ≤ 0.05. Correlation analysis of the normallydistributed variables was carried out by the Pearson method.Correlation between the variables was accepted with a correlationcoefficient of ≥ 0.5 or ≤ −0.5 and p ≤ 0.05. All statistical analyseswere carried out with SPSS software (Version 20, SPSS Inc.,Chicago IL, USA).

3. Results

Among 48 oxylipins measured, 44 hydroxy-, epoxy- anddihydroxy-FA could be detected and quantified in more than 95%

of the human serum samples (Table 2, SI Table S1). All oxylipinsvaried considerably between subjects. The measured concentra-tions for the analytes ranged over about three orders of magnitudefrom 50 pM for 11,12-DiHETE to 16 nM for 12-HETE. The concen-trations of other hydroxy-FA, ranged from 100 pM to 8 nM. Amongthese analytes, the metabolites of LA dominated and the concen-trations decreased in the order of LA4AA4EPA ≥ ALA. Theepoxy-FA were found in a similar concentration range of200 pM–6.5 nM. Depending on the parent PUFA, the epoxideswere found in the order LA4ALA4AA4DHA4EPA, similar tothe dihydroxy-FA levels, which were detected in a concentrationrange of 50 pM–11 nM. All LA metabolites were found at a similarconcentration of 3–8 nM, whereas the levels of hydroxy-, epoxy-,and dihydroxy-FA derived from the other PUFA greatly varied(Table 2). Aside from the dominating metabolite 12-HETE, theepoxy-FA 11(12)-EpETrE and 14(15)-EpETrE and 5-HETE were themajor serum AA derived oxylipins with concentrations of about1.5 nM. By contrast, the dihydroxy-FA 15,16-DiHODE was the majorALA derived metabolite (11 nM), followed by its correspondingepoxide 15(16)-EpHODE (4 nM). Similarly, the epoxy-/dihydroxy-FA metabolites at position n-3 were the major detected oxylipinsderived from EPA and DHA. For all three n-3 PUFA, the dihydroxy-FA at position n-3 was found at higher concentration than itscorresponding epoxide, whereas the epoxide/diol ratio for allother oxylipins was higher than one (Table 3).

3.1. Comparison of serum oxylipin levels between normo- andhyperlipidemic subjects

The differences in serum oxylipin levels derived from LA, AA,ALA, EPA and DHA between the normo- and hyperlipidemic groupwere minor (Table 2). Generally, a trend toward lower levels ofhydroxy-FA in the hyperlipidemic group was observed, being

Table 2Serum concentration of free oxylipins in total study population and normo- vs. hyperlipidemic groups.

Parent FA Oxylipin Total population (n¼40) Normolipidemic group (n¼20) Hyperlipidemic group (n¼20)

mean [pM]7SE mean [pM]7SE mean [pM]7SE P (normo vs. hyper)

n-6 FA LA Hydroxy-FA 9-HODE 59007510 60007720 57007750 0.781n

13-HODE 77007600 83007820 71007870 0.336n

Epoxy-FA 9(10)-EpOME 57007660 55007910 60007980 0.742n

12(13)-EpOME 65007640 64007900 66007930 0.895n

Dihydroxy-FA 9,10-DiHOME 35007530 38007510 33007940 0.668n

12,13-DiHOME 62007480 74007700 49007530 0.006n

AA Hydroxy-FA 5-HETE 16007130 17007220 15007130 0.557n

8-HETE 350713 370721 330715 0.123nn

9-HETE 17078.4 180710 160713 0.365nn

11-HETE 560773 6907130 420750 0.069n

12-HETE 15,50072,900 22,10075,200 890071900 0.025n

15-HETE 11007110 13007190 950782 0.088n

Epoxy-FA 8(9)-EpETrE 630762 580779 680797 0.426n

11(12)-EpETrE 14007140 13007170 15007230 0.472n

14(15)-EpETrE 13007130 12007140 14007230 0.403n

Dihydroxy-FA 8,9-DiHETrE 360727 270716 440745 0.001n

11,12-DiHETrE 560721 560733 560727 0.939n

14,15-DiHETrE 700722 690736 720726 0.483n

n-3 FA ALA Hydroxy-FA 9-HOTrE 400742 440768 370750 0.423n

13-HOTrE 390737 440765 340737 0.179n

Epoxy-FA 9(10)-EpODE 10007160 10007230 11007220 0.865n

12(13)-EpODE 530773 5207110 5407100 0.913n

15(16)-EpODE 37007380 38007490 37007590 0.941n

Dihydroxy-FA 9,10-DiHODE 120714 160720 85715 0.007n

12,13-DiHODE 140714 180721 100711 0.004n

15,16-DiHODE 11,0007950 11,00071000 990071,600 0.456n

EPA Hydroxy-FA 5-HEPE 330735 260731 400760 0.042n

12-HEPE 460789 6107160 310764 0.093n

15-HEPE 12077.9 11079.0 130713 0.140n

Epoxy-FA 11(12)-EpETE 170721 150727 200732 0.266n

14(15)-EpETE 210725 180732 240738 0.231n

17(18)-EpETE 320727 280735 370740 0.113n

Dihydroxy-FA 11,12-DiHETE 4873.5 4273.3 5476.0 0.089n

14,15-DiHETE 8375.6 7575.7 9279.5 0.131n

17,18-DiHETE 510734 470738.1 560755 0.184n

DHA Epoxy-FA 10(11)-EpDPE 700770 650797 7407100 0.515n

13(14)-EpDPE 470750 440768 490775 0.566n

16(17)-EpDPE 480751 440766 520779 0.413n

19(20)-EpDPE 11007110 9307130 12007170 0.292n

Dihydroxy-FA 10,11-DiHDPE 1807180 140715 220729 0.030n

13,14-DiHDPE 2007200 190716 210719 0.399n

16,17-DiHDPE 2707270 250719 300726 0.153n

19,20-DiHDPE 21007120 21007150 21007180 0.908n

All variables were normally distributed except for 8-HETE, 9-HETE and 12-HETE.n Significant differences in group means were determined by independent sample t-testnn Mann-Whitney U test for skewed distributed variables.

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–29 23

significant only for 12-HETE (Table 2). On the contrary, the 5-HEPEconcentration was significantly elevated compared to the normo-lipidemic group. For all epoxy-FA a trend to higher levels werefound in the hyperlipidemic group, while the levels of thecorresponding diols differed dependent on the parent FA. Thelevels of the LA and ALA derived dihydroxy-FA 12,13-DiHOME,9,10-DiHODE and 12,13-DiHODE were significantly lower, whereasthe concentration of all dihydroxy-FA derived from AA, EPA, andDHA were higher in the hyperlipidemic group (significant for 8,9-DiHETrE and 10,11-DiHDPE). When comparing the epoxide/diolratio on the individual level, no significant differences, except forthe ALA derived oxylipins were observed (Table 3). However,compared to the normolipidemic group, hyperlipidemic menshowed a trend to higher epoxide/diol ratios for all monitoredoxylipins, except for 10(11)-EpDPE:10,11-DiHDPE and 8(9)-EpE-TrE:8,9-DiHETrE.

Post-hoc tests for equivalence between the oxylipin pattern ofnormo- and hyperlipidemic men were carried out by evaluating

the 90% confidence intervals (CI) of the mean serum oxylipin leveldifferences of normo- and hyperlipidemic groups from meanlevels of the total study population. Only levels of 11,12-DiHETrEand 14,15-DiHETrE were considered as statistically equivalentsince the upper and lower CI of the mean differences were withina tolerable variance of 720% (SI, Table S2).

Other covariates such as BMI or TG/HDL ratio, a surrogatemarker for insulin resistance [44], were checked to possiblyinfluence oxylipin levels. Men with a high BMI (428) comparedto the group with a low BMI (o25) showed no significantdifferences in the oxylipin pattern. Similarly, no consistent shiftsin the oxylipin pattern were found when comparing groups withhigh and low (43.5; o2.2) TG/HDL ratio (SI Fig. S2–S3).

3.2. Correlation of serum oxylipins to their parent PUFA

Levels of free and non-covalently bound oxylipins in serumwere compared with the relative PUFA composition in erythrocyte

Table 3Serum Epoxide/Diol ratios. The serum concentration of epoxy-FA was divided by the concentration of the corresponding dihydroxy-FA on individual level. The mean and SDof the ratio is shown for the total study population and normo- vs. hyperlipidemic groups.

Parent FA Epoxide:Diol Total population (n¼40) Normolipidemic group (n¼20) Hyperlipidemic group (n¼20) P* (normo vs. hyper)

mean7SD mean7SD mean7SD

LA 9(10)-EpOME: 9,10-DiHOME 2.5470.43 2.0270.49 3.0670.69 0.22612(13)-EpOME: 12,13-DiHOME 1.2370.14 0.9770.16 1.4870.21 0.063

AA 8(9)-EpETrE: 8,9-DiHETrE 1.9470.19 2.1970.28 1.7070.25 0.19811(12)-EpETrE: 11,12-DiHETrE 2.5070.25 2.2870.27 2.7170.43 0.40614(15)-EpETrE: 14,15-DiHETrE 1.8870.21 1.7770.21 1.9970.36 0.596

ALA 9(10)-EpODE: 9,10-DiHODE 12.271.98 8.0572.08 16.373.15 0.03512(13)-EpODE: 12,13-DiHODE 4.3070.55 3.1470.51 5.4670.92 0.03515(16)-EpODE: 15,16-DiHODE 0.3770.02 0.3470.03 0.4070.04 0.226

EPA 11(12)-EpETE: 11,12-DiHETE 3.8970.44 3.5770.53 4.2270.72 0.47514(15)-EpETE: 14,15-DiHETE 2.6970.32 2.3870.35 3.070.54 0.34217(18)-EpETE: 17,18-DiHETE 0.6870.06 0.6270.07 0.7570.10 0.304

DHA 10(11)-EpDPE: 10,11-DiHDPE 4.7270.52 5.0770.75 4.3770.73 0.50613(14)-EpDPE: 13,14-DiHDPE 2.5170.28 2.3770.35 2.6470.43 0.62316(17)-EpDPE: 16,17-DiHDPE 1.8970.22 1.8270.29 1.9670.33 0.75619(20)-EpDPE: 19,20-DiHDPE 0.5670.07 0.4670.06 0.6570.13 0.164

All variables were normally distributed.n Significant differences in group means were determined by independent sample t-test.

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–2924

membranes, a robust measure for the PUFA status of an individual[42]. In erythrocyte membranes, the n-6 PUFA LA and AA domi-nated followed by the n-3 PUFA DHA, EPA and ALA (Table 1),which strongly varied among the subjects (e.g. RSD of EPA 40%).No significant differences in the PUFA composition of erythrocytemembranes were observed for LA, EPA and DHA levels, whereas inhyperlipidemic men AA levels were 7% lower and ALA levels were38% higher compared to the normolipidemic group (Table 1).

In order to investigate how parent PUFA level affects theformation of their corresponding oxylipins, the total study popula-tion was divided into two groups based on a high and lowpercentage of each PUFA in the erythrocyte membranes. As shownin Fig. 3, the serum oxylipin levels were consistently higher with ahigher status of their precursor PUFA. However, for the high vs. thelow LA, AA, ALA, and DHA status groups, a significant differencewas only observed for 12-HETE, 15-HETE and 15(16)-EpODE(Fig. 3). By contrast, all EPA derived oxylipins were significantlyhigher in the high EPA status group compared to the low EPAstatus group, except for 12-HEPE. For 5-HEPE, 15-HEPE, 11,12-DiHETE and 14,15-DiHETE a significant correlation (r ≥ 0.5) of theEPA status and the oxylipin concentration was found for the totalstudy population (Table 4). Additionally, a positive correlationbetween ALA levels in erythrocyte membranes and the ALAderived oxylipins 9-HOTrE, 13-HOTrE, 15(16)-EpODE, 9,10-DiHODE, 12,13-DiHODE and 15,16-DiHODE was observed in thehyperlipidemic group (Table 4). No correlation, neither in the totalstudy population, nor in the normolipidemic collective, was foundfor other parent PUFA (LA, AA, ALA and DHA) and their corre-sponding oxylipins. Similarly, no significant changes in the AAoxylipin levels between the groups with high and low ALA, EPAand DHA were observed (SI Fig. S1).

The ratios of AA derived HETE to the corresponding hydroxy-FAof other PUFA were correlated with the status of the precursorPUFA to evaluate changes in the LOX product pattern. For 12-LOX,the ratio of 12-HETE to 12-HEPE was compared between the groupwith high EPA vs. low EPA levels (Fig. 4). For 5-LOX, the ratio of 5-HETE to 5-HEPE was used. Finally for 15-LOX, the ratio of 15-HETEto 13-HODE, 13-HOTrE and 15-HEPE were compared to the statusof LA, ALA and EPA, respectively. For all PUFA and all LOX path-ways, the ratios were significantly decreased in the high PUFAgroups, except for 15-HETE:15-HEPE (Fig. 4). The ratio between5-HETE:5-HEPE and 12-HETE:12-HEPE negatively correlated withthe relative erythrocyte membrane EPA content in the total

population as well as in the normo- and hyperlipidemic groups(SI Table S3). Additionally, a correlation in the ratio of 15-HETE:13-HOTrE, a potential 15-LOX product of ALA, was observed in thenormo- and hyperlipidemic group. Compared to the normolipi-demic group, the hyperlipidemic group showed a trend towardlower HETE:hydroxy-PUFA ratios (SI Table S4). With a significantlower ratio of 5-HETE:5-HEPE, 12-HETE:12-HEPE and 15-HETE:15-HEPE, the product ratios of all LOX pathways were changed towardthe EPA product.

4. Discussion

While prostanoids and leukotrienes are well characterized,little information is available on the endogenous levels andbiological roles of n-6 and n-3 PUFA derived hydroxy-, epoxy-,and dihydroxy-FA in humans. Several studies demonstrate thatthese oxylipins act as potent modulators of biological processeswith opposing effects on inflammation and other biological func-tions. Thus, knowledge about their endogenous levels in humanblood and changes in different physiological and pathophysiolo-gical states is of pivotal importance in biology and medicine.

In this study, we present quantitative data on 44 oxylipinserum levels from a cohort of 40 subjects, which had normaland elevated serum triglyceride and cholesterol levels. We foundthat oxylipin concentrations varied strongly among subjects, con-sistent with all previous studies monitoring oxylipins in humanblood samples [11,12,25,36]. The observed concentrations ofhydroxy-, epoxy- and dihydroxy-FA in serum were well in linewith previously described levels of these oxylipins in plasma [11].The oxylipin concentration in plasma can be strongly affected bythe anticoagulant used [45]. Serum is a standard matrix which isused for the analysis of many biochemical parameters in blood andcan be obtained in highly reproducible quality. The deviationbetween the mean concentration in 40 human serum samplesanalyzed in this study compared to 70 plasma samples reported byPsychogios et al. [11] was lower than 50% and within the standarddeviations for 17 out of 33 oxylipins (SI Table S5). Giordano et al.described in the plasma of subjects (n¼14) almost the sameconcentration (deviation ≤ 3%) of the three AA deriveddihydroxy-FA as found in serum [46], while their reported EpETrElevels in plasma were 2–7 fold lower than the observed levels inserum as well as in plasma described by Psychogios et al. [11].

Fig. 3. Dependence of serum oxylipin levels on parent fatty acid status. Oxylipin levels derived from (A) linoleic acid (LA, 18:2 n-6), (B) arachidonic acid (AA, 20:4 n-6),(C) alpha-linolenic acid (ALA, 18:3 n-3), (D) eicosapentaenoic acid (EPA, 20:5 n-3) and (E) docosahexaenoic acid (DHA, 22:6 n-3) were compared in tertiles of the whole studypopulation with low and high PUFA percentage in erythrocyte membranes. All results are shown as the mean7SE. Significant differences were determined by independentsample t-test for normal distributed variables. 1Mann-Whitney U test was used for skewed distributed variables.

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–29 25

Comparing the human plasma levels of five oxylipins measured byGomolka et al., 15-HETE, 12-HEPE were found almost at the sameconcentration (800 pM and 380 pM), whereas the concentrationsof 5-HETE and 5-HEPE were 3–5 fold lower as in our study. Thehuman plasma oxylipin levels, observed in subjects (n¼5) shortlybefore cardiac surgery, reported by Strassburg et al., were overall2–5 times higher than concentrations obtained in serum [36]. Forexample, the mean concentration in plasma of 5-HEPE was foundto be around 1 nM, compared to our serum analysis (0.3 nM), andthe plasma analyses of Gomolka (0.06 nM) and Psychogios(0.2 nM) [11,35,36]. The comparison of oxylipin levels to earlierstudies demonstrates that the levels of most hydroxy-, epoxy-, anddihydroxy-FA in serum are comparable to those of plasma samplesand the deviation between serum and plasma levels is for mostanalytes not higher than those from two independent plasmaanalyses. Nevertheless, a dramatic formation of COX-1 metabolites,particularly thromboxanes, has to be taken into account whenanalyzing serum because of the blood coagulation process invol-ving the platelet activation and degranulation [11]. Similarly, wefound markedly higher 12-HETE levels of 1673 nM in serumcompared to 2–4 nM described in plasma [11,36], probably causedby platelet 12-LOX activated by blood coagulation.

As expected, the serum levels of hydroxy-, epoxy- anddihydroxy-FA were considerably lower, compared to plasma levels

after liberation of ester-bound oxylipins. Shearer et al. reportedplasma levels of oxylipins after saponification for a group ofsubjects (n¼10) before and after treatment with n-3 PUFA cover-ing the same set of hydroxy-, epoxy- and dihydroxy-FA [12].A comparison of the plasma levels before n-3 PUFA treatmentunveiled that the difference between the serum levels and thesum of free and liberated ester bound oxylipins in plasma,reported by Shearer et al., are strongly dependent on the oxylipin(SI Table S4). For example, the autoxidation product of AA, 9-HETE,is found at about 200 fold higher concentration after saponifica-tion indicating that this metabolite is predominantly esterified inplasma, and/or formed during saponification. By contrast, the(free) serum levels of PUFA dihydroxylated at position n-6 andn-3 (12,13-DiHOME, 14,15-DiHETrE, 15,16-DiHODE, 19,20-DiHDPE)were observed in the same range as in plasma after saponification.This indicates that these more polar compounds predominantlycirculate as free oxylipins. The corresponding epoxides, as well astheir constitution isomers (e.g. 9,10-DiHOME, 8,9-DiHETrE) seemto be predominantly esterified to more than 90% with a 10 foldhigher concentration found in plasma after conjugate cleavage (SITable 4). This is consistent with the observation that the epoxide/diol ratios of the serum levels of n-3 epoxy-FA are o1, while beingon the side of the epoxides for the n-6 and n-9 and all otherepoxy-FA (Table 3). The n-3 epoxy-FA are hydrolyzed by sEH at low

Table 4Correlation between serum oxylipin concentration and the percentage of theparent PUFA (LA, AA, ALA, EPA, and DHA) in erythrocyte membranes. Only thoseoxylipins are presented, where a significant correlation (correlation coefficient≥ 0.5 and p ≤ 0.05) could be observed. The analysis was carried out in the totalstudy population and independently in the in normo- and hyperlipidemicsubgroups.

Total population (n¼40) EPAHydroxy-FA 5-HEPE r 0.572nnn

15-HEPE r 0.524nn

Dihydroxy-FA 11,12-DiHETE r 0.517nn

14,15-DiHETE r 0.523nn

Normolipidemic group (n¼20) EPAHydroxy-FA 5-HEPE r 0.650nn

Epoxy-FA 11(12)-EpETE r 0.556n

14(15)-EpETE r 0.570n

17(18)-EpETE r 0.632nn

Hyperlipidemic group (n¼20) ALAHydroxy-FA 9-HOTrE r 0.740nnn

13-HOTrE r 0.757nnn

Epoxy-FA 15(16)-EpODE r 0.546n

Dihydroxy-FA 9,10-DiHODE r 0.846nnn

12,13-DiHODE r 0.749nnn

15,16-DiHODE r 0.517n

EPAHydroxy-FA 5-HEPE r 0.512n

15-HEPE r 0.511n

Dihydroxy-FA 14,15-DiHETE r 0.525n

17,18-DiHETE r 0.524n

All variables were normally distributed and correlation analysis was carried outaccording to Pearson.

n po0.05.nn po0.005.nnn po0.001.

Fig. 4. Serum hydroxy fatty acid ratios depending on parent fatty acid status. Ratios betwHETE:12-HEPE), and 15-LOX products (C—15-HETE:15-HEPE; 15-HETE:13-HODE; 15-HETand high PUFA status. All results are shown as the mean7SE of group measurement. Sig

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–2926

or medium rate compared to their regioisomers [17]. Thus, theobserved differences in the epoxide/diol ratios within the regio-isomers of the n-3 FA cannot be explained by differences in theirdegradation and may be caused by differences in their esterifica-tion. Oxylipin specific esterification seems to be a key factor in theregulation of the serum concentration of hydroxy-, epoxy- andhydroxy-FA, which warrants further research.

Strong differences in the levels of the different regioisomers ofthe epoxy-FA and corresponding dihydroxy-FA of each PUFA wereobserved. Particularly, it is striking that for ALA, EPA, and DHA then-3 epoxy-FA 15(16)-EpODE, 17(18)-EpETE and 19(20)-EpDPE andtheir corresponding dihydroxy-FA are present in serum in muchhigher levels than their n-6 and n-9 analogs. By contrast, the AA isalmost equally epoxidized at the n-6 and n-9 position (Table 2).This observation is consistent with the product specificity of theCYP2J and CYP2C enzymes [9,47], indicating that these are the CYPfamilies which form the epoxy-FA present in serum.

4.1. Correlation of serum oxylipins to their parent PUFA

The overall serum concentration of the different PUFA oxylipinsdecreased in the order LA4AA¼ALA ≥ DHA4EPA and correlateswell with the abundance of the PUFA in the human organism(Tables 1, 2). Subjects with a high PUFA status, determined aspercentage of the PUFA in the erythrocyte membrane, show a trendto higher serum levels of hydroxy-, epoxy-, dihydroxy-FA derived fromthis PUFA (Fig. 3). Hence, it can be concluded as a first approximationthat the endogenous formation of these oxylipins is primarily depen-dent on the availability of the substrate. However, no significant

een 5-lipoxygenase (5-LOX) products (A—5-HETE:5-HEPE), 12-LOX products (B—12-E:13-HOTrE) were compared between tertiles of the total study population with lownificant differences in group means were determined by independent sample t-test.

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–29 27

correlation between either LA or AA status and deriving oxylipinserum levels could be observed, except for 15- and 12-HETE (Fig. 3).Similarly, high endogenous levels of DHA did not lead to significantlyhigher levels of its epoxy- and dihydroxy-metabolites (Fig. 3). Thus, forthese three abundant PUFA, substrate availability seems to be only ofmoderate importance in the regulation of oxylipin formation.

By contrast, the formation of EPA derived oxylipins seems to bestrongly dependent on the availability of the substrate. Subjectswith a high EPA status showed significant higher levels of all EPAderived oxylipins except for 12-HEPE, which varied stronglyamong the subjects (Fig. 3). Given that the synthesis rate of EPAfrom ALA in humans is very low [3], the EPA status largely relies onthe diet. Numerous nutritional intervention studies have shownthat EPA levels in plasma and erythrocyte membranes increasesubstantially by supplementation with EPA- and DHA-rich fish oils[48]. Taken together, one can conclude that the EPA oxylipin levelsare directly affected by the diet. This hypothesis warrants furtherinvestigation particularly because several EPA oxylipins are highlyactive lipid mediators, e.g. the resolving anti-inflammatory resol-vins of the E series [49]. Shearer et al. already demonstrated thattreating healthy subjects (n¼10) with n-3 PUFA ethyl esters (4 g/day for 4 weeks) significantly increased the plasma levels ofseveral esterified and free EPA derived oxylipins (5-, 12-, and 15-HEPE, 14(15)-, and 17(18)-EpETE; 14,15-, and 17,18-DiHETE) [12].Moreover, Keenan et al. showed that the response of subjects on n-3-PUFA treatment depends on the basal n-3 PUFA status [50].However, no data about the influence of fish oil supplementationor other dietary interventions on the free oxylipin levels in blood(plasma or serum) have been reported so far.

The beneficial health effects of long chain n-3 PUFA arebelieved to be partly mediated by a replacement of AA at theactive side of LOXs and COXs, thus reducing the eicosanoidformation. In our study, we did not observe any significantdifferences in the AA oxylipins between subjects with a high anda low status of long chain n-3 PUFA (SI Fig. S1). However, whenregarding the ratio of LOX products of different PUFA based on thehydroxy-FA as pathway markers, significant changes wereobserved depending on the PUFA status of the subjects.A significant lower ratio of HETE and the correspondinghydroxy-PUFA product of 5-LOX, 12-LOX and 15-LOX was observedin subjects with a high status of LA, ALA, and EPA (Fig. 4, SI,Fig. S3). Particularly, the negative correlation of the HETE/HEPEratio with the EPA status (SI Table S3) shows that the LOX productpatterns are significantly influenced by the availability of substrateand thus could be modified with the diet. However, it should benoted that we observed LOX activity solely on the hydroxy-FA,being a product of the initially formed hydroperoxide. The lattercan be converted to different oxylipins aside from a reduction tothe hydroxy-FA. In order to evaluate the alteration of the LOXproduct pattern, a larger oxylipin pattern including leukotrienesshould be investigated in the future.

4.2. Comparison of oxylipin levels between normo- andhyperlipidemic groups

Similar to all previous studies which monitored oxylipins inplasma [11,12,36], the serum oxylipin levels, varied considerablyamong the human subjects. Aside from varying PUFA intake viathe diet, differences in lipid metabolism, such as hyperlipidemia,could be an underlying cause for this variation. However, whenanalyzing the oxylipin levels dependent on the triglyceride andcholesterol status, the same variation of hydroxy-, epoxy- anddihydroxy-FA was observed within the normolipidemic and hyper-lipidemic group. Only levels of six oxylipins were significantlydifferent between the groups, slightly exceeding the theoreticalnumber of three incorrectly identified differences (false discovery

rate of about 60 variables and an alpha error of 5%). Significantdifferences were observed for 12-HETE, 5-HEPE and fourdihydroxy-FA of different PUFA (Table 2). Since the major portionof 12-HETE is probably formed during serum preparation byplatelet 12-LOX, the difference is most likely caused by differencesin the blood coagulation between the groups and does not reflectan altered formation in vivo. The major route of dihydroxy-FAformation is the hydrolysis of epoxy-FA by sEH (Fig. 1). In order toanalyze the altered levels of dihydroxy-FA, levels were thereforecompared to those of the corresponding epoxy-FA. For all epoxy-/dihydroxy-FA of the different PUFA, except for 8(9)EpETrE and 10(11)EpDPE, a trend to higher epoxide/diol ratio in the hyperlipi-demic group was observed (Table 3). A high epoxide/diol ratiocorrelates with anti-inflammatory, vasodilatory and analgesiceffects, caused by inhibition of sEH using sEH inhibitors (sEHi)[51]. However, for most of the epoxide/diol ratios of the differentPUFAs the slightly higher epoxide/diol ratios in the hyperlipidemicgroup were based on higher epoxy-FA levels and thus do not implyan altered sEH activity in hyperlipidemic men (Table 2).

Interestingly, the strongest difference observed in hyperlipi-demic men, was the increased ALA status (Table 1). Taking theinfluence of PUFA status on the oxylipin formation into account,it is no surprise that this was accompanied by changes in theALA derived oxylipin levels. Interestingly, lower levels of 9(10)-DiHODE and 12,13(DiHODE) were found, leading to a significantincrease in the epoxide/diol ratio for these oxylipins in thehyperlipidemic group (Tables 1, 2). Moreover, the hyperlipi-demic group also showed a significant correlation between ALAstatus and ALA derived oxylipins (Table 4), whereas no suchcorrelation was found in the normolipidemic group. However, itseems unlikely that the observed changes in oxylipin metabo-lism are correlated to hyperlipidemia. Rather these changesseem to be caused by the elevated ALA status of the hyperlipi-demic subjects, being a confounding factor of the investigatedcollective of subjects. Similarly, the decreased HETE/HEPE ratiosin the hyperlipidemic group are rather correlated with asignificant decreased ALA/EPA ratio, than caused by elevatedblood lipid levels.

5. Conclusion

The analysis of serum samples for 44 oxylipins unveiled noapparent differences between normo- and hyperlipidemic sub-jects. Despite altered blood lipid levels, only very minor or nochanges in the serum concentration of hydroxy-, epoxy- anddihydroxy-FA were observed. This indicates that the formationand degradation of this class of lipid metabolites are undertight endogenous control, which is not significantly altered bymild metabolic diseases. However, the interindividual variancein the serum oxylipin concentration was high, which mightmask effects of hyperlipidemia on blood oxylipin levels. In thewhole collective of subjects, we observed a strong correlationbetween the EPA content in the erythrocyte membrane and theserum concentration of EPA derived hydroxy-, epoxy-, anddihydroxy-FA. Given that the synthesis of EPA from ALA in thehuman is low, this suggests, that oxylipin levels can be directlyinfluenced by the diet.

Acknowledgments

This study was supported by a Marie Curie Career IntegrationGrant to NHS, a Grant of the German Federal Ministry of Educationand Research (BMBF) to AH, a Kekule ́ Ph.D. fellowship of the Fondsder Chemischen Industrie to IW and grants of US National

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–2928

Institutes of Health (NIH), Environmental Health (NIEHS,P42ES004699 and R01ES002710) and Diabetes and Digestive andKidney Diseases (NIDDK, U24DK097154) and the West CoastCentral Comprehensive Metabolomics Resource Core (WC3MRC)to BDH. BDH is a George and Judy Marcus Senior Fellow of theAmerican Asthma Association.

Appendix A. Supplementary materials

Supplementary materials associated with this article can befound in the online version at http://dx.doi.org/10.1016/j.plefa.2013.04.001.

References

[1] G.C. Burdge, P.C. Calder, Conversion of α-linolenic acid to longer-chainpolyunsaturated fatty acids in human adults, Reprod. Nutr. Dev. 45 (2005)581–597.

[2] J. Whelan, The health implications of changing linoleic acid intakes, ProstagLeukotr Ess 79 (2008) 165–167.

[3] J.T. Brenna, N. Salem Jr., A.J. Sinclair, S.C. Cunnane, α-Linolenic acid supple-mentation and conversion to n-3 long-chain polyunsaturated fatty acids inhumans, Prostaglandins Leukotrienes Essent. Fatty Acids 80 (2009) 85–91.

[4] M. Plourde, S.C. Cunnane, Extremely limited synthesis of long chain poly-unsaturates in adults: implications for their dietary essentiality and use assupplements, Appl. Physiol. Nutr. Metabol.¼Physiologie appliquee, nutritionet metabolisme 32 (2007) 619–634.

[5] J. Cabo, R. Alonso, P. Mata, Omega-3 fatty acids and blood pressure, Br. J. Nutr.107 (Suppl. 2) (2012) S195–200.

[6] J. Delgado-Lista, P. Perez-Martinez, J. Lopez-Miranda, F. Perez-Jimenez, Longchain omega-3 fatty acids and cardiovascular disease: a systematic review, Br.J. Nutr. 107 (Suppl. 2) (2012) S201–213.

[7] R.M. Ortega, E. Rodriguez-Rodriguez, A.M. Lopez-Sobaler, Effects of omega3 fatty acids supplementation in behavior and non-neurodegenerative neu-ropsychiatric disorders, Br. J. Nutr. 107 (Suppl. 2) (2012) S261–270.

[8] O.D. Rangel-Huerta, C.M. Aguilera, M.D. Mesa, A. Gil, Omega-3 long-chainpolyunsaturated fatty acids supplementation on inflammatory biomarkers: asystematic review of randomised clinical trials, Br. J. Nutr. 107 (Suppl. 2) (2012)S159–170.

[9] C. Arnold, A. Konkel, R. Fischer, W.H. Schunck, Cytochrome P450-dependentmetabolism of omega-6 and omega-3 long-chain polyunsaturated fatty acids,Pharmacol. Rep. 62 (2010) 536–547.

[10] M.W. Buczynski, D.S. Dumlao, E.A. Dennis, Thematic review series: proteomics.An integrated omics analysis of eicosanoid biology, J. Lipid Res. 50 (2009)1015–1038.

[11] N. Psychogios, D.D. Hau, J. Peng, A.C. Guo, R. Mandal, S. Bouatra, I. Sinelnikov,R. Krishnamurthy, R. Eisner, B. Gautam, N. Young, J.G. Xia, C. Knox, E. Dong,P. Huang, Z. Hollander, T.L. Pedersen, S.R. Smith, F. Bamforth, R. Greiner,B. McManus, J.W. Newman, T. Goodfriend, D.S. Wishart, The human serummetabolome, Plos One 6 (2011), e16957, http://dx.doi.org/10.1371/journal.pone.0016957.

[12] G.C. Shearer, W.S. Harris, T.L. Pedersen, J.W. Newman, Detection of omega-3oxylipins in human plasma and response to treatment with omega-3 acidethyl esters, J. Lipid Res. 51 (2010) 2074–2081.

[13] A.R. Brash, Lipoxygenases: occurrence, functions, catalysis, and acquisition ofsubstrate, J. Biol. Chem. 274 (1999) 23679–23682.

[14] J. Bylund, T. Kunz, K. Valmsen, E.H. Oliw, Cytochromes P450 with bisallylichydroxylation activity on arachidonic and linoleic acids studied with humanrecombinant enzymes and with human and rat liver microsomes, J. Pharma-col. Exp. Ther. 284 (1998) 51–60.

[15] D.L. Kroetz, D.C. Zeldin, Cytochrome P450 pathways of arachidonic acidmetabolism, Curr. Opinion Lipidol. 13 (2002) 273–283.

[16] C. Morisseau, B.D. Hammock, Epoxide hydrolases: mechanisms, inhibitordesigns, and biological roles, Annu. Rev. Pharmacol. Toxicol. 45 (2005)311–333.

[17] C. Morisseau, B. Inceoglu, K. Schmelzer, H.J. Tsai, S.L. Jinks, C.M. Hegedus, B.D. Hammock, Naturally occurring monoepoxides of eicosapentaenoic acid anddocosahexaenoic acid are bioactive antihyperalgesic lipids, J. Lipid Res. 51(2010) 3481–3490.

[18] U. Garscha, T. Nilsson, E.H. Oliw, Enantiomeric separation and analysis ofunsaturated hydroperoxy fatty acids by chiral column chromatography-massspectrometry, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 872 (2008)90–98.

[19] E. Niki, Lipid peroxidation products as oxidative stress biomarkers, Biofactors34 (2008) 171–180.

[20] S. Mulugeta, T. Suzuki, N.T. Hernandez, M. Griesser, W.E. Boeglin, C. Schneider,Identification and absolute configuration of dihydroxy-arachidonic acidsformed by oxygenation of 5S-HETE by native and aspirin-acetylated COX-2,J. Lipid Res. 51 (2010) 575–585.

[21] T.H. Lee, K.F. Austen, Arachidonic acid metabolism by the 5-lipoxygenasepathway, and the effects of alternative dietary fatty acids, Adv. Immunol. 39(1986) 145–175.

[22] J. Yeung, M. Holinstat, 12-lipoxygenase: a potential target for novel anti-platelet therapeutics, Cardiovasc. Hematol. Agents Med. Chem. 9 (2011)154–164.

[23] A.W. Ford-Hutchinson, M. Gresser, R.N. Young, 5-Lipoxygenase, Annu. Rev.Biochem. 63 (1994) 383–417.

[24] K. Vang, V.A. Ziboh, 15-lipoxygenase metabolites of gamma-linolenic acid/eicosapentaenoic acid suppress growth and arachidonic acid metabolism inhuman prostatic adenocarcinoma cells: possible implications of dietary fattyacids, Prostaglandins Leukotrienes Essent. Fatty Acids 72 (2005) 363–372.

[25] J. Yang, K. Schmelzer, K. Georgi, B.D. Hammock, Quantitative profiling methodfor oxylipin metabolome by liquid chromatography electrospray ionizationtandem mass spectrometry, Anal. Chem. 81 (2009) 8085–8093.

[26] A.A. Spector, J.A. Gordon, S.A. Moore, Hydroxyeicosatetraenoic acids (HETEs),Prog. Lipid Res. 27 (1988) 271–323.

[27] W.F. Stenson, C.W. Parker, Prostaglandins, macrophages, and immunity, J.Immunol. 125 (1980) 1–5.

[28] J.R. Heidel, S.M. Taylor, W.W. Laegreid, R.M. Silflow, H.D. Liggitt, R.W. Leid, Invivo chemotaxis of bovine neutrophils induced by 5-lipoxygenase metabolitesof arachidonic and eicosapentaenoic acid, Am. J. Pathol. 134 (1989) 671–676.

[29] R. Kogure, K. Toyama, S. Hiyamuta, I. Kojima, S. Takeda, 5-Hydroxy-eicosapentaenoic acid is an endogenous GPR119 agonist and enhancesglucose-dependent insulin secretion, Biochem. Biophys. Res. Commun. 416(2011) 58–63.

[30] K.V. Honn, D.G. Tang, X. Gao, I.A. Butovich, B. Liu, J. Timar, W. Hagmann, 12-lipoxygenases and 12(S)-HETE: role in cancer metastasis, Cancer MetastasisRev. 13 (1994) 365–396.

[31] W.B. Campbell, J.R. Falck, Arachidonic acid metabolites as endothelium-derived hyperpolarizing factors, Hypertension 49 (2007) 590–596.

[32] I. Fleming, Vascular cytochrome P450 enzymes: physiology and pathophysiol-ogy, Trends Cardiovasc. Med. 18 (2008) 20–25.

[33] S.J. Thomson, A. Askari, D. Bishop-Bailey, Anti-inflammatory effects of epox-yeicosatrienoic acids, Int. J. Vasc. Med. 2012 (2012) 605101.

[34] T. Fromel, B. Jungblut, J. Hu, C. Trouvain, E. Barbosa-Sicard, R. Popp, S. Liebner,S. Dimmeler, B.D. Hammock, I. Fleming, Soluble epoxide hydrolase regulateshematopoietic progenitor cell function via generation of fatty acid diols, Proc.Natl. Acad. Sci. USA 109 (2012) 9995–10000.

[35] B. Gomolka, E. Siegert, K. Blossey, W.H. Schunck, M. Rothe, K.H. Weylandt,Analysis of omega-3 and omega-6 fatty acid-derived lipid metabolite forma-tion in human and mouse blood samples, Prostaglandins Lipid Mediators 94(2011) 81–87.

[36] K. Strassburg, A.M. Huijbrechts, K.A. Kortekaas, J.H. Lindeman, T.L. Pedersen,A. Dane, R. Berger, A. Brenkman, T. Hankemeier, J. van Duynhoven,E. Kalkhoven, J.W. Newman, R.J. Vreeken, Quantitative profiling of oxylipinsthrough comprehensive LC–MS/MS analysis: application in cardiac surgery,Anal. Bioanal. Chem. 404 (2012) 1413–1426.

[37] B. Inceoglu, K.M. Wagner, J. Yang, A. Bettaieb, N.H. Schebb, S.H. Hwang,C. Morisseau, F.G. Haj, B.D. Hammock, Acute augmentation of epoxygenatedfatty acid levels rapidly reduces pain-related behavior in a rat model of type Idiabetes, Proc. Natl. Acad. Sci. USA 109 (2012) 11390–11395.

[38] G.C. Shearer, J.W. Newman, Lipoprotein lipase releases esterified oxylipinsfrom very low-density lipoproteins, Prostaglandins Leukotrienes Essent. FattyAcids 79 (2008) 215–222.

[39] J. Poss, F. Custodis, C. Werner, O. Weingartner, M. Bohm, U. Laufs, Cardiovas-cular disease and dyslipidemia: beyond LDL, Curr. Pharm. Des. 17 (2011)861–870.

[40] W. Gade, J. Schmit, M. Collins, J. Gade, Beyond obesity: the diagnosis andpathophysiology of metabolic syndrome, Clin. Lab. Sci. 23 (2010) 51–61.

[41] S. Grundy, D. Becker, L. Clark, R. Cooper, M. Denke, J. Howard, D. Hunninghake,R. Illingworth, R. Luepker, P. McBride, J. McKenney, R. Pasternak, N. Stone, L.V.Horn, Third Report of the National Cholesterol Education Program (NCEP)Expert Panel on Detection, Evaluation, and Treatment of High Blood Choles-terol in Adults (Adult Treatment Panel III) National Cholesterol EducationProgram National Heart, Lung, and Blood NIH Publication no. 01-3670 2001,2002.

[42] W.S. Harris, C. Von Schacky, The Omega-3 Index: a new risk factor for deathfrom coronary heart disease? Prev. Med. 39 (2004) 212–220.

[43] N.H. Schebb, B. Inceoglu, C. Morisseau, K.C. Ahn, S.J. Gee, B.D. Hammock,Investigation of human exposure to triclocarban after showering, and pre-liminary evaluation of its biological effects, Environ. Sci. Technol. 45 (2011)3109–3115.

[44] T. McLaughlin, F. Abbasi, K. Cheal, J. Chu, C. Lamendola, G. Reaven, Use ofmetabolic markers to identify overweight individuals who are insulin resis-tant, Ann. Intern. Med. 139 (2003) 802–809.

[45] T.L. Goodfriend, T.L. Pedersen, R.J. Grekin, B.D. Hammock, D.L. Ball, A. Vollmer,Heparin, lipoproteins, and oxygenated fatty acids in blood: a cautionary note,Prostaglandins Leukotrienes Essent. Fatty Acids 77 (2007) 363–366.

[46] R.M. Giordano, J.W. Newman, T.L. Pedersen, M.I. Ramos, C.L. Stebbins, Effects ofdynamic exercise on plasma arachidonic Acid epoxides and diols in humanvolunteers, Int. J. Sport Nutr. Exercise Metab. 21 (2011) 471–479.

[47] C. Arnold, M. Markovic, K. Blossey, G. Wallukat, R. Fischer, R. Dechend,A. Konkel, C. von Schacky, F.C. Luft, D.N. Muller, M. Rothe, W.H. Schunck,Arachidonic acid-metabolizing cytochrome P450 enzymes are targets of{omega}-3 fatty acids, J. Biol. Chem. 285 (2010) 32720–32733.

J.P. Schuchardt et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 89 (2013) 19–29 29

[48] J. Neubronner, J.P. Schuchardt, G. Kressel, M. Merkel, C. von-Schacky, A. Hahn,Enhanced increase of omega-3 index in response to long-term n-3 fatty acidsupplementation from triacylglycerides versus ethyl esters, Eur. J. Clin. Nutr.65 (2012) 247–254.

[49] M.J. Zhang, M. Spite, Resolvins: anti-inflammatory and proresolving mediatorsderived from omega-3 polyunsaturated fatty acids, Ann. Rev. Nutr. 32 (2012)203–227.

[50] A.H. Keenan, T.L. Pedersen, K. Fillaus, M.K. Larson, G.C. Shearer, J.W. Newman,Basal omega-3 fatty acid status affects fatty acid and oxylipin responses tohigh-dose n3-HUFA in healthy volunteers, J. Lipid Res. 53 (2012) 1662–1669.

[51] B. Inceoglu, K. Wagner, N.H. Schebb, C. Morisseau, S.L. Jinks, A. Ulu, C. Hegedus,T. Rose, R. Brosnan, B.D. Hammock, Analgesia mediated by soluble epoxidehydrolase inhibitors is dependent on cAMP, Proc. Natl. Acad. Sci. USA 108(2011) 5093–5097.